Multi objective optimisation of water resources systems

Multi-Objective Optimisation of Water Resources Systems: A Shared Vision by Walter Godoy Thesis submitted in fulfilment of the requirement for the degree of Doctor of Philosophy College of Engineering and Science, Victoria University, Australia August 2015 Abstract Water resources systems are operated for many uses such as for municipal water supply, irrigation, hydro-electric power generation, flood mitigation, storm drainage, and for recreation Water resources systems may also serve as places of cultural and spiritual significance Decision-making in this context is inherently multicriterial, often requiring multi-disciplinary participation with a view to seeking an optimal solution or, at best, a compromise between conflicting interests for water Water resources planning involves a thorough understanding of not only the quantitative aspects such as the volumes of water harvested and released from reservoirs but also of the qualitative factors that underpin the shared vision for the operation of water resources systems for the benefit of all stakeholders The aim of this study was to develop a structured multi-objective optimisation procedure for the optimisation of operation of water resources systems considering climate change For this purpose, the integration of quantitative and qualitative information of water resources systems was achieved using a combined multi-objective optimisation and sustainability assessment approach as part of a three-phase procedure This procedure was tested through the preparation of optimal operating plans for a case study of the Wimmera-Glenelg Water Supply System (WGWSS), assuming a range of hydro-climatic conditions The WGWSS is located in north- western Victoria in Australia and is a multi-purpose, multi-reservoir system which is operated as a single water resources system; with many possible combinations of operating rules Phase (1) of the procedure involved the formulation of a higher order multi-objective optimisation problem (MOOP) for the WGWSS A higher order MOOP is defined in this study as a problem that is formulated with more than three objective functions The 18 objective functions of the MOOP were developed from four major interests for water identified in the WGWSS viz environmental, social, consumptive, and system-wide interests The 24 decision variables of the MOOP represented the complex operating rules which control the movement of water within the headworks The constraints of the MOOP, in terms of the physical characteristics of the WGWSS, were configured in a simulation model The formulation of the higher order MOOP demonstrated that the i procedure provided a means to explicitly account for all the major interests for water and to incorporate complex operating rules Phase (2) of the procedure involved the development of an optimisation-simulation (OS) model for the purposes of solving the higher order MOOP formulated in Phase (1) The optimisation engine was used to perform the search for candidate optimal operating plans and the simulation engine was used to emulate the behaviour of the system under the influence of these candidate optimal operating plans The setup of the optimisation engine was based on a widely used evolutionary algorithm and the setup of the simulation engine involved the replacement of an available simulation model with a surrogate model that had greater flexibility and stability in terms of changing from one operating plan to another Three hydro-climatic data sets were used to represent historic conditions and future climate conditions assuming a range of greenhouse gas emissions The setup of the optimisation engine was described in terms of the genetic operators (i.e selection, crossover, and mutation) and the optimisation parameters (i.e genetic operator settings, population size etc) Phase (3) of the procedure involved the development of an analytical approach which used the Sustainability Index ( ) to evaluate optimal operating plans The was used to aggregate the 18 objectives of the higher order MOOP, either separately in terms of the major interests for water, or collectively in terms of the sustainability of the WGWSS The was shown to have the flexibility to include a range of interests for water together with scaling characteristics that did not obscure poor performance The provided a simple means to rank optimal operating plans along the Pareto front with respect to all 18 objectives The Pareto front is the set of optimal trade-offs between the conflicting objectives Moreover, the was extended to incorporate stakeholders’ preferences for the purposes of selecting preferred Pareto-optimal operating plan(s) under the three hydro-climatic conditions mentioned earlier in Phase (2) The resulting Weighted Sustainability Index ( ) for the th stakeholder had all the benefits of the in terms of flexibility and scalability as described earlier Importantly, the key innovation of this procedure is that it combines the formation of Pareto