Stormwater untapping the potential. A rapid approach to assessing avoided drainage costs

Transcription

1 Stormwater untapping the potential. A rapid approach to assessing avoided drainage costs Sestokas, K* 1, Bishop, E 2, van Raalte, L 3, Watkinson, R 4, Marshall P and Stokes, D 5 As the population of Melbourne grows, so do opportunities to capture, store and reuse rainfall and stormwater at a property scale to assist with reducing future flooding and, therefore, future drainage infrastructure costs. This project estimates increases in impervious surface areas associated with Melbourne s forecast future growth, the traditional drainage infrastructure costs associated with managing this growth, and the likely effectiveness of on-site rainwater storage for avoiding some of the future flood mitigation costs. While the analysis was constrained by limited data and time, it provides a significant step forward in terms of describing an important external benefit associated with improved stormwater management at the property scale and suggests that the potential benefits for Melbourne are significant. 1 Melbourne Water, 990 Latrobe St, Docklands VIC 3008, Kristina.Sestokas@melbournewater.com.au. 2 DEPI, 8 Nicholson St, East Melbourne VIC 3002, Emma.Bishop@depi.vic.gov.au. 3 AECOM, Level 45, 80 Collins Street, Melbourne VIC 3000, Lucas.vanRaalte@aecom.com 4 Melbourne Water, 990 Latrobe St, Docklands VIC 3008,Rod.Watkinson@melbournewater.com.au 5 Spatial Economics, 5 Loch St, St Kilda West VIC 3182, marshalld@spatialeconomics.com.au, dstokes@spatialeconomics.com.au

2 D I. Introduction EVELOPMENT associated with urban growth increases the amount of hard surface area (imperviousness) and, under current approaches to development, this generally leads to increased volumes of stormwater runoff entering the drainage system and increased risk of downstream flooding. In many areas of Greater Melbourne, existing flooding impacts are already considered unacceptable and there is a strong desire across government, flood management agencies and the communities affected, to ensure that this impact is not exacerbated by future development. The authors of this paper embarked on an ambitious project (the project) to quantify the likely avoided drainage costs of two storage volumes (2.5kL and 5kL - representing two scenarios of available tank airspace) for all new developing and redeveloping properties within Melbourne s Urban Growth Boundary (UGB). The project included estimation of current and future levels of imperviousness across all catchments, an estimation of a comparative traditional drainage infrastructure cost, assessment of catchment characteristics, a review of technical studies into the effectiveness of on-site storage with respect to flooding, and finally, an estimation of the likely effectiveness of the two storage volumes for managing increased future flooding. The assessment was undertaken under significant time and data constraints but clearly demonstrated the potential benefits of capturing (and using) rainfall in new developments and redevelopments in terms of contributing to managing future flooding and thereby avoiding some future drainage costs. II. Background The purpose of this project was to determine the impact of capturing and storing rainwater at the lot on future drainage infrastructure costs. Other outcomes of managing stormwater at the lot, which were outside of the scope of this project, include impacts on existing flooding, liveability, reduced degradation of waterways and bays and reduced mains water consumption. 2

3 III. Dimensions of Flooding There are many different dimensions and scales of flooding. This section briefly introduces some of these and outlines the specific type of flooding considered in this project. A. The Physical Scales and Types of Flooding Flooding can vary from large to small in size, frequent to infrequent in terms of likelihood of occurrence, and low to extreme in terms of impact when an event does occur. Terminology relating to flooding also varies, from 'stormwater runoff', to 'nuisance' or localised flooding, to extreme flooding, usually depending on the perspectives of the different individuals involved in the discussions, and usually related to the size and frequency of the rainfall (and hence flood) event. Whilst WWCM approaches are generally discussed in the context of managing smaller, more frequent stormwater runoff or nuisance flooding events, the majority of the available studies on the impacts of on-site storages on flooding examined the larger flood events. For this reason, the team focused on increases to current 1% Annual Exceedance Probability (AEP) flood extents associated with MW's regional and council's local drainage infrastructure. This is also the event around which Victorian Planning Schemes, development advice, drainage scheme planning, and flood mitigation design are typically based. Examining only the avoided costs associated with mitigating increases to the 1% AEP flood extent introduces significant conservatism to the assessment as the available research does suggest that the impacts of distributed storages are likely to be much higher for the more frequent, smaller events. B. The Temporal Scales of Flooding: Existing and Future Prior to the 1970s, development around greater Melbourne occurred with only minimal regard to overland flows paths. As such, a significant proportion of Melbourne has been built in valleys and old creek beds, and large parts of the city are at risk of flooding. The size of the existing flooding problem is significant, with over 130,000 properties affected by flooding along MW's regional drainage systems (MW, 2014), and over 26,000 properties classified as being exposed to an Extreme risk of flooding (defined by MW's Flood Risk Assessment Framework (MW, 2010)). Numerous buildings are also known to be at risk of flooding above floor. Many other properties along Council's local drainage systems are also known to be at risk of flooding. 3

