M RC S EA FOR H YD ROPOW E R O N TH E M E KO NG M AI NS T R EAM IMPACTS ASSESSMENT. (Opportunities and Risks) Volume II: Main Report.

Transcription

1 M RC S EA FOR H YD ROPOW E R O N TH E M E KO NG M AI NS T R EAM IMPACTS ASSESSMENT (Opportunities and Risks) Volume II: Main Report 11 June 2010 The MRC SEA of Hydropower on the Mekong mainstream comprises 4 main phases: (i) scoping, (ii) baseline assessment, (iii) opportunities & risks assessment, and (iv) Avoidance, enhancement and mitigation assessment. The assessment of opportunities and risks comprises three volumes: (i) Vol I: Summary, (ii) Vol II: Main report, (iii) Vol III: Annexes & supporting materials This 3 volume report formally concludes the impacts assessment phase of the SEA and documents the development opportunities and risks presented by the 12 LMB mainstream projects in the context of sustainable development of the Mekong River Basin.

2 Disclaimer This document was prepared for the Mekong River Commission Secretariat (MRCS) by a consultant team engaged to facilitate preparation of a Strategic Environment Assessment (SEA) of proposals for mainstream dams in the Lower Mekong Basin. While the SEA is undertaken in a collaborative process involving the MRC Secretariat, National Mekong Committees of the four countries as well as civil society, private sector and other stakeholders, this document was prepared by the SEA Consultant team to assist the Secretariat as part of the information gathering activity. The views, conclusions, and recommendations contained in the document are not to be taken to represent the views of the MRC. Any and all of the MRC views, conclusions, and recommendations will be set forth solely in the MRC reports. This document incorporates impact analysis of issues raised in the consultation process as prepared by the consultant team and is to be considered a discussion draft only for the multi-stakeholder Regional Workshop in May 2010 in Vientiane. For further information on the MRC initiative on Sustainable Hydropower (ISH) and the implementation of the SEA of proposed mainstream developments can be found on the MRC website: and The following position on mainstream dams is provided on the MRC website in M R C p o s i t i o n o n t h e p r o p o s e d m a i n s t r e a m h y d r o p o w e r d a m s i n t h e L o w e r M e k o n g B a s i n More than eleven hydropower dams are currently being studied by private sector developers for the mainstream of the Mekong. The 1995 Mekong Agreement requires that such projects are discussed extensively among all four countries prior to any decision being taken. That discussion, facilitated by MRC, will consider the full range of social, environmental and cross-sector development impacts within the Lower Mekong Basin. So far, none of the prospective developers have reached the stage of notification and prior consultation required under the Mekong Agreement. MRC has already carried out extensive studies on the consequences for fisheries and peoples livelihoods and this information is widely available, see for example report of an expert group meeting on dams and fisheries. MRC is undertaking a Strategic Environmental Assessment (SEA) of the proposed mainstream dams to provide a broader understanding of the opportunities and risks of such development. Dialogue on these planned projects with governments, civil society and the private sector is being facilitated by MRC and all comments received will be considered.

3 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S A b ou t the MR C S E A o f H yd ro p o wer o n th e M e k on g m a i n s tr e am The Mekong River Commission (MRC) is an inter-governmental river basin organization that provides the institutional framework to implement the 1995 Mekong Agreement. The Governments of Cambodia, Lao PDR, Thailand and Viet Nam signed the Agreement on the Cooperation for the Sustainable Development of the Mekong River Basin. They agreed on joint management of their shared water resources by cooperating in a constructive and mutually beneficial manner for sustainable development, utilization, conservation and management of the Mekong River Basin water and related resources. Poverty alleviation as a contribution to the UN Millennium Development Goals is also a priority. The two upper states of the Mekong River Basin, the People's Republic of China and the Union of Myanmar, are dialogue partners to the MRC. In a region undergoing rapid change and economic growth, the MRC considers the development of hydropower on the Mekong mainstream as one of the most important strategic issues facing the Lower Mekong region. Through the knowledge embedded in all MRC programs, the MRC is conducting this Strategic Environment Assessment (SEA) to assist Member states to work together and make the best decisions for the basin. Twelve hydropower schemes have been proposed for the Lao, Lao Thai and Cambodian reaches of the Mekong mainstream. Implementation of any or all of the proposed mainstream projects in the Lower Mekong Basin (LMB) could have profound and wide-ranging socio economic and environmental impacts in all four riparian countries. This SEA seeks to identify the potential opportunities and risks, as well as contribution of these proposed projects to regional development, by assessing alternative mainstream Mekong hydropower development strategies. In particular the SEA focuses on regional distribution of costs and benefits with respect to economic development, social equity and environmental protection. As such, the SEA supports the wider Basin Development Planning (BDP) process by complementing the MRC Basin Development Plan (BDP) assessment of basin-wide development scenarios with more in-depth analysis of power related and crosssector development opportunities and risks of the proposed mainstream projects in the lower Basin. The SEA is being coordinated by MRC s cross-cutting MRC Initiative for Sustainable Hydropower (ISH) working with all MRC programmes. The SEA will directly enhance the baseline information and assessment framework for subsequent government review of project-specific EIAs prepared by developers. It will also inform how the MRC can best enhance its support to Member Countries when the formal process under the 1995 Mekong Agreement for prior consultation on any individual mainstream proposal is triggered (i.e. the Procedures for Notification, Prior Consultation and Agreement or PNPCA). The SEA findings will also inform steps that MRC programmes may consider in the next MRC Strategic Plan Cycle ( ) to help address the knowledge gaps and the key areas of uncertainty and risk concerning proposed mainstream developments. The SEA began in May 2009 and is scheduled to complete the final report and recommendations by mid This document is one of a series of documents arising from an intensive program of consultations in the Lower Mekong Basin and detailed expert analysis of the issues associated with developing hydropower on the Mekong mainstream. The intention is to consolidate SEA activities and progressively make conclusions and outputs available for public and critical review, so that stakeholder engagement can contribute to the SEA in a meaningful way. A full list of documents is available on the MRC SEA website. 3 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

4 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S CONTENTS 1 Introduction Overview Assessment approach Proposed LMB Mainstream Hydropower Description of the proposed mainstream projects Project type LMB mainstream project operations during wet and dry season Typical life-cycle for the proposed LMB mainstream projects Reasons behind the LMB mainstream hydropower proposals Regional cooperation, the MRC and the PNPCA process Energy & Power Regional distribution of power benefits Overview Distibution of Power benefits Other hydropower-related impacts Regional perspective of electricity sector impact Hydroelectric supply curves Hydro exporting countries Hydro importing countries Impact of mainstream projects on national energy polices Thailand and Vietnam I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

5 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S Lao PDR Cambodia Potential for collateral effects on electricity sectors Capital constraints Adverse signals for electricity conservation Design and operating aspects of mainstream projects Peaking operation Capacity factor Power transmission and local power supply Safety of mainstream dams infrastucture risk Mainstream hydropower development planning Conclusions regarding power sector impact Economic Systems Summary of past and future trends without LMB mainstream dams Expected macro economic trends with mainstream dam construction Capital investment Revenue generation investment stimulus Real exchange-rate appreciation and Booming Sector Effects Public debt sustainability in Lao PDR Conclusions Expected sectoral effects on economic trends Economic sectors Assests I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

6 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S Ecosystem services Expected distributional effects on economic trends Distributional effects between rural and urban areas Distributional impacts: poor and non-poor Distributional impacts: ecological zone Distribution of impacts by country Conclusion Hydrology and Sediment Strategic Issues Data sources and potential uncertainties Summary of past and future trends without LMB mainstream hydropower Stream power Water surface level changes Fate and transport of coarse sized sediment Fate and transport of fine sized sediment Expected direct effects of the proposed LMB mainstream projects Stream power Water surface level changes Fate and transport of coarse sized sediment Fate and transport of fine size sediment Terrestrial Ecosystems & Agriculture Strategic Issues Summary of past & future trends without LMB mainstream hydropower I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

7 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S 6.3 Expected direct effects of the proposed LMB mainstream hydro projects KBAs and Protected areas Areas inundated Access roads and transmission lines Indirect landuse impacts Impacts upon land use - forest cover and agriculture River Bank Gardens Other impacts upon agricultural land Sensitivity analysis of dam groupings Cascade of 6 dams upstream of Vientiane Middle Mekong dams Lower Mekong dams (Cambodian dams) Possible indirect links with other themes Summary of impacts on terrestrial ecosystems and agriculture Maps Aquatic systems Strategic Issues Summary of past and future trends without LMB mainstream hydropower Expected direct effects of the proposed LMB mainstream hydro projects Construction impacts Operation impacts Implications for the aquatic ecosystem I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

8 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S 7.4 Sensitivity analysis of dam groupings Cascade of 6 dams upstream of Vientiane Middle Mekong dams Lower Mekong dams Possible indirect links with other future trends Fisheries Strategic issues Thematic Overview Sensitivty analysis overview Cross-cutting conclusions Mainstream dam clusters Upstream cluster of dams Middle cluster of dams Downstream cluster of dams Overview Forecasted hydrological changes relating to Fisheries Wetlands, floodplains and fish productivity Long-distanCE migrants and Mekong fish production Migration patterns of dominant species Dominant species in mekong fish catches Gains in fish production from Dam reservoirs Estimates of productivity based on surface area, depth and flow Estimates of productivity based on surface area alone Sensitivity analysis of dam groupings I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

9 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S UPSTREAM CLUSTER OF DAMS Middle cluster of dams Downstream cluster of dams Social systems Summary of past & future trends without LMB mainstream hydropower Poverty Reduction Health & Nutrition Resettlement & land acquisition Opportunities and risks Dtrategic and institutional concerns regarding social impacts Impacts of proposed lmb mainstream hydro projects Location of impacts The difference between direct and indirect impacts, and how many people affected Direct and indirect impacts Equity of risks Transboundary impacts and conflict management Transboundary impacts Conflict management Knowledge gaps social systems Navigation Strategic issues Overview Subsistence users (small, medium) I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

10 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S Iincreased navigability: Decreased long-haul connectivity: Passenger transport (including ferries, tourism, cruises, etc) Pak beng to pak chom (zone 2) Ban koum and lat sua (zone 3) Stung treng and sambor (zone 4) Increased navigability: Decreased long-haul connectivity: Freight transport (small, medium and large) Pak beng to pak chom (zone 2) Ban koum and lat sua (zone 3); Stung treng and sambor (zone 4): Increased navigability: Decreased long-haul connectivity: Mekong delta Mining and resources Freedom of navigation Conclusion Climate Change Strategic Issues Summary of past & future trends without LMB mainstream hydropower Expected direct effects of the proposed LMB mainstream hydro projects I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

11 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S 11.4 Effects of the LMB mainstream projects on climate change - GHG emissions Total GHG emissions Emissions per power unit Emissions by ecological zone Emissions compared with other power sources Direct climate change effects on the mainstream projects Increased runoff, flow and flooding Risk of extreme events and dam failure indirect links between climate change, mainstream projects and other themes Reduced food security Reduced water quality Loss of biodiversity Constraints to poverty reduction Increased potential power capacity across the basis Sensitivity analysis of dam groupings Cascade of 6 dams upstream of Vientiane (zone 2) Middle Mekong dams (zone 3) cambodian dams (zone 4) Summary of impacts References Energy & power Economic systems Hydrology & Sediment I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

12 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C O N T E N T S 12.4 Aquatic & terrestrial systems Fisheries Social systems Navigation Climate change I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

13 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T I N T R O D U C T I O N 1 INTRODUCTION 1.1 OVERVIEW The SEA Impacts Assessment phase marks a critical point in the MRC SEA of hydropower on the Mekong mainstream. The impacts assessment phase spans two months from April June 2010 culminating in the SEA Regional Impacts Assessment workshop (19-20 May, Vientiane) and subsequent finalization of the impacts assessment paper. This draft presents the SEA team s preliminary findings concerning the opportunities and risks associated with the 12 LMB mainstream projects. The draft report is for critical review by all SEA stakeholders, including: (i) MRC bodies (ii) government agencies in the four LMB nations, (iii) hydropower developers and the private sector (ii) civil-society organizations in the LMB and internationally (iv) Mekong riparian communities and the representatives and public interest groups, and (v) donor partners and the finance community. The process of review, comment and then revision continues the consultative agenda of the SEA. The draft is being circulated for formal written submissions and placed online for wider exposure and responses. The main public forum for discussion on the draft is the SEA Regional Impacts Assessment Workshop. The impact assessment is divided into the 8 strategic themes used throughout the SEA process including the baseline assessment phase, namely: 1. Energy & power 2. Economic systems 3. Hydrology & sediment 4. Terrestrial systems 5. Aquatic systems 6. Fisheries 7. Social systems 8. Climate change The impacts assessment report will be finalized in early June 2010 in parallel with the preparation of the avoidance, enhancement and mitigation measures, which leads to a final regional workshop to consider the full SEA report. 1.2 ASSESSMENT APPROACH The assessment approach used in the SEA was described in the SEA Inception Report (IR) available online. 1 It examines the mainstream projects from a number of geographic perspectives: regional, national, hydro I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

14 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T I N T R O D U C T I O N ecological zones and then in groupings to better explore the opportunities and risks associated with mainstream hydropower development. REGIONAL: The Mekong River is commonly divided into its upper and lower reaches, based on arrangements for regional cooperation in sustainable development of the Mekong Basin. The SEA regional approach covers both, but focuses on transboundary concerns and the socio-economic and natural system linkages between the LMB countries. MRC and under the 1995 Mekong Agreement provide the institutional and planning context at the regional level, especially (i) the Basin Development Plan (BDP), and (ii) the procedures for Prior Notification/ Prior Consultation Agreement (PNPCA). NATIONAL: The LMB countries Cambodia, Lao PDR, Thailand and Vietnam and the effects of mainstream projects on each of their distinctive economies, and social and natural systems is the other key focus of the SEA. The SEA defined the key strategic issues of focus through an intensive program of ongoing national consultation. This means the 12 proposed projects are assessed against national interests and development priorities of the four LMB countries. HYDRO-ECOLOGICAL ZONES: The SEA impact assessment was further facilitated by viewing the projects within 6 hydro-ecological zones of the Lower Mekong River. The zones have distinctive bio-physical characteristics. They are defined based on: (i) hydrological regime, (ii) physiographic, (iii) land use, and (iv) existing, planned and potential resource developments (Table 1.1 and Figure 1.1). Table 1.1: Hydrological spatial zoning for the Mekong Basin ZONE IBFM FLOW ZONE DESCRIPTION PHYSICAL FEATURES Mainstream projects 1 Lancang River Major source is Tibetan plateau with precipitation, snowmelt, some glaciers in the headwaters Narrow steep gorges and bedrock confined single thread channel Zone of sediment production Drainage: Parallel, no tributaries, reticular Geology: Gorges, Ailao Shan Shear Zone Length ~1,900km with a fall of ~4,200m Existing, under construction and planned mainstream dams in China 2 Chiang Saen to Vientiane 3 Vientiane to Pakse Mountainous with large areas of remaining forest Steep narrow valleys with significant bed rock outcropping Significant meander in the basin and within channel features (islands, bars terraces) Shared Lao-Thai river bank Zone of sediment production Increasing influence of tributaries to flow Predominantly alluvial reach with widening cross-section and reduced meander Large flat-topped channel islands Shared Lao-Thai river bank Zone of sediment transport Drainage: parallel Geology: Gorges, Wang Chao fault zone Length ~1,100km with a fall of m Drainage: dendritic Geology: Gorges, central highlands & Khorat plateau Length ~600km with a fall of m Pak Beng, Louangprabang Xayaburi Pak Lay Sanakham, Pak Chom Ban Koum Latsua 4 Pakse to Kratie 3S contributions (>25% of Drainage: dendritic Latsua 14 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

15 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T I N T R O D U C T I O N 5 Kratie to Phnom Penh 6 Phnom Penh to South China Sea annual flow volume) Including wetlands of Simphandone, Khone Falls, Stung Treng, Kratie Khone falls shifts river from braided channel system to floodplain Mixed zone of sediment transport, production & deposition Cambodian floodplain and the Tonle Sap system Braided system of river channels Overbank siltation maintain flooding in the floodplain Seasonal reversal of flow in the Tonle Sap system Most inland extent of influence for saline intrusion and coastal process Zone of sediment deposition Braiding of the mainstream and complex network of canals Deltaic environment river fans out into a network of distributaries Extensive flooding and saline intrusion Geology: Gorges, central highlands Length ~ km, with negligible fall Drainage and dendritic form Length ~ km, with negligible fall Drainage: Distributaries Length ~ km with negligible fall (downstream) Don Sahong Thakho Stung Treng Sambor Note: adapted from MRC, 2005; MRC, unpublished GROUPING OF MAINSTREAM PROJECTS: The SEA adopts the MRC s Basin Development Plan (BDP) scenario framework. The BDP 20-year 2030 scenario and its two variants with/without the proposed LMB mainstream schemes is the basis for the SEA assessment of trends and impacts, also with reference to the baseline scenario. The baseline was the MRC Definite Future Scenario for 2015 (DFS). Both the DFS and PFS incorporate the influence of dams in the Lancang-Mekong portion of the basin in China on the LMB. The SEA also provides a sensitivity analysis which examined the LMB mainstream projects in three distinctive groups, namely the 20-year 2030 scenario with only: 1. The proposed cascade of 6 mainstream dams upstream of Vientiane in Lao PDR 2. The proposed middle-reach mainstream dams in Lao PDR, and 3. The proposed lower mainstream dams in Cambodia Each thematic chapter presents the findings of the impact assessment (opportunities and risks) according to those various geographic levels of focus. 15 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

16 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T I N T R O D U C T I O N Figure 1.1: Hydro-ecological zones of the Mekong Basin: LMB mainstream projects are indicated as orange dots 16 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

17 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W 2 PROPOSED LMB MAINSTREAM HYDROPOWER 2.1 DESCRIPTION OF THE PROPOSED MAINSTREAM PROJECTS PROJECT TYPE Hydropower projects generate electricity by utilising the stream power available within a river system and converting the kinetic energy of flow into electrical energy. For rivers with large flows, energy can be extracted from the flow itself, or an impoundment can be built to store potential energy and control generation through dam release leading to two distinct types of hydropower: storage and run-of-river projects. The proposed LMB mainstream dams are characterized as low head dams that span part of, or the entire mainstream channel in the Lower Mekong Basin. These projects would employ Kaplan or ship propeller-type turbines, which function effectively with low heads and high volumes of water to generate power and have efficiencies typically in the order of 85%. The gross heads of these projects would vary between 6.5 m to about 35 m, though this vertical head (the distance between the normal water level in the reservoir behind the dam and the tail water level below the dam) would vary significantly between monsoon periods (when the head is small but the flow is large) and low-flow seasons (when the head would be the greatest due to the much lower river flows and consequent lower tail water levels). The actual dam height as measured from the bottom of the foundations in the river bed to the top of the dam structures can be as high as 80 m. As a reference, the eight dams on the mainstream in China that are completed, under construction or planned vary from 67 m to 248 m head (for Jinghong and Xiaowan respectively). The 12 proposed mainstream developments are of three of three general types: 1. Dams that span the mainstream Mekong that involve limited inundation outside the mainstream channel: The 6 proposed dams from Pak Beng (the upper most dam in Lao PDR) to Pak Chom share similar broad characteristics. They typically span the Mekong river where it is 800-1,200m wide and has steeper valleys. The reservoirs forming behind these dams, typically 100 to 120 or more km would generally be confined to the river channel, except at confluences with major tributaries and other natural floodplain where the reservoir will flood land areas. The dams further downstream at Ban Koum and Lat Sua have wider channels and some floodplain areas that would be inundated. 2. Dams that span parts of the braided Mekong mainstream or diversion schemes: This includes the proposed Don Sahong project (240 MW) which is the only dam project which does not span the entire width of the Mekong mainstream, consequently the project is smaller than the others. It also includes the Thakho (60 MW) run-of-river diversion project that will divert a portion of one of the main braided 17 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

18 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W channels around Khone Falls, which does not involve a dam. It is in the same area as the proposed Don Sahong. 3. Dams that span the mainstream Mekong or several branches that involve significant inundation outside the mainstream channel: For the two proposed floodplain projects in Cambodia (Stung Treng and Sambor) the approach is to have concrete dams span the main river channels flanked by earth embankment dams totaling between 10 km and 18km respectively in length. The reservoirs forming behind these structures would inundate larger areas of floodplain with total areas ranging from 200 to 620 km2. Installed capacity range from 980 MW (Stung Trek) to 2,600MW (Sambor) with a maximum discharge through the turbines of 18,000 m3/s at Sambor (though Cambodia notes that a scaled-down version of Sambor is also feasible). Like the other projects, the storage capacity of these projects is limited to a few days or less and they would propose to use low-head Kaplan generation units. Project profiles: Profiles for the 12 mainstream projects were prepared during the SEA Inception phase (Volume 2); a summary of the salient features of the projects is presented in Table 1 below. A more general picture is run-ofriver projects do not have the capacity to regulate flow and rely on the daily flow rate to generate power; this makes output subject to weather-dependent flow conditions. Storage projects involve large impoundments and allow the operators greater control, however, also with greater disruption to the natural hydrological regime and other knock-on environmental and social consequences. Typically hydropower projects are not purely storage or run-of-river but are found on the spectrum that ranges between these two modes (see figure 3.1). For the LMB mainstream projects some, like Thakho are purely run-of-river, while others like Xanakham and Sambor have some potential for short-term storage. 18 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

19 Figure 3.1: Types of hydropower projects in the Mekong Basin the spectrum from storage type to run-of-river M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W 19 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

20 MAINSTREAM DAM LOCATION DEVELOPER EARLIEST POTENTIAL COMMISSION DATE DESIGN STATUS ENVIRONMENTAL ASSESSMENT STATUS Rated Head (m) Plant Design Discharge (m3/s) Installed Capacity (MW) Peaking Capability (MW ) Mean Annual Energy (GWh ) Firm Annual Energy (GWh ) Full Supply Level (mamsl ) Low Supply Level (Mamsl) Live Storage (mcm ) RESERVOIR AREA (km2) Length of dam (m) Height (m) M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W Table 3.1: Salient features of the LMB mainstream projects MANAGEMENT STATUS DESIGN SPECIFICATIONS DIMENSIONS Pak Beng Lao PDR Datang International Power Generation 2016 MoU, feasibility IEE submitted 31 7,250 1,230 1,230 5,517 4, Louang Prabang Lao PDR PetroVietnam Power Corporation 2016 MoU, feasibility Feasibility study, 40 3,812 1,410 1,412 5,437 4, , Xayaburi Lao PDR SEAN & Ch. Karnchang Public Co Ltd 2016 MoU, feasibility Feasibility and full ESIA submitted 24 6,018 1,260 1,260 6,035 5, Pak Lay Lao PDR CEIEC and Sino- Hydro 2016 MoU, feasibility IEE submitted 26 4,500 1,320 1,320 6,460 4, Sanakha m Lao PDR Datang International Power Generation 2016 MoU, feasibility Not yet 25 5, ,200 5,015 3, , Pakchom Lao PDR N/a 2017 MasterPla n Not yet 22 5,720 1,079 1,079 5,318 5, , Ban Koum Lao PDR Italian Thai Asia Corp. Holdings 2017 MoU, feasibility Not yet 19 11,700 1,872 1,872 8,434 8, Latsua Lao PDR Charoen Energy and Water Asia Co Ltd 2018 MoU, prefeasibility Pre-feasibility study submitted , ,668 1, , Don Sahong Lao PDR Mega First 2016 PDA, detailed planning Full EIA submitted, Additional studies 17 2, ,375 1, Thakho diversion Lao PDR CNR & EDL 2016 MoU, prefeasibility IEE submitted n/a n/a Channe l - 1,800m n/a Stung Treng Cambodia Vietnam Urban and Industrial Zone Development Inv. Corp N/a MoU, prefeasibility Not yet 15 18, ,870 2, , Sambor Cambodia China Southern Power Grid 2020 MoU, prefeasibility Pre-feasibility submitted 33 17,668 2,600 2,030 11,740 9, , I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

21 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W LMB MAINSTREAM PROJECT OPERATIONS DURING WET AND DRY SEASON Typically these dams will hold the water above what is normally the high flood level all year round and thus eliminate annual cycles of raising and lowering water levels. These reservoirs have a storage capacity in the order of a few days (less than 70mcm) and can operate continuously or under peaking. Supply levels in the reservoirs can be greater such as for a 5 m draw down from normal water levels in advance of major storm events. Figure 3.2 presents the reservoir inundation areas for the Xayaburi and Stung Treng projects as an example of the range of reservoir type in the group of proposals, all reservoir maps are located in the Annex. These projects each have an installed capacity of about 700 MW 1,400 MW, and a design discharge through all turbines of 3,000-12,000 m3/s. The capacity of the turbines is loosely comparable to the average annual flow expected in the river at the given reach. In the wet season the majority of the flow would be over the spillways. An example is shown for the Luang Prabang project below (figure 3.3), which is indicative for the other LMB mainstream projects except for Thakho. During the calendar year the low storage capacity means that the water levels in the reservoirs and downstream reaches will vary with season: January May/June: water levels in the reservoir will be kept close to full supply level and above the maximum observed water level with the dam passing the dry season flow through the turbines. Downstream water levels will be comparable to dry season water levels in the channel providing close to maximum available hydraulic head but minimum flow rates. During this period projects will be generating below rated capacity. June August: As the discharge in the river crosses the annual average the turbines will be generating at peak capacity. During this time there is the potential for peaking generation at a daily time step. Water levels in the reservoir would remain close to the full supply level with continuous operation and under peaking operations could fluctuate by 10m during the day. Downstream water levels with continuous operation would approximate the expected seasonal water levels without mainstream projects. Under peaking operations downstream water levels could vary by 4-8m between peak and off-peak periods. August October: As the river approaches peak flood, the turbines will not be capable of passing the increasing flows such that all the spill way and other gates would be opened. This would result in flood levels up and downstream of the dam with minor head difference and a reduction in the power generation potential of the project during this season. October November: As the flow drops towards the annual average there will be a period of optimal power generation with the potential for peaking operations. During this time water levels will be similar to the July-August block November December: then as the flow continues to drop to the dry season low, water levels up and downstream of the reservoir will experience fluctuations comparable to January to May above. 21 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

22 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W Figure 3.2: Reservoir inundation areas: (left) Stung Treng the flat topography results in a significant area of reservoir outside the mainstream channel. The reservoir has a length of 40.4km, surface area of 234km2 and volume of 1,550mcm directly affecting some 20 villages and towns; (right) Xayaburi the steep incised valleys of Zone 2 result in greater confinement of reservoirs to the river channel. The reservoir has a length of 96.8km, surface area of 61km2 and volume of 150mcm directly affecting some 35 villages and towns 22 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

23 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W Figure 3: Cross-section of a typical run-of-river power house: illustration of the Luang Prabang project TYPICAL LIFE-CYCLE FOR THE PROPOSED LMB MAINSTREAM PROJECTS At the soonest LMB mainstream projects could enter the Mekong system in 2015, with construction taking between 7 and 15 years. If multiple projects are approved and construction is staggered then there could be construction along the Mekong River for the next two decades. All projects are likely to be BOT (Built Operate Transfer) projects, which would allow the developer to operate the project for 25 years as a private venture before transferring ownership over to the national government. During this period the government would receive royalties on revenues generated by the project. After hand over, the project reverts to full ownership of the government and the management arrangements state that time would be decided which may include retaining the current operator under a management contract.. During the concession period operation and maintenance would be provide the private developer following terms of the concession agreement and regulations established by the national regulatory body. In general, certain increased expenditure at the outset can increase the design life of project components and help minimize future maintenance requirements and prolong component end-life. The initial costs are born by the developer. Reduced expenditure at the outset can increase maintenance and replacement costs, which are likely to be more significant at later stages in the project life after the 25 year concession period and hence borne by the government. The balance is subject to negotiation and part of the concession agreement. Typically all large dams go through refurbishment cycles were electrical and mechanical equipment is upgraded and modernized. The assessment of the economics of dams may use 23 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

24 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W a 50 to 100 year economic life. Most large dams exceed their economic life through cycles renovation and refurbishment. Figure 3.4 below presents an indicative time line of this phasing. 2 Figure 3.4: Long-term Phasing schedule for mainstream Mekong hydropower CONSTRUCTION OPERATIONS (INVESTOR) BOT HANDOVER PERIOD OPERATONS (GOVERNMENT) MAINTENANCE COMPONENT/INFRASTRUCTURE END LIFE REFURBISHMENT, UPGRADE or DECOMMISSIONING END PROJECT ECONOMIC LIFE Note: Phasing above is based on a 50year life cycle which is consistent with the time frame used by MRC to compute capital recovery costs. Depending on the operations and maintenance strategies the projects may last for longer than 50years 2.2 REASONS BEHIND THE LMB MAINSTREAM HYDROPOWER PROPOSALS The large annual discharge (505 km 3 ) and fall (4,800m) of the Mekong River combine to produce a hydrological regime with a large energy budget. Over geologic time the Mekong system has reached a state of dynamic equilibrium which utilises this energy to transport; water, nutrients, sediment, organic matter, aquatic biota; and carry out other important ecosystem functions making the Mekong Basin one of the most biologically diverse in the world and one of the most productive in terms of fisheries and agriculture. Initially, attempts to harness the energy of the Mekong and its tributaries were localised and relatively small-scale. From the 1950/60s hydropower storage reservoirs have been built on the tributaries of the Mekong River starting in Thailand. The preference for storage-type projects in the tributaries can be attributed to: (i) global experience with storage hydropower provided greater confidence in construction and operation techniques, and (ii) large reservoirs offered the ability to regulate stream flow allowing for a number of other benefits to human communities, including potential for multi-purpose use (e.g. irrigation off-takes form the reservoir), yearround water supply, control of dry season lows/droughts and wet season flooding. In the 1970/80s, the social and environmental impacts of hydropower with large storage reservoirs began to be better understood, especially in relation to the adverse effects of flow regulation and the inundation 2 For further information on the project economic life and actual operating life the World Commission on Dams (2000) report provides a view of issues such as refurbishment, repair and decommissioning. 24 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

25 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W of large areas of land. 3 This led to renewed interest in low head hydropower projects on the Mekong mainstream, which relied on the flow rather than storage head to generate electricity (Acres, 1994). In 1994, the Mekong Secretariat, United Nations Development Program and the Government of France undertook a study of the LMB Mekong mainstream to identify potential sites for low storage run-of-river projects. Eleven of the twelve projects considered in the SEA were considered in the 1994 study. Only Thakho which emerged as an alternative to Don Sahong is new. There are a number of engineering features of the LMB run-of-river projects that make them attractive from an operation and power production point of view: Large power outputs: The large flows in the Mekong River allow each project to generate between 2,000-12,000GWh/yr Short ramping rates: the projects can be brought online in a matter of minutes, compared to the much larger ramping rates for thermal power plants. Suitability for peaking: these ramping rates make peaking operation a possibility, which would allow operators to target energy production to times of the day when the buying price is highest These projects were then de-prioritized after the Acres Report (1994) because the large seasonal variability in Mekong flows and high sediment load of the river proved inhibitive to their feasibility as well as being unpopular with local communities. Three major developments in the basin have occurred since then which have out the LMB mainstream projects back on the table: 1. Upstream regulation of the flood pulse: changes to the hydrological regime from upstream hydropower development will disrupt the Mekong flood pulse and provide a more reliable yearround supply of water, increasing dry season flows by as much as 20% as far downstream as Kratie and as much as 50% at Kratie in the low flow months (Adamson, 2009). 2. Expected reduced sediment load: In the order of 40-60% of the Mekong sediment load originates in Yunnan. The Yunnan cascade will have individual trapping efficiencies of up to 95% and combined trapping efficiency of approximately 80%, which has benefits from an operational, maintenance and risk failure point of view. 3. Continued growth in energy demand for the Mekong region (as detailed in the baseline report) and the introduction of the private sector as a development force. These factors have facilitated a shift from exploring a number of options to identify a small number of candidate projects to take through to operation (as was the case in 1994), to the current situation where all 12 projects are being explored individually by 12 different developers. 3 See for example studies on the impacts of the Nasser Dam on the Nile ecosystem, delta, fisheries and agricultural soil fertility: Welcome, R.L River Fisheries. FAO Fisheries Technical Paper 262, or studies on the Mississippi River basin: Meade, R.H., and J.A. Moody Causes for the decline of suspended sediment discharge in the Mississippi River system, Hydrology Processes, vol 24, pp I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

26 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W 2.3 REGIONAL COOPERATION, THE MRC AND THE PNPCA PROCESS Planning for mainstream hydropower development is moving forward mainly on a project by project basis without overarching national plans for the use of the Mekong River in any of the four LMB countries or without an agreed regional plan for the use of the Mekong. The mainstream projects do not appear in the Thai and Lao national power plans. Equally significant, planning is moving forward within the confines of one sector in each country the power sector with limited involvement of other users of the Mekong River either development sectors or communities. However, there are various development planning levels, which provide an opportunity to support decision makers in adequately considering the risks and opportunities and broad trade-offs relating to the 12 projects in a river basin context. Central to these is the MRC BDP planning process. The SEA fits within these layers of planning and has the potential to contribute to each of them. The MRC Procedures for Notification, Prior Consultation and Agreement PNPCA established under the 1995 Mekong Agreement is one of the most significant regional instruments for consideration of mainstream hydropower power. 4 MRC Procedures for Notification, Prior Consultation and Agreement: Article 2 of the 1995 Agreement on Cooperation for the Sustainable Development of the Mekong River Basin calls on signatory countries to promote, support, cooperate and coordinate in the development of the full potential of sustainable benefits to all riparian States and to prevent wasteful use of Mekong River Basin water. In 2003, the LMB countries adopted Procedures for Notification, Prior Consultation and Agreement on uses of Mekong waters and related resources. That 2003 PNPCA Protocol and its 2005 Procedural Guidelines require Member States to notify each other through the MRC of proposed uses of the Mekong tributaries and mainstream so that the potential impacts on multi-stakeholder rights and interests can be assessed and optimal water use determined. For the mainstream proposals, the MRCS must take a proactive role to assist the Joint Committee in assessing whether the use is reasonable and equitable, whether greater benefits can be derived through cooperation and trade-offs and to ensure due diligence in the planning process. The PNPCA process requires that the Joint Committee aim to arrive at an agreement and issue a decision containing conditions relating to the proposed use (PCA ). During 1995 to 2008 there were 28 notifications all relating to developments on tributaries. The MRC took little action other than informing Member States. In early 2009, the situation changed with respect to mainstream proposals where initial information on a number of mainstream dam proposals was submitted to MRCS by Lao PDR, Cambodia and Thailand. The consultation part of the PNPCA has yet to be triggered, though it is anticipated in The MRCS was also instructed in by the Joint Committee to take two important steps to inform members and facilitate planning: (i) to conduct a strategic environmental assessment of all mainstream projects in the pipeline and, in parallel, (ii) to prepare Design Guidance for Mekong Mainstream Dams in the Lower Mekong Basin. Under the 1995 Mekong Agreement, the MRC needs to formulate a consistent approach to design and operation of mitigation measures of mainstream dams. Also, the Joint Committee 4 Other planning instruments are detailed in the SEA inception report and include: location project-specific planning, (ii) national & sector development plans, (iii) bilateral transboundary project planning, MRC Basin Development Plan, and (Iv) GMS sector and economic corridor plans 26 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

27 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T P R O P O S E D P R O J E C T O V E R V I E W will need to refer to certain design considerations during the prior consultation process for mainstream dams. The guidance is project specific, while the SEA is to explore the broader economic, social and environmental system implications of the projects collectively. During the Lao and Cambodia consultations on the scope of the SEA, government officials often asked where does the SEA fit in with the national project planning and assessment process? and what authority does the SEA have do Member States have to abide by its recommendations? In response to those questions: The Joint Committee will need to reach an agreed decision concerning the mainstream projects, individually and collectively. The SEA (like the Design Guidance) is advisory in nature it is intended to guide and help shape the Joint Committee s considerations and decisions under the PNPCA process that is its main function. The SEA is not being conducted as a formal requirement under the 1995 MRC Agreement and Protocols. It is a pilot to explore the potential usefulness of the tool in regional development planning. More specifically as a tool when needed to support the Joint Committee fulfill its role in guiding reasonable and equitable use of the Mekong waters; The SEA is not addressing an existing plan or one in preparation but a group of feasible project proposals for the same river; The projects are all in the planning stages so, in principle, the SEA can contribute to national planning and decision making with respect to the individual projects at the discretion of individual MRC Member States; Most of the projects have not yet been subject to EIAs or any form of cumulative impact assessment under national procedures so the SEA can help shape the requirements for those more specific studies. In addition, there are other planning platforms that can benefit from the SEA process and outcomes: The MRC Basin Development Plan is under preparation the SEA can contribute to that process. National power development plans are under review and preparation therefore open to influence. The GMS energy road map and strategy is now under review and revision, supported by the ADB GMS - RETA No 6440, and therefore the SEA could have an input to that process. 27 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

28 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y & P O W E R 3 ENERGY & POWER 3.1 REGIONAL DISTRIBUTION OF POWER BENEFITS OVERVIEW Under the energy and power theme in the SEA baseline assessment, the policies and motivations for hydropower development that are embodied in LMB country power policies were explored, along with the country-specific motivations for cross-border power trade. The baseline assessment also looked closely at the drivers and trends for electricity demand growth, the role of demand-side management in reducing the rate of demand growth, and the role of alternatives in the electricity supply mix considering the energy resource endowment of each country and current policies on fuel imports and power imports/exports. These issues and options were considered in the context of the energy resource base and the power development policy context for each LMB country, as well as the different viewpoints that have been expressed by the stakeholders in the open SEA consultation processes to date. This impact assessment for the energy and power theme focuses mainly on the regional distribution of power benefits, on the impact of power sectors in each LMB country from a utility economics perspective of least-cost supply and on physical infrastructure risk. The assessment is provided in mind of the scenarios used in the SEA assessment to integrate inputs from all the themes, namely: The BDP 20-year probable future scenario (PFS for the year 2030) with no LMB mainstream dams and with all LMB mainstream dams (with and with/out cases), and Three sensitivity cases that consider the 20-yr PFS plus, o Only the Lao northern cascade of LMB mainstream dams (ecological zone 2) o Only the Lao middle reach group (ecological zone 3) o Only the Lower group (ecological zone 4) DISTIBUTION OF POWER BENEFITS The results of the calculations on the regional distribution of power benefits for these two main scenarios (the BDP 2030 probable future scenario with/ without LMB mainstream dams) and the three sensitivity cases are provided in Table 4.1. The data show how hydroelectric power supply from tributary and mainstream LMB dams would be distributed among power systems in the four LMB countries. Table 4.1 thus illustrates the power benefit side of the opportunities and risk equation. 5 The other sections of this report assess in qualitative and quantitative terms the nature of the development opportunities and risks that LMB mainstream dams provide or pose in other sectors, as defined by the seven SEA themes. 5 The analysis is provided using the MRC Hydropower database that was employed also for the MRC BDP Scenario Assessment work. Thus the results on the estimation of power benefits (opportunities) are complementary with the MRC BDP. 28 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

29 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y & P O W E R Table 4.1 Regional Distribution of LMB Power Benefits for the 20-year Probable Future (with and without LMB mainstream dams) and sensitivity cases LMB Regional Distribution POWER SUPPLY (GWh/Year) PR0JECTED POWER EXPORT (GWh/Year) PROJECT INVESTMENT ($USM) SCENARIO 1 CAM LAO THAI VIE TOTAL CAM LAO THAI VIE TOTAL CAM LAO THAI VIE TOTAL Y-with MD 3,677 20,412 60,694 35, ,840 19,384 64, ,389 11,669 27, ,771 41, Y-w/o MD 1,703 9,038 26,206 16,346 53,293 1,618 28, ,189 1,268 11, ,771 15, Y- MD in zone 2 only 1,703 12,287 50,558 21,240 85,787 1,618 57, ,434 1,268 22, ,771 26, Y- zone 3 only 1,703 16,759 30,423 16,346 65,231 1,618 32, ,406 1,268 16, ,771 20, Y- zone 4 only 3,677 9,441 32,126 30,164 75,408 19,384 30, ,927 11,669 12, ,771 26,886 LMB Regional Distribution GROSS BENEFIT OF SUPPLY ($USM/Year) GROSS EXPORT REVENUE Estimated ($USM/Year) NET OVERALL POWER BENEFIT ($USM/Year) SCENARIO CAM LAO THAI VIE TOTAL CAM LAO THAI VIE TOTAL CAM LAO THAI VIE TOTAL Year with MD 782 3,555 5,284 2,559 12,179 1,250 4, , , , Y-w/o MD 362 1,574 2,281 1,193 5, , , , , Y- zone 2 only 362 2,140 4,401 1,550 8, , , , , Y- zone 3 only 362 2,919 2,648 1,193 7, , , , , Y- zone 4 only 782 1,644 2,797 2,201 7,424 1,250 2, , , ,506 Notes: 1 Scenarios are based on the MRC Basin Development Plan (BDP 20 year Probable Future snap shot) 2 Gross benefit of supply is based on avoided thermal costs in country where power is consumed 29 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

30 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R The difference in net overall economic power benefit between the 20-year probable future scenario with and without proposed LMB mainstream dams is about $US 3-4 billion annually by 2030, discounted ($ 2010). The economic power benefit depends heavily on the future energy mix that is assumed. It can be seen from Table 4.1 that the net overall power benefit is greatest for Lao PDR in all cases. This is due to a combination of factors, including the fact that Lao PDR has the largest number of potential sites for hydropower development and the relative cost of power from the different mainstream schemes, as is discussed further in Section that follows on supply curves OTHER HYDROPOWER-RELATED IMPACTS In the SEA baseline assessment report a number of non-power impacts associated with hydropower development in the LMB and issues raised by SEA stakeholders were quantitatively assessed. These are impacts that can have a regional distribution across the LMB countries, and to some extent, the Greater Mekong Sub-Region (GMS). These factors include (i) labor i.e. direct job creation (ii) contracts for goods and services like construction and electrical and mechanical equipment, and (iii) the contribution of LMB hydropower to regional GHG emission reductions from the power sector by virtue of avoiding or offsetting thermal generation beyond what can be achieved by demand-side management. Table 4.2 illustrates the scale and regional distribution of two such factors: expected wages from direct job creation, and gross GHG emission avoidance for the two scenarios and three sensitivity cases. Table 4.2: Regional Distribution of non-power impacts: Jobs and thermal avoided emissions LMB Regional Distribution DIRECT JOBS - WAGES DURING CONSTRUCTION 2,3 ($USM) TOTAL PRESENT VALUE OF DIRECT JOBS Wages duting Constuction & Operation ($USM) SCENARIO 1 CAM LAO THAI VIE TOTAL CAM LAO THAI VIE TOTAL Year with MD 2,769 7, ,007 3,334 9, , Y-w/o MD 345 3, , , , Y- zone 2 only 345 6, , , , Y- zone 3 only 345 4, , , , Y- zone 4 only 2,769 3, ,004 3,334 4, ,434 GHG EMISSION OFFSET4 (Million Ton CO 2 /Year) Ton/MWh by Country SCENARIO CAM LAO THAI VIE TOTAL Year with MD Y-w/o MD Y- zone 2 only Y- zone 3 only Y- zone 4 only Notes: 1. Scenarios are based on the MRC Basin Development Plan (BDP 20 year Probable Future snapshot) 2. Assumes wages only in the country where project is constructed (adjustments would be made for 2 LMB dams spanning Lao and Thailand Excludes indirect jobs / wages such as from services, mechanical & electrical equipment manufacture/supply, etc. 3. SEA baseline report considers the split between local and foreign jobs mostly from the GMS region 4. Gross GHG based on avoided thermal generation in the country where power is consumed. This is different than net GHG impact, which subtracts or takes potential for GHG emissions from reservoirs into account (a contoversial topic where as yet there is no consensus on the level of potential emissions from reservoirs of various kinds. 5. Specific GHG emission factor base on power generation mix in each country (i.e., type of thermal supply offset) 30 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

31 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R An integrated assessment of non-power impacts is provided in other thematic sections of this report, notably in the economics and climate change sections. 3.2 REGIONAL PERSPECTIVE OF ELECTRICITY SECTOR IMPACT HYDROELECTRIC SUPPLY CURVES An introduction to the concept of hydropower economic supply curves and their relationship to energy demand and alternative energy supply cost is helpful to better understand the benefits and risks that the proposed LMB mainstream schemes pose for the power sectors of LMB countries. Thermal power projects (natural gas, coal and oil-based) can be replicated at the same cost and size almost indefinitely to meet electricity demand growth. The physical limitation is fuel availability, which depends on the primary energy resource endowment of the country and appetite for fuel imports. This replication of power stations is not possible with hydroelectric projects. Each hydropower site is unique. The cost and annual electrical energy produced is different for each site and there are many possible project configurations. Therefore, to compare the electrical generation potential and cost of different sets of hydroelectric projects it is useful to refer to the supply curve of each set. The supply curve of a particular group or set of generation projects is a plot of electrical energy that is available below a certain cost level. Figure 4.1 shows the supply curves for two sets of LMB hydroelectric projects: One set corresponds to all new tributary and mainstream projects (i.e. projects that are planned or under construction) and the other set corresponds to all new tributary projects (i.e. excluding mainstream projects). 6 In Figure 1, the horizontal axis corresponds to annual energy (supply or demand) expressed in terawatt-hours. 7 The vertical axis corresponds to energy cost expressed in US dollars per MWh, a common trade unit for wholesale power. 8 Figure 4.1 also shows (using dashed arrows) the incremental annual electricity demand in Vietnam, Thailand, Cambodia and Laos that would be partially served by those projects. That is the annual demand in year 2025 that is not yet supplied by existing projects. 9 The incremental demand of Laos and Cambodia is very small compared to that of Thailand and Vietnam and also with reference to the power potential of new hydroelectric projects. The height of each dashed arrow represents the energy cost of alternative thermal energy in each country. The alternative energy cost in Vietnam and Thailand is much lower than in Cambodia and Laos and in effect indicates the limit of energy cost for hydroelectric projects in Laos and Cambodia that could competitively export to those power markets. 6 The environment and social mitigation costs incorporated in the supply curve analysis are costs provided by project proponents. These estimates have not been submitted to formal EIA / SIA scrutiny in national systems and may not reflect transboundary considerations. 7 1 terawatt-hour or TWh = 1 million megawatt-hours or MWh 8 Some readers may be more familiar with energy cost expressed in US Cents per Kilowatt-hour or KWh, the relationship is 1 USCts/KWh = 10$US/MWh 9 The individual Power Development Plans (PDPs) for GMS countries that were reviewed in the ADB RETA 2240 looking at regional power exchange, which is the basis for the considering the latest official forecasts in the LMB used in this SEA only provide load forecasts to I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

32 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R The horizontal distance between the two supply curves gives the difference in energy potential for any given level of energy cost. Since the lowest alternative energy cost is that of Vietnam (i.e. $73 /MWh) it can be assumed, as a reference, that 70 $/MWh is a reasonable maximum viable energy cost of energy for export. At this cost level, the supply curve for tributary projects shows a potential to contribute approximately 32 TWh of annual energy. The supply curve for all projects shows a potential contribution of 97 TWh. Thus, the mainstream dams add 65 TWh of annual hydroelectric energy competitive against alternative forms of generation in export markets. With this information it is possible to better explain the benefits and risks of mainstream projects to the different LMB national power sectors. Figure 4.1 Assessing the benefits of mainstream projects to the Power Sectors (supply curve) HYDRO EXPORTING COUNTRIES Cambodia and Lao PDR have very high alternative energy (thermal) costs and very small incremental electricity demand in relation to the potential scale and the cost of either set of hydroelectric projects (mainstream or tributary), as depicted by the supply curves in Figure 1. Thus, the differential energy potential and cost between tributary or mainstream developments is less important for supply in these smaller power systems. 10 In other words, the Lao and Cambodia power sectors may be largely indifferent to whether hydropower is supplied by mainstream or tributary projects, except if the mainstream schemes are the only option to obtain the quantum of hydroelectric power needed. To the extent that net revenue from hydropower exports can be reinvested in the power sectors of Cambodia and Laos, the difference in export revenue contributed by mainstream projects and tributary 10 At the same time shaving some form of hydropower supply is important as it implies a difference of two to four times in the cost of generation (thermal) to serve incremental demand. 32 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

33 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R projects could greatly improve access to affordable electricity. This is particularly so in Cambodia, where the grid is more limited. However, it is unlikely that the full differential benefit can be realized until after the debt on the projects is repaid, which probably means some 20 years after their initial date of operation. More specific comment on this aspect would require an evaluation of the terms of concession and power purchase agreements. Most proposed projects are only at the feasibility study stage (MOU), which as described in the baseline assessment report is a very early in the regulatory process HYDRO IMPORTING COUNTRIES In the power sectors of Thailand and Vietnam, the benefit of mainstream hydropower is derived from an inverse situation: low alternative energy cost (thermal) and high incremental electricity demand relative to the hydropower supply curves. As noted in the SEA baseline report, energy demand (GWh) in Vietnam and Thailand combined represents 96 % of LMB electricity demand. 11 Imports of power from LMB mainstream dams would only cover about 16 percent of the incremental annual demand of Thailand and Viet Nam. This contribution offers a cost saving of some 37 percent. Therefore at generation level, the main beneficial impact of power from LMB mainstream projects is to reduce the cost of incremental electrical energy supplied by about 6 percent. 12 Since generation is only about 50% of the cost of service to the final user (the rest is transmission and distribution costs), and since incremental demand growth is only about half of total demand, the actual benefit on total cost of service is probably in the order of 1.5%. This could still be very significant in absolute terms for the whole of the power sectors but unlikely to offer any major change the power utilities financial health, level of power services, or overall industrial competitiveness in the economy due to power supply. 3.3 IMPACT OF MAINSTREAM PROJECTS ON NATIONAL ENERGY POLICES THAILAND AND VIETNAM As the SEA baseline paper on the energy and power theme discusses several factors that shape national energy policies of LMB countries concerning cross-border power trade and their appetite for power imports. One key consideration is that policymakers in Thailand and Vietnam have been progressively relaxing limits on tolerable dependence on levels of imported power in their respective power systems. This has also been reflected in the gradual increase in the quantum of power for cross-border trade that is authorized under the bilateral power trade MOUs over the past decade between Lao PDR and Thailand (now 7,000 MW) and Loa PDR and Vietnam (now 5000 MW). Any continuation of this trend would enable an increasing proportion of imported hydropower from Cambodia and Laos. 11 Official demand projections from PDPs show that in both 2015 and 2025, energy demand (GWh) in Vietnam and Thailand combined will represent 96 % of LMB electricity demand. - Thailand: the power demand is projected to increase by a factor of 2.2 in the next 15 years, with an annual increase of peak demand of 2,600 MW per year in 2025 (equivalent to 3 new 800 MW gas-fired plants per year). - Vietnam: will catch up with Thailand demand in The power demand is projected to increase by a factor of 3.7 in the next 15 years, with an annual increase of 4,600 MW per year in 2025 (equivalent to 6 new gas-fired plants per year). 12 This excludes consideration of issues such as energy supply diversification, particularly important for the Thailand as the policy is to reduce dependence in natural gas as discussed in the baseline assessment. 33 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

34 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R From the supply curves (in Figure 4.1) it can be seen that mainstream LMB projects account for two thirds of the potential quantum of electricity available to import. Such higher tolerance for power imports from hydropower from neighboring counties also lessens the urgency and role of imported coal and nuclear generation in the LMB. These are med- to longer-term alternatives for large-scale power supply, which are identified in Thailand and Vietnamese power development plans (PDP). 13 In addition to cost, some observers feel these technologies (coal and nuclear) have disadvantages relative to hydro imports. This is in terms of strategic issues that range from public acceptance to international relations concerning reducing global GHG emissions (coal) to nuclear non-proliferation questions. The potential of mainstream dams is only 16% of incremental demand in Thailand and Vietnam combined LAO PDR Laos is an experienced hydro producer and has many tributary projects within a size range and energy costs comparable to those of mainstream projects. For example, Nam Theun 2 that has recently become operational has an installed capacity of 1070 MW. The average installed capacity of the nine proposed LMB mainstream dams in Lao PDR is 1085 MW, the largest being Ban Koum at 1,872 MW. Figure 4.2 shows the supply curves corresponding to hydroelectric projects in Laos and relative to its incremental demand. There is a large potential to produce economical electrical energy for domestic supply and export, whether the LMB mainstream projects are included in the supply mix or not. These conditions suggest that even without a foreign partner or export agreement, Laos can probably find the means to finance sufficient hydroelectric capacity to meet domestic demand. In the absence of mainstream dams opportunities Laos would clearly continue with a strategy of hydroelectric development exploiting its large inventory of attractive tributary projects consistent with its policies. Figure Impact of Mainstream Projects in the Laos Power Sector 13 As discussed in the SEA working paper on the energy and power theme the recent PDPs of Thailand and Viet Nam and ADB RETA (6220) show that expansion of power generation from coal in the LMB will be primarily based on import of coal from outside the greater Mekong subregion. 34 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

35 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R CAMBODIA There are very important differences between the outlook of Laos and Cambodia without development of mainstream hydropower schemes in their respective territories. Those differences are well illustrated in Figure 4.2 and the corresponding graph for Cambodia shown in Figure 4-3. Firstly, note the position of the supply curve of tributary projects relative to the incremental electricity demand. The potential for Cambodia to develop tributary projects with an energy cost attractive for export is only 5 TWh, which is about half of the Cambodian incremental energy demand by Secondly, since Cambodia has almost no existing hydroelectric generation, not only is it necessary to provide for affordable supply to meet the incremental electricity demand, but also to replace existing demand now met by expensive diesel generation. The total demand is also shown in Figure 4-3. It would need almost three times the energy potential of tributary projects to be met. The supply curve with all projects shows that after meeting all the domestic demand with hydroelectric power from tributary and mainstream projects, there is only about 5 TWh that could go to export, compared to nearly 75 TWh of exportable power in Laos. The implications therefore If Cambodia managed to develop its tributary projects for domestic supply only, that would only reduce it s extremely high energy cost by some 30%. Without mainstream projects the only reasonable energy strategy would be to develop coal plants and compete with Thailand and Vietnam for hydropower imports from Laos, which would be difficult. In addition, any solution to Cambodia s power demand requires a major expansion in the national grid. Figure Impact of Mainstream Projects to Cambodia Power Sector 35 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

36 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R 3.4 POTENTIAL FOR COLLATERAL EFFECTS ON ELECTRICITY SECTORS This section addresses two questions, namely: (i) could the large capital investments required for mainstream projects adversely affect other power sub-sectors by diverting needed capital from them, and (ii) would the lower energy supply cost derived from mainstream dams reduce incentives for energy conservation through demand side management CAPITAL CONSTRAINTS The magnitude of the equity investments and potential debt burden to develop the mainstream projects is probably beyond the possibilities of the power sectors of Cambodia and Laos as individual developers, even if supported by credible contracts with foreign off-takers. These projects can only be developed jointly by the host country and the export market country (or conceivably a third party foreign investor) under a sophisticated financial and trade arrangement, which may even transcend the electricity sectors to involve commercial commitments of a bi-lateral or regional nature. For the Nam Theun 2 tributary hydropower project, Lao PDR negotiated a concessional loan to finance its equity contribution. Because this is a long-term concessionary loan, the annually debt service payments are considerably less than the equity revenue Lao PDR receives. Thus there is no debt burden. Securing similar concessionary arrangements for LMB mainstream dams would likely be much more difficult for many reasons including perceptions of reputational risk for potential lenders due to the widely recognized controversial nature of the LMB mainstream proposals. Moreover, Information on how much equity and any consequent debt burden or relaxation of income taxes to finance debt is not available. It would make sense that the share of the burden on Cambodia or Laos would only be that warranted by the direct benefit to their power sectors. Therefore no negative impact should result from diverting cash flow from other development priorities. Naturally this assumes competent negotiation of development terms among all interested parties across the whole spectrum of financing considerations ADVERSE SIGNALS FOR ELECTRICITY CONSERVATION The SEA baseline assessment reviewed the current status and trends in demand-side management in the LMB. It focused on energy efficiency (EE) and demand-side management (DSM) progress and potential in Thailand and Viet Nam because these are the main markets for LMB mainstream hydropower schemes. The main conclusion was, with current trends, DSM can sever to reduce the rate of electricity demand growth in the medium-term to longer-term. In Viet Nam in particular, many enabling conditions for DSM institutional arrangements have yet to be put in place and structural change is a long-term proposition. All things considered, electricity demand is elastic with respect to the electricity tariff. Therefore, a reduction in the tariff sends an adverse signal for EE behavior and undermines consumer support for various DSM measures. The effect of mainstream projects on Thailand and Vietnam power cost would be to reduce coats in the order of 1.5%. While at first glance some may consider this detrimental to DSM, the assessment of EE/DSM in the SEA baseline indicates clearly the importance of the electricity tariff level overall in establishing appropriate tariff / pricing signals for EE/DSM, as well as the need to have appropriate tariff structures and time of day pricing as well as the many practical factors that come in play 36 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

37 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R to ensure electricity consumers in all sectors have real and improved access to energy efficient appliances/motors/ equipment by improving the supply chain. In Cambodia, where the effect of reduced tariffs will be most noticeable here probably would be an increase in electricity consumption (per connection) if tariffs were to drop by 30% or more as a result of lower cost supply from mainstream projects. However, that will most likely mean more use of electricity for productive purposes because electricity prices, even significantly reduced from current levels will still be much higher than neighboring countries and high relative to average household income levels. 3.5 DESIGN AND OPERATING ASPECTS OF MAINSTREAM PROJECTS PEAKING OPERATION All mainstream projects have very limited potential to regulate the flow of the Mekong River because their water storage capacity is small in relation to their design discharge. In terms of potential for peaking operation the more likely order of projects is Luang Prabang, Sambor, Pakbeng, Xalaburi, Don Sahong and Sanakham. These plants could store enough water for several hours of operation at full discharge. Other LMB mainstream schemes have lower potential for peaking. Nevertheless, during the dry season it is possible to operate most of the projects in such a way as to maximize generation during hours of high demand. However, regardless of final characteristics, there is nothing in the design of a hydroelectric project that prevents it from being operated at uniform discharge during all hours provided that the flow is not being modified upstream. The regulatory authorities that deal with the use of water have the responsibility to determine tolerable variations in flow regime and reflect such constraints when the MOU or a license to develop a project is granted. Not doing so can result in sub-optimal design and more expansive energy production CAPACITY FACTOR The plant capacity factor is defined as the fraction of the installed capacity of a project that is expected to be utilized on average. The mainstream projects have a rather wide range of capacity factors from 0.32 (Sambor) to 0.63 (Don Sahong). This is intriguing for a group of projects that would have to harmonize their operation and cater to similar power markets. The most likely reason for such differences is how developers have perceived the value of energy at the preliminary stage of design for all mainstream projects. Current designs and proposal have not yet been subject to the rigors of power purchase negotiation, and perhaps more important, securing project finance. It is at that stage that the real optimization of installed capacity (a key driver of cost) and hence capacity factor, often takes place. It is not uncommon for installed capacity to change quite drastically at more advanced levels of project design, after feasibility study. The final project may have very different capacity factors than those currently represented for projects at the feasibility study stage. 37 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