fronts for a range of hydro-climatic conditions with sustainability principles to deliver a practical tool that can be used to evaluate and select preferred Pareto-optimal solutions of higher order MOOPs for any water resources system ii Declaration “I, Walter Rafael Godoy, declare that the PhD thesis entitled ‘Multi-Objective Optimisation of Water Resources Systems: A Shared Vision’ is no more than 100,000 words in length including quotes and exclusive of tables, figures, appendices, bibliography, references and footnotes This thesis contains no material that has been submitted previously, in whole or in part, for the award of any other academic degree or diploma Except where otherwise indicated, this thesis is my own work.” Signature: Date: 20 September 2015 iii Acknowledgements I would like to express my gratitude to my family and friends who gave me the possibility to complete this thesis In particular:  My wife, Doris, and my children, Evelyn and Thomas, for their patience and support in times of much hardship during this research and preparation of this thesis This work in part is dedicated to them for my absence as a loving husband and father;  My supervisor, Prof Chris Perera, for his guidance in my research and tireless efforts in reviewing each chapter of this thesis Much appreciation is extended to Chris for his understanding of my personal struggles and in his belief that I was a worthy candidate;  My supervisor, Dr Andrew Barton, for the opportunity to apply for candidature and his belief in that my practical knowledge of water resources engineering was of valuable contribution to science Much appreciation is extended to Andrew for his strategic thinking in the application of this study to real-world water resources problems;  I would also like to thank the three examiners of this thesis (Prof D Nagesh Kumar, Prof George Kuczera, and an anonymous examiner) for their well considered comments which have greatly improved the quality of this thesis;  My mum and dad, Aida and Rodolfo, whom I know would be proud of the effort that has gone into this piece of work Much appreciation goes to my mum for her assistance with my family at times when I was absent This work, in part, is dedicated to them for instilling in me the belief that I can always better;  I thank the Australian Research Council, GWMWater, and Victoria University for the financial assistance provided to this research project I could not have pursued my PhD research if not for the scholarship funded by these organisations; and  My wife and sister-in-law, Claudia, for their assistance in the review and collation of the draft thesis for submission iv Table of Contents Abstract i Declaration iii Acknowledgements .iv CHAPTER INTRODUCTION 1-1 1.1 Background 1-1 1.2 Aims of the study 1-4 1.3 Research methodology 1-5 1.3.1 Phase (1) - Formulation of MOOP 1-6 1.3.1.1 Identification of major interests for water 1-6 1.3.1.2 Specification of objective functions, decision variables, and constraints 1-6 1.3.2 Phase (2) - Development of O-S model 1-7 1.3.2.1 Setup of optimisation engine 1-7 1.3.2.2 Setup of simulation engine 1-8 1.3.3 Phase (3) - Selection of preferred Pareto-optimal solution(s) 1-8 1.3.3.1 Design of an analytical approach to evaluate candidate optimal operating plans 1-8 1.3.3.2 Evaluation of optimal operating plans under a range of hydroclimatic conditions 1-9 1.3.4 Concluding remarks on methodology 1-9 1.4 Significance of the research 1-10 1.5 Innovations of the research 1-12 1.6 Layout of this thesis 1-13 CHAPTER MULTI-OBJECTIVE OPTIMISATION MODELLING IN WATER RESOURCES PLANNING - A REVIEW 2-1 2.1 Introduction 2-1 v 2.2 Water resources planning 2-3 2.2.1 Water resources systems 2-3 2.2.2 Moving towards sustainability 2-6 2.2.3 Future climate considerations 2-8 2.2.4 Systems analysis techniques 2-12 2.3 Multi-objective optimisation 2-14 2.3.1 Classical and non-classical methods 2-17 2.3.2 Optimisation-simulation modelling 2-18 2.3.2.1 Optimisation engine 2-22 2.3.2.2 Simulation engine 2-25 2.3.3 Higher order multi-objective optimisation problems 2-26 2.3.4 Selection of most preferred optimal solution 2-31 2.4 Summary 2-34 CHAPTER A SHARED VISION FOR THE WIMMERA-GLENELG WATER SUPPLY SYSTEM 3-1 3.1 Introduction 3-1 3.2 The Wimmera-Glenelg Water Supply System 3-6 3.2.1 The study area 3-6 3.2.2 The Wimmera-Glenelg REALM model 3-10 3.2.3 Stakeholders’ interests for water 3-12 3.2.3.1 Environmental 3-14 3.2.3.2 Social 3-16 3.2.3.2.1 Recreation 3-16 3.2.3.2.2 Cultural 3-18 3.2.3.2.3 Water quality 3-19 3.2.3.3 Consumptive 3-20 3.2.3.4 System-wide 3-22 3.2.4 Performance metrics 3-24 3.2.4.1 Reliability 3-25 3.2.4.2 Resiliency 3-27 3.2.4.3 Vulnerability 3-28 vi 3.3 A higher order MOOP for the Wimmera-Glenelg Water Supply System 3-29 3.3.1 3.3.1.1 Environmental 3-32 3.3.1.2 Social 3-32 3.3.1.3 Consumptive 3-33 3.3.1.4 System-wide 3-33 3.3.2 Decision variables 3-33 3.3.2.1 Priority of supply 3-35 3.3.2.2 Flood reserve volume 3-39 3.3.2.3 Share of environmental allocation 3-40 3.3.2.4 Flow path 3-43 3.3.2.5 