American University of Beirut, Beirut, Lebanon

Transcription

1 DSS for Wastewater Management at a River Basin Scale Mutasem El-Fadel 1,2, Majdi Abou Najem, Rania Maroun Dany Lichaa El-Khoury, Raja Bou-Fakr Eldin, and Renalda El-Samra 1,2 American University of Beirut, Beirut, Lebanon Short Abstract This paper presents a Decision Support System (DSS) developed to assist in prioritizing investment options in the construction and operation of wastewater treatment plants (WWTPs) at a river basin scale. The DSS allows the examination of what if scenarios in the context of surface water quality management in the basin. A sensitivity analysis was conducted on key indicators used in the prioritization process and example output analysis of a basic investment scenario in wastewater treatment plants. The DSS provided an evaluation tool to assess various management plans with optimal environmental benefit for cost-effective measures. Keywords: DSS, River Basin Management, Litani, Lebanon. Introduction The introduction of Decision Support Systems (DSS) for water-resources planning dates back to nearly two decades. Similarly, over the same period of time, there has been a growing interest in the application of geographical information systems (GIS) in river basin management and the development of general and specific interactive frameworks for river-aquifer simulation. This shared-vision creates an environment that will make possible for stakeholders and decision makers to incorporate their own data and vision in the overall system description while benefiting from experts' knowledge and experience. In Lebanon, the Litani river constitutes the most important freshwater resources with its basin being subject to various sources of environmental stresses raising the need of a support system for water quality management at a basin scale. This river stretches from the North Bekaa valley to nearly 170 km before discharging into the Mediterranean Sea, draining more than onefifth of Lebanon s area, with an estimated average annual discharge rate of 770 million m 3 from a 2,168 km 2 watershed. The Qaroun lake divides the Litani watershed into two sub-basins: downstream of the lake or the lower sub-basin where the river has been relatively unexploited until 2000 due to security reasons, and upstream of the lake or the upper sub-basin where more than 90% of the overall basin s population reside and therefore, forming the focus of this study (Fig. 2). Owing to its topographical characteristics and fertile soils, this area is noted for its agricultural activities and industrial clusters, particularly in the central Bekaa (El-Fadel et al., 2011). As a result, the Litani river and underlying aquifers which constitute the main sources of water for domestic, industrial, and agricultural use, are subject to environmental stresses including wastewater discharge, solid water disposal, irrigation return, and industrial effluents. The direct discharge of untreated domestic wastewater into the river and its tributaries is by far the main contributor to the degradation of water quality (USAID 2005; MoE, 2011). Consequently, the master plan towards improving water quality in the basin envisage amongst several other programs, the construction and operation of several wastewater treatment plants (WWTPs) along the river. This paper presents a DSS Page 1 of 8

2 designed to prioritize WWTPs and to examine what if scenarios related to surface water above the lake, with the aim of assisting decision-makers in adopting informed decisions regarding water quality management at the basin scale. evaluate and rank the proposed options. Expert and experience based performance indicators (Table 1) were defined for this purpose with corresponding weighing factors to analyze and rank the options. The Matrix Analysis uses a score-range of 0 to 100 for each indicator, and assumes a highest weighed score of 100 after all scores are normalized to 100 when the total cumulative score exceeds 100. Fig. 1 Geography of the Litani basin Methods The upper basin was divided into seven catchments each to be served by a WWTP that treats the wastewater generated from the towns within the catchment, and discharges the treated effluent into the Litani River or its tributaries (Fig. 2). Prioritization of investment options in these plants relied on user-defined environmental, economic, and financial criteria with the DSS targeting water quality and cost benefit analysis besides the prioritization process. Prioritization of Investment Options The prioritization of investment options in domestic WWTPs uses a Matrix Analysis to Fig. 2. WWTPs with towns served The performance indicators used in the Matrix Analysis include: 1. Load Reduction Efficiency (LRE, d, p): is a function of the influent load (per pollutant type) and the percent pollutant reduction. Error! Reference source not found. depicts the flow for the extraction of Pollutant Load Reduction indicator. Page 2 of 8