38 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R POWER TRANSMISSION AND LOCAL POWER SUPPLY Most of the power from any large hydro project, tributary or mainstream, goes into a project substation from where a transmission line takes it to the national interconnected grid or to regional grids for domestic supply and export. The transmission line from the project substation to the national grid usually has a voltage level (approximately 200 kv) that is lower than that of the primary grid (approximately 500 kv) but much higher than that of the distribution systems serving final users (anywhere between 5 to 20 KV). Since the regions near the Mekong River are more densely populated and more likely to be served by national or regional grids, the mainstream projects probably have less requirement for extensive high voltage transmission than tributary projects. In addition to the bulk power injected to the high voltage grids, small amounts of power from the plant must be locally available to meet the demand of the power station and its resident staff. This is supplied directly from the substation by a dedicated low voltage local line. It is not uncommon that nearby communities, if not already connected to the national grid, be supplied by this low-voltage line SAFETY OF MAINSTREAM DAMS INFRASTUCTURE RISK All dams represent a significant downstream risk where the potential for loss of life, economic loss and ecological damage from the failure of a dam is very large. LMB mainstream dams, while generally lower in height than tributary dams are no exception, since their failure can impact densely populated areas. Each developer of proposed LMB mainstream dams has designed structures for dam safety according to national regulations and regional practice. Reflecting the need for a consistent approach to dam safety among all LMB mainstream dams, the MRC recently developed detailed guidance for LMB country regulators and developers (MRC, 2009). This guidance is offered not only for the design stages, but also for the implementation and operation phases and is based on international best practice (Preliminary Design Guidance of Proposed LMB Mainstream Dams, Section 5, MRC, 2009). 14 The MRC guidance suggests that dam safety is of paramount importance for the individual dams proposed as well as the safety of any cascade as a whole. It notes the broader philosophy behind good practice today is the safe design, construction and operation of dams depends on more than engineering factors. 15 In this respect, the approach to dam safety must recognize that failure of a dam is a complex process that can include human error in design, construction and operation, maintenance and monitoring stages of projects. A systematic approach and management framework that accounts for the complexities of operation of dams and ensures that the appropriate institutional and communication arrangement are in place is fundamental to achieve and assure dam safety. Emergency preparedness programs are also of fundamental importance to engage downstream communities See FinalVersion-Sept09.pdf International standards include the World Bank Operational Policy 4.37: Safety of Dams and the International Commission on Large Dams (ICOLD) 15 Bulletin on Dam Safety Management, ICOLD, An emergency preparedness plan, specifies the roles and responsibilities of all parties when dam failure is considered imminent, or when expected operational flow releases threatened downstream life, property, or economic operations that depend on river flow levels. The plan itself is prepared during implementation and is in place before the projected initial filling of the reservoir. In particular, it is important to have a clear communication strategy to engage with stakeholders on dam safety issues and emergency preparedness activities that directly involve or affect them. 38 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

39 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R There are different modes and degrees of severity of dam failure. These range from failure events that affect dam facilities, but leave the main structures intact, to actual structural failure of the dam itself. From a design perspective, the safety of a dam is determined by a combination of factors including the magnitude of the flood event it can tolerate without risk of a catastrophic failure that could result in a devastating flood wave downstream. International organizations, engineering societies and local regulators establish strict guidelines on how such hydrological events should be selected and guidelines establish the probability of the flood event to be tolerated. Historically it was usually a 1-10,000 year flood for any dam that posed a risk to populated areas. The extreme events associated with such probabilities are derived from historical records of dams, but several decades ago it was realized that in the context of climatologic events, history is a not a reliable indicator of future behavior. Thus long before climate change became a household term a different criterion was established for the safe design of large dams in populated areas. This is based not in historical events, but on a hydro-meteorological calculation of the worst possible flood event at a particular site called the Probable Maximum Flood or PMF. The PMF can result in designs of spillways, for example, that are more than double the size of design to 1:10,000 flood event. It is important to note also that occurrence of a 1:10,000 or PMF flood event does not imply catastrophic failure of dam structures. Low-head dams can be designed withstand being overtop if the spillway design capacity is exceeded, but there would often be significant damage to equipment. The power station may be out of operation for a considerable time. Where a flood were to exceed the spillway design capacity, and depending on the type of dam, the outcome may be only limited local damage that is overshadowed by the wider damage that occurs in flood channels as the river reached extreme flow levels. In practice, the low-head Mekong LMB mainstream dams would certainly have to be designed to avoid complete failure. One example of such failure is the largest dam failure incident in North American history. This occurred in Quebec, Canada in 1989 in the Saguenay River 17 in 1989 recorded as the largest overland flood in 20th century in the country s history. In this case, eight dams were overtopped and thousands of people were displaced. No concrete dams or concrete rock-fill dams failed, but embankment dams did. There was some loss of life, considerable economic loss and massive ecological damage that was sediment related. The major damage was due to the overwhelming impact of the occurrence of the 1:10,000 year flood. Without a doubt, the highest priority for any agency responsible for regulating the development of LMB mainstream dams is to ensure that they all meet PMF criterion. Secondly, that the PMF is calculated at each site by the same methodology. In this respect the MRC s Preliminary Design Guidance calls for this approach. With the adoption of international best practice there will be an expert dam safety panel for each dam that reviews designs and actions at each step in construction and operation. The actions at each stage will be measured for conformance to international best practice and open to public view and scrutiny. In addition, the MRC s PDG identifies the need to evaluate the safety of the entire cascade of proposed LMB mainstream dams and the need for coordination and accountability arrangements for operation stages, if more than one LMB mainstream dam is eventually built I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

40 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R 3.6 MAINSTREAM HYDROPOWER DEVELOPMENT PLANNING The nature of the current development of hydropower in the LMB that involves soliciting private sector investment can be characterized as "opportunistic" in the sense that a potential developer negotiates directly with the licensing authority on the available sites, and subsequently, all developers (in many respects concerning the LMB mainstream dams) compete to be the first project to be approved. This is against the background of less certainty when the second or subsequent mainstream project would be approved as such projects depend on market conditions. Hydropower development in this context does not necessarily follow a prescribed plan of river basin optimization and strict economic merit. Rather it moves in response to the interest expressed by developers through initial MOUs that are a form of proxies to formal development licenses. In many respects this is not very different to what happens in other hydroelectric markets in the world that have more structured sets of regulations, licensing procedures and power pricing mechanisms. In those settings the benefits of river basin optimization can be set aside in the quest to attract foreign investors to this capital intensive industry. However, in the view of many the implications of Mekong LMB mainstream projects are too significant for opportunism to be the main determining factor in this case. The scale of these projects, the transboundary nature of their areas of influence and the uncertainties about their physical, social, environmental and economic impacts all call for a more cautious and rationale approach. There is ample room for optimism about the evolution of practices in the LMB. For example, after the LMB mainstream sites were studied (Acres Report, 1994), the 1995 Mekong Agreement was signed. The 1995 Agreement embodies the basin development plan (BDP) that looks at the development of hydropower in river basin context based on IWRM principles. The proposed LMB mainstream projects are also subject to Procedures under the 1995 Agreement, notably the PNPCA. At this time (2010) only the feasibility and EIA/SIA studies of mainstream proposals have been undertaken, or are at various stages of preparation. Moreover, this SEA is another example of MRC consideration of the mainstream proposals in a river basin context that will eventually inform the MRC PNPCA procedure, if and when it is triggered. 3.7 CONCLUSIONS REGARDING POWER SECTOR IMPACT Mainstream hydroelectric projects in Laos and Cambodia represent roughly two thirds of the total energy potential of hydroelectric projects identified in the LMB not yet operating or undergoing firm development. The overall power benefits to LMB countries are significant. The difference in net overall economic power benefit between the 20-year probable future scenario with and without proposed LMB mainstream dams is about $US 4 billion annually by It can be seen the greatest beneficiary is Lao PDR, in all cases (table 4.1). 40 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

41 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E N E R G Y A N D P O W E R Despite the fact that the energy from mainstream projects would be used regionally and setting aside considerations of energy supply diversity and export earnings, from a power supply perspective and the consideration of alternatives the development of mainstream projects more critical to the power sector of Cambodia. Several circumstances determine that the two mainstream projects in Cambodia (Sambor and Stung Treng) are important to Cambodia s power sector from a power supply perspective. Firstly, Cambodia has a very expensive generation system almost entirely dependent on imported oil. Thus, not only does Cambodia have to provide affordable power to meet incremental demand, but it also needs to replace its existing generation as much as possible. Secondly, Cambodia has a very small inventory of attractive tributary projects. The energy potential of these projects is not sufficient to meet incremental demands, let alone replace existing generation or export. Thirdly, Cambodia has virtually no experience in hydroelectric development or operation and thus will rely much more than Laos on foreign partnerships, which can only be attracted by mainstream projects that can enable power exports. It may be argued that the scale of electricity demand in Thailand and Vietnam and the cost of alternative forms of power supply (thermal) are such that development of LMB mainstream projects would have a minor impact on electricity prices, and not radically alter the energy supply strategies of those countries - applying least-cost criteria alone. For Lao PDR, the large inventory of economically attractive tributary projects means that the Lao hydroelectric industry can develop tributary projects for domestic use and export can continue at a healthy pace without mainstream projects. Nevertheless, the potential scale of annual export earnings would be significantly reduced, as shown previously in table 4.1 and the advantages of investing in stable and low-cost power supply on future especially when concession terms are completed would be forgone. 41 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

42 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S 4 ECONOMIC SYSTEMS Key strategic issues (relevant to hydropower) 1. What are the broad national and regional economic implications of large scale natural resource based development in the LMB countries? 2. What are the economic costs and benefits of sectoral development in the LMB countries? 3. What is the distribution of economic benefits between different areas, groups and sectors? 4.1 SUMMARY OF PAST AND FUTURE TRENDS WITHOUT LMB MAINSTREAM DAMS Three main areas pertinent to mainstream hydropower development were identified in the baseline economics analysis. These were macroeconomic implications of large scale natural resource development in the LMB (particularly for Lao PDR), sectoral economic trends in the LMB and distributional trends in the LMB. Macro-economic opportunities and risks relate to the large scale and rapidly increasing levels of investment in natural resources (i.e. hydropower, mining and plantation development) in particular in Lao PDR (and possibly Cambodia). This investment is largely composed of FDI. This potentially represents a significant boost for this small economy crowding in investment and increasing consumption across a number of sectors. Conversely, such rapid growth in these natural resource sectors potentially poses a risk to competing sectors through driving up relative price levels resulting in exchange rate appreciation. This has the potential to reduce the internal and external competitiveness of other sectors in the economy (such as agriculture and manufacturing). The baseline analysis also noted potential fiscal implications of increasing government debt burdens in Lao PDR as a result of funding the purchase of equity in hydropower projects. This greater debt burden needs to be balanced against greater government revenues generated from hydropower development. The sectoral analysis examined economic trends in fisheries, agriculture, construction, navigation, tourism, forestry, mining and industrial sectors and also looked at values of key environmental assets including aquatic plants, wetlands, flood control and saline intrusion control. In general sectors based on biotic natural resources are stable or in decline, whereas industrial use of natural resources and related sectors are expanding rapidly. The distributional analysis found that while overall population was likely to grow, rural populations in the basin were likely to remain relatively stable despite high natural growth rates as rapid rural-urban migration continues. Poverty rates in the basin are higher in remote upland areas and lower closer to the main stream and in larger urban settlements. Nevertheless, as population densities are low in upland areas the absolute numbers of poor are greater closer to the mainstream and in urban areas. This trend is increasing with rural-urban migration driven by declining natural resources bases in upland areas and increasing employment opportunities in lowland and urban areas. Despite rapid growth in industry and service sectors in terms of employment and livelihoods for agriculture and fisheries remain important for rural livelihoods across the basin. 4.2 EXPECTED MACRO ECONOMIC TRENDS WITH MAINSTREAM DAM CONSTRUCTION 42 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

43 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Macro-economic impacts of MSHP development are not likely to differ qualitatively from current trends. Additional FDI flows, revenues and state debt burdens are likely to be higher as a consequence of these projects. The following sections look at these potential impacts in greater detail CAPITAL INVESTMENT As noted in the baseline analysis, according to available data (from the MRC IKMP) investment in hydropower over the last two decades has been expanding rapidly. From 1990 to 2008 an estimated (nominal) USD 6.6 billion was invested in hydropower development in the LMB. Excluding mainstream dams from the investment figure is expected to reach USD 9.1 billion. In terms of annual average investment in hydropower this represents an increase from USD 365 million a year between to an expected USD 1.3 billion a year between The proposed investment in 12 mainstream hydropower projects, which are all large projects (with an average investment cost of around USD 1.9 billion 18 ) would add an estimated USD billion between 2016 and Based on available investment schedules, mainstream hydropower development would imply investment for the period of of around USD 1.5 billion a year. Figure 1 below gives an estimated investment schedule 19 for hydropower development in the LMB. Most of the funding for these developments is expected to come from sources external to Cambodia and Lao PDR. Figure 1: Estimated annual investment in LMB hydropower Source: ICEM based on IKMP hydropower database, MRCS 2008 (these figures do not include the Don Sahong or Thakho projects as information on these projects is not available in the database) 18 The two Cambodia projects, Sambor, and Stung Treng are by far the biggest with an estimated investment cost of USD 3.9 billion and USD 2.3 billion respectively, the 8 Lao PDR dams for which there is cost information have average investment cots of USD 1.6 billion. All figures are drawn from the IKMP hydropower database. 19 This schedule is based on the best available data. However, changes in design, other unforeseen construction contingencies and changes in price levels mean investment costs liable to significant changes. Moreover, the time schedule for these projects is likely to change also depending upon a range of contingencies. 43 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

44 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 2: Hydropower investment in Lao PDR Source: ICEM calculations based upon IKMP hydropower database 2008, WDI GDP figures 2009 and APERC GDP projections 2009 Figure 2 serves to emphasize the size of these investments relative to the GDP of Lao PDR. In the expected peak years of hydropower investment between 2016 and 2019, investments in mainstream hydropower alone may reach up to USD 2.5 billion annually, a figure which is likely to account for around a third of expected GDP at that time. This investment is in addition to any other investment in the country, other natural resource sectors such as mining and plantation development could also be a significant source of FDI. Inward investment is not particularly significant relative to the size of any of the other LMB economies. In commenting upon the impact of MSHP we assume the investment and revenues generated by these projects are additional. That is to say, these projects represent specific investment opportunities and attract investment to the host countries which they would otherwise not experience REVENUE GENERATION The 12 MSHP represent a significant opportunity for the generation of revenues for equity holders and host governments. For the 10 MSHP projects in Lao PDR gross revenues are likely to be in excess of USD 3 billion per year 20 and for the two Cambodian dams USD 1.3 billion per year. Hydropower projects have high capital investment costs relative to incremental (or recurrent) costs. As a result in the early years of operation a significant portion of the revenue is used to repay debts (in green in Figure 3 below) generated in the construction period (the black, grey, light blue and pink in Figure 3). However, once the bulk of the debts are paid off projects can generate large revenues net of financing and O&M expenditures. Revenues are 20 The figure for the 8 Lao PDR dams included in the database is USD 2.9 billion, the inclusion of the remaining 2 dams would in all likelihood bring this total to around USD billion. 44 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

45 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S generally divided into dividends, taxation and payments for water rights. Water rights payments are usually a unit charge on water resources and will therefore constitute a relatively fixed portion of revenues. Taxation and dividend payments are usually payable on profits and therefore increase as debt service demands decline. While Figure 3 gives a reasonable guide to revenue flows for typical hydropower projects, the structuring of the financing, equity and concession agreements are likely to vary for each of the MSHP projects. In particular, host governments are likely to take equity stakes in the projects which will increase revenues payable to governments (in the form of dividends), but are also liable to increase levels of public debt (to finance the equity), at least in the short term (see section below). Figure 3 21 : Typical cash-flow for hydropower projects Source: Tracebel Electricty and Gas International, presentation for the World Bank 2005 Thus a typical hydropower plant can be expected to generate considerable revenues for owners and government after the initial 5-10 years of operations when the bulk of the project financing has been paid off. MSHP projects are likely to be financed mainly through private sources (aside from a relatively small government equity stake referred to above) on a BOT or BOOT basis, with operating concession lasting for about 25 years. Government revenues in the first few years of operations can be expected to be a relatively small portion of the gross revenue. For example, net government revenues generated by NT2 are estimated to be around USD 80 million a year (or USD 2 billion over the period) between commissioning and the end of the concession agreement when ownership of the plant is transferred to the government of Lao PDR This is a generic guide to typical projects, depending on the structure of the financing, off-take and concession agreements details of cash flow may vary. For example, NT2 hydropower project has an additional revenue stream in the form of concession payments to the Lao PDR government. 22 World Bank (2005) Project appraisal document on proposed IDA grant for the Nam Theun 2 hydroelectric project. 23 These estimates of net revenue does not take into account indirect effects of the projects such as increase tax returns in other sectors due to investment stimulus effects and negative impacts such as increased debt service costs due to increased public sector debt and therefore risk premiums. 45 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

46 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Net government revenues in the short term from the predominantly privately financed MSHP projects will therefore be less than gross revenue figures, or figures on the economic value of these projects might suggest. Net revenues for government in the early years of these projects may be small relative to the size of the projects. Using NT2 as a guide we could expect revenues to be 20% of gross revenues for the concession period. Although as MSHP projects do not yet have recourse to concessional financing and technical assistance to support their development it is likely that net government revenues will constitute a smaller portion of net revenues than has been the case with NT2. Nevertheless, MSHP developments are expected to represent a significant source of increased government revenues. As recent IMF figures (Figure 4) revenues from the natural resource sector are an increasing important source of government revenues. Similarly, recent estimates from the World Bank 24 suggest hydropower projects now under construction in Lao PDR will lead to an increase in annual government revenues for USD 130 million to USD million by With mainstream hydropower these figures would increase substantially. Figure 4: Government of Lao PDR revenues and revenues from the natural resource sector (nominal USD) Source: IMF Country report No. 09/285, 2009 (Note: Resource revenues consist of royalties and taxes from mining and hydropower projects) As the discussion above suggests, the ability of host countries to generate net revenues from these (and other) hydropower projects depends upon, i) the capacity and political will to negotiate favorable PPA agreements with power importers and concession agreements with developers and financiers; and, ii) the ability of the host governments to effectively oversee the development and operation of the hydropower projects such that contractual obligations are respected by all parties and adequate government revenues are generated I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

47 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Finally, it should also be noted that additional government revenue of itself has no welfare implications. Rather additional revenue offers the potential for significant increases in development expenditures. Whether or not these are realized and whether or not they are effective in realizing welfare gains will be addressed in greater detail in the mitigation phase of the SEA. Any developmental gains due to increased hydropower revenues are not automatic and are dependent upon subsequent decisions made by government INVESTMENT STIMULUS Any economic stimulus effects are unlikely to be qualitatively different from those felt as a result of other large scale FDI investment in the region. However, any multiplier effects will be correspondingly larger due to the size of these developments. As with other hydropower development the extent to which development of these projects acts as a stimulus depends upon how much of the investment capital results in local consumption. As Figure 3 shows typically civil works (essentially construction activities) constitute a large portion of the expenditure on hydropower relative to other types of power generation which usually have larger engineering expenditures. This means that while a considerable portion of the investment will pass through Lao PDR and Cambodia in the purchase of engineering equipment from abroad (and possibly outside the LMB region) a considerable portion of the funds will be used in the purchase of local inputs such as construction materials and labour. Although in most cases skilled labour is likely to be imported. Nevertheless, as World Bank estimates of GDP growth due to natural resource investment over recent years (shown in Figure 5) illustrate the GDP impacts of hydropower development are likely to be significant, with hydropower currently under development expected to account for 3.9% GDP growth in 2010 and 2.3% of GDP growth between 2011 and 2015 (World Bank 2009). Figure 5 : Resource sector contribution to GDP growth ( ) Source: World Bank, Laos Economic Monitor 2009, Volume 14 (Note: * average for ) REAL EXCHANGE-RATE APPRECIATION AND BOOMING SECTOR EFFECTS The added magnitude of FDI and foreign exchange earnings from hydropower if not managed in a financially prudent manner may imply risks related to real exchange rate appreciation caused by inflationary pressures in non-tradable goods sectors. This may make tradable goods sectors, which are in competition with the booming hydropower sector less competitive. In the cased of Lao PDR this could be significant as the sectors at risk 47 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

48 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S (manufacturing and agriculture) are likely to be important for poverty reduction. This effect may therefore constitute another transmission mechanism through which the poor may be adversely affected from hydropower development (also affected through direct sectoral impacts see section Error! Reference source not found. below). This is essentially part of the wider implications of the stimulus or multiplier effect of the proposed investment. The magnitude of this effect will depend on the extent to which these projects are pass through or enclave projects which source inputs and receive investment from outside the host country. As already indicated in sections and above, while a large proportion of the investment capital will be used to purchase inputs aboard and possibly outside the region, the additional stimulus effect of the mainstream hydropower investments is likely to be significant. Moreover, net revenue generation will also increase the potential for greater government spending. In the case of a small economy like Lao PDR the magnitude of these effects is likely to be large relative to the size of the economy. The extent to which these revenues are likely to result in real exchange rate appreciation, or alternatively are used to ameliorate the negative impacts of real exchange rate appreciation on other tradable goods sectors will depend upon how the host government chooses to manage this revenue. There are two important aspects of this firstly, the extent to which the government has the capacity to manage these revenues, and secondly, the extent to which the host governments are under pressure to disperse these funds for various programmes which are unlikely to be of benefit to the tradable goods sectors and in fact represent expenditures which may lead to exchange rate appreciation. Thus it is difficult to draw definitive conclusions regarding the possibility of this effect. It is also important to note that if any real exchange rate appreciation takes place it is likely to have negative distributional consequences as sectors which are most likely to be adversely affected by this phenomenon are just those which are of long term importance for poverty reduction, in particular the agricultural and manufacturing sectors PUBLIC DEBT SUSTAINABILITY IN LAO PDR Host governments are likely to take an equity stake in these proposed hydropower developments. This will imply raising the national debt burden. In whatever way the financing is structured for these projects some state body will necessarily have at least some liability for this debt 25. Lao PDR in particular still faces a high risk of debt distress because of its large debt stock. In respect of debt sustainability, the significance of the proposed developments lies in the extra magnitude of debt they imply and to what extent any additional debt burden is likely to be off-set by state revenues generated by these projects. This will depend upon the structure of each individual project s finance for which there is little available information at present CONCLUSIONS The proposed hydropower developments do not represent a qualitative divergence from the macro economic issues described in the baseline report. Rather the proposed developments will represent a much higher magnitude of investment meaning the financial opportunities and risks are also likely to be much larger. The opportunities offered by these developments in terms of i) economic stimulus; and, ii) revenue generation while depending upon project details are relatively clear. The extent to which negative impacts through exchange rate appreciation and debt sustainability are likely to pose a problem will depend on the state management of the macro-economy and revenues generated from hydropower development. This will in turn be dependent upon 25 World Bank, Laos Economic Monitor 2009, Volume I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

49 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S both the political will and administrative capacity 26 of the governments to manage the economy such that it maximizes these benefits. This is closely related to overall mitigation measures and will be discussed in the mitigation, avoidance and enhancement section of the SEA. Moreover, as the power paper points out, the outcome of any agreement reached between off-take countries, host countries and developers will depend upon not just economic factors but also geopolitical considerations, and importantly the capacity of host governments to negotiate an agreement that is likely to be favorable to the host country. It is therefore not a foregone conclusion that, from a macro-economic and financial perspective that the impact of these projects will necessarily be benign for all sectors of the economy 4.3 EXPECTED SECTORAL EFFECTS ON ECONOMIC TRENDS Mainstream hydropower development is likely to have wide ranging effects across a number of economic sectors, economic assets and ecosystem services ECONOMIC SECTORS A number of sectors are likely to be significantly affected by mainstream dam construction. Sectors which are likely to experience significant impacts include fisheries, agriculture and forestry, tourism, navigation, construction, and mining and industry sectors. The likely impacts of the mainstream hydropower developments to these sectors have been estimated where adequate information has been available in terms of changes in output. For example, for paddy production losses and gains due to the developments have been included. Greater analysis of these changes is included in each sectoral theme paper. 26 A recent review of revenue management arrangements by the World Bank for NT2 cites systematic issues with financial administration capacity (Presentation on Nam Theun 2 Revenue Management Arrangements Water Week 2007, Mohib, A.S. World Bank) 49 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

50 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Table 1: Overview of economic risks and opportunities in key sectors Net loss/gain due to MSHP Description Indicative values Cause of loss/gain development -700,000 to -1.4 million tonnes Loss (capture USD 2.0 to 3.6/kg = USD 2,000 fisheries) 3,600/tonne -USD 1.4 billion to -USD 5.04 billion Fisheries Fish Riverbank garden production Gain (reservoir fisheries) Loss (marine fisheries from loss of nutrients to the sediment plume) Loss (riverbank gardens) 25,000 to 250,000 tonnes (63,000 most likely) USD 2,000 3,600/tonne USD 50 million to USD 900 million -4,535 tonnes of phosphates to marine area/year Replacement value of around USD 40 million/year -54% of river bank gardens loss in zones 2,3 and 4 USD 21 million/year Future trends with MSHP (narrative) Fish migration routes will be blocked and flood pulses will be disrupted, decline in fish populations likely to result. This is likely to be true both for migratory species and species which depend upon flood plains for Some of this loss may be off-set by the introduction of reservoir aquaculture but potential yields from this remain highly uncertain. Mekong delta marine fisheries estimated catch in 2008 was 563,000 tonnes worth between USD billion. Productivity of the fisheries in this area is closely related to the sedment plume and associated nutrients delivered by the Mekong river. No data available linking nutrient levels in sediment plume with fisheries productivity. Replacement cost of nutrients used as a basic indicator. Loss of river bank gardens due to inundation of long stretches of the mainstream river in zones 2, 3 and 4. This estimate does not include any difficulties in cultivating riverbank gardens downstream of the dams. Agriculture and forestry Loss (inundated paddy and transmission lines) - 7,962 ha of paddy - 22,475 tonnes of rice/year - USD 4.1 million/year Relatively small losses from inundation of paddy more than offset by gains resulting from increased irrigation associated with mainstream hydropower development. Paddy production Loss (value of nutrients (Phosphates) to agriculture) -3,400 tonnes of phosphates to flood plains/year Replacement value of fertilizer around - USD 24 million/year Any reduction in sediment load and flooding will lead to a decrease in associated nutrient replenishment. Measured as loss of phosphates due to sediment trapping at each of the MSHP dams. While productivity implications for agriculture could not be calculated, cost of artificial replacements given. 50 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

51 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Gain(increased irrigation) 17,866 ha of paddy 77,701 tonnes of rice/year USD million/year Irrigation projects associated with the hydropower developments are likely to improve land productivity and rice production in some areas. Tourism Tourism revenues Loss (degradation of natural resource base) N/A Some valuable environmental assets upon which burgeoning ecotourism industry is based will be degraded or lost due to changes in the hydrology and ecology of the mainstream resulting from these projects. Gain (HP project viewing) N/A Large hydro-electricity projects often attract (mainly domestic) tourism (for example Hoa Binh dam in northern Vietnam) Mainstream hydropower is likely to increase the navigability of the river as it will increase the depth of the river along significant stretches. However, this will be dependent upon designing dams such that they allow navigation. Which projects go ahead will affect the overall navigability of the river. Therefore, impacts on navigability and are dependent upon dam design Navigation Freight transport Passenger transport Gain (increased navigability) Gain (increased navigability) Loss (decreased longitudinal connectivity) N/A N/A N/A Even with navigation locks these projects will increase the time taken and probable costs of navigation Construction Sand and gravel extraction output Loss (reduced sediment load) N/A Unlikely to be significant in the short term. Aquatic plants Subsistence Loss (loss of habitat) N/A Changes in mainstream habitats will increase loss of currently economically important aquatic plants. Wetlands Clean water supply, plants for food and medicines, fuel wood, nutrient recycling, water purification wildlife habitats groundwater recharge, flood control, carbon sequestration, storm protection etc Loss (due to reservoir creation) between USD 4 million and USD 13.8 million per year (2000 prices) Most of the in-stream wetlands will be lost in zones 2, 3 and 4 with significant impacts on their productivity and the in all likelihood other ecosystem services they provide (for more information on the calculation of these values see background methodological paper) 51 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

52 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Flooding/floo d control Saline intrusion Nutrient replenishment, wildlife habitat, damage to goods and livelihoods Gain (reduction in flooding) N/A Some minor flood control effects Crop productivity N/A N/A No significant impact Source: All figures are derived from baseline and thematic papers 52 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

53 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S ASSESTS A range of fixed assets in and around the mainstream will be directly and indirectly affected by the development of the proposed mainstream hydropower projects. Fixed assets likely to be affected by hydropower development include the following: Houses and commercial property Residential infrastructure (water supply, electricity supply, waste water facilities etc) Industrial facilities Amenities and social infrastructure (schools, clinics etc) Roads and land transport infrastructure Moorings and docks Irrigation pumps and irrigation infrastructure 27 Land (agricultural and forestry) The social theme paper outlines the impacts in terms of population and households resettled, which reflect the current best estimates available at 63,112 people to be resettled in 6,847 households. Assuming that each household has a separate dwelling, this means 6,847 houses will need to be replaced due to the proposed development. Associated with the inundation of housing and settlements in impoundment area residential infrastructure and other amenities are also likely to be adversely affected. This includes electricity and water supplies, any waste water facilities and roads, as well as public amenities such as schools, clinics and government offices. In some places there are also industrial facilities close to the river positioned to take advantage of river transport. For example, there are numerous saw mills on the river in and around Pak Lai, many of these facilities are likely to be affected by raised water levels. These costs will be significant although at present no valuation data is available. Infrastructure actually in or on the river such as irrigation pumps and boat docks and moorings will also be affected. In this case the effects are likely to go beyond the upstream impoundment areas of the dams and also have effects down-stream due to changes in water level and sedimentation. In the case of irrigation pumps there are generally three types. Free-floating irrigation pumps placed on small wooden pontoons while allow the pump to access water throughout the year. While these pumps are common they typically small and provide water to a small local area. Other irrigation pumps are fixed upon the river bank and can be moved up or down depending on the water level in the river. Finally, the larger irrigation facilities are generally fixed and have a fixed inlet in the river. These last two types of irrigation pump are likely to need relocating as a result of hydrological changes resulting from dam construction. Up-stream facilities may be inundated; others may need to have their intakes modified to make better use of changed hydrological conditions. Down-stream fixed irrigation pumps may be subject to greater levels of erosion or deposition of sediments and need to be altered accordingly. If the dams are managed to provide peaking power changes in water levels on a daily basis will mean that irrigation systems may have to be managed differently. However, in the case of irrigation, more stable water levels in impoundment areas. 27 These are likely to be largely off-set by improvements in irrigation in the impoundment areas. 53 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

54 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Similarly, navigation infrastructure, and in particular docks and moorings for boats will be affected. Upstream some moorings and docks are likely to be inundated and will need to be reconstructed at a higher level. Depending on the management of the dams (i.e. if the water is drawn down in dry season to provide peaking power supplies) this may mean that for part of the day some of these facilities may not be accessible. Down-stream facilities may be subject to daily fluctuations in water levels (depending on the management of the projects) and changes in geomorphology including increased deposition in some areas and increased erosion in others, implying a need for remedial action (dredging to allow channel navigability or rip-wrap protection for facilities subject to erosion) (see hydrology theme). In general changes in the deposition and erosion of sediments due to the construction of the dams will mean significant changes to the geomorphology of the river. Together with the possibility of daily fluctuation in river levels due to management of the projects for peaking power, there is likely to be implications for all kinds of infrastructure in and one the river as well as assets based near the river which may be affected by declines in bank stability. This is likely to imply either the loss of assets or costs implied by improving protection around vulnerable locations. Finally, there is likely to be significant land lost in impoundment areas due to rising water levels. The terrestrial paper estimates the amount of land likely to be lost, these results are reported below. While the areas of land lost are not high relative to other hydropower projects which can have large impoundment areas, in some areas this composes a significant proportion of the available land, for example in zone 2, where much of the suitable agricultural land is close to the mainstream. This means that land loss may have a disproportionate effect on populations in this areas as alternative land is not available ECOSYSTEM SERVICES The baseline report mentioned two key ecological services related to mainstream flows saline intrusion control and flood control. These both imply important, if hitherto unquantified, economic values. However, the hydrology analysis suggests that the impacts of mainstream hydropower development on these services will be limited. Flow rates in the dry season will be effectively unchanged by mainstream development meaning they will imply no improvement in the control of saline intrusion. Wet season flooding will be slightly reduced, however, this change is marginal and there is inadequate information to enable this effect to be quantified. The third area of ecosystem services that was considered in the baseline were those services provided by the wetlands which include clean water supply, plants for food and medicines, fuel wood, nutrient recycling, water purification wildlife habitats groundwater recharge, flood control, carbon sequestration and storm protection amongst other things. Values for the wetlands were adopted from a transfer pricing exercise conducted in the by Schyut and Brander (2004), and arrived at an estimated value of between USD 4.2 million and USD 14.2 million/year (2000 prices). GIS analysis conducted for the aquatic systems paper suggests that the primary productivity of wetlands in the mainstream will be reduced in 735 Km 2 of the in-channel wetlands which will become permanently inundated. The value of the wetlands lost is estimated to be between USD 55 and USD 188 /ha/year. Meaning the loss of wetlands due to mainstream hydropower development is estimated to be between USD 4 million and USD 13.8 million per year (2000 prices). 54 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

55 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Nutrient flows in the basin also represent a valuable ecosystem service and have an important role to play in agricultural and fisheries (both fresh water and marine) productivity. These values have been addressed under the description of the fisheries and agricultural sectors. 4.4 EXPECTED DISTRIBUTIONAL EFFECTS ON ECONOMIC TRENDS The economic costs and benefits of the proposed developments will not be borne equally by all groups or in all locations in the LMB countries. This section examines the distribution of economic opportunities and risks associated with mainstream hydropower development across different groups and areas in the LMB. The distribution of benefits are examined in a number of different dimensions including geographical location, social group and urbanization level DISTRIBUTIONAL EFFECTS BETWEEN RURAL AND URBAN AREAS Opportunities and risks are likely to be unevenly distributed between rural and urban areas. However, both urban and rural areas are likely to realize opportunities and risks as a result of mainstream hydropower development. Urban and peri-urban areas which contain most of the manufacturing industries (such as the Bangkok and the surrounding areas in Thailand, Phnom Penh and Shinoukville in Cambodia and the South Eastern region around Ho chi Minh City in Vietnam), will be those that reap the benefits of cheaper and more reliable electricity (especially in the case of Cambodia where expensive electricity has been identified as a key barrier to competitiveness (see power theme paper and IMF 2009)). However, this benefit should not be overstated as in the case of Vietnam and Thailand power demand would be met through other forms of generation, even if costs were slightly higher given the small portion of over power supply the mainstream dams would account for, this would likely have a very small impact on power price (see power theme paper). Rural communities are unlikely to reap significant economic benefits from improvements in the power supply. While local communities may see the extension of electricity grids, in the immediate vicinity of the hydropower projects, project development is unlikely to have any significant effect on direct access to power supply for the more general rural population. The section on potential macro-economic impacts of mainstream hydropower development focusing mainly on Lao PDR - stresses uncertainty over whether potential macro-economic risks will be realized. Nevertheless, if booming sector effects are experienced this is likely to have a differential impact on rural and urban populations. In particular, non-tradable sectors (i.e. services) usually concentrated in urban areas are likely to perform better than sectors that compete with the booming sector or sectors that compete with imported goods. Conversely, the manufacturing sector, also based in urban areas may suffer from declining competitiveness. Rural areas may see a decline in the relative value of their agricultural and other products, posing a possible economic risk. Without undertaking a comprehensive and detailed modeling exercise it is not possible to evaluate the impact on different sectors and therefore differential impacts on rural and urban areas. Moreover, it is not clear that this effect can be ascribed to mainstream developments as any large capital developments are likely to imply similar effects. Finally, rural livelihoods are at much greater risk than urban livelihoods. As the analysis in the economics, social and fisheries paper have shown, rural livelihoods within the LMB are still largely dependent upon 55 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

56 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S the natural resource base. Throughout the basin, household rely on capture fisheries, agriculture (irrigated rice, rain fed rice, field crops and swidden agriculture), NTFPs and a number of other natural resource based activities, or ancillary services associated with them (e.g. boat and fishing net manufacture (see economic baseline paper)) for their livelihoods. In contrast urban dwellers tend to rely more heavily on industrial and service sector employment 28. This in turn means that any direct impacts of mainstream dam development such as the expected depletion of fisheries, are likely to be borne predominantly by the rural populations. However, these impacts may lead to secondary effects such as increased rural-urban migration as the depletion of the natural resource and increasing attenuation of rural livelihoods will constitute a significant push-factor in the decision to migrate to urban areas. This may, in turn, lead to increased urban poverty levels. Moreover, it is not clear how fish price levels will respond to depletion of fish stocks. A significant increase in price may affect urban and rural dwellers alike. In fact the urban poor may suffer more than rural populations in the event of increased food prices due to mainstream hydropower development DISTRIBUTIONAL IMPACTS: POOR AND NON-POOR Likely distributional impacts of the mainstream hydropower development are also likely to be unevenly distributed between the poor and non-poor. As the baseline report has shown poverty rates and distributions vary across the LMB. Higher poverty rates are usually found in remote up-land areas away from the mainstream, however, higher population densities mean that the absolute number of poor is higher in low-land areas closer to the mainstream. This trend has been increasing as increasing livelihood opportunities develop in low-land (and associated urban areas) coupled with the depletion of the natural resource base and attenuation of associated livelihoods in upland areas. While it is not possible to project poverty rate or distributions to 2030, these trends are, in all likelihood, expected to continue. Moreover, while overall poverty rates are expected to decline, poverty and increasing inequality is likely to remain a problem. The poor tend to be more vulnerable to adverse changes in environmental conditions because they have fewer assets, savings, skills and knowledge that can allow them flexibility in making necessary changes to their livelihood strategies in response to changes in environmental conditions. In the case of mainstream hydropower development the expected loss of fisheries is likely to be key impact on the poor. It is difficult to estimate the importance of fisheries to the poor in the LMB, in Cambodia it has been estimated that over 1 million people who in some way depend on these at risk fisheries (see social theme paper), however there is no information available indicating what proportion of them are poor. Assuming that the poverty rate amongst this group is the same as nationally, this means around 40% or 400,000 poor fisher folk s livelihoods are at risk. As most fisher-folk are from areas where poverty rates tend to be above the national average it is safe to assume that (with the possible exception of Lao PDR where upland areas are some of the poorest), amongst fisher-folk poverty rates are higher than national averages. Moreover, analysis in the baseline fisheries, economics and social assessment papers indicates that LMB populations are highly dependent upon fish as a source of protein. Meaning that aside from providing employment, this resource is also important from a consumption and nutritional perspective. Given the extent of the expected reduction in fisheries, it is likely to have a significant impact on the nutritional status of the poor 28 It should be noted that in some areas (for example the Mekong Delta), much of the manufacturing industry is dependent upon locally produced natural resources (e.g. the rice processing industry), so this distinction is not as clear cut as it may at first seem. 56 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

57 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S as they currently depend proportionately more on fish (and other aquatic animal) consumption than other groups and they are likely to be unable to diversify their consumption away to other food sources easily. The urban poor may be particularly at risk from any impact resulting in an increase in food price. This may be exasperated by increases in urban poverty due increased migration from rural areas due to the declining natural resource base there. In contrast the non-poor are not only less likely to be affected by the loss of capture fisheries, if they are, they are likely to be better able to cope with this loss as they have greater access to assets and resources DISTRIBUTIONAL IMPACTS: ECOLOGICAL ZONE Impacts of mainstream hydropower development are also likely to be differentiated by ecological zone. This section looks at the probable distribution of those impacts by zone. An estimated 77% of the 795km of zone 2 running from Chiang Saen to Vientiane will be effectively be a reservoir if the proposed developments go ahead. This will imply direct economic losses of a similar proportion of the wetland and their services in this area. Mainstream dam development in this zone will mean over 32,000 people will be resettled and nearly 76,000 directly affected 29 (see social impact assessment). Land and river bank gardens inundated in this zone will be of particular relevance as other suitable agricultural land in the area for potential replacement purposes is limited. While population density here is relatively low, in some areas of the Lao PDR side of the river there are communities which are highly dependent upon river resources although for much of the population in these areas upland agriculture is the predominant livelihood source. Poverty rates in Oudomaxai are relatively high but those in Xayabuory are quite low. On the Thai side of the zone livelihoods are more diversified apart from amongst certain river dependant groups, while poverty rate is lower than in Lao PDR. The expected change in the ecological systems due to hydropower development will have greater impact on those river dependant populations in these zones, and higher poverty amongst some groups will result in greater vulnerability to these changes. Zone 3 running from Vientiane to Pakse is significantly more populous than zone 2. The mainstream dams in this stretch of the river are expected to create impoundments for 159Km, or 22% of its length. In this area at least 935 people will be resettled 30 and 2,570 people will be directly affected. In the Lao PDR area of this zone populations dependent on fish, wetlands, riverbank gardens and NTFPs, as well as rice production. On the Thai side of the river livelihoods are mixed but mainly dependent upon agriculture although fisheries is an important livelihood source in some areas such as the Songkhram basin. Relatively high poverty levels in Lao PDR mean that the population there is potentially vulnerable to the impact of mainstream dams. Lower poverty levels on the Thai side of the river mean that populations there are much less vulnerable to these impacts. Zone 4 running from Pakse to Kratie has the lowest average population density of all the zones. Mainstream dam construction in this area will mean the inundation of around 143Km or 45% of the stretch. The population which will need to be resettled in this zone is estimated to be at least 31 around 29 These are conservative estimates by developers based upon estimates of populations who loose land. assets or other livelihood resources excluding fisheries (see social theme paper). 30 At least, as no figures for Lat Sua hydropower project were available at the time of writing. 31 Figures for resettlement and affected population were not available for the Thankho project, these estimates only include Stung Treng and Sambor. 57 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

58 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S 30,000 and directly affected populations are expected to be around 28,000. Populations in both the Lao and Cambodian stretches of zone 4 depend on fisheries and other river resources as an important part of their livelihood strategies. In the Lao areas populations are likely to be vulnerable as poverty rates are around 30-40% and populations depend heavily on river resources, meaning they are likely to be adversely affected by changes due to mainstream dam construction. Stung Treng in Cambodia has a poverty rate of less than 20% but that in Kratie is around 30-40%. Again high poverty rates and dependency of the population on aquatic resources suggest that these populations will suffer disproportionately from the any adverse ecosystem impacts resulting from the development of the mainstream hydropower dams. Zone 5 running from Kratie to Phnom Penh and Tonle Sap has much higher populations living around the mainstream. This is a highly diverse area encompassing the large urban area of Phnom Phen and Tonle Sap. Fish consumption in this area is relatively high, with wealthier households tending to consume more fish. Whereas the poor rely on fish and other aquatic animals as a critical source of protein and nutrition. Around Tonle Sap in particular, rural communities are highly dependent upon aquatic resources. It is estimated that some communities living on floating villages are almost wholly dependent upon aquatic resources. The IBFM social assessment report (2007) suggests that overall dependence on fisheries and other aquatic resources is high in rural areas but relatively lower in urban areas. Poverty rates vary widely between 20% in urban areas to 50-60% in outlying rural areas. The combination of poverty and fisheries dependence in rural communities in the zone suggests that any change in fisheries productivity will have a significant impact on these populations. Estimates suggest that up to 1 million people in this zone would be seriously affected by any reduction in fisheries resulting from the construction of mainstream dams. Zone 6 running from Phnom Penh to the sea is the most densely populated stretch of the river consisting largely of the irrigated rice paddies of the Mekong delta. The area is highly dependent upon the Mekong river for irrigation and nutrient replenishment during floods. Fisheries and shrimp production are also important. While employment in agriculture and fisheries is low compared to other areas in the basin, many of the industries which employ the population are directly dependent upon the natural resources of the delta (e.g. food and beverage manufacture). Higher income levels in the delta and a lower level of dependency on the aquatic resources likely to be affected by the mainstream dams means that impacts of the dams are likely to be less acute in the delta. However, even with relatively low poverty rates high population densities mean that the absolute number of poor is higher than in DISTRIBUTION OF IMPACTS BY COUNTRY CAMBODIA: Key beneficial impacts of mainstream hydropower development in Cambodia are likely to include increased foreign exchange earnings from power exports, increased direct investment in the hydropower facilities themselves (some portion of which is likely to enter into the domestic economy), increased government revenues (generated from taxation, licensing and equity shares in the projects), improved power supply and reduced power price (particularly important in the Cambodian context where power costs represent a key constraint on industry) and the generation of the employment opportunities in the construction, operations and maintenance of the hydropower facilities. Key negative economic impacts that Cambodia is likely to experience if these projects are developed are a highly significant reduction in capture fisheries which will have a significant macro-economic impact (given the importance of the Mekong fisheries to the Cambodian economy) and perhaps more 58 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

59 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S significantly an adverse poverty reduction impact especially in vulnerable populations in zones 4 and 5 (see above). In the case of the projects sited in Cambodia direct impacts due to loss of land, assets and other livelihoods are likely to be important. It is important to bear in mind that even if the Cambodian projects did not go ahead fisheries in Cambodia would be adversely affected by projects in Lao PDR/Thailand albeit to a lesser extent (see fisheries paper). The impact on food security and economic costs associated with increased malnutrition amongst vulnerable populations are likely to be high. Better analysis of these indirect costs needs to be conducted to allow a proper assessment of the costs of these developments. In short, Cambodia is likely to bear the brunt of the decline in fisheries due to the importance of this sector and the dependence of large sections of the population on fisheries for their livelihoods and as a key source of nutrition. Domestic hydropower projects will bring benefits but is not clear whether the financial and economic gain these may imply will offset the largely hidden costs borne by fisheries dependant populations. More research is needed to establish the likely magnitude of these costs. LAO PDR: Beneficial impact of mainstream hydropower development for Lao PDR, given that it will play host all or in part to 10 of the 12 proposed projects, include - as with Cambodia - increased foreign exchange earnings from power exports, increased direct investment in the hydropower facilities themselves (some portion of which is likely to enter into the domestic economy), increased government revenues (generated from taxation, licensing and equity shares in the projects) and the generation of the employment opportunities in the construction, operations and maintenance of the hydropower facilities. Domestic power supply is expected to be less of a benefit to a country which has as yet a very small manufacturing sector. However, there is the possibility that these hydropower projects may facilitate the development of mining projects in some locations. Many important negative impacts in Lao PDER relate to the changing hydrology of the areas within which the dams maybe constructed. Loss of production land, housing, other productive facilities, infrastructure and amenities are all likely to be significant in both upstream and downstream areas. In particular loss of river bank gardens and negative impacts on in-stream infrastructure due to changing water levels and increased erosion will be costs that are likely to be borne by local populations and local governments respectively. Loss of aquatic resources is likely to be significant for populations along the river, this impact is likely to be less wide spread than in Cambodia. Nevertheless, loss of aquatic flora and fauna, and fisheries productivity is likely to have highly significant if localized poverty and nutritional implications similar to those outlined for Cambodia. Indirect impacts though exchange rate appreciation may have negative implications for some sectors such as manufacturing and agriculture although this will depend upon the macro-economic management policies and capacities of the government. A key question for Lao PDR emerging from the discussion of both macro-economic and poverty impacts are the extent to which the government will be able to use net revenues from hydropower to address the uncompensated impacts of these developments. And related to this, more broadly, the extent to which the government is able to manage these revenues in such a way that they improve productive capacity and competitiveness in sectors which are important for poverty reduction (i.e. manufacturing and agricultural sectors). THAILAND: Key benefits for Thailand while sizable in terms of these projects are not particularly significant in terms of the overall Thai economy. While there are undoubtedly economic benefits from a cheaper and more stable electricity price from these projects given the size of Thai power demand it has 59 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

60 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S quite a small impact. Thai project developers will also reap benefits as will their suppliers (mainly construction firms, engineering firms and their employees), increased profits for these companies will also lead to greater tax returns. However, when considered against the size and diversity of the Thai economy these impacts do not seem particularly significant. Negative impacts are much more localized. As with Lao PDR and Cambodia key economic costs will be borne by river dependant populations especially fisher folk and those engaged in varied subsistence livelihood strategies. The north east of Thailand is the poorest in the country, however compared to the other LMB countries the population in the Thai portion of the basin is comparatively well off. This population also has greater opportunities to diversify livelihoods way from dependency on the river resource base. Therefore, while the initial impact on the river basin population is likely to be significant, this population is likely to be able to adapt to this impact more effectively that affected communities in the other riparian countries. VIETNAM: Key benefits for Vietnam are likely to be similar to those for Thailand, although the benefits of the project developments are likely to be somewhat less as fewer project inputs are likely to come from Vietnam. On the other hand, the benefits of the additional power supply are likely to be more significant - reflecting supply shortfalls and the overall size of Vietnamese power demand. However, it is difficult to argue that these beneficial impacts are particularly significant in the context of a Vietnamese economy that would be likely to continue its break-neck growth with or without these projects. Hydropower development in general, however, is likely to be highly significant for the Mekong delta as changes in seasonal flow rates, sedimentation and river ecosystems have wide spread effects. While it has been noted that the mainstream dams may have some limited effects in lessening flood-peaks and trapping some sediment, these are not likely to be significant for the delta beside the impact of the Chinese mainstream and tributary projects. The impact on fisheries is likely to be the key negative impact experienced by river dependent communities in Vietnam. As elsewhere in the basin capture fisheries are an important livelihood component. The significant loss of fisheries in the basin will therefore have serious implications for fisheries dependent livelihoods and nutrition in the Mekong delta. As in other areas the poor are likely to be most severely effected by these fisheries impacts as unlike land or other privately owned assets these represent a commons accessible to the poor CONCLUSION The distributional analysis shows that different groups, zone and countries are likely to experience very different streams of costs and benefits resulting from the impacts of the proposed mainstream hydropower dams. Rural populations, river dependant populations and the poor stand to lose most. In fact impacts are concentrated on some of the most vulnerable populations. Whereas the main the benefits will accrue to developers and their investors, to a lesser extent power consumers and host governments. These beneficiaries notably exclude the rural poor who are often not connected electricity grids. The distributional implications of these developments should be more fully considered. It is also worth nothing that to some extent the economic cost-benefit case for these developments may be beside the point for the host countries. This is because most of these benefits accrue to the power consumers and perhaps the developers. While developers may need to compensate for direct costs, it is unclear where the burden of the indirect costs will fall (such as poverty impacts, the impacts of changing hydrology and geomorphology on in stream infrastructure etc). If it falls to the host governments to 60 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

61 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S address these indirect impacts using state budgets, from the perspective of the host country, the calculation is not of the economic benefit of the power which is capture by foreign consumers but the uncompensated costs of the projects considered against the increase in budget revenues generated by the project. 61 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

62 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S 5 HYDROLOGY AND SEDIMENT 5.1 STRATEGIC ISSUES Stream power and energy transport in the Mekong Water surface level changes Coarse sized sediment fate and transport Fine sized sediment fate and transport 5.2 DATA SOURCES AND POTENTIAL UNCERTAINTIES The flow and water level figures quoted in the SEA (across all themes) are based on the model simulations undertaken as part of the BDP, with limited data available from Zone 1 (Yunnan province). Recent interactions with the Chinese authorities (ESCIR) indicate that there are some discrepancies between the predicted seasonal changes caused by the Yunnan cascade as modeled by the BDP and ESCIR (Table 6.1). The regulated hydrological regime will depend heavily on the reservoir operating strategy at each of the 8 dams in the Yunnan cascade and of the cascade as a whole. Therefore, there may be some adjustment to the predicted values for flow and water level, though the substantive conclusions are expected to hold. The SEA team has adopted the BDP modelling results for two reasons: (i) the predicted reduction of wet season flows (700cumecs) represents ~90% of the total active storage of the Yunnan cascade and so is considered suitable for the purposes of the SEA, (ii) to stream-line the SEA with the BDP scenarios and thereby increase the utility of the SEA. By basing the hydrological assessment on the BDP scenarios it has allowed the SEA to situate itself into a likely development context expected by 2030, however, it should be noted that the implications of the mainstream dams may differ significantly from those outlined below should development in the basin follow another path. Table 6.1: Changes to seasonal flows based on the MRC and ESCIR modelling Seasonal flow MRC ESCIR 3 Increase flow in dry season (m /s) 690 1,200 3 Decrease flow in wet season (m /s) 700 1,400 During the SEA Impacts assessment phase, ESCIR and MRCS have been abale to compare model outputs for scenarios with 4 dams built on the Lancang River. This new collaboration has shown better agreement between the MRCS and ESCIR data. 62 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

63 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.1: Regulation impacts of the Upper Mekong mainstream hydropower at Chiang Saen: comparison of ESCIR and MRCS modelling results: Comparison of the impacts of 4 dams built on the Lancang River using MRCS and ESCIR models. Sediment records have been collected from the Mekong River since the 1960s using two methods: (i) Suspended Solid Concentrations (SSC) collected since the 1960s, and (ii) Total Suspended Solids (TSS) collected since the 1980s. However, between sample sites and over the past years the field and laboratory methods employed and the frequency of sampling has been erratic. This has resulted in a large level of uncertainty and divergence in estimates of the Mekong sediment load. The MRC IKMP has initiated a program to consolidate the sediment database and ensure ongoing sampling becomes more consistent, however, for the SEA there remains some divergence in expert opinion on: (i) the total Mekong sediment load, (ii) the source of sediments, (iii) the grain size distribution of the sediment profile, and (iv) the proportion of bed load in the sediment regime. The SEA has provided estimates on these important questions supported by appropriate references, and has also indicated the range of opinion for certain key estimates to indicate the scale of divergence. However, in order to understand the implications of mainstream hydropower on the fate and transport sediments, it is critical that some of these information gaps are filled. 5.3 SUMMARY OF PAST AND FUTURE TRENDS WITHOUT LMB MAINSTREAM HYDROPOWER STREAM POWER Stream power (measured in units of MW/km or kw/m 2 ) is the rate at which energy is lost in moving over the bed of the river, and lost to turbulent flow dissipation. It links key hydraulic features of the Mekong system, including: power production, energy dissipation, geomorphology, flow turbulence and sediment transport. The length wise power dissipation along the Mekong River in Zone 2 and 3 ranges between 5 to 50 MW per km distance, depending on flow rate m 3 /s and average channel gradient. The stream power per unit area is this value divided by the width of the streambed, so that the narrower the reach the larger is the unit area stream power. This large variation results from the pulsing nature of the Mekong River which experiences dramatic changes in flow between wet and dry seasons seasons (Kummu et al, 2009). 63 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

64 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.3 plots the unit length power dissipation for the Mekong River under natural flow conditions demonstrating that there is an increase in dissipation from Zone 2 to Zone 5 proportional to the increase in flow rate. The vast majority (in the order of 96%) of stream power is applied at the bed and banks of river channels. 32 Consequently, stream power is sufficiently large in localized river reaches that under baseline conditions a substantial part of the bed load is thrown into suspension for short distances. Figure 6.2: Natural and human induced energy dissipation on the Mekong River: (left) Water power is being rapidly dissipated as the mainstream river flows over a bedrock outcrop, upstream of Luang Prabang, (right) Energy dissipation immediately downstream of a major irrigation intake, downstream of Phnom Penh. The water surface drops about 0.5 m because of power dissipation as the flow enters the canal. Stream power is important to almost all aspects of the river, including movement of coarse and fine sized sediment, the development of deep pools in the bedrock, channel geomorphology, bank erosion, and formation of mid-channel islands. Stream power is proportional to the water surface slope and the inchannel flow and is largest during the peak flood season, at maximum hydrograph levels. In the bedrock confined reaches (primarily Zone 2), energy stored in the river facilitates the seasonal migration of coarse and medium sized sediments through deep pools and between in-channel features (sand bars, islands, and siltation benches). In the alluvial reaches (Zone 3, 4) there is a similar seasonal migration of sediments through the deep pools, coupled with a complex process of erosion and deposition along the river banks. The 20 year scenario trend is for the peak stream power to shift downwards by between 10-30%, associated with smoothing of the annual hydrograph because of regulation by proposed dams/reservoirs with large storage (figure 6.3). Table 6.2 outlines the predicted reduction in average maximum daily flows for the 5 zones of the LMB. The largest reduction in maxima will occur in the upper reaches of the LMB (10-20% in Zone 2) with the change reducing in significance further downstream (5-10% in Zone 3, 4, 5 and ~5% in Zone 6). Further, the majority of the reduction occurs under the BDP Definite Future scenario suggesting that the driver of the trend is regulation of mainstream flows in Zone 1 by the Yunnan cascade of 8 hydropower projects. Climate change, with anticipated effects on future flood runoff from intense heavy rain systems, may offset or counteract this decline, but the relative magnitudes of the superimposed changes on the hydrograph cannot be quantified in this study. The expected reduction in stream power will reduce the efficiency of sediment transport but it is not expected to stop prevent the process. 32 Annandale, G.W Reservoir Sedimentation, Developments in Water Science No. 29. Elsevier Science Publishers B.V, Amsterdam The Netherlands 64 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

65 Pak Beng Luang Prabang Xayaburi Pak Lay Xanakham Pak Chom Ban Koum Lat Sua Don Sahong Stung Treng Sambor Average Wet season Power dissipation MW/km M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.3: Average rate of energy dissipation of the Mekong River during the wet season along each reservoir (MW/km): (dark blue) under the Baseline scenario; (light blue) under the 20Y foreseeable future without LMB mainstream dams scenario. The drop in available energy from the baseline to the foreseeable future scenario corresponds to the reduction in wet season peak floods caused by regulation of the hydrograph due to hydropower in China and the Mekong tributaries. In Zone 2 the drop in stream power is in the order of 30%, dropping to 10% in Zone 3 and 4. Note: the sharp decrease at Don Sahong is due to the reduced flow rate at this site as only a fraction of Mekong River flow passes through the Hou Sahong ,300 2,500 2,700 2,900 3,100 3,300 3,500 3,700 3,900 4,100 Distance from source (km) Baseline 20Y without mainstream Table 6.2: Changes to the Averaged maximum daily flow on the Mekong mainstream (source: BDP, 2009) BDP Scenario Chiang Saen Vientiane Pakse Kratie Tan Chau Chau Doc Baseline (BS) 6,476 12,121 27,333 36,297 20,764 5,605 Definite Future (DF) 5,512 10,848 25,919 34,793 20,336 5,406 20Y without LMB mainstream (20Yw/o) 5,403 10,705 25,655 33,909 19,885 5,256 % change DF from BS 15% 11% 5% 4% 2% 4% % change 20Y w/o from BS 17% 12% 6% 7% 4% 6% WATER SURFACE LEVEL CHANGES The low storage capacity of the LMB mainstream projects (in comparison to the annual flow of the river) means that the LMB projects will have a relatively minor impact on annual and seasonal flow rates. The projects will have a significant impact on water surface levels at, hourly, daily, and seasonal time scales. Water surface levels in the Mekong mainstream fluctuate relatively slowly, because of the large size of the river. Rates of change of water surface elevation are highest with the arrival of the flood pulse, and are typically up to about +/- 0.16m/day at Luang Prabang, +/- 0.11m/day at Pakse and about 0.09m/day at Stung Treng. Extremities of water surface levels are experienced seasonally, and follow a pattern that is 65 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