Storage maximum operating volume 3-48 3.3.2.6 Storage target and draw down priority 3-50 3.3.3 3.4 Objective functions 3-31 Constraints 3-54 3.3.3.1 Bounds on variables 3-55 3.3.3.2 Integer constraints 3-55 3.3.3.3 Statutory constraints 3-56 3.3.3.4 Physical constraints 3-56 Optimisation-simulation model setup 3-56 3.4.1 Simulation engine 3-58 3.4.1.1 System file 3-59 3.4.1.2 Input data 3-64 3.4.1.2.1 Hydro-climatic inputs 3-64 3.4.1.2.2 Water demands 3-66 3.4.2 Optimisation engine 3-66 3.4.2.1 Genetic operators 3-69 3.4.2.1.1 Selection 3-70 3.4.2.1.2 Crossover 3-71 3.4.2.1.3 Mutation 3-72 3.4.2.2 Optimisation parameters 3-73 3.4.2.2.1 Sensitivity analysis 3-75 3.5 Sustainability Indices for the Wimmera-Glenelg Water Supply System 3-77 3.5.1 The Sustainability Index 3-78 3.5.2 The Weighted Sustainability Index 3-83 vii 3.6 Summary 3-87 CHAPTER ANALYSIS OF OPTIMAL OPERATING PLANS USING THE SUSTAINABILITY INDEX ( ) 4-1 4.1 Introduction 4-1 4.2 A lower order MOOP - one user group 4-7 4.2.1 Problem formulation and model setup 4-7 4.2.2 Modelling results and discussion 4-8 4.2.2.1 Objective space 4-8 4.2.2.2 Decision space 4-13 4.2.2.3 Discussion 4-19 4.2.3 4.3 Conclusions 4-20 A series of higher order MOOPs – all user groups 4-21 4.3.1 Problem formulation and model setup 4-22 4.3.2 Modelling results and discussion 4-25 4.3.2.1 Objective space 4-25 4.3.2.2 Decision space 4-26 4.3.2.3 Discussion 4-36 4.3.3 4.4 Conclusions 4-39 A higher order MOOP for the Wimmera-Glenelg Water Supply System – all user groups 4-41 4.4.1 Problem formulation and model setup 4-41 4.4.2 Modelling results and discussion 4-42 4.4.2.1 Objective space 4-42 4.4.2.2 Decision space 4-46 4.4.2.3 Discussion 4-55 4.4.3 4.5 Conclusions 4-57 Summary 4-58 CHAPTER SELECTION OF PREFERRED OPTIMAL OPERATING PLANS UNDER VARIOUS FUTURE HYDRO-CLIMATIC SCENARIOS 5-1 5.1 Introduction 5-1 viii 5.2 A MOOP for the Wimmera-Glenelg Water Supply System under two plausible future GHG emissions scenarios 5-8 5.2.1 5.2.1.1 Run (A2) – The low to medium level GHG emission scenario 5-8 5.2.1.2 Run (A3) – The medium to high level GHG emission scenario 5-10 5.2.2 Modelling results and discussion 5-10 5.2.2.1 Objective space 5-10 5.2.2.2 Decision space 5-18 5.2.2.3 Discussion 5-27 5.2.3 5.3 Problem formulation and model setup 5-8 Conclusions 5-35 Selection of preferred optimal operating plan for the WGWSS 5-38 5.3.1 Stakeholder preferences 5-38 5.3.2 Post-processing results and discussion 5-42 5.3.2.1 Objective space 5-42 5.3.2.2 Decision space 5-45 5.3.2.3 Discussion 5-45 5.3.3 5.4 Conclusions 5-49 Summary 5-50 CHAPTER 6.1 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 6-1 Summary 6-1 6.1.1 Formulation of MOOP 6-3 6.1.2 Development of O-S model 6-6 6.1.3 Selection of preferred Pareto-optimal solution(s) 6-7 6.2 Conclusions 6-9 6.2.1 Additional benefits of using the Sustainability Index ( ) in higher order MOOPs 6-9 6.2.2 The results of the O-S modelling runs for the three hydro-climatic conditions (i.e the robust optimal operating plans) 6-10 6.2.3 The results of the selection process as applied to the robust optimal operating plans (i.e preferred optimal operating plans) 6-11 6.3 Recommendations 6-11 ix 6.3.1 Increasing the fidelity of the Wimmera-Glenelg REALM model Over the last decade there has been an increased level of fidelity in the configuration of environmental water demands (EWDs) in the Wimmera-Glenelg REALM model The term fidelity is used in this modelling context to refer to the degree of realism of a simulation model For instance, in a planning study undertaken in 2006 two EWDs were used to represent the environmental flow requirements for the entire system (i.e one for the Glenelg River and the other for the Wimmera River) These EWDs were configured as a seasonal pattern which were constant each year subject to the available water in the WGWSS In years of low water availability the seasonal pattern would be factored down and years of high water availability the seasonal pattern would be factored up (SKM, 2006) By comparison to the EWD setup in the (current) Wimmera-Glenelg REALM model, the level of complexity has increased in terms of the number of environmental flow sites (i.e four more sites); the variability in the seasonal pattern each year; and the management of the environmental water account in terms of the regulated and unregulated water that is used to supply these demands Moreover the basis of these EWDs has also increased in sophistication whereby environmental flow requirements place a greater focus on the frequency and duration of daily flow events (Godoy Consulting, 2014) Arguably, the next step in achieving higher levels of fidelity would be to convert the monthly operating rules within the Wimmera-Glenelg REALM model to a daily time-step with due consideration to the additional factors that arise in day-to-day operation Incorporating such higher fidelity attributes into the Wimmera-Glenelg REALM model would have represented a major development milestone and taken some time to complete For