4 Whilst the scale of existing ('legacy') flooding is significant, and it is likely that as urban areas are redeveloped over time distributed storages could have an impact on reducing legacy flooding in some areas, due to the limited time and studies available, this project examined only the avoided increases to flooding resulting from population growth and urban consolidation. It should be noted that whilst all new greenfield developments are now (since the 1970s) carefully managed and designed with stormwater and overland flow paths in mind, there is currently no consistent approach to infill developments, with these types of developments primarily being managed by Councils, subject to their individual policies and requirements. Improving rainwater management has the potential to significantly assist with managing future flooding, as well as providing a means to better manage impacts at the time that they occur. IV. How Much Future Flooding is there Likely to Be & What Would it Cost to Manage it? A. Estimating Increases in Imperviousness Metropolitan Melbourne is rapidly growing, with an increasing amount of development occurring within the existing established areas. This means an ongoing increase in imperviousness in all suburbs as houses/lots are subdivided, rebuilt or extended. Two methods for estimating increases in imperviousness were used. The first was undertaken by Spatial Economics (SE) (Marshall and Stokes, 2014) and used population growth estimates provided by DTPLI. This method determined increases in imperviousness in ten year increments to 2051 and was primarily undertaken to assist with the costing and economic analysis aspects of the project. The second method, undertaken by MW, was based on Planning Scheme Zones and development potential, and provided a basis from which to estimate long-term flooding impacts and traditional drainage requirements across catchments. Further information on the two approaches is provided in Appendix A. 4

5 B. Calculating the Costs of Managing Future Flooding in a Traditional Way 1. Across the regional drainage network Over the course of MW's Redevelopment Services Schemes (RSS) program ( ), over 50 investigations were undertaken to model and assess the potential impacts of urban consolidation on flooding across MW's regional drainage systems, and to design and cost infrastructure solutions to manage the impacts relative to existing 1%AEP flood extents. These studies provide a cost per additional impervious hectare within each catchment, calculated as the total cost of the drainage infrastructure divided by the total estimated increase in imperviousness (in hectares). The costs varied from $59,100 to $283,000 per additional impervious hectare, with an average rate of $151,880 (MW, ). These costs, grouped into average costs per municipality and adjusted to 2014 dollars, were then used as the basis for estimating a "traditional infrastructure approach" cost for managing future regional flooding due to population growth and urban consolidation. 2. Across local drainage networks In addition to the regional drainage assets discussed above, increased imperviousness is expected to impact on the costs of meeting drainage standards for catchments that are smaller than 60 hectares, which are the responsibility of local government. Unfortunately, detailed local government drainage upgrade costs are not easily obtainable, which made it difficult to correlate drainage infrastructure costs against increased imperviousness in the manner described for regional assets above. However, annual financial reports provide some indication of the year-on-year expenditure by individual councils. For each council, the spending on drainage assets was separated from road and other capital expenses as best as possible, and then correlated with increases in impervious area over matching time periods to provide an indication of the level of local government spending. Expenditures may lag or precede impervious area increases, meaning that the expenditure associated with a particular increase in imperviousness may not always occur in the same time period. However, as averages were used, this is not expected to have a significant impact on the estimated costs per additional impervious hectare. 5

6 Based on this assessment, the estimated local government costs range from $9,000 to $1,319,000 per additional impervious hectare depending on location, with more than half having expenditures falling between $100,000 and $400,000 per additional impervious hectare. V. What Benefits Could On-Site Storage Provide for Future Flooding? C. Literature Reviews to Understand Tank Effectiveness in Different Catchments A literature review was undertaken to understand the potential for on-site storages to contribute to managing increases in future 1%AEP flooding and to develop relationships between catchment characteristics and storage effectiveness. The review revealed that the factors that have the most influence on storage effectiveness, in the majority of cases, are: 1. catchment size, 2. catchment topography, 3. the proportion of existing and forecast development (in terms of imperviousness); and 4. the nature of the catchment outlet (free-flowing 'unconstrained' or 'constrained'). Based on the review, relationships to link these factors with the on-site storage volumes required to manage all future increases in imperviousness for a particular type of catchment were developed by MW. The relationships were then extrapolated to cover types of catchments where no specific study data was available, as presented in Appendix B. D. Classifying Catchments across Greater Melbourne & Estimating Tank Effectiveness In order to apply the relationships (described above) to all catchments within the UGB, MW assessed and categorised each catchment based on the above four key factors. This process involved calculating catchment areas, approximating times of concentration, and visually identifying and assigning an outlet condition type for over 500 catchments within the UGB. The proportion of each catchment within different municipalities was also calculated (for later cost estimation purposes), and existing and future impervious areas were 6