66 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S determined by the reach of the river, and the maximum/minimum flow for the season. The largest recorded water surface levels are from years with the largest recorded flood events, for the mainstream river. The relationship between the water surface level and the flow magnitude is described by the rating curve, and each river reach has its own individual rating curve. Fig 6.4: Changes to the Averaged annual hydrograph for Chiang Saen (left) and Pakse (right). The effects of upstream regulation can be clearly observed in the reduction of the hydrograph peak and an increase in the dry season flow.this predicted change is larger for upper reaches of the LMB and comparison of the Definite Future and 20Y scenarios for Chiang Saen indicates that the majority of this change is induced by the development of the Yunnan cascade. Seasonal changes in water surface levels exhibit a similar variability as defined by the river rating curve 7,000 6,000 5,000 4,000 3,000 2,000 1, Jan 1-Feb 1-Mar 1-Apr 1-May 1-Jun 1-Jul 1-Aug 1-Sep 1-Oct 1-Nov 1-Dec 30,000 25,000 20,000 15,000 10,000 5,000 0 Q Baseline (m3/s) Q Definite Future (m3/s) 20Y (m3/s) Q Baseline (m3/s) Q Definite Future (m3/s) Fig 6.5: Water level fluctuations along the Mekong River: (left) Irrigation intake upstream of Luang Prabang. The maximum height of the structure is set to be higher than the highest floods, and the bottom of the tower is set on bedrock, at a location that does not accumulate sand. Seasonal water level fluctuations in Zone 2 are typically 5-8m; (right) Government water level gauge at Kratie. The slope gauge is set at an angle, built in to the side of the concrete steps. Seasonal water level fluctuations for the Mekong mainstream are 15 to 20 m at this location Riparian communities and users of the the Mekong River depend on the seasonal and daily fluctuation in water surface levels for fishing, agriculture and transport. River bank inhabitants, such as fisherman living in floating homes, are used to river levels that fluctuate slowly, and there is typically many days of time to anticipate the onset of floods. The following changes are expected to the Mekong hydrology under the 20Y without LMB mainstream scenario. 66 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

67 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S FLOOD EXTENT There are two components to the Mekong floodplain: (i) the vast expanse of floodplain downstream of Kratie including the Tonle Sap and the Mekong Delta systems, and (ii) the smaller floodplains alongside the Mekong channel in Zone 2, 3 and 4, typically located at tributary confluences and reaches of flatter topography. 83% of the Mekong flooded area is located in Cambodia and Vietnam, these areas are important for regional biodiversity, as well as national agriculture and fisheries production. The majority of Cambodian flooded areas reach depths of more than 3m during the flood season, while Vietnamese areas reach depths of m. Approximately 750,000ha lie along the Mekong channel in Lao and Thailand, of which 64 69% flood to depths great than 3.0m. The Definite Future will see a typical reduction of ~250,000ha in flooded area, the majority of which will affect areas with flood depths greater than 3m. This will affect more than 15% of the flooded areas in Thailand and Lao, and less than 5% of the area in Cambodia and Vietnam. The Definite future represents the dominant change to the flood regime, with only marginal additional effects under the 20Y scenario (see Fig 6.6). Fig 6.6: Example of change in flooded area for different BDP scenarios (source: BDP, 2010): The dominant influence of the Lancang cascade on changes to flooded area reflects that the majority of storage capacity available with proposed hydropower development is contained within these 8 Upper Mekong dams FLOOD TIMING AND DURATION The LMB experiences four identifiable seasons which occur with regularity: 1. The low flow season begins at the end of the calendar year and extends through to mid/late May. Flow is maintained by snow melt in the Mekong headwaters and the minimal natural storage of the basin in the soil horizon, groundwater and wetlands; 67 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

68 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S 2. the transition to flood season is characterized by the first spates of the Monsoon season with an increase in both flow rate and flow variability. The transition season ends in mid July; 3. the flood season begins when flows exceed the long term average and is characterized by the flood peak in September and sometimes a smaller leading peak. The season finishes in early to mid- November when flow drops below the long term annual average; and 4. the transition to dry season as flood waters and surface water storage in the basin drains back to the mainstream. Table 6.3: Changes to the timing and duration of the Mekong Hydro-biological seasons: the Mekong hydrological regime is characterised by four distinct seasons, changes induced predominantly by hydropower development will see a major disruption in the timing and duration of these seasons with a reduction in the duration of the flood season and the complete disruption of the transition season which is important in triggering a wide range of ecosystem functions all the Mekong mainstream and its floodplains. STATISTICAL PARAMETER STATI ON SCENARIO Q(ave) m3/s CS VTE PK KT 2*Q(min) m3/s END OF DRY END TRANS. A END OF FLOOD 2*Q(min ) Q>Q(ave) Q<Q(ave) TRA NS A DURATION FLO OD DRY + TRA NS B BS 2,621 1, May 24-Jul 11-Nov DF 2,615 3, Jul 28-Jul 13-Nov Y 2,609 3, Jul 29-Jun 9-Nov Y w/o Same as 20Y BS 4,300 2, May 25-Jun 14-Nov DF 4,282 3, Jun 27-Jun 13-Nov Y 4,368 4,310 5-Jul 6-Jul 13-Nov Y w/o 4,368 4, Jun 1-Jul 13-Nov BS 9,390 3, May 16-Jun 8-Nov DF 9,319 4, May 14-Jun 7-Nov Y 9,043 5,551 1-Jun 22-Jun 6-Nov Y w/o 9,043 5,286 1-Jun 20-Jun 6-Nov BS 12,869 4, May 17-Jun 12-Nov DF 12,798 5, May 17-Jun 11-Nov Y 12,423 6,899 1-Jun 25-Jun 10-Nov Y w/o 12,424 6, May 19-Jun 10-Nov The trends expected to 2015 are: Onset : changes to the timing of the seasons is greatest in the upper reaches of the LMB where the Lancang flows contribute a greater proportion. The timing of Transition A from the dry to the flood season will be most affected, starting ~7-8weeks earlier at Chiang Saen and ~1 week at Kratie. This season is important in triggering seasonal patterns in the river s ecology (for example fish migrations). Upstream of Vientiane will also see an earlier onset of the flood season, though tributary flows from the left bank will reduce this to negligible variation downstream of Vientiane. Duration: Upstream of Pakse will experience a 2-4week reduction in the transition season from Dry to Flood, which will drop to ~1 week in the Mekong floodplain. The duration of the flood 68 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

69 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S season is not expected to be significantly affected except at the uppermost reaches of the LMB where the UMB flows still dominate wet season volumes. These cumulative reductions will see an increase in the duration of the combined Transition + dry season. Magnitude: dry seasonal flows will increase by 70% at upstream stations decreasing to a 10% at the Mekong Delta. Conversely, wet season flows will decrease by up to 18% at upstream stations decreasing to 2% change at the Mekong Delta. Cumulative: The cumulative effects of the changes to the hydro-biological seasons is a reduction in duration of the two transitions seasons combined with a reduced variability between the wet and dry seasons. These two factors serve to homogenize the Mekong hydrograph and subdue the prevalence of the flood pulse. This is most pronounced in the upper reaches of the LMB reducing downstream as the left bank tributaries begin to influence the hydrograph. The trends to 2030 will see an extension of the above. The major impact from the combined effect of the Yunnan cascade and the tributary developments will be the loss of the transition seasons resulting from a more even hydrograph (table 6.3). The spates and first flushes of the transition to flood play an important part in triggering key ecosystem functions of the Mekong system including spawning and migration of aquatic biota as discussed in the aquatic and fisheries themes and will no longer occur under the 2030 foreseeable future scenario. TONLE SAP The natural cycle of water surface fluctuations in the Mekong mainstream drives the flow in and out of the Tonle Sap system contributing ~57% of the inflow to the lake, and determines the extremities of water surface levels in Tonle Sap Lake which experiences average seasonal water level (WL) changes of 8m. Typically the area of the lake fluctuates between 2,200km2 and 13,200km2 between the wet and dry seasons (table 6.4). Table 6.4: interannual variation in the hydro-physical dimensions of the Tonle Sap Lake (MRC, 2006) TIME SERIES Average Range max WL [m] min Difference max 13,218 9,637-5,278 SA [km2] min 2,237 2,061-2,402 Difference 10,981 7,576-12,876 max V [km3] min Difference The present trend (with 20 year scenario) is for a reduction in the hydrograph maxima, and an increase in the hydrograph minima, associated with water storage in large capacity reservoirs. This will reduce the hydraulic gradient driving flow in and out of the Tonle Sap system and consequently increase the dry season inundated area while also reducing the wet season inundated area of the lake. By 2015, upstream regulation will seasonally alter water levels in the Tonle Sap Lake by +0.1m/-0.3m, which will decrease the maximum extent of flooding in the Lake by 3-5%, and increase dry season lake area by 5-8%. The decrease in wet season lake area will reduce the area of fringing forest flooded and 69 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

70 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S reduce the littoral zone of the lake both of which are critical aspects of the lake s high productivity. The increase in dry season lake area will further exacerbate the reduction in the Tonle Sap floodplain by permanently inundating some areas. Together these effects will induce a km2 reduction in floodplain subject to the seasonal flood pulse and oscillation between terrestrial and aquatic environments which represents 5-10% reduction in the floodplain area. This will reduce the productivity of the Tonle Sap System, but further study is required to quantify the magnitude of this reduction. Under the 2030 scenario these changes to the Tonle Sap system will undergo even more extreme change with an expected 900km2 reduction in the Tonle Sap flood plain area (Figure 6.7) Fig 6.7: Changes to the average monthly area of the Tonle Sap lake under: (i) baseline, (ii) definite future, and (iii) 20Y BDP scenarios. HOURLY AND RAPID FLUCTUATIONS IN WATER LEVELS The river does not display signs of rapid daily or hourly rapid fluctuations in the LMB at present. Under the 20 year scenario these changes may or may not be present, and will depend on how the tributary power dams are operated, and the effectiveness of a proposed re-regulating structure downstream of the China dams FATE AND TRANSPORT OF COARSE SIZED SEDIMENT There are two dynamic mechanisms driving sediment transport, depending on the grain size of the sediment load (Bagnold, 1979). Upward and turbulent dispersive forces are capable of keeping smaller particles in suspension allowing downstream transport in the water column (see next section), while heavier sediments rely on saltation to move in successive jumps along the channel bed (Figure 6.8). Both transport mechanisms are functions of flow, channel slope and channel dimensions. Coarse sized sediment e.g. coarse sand, gravel and larger sizes, is conveyed in the Mekong River mainstream as bed load, and moves downstream annually by distances of a few tens of meters to a few 70 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

71 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S kilometers. Prior to the closure of Manwan dam (1992), there were no anthropogenic and only minor natural obstacles (e.g. deep pools) to the transport of bed load in the mainstream. These deep pools did not stop the transport of bed load, but slowed downstream migration to the order of one pool crossing every water year (Conlan, 2008). Figure 6.8: Action of water on particles near the stream bed, rolling and saltating grains from Dunne and Leopold Stream power is fed from the moving water to the bed, and supplies energy to move coarse particles in a serious of jumps and temporary suspension. The effectiveness of saltation as a transport mechanism can be enhanced by fines sediments in the water column which can aid in the temporary suspension of coarse materials. For the Mekong River, bed load is poorly understood; no measurements have been made of bed load transport, and the best information is from calculations that have been found to apply in other river systems, based on the suspended load transport. Fig 6.9: Coarse sand deposits in Zone 2: (left) Pakbeng: coarse sand, resupplied and moved downstream during peak floods and; (right) Nam Ou: coarse and medium sand, resupplied and moved to the mainstream during peak floods. During the low flow and transition seasons sediment within these size fractions play an important role provding habitats for aquatic and terrestrial life. Diminishing amounts are predicted, because of trapping by existing and proposed dams in China and the Meong tributaries. ZONE 1-2: Present trends are for a reduction in the transport of coarse sized sediment as the effects are felt of the supply blockage caused by the existing Chinese dams. Chinese authorities estimate that between , 490MCM of sediment was trapped by Manwan, which corresponds to ~50Mt/yr, while Dachshaosan trapped ~30Mt over an 18month period ( ) (Walling, 2009). Estimates of trapping efficiencies for the Yunnan cascade agree that the entire Yunnan cascade will have a trapping efficiency of 75-81% and will trap 70-75Mt/yr (Kummu et al, 2009; Sarkkula, 2010). With the future 33 Dunne T., and L.B. Leopold Sediment production and transport. Chapter 17 of Water in Environmental Planning. W.H. Freeman and Company. 818 pp. 71 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

72 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S completion of the Yunnan cascade, the amount of river bed in Zone 1 able to resupply coarse sediment will be reduced and the average annual sediment load arriving at Chiang Saen will reduce from 90Mt/yr to 20Mt/yr. The reduced sediment will first manifest as erosion problems near Chiang Saen and then work progressively downstream. This downstream migration of the erosion zone will be slowed by the presence of deep pools in Zone 2 which typically require 1 water year to cross, such that it may take in the order of decades before coarse sized sediment is no longer supplied to the alluvial reach starting 40km to the north of Vientiane. Fig 6.10: Future coarse & medium sediment dynamics without LMB mainstream (image adapted from: BDP, 2010): (top) the alluvial reaches of the Mekong River have reached a state of dynamic equilibrium such that there is a progressive movement of coarse and medium-sized sediments downstream from Zone 2 to Zone 5. Zone 3 and 4 primarily act as zones of transport with Zone 5 as the main site of deposition for coarse/medium sized sediments (i.e. no net accumulation of sediments in Zone 3); (middle) the majority of sediment supply from Mekong headwaters will be trapped behind hydropower development in China increasing erosion of channel deposits and of the river channel at the top of Zone 3. As medium sized sediments are mobilised coarse sized sediments will remain armouring the channel. Further downstream the river will re-instate a dynamic equilibrium between erosion and deposition and the new balance is likely to see reduced deposition in Zone 5 over the next 15-20years. There may be some at present unquantifiable - additional drop in the transport of medium-sized sediments due to the reduction in fine sediments which are known to assist in keeping sand-sizes in suspension; (bottom) As available sediments in the river are depleted over the next 50years the supply of medium sized sediments to Zone 5 will decrease to zero, with a corresponding drop in deposition. The effects of erosion will be felt throughout Zone 3 with changes to the location of the thalweg and an increase in bank instability. During all time phases, there will not be any supply of sand-sized sediments to the Mekong Delta as the stream transport power will not be able to maintain suspension of these fractions past Zone 5. ZONE 3 ZONE 4 ZONE 5 ZONE 6 72 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

73 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S ZONE 3-4: For the river downstream of Vientiane, reductions in the transport of medium/coarse sand and finer sizes are felt first, as this is rapidly depleted from storage on the bed and banks of the river predominantly in the long alluvial reaches stretching from 60km north of Vientiane down to Kratie. As these medium-sized particles move downstream, the sedimentary nature of the river bed coarsens in response. This effect will first manifest at the upstream sections of the alluvial reaches causing progressive downcutting of the channel and moving downstream as the medium and finer sized sand deposits are depleted. Back-of-the-envelope estimates suggest that there is ~14,500MCM (~11,000Mt) of sand-sized sediment stored in this alluvial reach (Carling, 2009), so that at the downstream end of the alluvial zone the river is likely to maintain a dynamic equilibrium of erosion and deposition for another 15-30years. After this period there will be a reduction in the supply of medium and fine sized sands and will begin to affect the important zone of deposition between Kratie and Phnom Penh. This effect is much larger than predicted impacts caused by climate change over the 20 year scenario. ZONE 5-6: As the Mekong enters the floodplain there is a sharp reduction in gradient and the advent of an extensive floodplain characterised; by multiple wide channels with elevated banks, and; expansive flooded areas shaped by overbank flooding of sediment rich waters. Flood waters deposit sand size sediments in this zone, such that by Phnom Penh no significant proportion of sands remains in entrained in the water column (Wolanski et al, 1996). The supply of sand-sized particles to Zone 5 will reduce in response to sediment trapping of upstream hydropower. DEEP POOLS There are at least 335 deep pools along the thalweg of the Mekong mainstream (Conlan, 2008). These pools are dynamic features of the river channel consolidating coarse and medium-sized sediment movement into a pulse-like wave which transports sediment between adjacent zones of sediment storage over the monsoon flood cycle. The deep pools play an important role in regulating the downstream progression of sediment and building in-channel features such as islands and sand bars, which are important for the river s biodiversity. No information is available on the past trends in either the location or number of deep pools. By 2030 both the peak flows and sediment load will be reduced, however the reduction in peak flow is not expected to reduce the in-pool velocity enough to prevent entrainment of sediments deposited in the pool middle altering the flushing efficiency of the pools. At Chiang Saen the duration of the flood season will be shortened by ~4weeks (25% of historic flood duration), this will affect the time over which scour velocities are reached resulting in only partial scouring of the deeper pools and increasing the time-step over which the sediment wave completely passes through the pool. The reduction in sediment load because of the reservoirs in the 20Y without LMB mainstream is not likely to induce any immediate changes in the functioning of the deep pool as their buffering effect on sediment movement has resulted in a significant volume of sediment stored within the channel banks between pools (at crossings ). Therefore, in the short term the pools will continue to function naturally (with a potential reduced flushing efficiency) as crossings continue to migrate downstream. In the long term, upstream crossings will disappear as they translate downstream and are not replaced, such that the deep 73 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

74 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S pools will eventually become static holes and no longer dynamic components in the fate and transport storage of sediment and bed load FATE AND TRANSPORT OF FINE SIZED SEDIMEN T Fine particles (sometimes known as wash load) move in river systems in proportion to their availability in the river basin and its tributaries from erosion processes, and these particles frequently constitute the largest part of the transported sediment load. In rivers that have no dams, there is no hydraulically imposed limit to the transport of these fine materials. They pass down the river system easily and rapidly, and may go all the way to the river mouth in a single flood season. If significant amounts of the wash load are removed from the system by anthropogenic changes, it is almost impossible for the river to resupply the wash load to meet its previous natural magnitude. Fig 6.11: fate and transport of fines in the Mekong River: (left) Fine particles moving in suspension in the Mekong mainstream (nearfield), contrasted with relative absence of fine particles in flow entering from Nam Ou (farfield), low flow season, January Flow from left to right; (right) flood flow season: Fine particles moving in suspension across the Cambodian floodplain during a high water event, September Flow away from bridge (upwards and to left in photo). Over bank flows like this cover on average 18,000km2 of Cambodian floodplain and the deposition of fine sediments bringsan average of almost 4,000tonnes of nutrients per year to these areas. The load of suspended (fine sized) particles has been measured at a number of stations on the Mekong mainstream since the 1960s using two different methods (TSS and SSC). At Kratie the average annual sediment load is estimated at 165million tonnes/yr with estimates ranging between million tonnes/year. The majority of the load arrives with the peak in seasonal run off and two weeks before the flood peak (figure 6.12). The pulsing nature of the Mekong hydrograph results in a very large inter-annual variability of sediment loads, for example between the annual load at Kratie fluctuated between Mt/yr (figure 6.12). Figures quoted in this SEA assessment refer to the average values (unless otherwise stated) and should be interpreted in the context of large inter-annual variability. There is considerable divergence in opinion surrounding the origins of the Mekong sediments (c.f. the SEA Hydrology and Sediment baseline working paper). Marine sedimentary profiles taken from off-shore core samples suggest that 52% (86Mt/yr) of sediments originate from the geologic formation known as the Central Highlands which encompasses the highland areas from the Sre Pok River (Zone 4) to the Nam Hinboun (Zone 3), while 43% (71Mt/yr) of the sediment load orginates from the Lancang catchment (Zone 1) (Clift et al, 2004). Estimates based on sediment monitoring data at the MRC suggest that the contribution from Zone 1 is higher with averages ranging from 35 73% (~60-120Mt/yr respectively) 74 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

75 Annual sediment load (Mt/yr) M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S depending on the time-series range and the type of monitoring data used (c.f. SEA Hydrology and Sediment baseline working paper). Estimates for the 3 S basins (Sesan, Srepok and Sekong) range from 5 40% (8 66Mt/yr) (Carling, 2009; Sarkkula et al, 2010). Fig 6.12: The Mekong River sediment load: (left) High annual variation in the annual sediment transport load at Kratie, fluctuating between MT/y (adapted from: Sarkkula et al, 2010); (right) Luang Prabang station: temporal distribution of sediment supply over the 1962 water year compared to the hydrograph. The peak in sediment loads is aproximiately synchronous with the peak in run-off of the upstream catchment and occurs on average 2 weeks before the peak flood flows in the Mekong Rive (source: Walling, 2009) Kratie Sediment load (MT/yr) The SEA estimates a sediment load of 90Mt/yr at Chiang Saen, 84Mt/yr at Vientiane with the addition of ~25Mt/yr from the 3S basins and 56Mt/yr from the remaining catchments between the Nam Hinboun and the Se Done, based on three important conclusions: 1. The geologic formation termed central highlands extends across all left bank tributaries from the Nam Hinboun to the Srepok, while the hydrological term is generally attributed to the Sesan, Srepok and Sekong cacthments. This may have contributed to over-estimations of the 3S load. The upper limit of the load based on sediment monitoring was found this to approximate the load expected based on the areal proportion of the central highlands formation within the 3S basin. 2. The estimates for the Lancang River based on marine sedimentary profiles may under-estimate the actual contribution because a significant proportion of the sand sized contribution from China drops out of suspension before reaching the marine environment, It is difficulty to explain conclusively the drop in sediment load between Chiang Saen and Vientiane. In part it is likely due to the uncertainty in the sediment monitoring data, epecially the complete lack of information on bed load dynamics which may see a large component of the sediment load migrate downstream within the benthic boundary layer. No detailed sediment monitoring data was available further downstream than Kratie, however best estimates suggest that 75-85% of the Kratie Load enters the Mekong Delta ( million tonnes/yr) with ~9Mt/yr entering the Tonle Sap system and ~25Mt/yr depositing on the Cambodian floodplains of Zone 5 (Kummu et al, 2008). 34 The contribution of sand sized particles is dependent on time scales, and the Mekong River mout has migrated several hundred kilometers to the south east in the past 6,000years (Clift et al, 2004) 75 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

76 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S The present trend is for significant reductions in the transport of fine material, because of the operation of reservoirs with large storage, in China, and on major tributaries. In the 20Y foreseeable future scenario, the sediment loads in Zone 2 will drop by ~80%, while further downstream at Kratie the load is expected to halve (table 6.5). This will reduce the quantities of finer sized sediments arriving at the major sites of deposition in zones 5 and 6. A significant proportion of the Mekong suspended load is removed naturally from the river upstream of the delta, in the Tonle Sap system, and in the Cambodian and Vietnamese floodplains. Maximum transport of fine particle load is in the reach downstream of Stung Treng, to about Kampong Cham. MEKONG FLOODPLAIN & DELTA SEDIMENT DYNAMICS The Mekong delta represents the last terrestrial habitats of the Mekong basin before reaching the coast and is comprised of a freshwater zone and saline zone. Due to the complex interaction of tidal currents, overland flows and river hydrology the seasonal boundary between these two zones shifts by 50-60km. The thalweg in the delta is ~10m deep in the freshwater region (30-140km from mouth) and then drops to ~5m deep in the saline zone (0-80km from the river mouth) (Wolanski et al, 1996; 1998). The freshwater region is predominantly made up of floodplains used for agricultural production and deposition is largely overbank siltation as flood waters overtop the river and canal banks. The saline zone comprises coastal areas used for agriculture, aquaculture or remant mangrove systems. Sediment influx into the Mekong delta is primarily fine silt and clay (figure 6.13) (Wolanski et al, 1996). A back-of-the-envelope calculation based on the year 2000 flood found that velocities between Chau Doc and Can Tho are sufficient to keep the majority of the expected sediment size fractions in suspension. This is confirmed by field work undertaken during the 1996 high-flow season in which the mean flow velocity was ~1m/s, such that there is expected to be little in-channel deposition in the freshwater zone (Wolanski et al, 1996). The 2030 trend without LMB mainstream hydropower will reduce the transport capacity during the flood season, but this is not expected to be significant. As the river distributaries approach the sea, sediment dynamics begin to be influenced by the tidal regime and the influence of saline water on coagulation of cohesive sediments (figure 6.14). During the 1996 flood season the D50 floc size was ~40micrometers, which represented a 10-fold increase in particle size when compared to the uncoagulated washload of fines sediments travelling near the surface of the freshwater zone (Wolanski et al, 1996). The formation of floccs is aided by saline water and consequently deposition is concentrated near the river mouth and in the immediate offshore environment. In the order of 95% of the sediments arriving at the delta are deposited in the immediate offshore environment (within 20km from the coast) (Wolanski et al, 1996). The 2030 trend without LMB mainstream dams is for the the sediment load arriving on the delta floodplain to reduce by 12Mt/yr, and the offshore load to reduce by 47Mt/yr (table 6.5). This compares with a total sediment load in river of similar size (Mississippi River 35 ) of about 400 million tonnes about a century ago (prior to development). The suspended load transport of the Mississippi at the delta during the last 2 decades has been 36% of this value, associated with past construction of dams and bank stabilization structures. 35 Meade, R.H., and J.A. Moody Causes for the decline of suspended sediment discharge in the Mississippi River system, Hydrology Processes, vol 24, pp I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

77 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.13: Indicative changes to the sediment load composition for Kratie & Tan Chan: (left) Kratie: under the past and future without the mainstream dams sediments up to coarse sized sand are expected to be transported downstream, which is confirmed by the presence of the large sand and gravel extraction industries downstream of Kratie; (right) Tan Chau/Chau Doc: few medium and larger sized sediment is expected to be transported through the Cambodian flood plain such that the majority of the sediment currently entering the Mekong Delta are clay s and fine-sized sand. Table 6.5: Indicative changes to the fate of sediment downstream of Kratie: the 20Y foreseeable future is predicted to halve the sediment load arriving at Kratie primarily due to trapping by the dams in Zone 1 and in the 3 S basins. SITE OF DEPOSITION AVERAGE ANNUAL DEPOSITION VOLUME BDP Baseline Sediment [Mt/yr] 20Y Without LMB mainstream dams Sediment [Mt/yr] Kratie: annual sediment transport rate Cambodian floodplain Tonle Sap flood plain 9 5 Mekong Delta floodplain Mekong river mouth 5 3 Ca Mau Pennisula <1 0 Offshore coastal shelf (<20km from the coast) I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

78 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.14: Seasonal sediment transport pathways for the Mekong River estuary (Source: Wolanski et al, 1996): (a) High Flow Season: During the high-flow season a salt wedge forms ~5-10km from the delta mouth. The toe of the salt wedge is the zone of greatest in-channel sediment deposition. Under current conditions in the order of 3-5% of the delta sediments (~5Mt/yr) are deposited at the river mouth; (b) Low flow season: there is no salt wedge in the dry season, instead partially mixed estuarine conditions extend ~50-60km up river. Sediments deposited in shallow coastal waters (<20km offshore) during high flow season are re-introduced to the channel during the low-flow season concentrating at the turbidity maximum zone. NUTRIENT TRANSPORT The decrease in the fine particle component of the suspended load has vital implications for the transport of nutrients (phosphorus, nitrogen and trace amounts of important minerals) onto the floodplain, and into the Tonle Sap system. Fine particles have much more surface area per unit volume than do coarse particles, and fine particles carry the majority of the nutrients that are needed by living systems. Natural fertilization from floodwaters that is important to about 18,000 km 2 of the Cambodian floodplain and 5,000-10,000km 2 of the Mekong Delta floodplains (see Fuji et al 36 ), will be impacted by reductions in the concentration of suspended fine particles. Based on work in the Tonle Sap and Cambodian floodplain, Sarkkula et al (2010) estimate that ~130mg of Total P is transported with every kilogram of sediment, this shows agreement with field measurements made in An Giang province, where a ~190mg of Total P were attached to each kilogram of sediments (Nga, 1998). Using these estimates, in the order of 4,000tonnes of Total P is deposited on the Cambodian floodplain each flood season, and 1,500 tonnes in the Tonle Sap Lake, another 4,000tonnes is deposited on the Mekong floodplains, with some 17,000tonnes being transported to the estuarine marine 36 Fujii H., H. Garsdal, P.R.B. Ward et al Hydrological roles of the Cambodian floodplain of the Mekong River. Intl Journal of River Basin Management Vol 1, No 3 pp I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

79 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S environments off the delta coast (table 6.9, direct impacts section). Without the LMB mainstream dams these yields will be halved. Figure 6.13: Sediment dynamics of the Mekong Delta: (left) Sediment enters the Mekong Delta through the Mekong and Bassac channel. There are some 10,000km of canals in the delta, ~5,000km of which are primary canals. Overbank deposition on the delta flood plain is likely to be concentrated close to these canals. According to estimates there is little nett accumulation of sediments in the Mekong channels during the high-flow season, with a dynamic equilibrium between erosion and deposition (Wolanski et al, 1996). During high-flow season, as the river approaches the sea saline water exacerbate coagulation and induce deposition of ~3-5% of the delta s load, the remaining sediment is flushed out to sea and driven south along the deltaic coast with the current in the South China Sea, depositing within 20km of the shore line. A fraction of the load is transported as far south as the Ca Mau pennisula and complex interaction of the two tidal regimes induces settling such that the pennisula is growing in the order of 100ha/year. Mekong sediment supply <1 Ocean currents: During the high-flow season ocean currents in the South China Sea orient south against the Delta coast, reversing direction during the low-flow season. A well understood relationship for lake and river systems exists 37,38 between nutrient transport, primary production, aquatic invertebrate success, and fish growth and condition, showing that reductions in nutrients will have impacts on fisheries. Additional impacts from the decrease in the transport of fine sediment will also arise at the mouth of the Mekong, with impacts on the offshore fishery which is partly dependent on the nutrients transported to the ocean environment by the Mekong sediment regime. 37 Goldman C.R. and A.J. Horne Food-chain Dynamics, Chapter 15 of Limnology. McGraw Hill, pp Mulholland, P. J. and D. Rosemond Periphyton response to longitudinal nutrient depletion in a woodland stream: evidence of upstream-downstream linkage. Journal North American Benthological Society 11: I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

80 Pak Beng Luang Prabang Xayaburi Pak Lay Xanakham Pak Chom Ban Koum Lat Sua Don Sahong Stung Treng Sambor Average Wet season Power dissipation MW/km M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S 5.4 EXPECTED DIRECT EFFECTS OF THE PROPOSED LMB MAINSTREAM PROJECTS STREAM POWER The operation of the proposed mainstream reservoirs will cause a major and irreversible shift in the way stream power is dissipated in the river. Without LMB mainstream hydropower, the stream power that is naturally dissipated in the river is in the order of about 5 to 50 MW per kilometer distance, will be reduced to about zero MW per km during low and medium flows in the reaches that include the reservoirs upstream of the dams. At high flows the power dissipation in the reservoirs will be above zero, but will not be distributed in the same way as the present and future trend situation (figure 6.14). The length of river affected constitutes about 66 % of the total 1760 km river distance, from the proposed Sambor dam site to the top end of the reservoir behind the proposed Pak Beng dam site. The river will be transformed from a live river, to a series of impoundments with slow water movement, held back by the cascade of dams (figure 6.15). The only substantial parts that will remain live river reaches are from: (i) below the proposed Pakchom dam site downstream to 80 km southeast of Savannakhet, (ii) below the proposed Ban Koum dam site to the upper end of the reservoir behind the proposed Stung Treng dam, and (ii) downstream of Sambor. Fig 6.14: Changes to stream power of Zone 2 and 3: the development of hydropower on the Mekong mainstream will concentrate energy dissipation at the dam sites as the projects generate electricity. This will result in a decrease in energy dissipated along the channel bed of the reservoirs and reaches sufficiently far downstream of the dam wall. During low and medium flows there is likely to be a complete reduction in available stream power, with a 50-80% reduction during the wet season ,300 2,500 2,700 2,900 3,100 3,300 3,500 3,700 3,900 4,100 Distance from source (km) Baseline 20Y without mainstream Remaining 20Y scenario stream power with LMB M'stream hydropower 80 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

81 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S At the proposed dams, the stream power will be released at the turbines, in the tailraces, and (during spillage events) at the dam spillways. These engineered water power releases will be localized and concentrated, and will have no resemblance to the present and future trend of the Mekong river water flow. With a typical reservoir length of about 100 km, the accumulated power will amount to 500 to 2,000 MW at each of the dams, and, with the present proposed sizing of the turbines, it will be possible to capture a substantial part of this stream power for conversion to electrical energy, if construction proceeds (table 6.6.). The production and sale of electricity is the major opportunity that arises from the proposed projects, and is the reason why there is interest on the part of developers in proceeding. Table 6.6: Percentage of wet season stream power used directly in electricty production 39 : During low and medium flows the proportion of stream power utilised in electricity production will approach 100%, leaving minimal energy for eco-morphological processes.during the high-flow season, ~25-80% of stream power will be utilised in electricity production. Note: based on the operating water levels and dam characteristics for Don Sahnong and Lat Sua (c.f. SEA Inception Report Vol 2), back-of-the-envelope estimates suggest that they are designed to harness more energy than is likely available for the respective river reach converted to reservoir. 40 Subject to detailed investigation, these projects would likely require substantial engineering of the river channel to make them perform to design. 41 LMB mainstream dam site Reservoir Length (km) Baseline scenario 20Y without scenario Pakbeng % 57% Luangprabang % 75% Xayabuly % 79% Paklay % 74% Sanakham 90 31% 47% Pakchom 85 48% 72% Ban Kum % 25% Latsua 10 >100% >100% Don sahong 5 >100% >100% Stung Treng 45 37% 40% Sambor 90 48% 53% The change in the way stream power is distributed will have links with other natural processes, such as: Major changes to sediment transport of all sizes (c.f. next section) Functioning of deep pools (c.f. next section) Major changes in transport of nutrients and organic and woody debris (c.f. next section) Significant and irreversible changes in fisheries migration and passage (c.f. fisheries theme) Additionally, stream power changes will have links with risk factors for anthropogenic uses of the river, such as detriments to navigation and detriments to fishing opportunities. 39 Figures are based on a turbine efficiency of e = 85% 40 The stream power numbers presented in the SEA are indicative and based on average estimates of water surface gradients across long distances and should be refined for the specific reservoir in question to better explore this issue. 41 The Don Sahong feasibility study indicates that there is insufficient year-round hydraulic head in the Hou Sahong channel to allow the dam to operate at the design level. In order to provide the design power outputs, approximately km of upstream channel will need to be excavated to a depth of 6m (c.f. Don Sahong feasibility study report, page 3-19 & fisheries theme). 81 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

82 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Fig 6.15: Longitudinal profile of the Mekong River (Zone 2) with LMB mainstream projects:. The projects will also reduce the water surface slope by a factor of 2.5 causing a reduction in stream power (source: CNR, 2009) 82 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

83 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S WATER SURFACE LEVEL CHANGES For 7 of the proposed projects, the dams will be sufficiently high that water levels in the reservoirs behind the dams will be above the highest ever recorded river elevations for significant distances upstream. For example at the Pak Beng dam, the reservoir behind the dam will be at about 340 m elevation, and this is higher than the highest recorded flood level for a distance of ~90 kilometers upstream of the dam. Areas that were previously floodplains will be drowned by the proposed reservoirs. More than 5-10% of the river valley, in the total distance of 1760 km will be affected by receiving inundation at levels never experienced in the history of data collection for the river. Fig 6.16: Elevation of the water surface level upstream of the LMB mainstream projects example at Pak Beng (image source: CNR, 2009): The LMB mainstream dams will elevate the water surface to levels 5-10m above the levels expected in a 1 in 1,000 year flood event for ~10-90km upstream of the dam wall. At Pak Beng the Q1,000 water level is ~330masl, while the reservoir level at the dam wall during the same event is ~340masl. 42 Downstream of the dams, the changes in water surface levels are dependent on the operating strategy employed by the project. It is feasible that dams could be operated to induce very minimal changes from natural conditions (e.g. continuous operation) or drastic and rapid changes (e.g. peaking operation). The trade-off between maximising power production and minimizing changes to downstream river reaches will determine the operational strategy and is an important consideration for the Avoidance, enhancement & mitigation phase of the SEA. This paper explores the full range of impacts which are realistically possible. IRRIGATION Pump stations along the river may need rebuilding/relocating, to allow for operation at these new water levels. There are approximately 700 existing and proposed pump stations of various sizes drawing from 42 Note: a 1 in 1,000 year event has an ~10% probability of occuring over a 100years. 83 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

84 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S the Mekong River (Figure 6.17). About half (309) of these pump stations will be affected by the mainstream reservoirs (Table 6.7). Latsua will affect the greatest number of pump stations (~158), while Pak Chom and and Ban Koum have the largest area of irrigated land which could potentially benefit (13,928ha, 9,820ha respectively). The total loss of bank side growing areas from permanent inundation from the reservoirs in this distance will be a certain risk. Partial losses of bank side growing areas will apply to approximately an additional several hundred km of the mainstream river, associated with high water levels (c.f. terrestrial theme paper). Additional bank side growing areas that are present on tributary rivers, affected by Mekong river backwater inundation, will be at certain risk. Table 6.7: Pump stations in the reservoirs and 100km downstream from the LMB projects UPSTREAM DOWNSTREAM No pump stations Total Area Irrigated No pump stations wihtin Total Area Irrigated (20Y Proposed dam in reservoir (20Y scenario) 100km downstream scenario) (ha) (ha) Pak Beng Luang Prabang Xayaburi Pak Lay Xanakham Pak Chom ,928 Ban Koum 16 5, ,590 Latsua ,115 Don Sahong Stung Treng Sambor TOTALS 45 6, ,747 Pump stations located in the reservoirs will need to be removed to higher locations to remain operable; this represents a significant cost to the pump operators. The elevated water levels in the reservoir will reduce the pumping head of reservoir pumps reducing the cost of pumping to irrigation schemes serviced by these pumps, though given the magnitude of change in water levels, it is likely that many of these pumps will need to be resized if the benefits of reduced pumping heads are not to be off-set by reduced pumping efficiencies from wrongly sized machinery. The magnitude of this change for reservoir pump systems will also depend on the drawdown in the mainstream projects. If they are used for peaking operation then a drawdown in the order of 1-5m could increase the complexity of pumping operations and require changes to pump locations on a 12-hourly time step, or more sophisticated telemetry controlled operations. For projects downstream of the reservoirs the new flow regime imposed by the mainstream projects will induce migration of the river thalweg and would require many of these pump stations to be relocated. Further, water levels downstream of the dams are likely to continue to fluctuate at a seasonal scale, but could also fluctuate on a daily time-step depending on project operating strategy. Figure 6.18 illustrates this situation for the Ban Koum and Luang Prabang projects. 84 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

85 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Fig 6.17: Location of the 712 existing and planned pump stations on the Mekong River in relation to the proposed LMB dams 85 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

86 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T E C O N O M I C S Y S T E M S Figure 6.18: Example irrigation pumphouses affected by LMB mainstream hydropower: (left) Ban Koum: 16 pumping projects are located in the reservoir which will require relocation as well as potential improvements to pumping heads; (right) Luang Prabang: 10 pumping projects are located downstream of the reservoir and will need rolocation and different operating strategies as the changed flow relocates the thalweg and induces variable flows potentially at a 12-hourly time step. 86 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

87 WL (m) M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T RESERVOIR OPERATIONS Significant issues of rapid fluctuations of river flows and levels will arise from the proposed projects, because the time needed to open and close turbines is small, order of 10 minutes or less. Opening and closing turbines may arise from a variety of planned and unplanned circumstances, and breakdowns of plant and electrical transmission systems. Figure 6.19: Rating curve for Luang Prabang town: Under continuous operations at Luang Prabang dam, the town of Luang Prabang can expect water levels of ~8m. Under peaking operations with a 12hour cycle, water levels could oscillate between m with potentially very short transistion times Peaking water level fluctuation ~ 4.5m at daily time step Discharge (m3/s) Peaking Operation Qpeak = 5,600 m3/s WL ~ 10.0 m Continuous Operation Qcont = 3,810 m3/s WL ~ 8.0 m Off-peak Operation Qoffpeak = 1,900m3/s WL ~ 5.5m Rating Curve: Luang Prabang town (BDP BS) For example, based on a peaking strategy of 12hours, fluctuations of 3-5m could be expected at key towns downstream of the projects. Table 6.6 shows this for the case of Luang Prabang, Ban Koum and Stung Treng. The time elapse for a rapid fluctuation from opening the turbines at the proposed Luang Prabang dam site, for example, will be about 1 to 1 ½ hours to the city of Luang Prabang, and this will provide very little warning time for bank-side residents to prepare for inundation by rapidly rising river levels. Tributary channels experiencing a backwater, such as the Nam Khan, will also experience rapidly rising water levels from turbine start-up. Equally at risk will be downstream effects of rapidly falling levels, that may arise during the day or night when the electrical demand is low. In the low flow season and shoulder season, the dam operator/owner may decide to do peak power generation, with repeated fluctuations in the flow released during each 24 hour period. The best case scenario is that peak power generation will not be licensed, and that projects will be required to operate at continuous power production, with a resulting minimization of river level rapid fluctuations. 87 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

88 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Rapid fluctuations in river level will adversely affect the stability of river banks, as they will experience repeated wetting and drying. Enhanced erosion and collapse of river banks may be expected around the periphery of the reservoirs, and at specific locations downstream of the dams (see sediment transport). Table 6.8: implications of peaking operation on water level fluctuations for three sample projects LUANG PRABANG DAM (upstream of Luang Prabang town) Peaking Operation Averag e Daily flow Daily Flow Volume WLs from rating curve BAN KOUM PROJECT (70km upstream Pakse) Average Daily flow Daily Flow Volume WLs from rating curve Average Daily flow STUNG PROJECT Daily Flow Volum e WLs from rating curve (m3/s) (m3/d) (m) (m3/s) (m3/d) (m) (m3/s) (m3/d) (m) rated turbine flow 3,800 11,700 18,493 Mean Annual Flow (MAF) at dam 3,171 Rated turbine flow/maf 1.20 Off peak peak flow 1,900 91,200 5, ,800 9,247 peak flow 5,700 17,550 27, ,832 Continuous Operation 3,800 91,200 11, ,800 18, ,832 Additional flows (Dam site - town) Nam Ou , Nam Soung + Nam Khan 181 4, Flow rate at town/gauging station Continuous 4, , , , , , off peak 2, , , , , , peaking 6, , , Mean annual flow MAF (BDP BS) 3,702 9, ,232 13, ,136 Rated turbine flow/maf Fluctuation (peaking - offpeak) 3, , , Analysis of the the projected mean annual flow at the dam site and the estimates provided of active storage for each LMB mainstream dams, provides a ranking of the potential of the 12 projects for peaking operation (table 6.9). The implications of peaking operation will be felt within ~ km downstream. Peaking operations at Sambor or Kratie will take ~18hours for the wave to travel to Tan Chau and the large fluctuation in water surface will likely smooth out over this distance, such that there is no expected change to water levels or saline intrusion in the delta resulting directly from the LMB mainstream projects. Further unprecedented fluctuations in flow/level may arise from break-downs, malfunctioning or operator error in controlling the spillway gates. With a failure of the primary and back-up facilities to raise the spillway gates during the onset of the flood season, there is a significant risk of dam failure, and the release of flows of unprecedented magnitude at one or many of the dams. The best case scenario is that catastrophic releases of this type will not occur. A concerted effort will be needed for the establishment/enforcement and execution of a program of comprehensive dam safety reviews. 88 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

89 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Additionally station operating rules for staff at each project will have to be carefully prepared, and subjected to peer review prior to the start of operation (see Mitigation). Table 6.9: Ranking of LMB mainstream projects based on peaking potential Hydropower project Storage Reservoir Dam MAF Volume Storage/MAF Area height km3/yr km3/yr ha m Xanakham % 7, Sambor *** % 71, Luang Prabang % 6, Stung Treng % 23, Xayaburi % 5, Pak Lay % 7, Pak Beng % 8, Ban Koum** % 13, Pak Chom* % 5, Latsua** % 1, Don Sahong % Thakho * MAF estimated by nearest stream gauge (* = Vientiane, ** = Pakse, ***=Kratie) FLOODING Table 6.9 indicates that the LMB mainstream projects have minimal capacity to store water between seasons. Changes in timing of the seasonal hydrograph of the river resulting from the projects will be in the order of days (1-5days) for individual projects, but may be significant. For example with the 10 projects in place and operating, the time delay of the arrival of the flood at Kratie will be an additional 2 weeks during an average precipitation year, and the duration of the flood season will be reduced by less than one week. Flood flows also control important estuarine processes such as saline intrusion. The LB mainstream dams are not expected to have a significant direct effect on saline intrusion. TONLE SAP LAKE The major change from the BDP Baseline for flooding in the Tonle Sap will occur as a result of the combined effect of tributary and Yunnan storage. There is likely to be no appreciable change directly from the LMB mainstream projects. In summary, changes in water levels from the mainstream projects are expected to adversely affect market gardening along the Mekong River and in major and minor tributaries near the mainstream, the habitability of floating homes that are downstream of the dams, fish habitat, and the viability of water intakes/pump stations adjacent to the river and tributary mouths. 89 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

90 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T FATE AND TRANSPORT OF COARSE SIZED SEDIMENT There are highly likely risks that coarse sized sediment, e.g. coarse sand, gravel and larger sizes will accumulate in parts of the river where it has never accumulated in the past, for example in the headwater reaches of the ten proposed reservoirs. The amount of accumulation will depend on the sequencing of the construction of the proposed cascade, and on the contribution from immediately upstream tributaries. This component, together with medium sized sand, constitutes the majority of the bed load transport of the Mekong River. Essentially, the dams will create a series of obstacles to the bed load transport, and will cause delays, possibly of years or decades, in the downstream transport. With the expected localisation in stream power dissipation for the majority of the 1760 km distance from Pakbeng backwater to Sambor, there will be much reduced water power available to suspend particles and keep them moving. The result of this will be enhanced sedimentation, with the formation of deltaic type deposits at the head of each of the reservoirs. Deep pools that are in these head-of-reservoir areas will be in filled. The severity of this problem is difficult to compute, as the deep pools are present in the Mekong mainstream because of complex three dimensional river currents, and the local topography of the bedrock on the sides and bottom of the river. Much enhanced sedimentation from coarse particles should also be expected where medium and high gradient tributaries enter the proposed reservoirs. The well established deep part of the river channel (the thalweg ) that in many locations defines the international boundary (e.g. Thai-Laos border) will be lost in many locations, due to sediment accumulation and erosion. This may or may not pose problems in the future for establishing/verifying the exact location of the international border. There are risks that sediment will erode from areas where is has never eroded in the past, for example immediately downstream of the proposed dams, and in bank side locations where the river thalweg is pushed sideways by a tributary delta entering the mainstream. Immediately below each of the dams it is likely that the flow will scour the river bed, probably down to bed-rock at the majority of sites, and create new locations for the thalweg, depending on the flow alignment as it leaves the turbines and the spillways. For the low gradient alluvial reach (Zone 3) from upstream of Vientiane to near Savannakhet, there will be a trend towards down cutting of the river bed, with destabilization of banks. This trend will be caused partly by the proposed changes/dam construction in the river and its tributaries under the 20 year scenario, and the trend will be much accentuated by the proposed mainstream projects. The progression of this down cutting is hard to compute as there are insufficient data on parameters such as particle size distribution that are needed to establish what is stored in the river bed. Mitigation of bank erosion problems may be undertaken with engineering works such as rip-rapped embankments, but the cost of these per km is very high, and requires the proximity of a nearby rock quarry. For the lower river, Zones 5 and 6, there is low availability of coarse sediment in the mainstream, except at Kratie where there are dredges at work in the river channel extracting gravel sized material. In general, coarse sized sediment is the first fraction size to deposit, therefore the most upstream dam in a cascade or the first dam built will induce proportionately large accumulation of coarse sized materials 90 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

91 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T than dams downstream in the cascade. If all the dams were built, this would put Pak Beng, Ban Koum, Don Sahong and Stung Treng as having the most impact of coarse sediment transport, though it depends heavily on the sequencing. Downstream of the proposed LMB mainstream projects, coarse-sized sediment is likely to be the last sediment size to be transported and the erosion of smaller sizes will result in the armouring of the downstream river reach FATE AND TRANSPORT OF FINE SIZE SEDIMENT Decreases in the concentration of suspended fine material (fine sand, silt and clay particles) will arise in the mainstream channel, because some of this component will settle and be stored permanently on the bottom of the proposed reservoirs. With the expected low velocities in the reservoirs for the majority of the time, it is expected that sediment sizes equal to and larger than medium sand (about 0.3 mm grain size) will be trapped and settle to the bottom and backwater areas of the reservoirs. A proportion of the materials that are finer than medium sand will be carried to the bottom and permanently stored in the depositing sand layers. This trapping of sediment will primarily be an impact during the first decade of operation of the proposed mainstream dams, as it is likely that siltation will reach a long-term equilibrium fairly quickly (one to two decades) because the reservoirs are relatively small. Prior to reaching equilibrium, there will be more fine sediment entering the reservoirs than leaving them, and this will diminish the load that is carried by the downstream river. The magnitude of the deposition is hard to compute, and will depend on the sequencing of construction of the mainstream dams, and fine sediment contributions from tributary rivers that enter the mainstream immediately upstream of the reservoirs. Fig 6.20: Modelled relationship between LMB mainstream dam trapping efficiencies and reduction in Mekong sediment load (Source: MRC, 2010): With no LMB mainstream projects the sediment load at Kratie is expected to reduce by ~46%. This reduction increases rapid with LMB mainstream projects. A 5% trapping efficiency in the LMB mainstream projects could increase the expected reduction to almost 70%, and a trapping efficiency of 40% could result in a 90% reduction in sediment supply to Kratie. 91 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

92 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Figure 6.20 presents the relationship between reduction in sediment load and the trapping efficiencies of mainstream hydropower, which can be seen to be approximately logarithmic. The greatest reduction in sediment load occurs within the first 10 TE percentage points, after which further reductions in sediment load begin to plateau such that there are only minor reductions in sediment load when trapping efficiencies increase greater than 40% (MRC, 2010). Consequently, even modest trapping efficiencies from the LMB mainstream project will contribute significantly to the reduced sediment load expected by Computations have been done for the Mekong dams on the basis of reservoir trapping efficiencies from other sites worldwide (figure 6.21) and these have shown that the LMB mainstream projects have individual trapping efficiencies of between <1 60% (Kummu et al, in publication). Given that the majority of further reduction in sediment load by the mainstream projects will occur within the first 40 TE percentage points, the mainstream reservoirs will have a significant impact on the future sediment load. Figure 6.21: Mekong Basin hydropower trapping efficiencies with confidence intervals (Kummu et al, in publication). LMB mainstream projects are shown as squares on the left-hand side of the diagram To indicate the order of magnitude change, the current sediment load of the Mekong at Kratie is approximately 165million tonnes per year, the tributary projects and the Yunnan cascade are expected to reduce this by ~46% (i.e. a remaining tranpoosrt load at Kratie of 88Mt/yr). Taking the lower limit and upper limit of the potential reduction from the mainstream projects (Fig 6.20; 6.21) suggests that the 2030 sediment load at Kratie with mainstream dams could be in the order of 20-40million tonnes/year. This predicted reduction may be overestimated, because of major sediment flushing activities that are planned for the dams which may re-mobilise sediments in the immediate vicinity of the turbine intakes. Better predictions cannot be presented at the moment, because of insufficient data concerning the particle sizes of material in suspension in the river. Therefore, a conservative estimate is for the sediment load at Kratie to reduce to 42Mt/yr (figure 6.22). 92 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

93 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Figure 6.22: Approximate average annual suspended sediment transport balance: Under baseline conditions an average of 165Mt/yr arrives at Kratie and is then deposited downstream throughout the Mekong floodplains, the channel and on the Mekong marine shelf. 90Mt/yr originates from upstream of Chiang Saen and ~25Mt/yr from the 3S regions (totals in white squares).by 2030 wihtout the LMB mainstream the load at Chiang Saen and Kratie will drop to 20Mt/yr and 88Mt/yr respectively (light blue squares). This represents a halving of the sediment load without LMB mainstream projects. With the LMB mainstream projects this will halve again to ~42Mt/yr. The Zone 2 projects will trap ~50% of the load arriving from China, the Zone 3 projects will trap Zone 1, 2 and important contributions from the left-bank tributaries in Lao PDR, while Zone 4 projects will also trap sediments arriving from the 3S basins. 43 Chiang Saen Nong Khai LEGEND Approx. average annual sediment transport 84 BDP Baseline at station (Mt/yr) 20 BDP 20Y without LMB mainstream hydropower (Mt/yr) BDP 20Y with LMB mainstream hydropower (Mt/yr) Supply of sediment from subcatchment 9 25 Kratie Deposition of sediment 25 Quantitiy of sediment supplied/deposited (Mt/yr) under natural conditions 26 5 Proposed LMB mainstream hydropower project < Note: this figure does not describe in detail the sediment contributions of Zone 2. In this zone it is known that the tributaries contribute in the order of 12-14Mt/yr (Wang et al, 2009), however most analysis of sediment monitoring data also observes a reduction in sediment load between Luang Prabang and Nong Khai (Walling, 2009; Wang et al, 2009, Kummu et al, in publication, Sarkkula et al, 2010). No clear explanation is available for this and it could reflect: (i) issues with data, (ii) significant proportion of sand transport moves as bed load and so is not picked up in the monitoring, or (iii) there is nett deposition of sediments as the river morphology transitions from bed rock dominated to alluvial. At present there is insufficient data to clarify. 93 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

94 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T MEKONG FLOODPLAIN & DELTA SEDIMENT DYNAMICS The trend with the LMB mainstreams is to continue the reduction in fine sized sediment transport to the flood plains and delta. Sediment supply with the LMB mainstream dams will reduce by 75% and reduce deposition of sediments at the major sites (see table 6.9 below). Sediments are an important input into the delta environment providing: (i) a rich and free source of nutrients to the floodplain, aquatic and marine environments contributing the emergence of the Mekong Delta as one of the most important agricultural/aquacultural areas for Vietnam, and (ii) a constant supply of bulk material which ensures the integrity of the delta, its channel networks and coastline. A shift in the sediment balance will threaten both nutrient transport and delta stability. The MRC BDP indicates that the effects of reduced sediment supply to the delta will not be felt for 50years, based on the conclusion that the resupply of sediment from channel deposits in Zone 3 will continue to maintain sedimentary processes in the delta (Carling, 2009). However, assessment by the SEA team of the year 2000 flood confirms the predictions that no sand sized particles are transported to the delta and so the compensatory load suspended in zone 3 will not reach zone 6. Therefore the implications of reduced sediment loads at the delta are likely to be felt within years for the freshwater zone. Annex VIII in Volume III of the SEA Impacts assessment provides an overview of the assessment made for the year 2000 flood conditions. Figure 6.23: Indicative profile of suspended sediment profile with mainstream dams: (left) Kratie: reductions in sediment load will reduce the total load of sediment transport. The implications on grain size may be partially off-set in the short term as increased erosion of the mainstream channel bed and banks compensates for reduced loading and re-supplies some medium sized particles. In the long-term this off-set will not remain significant; (right) Tan Chau/Chau Doc: reductions in sediment load will see a significant reduction in the load of sediment arriving at the Delta with serious consequences for channel and coastal stability as well as nutrient supply to agricultural and fisheries activities. This cannot be off-set by increased upstream entrainment of coarser materials as the floodplain system does not have the capacity to entrain these larger sized materials into the delta. For the saline zone, the mainstream LMB projects pose a significant risk to the integrity of the delta coastline and the stability of the distributaries and deltaic fan which characterizes the delta. A ~75% reduction in deposition from the current situation (due to reduced loading) and no significant change in erosion rates will see increased impacts of erosion with knock-on effects for those living along the river channels and coastlines as well as in the sectors that utilize these resources, including: 94 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

95 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Coastal fishery zones will experience reduced primary production with implications for the whole marine fisheries. Other industries in the fisheries sector which rely on waste products of the marine fishery for feed pellets and stock (e.g. aquaculture) are also likely to be affected. The Aquatic food chain is likely receiving the majority of the fine sediment, because the amount of overland flooding is a minor fraction of the whole discharge. So the impacts of reduced sediments/nutrients will be felt most strongly in the delta primary production and fisheries. At present the delta marine fishery produces an average of 458,000tonnes/yr. Irrigation works may benefit from reduced sediment build-up primary canals in the long term, however, this is likely to be overshadowed by greater instability in natural and man-made channels. Coastal erosion and delta-building will undergo a shift in the current balance as existing coastal sediment dynamics are altered. Less sediment will wash along the eastern coast of the delta exacerbating the effects of erosion in these provinces and also reducing the rate of deltaic growth currently experienced by the Ca Mau peninsula. In the long term, erosion will become an even more serious problem and without mitigation measures the size of the delta could reduce. River and inland water way transport may be effected as channel stability reduces In-channel islands, which are heavily populated and amongst the most fertile zones of the delta, are likely to experience greater erosion at the upstream end affecting communities and industries It is more difficult to attribute time scales to the saline and marine zones, because little is known about the quantity and mobility of the delta s marine sedimentary deposits. Wolanski (1996; 1998) concluded that these marine sediments can be mobile and during the dry season the reversal in direction of littoral drift pumps marine sediments back into the Mekong River channel. This could potentially provide some buffer to the reduced sediment loading. Detailed study is required to quantify the time scales over which the reduction in nutrient supply will begint o impact productivity NUTRIENT TRANSPORT In the order of 4,000tonnes of Total P is deposited on the Cambodian floodplain each flood season, and 1,500 tonnes in the Tonle Sap Lake, another 4,000tonnes is deposited on the Mekong floodplains, with some 17,000tonnes being transported to the estuarine marine environments off the delta coast (table 6.9). Without the LMB mainstream dams these yields will be halved and with the dams they will be halved again, such that the total nutrient loading in Zone 5 and 6 will be 0.25% of the original baseline loading if the LMB mainstream dams are built. Additional impacts from the decrease in the transport of fine sediment will also arise at the mouth of the Mekong, with impacts on the offshore fishery which is partly dependent on the nutrients transported to the ocean environment by the Mekong sediment regime. The Mekong marine fishery catch has an annual average of ~458,000 tonnes over the past 13 years. Little is known about the relationship between marine productivity and nutrient dynamics and requires a significant amount of field research to improve understanding of this important link. 95 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