instance, it would have needed to be calibrated and validated over a range of climatic conditions in order to ensure that it was capable of replicating the behaviour of the system Kuczera et al (2009) highlight that one of the main modelling issues that arise when moving from monthly to daily time-steps is the need to more explicitly account for hydraulic constraints Moreover, Kuczera et al (2009) point out that the lack of travel time functionality is also evident in daily models and so this would also need to be addressed in order to avoid producing misleading impacts, particularly under climate change (Kuczera et al., 2009) Fortunately, one of the advantages of REALM is its ability to represent virtually any constraint imaginable using variable capacity carriers (Perera et al., 2005) These 6-12 types of carriers are essential for modelling complex storage operating rules and environmental flow rules as was further demonstrated by this study 6.3.2 Investigating potential developments to the optimisation process using the As explained earlier in Section 6.1, one of the main challenges in many higher order MOOPs is that the dominance test causes slow convergence to the Pareto front For instance, in the comparison of two very similar performing Pareto-optimal solutions, the solution that has at least one better performing objective would have the effect of dominating the other solution, assuming all other objectives of both solutions are equal This is the reason for the increase in the proportion of Pareto-optimal solutions to the population size in higher order MOOPs giving rise to the slow convergence to the Pareto front Given its ability to rank Pareto-optimal solutions, the could be used as part of the optimisation process in order to discard poorly ranked plans (e.g plans that have ) from the offspring population However consideration would need to be given to maintaining the population size constant at each iteration following the elimination of these poorly ranked solutions In addition to the ranking ability offered by the , it could be used to measure the diversity amongst Pareto-optimal solutions This ( ) attribute could be trialled in the NSGA-II as an extension to the niching strategy which solely works in terms of measuring solution diversity (i.e via the crowding distance metric) Such investigations into potential developments to the optimisation process would need to be undertaken with a clear rationale together with proven metrics to demonstrate the effectiveness of the 6.3.3 against other proven strategies Application to real-world planning study A true validation of the proposed multi-objective optimisation procedure would occur with its application to a real-world planning study This validation would encompass such areas as the elicitation of interests for water and stakeholders’ preferences; understanding the uncertainty associated with the inputs and parameters used to find 6-13 optimal operating plans; and proving that the optimal operating plans found are in close proximity to the Pareto front Whilst this study identified the four major interests for water through a desktop study of high quality information, this cannot replace the elicitation of actual interests for water that would be attained through a real-world planning study Similarly, the criteria by which to evaluate alternative operating plans would in all likelihood vary from agency to agency and from individual to individual The hierarchical approach to the structuring of the higher order MOOP would also be tested and opportunities for improving and streamlining such would be explored One such test could be to compare the optimal operating plans found by a MOOP which considered all objective functions versus another MOOP which had a collapsed or aggregated set of the same objective functions This would be analogous to comparing the 18-objective function MOOP presented in this thesis with a MOOP which considered the aggregation of these objective functions according to their respective interests for water (i.e environmental, social, consumptive, and system-wide interests) Moreover, the possibilities of increasing the fidelity of the simulation model and of using the as part of the optimisation process could be explored Similar to the elicitation of interests for water, preferences elicited from real-world stakeholders would also vary widely and efforts would need to be made to consolidate such preferences into workable information for input to the Again, this process of incorporating real-world attributes to the problem would have the potential to lead to improvements to the proposed Such elicitation of interests for water and stakeholders’ preferences could easily be under-estimated and under-valued by this study which used information from recently completed planning studies of the WGWSS The next major planning study in the WGWSS is scheduled to occur in 2019 This follows the recent completion of the review of water entitlement arrangements in 2014 (GWMWater, 2014) It is worth noting that this