3 Table 1. Matrix Analysis and performance indicators to evaluate and rank WWTPs 2. Pollutant Load Reduction/Increase from each Investment Option (LR/I, d, p): calculates the load reduction/increase by subtracting the total effluent load of each WWTP from the load that the communities discharge into the river if the WWTP was not built. 3. Population Downstream (PD d ): is used to give more weight to investment options benefiting larger upstream populations in the basin. Fig. 3. Flow of Pollutant Load Reduction indicator 4. Population Served at Year of Analysis (Y) (PS d ): gives more weight to the investment option with the greater population served. 5. Population Discharging into the Groundwater (GW d ): gives more weight to the WWTP that decreases the population discharging its wastewater into the ground (i.e. septic systems). 6. Construction and O&M Costs (CC d, O&MC d ): give more weight to the WWTP that has lower construction and O&M costs per m 3 of wastewater treated. 7. Direct discharge into the lake: gives full score for all investment options except that of the Qaraoun WWTP because of its adverse direct effect on the lake. The construction of the Qaraoun WWTP has its treated effluent discharged directly into the lake instead of the current practice of discharging into the ground. The assigned weights for each indicator, along with the equations adopted for score calculation are summarized in Table 2. Since the selection of weighing factors for the performance indices can be subjective, the DSS was structured to allow the user to enter the value of these weighing factors. In order to provide the user with some guidance as to what value could be assigned for a particular factor, a sensitivity analysis was conducted to assess the impact of the weights assigned to each indicator described above to score the WWTPs and rank them. Each weight was varied from the base value both positively and negatively by 5, 10, 25, 50, and 75 percent, while the rest of the weights were kept constant. The selected base case uses the weights assigned initially based on expert opinion and can be equally modified or ascertained based on field surveys. Water Quality The second DSS component focused on the estimation of water quality improvement associated with the implementation of the investment options, through the simulation of pollutant concentrations in the River and Lake. Based on the availability of hydrologic data, time series on monthly surface stream flow at three locations (Joub Jannine-6 year record, Berdawni-13 year record, and Page 3 of 8

4 Mansoura-21 year record) as well as the monthly lake inflow and outflow were coupled with the effluent discharge from proposed WWTPs to simulate corresponding impacts on several water quality indicators including biological oxygen demand (BOD), chemical oxygen demand (COD), suspended solids (SS), total Nitrogen (N), and total Phosphorus (P) for the river and lake water. For this purpose, a mass balance approach with decay factors where pertinent, was used to determine the levels of these indicators in surface water. The effluent discharge locations considered in the simulation correspond to the proposed sites of the WWTPs. The river is simulated with its main tributaries and Lake Qaraoun using a matrix formulation of the governing equations. The general configuration of the system is recognized by the DSS through a System Matrix that takes into consideration the upstream-downstream routing of all potential sites for proposed investment options in the form of WWTPs along the river. Indicator levels are determined and routed within the whole system by multiplying this matrix by the mass balance governing equations detailed below. River simulation Estimates of monthly wastewater discharge rates (m 3 /month), from proposed WWTPs, were determined by multiplying the total number of population connected to a particular plant by the average monthly discharge per capita in the Basin. Field measurements were used for calculating the pollutants' loadings as such: L p = V s x C Sp L p Loading of pollutant P (kg/month) V s Monthly wastewater discharge rate (m 3 /month) C Sp Pollutant level (kg/m 3 ) Pollutant concentrations in the river are dependent on the simulated investment option as expressed below. C P = L P / F if WWTP is not within the simulated investment option C P = (L P R P ) / F if WWTP is within the simulated investment option R P = L p x RF P C P Concentration of pollutant P (kg/m 3 ); L P Loading of pollutant P (kg/month); F Monthly stream flow (m 3 /month); R P Total pollutant reduction through treatment (kg/month), and RF P Pollutant reduction factor Various investment options ranging from the do nothing scenario (current situation) to do everything scenario (building all proposed WWTPs can be simulated. The impact of scenarios on surface water quality is evaluated by routing the loadings of pollutants indicators from discharge points through the system. The monthly averages of pollutants indicators are tabulated and presented graphically to assist in visualizing the corresponding temporal pollutant level fluctuations, as well as seasonal and annual trends and their relationship to hydrologic trends (seasonal/annual dry and wet hydrologic cycles). Water quality standards are included for comparison and determination of frequency of noncompliance and magnitude of exceedance of such standards. An option s performance or effect is determined through an index analysis (i.e. percent and magnitude of exceeding water quality standards). Therefore, simulated investment options can be compared to each other based on the frequency and magnitude of this percentage associated with adopting these options. Page 4 of 8