96 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T H Y D R O L O G Y & S E D I M E N T Table 6.9: Approximate annual average estimates of Mekong sediment and nutrients deposition: Under baseline conditions ~20% of the sediment load at Kratie deposits on the Cambodian floodplain (including Tonle Sap); 16% in the Mekong Delta floodplain, 3% at the river mouth and ~60% is transported into the marine environment where it deposits predominantly within 20km from the coastline. The reduction in sediment load at Kratie will see a proportionate reduction in the volume of deposition at each site downstream ANNUAL DEPOSITION VOLUME With LMB mainstream (assume net Without LMB mainstream maximal trapping efficiency of LMB BDP Baseline dams cascade of 10%) TE)total) = 75% Sediment [Mt/yr] Nutrient (Total P) [t/yr] Sediment [Mt/yr] Nutrient (Total P) [t/yr] Sediment [Mt/yr] Nutrient (Total P) [t/yr] SITE OF DEPOSITION Kratie: annual sediment transport rate , , ,594 Cambodian floodplain 25 3, , Tonle Sap flood plain 9 1, Mekong Delta floodplain 26 4, , ,039 Mekong river mouth Ca Mau Peninsula <1 32 <<1 14 ~0 8 Offshore coastal shelf (<20km from the coast) , , , I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

97 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E 6 TERRESTRIAL ECOSYSTEMS & AGRICULTURE 6.1 STRATEGIC ISSUES 1. Habitat loss and degradation, forest cover and protected areas 2. Changes to patterns of agriculture 3. Value of paddy agriculture in each zone 4. Changes to agricultural and land use patterns along the mainstream, especially river bank gardens 6.2 SUMMARY OF PAST & FUTURE TRENDS WITHOUT LMB MAINSTREAM HYDROPOWER The terrestrial ecosystems surrounding the Mekong as represented by the landuse and forest cover, start from extensive forest cover in Zones 1 and 2, which decreases markedly as the river passes through zones 3, 4, 5 and 6; agricultural land use becomes progressively higher percentage, especially in NE Thailand, southern Laos, and below Kratie in Zone 5 and in the Delta. Government policies tend to be towards intensification of agriculture, with increased irrigation in Laos and Cambodia. In NE Thailand water for further irrigation is a limiting factor, and availability of suitable land is a limiting factor in the Vietnam Delta. The terrestrial ecosystems around the Mekong river are recognized as being globally important in terms of biodiversity and of 2600 km of the Lower Mekong, lengths of over 1000 km are considered as Key Biodiversity Areas, but only about 100 km of the river actually lies within a nationally protected area. River bank gardens have been recognized as an important contributor to the livelihoods of riparian communities, and although these have not been systematically studied, their contribution in each zone has been estimated of the order of million US per year. 6.3 EXPECTED DIRECT EFFECTS OF THE PROPOSED LMB MAINSTREAM HYDRO PROJECTS KBAS AND PROTECTED AREAS One of the indicators of the biodiversity impact of the proposed mainstream dams is the extent to which they and the reservoirs affect Key Biodiversity Areas and nationally Protected Areas. Table 1 summarises this indicator. Of the 1,100 km of Mekong River which is considered to be part of a Key Biodiversity Area, about 860 km (about 80%) will be directly affected by the presence of the reservoirs, usually by inundation of riparian vegetation and wetlands along the river. This will change the terrestrial habitat of the flora and fauna using these areas, especially congregatory water birds, for which many of these KBAs have been notified. About 79% of zone 2 s KBAs will be impacted, and 100% of zone 3 s and of zone 4 s KBAs will be impacted by the reservoirs. 97 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

98 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 1: Impact of proposed dams and reservoirs on Key Biodiversity Areas (KBA) and Protected Areas in each Zone Zone Length of river channel Length of river considered as KBA Length protected Length impacted reservoirs by Relevant dams KBAs and protected areas directly impacted km km km km > Pak Beng, Louangprabang Sanakham, Pak Chom Ban Koum Upper Lao Mekong, Mekong channel upstream of Vientiane, Mekong channel near Pakchom. Pha Taem NP, Phou Xiang Thong NPA, Houay Kok/Houay Phalang IBA (includes Stung Treng Ramsar site) 5a b Whole of Tonle Sap 14,812 km Various wetlands in floodplain 27,425 ha 0 Total 2,600 >1, Latsua (d'nstream) Don Sahong Thakho Stung Treng Sambor Mekong Channel (Phou Xiang Thong - Siphandone), Siphandone, Mekong river (Kratie to Lao border), Stung Treng Ramsar site, proposed Dolphin special management zone 98 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

99 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E As indicated in the SEA Baseline report many of these KBAs are Important Bird Areas identified by BirdLife International and the Ramsar sites associated near the Mekong in Thailand and the Stung Streng Ramsar site in Cambodia. They are valuable because they are representative of the different ecosystems of the Mekong river valley as it passes downstream. The KBAs in Zone 2 are characteristic of the narrow, rocky, upland river with birds that are dependent upon the river, its rocks and sandbars river lapwings, practincoles etc. A large proportion of these areas will be lost. The KBAs in Zone 3 are also representative of the rocky river valley upstream of the Mun/Chi confluence, a unique stretch of river as it flows between Thailand and Laos. This is also the only stretch of river which lies in national protected areas in the two countries. Only about 150 km of the whole river is formally protected in a National Protected Area, and this length will be directly affected by Ban Koum reservoir which pass between Pha Taem National Park on the Thai side and Xiang Thong NPA on the Lao side. Although these areas were designated to protect the sandstone escarpments and forests, the river channel is a part of each country s protected area system. In Zone 4, the KBAs include Siphandone, the Stung Treng Ramsar site and the stretch of river between, Kratie and Stung Treng. This is a unique ecosystem, not found anywhere else on the same scale and diversity in the Mekong system, or in South-east Asia, or globally. It is very rich in biodiversity, with great eco-tourism value and potential. The Stung Treng Ramsar Site extends from Stung Treng town nearly to the Laos border. Unlike other Ramsar sites in Cambodia, Stung Treng is not within a formal protected area. This is exactly the area that would be inundated by the proposed Stung Treng dam. As a signatory to the Ramsar Convention, Cambodia has a commitment to conserve globally important wetlands. With this active consideration of the proposed Stung Treng dam, the Royal Cambodian Government should notify the Ramsar Convention to put the Stung Treng Ramsar site on the Montreux Record of sites under threat. It is highly likely that the reservoir would destroy the rationale and criteria for the Ramsar designation. In the event of the dam being built there would be a review of its designation, and if the inundation adversely changes the wetland habitats, birds and fish species it may be de-designated and lose its status as a Ramsar Site. Proposals have also been put forward to establish a dolphin conservation zone upstream of Kratie, which would be affected by the Sambor dam. These KBAs correspond directly to the environmental hotspots identified in the BDP assessment. The BDP assessment identifies the number of hotspots that might be affected, This SEA identifies the lengths of river lying within the KBA or protected area that will be affected. There are also indirect impacts on KBAs and protected areas downstream. These are principally related to the changes in the hydrology of the Tonle Sap system. A comparison of the extent of the surface area of the Tonle Sap modeled under different BDP scenarios, (see hydrology theme paper) shows that under the 20-year scenario with the mainstream dams the area inundated at the end of the dry season in May may increase by about 10% (250 sq km cf 2,500 sq km). Without the mainstream dams, there would be an increase of about 8% (200 sq km cf 2,500 sq km). In the wet season, September, the surface area is predicted to increase dramatically by about 700 sq km compared to the current average of 14,500 sq km. Most of this (about 650 sq km) is due to the shift in the hydrology caused by the Chinese and tributary storage dams, but there is a proportion due to the mainstream dams. The implication of this is that the Ramsar sites and flooded forests within the Tonle Sap system will be under threat. Higher water levels in both wet and dry season means that the flooded forests are squeezed they cannot expand in the landward side because of existing land use for agriculture, and the water side forests may no longer be viable because they will be inundated for longer. 99 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

100 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E AREAS INUNDATED The mainstream dams generally have small reservoirs associated with them, with the existing river channel making up the largest proportion inundated. The areas of wet season and dry season channels inundated has been covered in the aquatic ecosystems. Table 2 shows the areas of existing river channel, forest land and agricultural land inundated. The 12 mainstream hydropower projects will have a total of over 151,000 ha of reservoir, of which 29% lie in Zone 2, 16% in zone 3 and 55% lie in Zone 4. 92,000 ha of existing river channel will be inundated (61% of the total reservoir area - 59% in Zone 2, 84% in Zone 3 and 55% in Zone 4). Outside of the rive r channel, 23,000 ha of forest land (mostly degraded forest) will be inundated and 13,500 ha of agricultural land. Of the agricultural land inundated, only a total of 829 ha of irrigated land will be inundated. Several of the mainstream dam projects are being designed to provide facilities for irrigation, some of which are quite significant. The key dams here are Pak Chom, Ban Koum and Latsua, whilst information on the irrigation potential for the Cambodian dams is not yet available. Thus Pak Chom has plans for 11 irrigation schemes around the area, covering a total of 2,706 ha of which 1 scheme (217 ha) is in Laos and 10 schemes (2,489 ha) are in Thailand. At one stage Pak Chom was identified as the source of the offtake for the Thai mega-project for transfer of water to NE Thailand, but consideration of this is not included in this SEA. Ban Koum, also shared with Thailand, has a plan for the irrigation of a total of 7,870 ha of which 8 schemes are in Laos and 14 in Thailand. Latsua has similar sized irrigation plan for developing irrigation of 7,300 ha for 3 crops per year in Laos which is currently under feasibility study. When the productivity of the areas inundated is considered and compared with the proposed irrigation schemes associated with at least 3 of the dams, it can be seen from Table 5 that, in Zone 2, the lost annual productivity due to the inundation by the reservoirs would have a value of 1.15 million USD, 1.17 Million USD in Zone 3 and 1.77 million USD in Zone 4. When the proposed irrigation schemes are put in place, the value of the increased productivity is 1.89 million USD per year in Zone 2 and million USD per year in Zone 3. No figures are available for irrigation potential from the Cambodian dams. It can be seen that even with just three dams being used for irrigation as well as hydropower, the gains more than outweigh the losses in each zone. The development of irrigation schemes also requires additional services to improve the provision of inputs and marketing for farmers. The beneficiaries of such schemes are usually those who happen to own land in the areas to be irrigated, and are often not the same people who may lose their agricultural land in other parts of the inundated area. There may be a disproportionate impact upon the poor in the reservoir areas, who both lose their access to land and cannot share in the benefits of improved agriculture nearby ACCESS ROADS AND TRANSMISSION LINES The footprint of the dams not only includes the areas of land inundated, but also the provision of access roads for the construction and operation of the dams and the transmission lines to link the mainstream dams with the regional electricity grids. Table 3 shows the current state of information about these, showing that the additional transmission lines would require over 1,542 km of access corridor. Assuming that this corridor is 70 m wide, the total land occupied would be an additional 10,794 ha 44. Without the detailed analysis of routes it is difficult to say what proportion of forest or agricultural land that these corridors would traverse, or passage through protected areas, however an estimate can be made by taking the proportions of forest and cultivated land in the 50 km corridor on each side of the Mekong mainstream. This shows that a total of at least 7,886 ha 44 This figure is based upon the information from 7 of the hydropower schemes. It is possible that for all twelve of the schemes, the area occupied by transmission lines would be at least 15,000 ha. 100 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

101 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E of forest land and 2,286 ha of cultivated land would be affected in the 50 km corridor on either side of the Mekong. Map 1 shows the 500 kva transmission lines planned for the GMS into which the transmission lines from the proposed dams will connect. The losses in paddy production 45 as a result of the transmission lines are shown in Table 5 and range from annual losses of 0.36 million USD for Zone 2, 0.89 million USD for Zone 3 and 0.06 million USD for Zone 4. Note that these are low figures because data on lengths of transmission lines are not available yet. In terms of access roads that will have to be provided for construction of the dams, there is insufficient data at this stage to make a meaningful comment, although the distances are not anticipated to be very large for each dam. Road upgrades on the Laos side have already improved access for a number of dam sites, namely, Pak Chom and Sanakham, Ban Koum and Latsua, so additional access roads are likely to be purely local for these. In addition a number of dams will construct temporary bridges across the Mekong mainstream and some of the tributaries, e.g. Pak Beng and Louangbrabang (across the Nam Ou) and once constructed, the dam itself may provide crossing points across the Mekong INDIRECT LANDUSE IMPACTS The development of the hydropower schemes, with associated inundation, the transmission lines and access roads, will inevitably put pressure on adjacent landuses as resettled people move to new areas, and forest land is opened up for replacement agricultural land. These indirect pressures upon land use will develop over time as the different schemes are put into place. They are impossible to predict at this stage, but should be recognized as potentially significant indirect impacts. Without information on routes of transmission lines, it is difficult to say whether protected areas will be impacted, and the extent to which biodiversity corridors will be further fragmented. These are issues which should be addressed in the specific ESIAs of each hydropower scheme. 45 Losses due to the transmission lines assume that all the land under the transmission line will be paddy land. This will not be the case, but the estimate serves as a first order of magnitude) 101 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

102 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 6: Proposed transmission lines for the GMS with connections to proposed dams (Source: International Rivers, based on ADB, map to be revised) 102 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

103 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 2: Areas of land inundated by each reservoir associated with the proposed Mekong mainstream dams and associated irrigation schemes Zone no. Ecological Zone Dams Length of River Area of reservoir River channel inundated Land inundated Irrigation schemes proposed 1 China to Chiang Saen Chiang Saen to Vientiane Vientiane to Pakse Pakse to Kratie Chinese dams section reservoir Forest Agriculture Irrigated Laos Thailand Cambodia km km ha ha ha ha ha ha ha ha 2,120 Pak Beng ,700 7, , Louangprabang ,239 2,864 4, Xayaburi ,900 4, Pak Lay ,000 2,310 2, Sanakham ,000 2,000 not available 6,000 not available 0 Pak Chom ,354 6, not available 217 2,489 1 scheme 10 schemes TOTAL Zone 2 dams ,193 25,644 7,310 8, Ban Koum ,800 13,600 1, na 7,870 8 schemes 14 schemes Latsua ,700 7, , ,300 TOTAL Zone 3 dams ,500 20,600 2,274 1, Don Sahong Thakho N/A under consideration Stung Treng ,100 not available not available not available not available not available Sambor ,000 45,480 13,143 3,369 not available not available TOTAL Zone 4 dams ,390 45,573 13,332 3, Kratie to Phnom Penh and Tonle Sap Phnom Penh to the sea Lower Mekong Total 2, ,083 91,817 22,916 13, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

104 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 3: Land requirements for transmission lines and access roads for the mainstream dams (as far as is currently available) Zone no. Ecological Zone Dams Transmission line required Length Destination Total Area corridor Area of forest land Area cultivated land km ha ha ha Access roads km 1 China to Chiang Saen 2 Chiang Saen to Vientiane 3 Vientiane to Pakse 4 Pakse to Kratie 5 Kratie to Phnom Penh and Tonle Sap 6 Phnom Penh to the sea Chinese dams Pak Beng na Thailand na na na 15.6 Louangprabang 400 Vietnam 2,800 2, Xayaburi 220 Thailand 1,540 1, Pak Lay na Thailand na na na na Sanakham na Thailand na na na na Pak Chom 185 Thailand 1,295 1, na TOTAL Zone 2 dams 805 5,635 4, Ban Koum 434 Thailand 3,038 1,551 1,270 na Latsua na Thailand na na na na TOTAL Zone 3 dams 434 3,038 1,551 1,270 Don Sahong 21 Thailand Thakho 22 Laos Stung Treng na Vietnam na na na na Sambor 260 Vietnam 1,820 1, na TOTAL Zone 4 dams 303 2,121 1, Lower Mekong Total 1,542 10,794 7,886 2, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

105 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 4: Value of rice production lost from agricultural land inundated and gained from associated irrigation schemes and transmission lines Zone Unit China to Chiang Saen Chiang Saen to Vientiane Vientiane to Pakse Pakse to Kratie Kratie to Phnom Penh and Tonle Sap Phnom Penh to the Sea Paddy field area in 50 km corridor of river sq.km , , , , , Yield t/ha/yr Annual production t/yr 50, ,019 8,020, ,666 3,616,666 9,905, $US/kg US$ million , , With mainstream dams Inundated agricultural land ha , , , loss of production t/yr , , , Lost 0.2 $US/kg US $ million With associated irrigation Area of irrigation scheme developed ha ,696 15,170 na Yield t/ha/yr na Production t/yr ,436 68,265 na Value of annual production US $ Million na Losses due to transmission lines Areas of cultivated land lost in 50 km corridor ha , Yield t/ha/yr Lost production t/yr ,789 4, Value of lost annual production US $ Million (note that transmission lines lengths ar not available for all dams, so this is an underestimate of the losses) 105 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

106 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E IMPACTS UPON LAND USE - FOREST COVER AND AGRICULTURE A GIS analysis of the impacts of land use (carried out for this SEA) has considered the areas inundated by all of the dams and expressed in comparison to the total areas of different landuses within a 50 km corridor on each side of the Mekong. Maps 1 7 show the inundation areas nearest to some of the mainstream dam sites. It is clear from these maps that the inundation is largely within the river channel in most cases, except for the lowest two dams. Even near the dam sites, relatively little land is inundated. Table 6 shows this comparison for all of the dams proposed along the Lower Mekong. From this table it is clear that of the 1,371 sq km inundated the largest % losses due to the dams are wetlands, of which a total of 735 sq km will be modified into reservoirs. This makes up 54% of the area inundated and mostly consists of natural water bodies, i.e. the river channel. This loss amounts to about 7% of the wetlands within the 50 km corridor. The next largest land use type to be inundated is forest mostly evergreen and deciduous, with about 544 sq km inundated (about 40%). This amounts to about 0.5% of the forest land cover in the 50 km corridor. Agricultural land inundated is relatively small, making up about 64 sq km (under 5%) or 0.08% of the agricultural land in the 50 km corridor. The indication from these overall figures is that the different land uses inundated are relatively small in comparison to the wider area. Zone 2: When the different zones are considered, the cascade of dams above Vientiane, makes a significant impact upon the wetlands in Zone 2. Table 7 shows that 52% of all the wetlands in Zone 2 will be modified into reservoirs. This is a dramatic change also described in the aquatic impacts paper. The forest cover in zone 2 amounts to about 91 sq km, which is about 0.2% of the forest cover in the 50 km corridor on each side of the Mekong. The forest within the Mekong valley is largely secondary and often significantly degraded, with bamboo and other pioneer species moving into areas where the forest has been cleared. Some riparian vegetation remains, with mature trees in places, and this vegetation type would be the main casualty of forest losses due to inundation. It is unlikely that rare and endangered plant species would be lost. Agricultural land inundated amounts to nearly 10 sq km or 0.12% of the total agricultural land in the zone 2 50 km corridor. The dams in zone 3 (shown in table 8), will inundate 112 sq km. of wetlands amounting to about 6% of the total wetland area in the zone. About 0.1% of the forest land will be inundated and 0.02% of the agricultural land in the zone 3. Most of the forest in this zone is deciduous dipterocarp, which is well represented in the wider area. In Zone 4, the dams will inundate 335 sq km, or about 17% of the total wetland area in the 50km corridor of the zone. This is also a significant proportion implying a great change in the landscape character of the zone. This is made up about 35% of the natural water bodies, i.e. the river channels, and significantly about 3% of the flooded forests in the Zone 4. The flooded forests are made up of unique species that can withstand periodic inundation, and consist of rheophytic shrubs such as Homonia sp, Telactadium, and Rotula spp. and trees such as Anogeissus and Acacia spp. These would be lost in the areas where the proposed reservoirs would permanently inundate About 425 sq km of other forest will be inundated, making up about 1.63% of the forest in the corridor. It is not possible to distinguish beyond the major categories of deciduous and evergreen forest and bamboo, but it should be noted that the first type of forest to be lost will be the characteristic types that grow along the river banks, and these will not be replaced along the reservoir banks. River bank vegetation is clearly different from the trees further up the hillsides. 106 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

107 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Zones 5 and 6 will not be affected by the inundation caused by the dams. However, in these areas downstream of the dams, changes in surface water levels at different seasons may have effects upon the landuse patterns, forest cover and cropping, which are difficult to define or quantify. Table 5: Comparison of landuse in 50 km corridor along each bank of Lower Mekong with areas inundated by mainstream dams Land Use type Total all zones sq.km Total all dams sq.km % land use inundated Paddy field 62, Orchard Field crop 9, Swidden cultivation 2, Total cultivated land 75, % cultivated land 4.63 Industrial plantation 2, Forest plantation Bamboo forest 4, Coniferous forest Deciduous forest 51, Evergreen forest 58, Total forest cover 116, % forest Barren land 1, Grassland 6, Shrubland 11, Total open land 20, % open land 2.07 Flooded forest 1, Mangrove forest Marsh and Swamp Natural Water body 7, Aquaculture Total wetlands 11, % wetlands 53.5 Built-up area 5, % Built-up 0.13 TOTAL 228, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

108 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 7: Riparian vegetation and river bank gardens in the area affected by Louangprabang reservoir Figure 8: Degraded forest along the banks near Pak Lay 108 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

109 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 9: Mature riparian trees along the Mekong upstream of Sanakham dam site Figure 10: Riparian vegetation around the Ban Koum dam site 109 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

110 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 11: The character of the landscape around Ban Koum will be changed by the reservoir (view from Pha Taem National Park, Thailand) 110 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

111 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 6: Landuse types inundated by the Mekong mainstream dams in zone 2 Land Use type China to Chiang Saen Chiang Saen to Vientiane Zone 2 dams Zone 1 Zone 2 Pak Beng Louangprab ang Xayaburi Pak Lay Sanakham Pak Chom Total Zone 2 % of zone sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km Paddy field 500 3, Orchard Field crop 123 3, Swidden cultivation 177 1, Total cultivated land 815 8, % cultivated land Industrial plantation Forest plantation Bamboo forest 18 1, Coniferous forest Deciduous forest 4,993 16, Evergreen forest 3,117 24, Total forest cover 8,128 43, % forest Barren land Grassland Shrubland Total open land % open land Flooded forest Mangrove forest Marsh and Swamp Natural Water body Aquaculture Total wetlands % wetlands Built-up area % Built-up TOTAL 9,142 53, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

112 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 7: Landuse types inundated by the Mekong mainstream dams in zones 3 and 4 Land Use type Vientiane to Pakse Zone 3 dams Pakse to Kratie Zone 4 dams Kratie to Phnom Penh and Tonle Sap Phnom Penh to the Sea Zone 3 BanKoum Latsua Total Zone 3 % of zone Zone 4 Don Stung Sahong Treng Sambor Total Zone 4 % of zone Zone 5 Zone 6 sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km sq.km Paddy field 22, , ,910 19,810 Orchard Field crop 2, ,318 1,518 Swidden cultivation , Total cultivated land 25, , ,497 21,458 % cultivated land Industrial plantation 1, Forest plantation Bamboo forest Coniferous forest Deciduous forest 13, Evergreen forest 15, Total forest cover 31, % forest Barren land 1, Grassland ,067 1,566 Shrubland , Total open land 2, ,217 2,527 % open land Flooded forest Mangrove forest Marsh and Swamp Natural Water body 1, ,382 1,766 Aquaculture Total wetlands 1, ,714 3,069 % wetlands Built-up area ,654 % Built-up TOTAL 61, ,423 31, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

113 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Table 8: Estimates of losses of values of river bank gardens due to reservoirs in each zone No Ecological Zone China to Chiang Saen Chiang Saen to Vientiane Vientiane to Pakse Length of zone Length of zone affected by reservoirs Pakse to Kratie Kratie to Phnom Penh and Tonle Sap Phnom Penh to the sea River dependent rural population (2005) NA River dependent HH within 15 km River dependent HH affected by reservoirs No change % of HH using RBGs Number of HH per zone with RBGs affected by reservoirs Average area of RBG per HH Total area of RBG lost due to reservoirs Present Yield of vegetables produced Yield of vegetable produced after dams constructed Yield of Vegetables lost Total value lost per ha ha Tonnes US $ million 313,939 62,788 54, , ,891 12,997 1,651 11, ,343, ,636 59, , ,872 50,369 39,137 11, ,397 46,479 20, , ,669 4,346 3, ,581, ,390 No change 7 49, ,358 74,146 74, ,482,368 1,296,474 No change , , , , NB. These estimates do not include the losses due to difficulties of cultivating river bank gardens downstream of dams 113 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

114 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E RIVER BANK GARDENS When the dams are constructed and the reservoirs come up their operating levels, there will no longer be significant seasonal variation in the water levels in the reservoirs. There may be daily variation of one to two metres depending upon operating mode of the dam. This effectively means that no river bank gardening, which depend upon the seasonal recession of the water after the high flows, will be possible within the reservoir areas. In addition to this the daily flow variations below each of the dams may mean that river bank gardening downstream of the dams may be difficult and risky for at least 25 km downstream of each dam, if not further. Impacts of downstream river bank gardens have not been estimated. Using the baseline estimates developed for the values of river bank gardens in each zone, Table 9 shows the calculations in losses in production from river bank gardening as a result of the reservoirs. This comes to a total of USD million per year within zones 2, 3 and 4. This is most significant in Zone 2 and 3, with losses of about USD 9 million per year each. In Zone 2, where about 80% of the length of the river is converted to reservoir, so that the large majority of river bank gardening households will lose this livelihood source, and in Zone 3, where there is a higher density of households using river bank gardens, although the length of river bank affected is smaller. In Zone 5, immediately downstream of Sambor dam, river bank gardens may be affected by the daily fluctuating water levels, and the overall reduction in seasonal differences in flow. In Zone 6, river bank gardens are less important than floodplain agriculture, and the canal banks may be planted with trees rather than vegetables. However, cultivation on the in-stream islands especially fruit gardens, may be affected by changes in water levels, and availability of irrigation water. It is possible that alternative measures may be developed for reservoir bank cultivation, but it is unlikely that this will be as productive as normal river bank gardens, because they will no longer receive the nutrients brought down in the flood sediments. Figure 12: River bank gardens at i) Savannakhet, ii) Siphandone and iii) at Latsua dam site 114 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

115 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E OTHER IMPACTS UPON AGRICULTURAL LAND Apart from the areas of irrigated land inundated or created as a part of these schemes, it is possible that there may be a number of impacts upon agricultural land along the Mekong. Many of these are dealt with 115 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

116 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E in other sections, especially the hydrology and social sections and so are only highlighted here. These include: Increased efficiency of pumped irrigation as a result of generally higher water levels around the reservoirs. This is dealt with in the hydrology impacts paper, and may mean reduced pumping costs in certain locations. Variable water levels downstream of dams depending upon discharges, may mean more complicated pumping for irrigation that has to take into account variable levels Elevated water levels in reservoirs may mean an increase in the ground water levels around and downstream of dams. This may be beneficial for water sources based upon pumping from wells, since pumping efficiencies would be increased and costs reduced. However, increased ground water levels around dams, can also lead to mobilization of contaminants in the soil, for example arsenic, resulting in contamination of ground water sources. High arsenic levels have been recorded in Cambodia, especially around the Sambor dam and reservoir. In the Delta, it is likely that under the Definite Future with the Chinese and tributary dams and under the 20 yr scenario with the mainstream dams, there will be a decrease in the flooded area, in the duration of flood and the depth of flooding, compared to the baseline. There is a marginal decrease in these changes when the scenario with the mainstream dams is compared to without the mainstream dams. There is thus likely to be little difference in wet season cultivation in the flood plain. The increased dry season flows predicted for both the definite future and with the mainstream dams may have beneficial impact upon saline intrusion and acid sulphate release from soils. There is a marginal increase in the projected longer dry season with higher flows as a result of the mainstream dams. The BDP impact assessment predicts that the 20 year future without the mainstream dams will tend to decrease the areas affected by saline intrusion in the Delta by about 17%. The presence of the mainstream dams will not make a significant difference to this prediction. However, the Delta has a highly managed water system and the realization of these benefits in managing saline intrusion and acid sulphate soils may depend upon the effectiveness of this management. The fertility of the floodplains depends to a large extent on the nutrients associated with the fine sediments carried down by the river, and deposited in the flood plain. It is probable that there will be a significant cumulative impact upon sediment trapped by the dams, including the fine sediments, so that there could be a loss in overall fertility of the floodplains. However, as the BDP assessment points out, this loss may be to some extent be offset by the increase in nutrients in agricultural return waters and urban waste waters resulting from higher fertilizer use and urban growth. 116 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

117 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 13: Proposed irrigation schemes associated with the proposed Pak Chom project 117 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

118 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E 6.4 SENSITIVITY ANALYSIS OF DAM GROUPINGS In undertaking this sensitivity analysis, two questions are considered: What will be the critical changes arising from the group of dams in their own location compared to the overall Mekong terrestrial ecosystem and agricuture? What will be the critical changes arising from the group of dams on other parts of the Mekong terrestrial ecosystem and agricuture? CASCADE OF 6 DAMS UPSTREAM OF VIENTIANE If the upper cascade of six dams were installed alone, the changes to the landscape and terrestrial ecosystems of Zone 2 would be significant. Large parts of this zone are considered to be Key Biodiversity Areas, although none are actually protected as nationally protected areas. The cascade of dams will change the landscape in the Mekong valley, converting about 55% of the natural water bodies of the zone into inundated reservoir. Impacts upon the agricultural and forestry systems of the area are limited. There would be some relatively small losses of agricultural productivity, but this would be more than offset by the proposed irrigation schemes associated with Pak Chom dam. In terms of lost production of river bank gardens in Zone 2, the losses due to inundation are estimated at about 9 million USD per year which amounts to over 90% of the total Mekong river bank production in Zone 2. The dams in this cascade will have an influence upon the terrestrial ecosystem and agriculture in zones lower down in the river, resulting from the effective trapping of fine sediment with the loss of associated nutrients and fertility MIDDLE MEKONG DAMS The two dams in the middle reaches of the Mekong, Ban Koum, shared between Thailand and Laos, and Latsua which lies completely in Laos, have a significant impact upon the protected areas and key biodiversity areas in Zone 3. Ban Koum dam in particular is located between two national protected areas in Thailand and Laos, and whilst a small proportion of the terrestrial ecosystem will be inundated, the landscape in this sensitive and touristically important area will be changed significantly, from a seasonally variable river landscape to reservoir. The losses of inundated agricultural land are relatively small, and substantial irrigation schemes are associated with both dams, which will increase agricultural production in the zone by more then 10 times the lost production due to inundation. The losses in terms of production of river bank gardens are estimated at about 9 million USD per year. This is about 22% of the total production from Mekong river bank gardens in the Zone 3. The dams in Zone 3 will have some influence upon the terrestrial ecosystem and agriculture in zones lower down in the river due to the trapping of fine sediments and associated nutrients. In other respects, the terrestrial impacts of these dams are restricted to the zone in which they occur. 118 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

119 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E LOWER MEKONG DAMS (CAMBODIAN DAMS) As with the aquatic ecosystems impacts, the significance of the dams in the lower part of the Mekong in Cambodia, is that they would inundate one of the richest and most biodiverse areas of the entire Mekong system, an area which is of global importance to terrestrial as well as aquatic biodiversity. This is a unique area with interrelated aquatic and terrestrial ecosystems, serving a diverse range of land based flora and fauna. The Ramsar site established at Stung Treng, is important for its birds and mammals as well as its fish, and the area between Kratie and Stung Treng provides important habitat for hog deer and other mammals as well as the Irrawaddy Dolphin. There will also be loss of important flooded and riverine forests that are unique and can not be replaced or found elsewhere. In terms of agricultural land there will be a loss of a relatively small amount of agricultural production, but it is not known if there will be compensatory irrigation schemes. In terms of river bank garden production, about 45% of the current production will be lost, estimated at a value of 2.7 million USD per year. The Cambodian dams will have little influence upon the terrestrial ecosystem and agriculture in the zones above. However, Stung Treng and Sambor dams do have the potential to influence the terrestrial and agricultural ecosystems in the zones below, principally because of the influence upon the hydrology, flooding patterns and sediment distribution in the floodplain. The additional changes in hydrology are likely to be small in comparison to the changes occurring as a result of the Chinese and tributary dams, (i.e. the definite future) but these will affect the grassland ecosystems in the Cambodian floodplains and flooded forests in the Tonle Sap (see hydrology and aquatic impacts paper). 6.5 POSSIBLE INDIRECT LINKS WITH OTHER THEMES The impacts upon the terrestrial ecosystems and agriculture along the Mekong have a direct relevance to the social themes and the livelihoods of riparian communities. In particular the river bank gardens are often an essential component of riparian communities, providing both a source of income and essential subsistence crops vegetables, fruits, tobacco etc. In addition, the direct loss of agricultural land due to inundation by the reservoirs, will mean a loss of income and/or rice production will affect these riparian communities. Even though there may be overall increases in rice production due to new irrigation schemes in each zone, the distribution of these benefits will be limited to some of the riparian communities near only 3 of the dams, rather than being spread evenly over all of the dams. These have implications for both social and economic themes. The losses to biodiversity and the changing landscape of major portions of the river valley, in all three zones (2, 3 and 4) will have negative impacts upon the tourism potential and attractiveness of these zones. The losses of forestry resources and terrestrial resources due to the transmission lines may be locally important, but without specific alignments it is difficult to define their importance. 119 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

120 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E 6.6 SUMMARY OF IMPACTS ON TERRESTRIAL ECOSYSTEMS AND AGRICULTURE The Mekong mainstream dams will have a significant local impact upon the terrestrial ecosystems and agriculture in the areas of inundation. There will be little impact outside the areas of inundation and land take for transmission lines and access roads. The reservoirs inundate the river channel, with some smaller areas of forest and cultivated land inundated along the river banks for most of the dams. The two Cambodian dams differ in that they will flood larger areas, including forest and cultivated land. However, the reservoirs will change the landscape of the Mekong river valley, maintaining the water levels slightly above the current high flow levels, with little seasonal changes. This will have an impact upon the terrestrial biodiversity which is considered to be significant about half the length of the Lower Mekong has been recognized as Key Biodiversity Areas (even though only a small length in Zone 3 is a part of a National Protected Area), and the river above Stung Treng is a designated Ramsar Site. Significant parts of the Key Biodiversity Areas will be affected by the dams. Small areas of cultivated land will be lost, but the total loss in production will be more than recovered by irrigation schemes in three of the dams. River bank gardens in the reservoir areas will be lost, with some significant impacts on livelihoods of riparian communities. 6.7 MAPS Figure 14: Zone 2 - Extent of inundation on land use by Louangprabang dam (nearest the dam) 120 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

121 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 15: Zone 2 - Extent of inundation on land use by Pak Lay dam (nearest the dam site) Figure 16: Zone 2 - Extent of inundation on land use by Pak Chom dam (nearest the dam site) 121 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

122 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 17: Zone 3 - Extent of inundation on land use by Ban Koum dam, (nearest the dam site) Figure 18: Zone 3 - Extent of inundation on land use by Latsua dam 122 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

123 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T T E R R E S T R I A L & A G R I C U L T U R E Figure 19: Zone 4 - Extent of inundation on land use by Stung Treng dam Map 7: Zone 4 - Extent of inundation on land use by Sambor dam 123 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

124 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S 7 AQUATIC SYSTEMS 7.1 STRATEGIC ISSUES 5. Productivity of aquatic habitats in the Mekong 6. Biodiversity of aquatic habitats in the Mekong 7. The capacity of the Mekong s ecosystem regulating services purification and water quality 8. Value of Mekong River s cultural ecosystem services inspiration, recreation and tourism 7.2 SUMMARY OF PAST AND FUTURE TRENDS WITHOUT LMB MAINSTREAM HYDROPOWER The aquatic ecosystems of the Mekong are relatively natural at the moment, with high diversity of aquatic habitats rapids, deep pools, sandbars etc. that all contribute to the very high biodiversity in the river. There have been some changes in recent years, e.g. the development of two upstream dams in China, and on some of the tributaries in the LMB, that have begun to alter the hydrology and patterns of sediment discharge, so that the river morphology is beginning to change. As these developments increase in size and number, so this process of change will continue in the absence of the mainstream dams. Whilst the river is relatively clean and in good ecosystem health at present, there are increasing point sources of pollution, e.g. urban areas, and dispersed sources, e.g. agricultural run-off, these are mitigated by the large dilution effect of the river flow. However there are signs of decreasing water quality and these are expected to increase in the future with growth of population. The Mekong is recognized as having an immense cultural value for the riparian cities and communities and for tourism. The tourist attraction of a large, dramatic, near-natural river, a feature of the GMS tourism strategy, is expected to continue to increase in the future. 7.3 EXPECTED DIRECT EFFECTS OF THE PROPOSED LMB MAINSTREAM HYDRO PROJECTS The first task in this assessment of impacts of the mainstream dams is to develop an understanding of the key changes that are likely to take place as a result of their construction and operation. These are described generically and then cumulatively for all 12 hydropower schemes. Finally a sensitivity analysis is carried out on three groups of dams the cascade above Vientiane, from Vientiane to Siphandone, and from Siphandone to Kratie these relate to Zones 2, 3 and CONSTRUCTION IMPACTS If the dams are approved, as currently scheduled (MRC Hydropower database) the cascade upstream of Vientiane would be commissioned (i.e. constructed and operating commercially) by The dams at Sanakham, Pak Chom and Latsua are scheduled for commissioning between 2017 and Don Sahong and Thakho diversion are scheduled for completion by 2016 and Sambor by There is no date provided yet for Stung Treng. It is almost inevitable that many of these commissioning dates will slip, if 124 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

125 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S only because the approval process for the first set of dams will be prolonged into 2011 or However, for the purposes of this assessment, the construction of all 12 hydropower schemes is likely to extend for the next years. Therefore the construction impacts will be felt on the Mekong aquatic ecosystems for at least this period. There would in addition be a transition phase of at least 5 years after the dams are constructed, whilst the aquatic ecosystem adjusts to the changed conditions. It is therefore safe to say that construction impacts will be felt in different parts of the river to at least The key changes that will be felt during the construction period are: Increased release of sediment from construction activities, earth moving, tunneling, road construction along the river bank Increased risk of pollution from construction camps, stores and construction activities, including organic waste from sewage, spillage of oil and fuel, accidental release of toxic construction materials, with consequent kills of fish and other aquatic animals. Changes in flow as parts of the river are diverted past coffer dams and tunnels, and in the end during the filling of the reservoirs, although this will be relative short process for run-of river dams Changes in water quality as a result of breakdown of vegetation in the reservoirs, although this is expected to be relatively small because of the small footprints of the reservoirs, most of which lies in the mainstream channel Blockage of key sections of the river for fish movement and migration, both as a result of physical barriers, such as temporary and permanent dams, and as a result of declining water quality (especially sediment, with organic and toxic pollution) Blockage of key sections of the river for navigation especially in the reaches between Vientiane and Chiang Saen OPERATION IMPACTS As the hydropower schemes move into full operation, there are a number of changes that will affect the aquatic ecosystem, its productivity and biodiversity. These include: Seasonal changes in flow: the dams have very little active storage capacity and they will tend to pass on the inflows coming through them. The total storage capacity of all twelve hydropower schemes is something over 2,553 m.cu.m. The average yearly flow down the Mekong is 14,150 cu.m/sec. Thus the total active storage of the proposed mainstream dams would delay the flow into the Delta by about 2 days. In the dry season, when average flows are around 3,000 cu.m/sec at Kratie, the storage due to the mainstream dams might delay the first flushes of the river at the beginning of the wet season by up to 10 days. Of course there will be individual variations for the dams in different zones. The storage capacity of the Chinese dams and the large storage dams on the tributaries will have a much greater influence upon the seasonal flows. From a landscape point of view, the reservoirs will appear throughout the year like the river at the height of the flood season, with the channel full. The seasonal variation in flow patterns will be reflected in terms of the flow velocity through the reservoirs, being much the same as at present in the wet season but much slower in the dry season. There is the same volume of water spread throughout the whole channel of the river, rather than being confined to the dry season channel. 125 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

126 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Daily changes in flow: the operational rules for each of the dams is not yet clear; the decision to operate the dams in peaking mode rather than continuous generation will depend upon tariff agreements, and the environmental flow regimes required by the consenting agencies. A number of the dams will be operated as straight run-of-river dams, with little peaking, e.g probably Pak Chom, Ban Koum and Latsua. In these cases the water levels in the reservoir will depend upon the flow, and there will be little change in the downstream flows during the day. If the dams are operated in peaking mode, the downstream flows may be increased by several times the flow compared to the minimum flows released during non-peak times. The level of the water in the reservoirs may change by as much as 2 m during the day for many of the dams and up to 5 m for dams such as Pak Beng, Xayaburi and Stung Treng. During the high flow season, this will be of relatively little consequence as the flows downstream will generally be much greater than the design flows; however, during the dry season, the downstream flows released will become critical for the aquatic habitats, and the differences in levels in the reservoir will become more pronounced. Sediment movement and trapping: The predicted changes in passage of the sediment downstream as a result of trapping in Chinese and tributary dams shows that the sediment already in the system will continue to pass on downstream to the Delta, but will not be replaced to the same extent from sources upstream. Any changes due to the mainstream dams are superimposed upon the trapping of sediment that will take place in the Chinese mainstream dams and the storage dams on the tributaries. Run-of-river dams generally trap less sediment than storage dams, although a cascade of dams would cumulatively trap significant proportions (see section on hydrology and sediments); however, the trapping attributable to the mainstream dams may provide a tipping point of significant shift in the sediment dynamics of the Mekong. The changes in the hydrodynamics of the river caused by these mainstream dams will change the patterns of sediment movement and deposition, so that in general terms sediment will tend to build up at the head of the reservoirs in the dry season and be washed down into the main body of the reservoir with the flood waters. Deep pools at the top ends of the reservoirs may continue to operate as before, filling at the end of the wet season, and passing sediment on at the beginning of the dry season; however further down the reservoirs the deep pools will not flush out and will fill up with bedload and sediment. Immediately downstream of the dams there will be a tendency to scour the bed, as the potential energy of the water released in power generation is localized. This scour may extend for several kilometers downstream as the channel realigns itself. In the longer term the deep pools downstream of the dams will continue to operate as they have, but with less sediment passing through them. Hydrological changes over the next 20 years will have significant implications for the start and duration of the transition periods between dry and wet seasons, and between wet and dry seasons. In broad terms the Chinese dams and the storage dams on the tributaries will tend to delay the onset of the important ecological transition between dry and wet seasons, and will shorten this transition period, in some zones almost entirely (see explanation in the hydrological impacts paper). With the mainstream dams this trend is likely to be made more extreme. Vientiane at the end of Zone 2 may experience a delay in the onset of transition by up t a month, and the transition period be almost eliminated. The influence of the tributaries entering the river 126 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

127 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S in Zones 3 and 4, means that the onset is less delayed by about 2 weeks and overall as it reaches the delta by 3 weeks. The ecological implication of this delay in the onset of the transition from low to high flows is complex. In qualitative terms, the hydrological changes - early peaks in flows, natural chemicals washed out of the soils, local rainfall - often provide triggers for spawning of fish, for egg laying and hatching of eggs of amphibians and turtles. The close linkage between the timing of such events and availability of food, or the currents of the river carrying fish fry to the areas where they grow etc, is critical and if the timing is disturbed, breeding success may be very limited. If this goes on for a number of years it would lead to a crash in the populations of those affected species. The timing of the onset of the wet to dry transition is also important, and whilst hydrological modeling does not show much change in the date of the end of the flood season (usually early November), the duration of the transition period together with the dry season shows significant extension for 20 year scenarios with and without the mainstream dams. Usually the effect of the mainstream dams would be to extend this period from the end of the flood to end of dry season by up to 40 days in Zone 2, days in Zone 3, 4 and 5. During the transition period, as the waters recede, there is intense worm activity on the banks as they are exposed, breaking down vegetation and adding to the fertility of the river banks. The extended dry season, albeit with higher flows and water levels, will have ecological consequences that may change the balance of productivity and species diversity in the river. 127 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

128 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Map 1: Even in narrow river valleys, there are seasonally exposed in channel wetland areas Xayaburi dam site 128 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

129 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 1: Photo of Xayaburi Dam site. The water levels of the reservoir would be kept near this level IMPLICATIONS FOR THE AQUATIC ECOSYSTEM HABITAT CHANGE When the dams are finished and the reservoirs behind them are created, there would be a change in the aquatic habitats. In the areas inundated there would be losses of rapids and riffles and sandbars, and deep pools would tend to silt up. 996 km out of 2427 km of the Lower Mekong River (41% would be converted to run-of-river reservoir. 798 out of the 1100 areas of river over 10m deep would become part of reservoirs, with 38% of the areas occupied by pools of m deep affected, 45% of areas occupied by pools m deep affected and 59% of the areas occupied by pools over 30m deep. 93% of the area occupied by deep pools over 50 m deep will be affected by reservoirs. There would be an inundation of 76% of the total rock and rapids areas of the Mekong and 16% of the sand bars. These figures are summarized in Table 1. In each of the zones the following changes are likely: ZONE 1: No change ZONE 2: Out of a total of 795 km of river, 694 km (87%) would be converted from a free flowing river to reservoir with some daily fluctuation in water levels. There will be the usual flows through these 129 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

130 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S reservoirs, so there will be very little storage. This will mean that out of 460 sq km of the river channel, some 369 sq km (80%) would become part of the reservoirs. Of this about 50% is considered to be the dry season channel. Deep pools: 571 of the areas of river over 10 m deep will be affected, out of a total of 598 (95%). Most of these are between 10 and 20 m deep with a total area of 18.5 sq km out of sq km (85%); 71% of the total area of deep pools between 20 and 30 m deep and 25% of the pools deeper than 30 m will be affected by the reservoirs. Rocks and rapids: sq km of rocks and rapids in this zone will be inundated by the reservoirs out of a total of sq km (94%) Sand bars: sq km of sand bars in this zone will be inundated by the reservoirs out of a total of (57%) Figure 2: Photo of Pak Chom exposed river channel during recent low flows, showing diversity of wetland habitats 130 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

131 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Map 2: Area around Pak Chom (Zone 2), showing the seasonally exposed in-channel wetland areas, with the diversity of habitat. 131 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

132 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 3: Photo of Pak Chom area showing diversity of aquatic habitats at high flows ZONE 3: Out of a total of 713 km of river, 159 km (22%) would be converted from a free-flowing river to reservoir in this zone. Deep pools: 116 of the areas of river over 10 m deep will be affected, out of a total of 172 (67%). Most of these are between 10 and 20 m deep with a total area of 16 sq km out of 28.6 sq km (56%); 61% of the total area of deep pools between 20 and 30 m deep and 61% of the pools deeper than 30 m and 80% of the deep pools over 50 will be affected by the reservoirs. Rocks and rapids: 42 sq km of rocks and rapids in this zone will be inundated by the reservoirs out of a total of 66 sq km (75%) Sand bars: 1.17 sq km of sand bars in this zone will be inundated by the reservoirs out of a total of 55 sq km (57%) ZONE 4: Out of a total of 330 km of river in this zone, 143 km (43%) would be converted from free flowing river to reservoir. Deep pools: 111 of the areas of river over 10 m deep will be affected, out of a total of 181 (61%). Most of these are between 10 and 20 m deep with a total area of 13.4 sq km out of 25.3 sq km (53%); 72% of the total area of deep pools between 20 and 30 m deep and 81% of the pools deeper than 30 m will be affected by the reservoirs. 100% of the areas of deep pools over 50m deep would be affected by the reservoirs. Rocks and rapids: 3.1 sq km of rocks and rapids in this zone will be inundated by the reservoirs out of a total of 42 sq km (7%) 132 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

133 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Sand bars: 13 sq km of sand bars in this zone will be inundated by the reservoirs out of a total of 73 (18%) ZONE 5 AND ZONE 6: No change These changes in river morphology and habitat represent a very significant change in the landscape value and habitat diversity of the Lower Mekong. In Zone 2 this is taken to an extreme, where the cascade of dams from Pak Beng to Pak Chom will convert the river to a series of long thin reservoirs punctuated by a series of dams one leading into the next, with only a small piece of free-flowing river between Louangprabang dam site and the city, where the top of the Xayaburi reservoir starts. In this cascade the loss of particularly diverse habitats around Pak Chom and Sanakham may be very significant from an aquatic ecology point of view. In Zone 3, the extent of the reservoirs is not nearly so great, but the losses of deep pools and especially rocks and rapids is greater than might be expected, showing that Ban Koum reservoir especially actually inundates significant proportions of the important habitat diversity in this zone. In Zone 4, just under half the length of the river would be converted to reservoir, and a lot of the braided river channels and islands would be inundated. Whilst the rocks and rapids and sand bars are apparently less affected, there would be a considerable loss of the habitat diversity and landscape value, especially due to the Cambodian dams. CHANGES IN NET PRIMARY PRODUCTIVITY These changes in the river morphology and flow patterns will have an effect upon the primary productivity of the river. A first approximation of the order of magnitude of change in primary productivity as a result of the loss of seasonal exposure of in channel wetland areas has been made and is shown in Table 2. Whilst the actual amounts of NPP (net primary production) calculated in each zone should not be taken literally, the relative changes appear significant. For the whole of the mainstream of the Lower Mekong (excluding the Delta and Tonle Sap), the eleven mainstream reservoirs will cover about 38% of the length of the river (911 km) and reduce the area of river channel exposed during the dry season, which gives rise to varied in-channel wetland habitats, by about 48%. This may depress the NPP of the whole river system of the Lower Mekong by a range of 12 27% depending upon the rate of primary production ( Kg/m 2 /yr). When considering the separate zones, Zone 2, in which 77% of the length of the zone will be changed by the reservoirs, stands to lose about 82% of the seasonally exposed in-channel wetlands, and may lose between 27 53% of its NPP. In this zone, Pak Chom reservoir contributes the greatest loss of seasonally exposed wetlands, about 26%, followed by Xayaburi which contributes some 20% to this loss of productivity. The other dams contributions to this loss range from 11% (Louangprabang) to 15% (Sanakham). One direct use of primary production in this area is the collection of Mekong river weed, which forms an important livelihood opportunity for riparian communities. It is likely that under the new reservoir regime, the growth of this filamentous algae on the river bed in the dry season, would become very limited and could only be collected in some of the tributaries or above Chiang Khong/Houay Xai. Even in the areas downstream of the Louanprabang dam to the headwaters of Xayaburi reservoir, the growth of the algae will be limited by the daily fluctuations in the water levels and rates of flow. 133 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

134 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 4: Mekong river weed in Louangprabang morning market a livelihood source at risk? In Zone 3, about 22% of the length will be changed by the reservoir, with about 42% of the seasonally exposed in-channel wetland areas being lost, resulting in a loss in primary productivity of between 6 and 16%. Ban Koum in this zone stands out most significantly as contributing most of this loss of wetland area. In Zone 4, about 43% of the length of the zone will be affected by the reservoirs, with a loss of about 58% of the in-channel wetland areas. This might cause a reduction in the primary productivity of the area of between 15 and 32%. The two Cambodian dams are most significant in their contribution to these reductions, with Sambor almost double the impact of Stung Treng. The BDP analysis of changes in valuable wetland areas has focused on the wetlands alongside of the Mekong or in the floodplain. These are the inundated forests and grasslands and seasonal swamps and marshes, which are largely different from the in-channel wetlands described here. Under the BDP 20 year scenario with the mainstream dams, decreases of valuable wetlands of the order of 34% in Laos, 18% in Thailand, 2.4% in Cambodia and 0.1% in Vietnam. The large proportions affected in Laos reflects the relatively small areas of wetland associated with the Mekong floodplain compared to Cambodia and Vietnam. Many of these changes are predicted to occur as a result of alterations in the hydrology induced by the Chinese and tributary dams, i.e. the definite future scenario, and can not be directly attributable to the mainstream dams. 134 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

135 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Map 3: Ban Koum dam site and reservoir, showing diversity of river channel and aquatic habitats, including seasonally exposed river channels 135 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

136 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 5: Photo of the river near Ban Koum dam site (from Pha Taem National Park), showing seasonally exposed rocks and rapids. 136 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

137 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Map 4: The Don Sahong dam occupies a small area of Siphandone but would block a strategically important fish migration route 137 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

138 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 6: Fish traps on the Hou Sahong Figure 7: Flooded forests in the Stung Treng Ramsar site would be permanently flooded 138 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

139 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Map 5: Mekong River channel between Khone Falls and Stung Treng, showing area of inundation of proposed Stung Treng dam 139 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

140 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 8: A threatened species? Tourists watching dolphins near Kratie, Cambodia Zones 5 and 6 will be unaffected by these changes, since the in-channel wetland areas will continue to be seasonally exposed as before. Whilst a mathematical relationship between primary productivity and fish production would be very complicated to determine, there is a direct linkage through the food chain. It is reasonable to project that fish yields in each of the affected zones might change by similar proportions, although there will be an additional contribution to productivity from the inundation outside the mainstream channel, which has not be included in these estimates. In general, it is predicted that the productivity of the mainstream reservoirs from this source of in-channel wetlands could be reduced by between 30 and 60% compared to their current productivity. CHANGES IN AQUATIC BIODIVERSITY The changes in aquatic biodiversity as a result of these dams stem from several different factors: Changing habitat diversity with the river in the reservoirs upstream of the dams losing much of the diversity as shown above. In particular, the diversity of rapids, riffles and sand bars will be completely lost in the reservoirs, and deep pools will either tend to be filled in or become less dynamic. Changing habitat diversity downstream of the dams as a result of variable and unpredictable flow patterns, and extremes. The trapping of sediment in both the upstream dams and the mainstream dams will tend to reduce the passage of sediment and sandbars downstream overtime, although deep pools may still provide a dry season refuge. 140 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

141 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Loss of key spawning and feeding areas within the Mekong mainstream, especially the riffles and sandbars and the in-channel wetlands, both as a result of inundation, and changing flow patterns downstream of dams Barriers to fish movement and migration 46 and isolation of sections of the river from each other and from major tributaries. These barriers are accentuated in cascades, even if the dams are fitted with effective fish passages. In a cascade, each reservoir is effectively a separate unit, especially when there are no major tributaries entering the mainstream in between the head waters of the reservoir and the dam. In general reservoirs tend to be less biodiverse than rivers; this is often a reflection of the habitat diversity. For example, in Nam Ngum the reservoir fisheries catch about 20% less species than riverine fisheries in the Nam Ngum basin (Vattenfall, 2009) Construction activities that change sediments, water quality and flows such as to drive fish away, smother important downstream habitats, or cause mortality of aquatic organisms. It has been indicated in the Nam Ngum river basin monitoring that dam construction activities can depress fish catches by up to 50% and cause localized losses of up to 30% of fish species within km downstream of the dam. (Vattenfall, 2009) IMPLICATIONS FOR AQUATIC BIODIVERSITY OF THE MEKONG Overall In general the construction of dams, the reduction in the diversity of aquatic habitats by the inundation of the reservoirs, and the changing flow patterns will tend to reduce the overall biodiversity of the Mekong. The breaking up of the river into isolated segments will isolate populations, which may no longer be viable, leading to losses of some species. This will be reflected most obviously in the fish species, which may be reduced by up to half of the recorded species in some zones. It is unlikely that there will be any direct extinctions of fish species, but there will be local extirpation in most of the lengths of river converted to reservoirs. As a result of the changing habitat in the reservoirs, some species will thrive and their populations will increase, whilst others will be reduced. The species that will thrive are likely to be generalists that can breed within the body of the reservoir, and that do not require specialized habitat (e.g. rapids and riffles) or hydrological triggers to induce spawning. Changes in biodiversity may also be measured in mollusk species diversity the Mekong has the highest number of freshwater snail species in the world and other invertebrates that make use of the different aquatic habitats available. Amphibians depend upon the wetland pools left by the receding waters for breeding - the importance of the in-channel seasonally exposed wetlands for both habitat and species diversity and productivity can not be over-emphasised. River dependent birds will also be seriously affected, especially those that use the exposed sand bars and river banks for breeding, such as the River Lapwings and Pratincoles found in the upper 46 Detailed consideration of fish migration is provided in the fisheries theme. 141 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

142 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S reaches of zones 2 and 3. Loss of habitat diversity will cause a general reduction in populations, though it is unlikely that there will be any species extinctions as a result of the mainstream dams. More specifically for each zone: Zone 1: will be affected by the changing flow and sediment patterns from the Chinese dams, and the isolation from the rest of the Lower Mekong. It will no longer receive the long distance migrations of fish such as Pangasius krempfi and Anguilla marmoratus. Zone 2: will be one of the most affected zones in terms of biodiversity. The mainstream dams will cut the zone up into isolated segments, and since there is no major tributary entering the Mekong in these segments (apart from the Nam Ou, the Nam Xuang and the Nam Kham between Louangprabang dam and Xayaburi dam), most of these segments will become increasingly species impoverished, since spawning grounds will be inundated or inaccessible. The major in-channel wetlands in the Pak Chom area, extending up past Sanakham and Pak Beng, are considered to be important breeding areas and refuges for many species, representing important transition between the species in the Upper Mekong and the Middle Mekong. If this is lost, then species richness will be lost in both Zones 2 and 3. Zone 3: although this zone has only two dams, one of them Ban Koum is located at an area of high habitat diversity, especially for rocks and rapids and deep pools, in an otherwise rather simpler aquatic zone. Ban Koum reservoir could have an impact upon the species diversity of the zone, though not enough is known about the fish species there. Zone 4: with its varied habitat diversity, this zone has arguably the richest biodiversity of the Mekong mainstream. Upper part of Siphandone may be adversely affected by the flows from Latsua, although since this appears to be operated as one of the true run-of-river dams, this may not be so damaging. Despite damming only a relatively small channel of the Mekong beside Khone Falls, the key issue facing Don Sahong dam is fish migration. The dam would block the only channel that is used by fish for migration throughout the year at all water levels. Blockage of this channel could seriously impair the diversity of the fish species and their productivity on both sides of the Khone Falls. The two Cambodian dams would flood more extensive areas than the other dams and would reduce the diversity of habitats considerably, probably with the loss of many fish, reptile and mammal species. Even if fitted with fish passages, they would likely prevent the large fish migrations that currently characterize the zone. Zone 5 and 6: do not have dams planned and so will not involve major changes of habitat. The main impacts will be upon the reduced access of upstream areas for fish migrations by Sambor dam. This may eventually lead to significant declines in migratory species in the Tonle Sap and in the Delta, if these species are no longer able to reach alternative spawning sites and growing areas. CHARISMATIC AND ENDANGERED SPECIES Irrawaddy dolphin this critically endandgered species, living in Zone 4, below the Khone Falls down to Kratie, would be further threatened by the construction and operation of the mainstream dams in the Zone, and this might be the final threat which drives the extinction of this sub-population. Construction impacts in zone would cause massive disturbance to these sensitive 142 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