review process was supported by simulation modelling using REALM Moreover the recommendations of the study were largely concerned with improving system operation in terms of meeting the needs of social interests for water (i.e the preservation and restoration of recreation amenity); and environmental interests for water (i.e the development of collaborative 6-14 management plans for improving environmental watering arrangements between water agencies) In regards to the uncertainty associated with the inputs and parameters used to find optimal operating plans, it is recommended that an uncertainty analysis be included in the proposed multi-objective optimisation procedure in order to understand the implications of selecting one plan over another This could be undertaken in terms of quantifying the uncertainty of the data inputs and simulation and optimisation parameters used in the O-S model Having a better understanding of the uncertainty associated with the optimal operating plans found by the O-S model will provide the basis for more realistic trade-offs among Pareto-optimal plans Note that this uncertainty analysis would serve to compliment the use of future hydro-climatic projections in the proposed multi-objective optimisation procedure Whilst not a focus of this study, the search for optimal operating plans in a real-world study would need to extend beyond generations and demonstrate close proximity to the Pareto front, including a good level of diversity of plans along that front To that end, it would be recommended to exploit the distributed or shared memory parallel computing architectures that are available in order to provide the high computational effort required to evolve such optimal operating plans Additionally, this parallel processing approach would assist with addressing (in part) the issue of slow convergence that exists in many multi-objective evolutionary algorithms 6-15 7.0 References Agrell, P.J., Lence, B.J., and Stam, A (1998) “An Interactive Multicriteria Decision odel for ultipurpose Reservoir anagement: the Shellmouth Reservoir,” Journal of Multi-Criteria Decision Analysis, vol 7: 61 – 86 Alluvium (2013a) Glenelg River environmental flows study Report by Alluvium Consulting Australia for Glenelg-Hopkins Catchment Management Authority, Horsham, Victoria Alluvium (2013b) Wimmera River environmental flows study Report by Alluvium Consulting Australia for Wimmera Catchment Management Authority, Horsham, Victoria Anandhi, A., Srinivas, V.V., Nanjundiah, R.S., and Kumar, D.N (2008) Downscaling precipitation to river basin in India for IPCC SRES scenarios using support vector machine International Journal of Climatology 28: 401–420, DOI: 10.1002/joc.1529 Andreu, J., Ferrer-Polo, J., Pérez, M.A and Solera, A (2009) Decision Support System for Drought Planning and Management in the Jucar River Basin, Spain, 18th World IMACS / MODSIM Congress, Cairns, Australia 13-17 July 2009 Arnold, J.G., Williams, J.R., Srinivasan, R., and King, K.W (1999), SWAT: Soil and Water Assessment Tool, Blacklands Res Cent., Tex Agric Exp Stn., Temple, Tex Ashbolt, S., Maheepala, S., and Perera, B.J.C (2014) A Framework for Short-term Operational Planning for Water Grids, Water Resources Management, 28, 23672380 Australian Academy of Science, AAS (2010) “The Science of Climate Change: Questions and Answers.” Australian Academy of Science, Canberra, 24p Barlow, K.R (1987) Wimmera / Mallee Headworks System Reference Manual Rural Water Authority of Victoria, Wimmera Region, Horsham, Victoria 7-1 Bana e Costa, C.A., Antao da Silva, P., Nunes Correia, F (2004) Multicriteria evaluation of flood control measures: The case of Ribeira Livramento, Water Resources Management 18(3), 263-283 Bekele, E.G., and Nicklow, J.W (2005) Multiobjective management of ecosystem services by integrative watershed modelling and evolutionary algorithms, Water Resources Research, 41, W10406 Belton, V and Stewart, T.J (2002) “ ultiple Criteria Decision Analysis.” UK: Kluwer Academic Publishers, 372p., 2002 Bertsekas, D.P., (1991) Linear Network Optimisation: Algorithms and Codes MIT Press, Cambridge, MA Castelletti, A and Soncini-Sessa, R (2006) A procedural approach to strengthening integration and participation in water resource planning Environmental Modelling and Software, 21,(2006), 1455-1470 Coliban Region Water Corporation, Coliban Water (2012) “Wimmera system: Water Supply Demand Strategy 2011 to 2060 – Final Report,” Bendigo, Australia Dandy, G.C and Engelhardt, M.O (2006) Multi-objective Trade-Offs between Cost and Reliability in the Replacement of Water Mains Journal of Water Resources Planning and Management, March/April, 79-88 Deb, K (2001), Multi-Objective Optimization using Evolutionary Algorithms John Wiley & Sons, United Kingdom Deb, K and Agrawal, S (1999) “Understanding interactions among genetic algorithm parameters,” in Foundations of Genetic Algorithms 5, 265–286 Deb, K., Pratap, A., Agrawal, S., and Meyarivan (2002) A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II, The