7 assigned using MW's RSS calculator process (Appendix A). Further details on assumptions and opportunities to refine the methodology are provided in Appendix C. E. Estimating Tank Effectiveness across greater Melbourne Estimating tank effectiveness across the UGB involved linking the calculated 'required' on-site storage volumes (Section C) with the catchment characterisation data (Section D). A percentage effectiveness ratio, the ratio of proposed tank volume (either 2.5L or 5kL) to the required tank volume, was then attributed to each catchment. These ratios were then multiplied by the forecast increases in imperviousness to estimate the percentage of avoided imperviousness for each catchment, and hence each municipality (Appendix D). VI. So What Does this all Mean? What Costs Could Be Avoided? The final stage of the project involved bringing together MW's tank effectiveness data, SE's ten-yearly impervious area forecasts, rates of building stock turnover, and the traditional regional and local drainage infrastructure costs to estimate the total avoided costs of implementing the two stormwater management options (in comparison to Business As Usual, BAU). Under BAU, it was assumed that 30 percent of new developments installed rainwater tanks 6. Under both storage options, it was assumed that all new developments and redevelopments installed rainwater tanks. Based on the regional and local drainage infrastructure costs outlined in Section IV, the total cumulative cost of additional drainage infrastructure associated with managing population growth and urban consolidation by 2051 was found to be almost $9.8 billion, with a present value of $4.1 billion 7. Under the on lot storage scenarios, development projections were used to estimate when and where storage tanks would be installed. The reduced costs per increase in impervious area (described in Section IV) were then assigned to these developments. Based on this, the total cumulative cost for the 2.5kL option was calculated to be $6.4 billion out to 2051, with a present value of $2.7 billion7. The benefit of including tanks 6 Based on Victorian Building Authority building permit data 7 Discounted at 6% per annum 7

8 in all new developments (as opposed to the current rate of 30% of new homes) that, on average, have 2.5kL of available storage at the time of a storm, was therefore approximately $3.4 billion as a cumulative cost out to 2051, with a present value of $1.4 billion 7. VII. Conclusion This study clearly demonstrates the potential avoided future regional and local drainage costs associated with capturing and storing roof runoff as and when new developments and redevelopments occur. The present value avoided infrastructure costs estimated through this project range from $610 to $750 million in regards to local drainage infrastructure, and between $650 to $790 million for regional drainage infrastructure. Further review and refinement of these estimates are planned for the coming months. 8

9 Appendix A - Estimating Impervious Area Increases From Population Growth Estimates and Land Use Maps To estimate increases in imperviousness due to population growth, an assessment was first undertaken of previous population growth statistics, changes in dwellings ( ), different development types and their related increases in impervious surfaces, and land budgets. The assessment included the consideration of both residential and non-residential areas, as well as the categorisation of residential areas into dispersed, high density and broadhectare (representing currently undeveloped areas within the UGB) development regions. Numerous assumptions were made during the assessment, including that existing high density residential, commercial and industrial areas would be unlikely to increase in imperviousness in the future, that development in the 'dispersed' residential areas would include a mix of subdivisions, rebuilds and extensions, with specific redevelopment ratios applying to each municipality, and that vacant industrial land would develop in a similar manner to previous industrial developments in each municipality. Using the results of the retrospective analysis and new dwelling and population projections to 2051, forecast changes in imperviousness were then calculated in ten yearly increments for each municipality. From Planning Scheme Zones and Development Potential From 2004 to 2013 MW managed a Redevelopment Services Schemes (RSS) program, which collected financial contributions from developing properties to assist with future drainage infrastructure upgrades. Issues with collecting contributions from all developing properties, and difficulties scheduling capital works aligned with contributions collected and stormwater runoff impacts were some of the reasons the program was discontinued. MW's former RSS program utilised a GIS tool referred to as the "Impervious area calculator" to estimate increases in imperviousness across study catchments. This tool was utilised for this proect also, as it was able to provide a rapid estimate of existing and potential future imperviousness across the 500 plus catchments included in the assessment. The calculator assesses a modified Planning Scheme Zones layer (last updated in 2008), and assigns a current fraction impervious and future fraction impervious to each zone. It is only the residential type zones (including standard residential, rural residential, etc. zones), which are assigned an increase in fraction impervious, and this increase is linked to average existing property sizes within a