143 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S animals. The Don Sahong dam lies immediately above the small group living in the deep pool shared by Lao and Cambodia and construction activities would almost certainly drive them away. The other groups living in deep pools between the Khone Falls and Kratie, would be disturbed, and would eventually be separated from each other by the Stung Treng Dam. They would also suffer from changes in availability of fish as their main food source. However, it should be noted that river dolphin populations do survive in other modified rivers such as the Indus Dolphin, which continue to live in groups separated by the barrage headponds on the river Indus. However, these structures were built many years ago when the dolphin populations in the Indus were greater. The potential for the conservation of Irrawaddy Dolphin in the hydroelectric dam reservoirs on the Mekong would need to be thoroughly studied, before making predictions that they could survive. Giant Mekong Catfish is known to migrate out of the Tonle Sap and into the Mekong mainstream for spawning. A few individuals have been caught in recent years moving up past Khone Falls, i.e. through Zone 4 and into Zone 3. The other known area where they are caught is near Chiang Khong/Houay Xay at the top end of Zone 2 and it is presumed that spawning takes place nearby. The mainstream dams would prevent the long distance migration of the Giant Mekong catfish through Zones 5, 4 and 3, and this would probably lead to its extinction form the Tonle Sap, because it would be unable to reach its spawning grounds. If the fish caught near Chiang Khong are a separate sub-population migrating in the Upper Mekong, it is possible that these fish could survive in Zones 1 and 2, but if they are all part of the same population depending upon their migration through the length of the Mekong, then they would become extinct in the wild as a direct result of the mainstream dams. Siamese crocodile: the only remaining location on the Mekong mainstream where Siamese crocodiles may exist is in small tributaries within the Stung Treng Ramsar site. The construction and inundation of the Stung Treng dam would certainly lead to the local extirpation of these small populations. Other wild populations exist in the Cardamon mountains of Cambodia, which would not be threatened by the mainstream dams. Turtles are generally found throughout the Mekong, although increasingly rare and threatened by hunting. The mainstream in Zone 4 (in Cambodia) is seen as globally significant for Cantor s giant softshell turtle and may support the largest remaining breeding populations in the basin. All turtles are threatened by loss of their breeding habitat, the sandbars exposed during the dry season. As has been shown above many of the sandbars will be lost, submerged within the reservoirs, or reduced over time as larger proportions of the sediment load is trapped behind Chinese, tributary and mainstream dams. The mainstream dams would certainly contribute significantly to significant reductions in populations of most of the turtle species living in the Mekong. Large water birds there are a number of vulnerable and endangered large water birds living in and around the Mekong, especially in Zone 4 between Siphandone, Stung Treng and Kratie. These include various storks (painted and woolly necked), greater and lesser Adjutants, and ibises such as the Great Ibis, Black-shouldered Ibis. The endangered River Terns and the endemic Mekong wagtail are also dependent upon the river. Loss of habitat for nesting and feeding is the biggest threat to most of the river dependent species, and the changes in habitat resulting from the 143 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

144 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S mainstream dams, would increase this threat. The charismatic and vulnerable Eastern Sarus Crane will be less threatened by the mainstream dams because it is mainly associated with the flooded grasslands, though its range may be limited by reduction in the area of such grasslands Otters the three species of special interest in the Mekong basin the hairy-nosed otter (endangered), the smooth coated otter and the oriental small-clawed otter (both vulnerable) have been found in the Tonle Sap system and in the stretches between Kratie and Siphandone, Zone 4. They may also be found higher up in the river, certainly in areas where there are diverse aquatic habitats that provide shelter, suitable breeding places and food sources. The effect of the dams and reservoirs to limit this diversity and reduce the availability of suitable habitats would be an additional threat to these and other otter species. CHANGES IN WATER QUALITY The general impacts of these run-of-river mainstream dams on water quality comes mainly during the construction phase. There are three types of water quality issues associated with any large infrastructure development in or around water courses: Increase in sediment as a result of rock blasting and earthmoving. It is impossible to predict how much the sediment load in the Mekong will be increased as a result of construction, but it could be locally very significant at certain times during construction, especially if this coincides with low water flows, at times when the lowest sediment loads are being carried. The significance of increased sediment released downstream is that it may deposit on and smother gravel beds and riffles downstream that are important for fish spawning. Increase in organic matter as a result of uncontrolled discharge of waste waters and solid wastes from construction camps. Again largely a localized issue given the high dilution usually afforded by the Mekong waters, but this may become significant at times of low flow. This should be controlled by adequate waste water treatment and solid waste water disposal by the contractors. There is usually an increase in organic matter and increase in the oxygen demand of the water during the initial impoundment, as the remaining vegetation in the reservoir area breaks down. This is likely to be limited in the case of these mainstream dams since they are generally confined to the existing river channel, with small landward inundation. The only exception to this would be the Cambodian dams which would flood islands and larger areas of land that may need prior clearance before inundation. Accidental spillage of construction materials, concrete, toxic compounds and fuel and oils. Discharge of used engine oils from construction vehicle maintenance into water courses is a common occurrence. The risks of this would be substantially increased by the concentration of construction work of these dams over the next 15 years. In terms of operation, these mainstream dams are not expected to have a significant impact upon water quality. There is comparatively little storage, either dead or active storage and little opportunity for the accumulation of anoxic waters, which cause problems downstream when released. Similarly, unlike storage dams which can release significantly colder waters downstream, these dams will turbining waters from the upper layers of the reservoir, which should be fully oxygenated. 144 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

145 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S In general, the sediment trapping efficiencies of both the Chinese and tributary dams are predicted to reduce the sediment load in the Mekong from about 165 M tonnes per year to between M tonnes/year (SEA, Hydrology and Sediment section). The mainstream dams could reduce this further to M tones per year. This will significantly reduce the turbidity of the water of the Mekong, especially in the wet season, when it tends to be the most turbid. The position on this is not entirely clear, because the river will try and pick up additional sediments from the bed and banks as it passes downstream after each of the dams. Whilst high turbidity is often seen as an indicator of lower quality, in the case of the Mekong the high turbidity is an important characteristic of its sediment transport. The details of sediment flushing operations of the mainstream dams have not been developed at this stage, the MRC Technical Design Guidance for Mekong Mainstream Dams (MRC 2009) indicates that during flushing or sluicing operations, the reservoir is drawn down and behaves like a normal river reach transporting accumulated sediments through the dam. This results in concentrated sediment discharges, i.e. water with high sediment loads than normal (typically exceeding 100g/l commonly as much as 1000g/l), which can adversely affect water quality, particularly when done at sensitive times of year, e.g. during low flows. The MRC water quality parameters and thresholds indicate that a Total Suspended Solids concentration of more than 50 mg/l in the river water is considered Very Bad, although naturally high and increasing TSS concentrations are observed between upstream stations and Vientiane, where they can reach an average of 400mg/l. (MRC, 2007). The flushing releases from dams are significantly higher than this, and so need careful management not to impair the operation of downstream dams in a cascade, nor the remaining riverine ecosystem downstream of others. One other aspect of solids transport down a river system, that is often overlooked, is the floating matter logs, bamboos, plastic bags, dead animals etc. When operating with the spillway gates open, these items should pass through the dams and on downstream as usual, but when just passing water through the turbines, these will be collected on grills and removed. At lower flow situations floating items will be taken out of the system. This is a dam management issue that needs appropriate coordination between the dams and operation that follows accepted standards. There may also be cumulative impacts on water quality when the effects of the mainstream dams are combined with conditions of more intensive agricultural run-off, and discharge of waste waters from urban and industrial areas, both of which are expected to increase in the future. The BDP predicts that within the next 20 years, there is likely to be an 85% increase in nitrogen and a 100% increase in phosphorus in the run-off to the river, and a 35% increase in waste waters discharged to the river. In addition it is likely that there will be a significant increase in agricultural chemicals (herbicides and fungicides). It is possible that at times of low flow as the water passes through the reservoirs, there may be localized stretches of lower water quality, although the reservoirs are generally not located near urban areas, where this may occur. The reservoirs lower down the river system are more likely to have lowered water quality than those higher up the system. The increased flows released from the Chinese dams in the dry season would counteract this tendency for lowered water quality in the reservoirs during low flow seasons. 145 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

146 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 9: Timber industries along the Mekong near Sanakham a cause of pollution? Fig 10: Logs floating downstream are a management issue for dams 146 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

147 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S CHANGES IN THE CULTURAL ECOSYSTEM SERVICES As described in baseline section the cultural ecosystem values of the Mekong river can be divided into cultural and inspirational values, key festivals, recreational and Mekong tourism features. These are largely associated with its aesthetic and landscape value and strong associations with seasonal variation. On the spiritual and inspirational aspects, important naga myths are often linked to specific locations in the river such as deep pools and rapids, places where the landscape value is also high. As can be seen from the descriptions of the changes in aquatic habitats and landscape value due to the dams, such places are likely to experience changes, many of them becoming part of reservoirs and tending to lose their unique character. The fireball festival near the confluence with the Nam Ngum river on both Lao and Thai banks is associated with naga myths and may be affected by the changes in the seasonal flow patterns. If the dams are built and reservoirs created, this will result over time in a decline in the strength of the naga myths in these areas. Festivals such as the Giant Mekong catfish festival in Chiang Khong are clearly dependent upon the presence of the endangered fish. If the dams act as a barrier to exclude the migrating fish from the area or lead (along with other threats such as changing habitat and overfishing) to their extinction, then this festival will lose its original purpose. Boat racing festivals all along the Mekong celebrate the end of the flood season, or in the case of Phnom Penh, the change back to normal of the flows in the Tonle Sap river. If these mainstream dams are built, those towns and villages living along the banks of one of the new reservoirs will experience little seasonal change in height of the waters, although the flow rates will still change significantly. Figure 11: Celebrating the start of the Giant catfish fishing at Chiang Khong (Source: Zeb Hogan) 147 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

148 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S The recreational value of the river depends highly upon its aesthetic and landscape value, often for eating and relaxing along the banks. During the low flow season the river channel itself, the banks, sandbars and exposed rocks are used for picnicking, swimming, walking, angling, playing football etc. One of the implications of lower sediment flows coupled with higher dry season flow levels, is that exposed sandbars, such as those around Vientiane, may be much smaller in the future, offering less free recreational space and reducing its value. The mainstream dams will contribute to a small extent in this decline. The most economically realizable cultural value of the river is through Mekong tourism, the economic value of which is very big. As shown in the baseline each zone has a number of unique features natural e.g. the Irrawaddy dolphin and landscape features such as the Khone Falls, and tourists are currently attracted to the river because of its reputation as a wild river, passing through magnificent landscapes. The construction of the mainstream dams will have an undoubtedly impact upon this attractiveness, and over the next 10 years the boat tourism between Louangprabang and Chiang Khong will probably be lost almost entirely. Even when the dams are operating in Zone 2, and tour boats can pass though the navigation locks, the tourism experience will be very different. Whilst the boat tourism in Siphandone will be largely untouched, the developments at Don Sahong and especially the channel of the Thakho hydropower scheme around Khone Falls will have an impact upon the tourism experience there and will reduce the natural grandeur of the Falls. The disappearance of the Irrawaddy Dolphins would be a significant loss from the Cambodian and Lao tourism attractions, although the declining populations can not yet be attributed to the mainstream dams. The decision to start building the mainstream dams may be associated with an immediate boom in last chance to see type tourism on the Mekong, which would then be followed by at least a decade of disruption as the different dams are built. Once built and operating, perceptions and tours will adapt to take advantage of the changed conditions of the dams, locks and reservoirs. Dams and reservoirs are noted tourism attractions in their own right, attracting significant numbers of domestic as well as international visitors e.g. the Three Gorges Dam in China, Hoa Binh and Yali dams in Vietnam, and resorts such as on Sirindhorn reservoir in Thailand. Although the dramatic Mekong river landscape would be changed by the mainstream dams, and the river transport up the river would be punctuated by locks around the dams, the terrestrial landscape and culture of the Mekong countries will continue to attract international visitors. 148 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

149 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Figure 12: Tourism river traffic between Louangprabang and Chiang Khong/Houayxai would be disrupted during the dam construction phase 7.4 SENSITIVITY ANALYSIS OF DAM GROUPINGS In undertaking this sensitivity analysis, two questions are considered: What will be the critical changes arising from the group of dams in their own location compared to the overall Mekong aquatic ecosystem? What will be the critical changes arising from the group of dams on other parts of the Mekong aquatic ecosystem? CASCADE OF 6 DAMS UPSTREAM OF VIENTIANE If the upper cascade of six dams were installed but none of the other dams, the changes to the aquatic ecosystems of Zone 2 would be very significant, as has been shown above. Over 80% of the zone would be changed from a free flowing river to a regulated set of reservoirs that run into each other. Similar proportions of all the aquatic habitats (rocks and rapids, riffles, sand bars and deep pools) would be changed, with a consequent loss of breeding and spawning areas. The aquatic biodiversity of this zone would become seriously impoverished, the more so because there are few major tributaries entering the zone, which can provide alternative spawning areas. Productivity of this zone would also decrease, especially for Mekong river weed. The loss of critical diverse aquatic habitats above Vientiane, e.g. Pak Chom, which probably mark a transition point between the upper river and the middle reaches, 149 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

150 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S represents a loss of a unique area in the Mekong, possibly comparable in importance, though not in scale, to aquatic ecology as some of the other areas of aquatic habitat diversity lower down the system, e.g. in Siphandone. In terms of the influence that these dams have on other zones, there will be an immediate downstream impact zone, extending at least down to Vientiane as a result of daily variations in flow, and sediment trapping and flushing discharges. However, these will certainly have been balanced out by the time the river has passed through the rest of Zone 3. The biggest impact from an aquatic ecology point of view will be as a barrier to migration of fish. The lower part of Zone 2 acts as a transition between the upper reaches and middle reaches of the Lower Mekong and hence indirectly with the lower reaches, the delta and the sea. The biggest loss would therefore be upon connectivity between the sea and the Upper Mekong. Even if all the dams in the cascade are fitted with efficient and effective fish passages, the stretch of six dams in cascade over a distance of nearly 800 km represents an impossible barrier for the long distance migratory species MIDDLE MEKONG DAMS The two dams in the middle reaches of the Mekong, Ban Koum, shared between Thailand and Laos, and Latsua which lies completely in Laos, will have less an effect upon Zone 3 itself than the cascade above Vientiane has upon Zone 2. However, Ban Koum does occupy a stretch of the river which is distinct and ecologically significant in the context of Zone 3, with almost all the deep pools and rocky/rapid areas in the Zone. These two dams are intended to operate as near to run-of river as possible with minimum daily draw down, and so should have little impact downstream in terms of daily flow variation. However the influence of Ban Koum will be felt in the aquatic ecology as far downstream as Pakse, and the influence of Latsua will be felt well down into Siphandone, but probably not beyond Khone Falls. As before these two dams will act as a significant break in the connectivity of the mainstream between the lower parts of the Mekong and the middle and upper reaches. If these two dams were constructed with effective fish passage for both upstream and downstream migrations, this might be less important, but current designs of fish passage are unlikely to be effective for more than a few species. The relevance of fish passage in this section is not just relevant for the fishery in the mainstream, but also for the tributaries in southern and central Laos. It is likely that the fish productivity and biodiversity would be lost from these tributaries as a result of these two middle reach dams LOWER MEKONG DAMS The significance of the dams in the lower Mekong, in Cambodia is that they would inundate one of the richest and most biodiverse areas of the entire Mekong system, an area which is of global importance to aquatic biodiversity. This is a unique area with immense diversity of river morphology, aquatic habitats and landscape value, both in the Mekong system, but also in other major river systems. Because of the topography and nature of the river channel, the area of inundation would be much larger than the dams upstream, and cover many of the islands, deep pools, rocks and rapids and sandbars. They would almost certainly involve the loss of rare and endangered fish species, and most probably be the final threat for the Irrawaddy Dolphin. The area is an important spawning and breeding ground for many species, and it is not clear if alternative breeding grounds would still be available. 150 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

151 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S If these lower Mekong dams were to be constructed by themselves, the biggest impact to the system would be in terms of the connectivity of the system, especially for fish migration. The Mekong above Kratie to the Laos border and up the Khone Falls is an important destination for fish migrating out of the Tonle Sap. The combination of Sambor and Stung Treng dams would effectively stop this. Sambor dam would also stop the important fish migration route up the 3S rivers, especially the Sekong. The downstream flows from Stung Treng dam might also alter the ability of migrating fishes to navigate up the 3S rivers. The smaller hydropower schemes at Khone Falls Don Sahong and Thakho are of significance here for different reasons. Thakho HPP is true run-of-river involving a diversion or a proportion of water around the Khone Falls. It will have no effect upon fish migration. Don Sahong would effectively block the only channel that is known to provide a year-round route for migrating fish. Unless alternative and effective routes for fish passage are provided, this would be a barrier for some of the important small commercial species that use it during the dry season. However, the barrier effect of Don Sahong would probably become almost irrelevant if Sambor and Stung Treng dams are built. Downstream of Sambor, the aquatic ecology of the river below Kratie would be affected by changing daily flow patterns and sediment trapping and flushing. It is probable that the reservoirs in Sambor and Stung Treng would allow more sediment to drop out than the narrow in-channel upstream dams, so there are likely to be losses of sediment to the river downstream, that will not be rectified by sediment flushing POSSIBLE INDIRECT LINKS WITH OTHER FUTURE TRENDS As can be seen from the discussion above, the aquatic ecosystem theme is highly dependent upon the changes predicted by the hydrology and sediments theme. River morphology and aquatic habitat diversity result from the interaction between flows of water and sediments and the underlying geology of the basin. There would be link between these aspects through the river morphology to navigation. Navigation will be enhanced by the higher water levels of the reservoirs, and passage through the dams facilitated by navigation locks. However, there will be section of the river immediately downstream of the dams that may prove more difficult for navigation of larger boats, since the water levels may vary on a daily basis depending upon the mode of operation. The difficulties of maintaining navigation in the reach between Vientiane and Chiang Saen during the construction period of the cascade of dams has already been mentioned. For smaller fishing boats, the rapid change in flows resulting from peaking operations and sediment flushing, can be a significant safety issue, that can only be minimised by excellent communications of such operations to local river users. The change in the primary productivity of the aquatic ecosystems of the river will have implications for the overall productivity of the river in general and in the reservoirs themselves. It is unlikely that they will be as productive in terms of the fishery as Nam Ngum, which depends largely upon the nutrient inputs from a number of different tributary rivers. The cascade dams above Vientiane in particular have very few tributaries and are a very different type of reservoir. This has implications for fish catches and the livelihoods of the riparian communities. The decline in Mekong River weed in the mainstream of zone 2 will also have serious implications for the supply of this product and another source of riparian livelihood. 151 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

152 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S During construction, water quality issues and risks have implications for river users and sources of drinking water, i.e. affecting public health of riparian communities. During operation, these risks should be less, except during periods of sediment flushing, when drinking water sources will be put at risk by the high sediment concentrations. The changes to cultural ecosystem values of the river will affect the social, cultural and religious structure of communities along the river, especially those adjacent to the reservoirs or immediately downstream of the dams. The perception and willingness to pay for river based activities of visitors and tourists to the Mekong region will be affected, especially during the construction period, and tourism products and marketing will have to be changed once the dams and reservoirs have been created to re-develop river based tourism. This has both important livelihood and economic implications. 152 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

153 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Table 1: Profile of the Mekong River Channel in the different zones, featuring the areas occupied by the proposed mainstream dams Ecological Zone Characteristic Dams China to Chiang Saen Chiang Saen to Vientiane Vientiane to Pakse Pakse to Kratie Kratie to Phnom Penh and Tonle Sap Phnom Penh to the sea Headwaters and mountain river Upland river in steep narrow valley Thai/Lao midstream and tributaries Wetlands of Siphandone, Khone Falls, Stung Treng and Kratie, including significant tributaries Chinese dams Area of channel No of deep pools Area of deep pools Habitat features (Area > 10 m Rocks Plain Dry deep in dry >=10 m >=20 m >=30 m >= 50 m and Sand section season Wet season season depth depth depth depth Rapids bars km m sq km sq km sq km sq km sq km sq km sq km sq km 2,120 Length of River section Average width of river Pak Beng Louangprabang Xayaburi Pak Lay Sanakham Pak Chom TOTAL Zone 2 dams % of Zone Ban Koum Latsua TOTAL Zone 3 dams % of Zone , Don Sahong Thakho N/A Stung Treng 52 1, Sambor 86 2, TOTAL Zone 4 dams % of Zone Floodplains and the Great Lake 364 1, Mekong Delta, Tidal zone Lower Mekong River Total Lower Mekong Total in reservoirs % I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

154 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T A Q U A T I C S Y S T E M S Table 2: Potential Changes in primary productivity in the zones due to impoundment Characteristic Headwaters and mountain river Upland river in steep narrow valley Thai/Lao midstream and tributaries Wetlands of Siphandone, Khone Falls, Stung Treng and Kratie, including significant Length of River section Channel Areas Wet season channel area exposed in dry season (lost when inundated) due to dry season channel KgC/m2/yr due to wet season channel KgC/m2/yr Primary productivity - NPP due to area exposed in dry season KgC/m2/yr due to area exposed in dry season KgC/m2/yr Total NPP of lower rate for exposed areas Total NPP of higher rate for exposed areas Dry season channel km sq km sq km sq km TonsC/Yr TonsC/Yr TonsC/Yr TonsC/Yr TonsC/Yr TonsC/Yr 2125 Chinese dams Total Zone , , , , , ,403 TOTAL Zone 2 dams Area remaining free flowing ,632 27,320 23,376 54,544 39,008 70,176 Area inundated , , ,598 Total NPP in zone with dams 149, ,774 Difference due to impoundment 54, ,629 % change due to impoundment Total Zone , ,094 72, , , ,531 TOTAL Zone 3 dams Area remaining free flowing , ,872 42,263 98, , ,355 Area inundated , ,222 39,222 Total NPP in zone with dams 230, ,577 Difference due to impoundment 15,260 55,954 % change due to impoundment Total zone , , , , , ,093 TOTAL Zone 4 dams Area remaining free flowing , ,913 69, , , ,036 Area inundated , , ,817 Total NPP in zone with dams 276, ,854 Difference due to impoundment 48, ,239 % change due to impoundment tributaries Floodplains (figures exclude the Great Lake) No impoundment , ,026 82, , , ,470 Mekong Delta, Tidal zone No impoundment 225 Total Lower Mekong ,757 2, , , ,124 1,057, ,330 1,584,496 Total with impoundment ,288 1,151,674 Difference due to impoundment , ,821 % change due to impoundment I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

155 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S 8 FISHERIES 8.1 STRATEGIC ISSUES During the present SEA, stakeholder consultations and literature review led to the identification of three main strategic issues in relation to fisheries: i) long distance migration barrier effects, ii) loss of fisheries production and iii) loss of species. These are the three main issues dealt with in the present analysis. This analysis will first focus on the consequences of mainstream dams for fish production in the Mekong Basin. In order to detail impacts on a local level, dam projects will first be clustered in relation to the three main fish migration zones. For each zone, predicted hydrological conditions with and without Lower Mekong mainstream dams will be then reviewed, and the changes in flooded area will be specified. Since dam development will not have the same impact on long distance migrants (white fish) and on local residents (black fish), species composition basinwide will be reviewed and the proportion of species at risk will be characterized. Dam development also presents opportunities for reservoir fisheries and increased aquaculture production. This report will quantify the potential production gains from dam reservoirs. For each cluster of dams, this analysis will review: i) changes in hydrology, flooded area and aquatic habitats; ii) impacts of these changes on biodiversity; iii) impacts of these changes on local production (loss of species, loss of capture fish production), and iv) gains in reservoir production (reservoir fisheries and aquaculture production). Capture fish production losses in presence of dams will be compared to possible losses in absence of dams, and will be combined with expected reservoir fish production in order to provide a net estimate of changes in fish production by cluster and by zone. Mitigation measures aimed at facilitating migrations through dams and minimizing capture production losses will also be reviewed. The overall performance in the case of Mekong mainstream dams will be assessed. Ultimately, this report summarizes predicted trends in fisheries resources, with and without mainstream dams, in 10 and 30 years from now. 8.2 THEMATIC OVERVIEW An update about the relative share of black fish and white fish basinwide led to the conclusion that between 35% and 70% of the fish caught basinwide are long-distance migratory species vulnerable to mainstream dam development. The current level of knowledge does not allow a lower uncertainty range. Losses in capture fish production will be high even in absence of mainstream dams. The presence of 77 tributary dams in the basin by 2030 will result in obstruction of 37% of fish migration routes and loss of at least 250,000 ha of floodplains. However losses in fish production in absence of mainstream dams cannot be precisely quantified. Total fish production at risk from mainstream dam development ranges between 700,000 tonnes and 1.4 million tonnes per year (170, ,000 tonnes for Cambodia, 60, ,000 tonnes for Laos, 250, ,000 tonnes for Thailand and 240, ,000 tonnes for Vietnam). This estimate converges with the estimate resulting from the expert consultation organised by the MRC in 2008 (700, million tonnes). The lower range of the latter two estimates also converges with that of the BDP2 study of impacts on fisheries, i.e. 600, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

156 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S tonnes. So the most conservative figure agreed by all studies and experts, i.e. 600,000 tonnes of fish possibly lost, is equal to the annual inland fish production of all of West Africa. In other words, all experts and studies agree that building 11 mainstream hydropower dams implies an ecological and food security risk equivalent to depriving 15 west-african countries of all their freshwater fish. The production of reservoir fish could partly compensate for these losses. For mainstream dams, the reservoir surface area alone is the best single predictor of productivity. Productivity ranges from 20 kg/ha/year to 200 kg/ha/year, the variability being partly explained by factors such as depth of the reservoir or connection with rivers upstream. This large range illustrates the largely unpredictable nature of reservoir fish production. An analysis of surface areas of reservoirs created by mainstream dams (1500 km2) and of productivity leads to the conclusion that the highest fish production that can be expected from fisheries in mainstream Mekong reservoirs would be 30,000 tonnes. In fact the most likely production represents about 10,000 tonnes. The total amount of reservoir fish produced in all reservoirs of the LMB (mainstream + tributaries dams) would range between 25,000 and 250,000 tonnes, the most likely scenario being 63,000 tonnes. However this additional fish production is very small compared to predicted losses in capture fish production resulting from the obstruction of migration routes and changes in hydrology and habitats induced by hydropower dams. Overall the study also concludes that dams located upstream of Vientiane would have less impact on fishery resources than those located further downstream. The dams having by far greatest impact on fish resources would be Latsua, Stung Treng and in particular Sambor, because they would block the migration route of most species. The BDP2 has issued on 11th May 2010 a draft independent report on impacts of mainstream dams on fisheries resources. The findings of this report, briefly reviewed, are in line with the findings of the current assessment. Assessments of impacts of dams on biodiversity are more detailed in the SEA Fisheries report, whereas impacts of dams on capture fisheries and aquaculture production are more detailed in the BDP2 report. Possible losses in capture fish production, detailed in the present report, range between 700,000 and 1,400,000 tonnes. In 2008 panel of experts also estimated the possible losses at 700,000 1,600,000 tonnes (Barlow et al. 2008). The possible losses quantified in the BDP2 report amount to 600,000 tonnes. Whatever the discussion about the range of possible losses, it should be noted that the most conservative figure agreed on by all studies and experts, i.e. 600,000 tonnes of fish possibly lost, represents the annual inland fish production of the whole West Africa. 47 In other words, all experts and studies agree that building 11 mainstream hydropower dams implies an ecological and food security risk equivalent to depriving 15 west-african countries of all their freshwater fish. The present study and the BDP2 fisheries studies are also in agreement about the low fish production to be expected from reservoirs. The SEA study amounts this production to 25, ,000 tonnes, the most likely scenario being 63,000 tonnes, whereas the BDP2 study amounts this production to 16,000 64,000 tonnes. Several aspects in fisheries have not been covered by the current study nor by BDP2 study; these are: the impact of sediment retention by dams on water productivity and fish production; the changes in quality of fisheries products in the case of change in species composition, in particular the nutritional value of the newly dominant species and the possible impact of human diseases such as liver fluke present in multiple small cyprinids; the changing value of fisheries products, depending on changes in catch composition and on market demand for a scarcer product; the socio-economic issues and shifts of benefits between social groups following losses and changes in species composition. Ultimately, it should be kept in mind that both SEA and BDP2 studies are initial attempts to assess the impact of investments worth USD 18,847 million on a resource worth USD 2,100 3,800 million. The magnitude of possible 47 Benin + Burkina Faso + Chad + Côte d'ivoire + Gambia + Ghana + Guinea + Guinea Bissau + Liberia + Mali + Niger + Nigeria + Senegal + Sierra Leone + Togo; FAO statistics 156 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

157 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S impacts definitely calls for major investment in additional in-depth assessments of impacts of hydropower development on food security in the Mekong Basin. The fisheries impacts assessment is supported by Annexes II VI, which present additional assessments of the BDP scenarios, fish species and migrations. These annexes are contained within Vol III of the SEA Impacts Assessment SENSITIVTY ANALYSIS OVERVIEW For fisheries impacts, 3 clusters of dams have been examined: upstream cluster between Chiang Saen and Vientiane, middle cluster between Vientiane and Pakse, and downstream cluster between Pakse and Kratie. Table 1: Fisheries opportunities and risks for 3 key clusters of LMB mainstream projects Components of the HP plan (1) Upstream cluster of dams (2) Middle cluster of dams (3) Downstream cluster of dams Likely expected opportunities or risks caused by this component 262 species (22% of endemics) The migration barrier effect will apply to 40 sub-basins Hydrologically, the area is dominated by effects of Chinese dams a drop in the recruitment of local species is expected even in absence of mainstream dams. The contribution of these upstream species to the fish biodiversity of the basin is very important (in particular Balitoridae) 90% of river will be converted to reservoirs if 6 mainstream dams completed 41 species at specifically at risk, including the Giant Mekong catfish (at risk for extinction if dams completed) the risk of fish production losses in case the 6 upstream mainstream dams are built amounts to 130, ,000 tonnes. Reservoir fish production would range between 2,000 and 20,000 tonnes of fish, the most likely estimate being around 7,000 tonnes of reservoir fish per year in this zone. 386 species (29% of endemics) Composed of 40 sub-basins The Latsua dam would have much more negative impact on fish migrations than the Ban Kum dam because it would block access to the Mun/Chi system (70,000 km2). The Latsua dam would have the same impact than the Pak Mun dam on Mun-dependent fish species, plus additional impact on species migrating up the mainstream. The Ban Koum and Latsua mainstream dams would create 147 km 2 of reservoir, i.e. 23% of the mainstream between Pakse and Vientiane would be turned into a reservoir. This area can be expected to produce between 300 and 3,000 tonnes of reservoir fish, the most likely estimate being 330 tonnes The risk of capture fish production losses in case the 2 mainstream dams of the middle cluster are built amounts to 210, ,000 tonnes 669 species (14% of endemics) Composed of 24 sub-basins Migration barrier: six sub-basins affected, including the Sekong-Sesan-Srepok system (second largest in Mekong, area 78,648 km 2 ) If 11 reservoirs are built, 55% of the mainstream will be turned into a dam reservoir. If Cambodian dams are not built, only 48% of the mainstream would be turned into a lake The risk of fish production losses in case the 2 mainstream dams of the downstream cluster are built amounts to 220, ,000 tonnes The loss of floodplains inherent to these 2 dams corresponds to 3,000 to 12,000 tonnes of fish only The Stung Treng and Sambor mainstream dams would create 950 km 2 of reservoir. This area can be expected to produce between 2000 and 19,000 tonnes of reservoir fish, the most likely estimate being 47,000 tonnes 157 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

158 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S CROSS-CUTTING CONCLUSIONS Generic non fisheries-related conclusions In 2000, 20.6% of the Lower Mekong Basin was already barred by 16 dams and was inaccessible to fish species having to migrate to the upstream parts of the river network. In 2015, this area will have increased by 14% (from 164,000 to 188,000 km 2 ). In 2030, if 11 mainstream dams are constructed, 81.3% of the watershed will be obstructed and floodplain migrant fish will not be able to migrate further than Kratie (Sambor dam). If no mainstream dams are built in Cambodia, then 78.8% of the basin will not be accessible to long distance migrant fish. If mainstream dam development is limited to the 6 dams of the upstream cluster, then 68.7% of the basin will be barred. If no mainstream dams are built, the surface area inaccessible to long distance migrant fish is reduced to 37.3% of the watershed, despite the presence of 77 other dams on tributaries. The yearly loss in floodplain habitat and in productivity resulting from dam construction should be computed as the sum of losses in wet and in dry season. This corresponds to the surface area that will not be flooded any more in the wet season plus the surface area that will be permanently flooded in the dry season. Hydrological variations created by dam construction are detailed by the BDP2; however changes forecasted are averages by season and do not reflect daily variations. Important daily variability in downstream water levels following peak operation is a major problem for river ecology, fisheries and riverine livelihoods, as shown by the Yali dam where they have resulted in substantial losses and conflicts. Data on daily variations in flows downstream of planned mainstream dams are not available; however daily fluctuations in the level of the reservoir are expected to reach 2 meters for some projects and give an indication of the daily variability in downstream flows. Such level of variability is expected to have major effects on fish resources and on the environment in general and cannot be ignored in future analyses. If 11 reservoirs are built, 55% of the mainstream will be turned into a dam reservoir. If 6 dams are built between Chaing Saen and Vientiane, then 90% of the Mekong mainstream between these two points will be turned into a reservoir ecosystem. If Cambodian dams are not built but 9 other mainstream dams are, then 48% of the mainstream would be turned into a reservoir. An update about the relative share of black fish and white fish basinwide led to the conclusion that that between 35% and 70% of the fish production basinwide is made of long-distance migratory species vulnerable to mainstream dam development. The current level of knowledge does not allow a lower uncertainty range. A meta-review of reservoir fish production led to the conclusion that for mainstream dams, the surface area alone is the best single predictor of reservoir productivity and productivity ranges between 20 kg/ha/year and 200 kg/ha/year. This large range underscores the largely unpredictable nature of reservoir fish production. A combination of surface areas of reservoirs created by mainstream dams and productivity leads to the conclusion that the highest fish production to be expected from reservoir fisheries amounts to 30,000 tonnes basinwide. In fact the most likely production represents about 10,000 tonnes. 158 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

159 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S 8.3 MAINSTREAM DAM CLUSTERS The superimposition of maps of migration zones (Baseline Assessment, section 1.2) and of mainstream dam development leads the identification of three clusters of potential dams of significance to fish production: i) Upper Mekong, from China down to Vientiane (mountainous part of the River, altitude > 200masl); ii) Middle Mekong, from Vientiane down to Pakse (Khorat plateau; altitude between 100 and 200 masl); iii) from Pakse down to the sea (extensive wetlands and floodplains; altitude <100 masl). The first cluster includes all Chinese dams and 6 projects in the Lower Mekong Basin (from Pak Beng to Pakchom); the second cluster includes 2 dam projects in the mainstem (Ban Koum and Latsua); the last cluster includes Don Sahong and the two Cambodian mainstream projects. This clustering corresponds largely to the BDP2 scenarios for 2015 and 2030 (BDP 2010): 2015 definite future (no mainstream dams), 2030 w/o MS (no mainstream dams), 2030 w. LMD (only 6 Lao mainstream dams); 2030 w/o CMD (no Cambodian mainstream dams, 9 mainstream dams upstream of Cambodia) and year plan (11 mainstream dams). Table 9: Clusters of dams considered in the analysis of fisheries impacts Area Cluster Fish migration zone Dam Installed capacity (MW) Upper Mekong Lower Mekong Upstream cluster Middle cluster Downstream cluster Upper migration zone Middle migration zone Lower migration zone Operation (?) Reservoir area (km2) Length (m) Height (m) Gonguoqiao Xiaowan Manwan Dachaoshan 1, Nuoshadu 5, Jinhong Ganlanba Mengsong Pak Beng Luang Prabang 1, , Xayaburi 1, Pak Lay Sanakham , Pakchom , Ban Koum Latsua Don Sahong ha Stung Treng , Sambor 2, , I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

160 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Figure 20: Clusters of dams considered in the analysis of fisheries impacts UPSTREAM CLUSTER OF DAMS The upstream zone is mostly within China, and contains 262 species, including 22% of endemics (Baseline Assessment, section 1.2). The analysis below focuses more specifically on the IBFM zone nº2 (Chiang Saen to Vientiane). This zone is bounded by the Chinese border (upstream) and the Pak Chom dam (downstream), and corresponds to the upstream cluster of dams. In this zone the barrier effect on migrations will apply to 40 subbasins 48, in particular the large Nam Ou, Nam Mae Kok, Nam Tha, and Nam Khan basins. 48 Nam Mun, km2; Nam Chi, km2; Se Bang Hieng, km2; Nam Ngum, km2; Nam Cadinh, km2; Nam Songkhram, km2; Se Bang Fai, km2; Se Done, 7229 km2; Nam Nhiep, 4577 km2; Huai Luang, 4090 km2; Huai Khamouan, 3762 km2; Nam Kam, 3495 km2; H.Bang Koi, 3313 km2; Se Bang Nouan, 3048 km2; Huai Tomo, 2611 km2; Nam Hinboun, 2529 km2; Huai Som Pak, 2516 km2; H.Bang Bot, 2402 km2; Tonle Repon, 2379 km2; Nam Sane, 2226 km2; Nam Mang, 1836 km2; Huai Bang I, 1496 km2; O Talas, 1448 km2; Huai Bang Sai, 1367 km2; Nam Suai, 1247 km2; H.Ma Hiao, 990 km2; Nam Mang Ngai, 944 km2; Huai Bang Haak, 938 km2; Nam Thon, 838 km2; Huai Muk, 792 km2; Huai Thuai, 739 km2; Huai Bang Lieng, 695 km2; Huai Ho, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

161 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Figure 21: Main river basins in the upstream cluster zone MIDDLE CLUSTER OF DAMS The middle migration zone is characterized by 386 species and 29% of endemics (Baseline Assessment, section 1.2). This zone corresponds to the middle cluster of dams (Pak Chom dam upstream, Don Sahong dam downstream) and is composed of 40 sub-basins 49, in particular Nam Mun, Nam Chi, Se Bang Hieng, Nam Ngum, Nam Cadinh, Songkhram, Se Bang Fai and Se Done basins. km2; Hoaag Hua, 626 km2; H. Khok, 538 km2; Prek Mun, 476 km2; Nam Kadun, 456 km2; Nam Thong, 455 km2; H.Sophay, 186 km2; Phu Pa Huak, 132 km2. 49 Nam Mun, km2; Nam Chi, km2; Se Bang Hieng, km2; Nam Ngum, km2; Nam Cadinh, km2; Nam Songkhram, km2; Se Bang Fai, km2; Se Done, 7229 km2; Nam Nhiep, 4577 km2; Huai Luang, 4090 km2; Huai Khamouan, 3762 km2; Nam Kam, 3495 km2; H.Bang Koi, 3313 km2; Se Bang Nouan, 3048 km2; Huai Tomo, 2611 km2; Nam Hinboun, 2529 km2; Huai Som Pak, 2516 km2; H.Bang Bot, 2402 km2; Tonle Repon, 2379 km2; Nam Sane, 2226 km2; Nam Mang, 1836 km2; Huai Bang I, 1496 km2; O Talas, 1448 km2; Huai Bang Sai, 1367 km2; Nam Suai, 1247 km2; H.Ma Hiao, 990 km2; Nam Mang Ngai, 944 km2; Huai Bang Haak, 938 km2; Nam Thon, 838 km2; Huai Muk, 792 km2; Huai Thuai, 739 km2; Huai Bang Lieng, 695 km2; Huai Ho, 691 km2; Hoaag Hua, 626 km2; H. Khok, 538 km2; Prek Mun, 476 km2; Nam Kadun, 456 km2; Nam Thong, 455 km2; H.Sophay, 186 km2; Phu Pa Huak, 132 km I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

162 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Figure 22: Main river basins in the middle cluster zone The two dams of the middle cluster will have very different impacts on fish, since Latsua and Ban Kum dams are located upstream and downstream, respectively, of the mouth of the Pak Mun tributary (Figure 4). With an area of 119,707 km 2, the Mun/Chi River is the biggest hydrological basin the Mekong. Figure 23: Location of Ban Kum and Latsua dams and barrier effect on the Mun/Chi sub-basins 162 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

163 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S 8.4 DOWNSTREAM CLUSTER OF DAMS Figure 24: Main river basins in the middle cluster zone The lower migration zone contains 669 species and 14% of endemics (Baseline Assessment, section 1.2). This zone corresponds to the downstream cluster of dams (Don Sahong dam upstream, Sambor dam downstream) and is composed of 24 sub-basins 50. However in this lower migration zone only six sub-basins will be subject to the barrier effect of mainstream dams: Se Kong, Se San, Sre Pok, Prek Preah, Prek Krieng, and Prek Kamp basins. The Sekong-Sesan-Srepok system, converging upstream of Stung Treng, is the second largest hydrological sub-basin in the Mekong after the Mun/Chi system and has an area of km 2. The area downstream of Sambor dam, including the delta and the Tonle Sap sub-basin, is not affected by the barrier effect only by the hydrological modifications induced by these dams. CASE OF THE DON SAHONG DAM The SEA team recently obtained a copy of the Feasibility Study of the Don Sahong dam project, and copies of the public consultations of the Thakho hydropower project, whose site is also located in Khone Falls, less than 4km away from Don Sahong. The comparative analysis of these documents revealed some elements relevant to this SEA: 50 Delta, km2; Sre Pok, km2; Se Kong, km2; Se San, km2; St. Sen, km2; St. Mongkol Borey, km2; St. Sreng, 9986 km2; Siem Bok, 8851 km2; St. Chinit, 8237 km2; St. Baribo, 7154 km2; Prek Thnot, 6124 km2; St. Pursat, 5965 km2; Prek Chhlong, 5957 km2; Prek Te, 4364 km2; St. Staung, 4357 km2; St. Battambang, 3708 km2; St. Dauntri, 3696 km2; St. Siem Reap, 3619 km2; Prek Krieng, 3332 km2; Tonle Sap, 2744 km2; St. Chikreng, 2714 km2; Prek Preah, 2400 km2; St. Sangker, 2344 km2; Prek Kamp, 1142 km2; 163 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

164 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Hydrological analyses presented by the developer of the Thakho project highlighted the fact that very little water flows through Hou Sahong channel (3 to 5% of the Mekong flow depending on the season): The Don Sahong feasibility study confirms the insufficient natural level of water in Hou Sahong channel 51 to operate the dam at the expected level of 360 MW. As a consequence the project made plans for excavation upstream of Hou Sahong, in order to divert and attract water naturally flowing in the other channels. This excavation would be 5-6 meters deep and 1,200 to 1,800 meters long In the low season with a water level at the entrance to Hou Sahong of RL 71.0, say, the existing channel might have a discharge of 250 m3/s or even less. Clearly, therefore, it would be necessary to introduce appreciable changes to the entrance geometry (including lowering of the channel bed over hundreds of metres) to achieve the major increases in discharge in the Hou Sahong. (Don Sahong feasibility study report, page 3-11) In its natural state, the high bed levels in the upper reaches of the Hou Sahong would restrict flow into the channel, particularly in the low flow periods, and the power station would not be able to operate at its design capacity. To overcome this, the bed of the Hou Sahong will be excavated a maximum of 5 m deep for a length of about 2 km and there will also be a similar depth of excavation into the Mekong around the entrance to the Hou Sahong. (Don Sahong Environmental Impact Assessment, page 2-3) (Don Sahong feasibility study report,figure 12-1 sheet 2) These earthworks would generate more than 1.9 million cubic meters of excavation material: The excavation will be carried out in the following stages: Stage 1 (estimated volume of 1,600,000 cu. m.); Stage 2 (estimated volume of 300,000 cu. m); Stage 3 (estimated volume of 60,000 cu. m.) (Don Sahong feasibility study report, page 12-10) 51 hou = channel in Lao language 164 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

165 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S [1.9 million cubic meters of excavation material represent about 95,000 truckloads. A fraction of this material will be used for the construction of the dam and embankments but] there will be a requirement to dispose of more than a million cubic metres of surplus rock from these excavations. (Don Sahong Environmental Impact Assessment, page 2-98). As a comparison, one million cubic meters represent approximately a height of 150m of excavation material spread over the surface area of the Vientiane airport building (6500 m 2 ). At the excavation site, the river bed is made of hard rock. This is confirmed by geological surveys of the Thakho project. Excavating such hard riverbed implies the use of dynamite or explosives. At the upstream entrance to the Hou Sahong channel, a wide bar of massive rhyolite is present as seen on the aerial photographs. This also strikes east-west across the entrance and dips to the south. Drilling has confirmed its massive and hard nature. (Don Sahong Environmental Impact Assessment, page 4-3) The consequence of these works will be a diversion of water from the other channels into Hou Sahong. If it is as high as 1,500 m3/s (for RL 71 at the entrance to Hou Sahong), clearly diversion of a discharge of a similar magnitude into Hou Sahong would i) account for virtually all of the main river discharge; ii) reduce the discharge over Phapheng Falls to a very low value (Don Sahong feasibility study report, page 3-11) While it is recognized that the Khone Phapheng waterfall is best viewed at lower flows, the amount of reduction in low season flows, the peak tourism months, is critical. (Don Sahong Environmental Impact Assessment, page 4-6). The figure below illustrates approximately the intended earthworks at the mouth of Hou Sahong: As a part of its assessment of impacts of mainstream dams, this SEA notes that: the ecological impact on fisheries and aquatic ecology of blasting and excavating more than 1.9 million cubic meters from the river bed has not been mentioned in the Environmental Impact Assessment of the Don Sahong project; the ways to dispose more than one million cubic meters of excavation material have not been specified; the impact of the water diversion planned on the discharge in Hou Phapheng and on Khone Phapheng waterfall has not been detailed. Will the reduced discharge in Hou Phapheng be enough to keep, in dry season, the visual aspect of a site known as the biggest waterfall in Southeast Asia? Will the reduced discharge in other channels be enough to allow fish migrations and fishing at critical times of the year? the impact on tourism and downstream areas of the above plans has not been discussed. 165 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

166 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S OVERVIEW Table 2 show the proportion of the Lower Mekong watershed located upstream of the dams being considered. This confirms that three dams, namely Ban Kum, Latsua and Stung Treng, have a proportionally greater share of watershed than others. Table 10: Watershed area upstream of each mainstream dam Cluster Project name Watershed area upstream % of LMB watershed upstream of the dam (km 2 ) 1 Pakbeng 218, Luangprabang 230, Xayabuly 272, Paklay 283, Sanakham 292, Pakchom 295, Ban Kum 418, Latsua 550, Don Sahong 553, Stung Treng 635, Sambor 646, LMB 795,000 Source: MRC data (BDP2) Figure 25: Mainstream dams and corresponding upstream watershed area In the Mekong Basin long-distance fish migrations occur at a large scale between downstream floodplains and the upstream sections of the Mekong and its tributaries (Baseline Assessment, section 5.4). For this reason the proportion of the upstream section of the Mekong watershed blocked by dam gives an indication of the surface area or habitat that will become inaccessible to migrant fish. In order to assess that surface area, dams have been classified by watershed then by river; the area blocked by the dam located furthest downstream is inaccessible to fish. This area was quantified for each basin and for 3 periods of time: 2000 (baseline), 2015 (Definite future) and For 2030, the scenarios detailed are i) no mainstream dams; ii) 6 mainstream dams in the upstream cluster; iii) no mainstream dams in Cambodia, and iv) 11 mainstream dams (details in Annex 1). The table below summarizes the findings: 166 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

167 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Table 11: Surface area of the Lower Mekong Basin blocked by dams under different scenarios S1. Baseline Area S2. Area S2 S3. No S1 Definite MS future dams Nb of dams / km 2 obstructed for migrations % of LMB obstructed for migrations LMB area (km 2 ) = 795, S4. 6 MS dams in upper LMB Area S4 S5. No Cam MS dams Area S5 S6. All 11 MS dams Nb of dams / km 2 obstructed for migrations % of LMB obstructed for migrations Km 2 obstructed specifically by LMB mainstream dams S4 - S3 = S5-S3 = S6 - S3 = % of LMB obstructed specifically by mainstream dams Sources: Description of BDP2 scenarios dated Conclusions: In 2000, 20.6% of the Lower Mekong Basin was already barred by 16 dams and was inaccessible to fish species having to migrate to the upstream parts of the river network. In 2015, this area will have increased by 14% (from 164,000 to 188,000 km 2 ). In 2030, if 11 mainstream dams are constructed, 81.3% of the watershed will be obstructed and floodplain migrant fish will not be able to migrate further than Kratie (Sambor dam). If no mainstream dams are built in Cambodia, then 78.8% of the basin will not be accessible to long distance migrant fish. If mainstream dam development is limited to the 6 dams of the upstream cluster, then 68.7% of the basin will be barred. If no mainstream dams are built, the surface area inaccessible to long distance migrant fish is reduced to 37.3% of the watershed, despite the presence of 77 other dams on tributaries. Since many dams on tributaries already exist or are planned, mainstream dams will not be the only cause of the barrier effect on fish migrations. The proportion of the total impact attributable to mainstream dams can be quantified by subtracting the area barred by tributary dams from the overall area barred. These results indicate that the cluster of 6 mainstream dams in upper Laos would obstruct 31% of the migration routes (but not of the migration themselves), that all LMB mainstream dams would obstruct 44% of the Lower Mekong Basin, and that in the presence of 9 other mainstream dams the Cambodian mainstream dams alone would obstruct only 3% of the basin, because most of the basin area between Sambor and Latsua (5 watersheds including Sekong, Sesan and Srepok watersheds) would already be obstructed by 21 tributary dams. Area S3 Area S6 8.5 FORECASTED HYDROLOGICAL CHANGES RELATING TO FISHERIES The dams planned will bring about a number of hydrological changes of importance for fish: i) flow reduction in the wet season and floodplain area: the Mekong system has the most productive fishery of the world because of its large floodplain and its hydrological variability. If the flow is lower in the wet season, in some places water might not reach the level of the banks and thus not spill over in plains, or not spill as far as before, which would mean a loss of habitat for fish; ii) flow increase in the dry season: some land areas that used to be dry in the dry season will become permanently flooded, and thus will lose the productivity that results from the seasonal inundation (Junk et al. 1989). As a consequence, the yearly loss in floodplain habitat and in productivity resulting from dam construction should be computed as the sum of losses in wet and in dry season (loss in variability). This corresponds to the surface area that will not be flooded any more in the wet season plus the surface area that will be permanently flooded in the dry season (double ring illustrated in Figure 7). 167 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

168 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Figure 26: Change in flooding due to dams and loss of productivity Floodplain before the dam Dry season water level Wet season water level Floodplain after the dam Area not flooded any more in the wet season New area permanently flooded in the dry season Areas characterized by loss of productivity (by loss of flood pulse) Source: Baran et al. in press. In the past years a number of studies have attempted to quantify these changes. For this SEA the reference predictions used are those of the on-going BDP2 analysis, since they based on the most updated and accurate set of scenarios. However we also included, for information, the predictions from three other large modelling studies: i) the WorldBank Mekong development scenarios (Podger et al. 2004); ii) the Nam Theun 2 Cumulative Impact Analysis (Norplan and Ecolao 2004); and iii) the BDP 1 scenarios for strategic planning (MRC 2005). The results of these respective studies are summarized below for comparison with the BDP2 scenarios. All former studies are characterized by a small number of dams, and thus can only be compared with the BDP scenario. Table 12: Development scenarios in recent hydrological modelling studies Scenarios WorldBank 2004 Nam Theun BDP BDP /2010 High development (25 dams basinwide) 2025 (28 dams basinwide) High development (21 dams basinwide) 2015 Definite future (47 dams basinwide) 2030 no LMB mainstream dams (77 dams basinwide) 2030 with 6 Lao mainstream dams (83 dams basinwide) 2030 with 9 mainstream dams not in Cambodia (86 dams basinwide) 2030 with 11 mainstream dams (88 dams basinwide) The hydrological consequences forecasted by the different studies are detailed in the table below. The results from the on-going BDP2 modelling work have been extracted from detailed hydrological forecasts provided, from Technical notes 1, 2 and 4 dated February 2010 and corresponding presentations ( 168 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

169 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Table 13: Hydrological changes forecasted by different studies for 5 scenarios Discharge in dry season (% change from baseline) Definite Future 20Y w/o MD 20Y with LMD 20Y no CMD 20Y , no mainstream dams 2030, 6 mainstream dams 2030, 9 mainstream dams 2030, 11 mainstream dams WL in wet FA, wet Discharge, dry WL, wet Discharge WL, wet FA, wet Discharge, WL, wet FA,wet Discharge WL in FA in wet season (change season (% season (% season in dry season season dry season season season in dry wet season (% from baseline change change from (change season (% (change (% (% change (change (% season (% season change in m) from baseline) from change from change from from change change (change from baseline) baseline from baseline from baseline) baseline from from from baseline) in m) baseline) in m) baseline) in m) baseline) baseline) baseline in m) Luang Prabang BDP1: +55 BDP1: -1.5 BDP2: BDP2: 0 BDP2: BDP2: BDP2: BDP2: -0.6 Nong Khai/ WB: +47 to 73 WB: -0.5 to 1.0 BDP2: BDP2: - BDP2: +54 BDP2: - Vientiane BDP2: BDP2: -0.5 Savannakhet/ WB: +30 to 52 WB: -0.4 to 0.7 BDP2: +31 BDP2: - BDP2: BDP2: - Pakse NT2: +135 NT2: -1.6 BDP1: +33 BDP1: -0.7 BDP2: BDP2: -0.3 BDP2: 32.6 BDP2: 59.7 BDP2:? BDP2: BDP2: 54.8 BDP2:? BDP2: Kratie WB: +22 to 45 WB: -0.4 to 0.8 BDP2: BDP2: BDP2: BDP2: 27.4 BDP2:? BDP2: +28 BDP2: NT2: +125 BDP1: -0.8 BDP1: +27 BDP2: -0.3 BDP2: Tonle Sap NT2: NT2: BDP1: -0.4 NT2: -8.7 BDP1: -3.4 BDP2:? BDP2:? BDP2:? BDP2:? BDP2:? BDP2: BDP2: BDP2:? BDP2: BDP2: -5 (Cambodia and LMB) Tan Chau BDP1: +26 WB: -0.2 to 0.3 BDP2: BDP2: - BDP2: BDP2: - BDP2: 13.7 BDP2:? BDP2: BDP2: BDP2: BDP1: -0.2 BDP2: -0.3 BDP2:? BDP2: BDP2: BDP2: BDP2: BDP2: -7 (Cambodia and LMB) Whole Cambodia Whole Basin BDP2: - 104,000 ha BDP2: - 251,000 ha BDP2: - 142,000 ha BDP2: - 309,000 ha BDP1: Basin Development, phase 1; BDP2: Basin Development Plan phase 2; NT2: Nam Theun 2; WB: WorldBank Q: discharge; WL: Water level; FA: Flooded area 169 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

170 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Conclusions: In the Definite Future scenario, when comparing to the 2000 baseline, BDP2 (technical notes 1 and 2 and presentations dated February 2010) predicts that dry season discharge will increase by 45% in Luang Prabang, 32% in Pakse, 22% in Kratie and 13% in Vietnam. These estimates are close to estimates from BDP1 and WorldBank studies, within a range of about 10%. This is surprising since the number of dams integrated in the BDP1 and WorldBank studies is only about half of those integrated in the BDP2 studies, which would indicate that an additional 19 to 26 dams in the basin would not make much difference for discharge levels in the dry season. Changes predicted by the Nam Theun 2 study are much more dramatic, with a dry season discharge doubling in Pakse and Kratie. In the wet season, the variable common to all studies is the water level, not discharge. According to BDP2, wet season water level is expected to be reduced by 30 to 60 cm maximum, depending on places. Previous studies tend to forecast that the reduction of water level will be higher by 10 to 90 cm. None of the former studies analysed scenarios for 2030, and thus cannot be compared to BDP2. BDP2 forecasts that twenty years from now, in the absence of mainstream dams, discharge in the dry season will be between 13% and 65% higher than in 2000 (higher values upstream). With 6, 9 or 11 mainstream dams in the Lower Mekong Basin the range will be approximately the same (13% to 58%) because most of these mainstream dams are considered run-of-the-river. The same applies to water level in the wet season, with a decrease of between 20 cm and 70 cm maximum depending on the location (more losses upstream), for all scenarios. It is important to note that hydrological changes forecasted by the BDP2 are averages by season which do not reflect daily variations. Important daily variability in downstream water level following peak operation is a major problem for river ecology, fisheries and riverine livelihoods, as shown by the Yali dam. Downstream of Yali dam in Cambodia, until at least 2003, daily fluctuations ranging between 50cm and one meter have resulted in dramatic losses in habitat, fish resources, livestock, and at least 39 casualties downstream of Yali dam (Fisheries Office of Ratanakiri Province, 2000;McKenney 2001; Baird et al. 2002, Lerner 2003, Baird and Meach 2005, Wyatt and Baird 2007). Data on daily variations in flows downstream of planned mainstream dams are not available; however expected daily fluctuations in the level of the reservoir (i.e. upstream) are indicated for some projects and give an indication of the daily variability in downstream flows. In the case of Luang Prabang, Latsua and Stung Treng dams, two meters of daily fluctuation correspond respectively to 87, 23 and 428 million cubic meters (see Table 10). Such level of variability is expected to have major effects on fish resources and on the environment in general and cannot be ignored. However the complete absence of data about this phenomenon did not allow factoring it into the current impact analysis. The latest BDP2 scenarios for 2015 predict a loss of 1040 km 2 of floodplains in Cambodia compared to the situation in 2000, and of 2510 km 2 for the whole Mekong Basin (-5%). In 2030, with 11 mainstream dams, a total loss of 1420 km 2 and 3090 km 2 is forecasted in Cambodia and in the whole Basin respectively (-7%). It is not clear whether the loss of floodplain area forecasted by the BDP2, which seems relatively limited (7% loss maximum after the total number of dams has increased from 16 dams in 2000 to 88 in the MD scenario), includes the double ring detailed above. Regarding the Tonle Sap area (an area that generates 60% of the Cambodian fish production), some additional forecasts are proposed by the BDP2 (BDP2 2010): 170 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