Institute of Electrical and Electronics Engineers Transactions on Evolutionary Computation, 6(2): 182-197 Department of Sustainability and Environment, DSE (2011) Western Region Sustainable Water Strategy Victorian Government, Melbourne, Australia 7-2 Department of Sustainability and Environment, DSE (2012) Wimmera-Glenelg Post Irrigation Cap Model, Accreditation for the Murray Darling Cap, 23 October 2012 Emsconsultants (2009) “Consultations with Indigenous Groups for the development of the draft Western Region Sustainable Water Strategy,” prepared for the Department of Sustainability and Environment, Melbourne, Australia Erlanger, P and eal, B (2005) “Framework for urban water resource planning,” prepared by Sinclair Knight Merz Pty Ltd for Water Services Association of Australia, Melbourne, Australia eWater (2012) Introduction to eWater Source, brochure for training workshop Fleming, P.J., Purshouse, R.C., and Lygoe, R.J (2005) Many-Objective Optimization: An Angineering Design Perspective, Springer-Verlag Berlin Heidelberg, 14-32 Gutteridge Haskins and Davey, GHD (2011) Report for Wimmera-Glenelg REALM Model Update: Volumes & - Model Setup Report, prepared for the Department of Sustainability and Environment, Melbourne, Australia Godoy Consulting, (2012) Wimmera-Glenelg REALM Model – 2011-12 Data Update Prepared for GWMWater, Melbourne, Australia Godoy Consulting, (2013) Wimmera-Glenelg REALM Model – 2012-13 Data Update Prepared for GWMWater, Melbourne, Australia Godoy Consulting, (2014) “Environmental Water Demand modelling for Glenelg and Wimmera catchments,” prepared for Department of Environment and Primary Industries, Melbourne, Australia Godoy, W., Barton, A., and artin, J (2009) “Modelling the Future Use of Reservoirs: A Case Study Exploring the Impact of Operating Rules and Climate Change,” Australian Journal of Water Resources, vol 13(2): 113-120 Godoy, W and Barton, A.F (2011) Modelling the environmental water reserve: A case study exploring the effects of the environment’s water entitlement in a complex water supply system Australian Journal of Water Resources, 14(2), 157-167 7-3 Godoy, W., Barton, A.F., and Perera B.J.C (2011) “A Procedure for Formulation of Multi-Objective Optimisation Problems in Complex Water Resources Systems,” MODSIM2011, 19th International Congress on Modelling and Simulation, Perth, Australia, CD Godoy, W., Barton, A.F and Perera B.J.C (2012) Multi-Objective Optimisation Method for Water Resources Management: A Multi-Reservoir System Case Study, Hydrology and Water Resources Symposium, Sydney, Australia Goldberg, D.E (1989) Genetic algorithms in search, optimization, and machine learning, Addison-Wesley, Reading, Mass Grampians Wimmera Mallee Water, GWMWater (2011) Storage Management Rules for the Wimmera-Glenelg System Headworks, prepared for Department of Sustainability and Environment, Melbourne, Australia Grampians Wimmera allee Water, GW Water (2012a) “Lake Lonsdale anagement Plan,” Horsham, Australia Grampians Wimmera Mallee Water, GWMWater (2012b) “Lake Fyans anagement Plan,” Horsham, Australia Grampians Wimmera Mallee Water, GW Water (2012c) “Water Supply Demand Strategy,” Horsham, Australia Grampians Wimmera Mallee Water, GWMWater (2013) Fact sheet: Bulk Entitlement Operations Review – Storage Management Objectives, Horsham, Australia Grampians Wimmera Mallee Water, GWMWater (2014) Bulk and Environmental Entitlements Operations Review (Wimmera and Glenelg Rivers) - Summary Report, Horsham, Australia Graymore, L ., Wallis, A ., and Richards, A.J (2009) “An index of regional sustainability: a GIS-based multiple criteria analysis decision support system for progressing sustainability,” Ecological Complexity, vol 6, 453-462 7-4 Hajkowicz, S And Higgins, A (2008) A comparison of multiple criteria analysis techniques for water resource management, European Journal of Operational Research, 184 (2008), 255-265 Hamstead, , Baldwin, C., and O’Keefe, V (2008) Water allocation planning in Australia – Current practices and lessons learned, National Water Commission, Australian Government HydroTechnology (1995) Development of REALM model of Wimmera-Mallee Stock and Domestic Water Supply System Report prepared for the Rural Water Corporation and the Office of Water Reform, Melbourne, Australia Intergovernmental Panel on Climate Change, IPCC (2000) IPCC Special Report: Emissions Scenarios – Summary for Policymakers, A Special Report of IPCC Working Group III, 27p Intergovernmental Panel on Climate Change, IPCC (2001) Climate Change 2001: Synthesis Report A Contribution of Working Groups I, II, and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change [Watson, R.T and the Core Writing Team (eds.)] Cambridge University Press, Cambridge, United Kingdom, and New York, NY, USA, 398 pp Intergovernmental Panel on Climate Change, IPCC (2007) IPCC Fourth assessment: synthesis report – summary for policymakers, 6–9 http://www.ipcc.ch/pdf/assessment-report/ar4/syr/ar4_syr_spm.pdf (accessed on November 2012) Intergovernmental Panel on Climate