5 Indicator Weight Score Load Reduction Efficiency Score (LRE, d, p) = (WWTP LRE,d,p / Highest WWTP LRE,d,p ) x 100 (LRE, d, p) Pollutant Load Reduction /Increase from each Investment Option (LR/I, d, p) 25 points (5 points for each of the five pollutants) 25 points (5 points for each of the five pollutants) Where: Score (LRE, d, p): Score for the load reduction efficiency indicator of pollutant p and investment option d. WWTP LRE,d,p: Percent load reduction efficiency of pollutant p by investment option d (from the total pollution p generated and discharged into the river). Highest WWTP LRE,d,p: Highest percent load efficiency reduction of pollutant p by all the investment options (from the total pollution p generated and discharged into the river Score (LR/I, d, p) = (L R/I,d,p / Highest L R/I,d,p ) x 100 L R/I,d,p = Pollutant Load p/capita.day x PGF Y,2005 x [(P (before, 2005) ) ((P (after, 2005) x LRF d,p )] Where: L R/I,d,p: Load Reduction/Increase for pollutant p in investment option d. Pollutant Load p/capita.day: Load (kg) of pollutant p per capita per day. PGF Y,2005: Population Growth Factor for design year Y, given population at base year With n being the number of years (Y-2005) and PGR being the population growth rate with PGF Y,2005 = (1+PGR) n P (before, 2005): Population connecting to the sewer line (discharging into river) in 2005 assuming WWTP is not installed (current situation). P (after, 2005): Population connecting to the sewer line (discharging into river) in 2005 assuming WWTP is installed. Population Downstream 15 points Score (PD, d) = (PD d / Highest PD d ) x 100 (PD d) Population Served at Year of 5 points Score (PS, d) = (PS d / Highest PS d ) x 100 Analysis (Y) (PS d ) Population Discharging into the 7.5 points Score (GW, d) = (GW d / Highest GW d ) x 100 Groundwater (GW d ) Construction Costs 12.5 points Score (CC, d) = [(Highest CC CC d ) / Highest CC) x 100 (CC d, ) O&M Costs 12.5 points Score (O&MC, d) = [(Highest O&MC O&MC d ) / Highest O&MC) x 100 (O&MC d ) Direct discharge into the lake 2.5 Table 2. Assigned weights and scores of the performance indicators Lake simulation Average monthly BOD levels in Lake Qaraoun are also estimated using a mass balance approach that is based on a volumebalance state equation as expressed below: S t+1 = S t + I t O t L t Sp t S min S t S max Sp t = 0 if S t + L t O t L t Smax Sp t = (S t + L t O t L t ) - S max if S t + L t O t L t Smax S t lake storage volume in month t; I t Inflow to the lake during month t; O t Outflow (controlled output) from the lake during month t; L t Losses in month t (mainly evaporation and seepage); S min Minimum storage level in the lake; S max Maximum storage level in the lake; Sp t Spill (uncontrolled output) during time stage t. Clearly, the state equation is recursive whereby the value of S t+1 represents the start-of-month storage in the next time step. The minimum and maximum storage constraints are recognized by testing for S t+1 against their values for each time step. If the calculated value of S t+1 is larger than S max, the difference is considered as a spill and the Page 5 of 8

6 next month storage is set at S max. In cases where S t+1 is less than the minimum allowable storage S min, the next month storage (S t+1 ) is set at the minimum storage level and the outflow O t released during month t is adjusted accordingly. Similarly, the monthly mass balance of BOD load in the lake is defined by: BOD t+1 = BOD t + BOD I t BOD O t BOD I t = C BOD-t x I t BOD O t = BOD C t x O t BOD C t = [(BOD t + BOD t+1 ) /2] /[(S t +S t+1 )/2] BOD t Load within the lake storage during month t; BOD I t Load received by the lake with the inflow I t during month t; BOD o t BOD load leaving the lake with the outflow O t during month t; and C BOD-t Average BOD concentration in the river during month t. A decay factor is used in the lake simulation due to the relatively long residence time of water. The final value for the BOD level in the lake during month t is determined as: BOD t-final = BOD t x (1 D t ) D t Monthly BOD decay factor in the lake. Similar to river simulations, the monthly averages of the BOD indicator are tabulated and presented graphically to assist in visualizing the corresponding temporal pollutant level fluctuations, as well as seasonal and annual trends and their relationship to hydrologic trends (seasonal/annual dry and wet hydrologic cycles). The BOD standard is included for comparison and determination of frequency and magnitude of non-compliance or exceedance of accepted standards. Cost Benefit Analysis (CBA) The purpose of the third component of the DSS is to provide decision makers with an estimated range of capital and operational expenditures (CAPEX), as well as expected conservative benefits resulting from potential water quality improvement. The cost-benefit analysis of various options defines the level of investment in terms of CAPEX (capital and operational expenditures) which will be a function of the technology adopted and the capacity in terms of population served. It also estimates the benefits corresponding to health and irrigation practices associated with the improvement in water quality using the averted cost approach due to the implementation of any particular investment options. Averted health-related costs focused on premature mortality and morbidity as outcomes associated with water-borne illnesses, namely diarrhea and typhoid (Refer to El Fadel et al for details). Premature mortality was expressed in terms of DALYs (Disability Adjusted Life Years). Morbidity was expressed both in terms of 1) DALYs to reflect the value of pain and suffering; 2) medical expenditures; and 3) opportunity cost of the time spent being sick or lost work days. The methodology adopted for averted irrigation costs, correspond to the installation of water filters to avoid the clogging of drip irrigation infrastructure due to eutrophied water in irrigation canals. Thus, the CBA consisted of comparing the CAPEX of investment with ranges of benefits accrued from corresponding environmental improvements over a userdefined design period, normally 20 to 25 years for domestic WWTPs. It is assumed that the benefits accrued are the result of the averted damage of pollution. If no investment option is implemented, the damage is assumed to grow proportionately to the population growth rate. This is normally a conservative approach particularly when the ecosystem assimilation capacity is exceeded. It is also assumed that Page 6 of 8