10 1hectare grid area. The relationships between property size and the potential for redevelopment and increases in imperviousness were developed as part of the establishment of the RSS program, with input from GHD consultants. No further changes to the calculator or imperviousness relationships were made for the purposes of this project. Appendix B: Literature Review Findings Table 1. Required Storage Volumes (L) to Manage Increased 1%AEP Flooding The reports and studies that contributed to the data in Table 1 are listed below in Table 2.

11 Table 2. Studies and Reports reviewed to Develop Tank Effectiveness Relationships Report title Author *Effect of rainwater tanks on storm water discharge for different storm durations (Intensity) Study on the Combined Effects of OSD and Rainwater Tanks on the Upper Parramatta River Catchment at Varying Sub-Catchment Scales. Report for the Upper Parramatta River Catchment Trust (dated August 2002) Ruwan Jayasinghe; March 2002 Cardno Willing (2002) Integrated Stormwater Treatment and Harvesting: Technical Guidance Report) Mitchell et al, 2006 Improving the Management of Urban Runoff using On-Site Detention Paper for the 46 th Floodplain Management Authorities Conference *Preliminary Assessment of On-Site Retention - Shakespeare Grove Main Drain Rainwater Tank Evaluation Study for Greater Melbourne *Can allotment-scale rainwater harvesting manage urban flood risk and protect stream health? Carse, J, Lees, S, Phillips, B.C and Goyen, A.G (2006) GHD; September 2007 Coombes, P (Bonacci, August 2008) Matt Burns, et al; Novatech 2010 Oakleigh North Drain RSS Pilot Study GHD; July 2011 *Investigation of rainwater tanks and the impact on stormwater infrastructure sizing DRAFT report (not yet finalised) *Investigation of Rainwater Tanks and the Impact on Stormwater Infrastructure Sizing in High-Density infill scenarios DRAFT report (not yet finalised) *Shakespeare Byron Tanks Modelling DRAFT results 2014 DCE, November 2011 DCE, April 2014 BMT WBM; current study Impacts of Rain Tanks in Clyde Creek Study for OLV Neil M Craigie Pty Ltd, October 2013 Literature review of the links between flooding and water sensitive urban Alluvium, 2013 design assets. For the Dandenong IW and PB Study (memo dated 31 July 2013) 1 *Fishermans Bend Study, MW internal study and memo 2014 * Key studies used to develop tank effectiveness relationships

12 Appendix C: Tank Effectiveness Assessment Methodology Effectiveness of tank storage More studies that model the effectiveness of rainwater tanks in specific catchments would improve the robustness of the analysis, and help to reduce assumptions made in the development of relationships shown in Table 1. The catchment specific tank sizes that were based on studies are shown in black in Table 1. Where studies looked at the impact of storages on flows only (ie not flooding) the following equation was used to estimate an equivalent change in fraction impervious for that catchment:. Many of the studies (though not all) also only considered what benefit a set available storage volume could provide across a catchment ie no consideration of preceding rainfall conditions, which may affect available storage volumes or consideration of tank configurations to assure the availability of the required storage volume, such as leaky tanks, etc. This would be a key consideration in appropriately designing and installing storages to ensure the availability of space for the capture of roof runoff for flood mitigation. Where studies did consider long term rainfall sequences, the majority clearly demonstrated the benefits of a leaky tank type system to ensure the required storage volumes were available at the start of a storm. Catchment boundaries Preliminary analysis was based on MW catchment boundaries at various scales across the Greater Melbourne region (ie at smaller drainage network scales where more detailed data was available to larger creek basin scales in less developed areas or where limited existing data was available). Many catchment areas straddling the UGB, as well as catchments draining satellite growth areas detached from the main body of urban development were excluded from the analysis. Some border catchments were included if the majority of the area was located within the urban growth boundary, even if they included areas outside of the UGB. Inclusion of the outside areas was assumed to make negligible difference to the overall avoidable cost estimates due to minimal increases in imperviousness outside of the urban growth boundary. Areas excluded from within the UGB could potentially increase impervious hectare estimates, and also avoided costs, however, were not considered given time constraints and other uncertainties.