Change, IPCC (2014) http://www.ipcc.ch/ Janssen, R (2001) On the use of multi-criteria analysis in environmental impact assessment in the Netherlands, Journal of Multi-Criteria Decision Analysis, 10, 101109 Jones, R.N and Durack, P.J (2005) Estimating the Impacts of Climate Change on Victoria’s Runoff using a Hydrological Sensitivity odel, CSIRO Atmospheric Research, Victoria, Australia 7-5 Juwana, I., uttil, , and Perera, B.J.C (2012) “Indicator-based water sustainability assessment – A review,” Science of the Total Environment 438 (2012), 357-371 Knowles, J and Corne, D (1999) The Pareto archived evolution strategy: A new baseline algorithm for multiobjective optimization, Proceedings of the 1999 Congress on Evolutionary Computation, Piscataway, NJ: The Institute of Electrical and Electronics Engineers Press, 98-105 Kollat, J.B., Reed, P.M., Maxwell, R.M (2011) Many-objective groundwater monitoring network design using bias-aware ensemble Kalman filtering, evolutionary optimization, and visual analytics, Water Resources Research, 47, W02529, doi:10.1029/2010WR009194, 2011 Kuczera, G., 1992 Water supply headworks simulation using network linear programming Advances in Engineering Software 14, 55–60 Kuczera, G., Cui, L., Gilmore, R., Graddon, A., Mortazavi, S.M., and Jefferson, C (2009) Addressing the shortcomings of water resource simulation models based on network linear programming, 32nd Hydrology and Water Resources Symposium, 30 November-3 December 2009, Newcastle, Australia Kuczera, G and Diment, G.A., (1988) General water supply system simulation model: WASP Journal of Water Resources Planning and Management, American Society of Civil Engineers, 114 (4), 365–382 Kularathna, M.D.U.P., Rowan, T.S.C., Schultz-Byard, H., Broad, D.R., McIver, D., Flower, D., Baker, B., Rhodes, B.G., and Smith, P.J (2011) Multi-Objective Optimisation using Optimizer WSS to support operation and planning decisions of Melbourne Water Supply System, 19th International Congress on Modelling and Simulation, 12-16 December 2011, Perth, Australia Kumar, D.N., Raju, K.S., and Ashok, B (2006) Optimal Reservoir Operation for Irrigation of Multiple Crops Using Genetic Algorithms, Journal of Irrigation and Drainage Engineering, American Society of Civil Engineers, March/April 2006, 123129 7-6 Kumar, D.N and Reddy, M.J (2007) Multipurpose Reservoir Operation Using Particle Swarm Optimization, Journal of Water Resources Planning and Management, American Society of Civil Engineers, May/June 2007, 192-201 Labadie, J.W., Bode, D.A., Pineda, A.M., (1986) Network model for decision support in municipal water raw water supply Water Resources Bulletin 22 (6), 927–940 Labadie, J.W (2004) Optimal Operation of Multireservoir Systems: State-of-the-Art Review Journal of Water Resources Planning and Management, 130(2), 93-111 Linsley, R.K., Franzini, J.B., Freyberg, D.L., and Tchobanoglous, G (1992) WaterResources Engineering, McGraw-Hill series in water resources and environmental engineering, Fourth Edition, Singapore Lotov, A.V., Bourmistrova, L.V., Efremov, R.V., Bushenkov, V.A., Buber, A.L., and Brainin, N.A (2005) Experience of model integration and Pareto frontier visualization in the search for preferable water quality strategies, Environmental Modelling & Software, 20, 243-260 Loucks, D.P (1997) “Quantifying trends in system sustainability,” Hydrological Sciences-Journal-des Sciences Hydrologiques, vol 42(4): 513 - 530 Loucks, D.P., and Gladwell, J.S (1999) “Sustainability Criteria for Water Resource Systems,” Cambridge University Press, ewYork, United States of America Maier, H.R., Kapelan, Z., Kasprzyk, J., Kollat, J., Matott, L.S., Cunha, M.C., Dandy, G.C., Gibbs, M.S., Keedwell, E., Marchi, A., Ostfeld, A., Savic, D., Solomatine, D.P., Vrugt, J.A., Zecchin, A.C., Minsker, B.S Barbour, E.J., Kuczera, G., Pasha, F., Castelletti, A., Giuliani, M., Reed, P.M (2014) Position Paper - Evolutionary algorithms and other metaheuristics in water resources: Current status, research challenges and future directions, Environmental Modelling & Software, 62 (2014), 271-299 McCuen, R.H., Knight, Z., Cutter, A.G., (2006) Evaluation of the Nash–Sutcliffe efficiency index, Journal of Hydrologic Engineering, 11 (6), 597–602 McMahon, T.A., Adeloye, A.J., and Zhou, S-L (2006) Understanding performance measures of reservoirs, Journal of Hydrology, 324 (2006), 359-382 7-7 Mooney, C and Tan, P.