7 the damage converges to zero when effluent discharge standards are met. Finally, it was assumed that the damage cost of pollution is following a linear function between a maximum for the Do Nothing Scenario and zero for a pollutant level below recommended standards. Results The DSS allowed for testing various WWTP investment scenarios in terms of impact on improving water quality in the basin with associated financial benefits in the form of cost-benefit analysis and water quality in chart and GIS format (Fig. 4). For demonstration purposes, a scenario for investment in all WWTPs was simulated. The results indicated that this scenario is most desirable in terms of return on investment, which can be attained in 17 years after investment. Similarly, this scenarios yielded a maximum frequency of exceedance of the BOD and total Nitrogen standards of 13 and 8 percent respectively. The GIS interface showed that the maximum BOD concentration in the river and the lake is 33 mg/l, which is significantly lower than the simulated levels under the do nothing scenario (548 mg/l). As for the sensitivity analysis of the weighing factors, the simulated results showed that the outcome of the Matrix Analysis has limited sensitivity to the selected weights, whereby variations of less than 10 percent yielded no impact on the overall ranking. While the percent pollutant removal and the relative pollutant removal indicators exhibited the most impact on the overall ranking, the unit construction cost and the operation and maintenance cost indicators had minimal impact on the ranking of the WWTPs. a) Cost benefit analysis b) Water quality level in chart format c) Water quality levels in GIS interface Fig. 4. Typical outputs of DSS Scenario analysis Discussion and Conclusions Investing in upstream WWTPs will improve the quality of the water throughout the upper basin and will avert more socio-economic costs than downstream, investment. Total nitrogen levels did not drop significantly upon Page 7 of 8

8 investing in WWTPs since secondary treatment is adopted which have a low removal efficiency for nitrogen and phosphorous. While at the level recorded the latter are beneficial to soil and plants, they have been associated with algae growth causing damage to irrigation equipment and creating a perception of pollution in the area. Worldwide, the intricacy of river basin management is growing as the uses of water and the objectives to be fulfilled continue to increase. To guarantee a successful planning and operational management of such systems, the application of expert system tools becomes of utmost importance. A DSS can help decision-makers and funding agencies to address specific questions and needs by facilitating the use of models and databases in an interactive way. It constitutes a step towards meeting user needs and a framework that allows the user to define developmental requirements for integrated basin management, particularly with regards to long term monitoring, and wastewater and water quality management. USAID (United States Agency for International Development) Final and Survey Reports. BAMAS project (Litani Basin Management Advisory Services). Development Alternatives Inc. (DAI), Water & Environmental Sustainable Solutions (WESS), Khatib & Alami, Beirut, Lebanon. Acknowledgments The development of the DSS was funded through the US Agency for International Development (USAID) under the BAMAS project (Litani Basin Management Advisory Services). Disclosures The authors have nothing to disclose. References El-Fadel M, Maroun R, Aldeen RBF, El- Khoury DL and Najm MA. Lebanon: Health Valuation of Water Pollution at the Upper Litani River Basin. (2011). In: Nriagu JO (ed.) Encyclopedia of Environmental Health, 3, Burlington: Elsevier. UNDP, United Nations Development Program. (1970) Liban etude des eaux souterraines. Government of Lebanon, Beirut, Lebanon. MoE, Ministry of Environment, (2011) UNDP, United Nations Development Program., Elard, Business Plan for the Protection of the Litani River Basin, Beirut, Lebanon. Page 8 of 8