13 Further to the above, many larger catchments were also not broken into smaller catchments for the estimation of impervious area increases and reductions. This could mean that many larger areas were classified as receiving no (or minimal) benefit from the proposed storage volumes, when a more detailed assessment might have shown that smaller drainage catchments within the larger area could in fact be classified as benefiting from the proposed storages and hence avoiding additional future drainage costs. Catchment areas could be broken-down into smaller sub-catchments to improve the resolution of analysis. Outlet condition Preliminary analysis was based on a rapid appraisal of outlet condition that considered whether the outfall of each catchment was constrained (eg. connecting into a downstream piped system) or unconstrained (eg. out letting to a bay or large waterway). The required storage volume for catchments with constrained outlets was assumed to be double that of an unconstrained catchment of similar area, time of concentration and fraction impervious increase (see Table 1). This assessment could be improved by modelling the effectiveness of tanks in two catchments with similar characteristics overall, but different outlet conditions. Time of concentration (as an indicator of critical storm duration) Time of concentration was estimated for each catchment based on the following: 1) A roof to gutter time of 7 minutes; 2) Average overland velocity (m/s) estimated using the Colebrook-White and Darcy-Weisbach equations with the following assumptions: o nominal pipe diameter of 450mm o 1.5 mm K roughness coefficient o an estimate of average catchment slope adjusted by a factor of two thirds to reflect generally steeper upper reaches, and flatter lower reaches.

14 o 1.5 multiplier of water travel time to reflect the generally slower travel times associated with overland flows in 100yr ARI events o a 2 times multiplier for catchments with an existing FI of less than 0.2 to allow for slower overland flow times in more rural (pervious) catchments. 3) catchment specific area; and 4) an average 4:1 catchment length to width ratio (to estimate catchment length from catchment area, hence water travel times), except for catchments visually identified as not generally conforming to this assumption, in which case an actual catchment length (representing the flow path) was measured. Catchments were grouped based on time of concentration as follows: (i) (ii) (iii) tc < 1hr 1hr < tc < 2hrs tc > 2hrs Further refinements to the above process could include: 1) Measuring catchment lengths (hence water flow path lengths) in all catchments, 2) Accessing detailed flood modelling study outputs to ascertain critical storm durations for each modelled catchment (ie the duration of storm that results in the highest level of flooding at different points within a catchment). Fraction impervious Changes in impervious area for all catchments were calculated based on an initial and ultimate fraction impervious from the (former) RSS impervious area calculator. An improvement to the adopted methodology would be to consider current (2014) Planning Schemes zones to estimate existing impervious fractions and areas. The coverage of the Planning Schemes zones layer within the UGB could also be reviewed to improve the assessment methodology.

15 Catchments were grouped based on increases in impervious area as follows: 0-17%, 18-40%, > 40%. For the analysis, further interpolation between impervious area increases (expressed as a percentage increase) and catchment areas was undertaken to estimate required tank sizes. Assignment of impervious area reductions to Municipalities The adopted assessment approach only considered the geographical distribution of drainage catchments across municipalities. A further refinement to the assessment approach could be to further consider Planning Scheme zone distributions within a drainage catchment, and therefore relative to a particular municipal boundary. Given the scale of the avoided costs project (ie greater Melbourne) this was not considered relevant for this study, however, if more detailed assessments were to be completed at a regional or municipal level this refinement would help to improve the accuracy of the estimates. Appendix D: Tank Effectiveness Results Table 2. Tank Effectiveness Results per Municipality

16 References DEPI (2014) Draft Victorian Floodplain Management Strategy, Department of Primary Industries, June DEPI (2013) Melbourne s Water Future. Published by the Victorian Government Department of Environment and Primary Industries, November DTPLI (2014) Plan Melbourne. Published by the State of Victoria, Department of Transport, Planning and Local Infrastructure, May Melbourne Water ( ) Redevelopment Services Scheme investigation reports. Numerous reports prepared by various consultants for Melbourne Water. Melbourne Water (2007) Port Phillip and Westernport Region Flood Management and Drainage Strategy. Melbourne Water Corporation, Victoria Melbourne Water (2010) Flood Risk Assessment Framework. [online], Accessed 22 August 2014 from Melbourne Water (2014) Draft Directions Consultation Paper for a new Port Phillip and Westernport Region Floodplain Management Strategy. Unpublished document under development. Marshall, P and Stokes, D (2014) Impervious Surface Area Projections Metropolitan Melbourne: 2011 to 2051, June 2014