171 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Reduction of the total flooded area by 60,000 ha (4.5 %) in an average year, and as much as 100,000 ha (9%) in a dry year; Reduction of the area of flooded forest by 5,000 ha (1.1%) in an average year to 23,000 ha (5.3 %) in a dry year; Reduction of the area of inundated grasslands by 8,500 ha (3.2%) in an average year to 25,000 ha (10 %) in a dry year; Reduction of the area of flooded marshes by 3,000 ha (1.0%) in an average year to 5,500 ha (1.8 %) in a dry year; Reduction of the area of flooded rice fields of 41,000 ha (18%) in an average year and 48,000 ha (28 %) in a dry year; Reduction of flood depth of just over 0.5 m in an average and dry year; Reduction of flood duration of the flooded forest area by generally less than 2 weeks in an average year, but up to 1 month in a dry year; Reduction in flood duration by generally less than 1 month in an average year in 70% of the inundated grassland area, but an increase of flood duration with up to 1 month in 25% of the area; Increase of the water level in the dry season with about 30 cm, resulting in a volume increase of 780 MCM, or an increase of over 50%; Reduction of sediment inflow in the system of at least 8 to 13%. Table 14: Mode of operation of the planned mainstream dams and expected daily fluctuation in the reservoir level (Source: SEA Inception Report Vol 2) Mode of operation Daily fluctuations in the reservoir (m) Pak Beng NA NA Louang Prabang peak load h/day -2m Xayaburi NA 0 Pak Lay Peak load 8-10 hours/day 1-2 m Sanakham NA NA Pakchom Continuous - 2m Ban Koum Continuous NA Latsua Peak load >16 hours/day -2m Don Sahong Continuous NA Stung Treng Continuous 2m Sambor NA small 8.6 WETLANDS, FLOODPLAINS AND FISH PRODUCTIVITY Changes in floodplain areas highlighted in Table 5 imply changes in related fish production. This relationship between wetland or floodplain area and fish yield has been studied in detail and the comprehensive reviews done by Hortle (2007) and Hortle et al. (2008) are summarized below. Welcomme (1985) estimates that floodplain area can predict 70% of floodplain river productivity, and kg/ha/year is a typical range for tropical floodplain rivers. In the LMB, however, natural productivity exceeds that of many other tropical floodplains, and is estimated to range between 25 and 630 kg/ha/year, with a mean yield of 119 ± 25 kg/ha/year (average of 18 studies, details in Hortle et al. 2008, 171 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

172 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S p. 39). This value is lower and more accurate than the 230 kg/ha/year mentioned in Baran et al. (2001) and Sverdrup-Jensen (2002) 52 ). To reflect the variability in floodplain productivity, Hortle (2007) used low, medium, and high values of productivity per hectare (respectively 50, 100, and 200 kg/ha/year) and confirmed this range later on (Hortle et al 2008), insisting on the middle value (119 ± 25 kg/ha/year) as the most likely average figure. Notes: when relating floodplain productivity, floodplain area and total production, the most recent wetland surface area estimates available and published are those detailed in Hortle et al. (2008, p. 40), updating those of Hortle there is confusion in the literature between wetlands, floodplains and rice field areas. Multiple authors do not differentiate between floodplains and wetlands, although Hortle et al. (2008) distinguish between wetlands, rice fields, swamps, flooded natural vegetation and permanent water bodies. However in the surface areas detailed by these authors, rainfed rice fields (RRF) are lumped with flooded rice fields, although it is acknowledged that the productivity of the latter (around 200 kg/ha/y) is double that of the former. in the unofficial IBFM report nº 8 (King et al. 2005), fish productivity per hectare is not detailed, but a non-linear relationship is assumed and it is estimated that a 10% reduction in floodplain area would result in a 20-30% reduction in fisheries productivity. This study also points out that loss of production in due to reduced floodplain area can be exacerbated by a delayed flood duration (issue highlighted and detailed in Baran et al. 2001, 2005, and Kurien et al. 2006) 8.7 LONG-DISTANCE MIGRANTS AND MEKONG FISH PRODUCTION A number of Mekong fish species migrate between floodplains and tributaries, but details about where they migrate to have never been summarized. For this study we reviewed existing information to characterize the migration of as many species as possible, and combined this information with the contribution of these species to total catches. The methods are detailed below, and results will be detailed for each dam cluster in a following section MIGRATION PATTERNS OF DOMINANT SPECIES This synthesis is initially based on migration maps available for 23 species in the Mekong Fish Database (MFD 2003, see Annex 2). On each map five barriers have been represented: i) Sambor blocking migrations upstream and towards the 3S system; ii) Stung Treng/Don Sahong blocking migrations through Khone Falls; iii) Latsua blocking migrations towards the Mun/Chi system; iv) Ban Kum blocking migrations towards Vientiane, and v) the upstream cluster blocking access to the upstream migration zone. This analysis of 23 fish taxa was complemented by a synthesis of all ecological information published in Mekong Fish Database and in FishBase, obtained by merging these two databases (FishBase having more 52 yet figures of total production in these studies are close to Hortle s figures because the surface area of wetlands was updated. 172 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

173 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S information on the taxonomy and biology of the species, and Mekong Fish Database on their ecology and distribution). This combination provides information (with more or less details depending how well a given species is known) for 768 species. An example of the information synthesized is given in Annex 3 for two well-known migrant species. The analysis focused on six main upstream migration patterns : i) fish migrating from floodplains up to Kratie/Sambor; ii) fish migrating to the 3S system (Sekong-Sesan-Srepok); iii) fish migrating through Khone Falls; iv) fish migrating to the Mun/Chi system; v) fish migrating upstream of Pakse; vi) fish migrating upstream of Vientiane. Ultimately a total of 46 species displaying particular migration patterns or critical habitats in the different zones describe above were identified. The matrix of species by zone (Annex 4) was complemented by the contribution of each species to total catches basinwide, based on data detailed in section 5.2.1). A summary of that information is detailed in Table 7. Table 15: Summary of the migration patterns of 43 species dominant in catches 53 Number of species % of species reviewed All species reviewed Migratio n up to Kratie Migration to the 3S system Migration through Khone Falls Migration to the Mun/Chi system Migration up to Vientiane Migration upstream of Vientiane Note: This review only reflects the status of our knowledge about migrations in relation to a few key locations in the basin, as mentioned in the scientific literature and in databases. This review is also based on information available in written form in the two databases cited; there is additional information about migrations available but it is coded (e.g. migration timing of Hypsibarbus malcolmi in multiple locations in the Basin) and the 10 days of work available for this review of impacts did not allow systematically analysing that raw information. Some other species living in the mainstream such as Gyrinocheilus aymonieri represent 1.63% of catches basinwide but they do not exhibit long-distance migration patterns, and although they might be at risk of mainstream dam development, they have not been covered in the present analysis Conclusions: This analysis focused on 43 white fish species whose long-distance migration patterns are well known, and making up to a third of the fish catch basinwide. Out of these species, all exhibit a migration pattern between downstream floodplains and the Mekong mainstream by Kratie; 95% of them (= 28.5% of the catch basinwide) migrate through Khone Falls; and about two-thirds undertake a migration between Khone Falls and upstream (towards tributaries of the middle Mekong migration system, or upstream of Vientiane). Several limitations to this approach must be highlighted: i) upstream of Kratie it is impossible to assess how much of the catch the migrant species identified represent (what is known is only their contribution to the overall catches); to do so it would be necessary to know the catch per river stretch; this information is available in raw MRC data (see section 5.2.1) but has never been analysed so far; ii) the fact that a 53 This analysis assumes that the downstream floodplains in Cambodia and Vietnam are a starting point, and considers only upstream long distance migrations. Migration within sub-systems, from floodplains to local tributaries or lateral migrations between close habitats are not reflected here. 173 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

174 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S species migrates to a certain zone does not mean that it is specific to that zone (thus 22 of the 25 species migrating to the 3S system are also found upstream of Vientiane) nor that access to this zone is absolutely necessary to the reproduction of that species; iii) for some species there could be sub-populations able to complete their life cycle within segments of the basin (e.g: between Cambodian floodplains and the 3S system, between Thai floodplains and the upstream Mekong) without having to migrate through the whole basin. Figure 27: Number of species of long-distance migrants found in the sections of the LMB possibly barred by mainstream dams DOMINANT SPECIES IN MEKONG FISH CATCHES A debate has recently surfaced about the species composition in catches basinwide, and the relative proportion of black fish in the overall catch. This is important because the impact of dams is likely to be much less severe for black local fish than for white migrant fish (Baseline Assessment, section 6.2). We review below some approaches that either identified dominant species in overall catches, or quantified the share of black fishes in overall catches. Note: strictly speaking, the term of black fish refers to floodplain fishes only, and thus to species found in floodplain or wetlands. This term should not be used for non-migratory species found upstream of tributaries and in stream environments. 174 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

175 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S METHOD 1: CATCH MONITORING IN THE MAINSTREAM Figure 28: Site locations of the AMCF fishermen s catch monitoring survey. Fish abundance and diversity have been monitored basinwide based on gillnets operated by local fishermen (Starr 2008). This monitoring undertaken by the MRC Assessment of Mekong Capture Fisheries Project was carried out in the four LMB countries between 2002 and At each selected site three fishermen using gillnets recorded species caught daily, in different types of habitats: the main channel of the Mekong River, deep pools in the mainstream and along three tributaries, and in the delta. Data used in the analysis below correspond to the December 2003 November 2004 period. Data from this monitoring have been processed and made public for the first time in Halls and Kshatriya (in press). This study reflects catches of one specific gear only (gill nets) and is restricted to large rivers (mainstream and main tributaries), without covering floodplains and wetlands. Despite a subsequent underestimate of the production of black fish, this study is by far the most detailed source of information available to date when dealing with fish catches at the species level and basinwide. The AMCF survey resulted in reports about a total of 233 species of fish belonging to 55 families. Twentytwo species (9.4% of the species richness sampled) were identified as black fish and 150 species (64.4%of the total richness) were identified as white fish. Among those, 58 white fish species (24.9% of the species richness sampled) representing 38.5% of the total catch were considered highly vulnerable to dam development (the criteria for classification, the 58 species, and their share in catches basinwide are detailed in Annex 5. From this assessment it can also be deduced that = 61.5 % of the catch is made of species that are not highly vulnerable to dam development (this latter category includes black fish and other species). METHOD 2: SURVEYING EXPERTS In the first half of 2007, the MRC Fisheries Programme gathered an expert panel consisting of 13 fisheries scientists from Lao PDR, Cambodia and international research organizations operating in the LMB to provide an estimate of the size and value of the migratory fish resource in the LMB (details in Barlow et al. 2008). Experts were asked What percentage of the total yield from the capture fishery in the LMB is 175 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

176 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S white fish (that is, those that are highly migratory)? The combined answer was that migratory fish resources vulnerable to mainstream dam development comprise 71% of the fisheries yield in the LMB. METHOD 3: CATCH STATISTICS IN CAMBODIA In 2000, van Zalinge and his colleagues, summarizing several years of fish monitoring by the MRC/DoF/DANIDA project "Management of the Freshwater Capture Fisheries in Cambodia" (Diep Loeung et al. 1998), reached the conclusion that longitudinal migrants constitute about 63 % of the total catch taken by fisheries in the Tonle Sap area (van Zalinge et al. 2000). This proportion might somehow be an overestimate since the fishery best monitored was the dai fishery targeting white migrant fish, whereas the species composition in dispersed small-scale and familial fisheries and in the non-transparent large scale fishing lots was not well known. METHOD 4: UPDATE INTEGRATING BLACK FISH PRODUCTION IN RICE FIELDS In 2008, Hortle et al. looked in detail that the production of rice field fisheries in Cambodia, and found that 87.7% of the catch in rice fields is made of black fish, and that rice fields represent 86.1% of wetlands in the Lower Mekong Basin 54. We used this update to determine the contribution of black fish to fish production basinwide; the steps of this assessment are detailed below: Wetlands = floodplains + rice fields + other wetland types Rice fields = floodplain rice fields (high fish productivity) + rainfed rice fields (RRF, lower fish productivity) Other wetland types = permanent water bodies + aquaculture areas + swamps + flooded forest/grassland/shrubs Surface area of wetlands = surface area of floodplains (including floodplain rice fields) + rainfed rice fields + permanent water bodies + aquaculture + swamps + flooded forest/grassland/shrubs Surface area of rainfed rice fields = surface area of wetlands - floodplains - permanent water bodies - aquaculture - swamps - flooded forest/grassland/shrubs Surface area of Wetlands in the LMB: 184,900 km 2 (Hortle et al p. 40) Surface area of floodplains in the LMB: 50,152 km 2 (TKK & START-RC 2009 p. 22). Surface area of rice fields: 159,200 km 2 (Hortle et al p. 40) Surface area of permanent water bodies: 13,800 km 2 (Hortle et al p. 40) Surface area of aquaculture zones: 2,400 km 2 (Hortle et al p. 40) Surface area of swamps: 2,200 km 2 (Hortle et al p. 40) Surface area of flooded forest/grassland/shrubs: 7,300 km 2 (Hortle et al p. 40) Surface area of rainfed rice fields = 109,000 km 2 54 However a detailed analysis of productivity by wetland type, based on the reviews detailed in Hortle (2007) and Hortle et al. (2008), and detailed in another section of this report, leads to the conclusion that rice fields produce around 100 kg of fish per hectare and per year, whereas floodplains produce around 200 kg per hectare and per year. 176 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

177 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Production of RRF = productivity of RRF x surface area of RRF = Y% of (black fish) RRF Productivity of RRF: around 100 kg/ha/y (Hortle et al p.41) Surface area of RRF: 109,000 km 2 = 10,900,000 ha Fish production of rainfed rice fields = 1,090,000 tonnes Percentage of black fish in rainfed rice fields: 88% (Hortle et al, p. 19) Percentage of white fish in rainfed rice fields = = 12% (Hortle et al, p. 19) Production of black fish in rainfed rice fields = 1,090,000 x 0.88 = 959,000 tonnes Production of white fish in rainfed rice fields = 1,090,000 x 0.12 = 131,000 tonnes Production of fish basinwide: 2.1 million tonnes (based on consumption studies; Baseline Assessment section ) Production (tonnes) of fish basinwide = production of Flood Plains + production of Rainfed Rice Fields Production of Flood Plains = Production basinwide (2,100,000 t) - Production of RRF (1,090,000 t) Production of Rainfed Rice Fields = 1.01 million tonnes (made of white fish and black fish) In the Tonle Sap (i.e. floodplain system), the proportion of white fish reaches 63% (van Zalinge et al. 2000) and thus the proportion of black fish is = 37% Production of white fish in floodplains = 1,010,000 x 0.63 = 636,000 tonnes Production of black fish in floodplains = 1,010,000 x 0.37 = 374,000 tonnes Total production of black fish: 959,000 tonnes from rainfed rice fields + 374,000 tonnes from floodplains = 1.33 million tonnes, out of 2.1 million tonnes = 63% Production of black fish basinwide = 1,010,000 x 0.63 = 636,000 tonnes Percentage of black fish basinwide = 63% So based on estimates of black fish proportion in rainfed rice fields and in actual floodplains, and on the respective proportion of each habitat type in the LMB, there would be 63% of black fish in the total fish production of the Mekong Basin. As a consequence, migratory fish resources vulnerable to mainstream dam development would represent 37% of the fisheries yield in the LMB. OVERVIEW From the table below it can be concluded that between 35% and 70% of the fish production basinwide is made of long-distance migratory species vulnerable to mainstream dam development. The current level of knowledge does not allow a lower uncertainty range. Applying these figures to the estimates of total fish production detailed in the Baseline Assessment (section and Table 9 below) with a focus on catch estimates based on consumption studies since they are the most robust shows that the total fish production at risk of mainstream dam development ranges between 700,000 tonnes and 1.4 million tonnes (170, ,000 tonnes for Cambodia, 60, ,000 tonnes for Laos, 250, ,000 tonnes for Thailand and 240, ,000 tonnes for Vietnam). 177 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

178 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Table 16: Percentage of the catch made of species not highly vulnerable to dam development (black fish or others) Percentage of the catch made of fish vulnerable to dam development Quality of the estimate Mainstream catch monitoring 38.5 ** Expert judgment 71 * Catch statistics in Cambodia 63 * Integration of rice fields 37 ** 8.8 GAINS IN FISH PRODUCTION FROM DAM RESERVOIRS Utilization of dam reservoirs for fish stocking is often mentioned as a way to increase fish production; this production depends on the characteristics of the reservoir considered. Table 10 details the specifications of reservoirs created behind each of the 11 mainstream dams; it was generated by the SEA GIS team, based on a digital terrain model and the specifications of each dam. Table 17: Characteristics of each mainstream dam reservoir Reservoir Area (km 2 ) Average depth (m) Maximum volume (mcm) % of total volume at depth < 2 m < 5 m Pak Beng Louang Prabang Xayaburi Pak Lay Sanakham Pakchom Ban Koum Latsua Don Sahong Stung Treng Sambor The proportion of the total volume of the reservoir between the surface and -2m or -5m characterizes the shape of the reservoir, and subsequently its fish productivity since the latter is concentrated in the water closest to the surface, with deeper waters being less productive (Bernacsek, 1997). According to Bernacsek (1997), assessing production potential requires data about annual affluent flow volume in each reservoir. This information was derived, for each main scenario, from annual discharge volumes in dry and wet seasons in the upstream hydrological station nearest of each reservoir. 178 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

179 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Table 18: Annual effluent flow volume in each reservoir (million cubic meters) Reservoir Upstream station Baseline Definite future 2030 Pak Beng Chiang Saen Louang Prabang Louang Prabang Xayaburi Louang Prabang Pak Lay Louang Prabang Sanakham Louang Prabang Pakchom Chiang Khan Ban Koum Mukdahan Latsua Pakse Don Sahong Pakse Stung Treng Pakse Sambor Kratie The yields obtainable from reservoir fisheries vary considerably between reservoirs and depend on size and on multiple other factors. Small and shallow reservoirs are considerably more productive than large and deep ones (Bernacsek 1997). Productivity of reservoirs in southeast Asia ranges between a maximum yield of 200 kg/ha/year (China, Vietnam) and a few kilograms per hectare per year in Thailand, Indonesia or in Malaysia (Bernacsek 1997, Jackson and Marmulla 2001). Even with regular stocking, yields from large reservoirs in Thailand have been consistently below 200 kg/ha/year (Amornsakchai et al. 2000). In Pak Mun the reservoir fish production was expected to amount to 220 kg ha -1 but it actually reached only about 10 kg.ha -1 (Amornsakchai et al. 2000,. Jutagate et al. 2001) In India, large- and medium-size reservoirs have been found unsuitable for self-sustaining production, and depend on continuous stocking (Bernacsek 1997a; Jackson and Marmulla 2001). Overall, the large uncertainty throughout Asia in yields achievable in newly created reservoirs seriously hampers the credibility of reservoir fish production predictions (Bernacsek 1997). Generally speaking, during the first ten years after impoundment, fish in reservoirs benefit from a high primary production and catches are very high, but this period is followed by a progressive then sharp decline (review in Baran et al. in press). In addition to biological issues, varying degrees of reservoir management (Jutagate et al. 2006) and socio-economic issues (access rights, availability of fingerlings, market competition, etc) are often another major reason behind the failure of reservoir fish production systems (De Silva and Funge-Smith 2005). For instance, South America has a long experience of dam development and results from stocking in large reservoirs in South America have been meager or null (review in Quirós, 1999). Nam Ngum in Laos seems to be an exceptional case: the reservoir is still productive (13.8 kg/ha/year, around 600 tonnes/year, Bernacsek 1997) after three decades of exploitation. The reasons behind this sustained productivity are not well understood. Other reservoir fisheries in the region have on the contrary been quite disappointing, and there is a risk that a success story such as Nam Ngum will give a false impression that reservoir fisheries are a very productive option likely to compensate for the loss of capture fish resources ESTIMATES OF PRODUCTIVITY BASED ON SURFACE AREA, DEPTH AND FLOW Bernacsek, in his extensive review of more than 26 large dam fisheries in the Lower Mekong commissioned by the MRC in 1997, showed that the potential productivity of a reservoir was best predicted by its surface area, its depth, and the water inflow in that reservoir. 179 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

180 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S C = 1.877A -12D I where C = catch (tonnes/year), A = surface area (km 2 ), D = depth (m) and I = affluent flow (cmc/year) Productivities calculated in this manner for mainstream dam reservoirs, using information summarized in Tables 9 and 10, ranged from 390 to >>1000 kg/ha/year, which are very high values inconsistent with those previously observed (Hortle, 2007). This may be due to the fact that the dataset used to derive the formula included dams characterized by flows much lower than those of mainstream dams. This result indicates that due to their large annual flow volumes, mainstream dam reservoirs appear to fall outside the linear range of this formula ESTIMATES OF PRODUCTIVITY BASED ON SURFACE AREA ALONE. By default, the surface area alone is the best single predictor of reservoir productivity. The total production of each mainstream reservoir was estimated using surface areas (Table 9) and three productivity levels: low (20 kg/ha/year), medium (50 kg/ha/year), and high (200 kg/ha/year). This large range reflects experience throughout Asia (Sricharoendham et al., 2000; Mattson et al., 2000; Jutagate et al., 2001) and underscores the largely unpredictable nature of reservoir fisheries detailed above. The corresponding estimates for mainstream dam reservoirs, weighed by the shape of the reservoir (deep reservoirs having a low productivity by hectare) are presented in Table 12. Conclusions: At best the maximum fish production to be expected from reservoir fisheries amounts to 30,000 tonnes basinwide. In fact when the shape of reservoirs is taken into account, the most likely production represents about 10,000 tonnes for the 1500 km 2 of reservoir area created by mainstream dams. This value should be compared with to the 700, million tonnes of capture fish production at risk of mainstream dam development (section 5.2.5). Table 19: Predicted production range in mainstream dam reservoirs. The most likely production, based on reservoir depth and shape, is highlighted in yellow. The most likely scenario is based on the assumption that reservoirs having more than 50% of their volume comprised between 0 and 2m are very productive (200 kg/ha/year), whereas reservoirs have a medium productivity (50 kg/ha/year) if 20-50% of their volume is comprised between 0 and 2m, and they are poorly productive (20 kg/ha/year) if less than 20% of their volume lies within the 0-2m layer. Reservoir Area (km2) Average depth (m) % of total volume at depth,2m Low productivity scenario (tonnes/year) Medium productivity scenario (tonnes/year) High productivity scenario (tonnes/year) Most likely scenario (tonnes/year) Pak Beng Louang Prabang Xayaburi Pak Lay Sanakham Pakchom Ban Koum Latsua Don Sahong Stung Treng Sambor Total I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

181 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S 8.9 SENSITIVITY ANALYSIS OF DAM GROUPINGS UPSTREAM CLUSTER OF DAMS CHANGES IN HYDROLOGY Upstream of Vientiane, in 2015 as well as in 2030, discharge will be 40 to 60% higher in the dry season than in However, most of the change is due to Chinese dams and the additional changes due to Lower Mekong mainstream dams are limited. In the wet season water level will be around 50 cm lower than compared to the baselie (again, no major difference between 2015 and 2030). So from a seasonal perspective, mainstream dams in the upstream cluster will not have a major impact on the hydrology of that area already driven by Chinese and tributary dams. However the daily variability in water levels created by several of these dams operation in peak mode (see table 6) might be very substantial, but could not be documented here due to lack of data. EXPECTED LOSSESS WITHOUT MAINSTREAM DAMS This section of the river is characterized by typical stream species living in riffles and fast flowing waters (Balitoridae, Cobitidae, etc; Baseline Assessment section 1.4). In case of the Definite Future scenario (absence of LMB mainstream dams), the increased dry season discharge due to Chinese dams is expected to alter riffles and shallow habitats used by species for breeding and as nurseries. The construction of 17 dams on tributaries in the upstream cluster (-46,000 km 2, see Annex 1) will also reduce fish habitat. As a consequence, a drop in the recruitment of local species is expected even in absence of mainstream dams in the upstream cluster. The contribution of these upstream species to the fish biodiversity of the basin is very important: with 93 species, Balitoridae represent the second most species-rich family in the Mekong, after Cyprinidae. EXPECTED LOSSESS WITH MAINSTREAM DAMS: The extensive South American experience of extensive damming in large river basins summarised by Quiros (2004) indicates that the loss in capture fish production amounts to at least 50% 55. Based on this experience we make below the assumption that losses in capture fisheries could reach 50% in case 6 upstream mainstream dams only are built, 60% if 9 mainstream dams are built (from Pak Beng down to Don Sahong) and 70% if all 11 mainstream dams are built. POSSIBLE FISH BIODIVERSITY LOSSES: LMB mainstream dams in the upstream cluster will have two main consequences: barrier effect for migrant fish and modification of riverine habitats by creation of reservoirs. The 6 upstream dams would create 403 km 2 or 715 linear kilometers of reservoirs, i.e. slow lacustrine, largely deoxygenated and stratified water in lieu of the former running waters (the two deepest reservoirs of the basin, Luang Prabang and Sanakham -average depth of 13m and 54m respectively, see Table 13-, would be located in this zone). 55 Quiros indicates that after a decade the fish production drops well below 50% of its original state 181 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

182 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Table 20: Length of reservoir created in the upstream cluster in relation to river length Dam Reservoir length (km) Pak Beng 180 Luang Prabang 150 Xayaburi 100 Pak Lay 110 Xanakham 90 Pak Chom 85 Total length of reservoirs (km) River Length (km) % of mainstream turned into a reservoir According to existing species records (MDF 2003, Dubeau 2004), the zone between Vientiane and the Chinese border includes 189 fish species. Out of these, an analysis of the MRC Mekong Fish Database records shows that at least forty-one species are not found elsewhere in the LMB (Annex 5). Mainstream dams will trigger a significant change in habitat, but stream species will also be subject to the impact of the other dams on tributaries. One species (Acheilognathus deignani) provides an extreme example of the outlook for some stream fishes in this zone: this species is found in the mainstream and in the Nam Ou only; regardless of plans in the mainstream, in the Nam Ou Basin there are plans for 11 other dams (projects Nam Ou 1 to 7, to be operational between 2013 and 2015, with a height comprised between 47 and 147 m, plus Nam Nga, Nam Phak, Nam Ngao and Nam Pok projects, planned after 2017, 5-69 m high). Table 21: Catch of migrant white fish vulnerable to mainstream dam development Total catch (estimate based on fish consumption studies) Yield at risk under the assumption of 35% of vulnerable species Yield at risk under the assumption of 70% of vulnerable species Cambodia Lao PDR Thailand Viet Nam Total 481, , , ,118 2,062, ,500 58, , , , , , , ,500 1,443,500 Conclusions: In the upstream migration zone biodiversity is clearly at risk. Following the construction of 6 mainstream dams in this area, 90% of the river stretch between the Chinese border and Vientiane would be turned into a reservoir. At least 41 species are threatened by a severe alteration of their habitat. By comparison this number corresponds for instance to about half the total freshwater fish fauna of the United Kingdom (99 species). The family most exposed would be Balitoridae (river loaches), with about 10% of its 93 Mekong species at risk. The iconic, endemic and critically endangered Mekong Giant catfish would also be much at risk of total extinction since its main breeding area is located in this area, near Chiang Saen. However since 17 other dams blocking access to 46,000 km 2 upstream of tributaries are also planned in the area, habitat alteration and impact on fish biodiversity are not specifically due to mainstream dams. There is no information as to whether any of the 41 species specifically threatened can survive in reservoirs. 182 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

183 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S This estimate is very close to that given by Barlow et al. in 2008 ( million tonnes at risk). Our review integrates two methods used in Barlow et al. ( ) but not the more approximate third one (it is the latter method that provided the upper range of Barlow s estimate, i.e million tonnes of fish at risk). On the contrary the present review integrates two more approaches based on field data, reflects recent findings about the importance of black fish in catches, is explicit about the amount of white fish and black fish in each method, and integrates error ranges in both percentages of long distance migrant fish and total catch. POSSIBLE FISH PRODUCTION LOSSES (2030, 6 MAINSTREAM DAMS IN THE UPPER LMB): In this upstream cluster zone, the local fish production amounts to 40,000 60,000 tonnes per year (Baseline Assessment, section 2.4.2), which represents 2-3% of the most likely fish production basinwide (i.e. 2.1 million tonnes, Baseline Assessment section 2.1.3) or 3-5% of the fish production along the mainstream (Baseline Assessment, Table 28). At the local level, in absence of other mainstream dams and based on the assumptions detailed under section (i.e. 50% losses in capture fisheries), the fish losses in catches of the local fishery would amount to of 20,000 to 30,000 tonnes of capture fish. At the Basin level, in absence of other mainstream dams, but in presence of the other tributary dams planned by 2030, 31% of the Basin will be specifically obstructed by the dams of the upstream cluster (scenario S4 in Table 3). However the 27 fish species will still find habitat for breeding and feeding in the ( ) = 31.3% of the basin not obstructed by any dams. The analysis detailed in Section 5 shows that at least 27 species of long-distance migrants representing about 18% of catches basinwide, i.e. around 380,000 tonnes, use this section of the river. When a correction factor is introduced to reflect the importance of black fish in catches basinwide and the real share of long-distance migrants vulnerable to the barrier effect of dams, i.e % of the overall stock (see Section 5.2.5), the conclusion is that the risk of fish production losses in case the 6 upstream mainstream dams are built amounts to 130, ,000 tonnes 57 POSSIBLE FISH PRODUCTION GAINS DUE TO DAMS: The mainstream dams of the upstream cluster will create 403 km 2 of reservoir. This area can be expected to produce between 800 and 8,000 tonnes of reservoir fish, the most likely estimate being 3,000 tonnes. This production will be supplemented by that from dam reservoirs on tributaries, but there is no information about the cumulative surface area of these projects on tributaries, so their potential reservoir production cannot be precisely assessed. As a crude alternative estimate, one can note that the surface area of exisiting reservoirs in northern Laos and Thailand ranges between 10 and 80 km 2 (Bernacsek 1997); if we take 40km 2 as an average, this corresponds to around 700 km 2 for the 17 dams planned, and depending on the level of productivity, the corresponding reservoir fish production would range between 1,400 and 14,000 tonnes, which, added to the production of mainstream reservoirs, would represent between 2,000 and 20,000 tonnes of fish, the most likely estimate being around 7,000 tonnes of reservoir fish per year in this zone. Conclusions: 56 in the case of method 1 in Barlow et al. s paper, two mistakes had to be corrected: i) the estimate of LMB total in Hortle (2007) is actually 2,062,000 tonnes, not 1,860,000 tonnes, and ii) a typo in the result of 1,860,000 x 38.5% 57 [380,000 tonnes x 35%] - [380,000 tonnes x 70%] 183 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

184 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S In the upstream cluster zone, the fish production represents 5% maximum of the Mekong fish production. In this area covering 123,700 km 2 there might be 23 hydropower projects by , which means that the river network would be largely obstructed by dams, and that the local habitat of the 189 local fish species will change drastically. Following dam development a loss of fish production and of fish biodiversity is to be expected. At the local level, the loss of capture fish (20,000 42,000 tonnes) and the additional production of reservoirs (most probably around 7,000 tonnes) would result in a net loss of 13,000 to 35,000 tonnes of fish supply. This would be a very substantial risk for food securlity, corresponding to 15% to a third of the whole annual meat production of the country 59. At the Basin level, the construction of mainstream dams in the upstream cluster would reduce fish production in the whole Mekong Basin by 130, ,000 tonnes MIDDLE CLUSTER OF DAMS CHANGES IN HYDROLOGY In 2015, according to the BDP2, between Vientiane and Pakse, discharge will be one third higher in the dry season than in These forecasts are roughly in line with those of BDP1 and the WorldBank, but not with those of the Nam Theun 2 project that predicts more than doubling of the flow in the dry season. BDP2 forecasts are almost identical for the other scenarios (2030 with 0, 6, 9 and 11 mainstream dams), with just a few percents variation in wet season discharge compared to the situation forecasted in 2015, i.e. dry season discharge increased by a third compared to During the monsoon season, according to BDP2, water level will be 30 cm lower in 2015, although there is a discrepancy with other studies that forecast between -40cm to -1.6 m. For the 2030 scenarios wet season water level is expected to decrease by 40 to 50 cm maximum, whatever the number of dams on the mainstream. Like in the upstream cluster, the very substantial daily variability in flows to be expected from Latsua dam planned to operate in peak mode (2m daily variability in its reservoir level) could not be detailed nor analysed by lack of data. EXPECTED LOSSESS WITHOUT MAINSTREAM DAMS Given the absence of taxonomic records in the mainstream between the two dam project and Vientiane, it is difficult to identify species that are dependent on the mainstream in this section of the river. 58 Nam Beng (30 MW; 2014), Nam Dong (1 MW; 1970), Nam Ko (2 MW; 1996), Nam Long (5 MW; 2013), Nam Nga (98 MW; 2017), Nam Ngay (1 MW; 2002), Nam Ou 1 (180 MW; 2013), Nam Ou 2 (90 MW; 2014), Nam Ou 3 (300 MW; 2013), Nam Ou 4 (75 MW; 2014), Nam Ou 5 (108 MW; 2013), Nam Ou 6 (210 MW; 2014), Nam Ou 7 (180 MW; 2015), Nam Pha (147 MW; 2016), Nam Suang 1 (40 MW; 2016), Nam Suang 2 (134 MW; 2016), Nam Tha 1 (168 MW; 2013), plus the 6 mainstream projects I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

185 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S In absence of mainstream dams, in ,000 km 2 (23.6% of the LMB) would be barred by dams on tributaries anyway, and in ,000 km 2 of watershed, i.e. 37.3% of the LMB, would be obstructed (see Annex 1). This would represent a loss of 4 and 17% of connectivity compared to the baseline situation in This loss implies a loss of access to breeding areas upstream of tributaries, and a subsequent loss of productivity to be expected even without mainstream dams. A diversity of other factors detailed in the Baseline Assessment (section 6.2 and Table 43) and unrelated to LMB mainstream dams (e.g.; loss of connectivity in floodplains, increased fishing pressure, etc) will contribute to fish catch reduction. However insufficient knowledge about resilience of fish resources and lack of accuracy in multidimensional forecasting make it impossible to predict the extent of these losses. Another degree of uncertainty is introduced by the management of the Pak Mun dam (6 Km upstream of the confluence with the Mekong River): until 2001 dam gates were permanently closed, then open until 2002, then open 4 months a year until 2007, then they were permanently closed again. The Pak Mun dam controls access to the Mun/Chi sub-basin (120,000 km 2 ) and blocks migrations between the mainstream and this basin, and its significant impact on fish production has been reviewed in Amornsakchai et al. (2000). According to this study, the closure of the Mun River impacted in particular 17 migratory fish species, including 15 catfish species, whose migration routes were blocked, and whose spawning ground habitats upstream were modified. Jutagate et al. (2001) also found that after Pak Mun dam closure, only 96 fish species remained out of the previous 265 species. EXPECTED LOSSESS WITH MAINSTREAM DAMS: POSSIBLE FISH BIODIVERSITY LOSSES: With the 2 dams of the middle cluster, 165 kilometers of river, i.e. 23% of the mainstream between Pakse and Vientiane, would be turned into a reservoir. This would have a definite impact on biodiversity, although the magnitude of this impact could not be quantified here. Table 22: Length of reservoir created in the middle cluster in relation to river length Reservoir length Dam (km) Ban Koum 155 Lat Sua 10 Total length of reservoirs (km) River length (km) % of mainstream turned into a reservoir The Latsua dam, although located only 34 km below the Ban Kum dam, would have much more negative impact on fish migrations and production than the latter because it would block access to the Mun/Chi system (70,000 km2). With 270 species (Baseline Assessment, section 1.4), biodiversity in the Mun/Chi system is very high and this basin is subject to intensive fish migrations. The Latsua dam would have the same impact than the Pak Mun dam on Mun-dependent fish species, plus additional impact on species migrating up the mainstream, and the construction design does not mention sluice gates. Given the absence of taxonomic records in the mainstream between the two dam project and Vientiane, it is difficult to identify species that are dependent on the mainstream in this section of the river. However some species identified thanks to the analysis detailed in Section 5.1 are specifically at risk (Table 15). If Ban Kum dam is built, 28 of the long distance migrant species have an alternative in the 3S system (except if Lower Sesan 4 is built) and in the Mun/Chi system (except if the Pak Mun dam is closed). Four species do not migrate towards the 3S system: Boesemania microlepis, Cyclocheilichthys enoplos, 185 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

186 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Pangasius polyuranodon, Pangasius sanitwongsei; they represent at least 1.7% of the fish production. If Latsua or Stung Treng dams are built, 26 of these species have an alternative in the 3S system (except if Lower Sesan 4 or the 20 other dams considered in these 3 watersheds are built). Table 23: Some particular fish species related to the Mun / Chi system Species closely associated with the Mun / Mekong confluence Species related to the Mun system Family Scientific Name Information available Found only in large rivers of the Middle Mekong Basin. Most common along the Thai-Lao border at the mouth of the Mun River; it also used to occur in Mun and Songkhram Rivers, but is Cyprinidae Aaptosyax grypus now an extremely rare species An uncommon fish in the Mekong. Occurs just upstream from Cyclocheilichthys Khone Falls at the mouth of the Mun River. Also recorded from Cyprinidae heteronema the Great Lake. Cyprinidae Pangasiidae Pangasiidae Cyprinidae Cyprinidae Pangasiidae Pangasiidae Mystacoleucus chilopterus Helicophagus waandersii Pangasius kunyit Cyclocheilichthys furcatus Mekongina erythrospila Pangasianodon gigas Pangasius conchophilus Distribution: Mekong Basin in Laos and Thailand (mouth of Nam Mun) and possibly in Yunnan. Recorded from the Mae Kok at Chiang Rai and Mae Poon In the Mun River, the species migrates upstream from the beginning of the rainy season to the end of August and move back downstream from late September to November. Distribution: found basin wide in the mainstream of the Mekong. This species is impacted by the Pak Mun Dam, because it blocks their migration route and the fish pass is not working properly, and the rapid habitats where the fish used to spawn have been inundated by the head pond. Distribution: the northern boundary for this species is at Chiang Saen, however the main distribution range is from Nakhon Phanom to Kandal Province in Cambodia. Now ery rare in the Mun River Known from the Middle Mekong along the Thai-Lao border to the Tonle Sap. Recorded from the confluence between the Mun and the Mekong Rivers and from the Mekong at Ban Tha Kai 21 kilometres downstream from Mukdaharn. In the Mun River, the species migrates upstream from the beginning of the rainy season to the end of August and move back downstream from late September to November. Occurs from Chiang Saen to Pak Lay, however from Chiang Khan to Paksan the species was not reported. Occurs again at Thakhek and downstream to Sambor. Also recorded from Xe Bangfai, Nam Theun, and Chi Rivers. May spawn in the delta or mouth of the Mekong; other spawning grounds have been identified in the mainstream of the Mekong River near Chiang Rai, and in the Mun River at Ubon Ratchathani. An important spawning ground appears to be in the Mekong mainstream somewhere between Kompong Cham and Khone Falls and in rapids and riffles of the Mun river. Distribution: the distribution range is from the Mekong Delta all the way along the Mekong to Chiang Saen. Source: MFD 2003 POSSIBLE FISH PRODUCTION LOSSES (2030, 9 DAMS, NO CAMBODIAN MAINSTREAM DAMS): Under this scenario 78% of the Basin would not be accessible to migrant fish coming from downstream floodplains, and 41% of the Basin would be specifically obstructed by mainstream dams (Annex 1). Overall, 41 species are known to migrate though Khone Fall on their way to the upstream part of the LMB (Annex 4). These species represent 28.5% of the overall catch (2.1 million tonnes), i.e. about 186 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

187 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S 600,000 tonnes. When a correction factor is introduced to reflect the importance of black fish in catches basinwide and the real share of long-distance migrants vulnerable to the barrier effect of dams, i.e. 35 to 70% of the overall stock (see Section 5.2.5), the conclusion is that the risk of capture fish production losses in case the 2 mainstream dams of the middle cluster are built amounts to 210, ,000 tonnes 60. This approach leads to an estimate much lower than that detailed in Section 5.2.5; thatissue is detailed in Section POSSIBLE FISH PRODUCTION GAINS DUE TO DAMS: The Ban Koum and Latsua mainstream dams would create 147 km 2 of reservoir. This area can be expected to produce between 300 and 3,000 tonnes of reservoir fish, the most likely estimate being 330 tonnes. This production will be supplemented by that from the 77 dam reservoirs on tributaries, but there is no information about the cumulative surface area of these projects on tributaries, so their potential reservoir production cannot be precisely assessed. As a crude alternative estimate, one can note that the average surface area of the 36 LMB reservoirs listed in Bernacsek (1997) amounts to 143 km 2. For 77 dams, the cumulated surface area would be 11,024 km 2. Depending on the level of productivity, the corresponding reservoir fish production would range between 22,000 and 220,000 tonnes, the most likely estimate being 55,000 tonnes. Conclusions: The Latsua dam would have much more negative impact on fish migrations and production than the Ban Kum dam because if would block access to the Mun/Chi system. It would then reiterate the impacts of the Pak Mun dam on the Mun River, in addition to those on the mainstream. Under this scenario 78% of the Basin would not be accessible to migrant fish coming from downstream floodplains. The risk of capture fish production losses in case the 2 mainstream dams of the middle cluster are built amounts to 210, ,000 tonnes (probably an underestimate). However the overall production or fish in reservoirs basinwide can be expected to reach 55,000 tonnes. This would result in a net balance of 155, ,000 tonnes of fish production lost in case the Ban Koum and/or Latsua dams are built DOWNSTREAM CLUSTER OF DAMS CHANGES IN HYDROLOGY In 2015, according to the BDP2, in the downstream migration zone of the LMB, from Pakse down to the sea, the dry season discharge would be 13 to 22% higher than in Like in other comparisons, these results are similar to those of the WorldBank and BDP1 forecasts, but much lower than the Nam Theun 2 predictions (+125%). In the wet season in 2015, average water level would be around 30 cm lower than in In terms o f discharge or water level, BDP2 forecasts are almost identical for the other scenarios (2030 with 0, 6, 9 and 11 mainstream dams), with just a few percents variation compared to the situation forecasted in [600,000 tonnes x 35%] - [600,000 tonnes x 70%] 187 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

188 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S In terms of changes in floodplains, when compared to 2000, BDP2 predicts a loss of 250,000 ha of floodplains in the whole Basin by 2015 and a loss of 309,000 ha of floodplains by 2030 (-5 and -7% respectively). It is not clear whether this forecast integrates the notion of double ring (detailed in section 3) due to the increase of 30cm of water level in the dry season. Additional forecasts include, for the Tonle Sap area, a reduction of flood duration by 2 weeks to 1 month and a reduction of sediment inflow in the system of at least 8 to 13%. EXPECTED LOSSESS WITHOUT MAINSTREAM DAMS The factors that apply to the downstream cluster of dams are the same as those detailed in section 7.2.2: - in absence of mainstream dams, in ,000 km 2 (23.6% of the LMB) would be barred by tributary dams; in ,000 km 2 i.e. 37.3% of the LMB, would also be obstructed by dams on tributaries (see Annex 1). This loss of 4 and 17% of connectivity compared to the baseline situation in 2000 implies a loss of natural fish productivity even without mainstream dams. - multiple factors detailed and unrelated to LMB mainstream dams (e.g.; loss of connectivity in floodplains, increased fishing pressure, agricultural development in the Tonle Sap and subsequent pesticide inputs, reduced sediment inflow due to 77 tributary dams modifying productivity of the estuarine and costal zone, etc) will contribute to fish catch reduction. However the complexity and interactions between these issues make prediction of losses in capture production losses impossible. The fact that 250,000 ha of floodplains will be lost by 2015 is a third main factor influencing fish production even in absence of mainstream dams. This loss corresponds to a loss of capture fish production comprised between 13,000 and 50,000 tonnes 61 (see Section 4). EXPECTED LOSSESS WITH MAINSTREAM DAMS: POSSIBLE FISH BIODIVERSITY LOSSES: The Baseline Assessment (section 1.4) has shown that 204 species are found in the mainstream at the level of Kratie, and 168 species are found below of Khone Falls. Out of these, the analysis of migration patterns described in the Mekong Fish Database (Annex 4) has identified 43 species undertaking long-distance migrations through the mainstream. It is clear that a number of additional species would be impacted by the transformation, between Kratie and Pakse, of 42% of the mainstream into reservoirs, but the exact number of species at risk could not be specified during this study. Table 24: Length of reservoir created in the downstream cluster in relation to river length Dam Reservoir length (km) Don Sahong 5 Stung Treng 45 Sambor 90 Total length of reservoirs (km) River length (km) % of mainstream turned into a reservoir [250,000 x 50 kg/ha/y] [250,000 x 200 kg/ha/y] 188 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

189 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S Overall, the analysis of tables 13, 14 and 16 shows that if 11 reservoirs are built, 55% of the mainstream will be turned into a dam reservoir. If Cambodian dams are not built, only 48% of the mainstream would be turned into a lake. Table 25: Linear length of reservoirs and total length of the mainstream Total length of reservoirs (km) 1020 River length (km) 1838 % of mainstream turned into a reservoir 55 POSSIBLE FISH PRODUCTION LOSSES (2030, 11 MAINSTREAM DAMS): following this scenario 81.3% of the Basin river network and fish migration routes would be obstructed by dams, 44% of this obstruction being due specifically to mainstream dams Overall, 43 species are known to migrate between downstream floodplains where they feed and upstream habitats where they breed (Annex 4). These species represent 29.8% of the overall catch, i.e. about 630,000 tonnes. When a correction factor is introduced to reflect the importance of black fish in catches basinwide and the real share of long-distance migrants vulnerable to the barrier effect of dams, i.e. 35 to 70% of the overall stock (see Section 5.2.5), the conclusion is that the risk of fish production losses in case the 2 mainstream dams of the downstream cluster are built amounts to 220, ,000 tonnes 62. This approach leads to an estimate much lower than that detailed in Section In the latter the specific impact of a subset of dams is not detailed, but the overall conclusion is that if 81% of the LMB are obstructed by dams (cf. Annex 1), then 700,000 to 1.4 million tonnes would be at risk. The difference between these two assessments can be explained by the fact that the latter refers to all dams of the LMB (scenario 2030, 11 mainstram dams) whereas the former refers oly to the specific impact of the dams of the middle cluster. Another factor is the loss of floodplain area. This loss involves all species using the floodplains, and not only migrant white fish. Therefore this loss is to be added to the above loss. BDP2 estimates that 309, ,000 = 58,000 ha of floodplain would be lost in case of mainsream dam development (the 251,000 ha being lost even in absence of mainstream dams; see Table 5). This loss of floodplains corresponds (see Section 4) to 3,000 to 12,000 tonnes of fish 63. POSSIBLE FISH PRODUCTION GAINS DUE TO DAMS: The Stung Treng and Sambor mainstream dams would create 950 km 2 of reservoir. This area can be expected to produce between 2000 and 19,000 tonnes of reservoir fish, the most likely estimate being 47,000 tonnes. This production will be supplemented by that from the 77 dam reservoirs on tributaries, but there is no information about the cumulative surface area of these projects on tributaries, so their potential reservoir production cannot be precisely assessed. Like in section 7.2.4, one can note that the average surface area of the 36 LMB reservoirs listed in Bernacsek (1997) amounts to 143 km 2. For 77 dams, the cumulated surface area would be 11,024 km 2. Depending on the level of productivity, the corresponding reservoir fish production would range between 22,000 and 220,000 tonnes, the most likely estimate being 55,000 tonnes, to be supplemented by the 2,000 19,000 tonnes from the two dams of the downstream cluster. 62 [630,000 tonnes x 35%] - [630,000 tonnes x 70%] 63 [58,000 ha x 50 kg.ha/year = 2900 tonnes] [58,000 ha x 200 kg.ha/year = 11,600 tonnes] 189 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

190 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T F I S H E R I E S If all mainstream dams are taken into account, then the overall reservoir fish production would be: from tributary dams: 22,000 55, ,000 tonnes (see above) from mainstream dams: 3,000 8,000 30,000 tonnes resulting in a range comprised between 25,000 and 250,000 tonnes, the most likely production being 63,000 tonnes. Conclusions: Due to the presence by 2030 of 77 tributary dams in the basin, resulting in 37% of the fish migration routes obstructed and at least 250,000 ha of floodplains lost, losses in capture fish production will be high even in absence of mainstream dams; however they cannot be quantified. In presence of mainstream dams in the downstream cluster (between Pakse and Kratie), 55% of the length of the mainstream would be turned into a reservoir. This would have profound impacts on biodiversity, in particular on 43 species having to undertake long-distance migrations and making up to 220,000 to 440,000 tonnes of fish in the downstream migration zone of the Lower Mekong Basin. An alternative estimate amounts the possible loss for the overall basin to 700,000 to 1.4 million tonnes. An additional 3,000 to 12,000 tonnes of capture fish losses would follow the reduced surface area of floodplains resulting form the constuction of mainstream dams. These losses could be partly compensated by the production of reservoir fish. The total amount of reservoir fish produce would range between 25,000 and 250,000 tonnes, the most likely scenario being 63,000 tonnes. 190 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

191 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S 9 SOCIAL SYSTEMS 9.1 SUMMARY OF PAST & FUTURE TRENDS WITHOUT LMB MAINSTREAM HYDROPOWER Major land cover transformations have already occurred in all countries in many parts of the Lower Mekong Basin (LMB) over the past 30 years or so. The nature and extent of these transformations are both natural and man-made. Combined pressure on Mekong communities has been observed from many different directions, whether these pressures are policy, environmental or local unsustainable resource use practices. Natural factors including flooding, drought, and increased saline intrusion (particularly in the Mekong delta). Man-made factors include changes in agricultural practices, urban growth, land conversion (including infrastructure development, poor fishing practices, urban and agricultural waste discharges into the river, riverbank erosion from soil removal from wetlands for biofertiliser, or sand and gravel from river banks and delta), and destruction of forests and mangroves. Such activities to date have had both positive and negative impacts on the socio-economy and future food security of LMB populations, with associated consequences for key issues relating to Mekong mainstream dams identified under the Social Component theme POVERTY REDUCTION Past and current trends in the LMB demonstrate progress towards meeting MDG poverty alleviation targets on some indicators but regression on others. Progress is variable from one LMB country to another, depending on how national strategies are developed and applied. For example, the target to cut the under-5 mortality rate by 2/3rds between 1990 and 2015 is likely to be met, while targets related to environmental sustainability will not only not be met but become worse, particularly in relation to loss of forest cover, water supply and sanitation 64. As there is high livelihood dependence in all LMB countries on water and land resources - the highest in Laos, the lowest in Thailand the current rapid depletion rate of the natural resource base remains a serious concern for future efforts towards poverty alleviation. The normal livelihood base is one which diversifies both seasonally and occupationally. Increasingly, options for traditional diversification are becoming more limited. The most vulnerable are those with limited coping mechanisms and low occupational or income source diversity. The shift to alternative livelihoods less rural and less dependent on natural resources takes time, with some households adjusting more successfully than others. Thailand has demonstrated greater movement than other countries in this respect, but it has been achieved gradually over many years and more markedly in some provinces than in others. National revenues from hydropower in the region are increasing, but their association with poverty alleviation remains to be clarified. Only if this association can be better explained by national hydropower strategies can a certain confidence level be gained that the Mekong mainstream dams will assist LMB countries to reach their MDG targets, rather than setting back the timing of this achievement by many years. 64 "Development Asia: Special Report, Millenium Development Goals", Asian Development Bank, Year II, No. V, October-December I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

192 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S HEALTH & NUTRITION The status of these issues in the LMB is directly associated with (i) ease of access to adequate health infrastructure and personnel; (ii) drainage and clean water resource management with its associated health and sanitation consequences; (iii) knowledge and awareness levels, which may be associated with relative vulnerability to food insecurity; and (iv) access to free sources of high nutritional value from natural resources, such as fish, non-timber forest products, and wild game. Stunting and wasting are characteristics of malnutrition more common in Lao PDR and Cambodia than in Thailand and Vietnam, affecting both life expectancy as well as children's health each important measures of food security and quality of life, and significantly affecting a country's ability to be economically productive. Public expenditure on health in all LMB countries is uneven, and while Thailand has removed clean water supply and sanitation from its MDG targets (having achieved this by 2007), remaining LMB countries retain the target and have some way to go before achieving it. Some health and nutrition issues can be addressed by improved financial resource allocation, but others are associated with ease of access to the natural resource base RESETTLEMENT & LAND ACQUISITION Numerous gaps remain in land acquisition and compensation policy and procedure compared to international best practice for all LMB countries. Lack of consistent national or transboundary mitigation frameworks present challenges to achieving policy equity in project implementation, while limited human capacity and/or political will to effectively monitor developers and require them to satisfactorily meet policy commitments, remain obstacles to socially equitable resettlement practice. Other land expropriation practices in the LMB such as village resettlement for administrative consolidation, land conversion and urban growth, coupled with unsustainable land and river usage, are already causing communities to lose much of the natural resource base on which their livelihoods depend. This introduces a potential "double jeopardy" context whereby communities affected by one impact may become even more vulnerable when affected by another, and at higher risk of impoverishment from Mekong mainstream dams impacts. All LMB countries agree that the social and political economy of resettlement in relation to hydropower is one of the most serious strategic transboundary issues facing hydropower development in the area. 9.2 OPPORTUNITIES AND RISKS All 12 Mekong mainstream dams will present certain opportunities and create certain risks. The extent of how opportunities might be maximised to the widest benefit of LMB communities, particularly riparian communities, and how risks might be managed, particularly for affected people, depends on whether each LMB country feels adequately informed about key issues and has sufficient information as well as experience on which to base its decisions. Each LMB country must feel confident that its internal (at national, provincial and district levels), as well as its transboundary, policy, regulatory and administrative frameworks, are robust enough to manage known opportunities and risks, as well as the many currently unknown or unknowable ones. 192 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

193 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S During the impact assessment workshop, the four national working sessions concluded that opportunities and risks ranged in importance, depending both on the type of issue the working groups had previously prioritised, and depending on respective national perspectives (Table 1). Only Lao PDR, the country with the largest number of proposed mainstream dams, saw any positive opportunities for most of the Social Theme priorities. Thailand identified some opportunities for positive changes in health and nutrition status. All countries, without exception, considered all Social Theme priorities to range from very severe to severe risk. Table 1: SEA Impacts Assessment Workshop Outputs: National Working Session Priorities THEME ISSUE LAO PDR Hydrology and sediment Changes in patterns of maximum water levels, rates of rise and predictability Changes in sediment transport and deposition Changes in nutrient transport Terrestrial Habitat loss and degradation ecosystems and Changes in Land use agriculture Changes in irrigated agriculture Changes in River bank gardens Aquatic Change in productivity of aquatic ecosystems habitats Changes in populations of rare and endangered species Fisheries Social systems Changes in water quality Changes in long distance migration Changes in fish species biodiversity Changes in fish production Changes in poverty and natural resource based livelihoods Changes in health and nutrition Social effects of resettlement, land acquisition and loss of access Changes in cultural values and patterns Economics Contributions to national economy - Export earning Contributions to national economy - Foreign Direct Investment Contributions to local economies (district and community level Energy and Achieving energy security Power Meeting national energy demands Climate change Meeting local energy needs Relative emissions of green-house Gas Direct impacts of climate change on hydropower projects - extreme events & dam security Combined effect of climate change and mainstream dams on food security CAMBODIA THAILANDVIETNAM Uncertainty No clear relationship Not relevant Not relevant Not relevant 193 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

194 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Opportunities and risks related to dam construction were presented by the SEA team at the Impact Assessment Workshop (IAW) in May 2010, and apply to all 6 hydro-ecological zones identified in the Lower Mekong Basin (LMB) by the Mekong River Commission (MRC). Based on IAW feedback, perspectives on social opportunities and risks were updated and are outlined in Table 2. Table 2: Opportunities and Risks for Key Social Theme Topics Topic 1: Poverty Reduction and the Natural Resource Base in the Lower Mekong Basin Opportunities Risks Improved public infrastructure (schools, Loss of land and river livelihood resource base health) without sufficient replacement/compensation Improved all-weather road access Market access improved with roads & bridges Replacement land of equal size and productivity increasingly unavailable Revenue benefits directed towards poverty Some socio-economics are irreversible and alleviation (guarantees need validating) Enhanced project-related work and service provision opportunities (guarantees needed) Improved household access to electricity (guarantees needed) irreplaceable (e.g. fisheries) Psycho-social adjustment difficulties, particularly for ethnic minorities Ethnic minorities already poor and vulnerable, already more dependent on the natural resource base for livelihoods, and therefore at greater risk of greater impoverishment Increased land values and associated exclusion of poorer communities Revenue benefits not equitably shared Developer has little or no commitment to social and environmental mitigation and may not provide direct poverty alleviation benefits for affected communities Pace and intensity of economic development happening faster than local and national capacities to deal with it Topic 2: Health and Nutrition Good health programmes can lead to significant reduction of chronic complaints (e.g. parasitic infections) and avert major outbreaks of disease Commitment by developer to improve capacity of local health staff & village health workers, and to develop IEC on different health issues Infrastructure improvements in water supply, wastewater disposal and sanitation Improved knowledge and awareness about clean water use and sanitation Road construction facilitating access to health services Improved infrastructure providing better access to goods, services and markets Improved access to facilities, e.g. hospitals, clinics, schools, electricity supply, all-weather Elevated groundwater levels leading to waterlogging & higher risk of vector disease transmission (malaria, dengue, filariasis) if drainage and sanitation programmes not implemented Reduced access to free wild foods (forests, fisheries, wetlands) limiting available sources of high nutritional intake Risk to life and property, as well as limiting access, from sudden changes in water flows Unanticipated flooding and poor preparedness leading to loss of life, property and assets Poorly managed health interaction with construction workers leading to higher incidence of STDs/HIV/AIDS transmission to local communities Topic 3: Resettlement, Land Acquisition, Accessibility Loss of land, assets, homes, livelihoods, without adequate restoration Benefits grabbing by strong groups and vested interests vulnerable groups squeezed out 194 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

195 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S roads Developers investing in project areas providing financial inputs to localised poverty reduction Double-jeopardy for vulnerable households previously affected by land sequestration Loss of spiritually and culturally significant locations (spirit forests, cultural sites) Limited commitment by developer and failure to provide adequate and timely funds leading to greater adverse impacts than necessary 9.3 STRATEGIC AND INSTITUTIONAL CONCERNS REGARDING SOCIAL IMPACTS Overarching strategic and institutional issues need to be described first. These issues will influence the efficacy of risk management with respect to the 12 dams. One important finding of the SEA is that many possible impacts on the Mekong river as a result of the proposed dams are, as yet, not well understood because there has not been adequate research and information in several key areas. Knowledge gaps are particularly apparent with respect to sediment loss and navigation, as well as to cultural, social and livelihood profiles of riparian communities in different impact zones. Without reliable information about where sediment loss will occur and what its consequences could be, the number of people affected, where they will be affected, and the type and scope of impacts, cannot at this time be assessed (see Section 1.6.3). A second finding is that the proposed 12 dams present the sort of situation that no-one has the experience of dealing with before, either in the region, or in the world. All LMB countries have reported that no clear and well-understood national or transboundary framework exists which is currently capable of dealing with the issues arising from the dams, nor to address potential conflicts which they might generate. There is no overall strategy for selecting the hydropower projects for development. The private sector is currently deciding which Project is to be done on a first-comefirst-served basis and there is no proper consideration of how projects should be sequenced (see Sections and 1.7) Several of the same districts and provinces may experience multiple impacts at different times from the same, or from different developers - for example, Pak Lay district will be affected by construction activities, by headpond filling, and by downstream impacts of Pak Lay dam, as well as by downstream operations of Xayaboury dam. Not only will this result in difficulties assigning impacts to a particular project, but because all dams will be constructed over a period of 20 years or so, some riparian districts can expect multiple and repeated impacts with constant sets of disruptions over this period. If one developer attributes an impact to another developer, the dispute about who should pay for what may rumble on for years without resolution, leaving affected communities stuck in the middle with no redress (see Section 1.6.2). Provincial and district authorities of each country, who will bear the brunt of dealing with the day-to-day issues related to each dam, in many instances have no experience of implementing national social and environmental safeguard policies, let alone dealing with transboundary issues. Even if agreements on individual projects are reached at national and international levels, translating these into actions on the ground require very substantial capacity development (see Section 1.7). Very high potential for social conflict exists over the impacts of the 12 dams, both within countries, and between different countries (see Sections ). 195 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

196 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Lastly, different developers have different levels of commitment to provide adequate funding for social and environmental safeguards. Common safeguards standards should be developed by the LMB countries, and mandatory due diligence conducted by the MRC on each developer against these agreed standards before approval of any proposed development (see Sections and 1.7). 9.4 IMPACTS OF PROPOSED LMB MAINSTREAM HYDRO PROJECTS LOCATION OF IMPACTS The Mekong River Commission (MRC) 65 classifies a 15km wide corridor on either side of the Mekong River as the highest impact area for hydropower on the mainstream. Six different hydro-ecological zones were identified by the MRC, and the SEA team asked to describe impacts according to each zone. These impact zones are outlined in Table 3 and shown on Map 1, which also reflects zonal poverty rates identified by the MRC. During the IAW, feedback indicated a preference for impacts to be described by country instead of zones. Vietnam questioned the relevance of a 15km corridor concept for its flood delta, because the extent of inundation is not well defined. Additionally, the delta has an extensive canal network delivering fresh water to the majority of the delta area for both agriculture and domestic purposes. Thailand also questioned the relevance of hydro-ecological zones, particularly as they have no relation to how each LMB country approaches its policies, legislative frameworks and implementation methodologies. While two countries on either side of the Mekong river may share very similar ecological, hydrological and even cultural constituents, the way in which poverty alleviation, health, nutrition, land acquisition and resettlement, are dealt with in the respective countries may be quite different, with consequences for impacts assessment. Nonetheless, there remains some relevance for the Social Theme, particularly in relation to cultural identity. Different zones support different ethnic groups and by association, different ways of life which are expressed through cultural identity. Forced changes to ways of life will consequently affect local interpretations of identity. The MRC's zones therefore present an initial useful framework in which potential impacts on culture and ethnicity may be understood. However, an effort has also been made to sub-divide zonal impacts by country, for easier reading by different countries. Table 3: Hydro-ecological zoning of the Lower Mekong Basin and location of proposed Mekong mainstream hydropower projects Zone Location Proposed Hydropower Project Total No. Dams 1 China Chiang Saen Chinese dams 3 operational, 1 in construction, 4 proposed 2 Chiang Saen - Vientiane Pak Beng, Louang Prabang, Xayaboury, Pak Lay, Xanakham, Pak Chom 6 3 Vientiane - Pakse Ban Koum, Latsua 2 4 Pakse - Kratie Don Sahong, Thakho, 65 Mekong River Commission, "Integrated Basin Flow Management Progress Report (IBFM)", 2007, and in the MRC Technical Paper No. 30, Social Impact Monitoring and Vulnerability Analysis (SIMVA) in the mekong Corridor: Report on a Regional Pilot Study, March I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

197 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Stung Treng, Sambor 4 5 Kratie Phnom Penh & Tonle Sap 0 6 Phnom Penh Mekong delta & sea I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

198 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Map 1: Hydro-ecological Zones in the Lower Mekong Basin, and National Poverty Rates Source: Mekong River Commission, "Integrated Basin Flow Management Progress Report (IBFM)", I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

199 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S THE DIFFERENCE BETWEEN DIRECT AND INDIRECT IMPACTS, AND HOW MANY PEOPLE ARE AFFECTED BY EACH It is important to analytically distinguish between direct and indirect impacts. Direct impacts are defined in this document as experienced by those losing land, property, assets, income, livelihoods, and access to other communities or to livelihood resources. It covers the whole range of impacts that might be experienced by communities living and working in the impact zones of a hydropower project's construction, impoundment and operational area. Even if people are not affected by physical relocation, they can be affected by loss of critical assets, such as land, which affects their livelihoods, they can be affected by loss of access to a facility or resource, they can also be affected by the consequences of a dam which might lead to incomers coming into the area, land-grabbing, or just being priced out of an area because provision of improved facilities can lead to rising land and property values. Many direct impacts will be the same for all people by every dam, irrespective of location. However, the intensity of impacts may be experienced differently at different parts of the river system. All dams will require land acquisition for construction sites, associated facilities (such as access roads, contractors' camps, etc.), and headponds, which will result in asset loss and/or relocation of households. Specific asset loss may be permanent, temporary, or seasonal. Current estimates of total number of relocation impacts caused by all 12 dams are: 131 villages, 6,847 households, and 63,112 people. Preliminary overall estimates of total people directly affected is 106, This figure is based on the assumption that a developer will opt for the least-impact technical design. Table 4 summarises the numbers of directly affected people by the 12 dams. These figures are considered preliminary and conservative in the absence of more definite information. To date the main international focus of Mekong mainstream dams impacts has been on water resources and fisheries. These indeed will be severe. But the impacts on land loss and usage will also be very severe, permanent, and irreversible, and this fact should not be lost in the broader discussion. Loss of traditional land usage results not just in loss of a livelihood asset, but can lead to permanent changes in culture and identity, affecting as it does different communities' ways of life which are defined and articulated to others through the way in which they interact with and depend on, that land. Indirect impacts are sometimes more difficult to define and to estimate in scope. Indirect impacts can include positive aspects, such as new project related employment opportunities. Or negative aspects, such as pollution created by contractor s camps or workers, loss of access, cumulative loss of naturalresource based livelihoods, and health risks associated with worker's camps or groundwater saturation. Or it can include unanticipated impacts before or after operations. Distinguishing between direct and indirect impacts is important, as it frequently defines who is entitled to compensation and support and who is not. However, people affected by indirect impacts are also entitled to compensation and mitigation measures, though the responsibility for undertaking this is often either disregarded, or presents too large a scope for stakeholders to agree how to assign it. This is particularly relevant to indirect impacts of the proposed 12 Mekong mainstream dams, where transboundary and cumulative consequences raise issues of accountability that none of the LMB countries have previously had to address. 66 More discussion on numbers of people directly and indirectly affected by the 12 dams is included by Zone in Section It should be noted that all estimates of affected persons are preliminary and probably conservative at this stage because some developers did not respond to the SEA team on queries of numbers of affected people, some EIAs/SIAs were under way or not available to the team, and no Resettlement Plans for any project were available. 199 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

200 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Table 4: Preliminary Totals of People Directly Affected by the Mekong Mainstream Dams No. Dam Name Total Affected Villages Total Affected HHs Total Affected Persons Number of Resettled Villages Number Resettled HHs of Number of Resettled Persons 1 Pakbeng (1) 57 6,831 35, ,700 2 Louang Prabang (2) 36 2,516 12, ,966 3 Xayaboury (3) 29 1,988 4, ,130 4 Pak Lay (4) 27 1,079 19, NA 6,129 5 Sanakham (2) , ,000 6 Pak Chom (2) Ban Koum (2) Lat Sua (2) 0 NA NA NA NA NA 9 Don Sahong (2) Thakho (2) Stung Treng (2) 21 2,059 10, ,059 10, Sambor (2) NA NA NA Preliminary Totals Data Sources: NA=Not Available. * indicates figures from 1994 study by Compagnie Nationale du Rhone, Acres International Ltd. & Mekong Secretariat Study team. No updated information available to SEA 1. Data from Initial Environmental Examination (IEE), Pak Beng Hydropower Project, Lao PDR, December 2008, Earthsystems, Norconsult & SEA Inception Report, Vol. 2, Project Profiles 2. SEA Inception Report, Vol. 2, Project Profiles 3. Final Report, Social Impact Assessment of Xayabouri Hydroelectric Power Project, Lao PDR, August 2008, Team Consulting Engineering & Management Co. Ltd., Ch.Karnchang Public Company Ltd. & SEA Inception 4. Initial Environmental Examination (IEE), Pak Lay Hydropower Project, Lao PDR, June 2008, Earthsystems, Norconsult, CEIEC & Sinohydro Joint Venture. Figures taken are for the maximum impacts downstream option. Indirect impacts are likely to affect those living or working within access (i.e. 15kms) of the Mekong mainstream, its tributaries and wetlands, but who are not expected to lose land or housing. The total population within this corridor in Laos, Thailand and Cambodia is estimated to be 29.6 million people 67. Of these, local riparian communities are normally the most exposed to indirect impacts, namely district populations within a 5km reach of the Mekong mainstream. This amounts to more than 2 million people at highest risk in the 47 districts in the immediate headpond, construction and downstream of the 12 dams. The LMB population at risk of indirect impacts in the Vietnamese delta totals 14 million people 67 Executive Summary, p. xvii, and Table 4, Zone Populations, MRC Technical Paper No. 30, SIMVA, March 2010, op cit 200 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

201 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S (13,849,801), with the highest proportion of its population in the LMB 68. The Vietnamese team in the IAW rejected the concept of a 15km impact corridor as relevant for the delta area. Table 5: Populations of Directly Affected Districts No. Source: SEA Baseline Study, Social Component Dam Name No. Affected Provinces No. Affected Districts Total Affected District Population 1 Pakbeng ,659 2 Louang Prabang ,204 3 Xayaboury ,198 4 Pak Lay ,544 5 Sanakham ,974 6 Pak Chom ,189 7 Ban Koum ,140 8 Lat Sua ,160 9 Don Sahong , Thakho NA NA NA 11 Stung Treng , Sambor ,610 Totals ,094,749 Table 6: Zone Populations Zone Total Urban Rural %Rural Zone 2 1,351, ,599 1,036,636 77% Zone 3 4,013, ,800 3,087,194 77% Zone 4 900, , ,715 87% Zone 5 9,574,444 1,495,389 8,079,055 86% Zone 6 13,849,801 1,816,225 12,033,576 87% Total 29,689,795 4,664,619 25,025,176 Source: MRC Technical Paper No. 30, SIMVA, March 2010, Table 5 Additional indirect impacts will be felt on fisheries, where a full understanding of total numbers of people affected has yet to be reached. The current estimate provided by MRC's BDP is that at least 1 million fisheries-dependent people will be affected in Cambodia alone. As indirect impacts become further distanced from an actual project site, so their relationship to the cause becomes more difficult to identify. For example, riverbank erosion is currently a normal feature of seasonal changes to the flows of the Mekong river and its tributaries. The extent of directly attributable erosion impacts to Mekong mainstream dams will be difficult to determine without reliable and regular monitoring procedures. 68 Ibid 201 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

202 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Determining the extent of attributable indirect impacts to Mekong mainstream dams becomes more challenging when taking pre-existing conditions into consideration, particularly for downstream countries such as Vietnam. For example, climate change will increase the volume of water reaching the LMB delta with a resulting reduction in saline intrusion. Yet the incremental effects of sea level rise, increased incidents of storms and tidal influences, will act to increase saline intrusion, though without modelling this scenario, the extent is as yet unknown. Mainstream dam projects will not directly affect average annual salinity in the delta through flow changes, but they may affect the timing of saline intrusion during the dry season. Direct impacts are relatively simple to determine, but because the interaction of variables is so complex, indirect impacts are much more difficult to conclude. Furthermore, cumulative impacts may take some time to make themselves known, such as erosion in the Vietnamese delta and consequences for agriculturally-dependent households. In the absence of reliable data, many indirect impacts cannot yet be fully appreciated. Knowledge gaps relating to the Social Component are identified in Section Numbers of indirectly affected people also depends on risk management by developers and line agencies of the LMB countries. Poor management of dams and erratic water releases will increase the number of affected people. For example, if there are operating failures on Latsua or Ban Koum, the 76,368 population of Pakse is at high risk. Numbers also depend on individual developers' willingness to implement remedial or mitigation measures in a timely fashion. For example, if health, drainage and sanitation programmes are not implemented adequately, there is a proportionate risk of higher numbers of indirectly affected people from spread of vector disease DIRECT AND INDIRECT IMPACTS As Zone 1 (China to Chiang Saen) was not included in the SEA evaluation, no direct impacts can be presented. Nonetheless, there are some lessons learned with respect to the three operational Chinese dams of Manwan, Dachaoshan, and Jinghong. These lessons have particular relevance for dams in Zone 2 and will be reflected in this section. Zone 2: Chiang Saen to Vientiane, incorporating 6 dams of Pak Beng, Louang Prabang, Xayaboury, Pak Lay, Xanakham and Pak Chom. The 6 dams in this Zone will affect 10 provinces and 32 districts in Thailand and Lao PDR. Zonal population totals just over 1.3 million people (1,351,35), of which 77% is rural 69. The majority of directly affected population is Lao, many of whom are ethnic minorities living below the poverty line and highly dependent on the natural resource base. No figures for directly affected people in Thailand are available at this time. This cascade of 6 dams will directly affect the largest number of people of all Zones, totalling an estimated 76,290 people. Zones 1 and 2 are characterised by particularly high poverty incidence, high proportions of ethnic minorities in relation to overall district populations, levels of stunting and wasting in children above national averages, large proportion of poor districts in Laos, and high levels of food insecurity. Districts have very limited health and educational facilities, and all season road access is very poor. Land is steep, paddy land areas are scarce, and communities are heavily dependent for their livelihoods on upland swidden cultivation and on the natural resource base, including rivers, streams and wetlands. 69 All Zonal population figures in this section are taken from MRC Technical Paper No. 30, SIMVA, March 2010, Table I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

203 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S As the land flattens out towards the Vientiane plain, the ethnic composition of the population moves from a preponderance of ethnic minorities to more Lao/Thai groups. The population increases and depends increasingly on fixed agriculture. Seasonally cultivated riverside gardens become more important to local livelihoods. Health and educational facilities are more accessible and numerous, and there are higher levels of literacy in the district populations. Vientiane province has the lowest national poverty rate in Laos. Table 7: Direct and Indirect Impacts in Thailand and Lao PDR, Zone 2 Thailand Lao PDR Direct Impacts from Pak Beng and Pak Chom Direct Impacts from all 6 dams i. Affecting 3 provinces and 11 districts i. Affecting 7 provinces and 21 districts ii. Numbers of directly affected people in Thailand ii. Affecting 161 villages, 13,321 households, not available 76,290 people i. Potential for permanent loss of home and iii. Some dams affecting high proportions of non-agricultural land already very poor district populations ii. Permanent loss of agricultural land & iv. Permanent loss of home and non-agricultural productive trees land iii. Potential for permanent loss of community v. Permanent loss of agricultural land & resources (e.g. schools, temples) productive trees iv. Permanent loss of access to forest resources vi. Permanent loss of community resources (e.g. v. Permanent loss of accessible grazing land schools, temples) vi. Loss of natural-resource based livelihoods vii. Temporary and permanent loss of ethnic vii. Seasonal loss of land (riverbank gardens) viii. Seasonal loss of fisheries minority common property resources (e.g. grazing land, NTFPs) ix. Improved irrigation opportunities from Pak viii. Loss of natural-resource based livelihoods Chom, with resultant potential for higher ix. Seasonal loss of land (riverbank gardens) productivity and improved agricultural incomes x. Seasonal loss of fisheries xi. Permanent loss of river-based alternative livelihoods (e.g. loss of aquatic products due to increased turbidity such as river weed (kai) which is both nutritionally and commercially valuable; loss of access for gold panning, sand and gravel extraction at Pak Lay etc.) xii. Risk of "double jeopardy", relocation and asset loss for communities already affected by prior relocation xiii. Risk of operational erosion leading to additional unanticipated relocation Indirect impacts i. Loss of cultural festivals such as the Giant i. Loss of trans-mekong river access Mekong Catfish festival ii. Permanent reduction of seasonal fisheries ii. Loss of trans-mekong river access xiv. Loss of basis for way of life having iii. Permanent reduction of seasonal fisheries subsequent consequences for cultural identity iv. Benefits from improved infrastructure access iii. Benefits from improved infrastructure access In proportion to overall population figures, some dams will have a more severe direct impact on rural communities than others. Where dams will have an impact on districts with higher incidences of poverty, the proportional impact will be more severe than in relatively prosperous districts. Pakbeng has the highest potential to adversely affect the poor, especially in Lao PDR. This dam will affect 8 Lao districts, of which 7 are officially classified as poor or very poor. Pakbeng will also directly affect a much higher 203 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

204 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S proportion of district population than any other dam in Zone 2, at 15.8%. This compares to Louang Prabang, which will affect 8.1% of total affected district population, Xayaboury a lower impact at 2.1% of district population, Pak Lay 6.7%, Xanakham 2.5% and Pak Chom an estimated 0.1%. One lesson learned from the experiences of the Chinese dams, particularly the Manwan 70, has implications for dams in the upper part of the Zone 2 cascade. Slope erosion in the Manwan dam area with similar steeply sided mountainous valleys to Zone 2, has already resulted in unanticipated direct impacts of landslides and mudflows, plus frequent torrential flooding, requiring additional relocation of 1,000 people at extra cost, and leaving a further 8 townships under threat. Part of the reason for erosion has been attributed to customary lifestyles of upland ethnic minorities, who have moved further up hill slopes rather than sever their attachment to their customary land and livelihood methods, and have cleared land on steeper elevations for swidden cultivation, further contributing to erosion risks. But this is not the only identified cause; development of associated facilities such as roads and irrigation works have destabilised the fragile hill slopes, as has increased river flows. A combination of cultural and technical factors have played their parts in creating an additional impact burden for provinces and people alike. Many riparian districts in Zone 2 have a high percentage of their land above a 16% slope. All Lao riparian districts in this Zone, with the exception of Kenthao and Vientiane, have between 26%-50% of fragile land slope. Five riparian districts falling into impact areas (Paktha, Pakbeng, Nga, Hongsa, Nan and Chomphet) have an even higher percentage (51%-75%) of steep land as a percentage of total district land areas 71. Additional risk of erosion and unanticipated livelihood loss and relocation for Zone 2 dams could therefore be high, particularly as flatter, cultivable land is extremely scarce, what there is will either be lost to the dams or is already cultivated by other groups, and finding replacement land of similar productivity will be extremely difficult. This could push developers to resort to cash compensation. This solution is certainly an easier option for developers where land is routinely undervalued for compensation purposes, but it will not prevent swidden cultivators from continuing traditional livelihood lifestyles. Instead it will push the ethnic groups dependent on such lifestyles into clearing more marginalised and poorer land, leading both to greater erosion risks for the hydropower projects as well as greater impoverishment for affected communities. Furthermore, the notion of cash compensation for common property usage is extremely difficult to define in cash terms, and certainly an alien concept for upland ethnic groups. The SEA baseline outlined the spread of different groups in Zone 2. In the upper and steeper part of the Zone, the Thai side of the Mekong river consists primarily of Thai population, though with small groups of Leu, Lahu, Hmong, Khmu and Lao. On the Lao side, districts in the impact areas have a much higher proportion of ethnic minorities, particularly Khmu, Leu and Hmong. On both sides of the river, these ethnic groups have distinct lifestyles, more heavily dependent on natural resources and the rhythm of the rivers and wetlands, with strong spiritual attachments to certain areas, particularly forests. Communities 70 He Daming, Zhao Wenjuan, Chen Lihui, "The Ecological Changes in Manwan Reservoir Area and its Causes", Proceedings abstract, International Conference on Advances in Integrated Mekong River Management, th October 2004, RR2 Research Group & Mekong River Commission: Yu Xiaogang, Jia Jiguo, "An Overview of participatory Impact Assessment for Manwan Hydropower Station in Lancang River", no date 71 Map 8, Districts with Fragile Lands, "District Vulnerability Assessment: Lao PDR", United Nations, World Food Programme, February I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

205 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S have less sense of individual land and property ownership, and view the natural resources on which their livelihoods depend as common to all and free in their benefits. Efforts to change these lifestyles, particularly prevention of rotational swidden agriculture, have already been underway for many years in Laos. This has been effected through compulsory relocation programmes targeted at ethnic minorities. Some ethnic minority communities have already been relocated once or twice in the preceding 10 years (e.g. a Hmong village in Pak Beng impact zone), and are already among the most disadvantaged in terms of poverty levels and poor social conditions. Households in Ban Houay Xong, Nan district, potentially affected by the Xayaboury hydropower project, were moved from the uplands to the lowlands in the mid- 1990's but placed into an area which frequently flooded, and after 7 years were obliged to relocate themselves twice with no outside assistance to try and re-establish their village and livelihoods again. Land acquisition will result in this village being displaced for the fourth time in 15 years. Repeated compulsory relocation within a relatively short period of time is one of the most impoverishing acts that can occur. The SEA terms it "double jeopardy", where communities have to re-establish their social, cultural and economic lives not just once, but repeatedly. An estimated total 72 of 8,418.5has of agricultural land and 6,523has of forests including spirit forest, will be lost in Zone 2. Cultural artefacts, such as cemeteries and temples will also be lost. Although the Pak Ou caves will not be inundated, it is thought that access to this popular tourist site will be more restricted, with associated impacts on tourism livelihoods. With flow changes to the river, there will be an earlier loss of riverbank garden cultivation, which is an increasingly important part of the domestic economy further down the Mekong in Zone 2 as the land flattens out towards the Vientiane plain. Impacts on fisheries may result in loss of cultural events associated with the Mekong river's life, such as the Giant Mekong Catfish festival in Chiang Khong, which is dependent on the survival of the species. Positive impacts could include improved opportunities for irrigation on the Thai side from the Pak Chom dam, with higher productivity potential leading to improved agricultural incomes. Other positive impacts could include improved trans-mekong trasnsport if the dams are designed to provide publicly available road access. Failing this, the dams could have the opposite effect of limiting trans-mekong access with increased velocity of flows preventing smaller river craft from crossing or using the river. Zone 3: Vientiane to Pakse, incorporating 2 dams of Ban Koum and Latsua. Affecting Thailand and Lao PDR. Zonal population is over 4 million people (4,013,994), 77% of which is rural. Zone 3 is characterised by selective patches of poverty Thailand has some of the poorest districts in the Kingdom, and Laos also has some very poor districts. However, in both countries, poverty incidence reduces as proximity to the Mekong river increases. Champassack province has the lowest poverty rate of southern Lao provinces. Riverbank land in this Zone is highly productive and intensively cultivated. A significant impact will be seasonal loss of riverbank gardens due to increased flow regimes with associated impacts on livelihoods. Riparian land in this Zone is among the most expensive and productive in both countries. Finding replacement land of equal productivity will not be such a constraint as in Zone 2, but it will be difficult to find it in such proximity to relocation sites that may satisfy affected households. The two dams will 72 These figures are minimum totals as of this report's date. All land acquisition data is drawn from project-specific IEEs or from SEA team questionnaires to developers. Two developers did not complete the estimate for land acquisition. Estimates of acquired land provided by developers only relate to those directly affected by relocation, and do not include land acquisition for associated facilities such as access roads, transmission lines, etc. Total land loss may thus be higher. 205 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

206 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S acquire 1,667.6has of agricultural land, of which 332has is irrigated. Ban Koum is not estimated to acquire forest land, but Lat Sua will acquire 4kms² of forest. Transportation is extremely important for this stretch of the Mekong river. Many small craft owners earn their living from navigating within national river systems, as well as across the Mekong itself. It is likely that related livelihoods will be severely adversely affected by changes in flow which will make it increasingly difficult for small craft to navigate safely. As limited data was available for Lat Sua, it is not possible to estimate the scope of impacts on riparian district populations. Ban Koum is estimated to directly affect 0.6% of district populations. Communities are almost entirely of Lao and Thai-Lao ethnicity. No significant impacts are expected on cultural or historic sites in Zone 3, though riverside temples and sacred trees are at risk from increased erosion. As the river flattens out and flows towards Pakse, riparian districts are traditionally prone to regular seasonal floods. These can be mitigated through operational control of Zone 3 dams. However, operational risks also exist, depending on management competence. Backwater in event of failure of floodgate opening could result in flooding in Pakse with consequent loss of life, property and assets. The extent and rapidity with which gates may be opened at Ban Koum will also affect both livelihoods and safety in Pakse. The population at risk in Pakse is 76,368, with a population density of 611 persons per square km. Dependence on fisheries in Zone 3 for commercial, as well as subsistence purposes, is much higher than in Zone 2. Champassack is the top producing province in Lao PDR for both fish culture (8,300 tonnes) and fish capture (5,600 tonnes) 73. Crop production, both irrigated and rain fed, is combined with livestock rearing, fisheries, gathering of NTFPs, and aquatic products. Due to its proximity to Thailand, remittances play an important part in the Champassack domestic economy. Table 8: Direct and Indirect Impacts in Thailand and Lao PDR, Zone 3 Thailand Direct Impacts i. Affecting 1 province and 4 districts, some of the poorest in Thailand, but poverty levels lower immediately adjacent to the Mekong river ii. 1 village affected and 29 households. Numbers of directly affected people in Thailand not available, but the following direct impacts can be anticipated: iii. Permanent loss of home and non-agricultural land iv. Permanent loss of agricultural land & productive trees v. Permanent loss of community resources (e.g. riverbank temples) vi. Permanent loss of accessible grazing land vii. Loss of natural-resource based livelihoods Lao PDR i. Affecting 2 provinces and 5 districts, some of the poorest in Laos, but poverty levels lower immediately adjacent to the Mekong river ii. Affecting 3 villages, 158 households, 935 people 74 iii. Permanent loss of home and non-agricultural land iv. Permanent loss of agricultural land & productive trees v. Loss of irrigation pump stations vi. Permanent loss of community resources (e.g. riverbank temples) vii. Loss of natural-resource based livelihoods viii. Seasonal loss of land (riverbank gardens) ix. Seasonal loss of fisheries 73 MRC Technical Paper No. 30, SIMVA, March 2010, op cit 74 Figures of people not affected by relocation but directly affected unavailable for Lat Sua. Ban Koum figures for affected people include both Thai and Lao totals. 206 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

207 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S viii. Seasonal loss of land (riverbank gardens) x. Poor operational management leading to high ix. Seasonal loss of fisheries risk of unanticipated flooding impacts with x. Poor operational management leading to high consequent loss of land, property, assets and lives risk of unanticipated flooding impacts with xi. Riverbank cultural assets threatened (e.g. consequent loss of land, property, assets and lives temples, sacred trees) xi. Riverbank cultural assets threatened (e.g. xii. Improved irrigation opportunities with higher temples, sacred trees) productivity potential leading to improved xii. Improved irrigation opportunities with higher agricultural incomes productivity potential leading to improved agricultural incomes Indirect Impacts i. Elevated groundwater levels could benefit i. Elevated groundwater levels could benefit domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease ii. Loss of livelihoods for small river craft owners ii. Loss of livelihoods for small river craft owners iii. Reduction of fisheries and aquatic products for iii. Reduction of fisheries and aquatic products for subsistence and livelihoods subsistence and livelihoods Zone 4: Pakse to Kratie, incorporating 4 dams of Don Sahong and Thakho in Lao PDR, and Stung Treng and Sambor in Cambodia. Zonal population is less than 1 million people (900,321), 87% of which is rural. Three of the four dams in Zone 4 (Don Sahong, Stung Treng, Sambor), will have the most detrimental indirect impacts of all the 12 mainstream dams, not only within Zone 4, but cumulatively and indirectly downstream in Zones 5 and 6, reaching to the Mekong delta. Indeed, the further downstream the dams are sited, the higher and more extensive the adverse impacts. The 3 largest dams (Don Sahong, Stung Treng, Sambor) will directly affect 3 provinces and 8 districts. Thakho is reported as not affecting anyone, either by relocation or acquisition of land and assets 75. Don Sahong will affect a small proportion of affected district populations at 0.02%, but almost all of these districts are classified as first and second priority poorest in Laos. Stung Treng at 17.5% and Sambor at 13.1% will have the highest direct impacts on the largest percentage of affected district populations than any of the other 12 dams, with the exception of Pak Beng. This is particularly worrying for Cambodia, as these two dams will have a proportionately higher impact on the poor in the two provinces of Stung Treng and Kratie, both of which have the highest poverty rates in the country at 46% each. In Stung Treng province, the site for the Mekong mainstream dam is at the confluence of the Mekong river with three major tributaries, the Sekong, Sesan and Sre Pok, which additionally have associated hydropower development proposed. The confluence of these rivers has created a vase wetland ecosystem rich in biodiversity, on which an estimated 90% of the provincial population is dependent 76. There are both permanent and temporary settlements, some seasonally established to take advantage of the annual fish migration, and some increasingly permanent and populated by landless people from other parts of the province. In Stung Treng, reports are that poor households have a higher dependence on fisheries than better-off households, with fisheries contributing more than 30% more of poor households' income than of better-off households. Impacts on this livelihood source will therefore have a 75 Thakho Hydropower Project, Rapid Initial Environmental Examination, August 2009, Worldwide Fund for Nature (WWF) 76 David Allen, William Darwall, Mark Dubois, Kong Kim Sreng, Alvin Lopez, Anna McIvor, Oliver Springate-Baginski, Thuon Try, "Integrating people in conservation planning: an integrated assessment of the biodiversity, livelihood and economic implications of the proposed special management zones in the Stung Treng Ramsar Site, Cambodia", IUCN Species Programme, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

208 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S disproportionate impact on the poor. One ethnic group in particular, the Cham (Muslim Khmer) is almost totally dependent on fisheries for their livelihoods, and as such, have developed a range of fishery skills and knowledge superior to other ethnic groups. They tend to be semi-nomadic, travelling to Stung Treng with the onset of the rainy season 77. The MRC estimates that in Cambodia alone, more than 1 million fisheries-dependent people could lose their livelihoods, including in the Tonle Sap where an estimated 14% of surveyed households defined their main occupation as fishing 78, but where the vast majority of its population derives secondary or associated livelihoods, as well as subsistence, from fisheries. Given that the number of poor has already been increased by prevailing land sequestration practice for commercial concessions, the risk of double jeopardy for both directly and indirectly affected people in Stung Treng and Kratie is extremely high. Stung Treng is reported to have the highest level of level of compulsory land sequestration for distribution to concessions holders. It is not over-inflating claims to say that both the Stung Treng and Sambor would create a situation of extreme crisis for the populations of affected provinces, and could provoke an emergency food security situation for the poor. These two dams have the highest potential to seriously worsen the incidence of poverty in Cambodia. All reports on LMB food security acknowledge that rice sufficiency, either through cultivating oneself or by having enough money from other sources to buy it, is a primary way in which communities define food security, which in turn depends on seasonal factors. The FAO notes that the main causes of malnutrition in Cambodia include low food intake and chronic illness, caused primarily by household food insecurity, and lack of clean water and sanitation 79. Lack of access to land in Cambodia is a critical factor contributing to poverty levels, not just in relation to individual ownership and use, but to commons also. The WFP also observes 80 that man-made factors have become increasingly important in presenting serious risks to sustainable economic growth and improved food security in Cambodia. These factors include: rapidly increasing population, decreasing access to land, weak institutional capacity, poor infrastructure, social exclusion and increasingly uneven distribution of wealth. These key points were also observed by stakeholders during SEA discussions for all countries, and were not confined to Cambodia alone. As land flattens out on the Mekong mainstream, other area-specific indirect impacts may be associated with Don Sahong, Stung Treng and Sambor. As the baseline indicated, health issues related to vector borne diseases, poor water management and sanitation, and higher levels of poverty-related symptoms such as wasting and stunting, are more evident in provinces and districts further downstream, particularly in southern Laos and Cambodia. In Laos, all Zone 4 riparian districts have at least 61%-80% of households with poor sanitation, while 3 districts (Pakse, Soukhouma and Mounlapamok) show an even worse situation with between 81%-90% with poor sanitation 81. Any changes in groundwater elevations and flows may result in increased waterlogging, with likely indirect impacts of increases in already serious water and sanitation-related health problems in these areas. 77 Thuon Try and Marcus Chambers, "Situation Analysis: Stung Treng Province, Cambodia", UNDP, IUCN, MRC GEF-funded programme, Mekong Wetlands Biodiversity Conservation and Sustainable Use Programme, MRC Technical Paper No. 30, SIMVA, March FAO, Nutrition Country Profiles, Cambodia, p. 16, "Integrated Food Security and Humanitarian Phase Classification (IPC), Pilot in Cambodia, Final Report", World Food Programme, April Map 9, Households with Inappropriate Sanitation, SEA Baseline, Social Component 208 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

209 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S An estimated total agricultural land to be acquired for the 3 main dams is 3,412.9has, the largest acquisition by far being for Sambor (3,369has). Substantial forest areas will also be lost, totalling 13,332.1has, again, the majority acquired by Sambor (13,143has). Tourism is an important livelihood contribution in the Stung Treng Ramsar site as well as at Khone Falls, and related livelihoods may be adversely affected by some tourism sectors who see the natural beauty of locations as well as interactions with the rare Mekong dolphin. However, other sectors may see enhanced tourism at dam sites. Table 9: Direct and Indirect Impacts in Lao PDR and Cambodia, Zone 4 Lao PDR Cambodia Direct Impacts i. Affecting 1 provinces and 4 districts i. Affecting 2 provinces and 4 districts ii. Affecting 4 villages, 14 households, 66 ii. Affecting 21 villages 83, 3,079 households, people 82 29,651 people iii. Permanent loss of home and non-agricultural iii. Permanent loss of home and non-agricultural land land iv. Permanent loss of agricultural land & iv. Permanent loss of agricultural land & productive trees productive trees v. Permanent loss of community resources (e.g. v. Permanent loss of community resources (e.g. riverbank temples) riverbank temples) vi. Permanent loss of accessible grazing land vi. Permanent loss of accessible grazing land vii. Loss of natural-resource based livelihoods viii. Seasonal loss of land (riverbank gardens) ix. Permanent and seasonal loss of fisheries i. Elevated groundwater levels could benefit domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease such as already endemic and increasing dengue and filariasis ii. Loss of livelihoods for small river craft owners iii. Reduction of fisheries and aquatic products for subsistence and livelihoods iv. Severe transboundary effects on Cambodia with blockage of migrating fish passages, decline in fisheries and resultant impacts on fisheriesdependent households livelihoods and domestic economy vii. Loss of natural-resource based livelihoods viii. Seasonal loss of land (riverbank gardens) ix. Permanent and seasonal loss of fisheries x. Disproportionate impacts on the poor who are more livelihood-dependent on fisheries in Stung Treng & Kratie xi. Double jeopardy for people affected by prior land sequestration Indirect Impacts i. Elevated groundwater levels could benefit domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease such as already endemic and increasing dengue and filariasis ii. Loss of livelihoods for small river craft owners iii. Reduction of fisheries and aquatic products for subsistence and livelihoods iv. Severe downstream effects with blockage of migrating fish passages, decline in fisheries and resultant impacts on fisheries-dependent households livelihoods and domestic economy v. Major impact on Tonle Sap fisheries 82 This is only a developer's estimate of the number of people relocated, not total people affected by loss of land and assets. 83 Numbers of villages affected not available for Sambor 209 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

210 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S v. Loss of tourism in Khone and Phapheng Falls livelihoods and related processing industries vi. Impacts on Tonle Sap floating homes with rapidly changing water flows vii. Loss of tourism opportunities in Stung Treng Ramsar site viii. High risk of major loss of life, livestock and assets without establishment of effective early flood warning system Zone 5: Kratie to Phnom Penh and including the Tonle Sap, no dams in this section, and no direct effects. All impacts in Zone 5 are indirect, but no less serious for that, and these increase further downstream. Very high poverty levels in Zone 5 provinces can be observed. The critical impacts for Zone 5 will be on fisheries and wetlands. These are not addressed in this part of the SEA, but are dealt with under the Fisheries and Aquatic components. The MRC estimates that at least $1 million people in Cambodia alone will have fisheries-dependent livelihoods affected Table 10: Direct and Indirect Impacts in Cambodia and Vietnam, Zone 5 None Cambodia Direct Impacts None Indirect Impacts i. Elevated groundwater levels could benefit domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease such as already endemic and increasing dengue and filariasis ii. Water intakes and pump stations needing relocation and/or adjustment to cope with changing water flows iii. Floating homes in Tonle Sap affected by water flow changes iv. Severe downstream effects with blockage of migrating fish passages, decline in fisheries and resultant impacts on fisheries-dependent households livelihoods and domestic economy v. Major impact on Tonle Sap fisheries livelihoods and related processing industries vi. Disproportionate impacts on ethnic minorities, particularly Cham and Cambodians of Vietnamese descent, highly dependent on fisheries vii. High risk to loss of life, livestock and assets in the absence of effective early flood warning systems viii. Seasonal livelihoods activities needing adjustments, e.g. Kratie flood arrival delayed c. 2 weeks, duration reduced by c. 1 week Vietnam i. Cumulative downstream impacts in Vietnam from altered river flows and reduction in sediments and nutrient loads reaching the delta, leading to incremental reduction of agricultural land productivity, rise in agricultural costs, reduced fisheries production Many of the same risks for enhanced poverty in Zone 5 exist as for Zone 4. Seasonal flooding already creates seasonal increases of water-related diseases, and further uncontrolled waterlogging due to elevated groundwater levels caused by upstream dams could contribute to the increase of vector 210 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

211 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S disease. Arsenic in groundwater is also a little-known phenomenon in Cambodia and Vietnam, as well as in southern provinces of Lao PDR. A risk assessment was carried out by the World Health Organisation (WHO) 84, which identified arsenic contamination of groundwater in the Vietnam Mekong river delta, as well as in approximately 1600 villages in 6 provinces in Mekong river floodplains. The report indicated uncertainties in the number of people potentially currently affected in Cambodia, as well as the degree of exposure. Laos was also affected but in a very small way compared to Cambodia and Vietnam, and exposure in Mekong river riparian areas has not been identified. The Mekong river changes rapidly in Cambodia and Vietnam in response to river stage, erosion or deposition of sediments. Groundwater arsenic contamination is strongly correlated to these features. Land use changes or changes in groundwater flows, can result in arsenic intrusion into areas where no prior contamination occurred. Changes in river water levels can cause a significant change in groundwater flow directions, and thus induce migration of contaminated groundwater in a new, formerly non-contaminated area. This could pose a major health threat to affected populations. Once again, rapid changes in water flows into and out of the Tonle Sap could affect those communities living in floating homes, and will require changes to timing and patterns of livelihood activities and access. In particular, fisheries-dependent households of ethnic minorities such as Cham and Cambodians of Vietnamese descent, will be more affected, as ethnicity here coincides with predominant livelihood opportunities for historic reasons 85. These minorities generally do not own land. However, even other minorities, as well as Cambodians owning land, will be affected due to prior sequestration of their land for concessions, forcing many to fall back on fisheries as an alternative livelihood source. Zone 6: Phnom Penh to the Mekong Delta and the sea, no dams in this section, and no direct effects. Indirect impacts may be observed, but it is likely that these will be cumulative and not fully emerge until several years into operations. Likely indirect impacts may be experienced in upstream locations of the Vietnamese delta through higher risk of erosion of channels and irrigation systems. This may result in higher costs of operation and maintenance, and river and waterways transport could be affected by channel instability with consequent livelihood impacts. Erosion from increased water flows will further result in reduced areas of agricultural land, which will have a major impact on the ability of Vietnam's most productive area to generate the amount of rice and agricultural products that it does at present. The mainstream projects will contribute to a reduction in nutrient loads reaching the delta, more fully described under the Hydrology & Sediment component of the SEA. Loss of suspended sediments and nutrients will lead to an incremental reduction in productivity of agricultural lands, which are regularly flooded during the wet season. However, many nutrients are already lost through embankment construction and flood protection, and therefore the full extent of nutrient loss may not be attributable to mainstream dams. 84 "Research needs for household level treatment to remove arsenic and fluoride in drinking water in S. E. Asia", David Fredericks, WHO, no date ( ) 85 MRC SIMVA March 2010, op cit, p I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

212 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Nutrient loss will also reduce productivity of both fresh water and marine aquatic systems, leading to reduced fisheries production for both Cambodia and the Vietnam delta. Reduced productivity will also have knock-on effects for the aquaculture sector, relying as it does on trash fish. Risks may therefore increase for the local population in facing livelihood vulnerability over the years. Without maps of areas potentially at risk and with the extent of inundation area as yet not well defined, it is not possible to provide further information of what communities, how many and where, may be affected. This is a knowledge gap that needs filling. Nonetheless, given that the delta is the most densely populated zone in the LMB, and where the proportion of underweight and stunted rural children is higher than the national average, communities here are threatened with increasing impoverishment from land fertility and productivity loss from land erosion & sediment loss. It is recommended that a monitoring baseline of these communities be established to understand the scope of potential future impacts. Table 11: Direct and Indirect Impacts in Cambodia and Vietnam, Zone 6 None Cambodia Direct Impacts None Indirect Impacts i. Elevated groundwater levels could benefit domestic water supply and more varied irrigation options, but high risk of waterlogging and increased vector disease such as already endemic and increasing dengue and filariasis ii. Water intakes and pump stations needing relocation and/or adjustment to cope with changing water flows iii. Severe downstream effects with blockage of migrating fish passages, decline in fisheries and resultant impacts on fisheries-dependent households livelihoods and domestic economy iv. High risk of major loss of life, livestock and assets without establishment of effective early flood warning system Vietnam i. Higher risk of erosion of water channels affecting agricultural productivity and water-way transport ii. Erosion leading to increased costs of channel maintenance and upkeep iii. Loss of annual nutrients through flooding leading to incremental agricultural productivity loss iv. Nutrient loss also affecting fresh water and marine fisheries, with reduced productivity and knock-on effects for aquaculture v. Losses in agriculture and fisheries leading to increased poverty and limited livelihood opportunities 9.5 EQUITY OF RISKS All LMB countries identified topics under the Social Component as being almost entirely negative. Cambodia, Vietnam and Thailand all ranked the dams as having very negative impacts on (1) Changes in poverty and natural resource based livelihoods, and negative or very negative impacts (2) Social effects of resettlement, land acquisition and loss of access. Laos was split between very negative and very positive impacts for these two topics. All LMB countries ranked dam impacts on (3) Changes in cultural values and 212 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

213 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S patterns, as negative or very negative. Again, all LMB countries ranked dam impacts on (4) Health & nutrition as either negative or very negative, but Laos and Thailand also saw some very positive opportunities for this topic. The conclusion to draw from national discussions is that only Laos sees any equity of benefits versus risks, whereas the other three LMB countries universally see mainly risk. The relative importance of each impact is difficult to quantify, with each sector and each country considering its own priorities and concerns as the most important, while viewing impacts of other sectors or countries not important enough to constitute an insuperable objection to construction of a particular project. Achieving equitable socio-economic tradeoffs may prove impossible. For example, does agriculture count as much as industry for a livelihood option? Do poverty risks to directly affected people outweigh national poverty alleviation opportunities through increased hydropower revenues? Are severe fisheries losses in one country with consequent livelihood impacts on a large number of people justified by national development aims of another country? Achieving equity between social costs and benefits is a difficult process, further complicated by competition between different drivers of progress and different interests. All indications are that the poorest communities will be those most adversely affected by the 12 Mekong mainstream dams. Those at highest risk are the poor and vulnerable, particularly from ethnic minority households directly affected by the 6 northern cluster of dams (Zone 2), most notably by Pak Beng and especially on the Lao side, as well as rural communities in the southern cluster of dams (Zone 4), particularly Stung Treng and Sambor, and especially in Cambodia's Stung Treng and Kratie provinces, and fisheries-dependent households in the Tonle Sap. The main reasons why poorest communities are at highest risk, and particularly in these two Zones, are: Zones 2 and 4 have the highest prior poverty levels with the largest number of ethnic minorities, who are already disproportionately poor in relation to overall district populations. Seven of the eight districts affected by Pak Beng dam are already categorised high priority poor districts by the Government of Lao PDR Almost half the populations of Stung Treng and Kratie are classified as poor Poor households are more dependent on the natural resource base than better-off households, particularly among ethnic minorities who are heavily dependent on upland swidden agriculture (Zone 2 Lao PDR) and fisheries (Zones 4 and 5, Cambodia) Many ethnic minorities in Laos and rural communities in Cambodia have already lost their access to customary agricultural methods and locations through government land sequestration policies. These will be again affected by compulsory land acquisition, making them even more vulnerable to increased impoverishment Ethnic minorities in Laos have already shown indications of cultural and psycho-social stress from compulsory relocation, as have Cambodian communities, which can only be expected to deepen with dam impacts Many ethnic minorities leading traditional, resource-dependent lifestyles, find it particularly difficult to adjust to new economic structures and different ways of life Dams in Zones 2 and 3 present high risk of impoverishment because of the geographical relationship of one dam to another, and because there has been no consideration given to sequencing of 213 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

214 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S construction and operation activities. Many districts in the impact areas of these 6 dams are not just affected by one dam at one time, but they may be affected again and later on by upstream or downstream impacts of another dam. Some households and communities can thus expect multiple and sequenced impacts spreading over many years. For example, the reservoir area of Louang Prabang dam extends back to the downstream area of Pakbeng, while its downstream area is also affected by the upstream area of Xayaboury dam. The concept of vulnerability is not just linked to the ability to resist project-induced impoverishment, but also to resist social impoverishment. Social impoverishment includes that which disrupt homogeneous and co-dependent communities by forced displacement from locations where they have developed economic, cultural, spiritual, and social relationships. When compensation is restricted to cash alone, this leaves relocation choice to individual households, who may not be able to replicate their social structures and who are removed from their cultural and spiritual ties, and completely fails to account for loss of common property resources. Community institutions and social networks are weakened, cultural identities and the potential for mutual self-help are diminished or lost 86. The distribution of risks of both direct and indirect impacts from the 12 Mekong mainstream dams are not equitable. Inevitably in many riparian provinces, Mekong mainstream communities depend heavily on fisheries and aquatic food sources for both consumption and livelihood, compared with communities at a distance from river sources. While the sources of these fisheries and aquatic foods are varied, the most important is the Mekong river, accounting for an approximate 37% of riparian communities' fisheries and 39% of the Tonle Sap fisheries alone 87. Any changes to these resources would have severe and long term consequences for riparian communities' protein intake, food security and overall health status. Ethnic minorities in Laos and Cambodia are particularly vulnerable, highly dependent as they are on the natural resources of the river, and in Laos, on common property resources. They are even more vulnerable in terms of impacts, as such communities have little power over national decision-making processes, and may already be socially and economically marginalised within countries. While Thailand and Vietnam have relatively well developed systems of social support and assistance, the two countries where the most critical direct and indirect impacts of hydropower development will be experienced still face major challenges in both capacity and resource limitations, to ensure that poor and vulnerable households who could be directly and indirectly affected by the 12 mainstream dams, not only receive what are their statutory rights to compensation and mitigation measures, but can share in any benefits that might arise from hydropower development. The scope, methodology and competence of compensation, mitigation and avoidance measures, will directly increase or decrease equity of risk. Thus the reputation and track record of proposed dam developers is critical to know before assessing the degree of risk of impoverishment to which affected communities might be exposed. To address this under hydropower development requires collective, transboundary and coordinated action which seeks a balance between poverty alleviation, economic development, social and ecological integrity. It is therefore not just a question of national and transboundary balancing of opportunities and risks, but more clearly identifying responsibilities and obligations on the part of line agencies, developers and oversight agencies, to ensure that the poor are not rendered poorer, and that equity of any development reaches such households in such a way that it is a right, not a handout. 86 World Bank, (2001) Operational Policy 4.12, Involuntary Resettlement, paragraph 1 87 Mekong River Commission, SIMVA, November I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

215 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S 9.6 TRANSBOUNDARY IMPACTS AND CONFLICT MANAGEMENT TRANSBOUNDARY IMPACTS Substantial transboundary impacts may be experienced where watersheds cross national boundaries. If watershed environmental loss occurs in one country due to activities such as illegal logging or land clearance for foreign concessions, this may have erosion consequences, with subsequent sedimentation impacts on reservoirs in another country. In situations such as these, without prior transboundary discussion and agreement, it will be more difficult to assign responsibility for indirect impacts, and therefore to agree budget allocations and remedial measures. Ideally, the situation should not arise in the first place, with transboundary agreements in place beforehand. There are transboundary risks associated with different types and principles of compensation and mitigation measures being applied by different developers in a single country, as well as by any one developer creating impacts in more than one country. The extent of risk depends on: 1. whether a developer is willing to change technical design of a dam to minimise impacts; 2. whether national policies and strategies have a good chance of being applied in practice, not just in principle; 3. familiarity levels in provinces and districts of compensation and mitigation policies, legislation and related implementing procedures; 4. ability and/or willingness of national agencies to monitor resettlement and livelihood restoration activities, and ability to insist on changes and/or compliance with agreements; 5. extent of preparation, competence and budget allocation by the developer to treat resettlement issues with the same professional seriousness with which engineering issues are addressed 6. whether a process of adaptive management is accepted and applied to address issues which could not have been anticipated during the planning period. The issue of resettlement is normally assigned to national policy implementation, and therefore assumed to have no transboundary impacts. This assumption is incorrect with respect to proposed Mekong mainstream dams. There are several reasons for this. The first is the broader definition of "resettlement" that is now applied by international financing agencies, as described in the baseline. Secondly, mainstream run-of-river dams or barrages have land acquisition and livelihood impacts on both sides of the river, as well as upstream and downstream of construction sites. Where the river forms an international boundary, there will consequently be impacts in both countries, and downstream impacts will be experienced across national boundaries. Impacts can, however, often be difficult to assign as national or transboundary. In the SEA's national scoping workshop (July 2009), it was pointed out that many current transboundary complaints remain unresolved. For example, Cambodia accuses Laos of downstream problems caused by waste disposal and contamination of waterways by plastic bags. Laos counters these with accusations of its own that Cambodian boats illegally cross boundaries and cause the problem. What this demonstrates is not just that there are grievances, but that there is currently no process or framework in place to deal with accusations and counter-accusations of who is responsible for what. This has implications of transboundary impacts related to hydropower activities. 215 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

216 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Strategic transboundary issues relating to Resettlement impacts include: Variability of LMB country legislation relating to potential social inequity of treatment of Affected Persons (APs). Each LMB country has its own land acquisition and compensation laws and policies. Some are more comprehensive than others. Although impacts may be the same on both sides of the river, actual compensation and mitigation measures may be different, leading to social inequity in treatment of people affected in the same way by the same project but living on opposite banks of the Mekong river, or living upstream or downstream of the construction site. International financing safeguard standards. If any developer seeks financing from an International Financing Institution (IFI) which has either developed its own safeguard policies or subscribes to the Equator Principles, common land acquisition and compensation, mitigation and livelihood restoration standards are required to be applied to the highest, rather than the lowest, levels irrespective of individual country systems. Definition of locations deemed affected. Not every country nor every developer recognises and accepts that mitigation measures, compensation, and livelihood restoration, apply to impacts experienced in all 3 locations outlined above. Most consider the construction site and land lost to impoundment as being the extent of responsibility for compensation, and that this can be addressed through cash payments. Consequently associated impacts, such as downstream impacts, health consequences of elevated groundwater levels adjacent to headponds, and livelihood restoration, may be left out of the resettlement planning equation. Minimal safeguards standards approach. Where a minimal standards approach can be applied, a developer may disregard even national land acquisition and compensation standards, and national agencies themselves may be unwilling or unable to monitor their application, or monitors in one country may be more effective than in another country on the same project. This also could lead to further substantial inequities for affected people, with one country more effectively protecting the rights of its citizens than another. Consistency of developer's approach needed. If all countries support the premise that the developer is responsible for compensation and mitigation measures, a consistent approach throughout the LMB countries is necessary concerning the developer's responsibilities towards a land acquisition, compensation, mitigation measures and livelihood restoration programme, which does not depend solely upon cash compensation, but is approached as a development project in its own right. Political unrest caused by social discontent among affected people in one country could cause construction delays, causing associated construction delays in the transboundary country and resulting in overall higher construction costs for the developer, delays in meeting power purchase agreement deadlines and possible consequent financial penalties. Disease does not recognise national boundaries. Existing health problems which could be exacerbated by construction and operational activities, and development of new problems, need a coordinated transboundary approach to anticipate and address. Natural phenomena, such as naturally-occurring arsenic in groundwater, are also not restricted to one country. Again, a coordinated approach is required to address potential risks to human health. 216 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

217 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S Transboundary watershed management may be required in some locations to avoid additional operational resettlement impacts due to erosion, particularly in Zone 2 locations between Thailand and Laos, and in Zone 4 between Laos and Cambodia. Those engaged in the illegal activity of human trafficking routinely circumvent national policies and procedures. Improvement of communications and transport networks require associated improvement of transboundary frameworks to deal with this. Lastly, operational procedures need transboundary agreements on dam safety and downstream flood preparedness. A large dam does not need to fail to have severe consequences, and notification procedures and preparedness are essential, as are downstream early warning procedures in case of sudden water releases, particularly in densely populated areas such as the lower reaches of the LMB. The resettlement component of any project is often the weakest component in a dam developer's armoury. This is unwise at best, as it risks contributing to substantial cost overruns, implementation delays and more adverse impacts on a larger number of people. At worst it can lead to considerable social and political unrest, and is a high risk approach for any LMB country CONFLICT MANAGEMENT There are four levels of potential conflict which could arise over the various dams: 1. between an aggrieved person and a particular developer. Where a dam is not in a transboundary location, this can be addressed by conflict resolution procedures developed in project-specific resettlement plans, as well as by national and local redressal mechanisms; 2. between aggrieved persons in different countries and a single developer in transboundary locations (e.g. dams on sections between Thailand and Laos, and for dams immediately on the Lao/Cambodian border), who may see lack of equity in the way the developer treats compensation and mitigation processes for people affected in the different countries; 3. between different developers who may have separate or sequential impacts on the same communities or districts, and who may refuse to accept responsibility for those impacts. There is currently no mechanism to deal with this situation; 4. transboundary disagreements. There is also currently no grievance redressal framework to deal with these last three types of potential conflict. At present, there are no frameworks in place to deal with any of these conflict scenarios, with the exception of national legislation and policies dealing with grievances between directly affected persons and a developer. All LMB national processes regarding broader conflict resolution needs are deficient at present, as they have never been required to address cumulative and overlapping impacts by different developers on the same people in the same or in neighbouring countries. The potential for conflict is already demonstrated by the experience of multiple hydropower developer activities in river basins even within national boundaries. In the Nam Ngum river basin, an upstream hydropower developer whose activities affected the domestic water supply of a village prior to its relocation under a second hydropower developer further downstream, took some time to accept its 217 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

218 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S responsibility for restoring the domestic water supply. This kind of "pass-the-parcel" over assigning responsibility for overlapping impacts can be expected between the 12 developers on the Mekong mainstream also, leaving affected communities in limbo with no option for redress. Where this sort of situation may arise between countries, significant deterioration in relations between those countries may be anticipated KNOWLEDGE GAPS SOCIAL SYSTEMS Very substantial knowledge gaps remain which continue to make it difficult to fully appreciate the full scope of both direct and indirect impacts of the proposed 12 Mekong mainstream dams on communities in the LMB. These gaps are both social, economic and technical. 1. Socio-economic systems in the LMB. While the MRC undertakes and updates SIMVAs in selected locations in the LMB, these have focused on LMB reliance on water resources, rather on land-based riparian natural resources such as agricultural land, forests and riverbank gardens. There is very little information available or analysis undertaken on the complexity of interactions and reliance of different riparian populations on such resources in the LMB, and the way in which the Mekong river enables this reliance. The exception to this is in the Tonle Sap, which has been socially and economically well studied, as has the Stung Treng Ramsar site. But extensive knowledge gaps remain concerning social structures, cultural identity, economic status, and their relationships to the degree of reliance on natural resources, in Zones 2, 3, 4 and 6. There is almost no information available at all for Zone 1. The MRC's recent SIMVA (2010) acknowledges that its relatively small sample size and the fact that long stretches of the Mekong river could not be covered by its social assessment team, means that the survey should not be considered definitive or truly representative of the whole Mekong corridor 88. Updates and enlargement of a long-term Social Impact Monitoring process are now under way. Additional support should be provided to enhance public understanding of riparian community ways of life in more detail, to provide improved understanding of the consequences for any changes to those ways of life. 2. Fisheries and livelihoods. While comprehensive data is available on fish species and their distribution and movement along the LMB system, and on the commercial value of fisheries in Cambodia, there are gaps in knowledge and information on the distribution of subsistence versus commercial fisheries and the relative importance of both to riparian communities in other LMB countries and in different parts of the Mekong mainstream, particularly in Zones 1-4. It is difficult to estimate the full scope of social and economic impacts on local communities over loss of fisheries without further research. A fisheries survey undertaken by the MRC in Louang Prabang is an important contribution to this knowledge base 89, but many changes have occurred in Laos since 2000, and updated information is needed in these sort of upland riparian communities on both the Lao and Thai sides, as well as in communities in other parts of the Mekong mainstream. "The magnitude of the social impacts through impacts on fisheries is dependent on what those impacts will be. There is a big difference between a 20% reduction and a 50% 88 MRC SIMVA March 2010, op cit, p. xxiv 89 Fisheries Survey, Luang Prabang Province, NAFRI/MRC, October I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

219 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S reduction in the fish catches in Tonle Sap and these two possibilities would accordingly impact differently on livelihoods. It is likely that the large-scale and medium scale fishing people would be hardest hit, whereas subsistence fisheries could increase effort and/or fish in alternative habitats to the main river system" 90. In particular, the impacts on fisheries to Lao and Thai riparian communities if only the northern (Zone 2) cascade of dams are constructed, are not well understood, though the significantly negative impacts on fisheries is better understood in relation to dams in the lower reaches of the Mekong, particularly the Don Sahong, Sambor and Stung Treng dams. 3. Due diligence on developers. Normally developers are expected to undertake extensive prior social and economic assessments in the form of Social Impact Assessments, Social Management Plans, Resettlement Plans, etc. While some developers comply with national requirements to present a clearer picture of the socio-economic situation in the impact areas of a proposed development, others fail to do so, or do so inadequately. Deficiencies in developers' obligations to create an adequate knowledge base to inform national and LMB governments about possible scope of impacts, in turn inhibits MRC's ability to effectively review any particular dam. Comprehensive safeguard documents in line with standards and procedural compliance requirements of national policies should be fully available to MRC before any consideration of specific dams, and in turn, MRC should have the institutional capacity to review such documents prior to any decision. Some hydropower developers have experience of implementing hydropower projects in the region, while others have experience in other countries. Due diligence of proposed developers' track performance in applying national social and environmental safeguards should be conducted by the MRC as part of its review process. 4. Sedimentation and Water Flow. Proposed reservoirs of the 12 mainstream dams will cause a major and irreversible shift in the way stream power is dissipated in the Mekong river. There remain critical data information gaps on sedimentation and water flows, with the result that indirect, cumulative and long term impacts, particularly in the lower reaches of the LMB, particularly the Cambodia floodplain, the Tonle Sap, and the Vietnamese delta, cannot be fully estimated at this time. Consequently the Social Component cannot determine the scope of social and economic impacts on communities living in areas which could be affected by alterations in flooding patterns, damage to infrastructure such as irrigation intakes or boat docks. The MRC's BDP also notes that the complex issue of water flows and nutrient load changes in the LMB, particularly on the Tonle Sap natural flow reversal, needs further study Climate Change. Impacts from other influences, such as climate change, and their interaction with possible impacts from dams, are currently less understood. Consequently, cumulative social and economic impacts on LMB communities cannot be determined at present. Bridging these knowledge gaps are critically important, not only to reach a better understanding of the pre-mainstream dam situation, but also to determine the extent to which changes in the socio-economic situation of Mekong riparian communities can be attributed to impacts of the different dams, especially 90 Mekong River Commission, Basin Development Plan Programme, Phase 2, Technical Note, Initial Findings from Assessments, January Ibid 219 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

220 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T S O C I A L S Y S T E M S to indirect impacts, and what are pre-existing and continuing cumulative effects of other activities occurring along the Mekong river. This also has financial consequences for the dam developers, as it will influence the scope of compensation requirements and mitigation measures they will be required to undertake. 220 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

221 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N 10 NAVIGATION 10.1 STRATEGIC ISSUES 1. What opportunities are provided by construction of mainstream dams for development of navigation activities on the Mekong River? 2. What threats do the mainstream dams pose to long-haul and cross border transportation on the Mekong River? 3. What are the minimum requirements, standards and guidelines that should be adopted for the design, construction, maintenance and operation of ship locks? 10.2 OVERVIEW The construction of mainstream dams for hydropower dams provides an opportunity to improve the navigability of the Mekong River by providing more reliable and consistent water depths that will facilitate larger vessel capacities. At the same time, the construction and operation of mainstream dams presents a threat to long-haul and cross-border river transportation on the Mekong River. Development of Mekong navigation in the upper reaches depends on the incorporation of appropriate navigation facilities into hydropower developments as well as designing the hydropower cascade to ensure a reliable channel is available between major ports If the mainstream dams proceed, then article 9 of the Mekong Agreement requires that navigation locks are incorporated into hydropower developments. It is also logical that the design, construction, maintenance and operation of all ship locks along the river should be subject to common standards and guidelines. Lock dimensions must accommodate traffic increases over a 50-year timeframe. Transportation of heavy cargo, such as mined products, containerised waterborne transportation and the introduction of inland cruise vessels will inevitably increase in years to come and the size of ship locks needs to accommodate this expansion SUBSISTENCE USERS (SMALL, MEDIUM) PAST TRENDS AND CURRENT SITUATION: There has been a decline in small and medium users for transport on the Mekong River over the last ten years with the improvement of roads and access to public road transport and private vehicles. However 221 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

222 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N the Mekong River is still an important means of transportation for a large number of riparian communities. Fig 11.1: Mekong River an important means of connectivity for small riparian communities, Lao PDR Local communities continue to use the Mekong River as an important means of transport; linking communities and villages for trade, social and economics means. And for poor rural communities boats provide an affordable and easily accessible means of transport which is environmentally friendly. Small boats carry agricultural products to markets and provide access to schools, health care and other social services. Subsistence users still use the Mekong River from Pak Beng to Pak Chom (Zone 2), Ban Koum and Lat Sua (Zone 3) and Stung Treng and Sambor (Zone 4). FUTURE TRENDS WITH HYDROPOWER: Countries in the Greater Mekong Subregion will continue to invest in roads and road transport which will lead to a decline in small and medium users on the Mekong River as seen in the last decade. However with the further development of agriculture, aquaculture and sustained population growth in Lao PDR and Cambodia could result in higher population density on the Mekong River. This coupled with improved navigability may lead to a higher demand for small and medium waterway users in the future INCREASED NAVIGABILITY: Increased navigability will improve safety for smaller vessels as there will be less currents, rapids and more consistent navigation conditions all year round for all navigation users on the Mekong mainstream. Increased navigability may encourage more small and medium users to transport goods to nearby villages to increase trade and improve connectivity between riparian communities. Navigation in the future should be promoted as an efficient and environmentally friendly means of transporting goods and people in the Mekong Basin. 222 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

223 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N DECREASED LONG-HAUL CONNECTIVITY: Subsistence users do not commonly travel long distances, however they are currently able to move freely along the Mekong River. In all flow zones there are a number of small riparian communities that use small vessels to transport construction materials, agricultural products and people from one village to another. The mobility of riparian communities that are within 15 kilometres of proposed hydropower dams may be impacted during the construction phase and operation of the of hydropower dams. There will be phases during construction of the mainstream dams when vessels will not be able to pass freely. Smaller vessels may not be permitted to use the Ship Locks on the Dams or incur delays as the locks may be required to wait for a number of vessels to enter the lock prior to releasing the vessels into the mainstream. Subsistence users could be required to pay fees to use the ship locks, this may make the Mekong River no longer economical for small and medium users. Further research needs to be done to determine the numbers of small users within 15 kilometres of proposed hydropower projects and how the construction phase and operation of the dam will impact on their mobility and livelihoods PASSENGER TRANSPORT (INCLUDING FERRIES, TOURISM, CRUISES, ETC) PAK BENG TO PAK CHOM (ZONE 2) PAST TRENDS AND CURRENT SITUATION: Tourism is one of the fastest growing sectors in the Greater Mekong Subregion. International tourism is expected to develop. The upper values in these forecasts represent the situation of good cooperation between the GMS-states, which is mainly in the field of easing of border crossing travel. The number of international tourist arrivals is expected to increase by between 7% and 12% per year, and so will the cruise sector in the Mekong Basin. This gives opportunities for more cruises and for further diversification into higher value segments. (Vrenken, 2008) The main navigation activity in the Upper Mekong in Lao PDR is passenger transport and cruises. In the last 10 years passenger transport has been dangerous due to rapids and low water levels, passenger transport has been restricted to small slow boats with shallow draught to accommodate the low water levels in the dry season. 223 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

224 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N Fig 11.2: Hazardous navigation conditions north of Vientiane FUTURE TRENDS WITHOUT HYDROPOWER: Tourism in the upper Mekong continues to grow. Between 20,000 and 25,000 tourist cruise passengers a year take trips on this part of the river. More than 85 per cent of these people travel between Houei Sai and Luang Prabang. This trip is an important component of the increasingly popular link between Chiang Mai and Luang Prabang. Cruise operators in the Upper Mekong have been able to increase their activity in the recent years. The development has been most significant for relatively new tour operators. These operators serve a higher income tourist segment than local operators and therefore also have the potential to create more spinoff in higher class tourism in villages along the Mekong. Currently navigation between Houei Say and Luang Prabang is still hazardous, particularly with the low water levels in 2009/2010 leading to a number of accidents reported to the Mekong River Commission. The Chinese government has recently agreed to provide the Government of Lao PDR (GoL) with 15 million USD to further improve navigation conditions on the river between Houei Say and Luang Prabang. It is expected that there will be increased investment in river cruises by local and international cruise 224 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

225 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N companies. International cruises to and from China and Thailand to Luang Prabang are expected to increase significantly. Fig 11.3 Passenger transport vessels operating between Luang Prabang and Houei Say, Lao PDR FUTURE TRENDS WITH HYDROPOWER: If all six proposed mainstream dams are constructed north of Luang Prabang then the cascade of dams will provide sufficient water depth and navigability all year round and 24 hour navigation. Night navigation is very attractive to cruise operators and as it allows other tourism activities to be integrated into tourism packages during the day and cruising at night. This would see the introduction of a number of large international tourist cruises to and from China and create further tourism opportunities for riparian countries. A new potential of cruise services will be unlocked if navigability improves in a way that the whole stretch between Vientiane and Yunnan can be navigated, This will make cross-border tourism in the Lancang- Mekong area more attractive and can reinforce other ongoing initiatives for the development of tourism in this region. Proposed International cruises from Jinghong, China and Luang Prabang Lao PDR will require further investment in fleet and port infrastructure to ensure safe and sustainable navigation on the Mekong River. Investments will need to be made in improving customs and immigration services. 225 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

226 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N If only one or a few of the proposed Hydropower dam are constructed then there will only be partial accessibility and limited improvements to navigability for passenger transport BAN KOUM AND LAT SUA (ZONE 3) PAST TRENDS AND CURRENT SITUATION: Currently, downstream of Savanakhet up to the Khone Falls navigation is very limited. The transport of passengers has declined in Flow Zone 3 due to improved roads and the opening of the Second Thai-Lao Friendship Bridge from Mukdahan to Savannakhet in FUTURE TRENDS WITHOUT HYDROPOWER: Channel improvement, aids to navigation and more investment in vessel and port infrastructure may provide opportunities for investment in increased cruises and passenger transport between Vientiane and Savanakhet. FUTURE TRENDS WITH HYDROPOWER: The proposed construction of two dams at Ban Koum and Lat Sua will only produce limited opportunities for increased navigability STUNG TRENG AND SAMBOR (ZONE 4). PAST TRENDS AND CURRENT SITUATION: The passenger traffic into and out of Stung Treng port has been declining in recent years. The Cambodian Master Plan in 2004 forecasted passenger traffic to and from Stung Treng to only grow within the range of -2% and +2% per annum for the next 20 years. FUTURE TRENDS WITHOUT HYDROPOWER: The signing of a Treaty on Waterway Transportation between Cambodia and Viet Nam has opened up Mekong waterways to a range of new possibilities for generating trade and tourism revenue. Through this treaty, the Governments of Cambodia and Viet Nam have agreed to allow all waterway users to freely cross borders for the transport of cargo and passengers. Stung Treng is not included in the agreement, however transport from Kampong Chan is expected to increase significantly and cruise companies are currently undertaking feasibility studies to increase passenger services in Flow Zone 4. FUTURE TRENDS WITH HYDROPOWER: In the Lower Mekong between the Khone Falls and Kampong Cham: 130km of this 280km stretch would be improved by the construction of mainstream dams downstream of the Khone Falls. However the dams at Stung Treng and Sambor would not create the same opportunities for increased navigability as in the Upper Mekong as the Khone Falls still poses a major obstacle for navigation activities in the Lower Mekong. The construction of the Cambodian dams will only marginally increase navigability for cruise 226 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

227 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N services and passenger transport. A more stable flow regime will improve safety and reliability for cruise operators which will see increased future investments in the cruise sector INCREASED NAVIGABILITY: The tourist industry is valuable for the economies of each of the riparian countries. For a good functioning of the cruises, navigation conditions need improvement in the dry season already. There is much potential additional value if navigation can be ensured on the complete Lancang Mekong stretch between Jinghong and Luang Prabang. Increased navigability will provide better opportunities to operate cruises all year round and ensure a higher level of safety and reliability. There may be an opportunity to adopt night navigation, which would ensure cruises could operate all year round and 24 hours a day increasing efficiency. Less rapids and decreased currents will lead to a decrease in operator fuel costs to operators and reduce greenhouse gas emissions DECREASED LONG-HAUL CONNECTIVITY: The construction of hydropower dams will have potential negative impact on passenger transport during the construction phase and operation of the hydropower dams. The construction phase of mainstream hydropower projects will restrict the connectivity of passenger transport as there will be times when the navigation channel is obstructed and vessels cannot pass freely or without risks to vessels, and passengers. Currently the most popular passenger transport in the Upper Mekong is to and from Luang Prabang. Passenger transport vessels and cruises will be required to enter ship locks at the mainstream dams causing delays and inconvenience for passengers. Hydropower developers will need to ensure ship locks will be operational to accommodate cruise vessels and 24 hour navigation FREIGHT TRANSPORT (SMALL, MEDIUM AND LARGE) PAK BENG TO PAK CHOM (ZONE 2) PAST TRENDS AND CURRENT SITUATION: In 2000, prior to the signing of the Quadrangle Agreement on Navigation between China, Myanmar, Thailand and Lao PDR many parts of the Mekong River were not navigable for Freight Transport in the Upper Mekong. River regulating works between Jinhong China and Chiang Saen, Thailand has improved the navigability of the Upper Mekong and resulted in increased trade between the two countries. 227 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

228 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N Fig 11.4 Chinese vessels loading cargo at Chiang Saen Port, Thailand The total volume of the freight traffic between Thailand and Yunnan over the Mekong River was about 260,000 tonnes in the fiscal years of 2006 and By far most of this traffic nearly 180,000 tons - goes through the Port of Chiang Saen. Traffic volumes increased from nil when the Quadripartite Agreement was signed in 2000 and showed a steady increase up until The graph in (Fig 11.5) shows the steep increase from 2004 to 2005, a constant level afterwards and a slight decline in 2008 to about 154,000 tonnes in the fiscal year 2008 (October-September). This decline is mainly due to low traffic volumes in August 2008, which can be explained by the extremely high water level in this month. Apart from this cargo in ,366 vehicles were transhipped in Chiang Saen and nearly 7,000 passengers passed through the port (Vrenken. 2008). Cargo operations in the Upper Mekong has decreased significantly in Lao PDR, due to road construction and companies opting for road instead of using cargo vessels in the dangerous navigation conditions north of Luang Prabang. The domestic trade is predominantly agricultural products, consumables and arts and crafts from local communities for sale in Luang Prabang. Freight transport in other sections of the 228 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

229 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N Upper Mekong in Lao PDR has also gone through a decline in demand and today is characterised by low productivity. Figure 11.5: The development of cargo transport between Chiang Saen port and Yunnan ports FUTURE TRENDS WITHOUT HYDROPOWER: The Mekong River is already an important link in the transit chain between Kunming and Bangkok with about 300,000 tonnes of good shipped via this route each year. The volume of this trade is expected to increase by 8-11 per cent per year (Vrenken 2008). The development of the Chiang Saen Port II will provide even further opportunities for economic growth and trade between Thailand and China. Increased navigability may provide new opportunities for inland waterway transport in the Upper Mekong. The Thailand Government has allocated a total budget of approximately 1,500 million baht for the construction of Chiang Sean Port II, which is expected to be completed in The Thailand Marine and Harbour Administration is confident that the completion of Chiang Sean II Port will contribute to further international trade and transportation among the Contracting Parties to the Agreement on Commercial Navigation on the Lancang-Mekong River, which will in turn benefit the Greater Mekong Sub-region (GMS) as a whole. 229 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

230 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N Fig 11.6: Chiang Saen Port II Development, Thailand Mekong River Flood way Kok River Flood Plain Birth with Ramp (N) Port Office Quay Wall Birth with Ramp (S) Tank Farm The Mekong River Commission has contributed 2.5 million (USD) for condition surveys, maps and the installation of navigation aids to improve the safety of waterborne transport in the Upper Mekong. Additionally, the Chinese Government has approved a 15 million (USD) loan to the Lao PDR Ministry of Public Works and Transport for further improvement of the Mekong River between Houei Say and Luang Prabang. This investment is a clear indication that China intends to increase trade and tourism with Lao PDR under the Agreement on Commercial Navigation on Lancang-Mekong which allows vessels of the contracting parties to sail freely between Simao in the People s Republic of China and Luang Prabang in Lao People s Democratic Republic. The preliminary announcement was confirmed in April 2010; currently the scope of work has not been confirmed. Channel improvement, aids to navigation and more investment in vessel and port services will significantly improve the effectiveness of inland waterway transport and provide economic opportunities for riparian communities. FUTURE TRENDS WITH HYDROPOWER: The construction of all 6 proposed mainstream dams in from Pak Beng to Pak Chom will provide the best case scenario for increased navigability on the Mekong River. Barges of up to 1000DWT with a draught of 2.5m will be able to operate from Jinghong, China to Vientiane, Lao PDR. Increased navigability will provide new opportunities for inland waterway transport for freight transport. Accessibility to and from 230 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

231 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N Luang Prabang is the most important consideration for increasing trade and economic opportunities between China and Lao PDR. If only one or a few of the proposed Hydropower dam are constructed then there will only be partial accessibility and limited improvements to navigation conditions BAN KOUM AND LAT SUA (ZONE 3); PAST TRENDS AND CURRENT SITUATION: Currently, downstream of Savanakhet up to the Khone Falls navigation is very limited. The transport of goods and passengers has declined in Flow Zone 3 due to improved roads and investment in road freight transport. The opening of the Second Thai-Lao Friendship Bridge from Mukdahan to Savanakhet in 2006 has resulted in the decline of ferries carrying cargo between Thailand and Lao PDR. FUTURE TRENDS WITHOUT HYDROPOWER: Increased navigability may provide further opportunities in the future along with the further development of mining, resources and agricultural sectors. The development of Mekong river ports between Vientiane and Savanakhet was also proposed, with similar status as Lancang-Mekong navigation improvement for development before This project may be rejuvenated in the future with increased navigability and opportunities arising in the mining and resources sector. FUTURE TRENDS WITH HYDROPOWER: On the 450km stretch between Vientiane and Savanakhet, it may be possible to operate year-round, with vessels of 100DWT in capacity and, in all likelihood capacities could be increased to 500DWT for 6 to 8 months of the year. A long-term objective of a 2.5m draught seems realistic if certain additional smaller hydropower projects are carried out in this part of the Mekong River. If increased navigability is increased between Vientiane and Savanakhet then this may increase opportunities for navigation activities in Flow Zone STUNG TRENG AND SAMBOR (ZONE 4): PAST TRENDS AND CURRENT SITUATION According to the Master Plan for Waterborne Transport on the Mekong River System in Cambodia 2004 local cargo traffic is forecast to grow between 2 and 6% over the next 20 years. However, the port s cargo throughput could be augmented by additional cement volumes which might be expected to be attracted to IWT following the commencement of domestic cement production in In total, the port s cargo throughput could reach 12,000 20,000 tonnes by Increased cement tonnages may require an increase in average vessel capacity (possibly of up to 200 DWT, if such an increased capacity can be accommodated on a year-round basis in the future). FUTURE TRENDS WITHOUT HYDROPOWER: Formulation of the Treaty on Waterway Transportation between Cambodia and Viet Nam aims to improve the legal framework for encouraging freedom of navigation in the Mekong River system, thereby 231 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

232 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N implementing Article 9 of the 1995 Agreement on the Cooperation for the Sustainable Development of the Mekong River Basin. The new treaty will open up Mekong waterways to a range of new possibilities for generating trade revenue. Through this treaty, the Governments of Cambodia and Viet Nam have agreed to allow all waterway users to freely cross borders for the transport of cargo. FUTURE TRENDS WITH HYDROPOWER: The Lower Mekong between the Khone Falls and Kampong Cham: 130km of this 280km stretch would be improved marginally by hydropower developments downstream of the Khone Falls. However the dams at Stung Treng and Sambor would not create the same opportunities for increased navigability as in the Upper Mekong. Khone Falls still poses a major obstacle for navigation activities in the Lower Mekong. In the long term, the most promising solution would be to plan a high capacity channel (3m draught) linking a multimodal hub near the Khone Falls to Kampong Cham and the delta. A by passing canal could also be considered if traffic increases strongly. (Mathurin, 2008) INCREASED NAVIGABILITY: Increased navigability may lead to opportunities for inland waterway transport. A general increase in demand for freight transport will benefit inland waterway transport. Assuming that market shares will sustain, the volume of inland waterway transport will also increase with nearly 10 percent per year for freight.vrenken). If properly planned and designed, the navigation conditions good very much be improved, even to the effect that barges of more than 1,000 DWT could be brought in. The construction of all 6 proposed mainstream dams in from Pak Beng to Pak Chom (Zone 2) will provide the best case scenario for increased navigability on the Mekong River. Barges of up to 1000DWT with a draught of 2.5m will be able to operate from Jinghong, China to Vientiane, Lao PRD. Increased navigability will provide new opportunities for inland waterway transport in Lao PDR for freight transport.. It should be noted that if only one or a few of the proposed mainstream dams between Pak Beng to Pak Chom are constructed then there will only be partial accessibility and limited improvements to navigation conditions. The construction of the mainstream dams in Ban Soum and Lat Sua (Zone 3) and Sambor and Stung Treng (Zone 4) will provide only limited opportunities for increased navigability. Further investment would be required to improve the navigation channel and development of river ports to improve connectivity on the 450km stretch of the river between Vientiane and Savanakhet, Knone Falls still provides a major obstacle for inland waterway transport DECREASED LONG-HAUL CONNECTIVITY: Freight transport will be required to enter ship locks at the mainstream projects, causing delays and reducing the efficiency of inland waterway transport. However for industrial companies, reliability of supply is often more important than speed of transport. Hydropower developers will need to ensure ship locks are operational, reliable and can accommodate 24 hour navigation for freight transport. Ship locks will have more of a negative impact on passenger transport. 232 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

233 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N The construction phase of mainstream hydropower projects will restrict the connectivity of freight transport on the Mekong River. There may be phases when the navigation channel is obstructed and vessels cannot pass freely or without risks to vessels, workers and the environment MEKONG DELTA PAST TRENDS AND CURRENT SITUATION: River and sea ports in the Mekong Delta are one of the most significant trade regions in Viet Nam. In the Mekong Delta almost 70 per cent of goods; rice, construction materials and consumables are transported by water. Trade in the Mekong Basin continued to experience significant growth despite the economic downturn, the Mekong and Bassac rivers have become very important trade corridors in Viet Nam. On the other hand in Cambodia there was a slight decrease in cargo throughput in 2008 and 2009 in Phnom Penh Port after enjoying steady growth since 2004; this can be attributed to the global financial crisis as demand for the export of garments in US. FUTURE TRENDS: In 2009, trade in the Lower Mekong Basin received a significant boost with the opening of a new deepwater port at Cai Mep in Viet Nam. This new port has generated a renewed focus on the Mekong River as a trade route. The Cai Mep container terminals can accommodate vessels with a draught of 15.2 m, equivalent to the largest container ships in the world. These mother vessels sail directly to Europe or the United States, which means that goods can be shipped internationally to and from Phnom Penh with only a single trans-shipment at Cai Mep MINING AND RESOURCES Foreign Direct Investment continues to increase in Lao PDR and Cambodia in the Mining and Resources Sector. Representatives from Chinese Mining Operations have already met informally with the MRC Navigation Programme discussing future navigation opportunities relating to the transportation of Bauxite, Iron Ore and construction materials from and to their future mining activities. It is proposed that approximately 5-10mt/A could be transported on the Mekong River and incorporated into a multi-modal transport chain with road and rail, depending on the future railway route in Lao PDR and further enhancements in inland waterway transport. More ships with bigger capacity will be required to transport minerals and construction materials. Ships. Vessels of up to 1000DWT and larger would be required for the transport of minerals and construction materials. Further investments will be required into improved waterway infrastructure and capacity of current vessel fleets. Increased navigability and all year round navigation will be an incentive for riparian governments to invest in Waterway transportation and promote multi-modal transport solutions to the mining and resources sector. 233 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

234 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N 10.8 FREEDOM OF NAVIGATION Freedom of Navigation should not be compromised during the construction phase and the operation of hydropower dams. Ships locks must be appropriate dimension and operational. The Agreement on the Cooperation for the Sustainable Development of the Mekong River Basin, ratified in 1995 by the four riparian countries of the Lower Mekong River Basin - Cambodia, Lao PDR, Thailand and Viet Nam - gave the overarching role of MRC as assistance in regional cooperation and policy development. A common interest in increasing international trade through inland water transport was the reason why the MRC signatories chose to have a separate article in the 1995 Agreement on Freedom of Navigation. That is why Article 9 of the 1995 Agreement gives the MRC a specific mandate to promote water transportation and to encourage freedom of navigation on the mainstream of the Mekong River. In this respect, Guidelines have been established for the planning and design of hydropower projects on the Lower Mekong Mainstream, focusing on environmental and transboundary issues. It is recommended that the envisaged hydropower developments include, at first stage, ship locks accommodating medium term traffic (5 million tons/per annum) and allow for future lock system expansion, at second stage (when the level of traffic justifies it). The transit capacity of the first stage of lock implementations thus is consistent with next Upper Mekong improvement phase, decided in the framework of the Quadrangle Navigation Agreement 9 (Mathurin, 2008) CONCLUSION The following conclusions can be made on the impacts of mainstream hydropower dams on Navigation and inland waterway transport: Increased Navigability 1. The construction of the six mainstream dams between Pak Beng to Pak Chom will provide opportunities for the future development of passenger and freight transport from Vientiane, Lao PDR to Jinghong China. 2. The construction of the mainstreams dams in Sambor and Stung Treng will provide opportunities for the development of navigation between Khone Falls and Phnom Penh. 3. The construction of the mainstream dams in Ban Soum and Lat Sua will provide only limited opportunities for developing navigation. 4. Channel improvement and implementation of suitable ship locks must be associated with port modernisation, good links with road and (possibly) rail networks and trade facilitation. Decreased Long-haul River Connectivity 234 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

235 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T N A V I G A T I O N 5. Decreased connectivity will be a negative impact of the six mainstream dams between Pak Beng to Pak Chom for passenger transport if suitable locks are not operational and maintained effectively. 6. The construction of mainstream dams will reduce connectivity for freight transport. 7. The construction of all mainstream dams will have negative impacts on subsistence users on the Mekong River. Freedom of Navigation 8. Freedom of Navigation will be impeded during the construction phase of mainstream dams. 9. The operation of mainstream hydropower dams will impede Freedom of Navigation, Article 9 of the MRC 1995 Agreement if suitable ship locks are not operational and maintained effectively. Ship Locks 10. High lock reliability and maintainability is an essential prerequisite for developing stable waterborne traffic. 11. Sediment build up behind the mainstream dams may impact on the navigational channel and the entrance to ship locks I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

236 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E 11 CLIMATE CHANGE 11.1 STRATEGIC ISSUES 1. What changes are foreseen in climate and hydrological variability and extremes? 2. What implications will those changes have for natural and social systems in the basin? 3. What implications will those changes have for development sectors in the basin including hydropower? 11.2 SUMMARY OF PAST & FUTURE TRENDS WITHOUT LMB MAINSTREAM HYDROPOWER Already, climate changes in the Mekong region are influencing ecosystems, livelihoods and development through changes in regular weather ie daily, seasonal and annual patterns and through irregular extreme events. The main influences (and indicators of change) are temperature, rainfall and runoff, sea level, tidal fluctuations and extreme events such as storms, floods and drought. Over the past 3 to 5 decades, trends of increasing mean annual temperature have been recorded in each LMB country. Most notable is the increase in variability from one year to the next. The trends in rainfall are less consistent with increasing variability and extremes between wet and dry in Laos and Cambodia, a decrease in rainfall in Thailand, and decreases in most localities in the north of Vietnam with increases in most areas of the South during all seasons. During winter in Vietnam overall rainfall fell by 23%. Seasonal changes are important, with most increases in rainfall occurring during the wet season. All countries have experienced decreasing rainfall during the dry season with aggravated drought and water stress situations in many catchments. Future climate to 2030 is projected to include steady increases in mean basin temperature by 0.8 C. Greater increases are expected in northern zones of the basin up to 1.4 C in Yunnan Province. Annual rainfall would increase by 13.5% (0.2m) mainly due to increases during the wet season (May to Oct). Dry season rainfall will increase in northern zones (1 and 2) and decrease in southern zones (3 to 6 ie from Vientiane to the Delta). The overall disparity between wet and dry seasons will increase especially in zones 3 to 6. Since the SEA baseline was prepared, MRC has conducted valuable additional modelling on the effects of climate change on rainfall, runoff and flow. This summary section on past and future trends with climate change expands the earlier baseline report by analysing the relevant data from the MRC modelling and comparing findings with the earlier CSIRO study. Overall the trends from both the CSIRO and MRC modelling are similar. Specific differences in reporting are due to differences in methodology and focus as follows: 236 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

237 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E CSIRO study (2008) Focussed on one IPCC scenario A1B as it represents a mid-range scenario in terms of development impacts on GHG emissions. Modelled mean temperature and rainfall for six IPCC scenarios (Figure 1) As the past baseline, constructed a 1901 to 2002 dataset by interpolating observed values. Used 11 Global Climate Models and took averages Focussed most analysis on one future time slice 2030 MRC modelling Used A2 and B2 scenarios which to 2050 at least are expected to release less GHGs than A1B Modelled temperature and rainfall for two scenarios Adopted the BDP baseline scenario of observed values for 1986 to 2000 Used 2 recent GCMs Expressed results in six blocks of years from 2010 to 2060 (Figure 2) The CSIRO work showed that temperature and rainfall projections for all IPCC scenarios in the Mekong Basin only begin to differentiate significantly after 2030 (Figure 1). Figure 1: Projected mean temperature and mean annual rainfall for the Mekong Basin for different IPCC scenarios at 2030, 2050 and 2070 (CSIRO 2008) CSIRO projected total annual runoff from the basin in 2030 to increase by 21% mainly during the wet season with the annual discharge at Kratie increasing by 22% with increases in all months but mainly the wet season. This projection takes into account estimates for water use by future populations in the basin and irrigation. CSIRO found that flooding is projected to increase throughout the basin with downstream zones affected most. For example at Kratie (zone 5) the annual probability of extreme wet flood events will increase from 5% (ie historic conditions) to 76%. It will increase to 96% in the wet season. The duration of flooding will increase in this zone with an earlier onset. The maximum and minimum area and water levels in Tonle Sap would increase annually. The annual average flooding in Delta would also increase by 3,800km 2. The impacts of this increase in flow and flooding could be expected to be greatest on the mainstream Mekong due to cumulative contribution from tributaries. MRC compared changes in average annual flows due to climate change against the MRC baseline scenario (no new developments) (Figure 2). 237 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

238 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Figure 2: Percentage change in mean annual flow due to climate change against the BDP baseline ( ) Source: ICEM analysis of MRC modelling data Figure 2 shows that: is projected to have an mean increase in annual flow of ~23% in Zone 2, dropping to an increase of ~15% in zones 3 to 5 The CSIRO work is consistent with the upper range of estimated increase. Given the projections are comparing AB1 with A2 and B2 that result would be expected. The overall average increase against the baseline across the 40 years from 2020 to 2050 is ~10%. Despite rainfall increases, zones 3 to 6 are projected to experience reduced rainfall and runoff during the dry season. Southern Laos, Northeast Thailand, Central and Southeastern Cambodia, including the Tonle Sap catchment and the Delta region are still susceptible to high water stress during the dry season. More dry spells and a greater severity in drought periods are expected. The past trend of increasing variability with greater extremes between wet and dry seasons, especially in southern and eastern areas, it projected to continue. The projected 2030 increases in temperature, rainfall and runoff with more extreme climate events will influence the productivity of economic sectors and livelihoods. Overall agricultural productivity will increase in the basin (around 3.6% by 2030) but food security will decrease, despite the increasing areas under irrigation. Those decreases are due to reduced dry season rainfall and runoff in central and southern zones, increasing populations and reduced production in excess of demand and increasing saline intrusion in the Delta due to storm surge and tidal influences and decreases in dry season rainfall and runoff. 238 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

239 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Overall fish biodiversity and stability in fisheries sector production is expected to decrease in the basin despite some climate change benefits of increasing flooded area and nutrient loading. The decreases are due to the complex interplay between decreased agricultural productivity and food security increasing demand and pressure on fish populations, increased riparian populations, reduced fish migration and aquatic biodiversity in zone 1 and in Mekong tributaries due to dam and infrastructure construction, and reduced and disturbed habitat due to a combination of climate change and development. The benefits to productivity of increased nutrients due to increased runoff and erosion with climate change may be offset by reduced sediment due to China and tributary dams, especially in the central highlands of Vietnam. Hydropower: Overall the hydropower sector will benefit from climate changes from increased capacity in basin catchments, but there are risks. Increased rainfall, runoff and flow throughout basin would increase potential capacity of hydropower generation in both the tributaries and mainstream. for hydropower. Some catchments will experience very high increases in runoff and water volume possibly beyond the capacity of existing tributary dam schemes creating risk of failure and need for retrofitting. Increase in extreme wet events and incidence of flood events brings a risk of catastrophic failure (climate change may turn a 1 in 10,000 year flood risk into a more regular event for example to a 1 in 1,000 flood). Livelihoods: Aquatic and terrestrial natural systems are under increasing stress in the Mekong basin. While there are benefits, overall climate change will increase that stress by increasing the need to make agriculture more productive and extensive and by increasing pressure to exploit aquatic resources. Overall reductions in fish habitat, feeding and nursery areas and increasing water stress in some catchments and the frequency and intensity of drought periods will all have knock-on effects on livelihood activities. Other developments, such as hydropower dams, intensify natural system stress and the negative effects of climate change. Climate changes such as temperature and rainfall increases and increased incidence of flooding will also increase health risks which would reduce labour productivity and increase levels of poverty EXPECTED DIRECT EFFECTS OF THE PROPOSED LMB MAINSTREAM HYDRO PROJECTS Planning and design of the LMB mainstream projects has not taken climate change into account. Variation in modeling results and unreliable baseline information increase the uncertainties of climate changes and their effects. Yet, even if the most modest projected changes are considered rather than ranges between extremes the relationships between the proposed mainstream projects and climate change is likely to be significant. The effects fall into three categories: (i) the direct effects of the projects on climate change, (ii) the direct effects of climate change on the projects and (iii) the complex inter-linkages between climate changes, the mainstream dams and other sectors. The effect of the projects on climate change is through the emission of green house gases and clearing of forests which act as carbon sinks. The direct effects of climate change on the projects are through increased runoff and flow and in extreme flood events. 239 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

240 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E 11.4 EFFECTS OF THE LMB MAINSTREAM PROJECTS ON CLIMATE CHANGE - GHG EMISSIONS 92 Climate change has mitigation and adaptation dimensions. Mitigation as it relates to human induced climate change is concerned with the net potential impact of the mainstream dams and hydropower generally on regional GHG emissions. It is well document by the UN Clean Development Mechanisms (CDM) that there are two primary consideration related to hydropower and GHG mitigation (i) thermal power emissions (coal, oil and natural gas) avoided when hydropower displaces thermal generation, and (ii) various potential GHG emissions including co2 and methane from reservoirs. In Section 4 of this report under the energy and power theme the calculation of thermal avoided emissions (or gross emission reduction are presented). The calculation is base on the specific type of thermal project that hydropower or hydropower imports would displace in each country and is relatively straight forward. The CDM also provides methodologies to make this calculation. The results indicated that the BDP 20-year probable future (PF) scenario would result in gross GHG emission reductions of some 42 million tones of CO2 annually in the year The 20-year PF scenario with all LMB mainstream dams would result in gross GHG reductions of some 94 million tonnes of CO2 annually in The balance of this paper focuses on estimating the GHG emissions that LMB reservoirs for hydropower schemes may produce, recognizing that this is a much more complex undertaking than calculating gross emissions because of the bio-physical process and a highly contested nature of the results to date world wide as documented by the World Commission on Dams in 2000 and more recently by the International Rivers Network and the International Hydropower Association. Reservoir area is an important factor influencing total GHG emissions from hydropower projects in the Mekong Basin. Only 11% of the region s existing and planned reservoir area to 2030 (i.e. BDP 20 Year Scenario with mainstream dams) is located in the UMB all along the mainstream Mekong. Proposed LMB mainstream projects cover 27% while tributary projects will make up 62% of total reservoir area. That difference in reservoir area is reflected in the projected GHG emissions to 2030 (Table 1). The ecological zones in which reservoirs are located are also influential especially slope, temperature and land cover characteristics. Mainstream reservoirs located in the more temperate mountainous areas of zone 1 in Yunnan Province (UMB) release less GMGs than mainstream dams proposed for Laos ie in tropical forest zones 2 and 3, which in turn release less than proposed mainstream projects in Cambodia in the tropical lowland and flood plain areas of zone 4 (Figure 3) TOTAL GHG EMISSIONS The two proposed projects in zone 4 Sambor and Stung Streng with their relative large lowland reservoirs, would contribute 47% of GHG emissions from the existing and proposed mainstream projects 92 ICEM collaborated with Matti Kummu of the Water & Development Research Group, Aalto University, Finland on the GHG emission estimates. The full results of the emissions work is to be published in Kummu, M., Räsänen, T. and Varis, O. submitted Greenhouse gas emission estimations for existing and planned reservoirs in the Mekong Basin. Environmental Science & Technology 240 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

241 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E from Yunnan Province to the sea (Figure 3). Those two projects would release close to 80% of average total GHG emissions from the eleven proposed LMB mainstream dams with Sambor releasing more than any other Mekong Basin hydropower reservoir (Figure 4). Figure 3: Total emissions from existing and proposed mainstream Mekong hydropower projects 2030 TOTAL EMISSIONS [t/yr - CO2eq (100-yr)] Jinghong 3% Dachaoshan 1% Xiaowan 3% Manwan 0% Latsua 1% Ban Koum Gonguoqiao7% 0% Nuozhadu 16% Sambor 35% Pakchom 3% Sanakham 3% Paklay 3% Xayabuly 2% Luangprabang 4% Pakbeng 5% Don Sahong 0% Stung Streng 12% Figure 4: Average total GHG emissions from proposed LMB mainstream hydropower reservoirs by hydro-ecological zone 2030 (modified from Kummu et al. submitted) AVERAGE TOTAL EMISSIONS [t/yr - CO2eq (100-yr)] Pakbeng Latsua Sambor Zone 2 Zone 3 Zone 4 Sambor Stung Streng Don Sahong Latsua Ban Koum Pakchom Sanakham Paklay Xayabuly Luangprabang Pakbeng 5,522 27, , , ,787 94,572 86, , , ,644 1,256, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

242 Sambor Nam Theun 2 Nam Ngum 1 Nuozhadu Xekong 4 Ubol Ratana Stung Streng Nam Pha Theun hinboun exp. (NG8) Xekaman 1 Duc Xuyen Nam Ngum 2 Battambang 1 Nam Theun 1 Ban Koum Sirindhorn Pursat 1 Nam Tha 1 Pakbeng Nam Suang 2 Nam Ngiep 1 Lower Sesan 2 + Lower Srepok 2 Xe Kong 5 Se Sesan 4 Plei Krong Luangprabang Yali Xepian-Xenanmoy Pakchom Xiaowan Sanakham Jinghong Lam Ta Khonk P.S. Nam Ou 7 Paklay Nam Lik 2 Xayabuly Buon Tua Shrah Nam Ngum 3 Houayho Xe Kong 3up Pursat 2 Nam Beng Nam Ou 6 Nam kong 1 Nam Lik 1 Nam Leuk Sre Pok 3 Nam Ngum 5 Nam Mang 3 Latsua Hoya Lamphan Upper Kontum Nam Suang 1 Dachaoshan Nam San 3 Xekaman 3 Manwan Nam Pung Gonguoqiao Don Sahong Nam Phak [t/yr - CO2eq (100-yr)] M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Even when all existing and proposed mainstream and tributary projects in the Mekong Basin are considered, Sambor and Stung Streng stand out as two of the seven top GHC emitters (Figure 5). Figure 5: GHG emissions from existing and planned hydropower project in the Mekong Basin (2030): LMB mainstream dams are presented in orange Mekong Basin hydropower: Emissions total 1,300,000 1,250,000 1,200,000 1,150,000 1,100,000 1,050,000 1,000, , , , , , , , , , , , , , , , , , ,000 50,000 0 (modified from Kummu et al. submitted) Given there are 90 existing and proposed LMB tributary reservoirs compared to 11 LMB mainstream projects, it is understandable that total tributary to mainstream project GHG emissions are 3:1 (Figure 6). Yet, on an average project by project basis the ratio is reversed with average tributary project emissions 64% the level of mainstream project emissions. Figures 6 and 7: Total and average GHG emissions from existing and planned Mekong Basin hydropower projects TOTAL EMISSIONS [t/yr - CO2eq (100-yr)] MAINSTREAM TOTALS YUNNAN TOTALS TRIBS TOTALS AVERAGE EMISSIONS [t/yr - CO2eq (100-yr)] MAINSTREAM TOTALS YUNNAN TOTALS TRIBS TOTALS TRIBS TOTALS 7,829,578 TRIBS TOTALS 173,991 YUNNAN TOTALS 828,073 YUNNAN TOTALS 138,012 MAINSTREAM TOTALS 2,699,748 MAINSTREAM TOTALS 269, I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

243 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E EMISSIONS PER POWER UNIT The difference in performance is more pronounced when average emissions per unit of power generated is calculated. The LMB mainstream projects would emit close to 4 times more GHGs/GWH than the UMB mainstream projects. Existing and planned LMB tributary dams on the other hand, would emit ten times more GHGs/GWH than the LMB mainstream projects (Table 1 and Figure 8). Most of the tributary reservoirs involve the clearing of dry or wet tropical forest and a reduction of carbon sink capacity in the basin (Figure 7). Relatively little forest is lost in the LMB and UMB mainstream reservoirs. That marked difference in GHG emissions/gwh between LMB tributary and LMB mainstream projects is explained by the much higher flow and generation capacity on the mainstream. On average the mainstream projects would generate 48 power units/km 2 compared to 14 power units/km 2 for tributary projects. Table 26: Total reservoir area, total annual energy, average GHG emissions and average GHG emissions per unit of power for the three main reservoir groups. Group UMB mainstream LMB mainstream Area Annual energy km 2 GWh Avg. emissions 10 3 t/yr-co 2 eq(100-yr) , ,500 2, Tributary ,400 7, TOTAL ,700 11, (modified from Kummu et al. submitted) Avg. emissions per unit of power t-co 2 eq(100-yr)/gwh 243 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

244 Battambang 1 Pursat 2 Duc Xuyen Theun hinboun exp. (NG8) Nam Pung Nam Ngum 1 Nam Pha Nam Beng Pursat 1 Plei Krong Xekaman 1 Nam Suang 2 Nam Tha 1 Xekong 4 Buon Tua Shrah Nam Leuk Nam Mang 3 Nam Lik 2 Nam Lik 1 Nam Theun 2 Xe Kong 5 Houayho Nam Ngiep 1 Nam Theun 1 Nam Ou 7 Nam Ngum 2 Nam Suang 1 Xe Kong 3up Nam kong 1 Se Sesan 4 Hoya Lamphan Stung Streng Sambor Xepian-Xenanmoy Lower Sesan 2 + Lower Srepok 2 Nam Ou 6 Nam Ngum 5 Nam San 3 Nam Ngum 3 Yali Pakbeng Sre Pok 3 Ban Koum Luangprabang Nuozhadu Sanakham Pakchom Upper Kontum Paklay Xekaman 3 Nam Phak Xayabuly Jinghong Latsua Xiaowan Dachaoshan Don Sahong Gonguoqiao [t/yr - CO2eq (100-yr)/GWh] M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Figure 8: Emissions per unit of power for existing and planned hydropower project to 2030 in the Meking Basin: LMB mainstream dams are presented in orange 2,200 Mekong Basin hydropower: Emissions per PU 2,000 1,800 1,600 1,400 1,200 1,000 average PU emissions from a coal plant are 921 t/yr - CO2eq (100-yr)/GWh (modified from Kummu et al. submitted) 244 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

245 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Figure 9: Total reservoir area per land cover affected in the Mekong Basin (Source: Kummu et al. submitted) Of the LMB mainstream projects, Sambor and Stung Streng in hydro-ecological zone 4 emit on average four to six times more GHGs per power unit than the nine projects in Lao PDR (Figure 8) EMISSIONS BY ECOLOGICAL ZONE The SEA adopted a hydro-ecological zonation of the mainstream Mekong and project that through the basin according to tributary sub-catchments feeding into each mainstream zone. Total emissions for mainstream projects were estimated for each hydro-ecological zone and by all measures zone 4 in Cambodia becomes the highest GHG emitter (Figure 10). 245 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

246 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Figure 10: Average total emissions for each unit of power by hydro-ecological zone AVERAGE TOTAL EMISSIONS per PU [t/yr - CO2eq (100-yr)/GWh] Pakbeng Latsua Sambor Zone 2 Zone 3 Zone 4 Sambor Stung Streng Don Sahong Latsua Ban Koum Pakchom Sanakham Paklay Xayabuly Luangprabang Pakbeng ZONE M'stream projects No AVERAGE EMISSIONS per PU [t/yr - CO2eq (100-yr)/GWh] M'stream Average (modified from Kummu et al. submitted) To compare that performance with tributary projects and according to FAO land cover cateories, all existing and proposed projects in the Basin were overlaid on the land cover classification: Temperate mountain system (TeM) Subtropical mountain system (SM) Tropical mountain system (TM) Tropical dry forest (TAwb) Tropical moist deciduous forest (TAwa) Tropical rainforest (TAr) Figure 9 shows that all the LMB mainstream projects fall within three FAO land cover categories Tropical dry and moist deciduous forest and rainforest. Annual average emissions per power unit are highest in the tropical dry forest area. The same applies to emissions from tributary projects EMISSIONS COMPARED WITH OTHER POWER SOURCES Mainstream project emissions per power unit are low when compared with other dominant sources of power such as coal, gas and diesel (Figure 10). The average PU emissions from a coal plant is 921 t/yr - CO2eq (100-yr)/GWh. Sambor and stung treng have emissions of t/yr - CO2eq (100-yr)/GWh, and these are the highest for the LMB mainstream projects. Some of the LMB mainstream projects emit less than solar. All emit less per power unit than more conventional sources. Some tributary dams stand out emitting more than coal, diesel and natural gas (Figure 8): 1. Battambang 1 >2,000t/yr - CO2eq (100-yr)/GWh 2. Pursat 2 > 1,600 t/yr - CO2eq (100-yr)/GWh 3. Theun-Hinboun exp. (NG8) Also, there is a group of three tributary projects which have emissions comparable to the range for coal: 246 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

247 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E 1. Nam Pung 2. Nam Ngum 1 3. Nam Pha Figure 11: Annual emissions per power unit for each reservoir and ecological zone (modified from Kummu et al. submitted) In summary, because of their large capacity, location and reservoir size, the mainstream projects in Cambodia have high total GHG emissions relative to other hydropower projects in the basin. Yet, overall, the mainstream dams are more attractive than most alternative power sources, other than wind and nuclear, for their relatively small impact on climate change. 247 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

248 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E Figure 12: Comparison of average emissions per power unit between power sources[t/yr-co2eq (100- yr)/gwh] UMB mainstream Wind Nuclear Tree plantation LMB mainstream Solar LMB Tributaries Natural gas Heavy fuel oil Col Lignite , ,000 1,200 1, DIRECT CLIMATE CHANGE EFFECTS ON THE MAINSTREAM PROJECTS INCREASED RUNOFF, FLOW AND FLOODING The most important climate change effects on the mainstream projects would be (i) the increase in runoff during the wet season bringing with it increases in sediment load, (ii) an increase in annual average flow in the range of 9% (MRC) to 22% (CSIRO) taking into account UMB and LMB mainstream and tributary dams and (iii) an increase in the incidence, depth and duration of extreme flood waters. The MRC modeled increases in flow are for monitoring stations on the Mekong mainstream so the LMB mainstream reservoirs would definitely be affected. MRC modeled 40 years to 2050 and averaged climate and hydrological parameters over that period. With the BDP 20 YR scenario including the LMB mainstream projects and additional tributary reservoirs, MRC confirmed the broad trends of the CSIRO projections of increasing wet and dry season mainstream flow close to 9% annual average, less than the CSIRO figure. MRC found that flooding incidence would increase with a 12-82% change in depth in the floodplain for A2 and a 22% increase in flood duration. There would be an increase in areas in the Delta affected by saline intrusion in the range 249 to 1,882 km 2 or a 1.4% (B2) to 10.5% (A2) increase. Unlike the UMB mainstream reservoirs which can store and release large volumes of water, the LMB mainstream projects would have no capacity to moderate the increases in flow and flooding due to climate change. Under the LMB 20 year scenarios, there are 70 tributary dams (30 additional to the Definite Future Scenario) with a live storage of 20,185 MCM or 4.2% of Mekong water and 6 Chinese dams with a total active storage making up 13.5%. The impact of climate change on the frequency of flood events of different magnitude is not well understood, but all studies to date project increases in frequency and magnitude. Depending how the cascades of UMB and LMB tributary dams are managed, the expected increases in flow due to climate change could be moderated to some extent or increased at any point during the wet season. LMB mainstream dams could be subject to seasonal and extreme event peaks in flow well beyond their normal operating design standards. 248 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

249 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E There has been no modeling of the sediment load implications of climate change for the LMB mainstream dams. With a projected 21% annual increase in runoff from basin catchments, soil erosion and erosion of river banks and channels is likely to increase affecting sediment load and water quality in mainstream reservoirs. Increased runoff and expected use of pesticides in the basin could also lead to increased levels and accumulation of chemicals in mainstream reservoirs RISK OF EXTREME EVENTS AND DAM FAILURE Annex VII (Volume III) provides the details of the analysis of likelihood of extreme events with climate change and how that relates to the mainstream dams. Comparing the historic and future climates the following conclusions can be drawn: The historic 100 yr event is smaller than the 1 in 10yr event with the projected climate change at each station analysed. This means that an event which occurred every 100years is likely to occur more than once every 10 years. The historic 1,000yr event is comparable in size to the 1 in 100yr event at each station with climate change. This means that the 1 in 1,000yr event will become ~ 1 in 100year event. The historic 1 in 10,000yr event is comparable to the 1 in 1,000year event with project climate at each station. This is more pronounced for downstream stations Table 2: comparison of changes to the magnitude of extreme events for the same return period EXTREME VALUE DISTRIBUTION Historic Return period flow (EV dist) EXTREME VALUE DISTRIBUTION Project 2030 Return period flow (EV dist) with CC Station 10yr 100yr 1,000yr 10,000 10yr 100yr 1,000yr 10,000 Chiang Saen 12,252 14,551 16,808 19,061 13,209 15,769 18,282 20,790 Luang Prabang 17,137 19,912 22,637 25,357 18,783 22,362 25,876 29,384 Vientiane 18,670 21,285 23,852 26,414 19,692 22,745 25,742 28,734 Pakse 40,842 45,344 49,765 54,177 43,459 49,149 54,734 60,311 Kratie 56,254 62,934 69,493 76,040 59,000 66,886 74,629 82,358 Assuming a mainstream project design life of 100years: Projects designed for a 1 in 100 year event will see the probability of this event occurring over the design life increase from 63% to ~100% (e.g. Pak Chom, Ban Koum) Projects designed for a 1 in 1,000 year event will see the probability of this event occurring over the design life increase from 10% to 63% (e.g. Pak Lay) Projects designed for a 1 in 10,000yr event will see the probability of this event occurring over the design life increase from 1% to 10% (e.g. Luang Prabang, Xayaburi, Lat Sua) Table 3: Risk of extreme event occurrence associated with return periods (T) for an assumed project life of 100yrs T ,000 2,000 5,000 10,000 50,000 RISK % % % 86.74% 63.40% 18.14% 9.52% 4.88% 1.98% 1.00% 0.20% 249 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

250 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E 11.6 INDIRECT LINKS BETWEEN CLIMATE CHANGE, MAINSTREAM PROJECTS AND OTHER THEMES There is an array of important relationships when climate change, the LMB mainstream dams and other major trends in the region are overlaid REDUCED FOOD SECURITY There are various trends which have a synergistic effect in reducing food security in the LMB catchments linked to the mainstream projects. The population of the LMB is expected to double by Food demand would double from ~17 million tonnes for 2000 to ~ 33 million tonnes for the 2030 projected population (CSIRO 2009). With climate change overall agricultural productivity in the LMB is expected to increase by 3.6% due mainly to increased rainfall. Yet, some catchments linked to the LMB mainstream projects will experience reduced rainfall during the dry season, and reduced productivity. Overall, CSIRO conclude that excess production above food demand will be reduced by 56% to ~11 million tones. Even for those catchments with increased productivity, population increases will maintain or create food scarcity situations.93 Another factor working to reduce agricultural productivity will be the effect of the China and LMB mainstream dams on sediment load in flood waters. The China dams are projected to take out some 50% of the total sediment reaching the sea. The LMB project will take out another 50% of that load. Reduced sediment and decreasing overbank and floodplain flooding due to the China dam operations would have immediate impacts on agricultural productivity in affected areas. The extent to which increased sediment entering the mainstream from tributaries in 2030 will compensate for this loss has still to be modeled. In those situations riparian urban and rural communities would look increasingly to the river to offset shortfalls in agricultural produce. This SEA has concluded that, during both construction and operation, the mainstream dams will reduce fish populations, diversity and overall fisheries sector productivity. Therefore the mainstream dams will further aggravate growing food scarcity in the LMB. Agricultural productivity will benefit from increased access to irrigation water from mainstream and tributary reservoirs. Yet, CSIRO found that the irrigation requirement for crops grown in the dry season would be greater for all catchments under the likely future climate a demand which would need to be met through the increased runoff and reservoir water access. Irrigation systems would need to be designed to meet the increased crop water demand REDUCED WATER QUALITY The LMB mainstream reservoirs are relatively shallow when compared to the UMB projects (ie 10-60m compared to m). Even so, 30-90% of their total reservoir volume is dead storage which during the dry season will affect the rate of flow in the river and turnover of the water body. That will be so especially for the large dams in Cambodia where the river is losing its velocity and dead storage is greater. With climate change the total sediment and organic load entering the mainstream from tributaries would 93 Food scarcity here means a deficit between availability and demand for agricultural production within a catchment (CSIRO 2008) 250 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

251 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E increase. That additional load would be moderated as more tributary dams are built, but - when combined with increasing populations and land use changes and disturbance overall sediment and organic loads reaching the mainstream are likely to increase. Falling agricultural productivity is likely to lead to increased application of agricultural chemicals from relatively low levels now. Also industries are expected to develop along the mainstream and tributaries within a 50 km radius. Those industries and the expanding number and size of urban settlements are likely to increase the pollution loading in the mainstream. The combination of increased sediment and organic loading with increased chemical and organic wastes in certain reaches of the mainstream within reservoirs will act to concentrate and accumulate pollutants bound to organic material, in sediments and in solution. Already, this is becoming a significant problem in the Mekong Delta adjacent to and downstream of industrial and urban areas. Anything structure which impedes flushing and turnover of water volume and sediment will aggravate this growing problem. The mainstream dams would act to concentrate and accumulate pollutants during the dry season. During the wet, thorough flushing is likely although not certain in some reservoirs. For the northern cascade in Zone 2, for example, the full supply level is much higher than the highest recorded water levels near the dam wall (almost twice as high at the dam wall for Luang Prabang). Depending on the location of deep pools, the depth of the water column could be 30m (dam wall) m (deep pool) in which case there may also be mixing and flushing constraints during the wet LOSS OF BIODIVERSITY In aggregate, it is likely that climate changes will lead to losses in aquatic and terrestrial biodiversity. Overall reductions in aquatic habitats and shifts in terrestrial ecosystem conditions will combine with human induced changes and disturbances to reduce the basin s capacity to adapt to change. The hardening of land surfaces and of river banks and channels in particular will constrain natural system adaptation to climate change. The mainstream developments would be a significant contributing force in that progressive hardening process. They would act to simplify the aquatic and surrounding terrestrial systems, reducing their diversity, resilience and stability in the face of shocks and stresses CONSTRAINTS TO POVERTY REDUCTION Reduced food security, water quality and the diversity and stability of natural systems would make conditions for poor riparian communities more difficult. Losses in aquatic and agricultural productivity would increase food prices and access. Losses in fisheries productivity would limit livelihood and subsistence opportunities. Natural system instability, including landslides, bank collapse, soil erosion and degradation of biodiversity would all affect the poor first and most significantly. The mainstream dams could bring benefits to local poor communities but, when combined with climate change and other trends, a number of influences of the projects would work against poverty reduction INCREASED POTENTIAL POWER CAPACITY ACROSS THE BASIS The indirect relationships between climate change and mainstream projects also needs to be overlaid national and LMB power demand. The increased runoff in all LMB catchments with climate change would increase hydropower potential in tributaries and in the mainstream through increased river flow. The mainstream hydropower projects are being considered as part of LMB country effort to meet growing 251 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

252 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E power demand. While the effects of increased runoff on power capacity of tributaries have not been modeled in detail, estimates would suggest that it has the potential to offset the 16% of regional power production estimated for the mainstream projects. Many other factors would come into play in realizing that potential, most notably economic and technical feasibility SENSITIVITY ANALYSIS OF DAM GROUPINGS In undertaking this sensitivity analysis, influential factors are the number of tributaries by hydroecological zone and their flow contribution to the mainstream in addition to the area and geology of their catchments (Figures 11 and 12). Related factors include climate change effects on agricultural productivity in various tributary catchments, effects of the different project groupings on local fish migration and fisheries productivity, and the likely concentrations of urban and industrial development and populations within reservoir catchments CASCADE OF 6 DAMS UPSTREAM OF VIENTIANE (ZONE 2) Increased flow: Total tributary contribution to water flow in this zone is 14% (Figure 12). It will be the zone least affected by projected increases in runoff due to climate change. That contribution is expected to be modulated by a relatively large number of planned tributary dams. Food security: 77% of the mainstream within Zone 2 will be changed by the six reservoirs, with an 82% loss in seasonally exposed in-channel wetlands and a significant loss in aquatic systems productivity and diversity. All catchments in this zone are expected to suffer losses in agricultural productivity and food deficits. Water quality: The number of dams will have localized effects on flow especially during the dry season which in those reaches of the river would aggravate the concentration and bio-accumulation of anticipated increases in agricultural, industrial and urban chemical pollutants MIDDLE MEKONG DAMS (ZONE 3) Increased flow: Zone 3, with the largest area and a total tributary inflow of 38% will be most affected by projected increases in runoff and sediment due to climate change. Proposed mainstream projects in this zone will be exposed to significant increases in annual volumes of water and in frequency and magnitude of wet season extremes in flood events. Food security: This zone has some of the catchments most seriously affected by reductions in dry season rainfall and agricultural food deficits. In Zone 3, about 22% of the mainstream length will be changed by the reservoir, with about 42% of the seasonally exposed in-channel wetland areas being lost leading to overall reductions in aquatic systems productivity. Water quality: The large areas of agricultural land in this zone, the most numerous settlements and increasing pockets of industrial development create potential for relatively high pollution levels in runoff entering mainstream reservoirs. 252 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t

253 M R C S E A V O L U M E I I M A I N I M P A C T S A S S E S S M E N T R E P O R T C L I M A T E C H A N G E CAMBODIAN DAMS (ZONE 4) Increased flow: 27% of flow is contributed to this zone of the river from three major tributaries in relatively steep terrains. The velocity of runoff in extreme rainfall events will be modulated by intensive development of tributary dams. This zone includes catchments in the Vietnam Central Highlands that contribute around 15% of total sediment to the river. That relatively high sediment load and the projected increase due to climate change will be reduced by the tributary cascades. Food security: In Zone 4, about 43% of the mainstream length will be affected by the reservoirs, with a loss of about 58% of the in-channel wetland areas with a very significant impact on biodiversity and fisheries productivity. This zone would experience severe food deficits. Water quality: Locally sourced pollution would not be as serious in this zone but the contributions from upstream might become a problem especially for Stung Streng reservoir. Figure 13: Mean annual flow contributions to the Mekong River 11.8 SUMMARY OF IMPACTS The relationships between the proposed mainstream projects and climate change is likely to be significant. There would be direct and indirect effects. The direct effects are two way: (i) the contribution of the projects to climate change through green house gas emissions and (ii) the direct effects of climate change on the projects through increased runoff and flow and in extreme flood events. 253 I C E M I n t e r n a t i o n a l C e n t r e f o r E n v i r o n m e n t a l M a n a g e m e n t