-L (2012) South Australia’s River values in water planning, Journal urray: Social and cultural of Hydrology (2012), http://dx.doi.org/10.1016/j.jhydrol.2012.04.010 Mortazavi, S., Cui, L., and Kuczera, G (2009) Application of Multiobjective Optimisation Methods for Urban Water Management: A Case Study for Canberra Water Supply System in 32nd Hydrology and Water Resources Symposium, Newcastle, Australia Murray-Darling Basin Authority, MDBA (2011) Note to file on exercise of delegation under the Water Act 2007 (Cth), Trim Number D11/32122 National Water Commission, NWC (2014) http://www.nwc.gov.au/nwi/objectives Neal, B., Sheedy, T., Hansen, B., and Godoy, W (2005) Modelling of Complex Daily Environmental Flow Recommendations with a Monthly Resource Allocation Model 29th Hydrology and Water Resources Symposium, 21 – 23 February 2005, Canberra, Australia Nicklow, J., Reed, P., Savic, D., Dessalegne, T., Harrell, L., Chan-Hilton, A., Karamouz, M., Minsker, B., Ostfeld, A., Singh, A., and Zechman, E (2010) State of the Art for Genetic Algorithms and Beyond in Water Resources Planning and Management Journal of Water Resources Planning and Management, July/August, 412-432 Oliveira, R., and Loucks, D.P., (1997) Operating rules for multireservoir systems, Water Resources Research, 33(4), 839-852 Palmer, R.N., Werick, W.J., MacEwan, A., Woods, A.W (1999) Modeling Water Resources Opportunities, Challenges, and Trade-offs: The Use of Shared Vision Modeling for Negotiation and Conflict Resolution, Journal of Water Resources Planning and Management, American Society of Civil Engineers, 1999, 1-13 Perera, B.J.C., James, B., and Kularathna M.D.U.P (2005) Computer Software Tool REALM for Sustainable Water Allocation and Management Journal of Environmental Management, 77, 291-300 7-8 Razavi, S., Tolson, B.A., and Burn, D.H (2012) Review of surrogate modeling in water resources, Water Resources Research, 48, W07401, DOI:10.1029/2011 Reed, P., Minsker, B.S., and Goldberg, D.E (2000) Designing a competent simple genetic algorithm for search and optimization, Water Resources Research, 36(12), 3757-376 Sachindra, D.A., Huang, F., Barton, A., and Perera, B.J.C (2012) Least square support vector and multi-linear regression for statistically downscaling general circulation model outputs to catchment streamflows, International Journal of Climatology DOI: 10.1002/joc.3493 Sachindra, D.A., Huang, F., Barton, A., and Perera, B.J.C (2014a) Statistical downscaling of general circulation model outputs to precipitation – part 1: calibration and validation International Journal of Climatology DOI: 10.1002/joc.3914 Sachindra, D.A., Huang, F., Barton, A., and Perera, B.J.C (2014b) Statistical downscaling of general circulation model outputs to precipitation – part 2: biascorrection and future projections DOI: 10.1002/joc.3915 Sandoval-Solis S., cKinney D C and Loucks D P (2011) “Sustainability Index for Water Resources Planning and anagement,” Journal of Water Resources Planning and Management, 137(5), 381-390 Schaffer, J.D., Caruana, L.J., and Eshelamn, R (1989) A study of control parameter affecting online performance of genetic algorithms for function optimisation, in: Proceeding of Third International Conference on Genetic Algorithms, 4-7 October, George Mason University, California, 1989, 51-60 Schwefel, H.P., (1995) Evolution and optimum seeking, Wiley, New York Sinclair Knight Merz, SKM (2004) Wimmera-Mallee simulation model – annual update methodology Prepared by SKM for Department of Sustainability and Environment Sinclair Knight Merz, SKM (2007) PRIDE User Manual Prepared by SKM for the Department of Sustainability and Environment, Melbourne, Australia 7-9 Sinclair Knight Merz, SKM (2010) REALM Routing - Progress to Date, in Project No VW03771.: Melbourne, Australia Sinclair Knight Merz, SKM (2011) Explicit Environmental Water Requirement Modelling in REALM - User manual and Goulburn River Modelling Report, Prepared by SKM for the Department of Sustainability and Environment, Melbourne, Australia Sinha, A., Saxena, D.K., Deb, K., Tiwari, A (2013) Using objective reduction and interactive procedure to handle many-objective optimization problems, Applied Soft Computing, 13, 415–427 Siriwardene, N.R and Perera, B.J.C (2006) Selection of genetic algorithm operators for urban drainage model parameter optimisation, Mathematical and Computing Modelling, 44 (2006), 415-429 Srinivas, N and Deb, K (1995) Multiobjective optimization using nondominated sorting in genetic algorithms Evolutionary Computation, 2(3), 221-248 Van Veldhuizen, D.A and Lamont, G.B (2000) Multiobjective Evolutionary Algorithms: Analyzing the State-of-the-Art Massachusetts Institute of Technology, Evolutionary Computation, 8(2), 125-147 Victorian Environmental Water Holder, VEWH (2013) “Seasonal Watering Plan 201213,” Department of Environment and Primary Industries, Melbourne, Australia Victorian Environmental Water Holder, VEWH (2014) “Seasonal Watering Plan 201314,” Department of Environment and Primary Industries, Melbourne, Australia Victorian Government Gazette, VGG (2010) “No S 446 Friday 29 October 2010 Special” Authority of Victorian Government Printer, Melbourne, Australia, 1-80 Wannon Region Water Corporation, Wannon Water (2012) “Water Supply Demand Strategy 2012-2060,” Hamilton, Australia Wu., W., Maier, H.R., and Simpson, A.R (2010) Single-Objective versus Multiobjective Optimization of Water Distribution Systems Accounting for Greenhouse Gas Emissions by Carbon Pricing, Journal of Water Resources Planning and Management, American Society of Civil Engineers, Sept/Oct 2010, 555-565 7-10 Zitzler, E & Deb, K (2000) “Comparison of ultiobjective Evolutionary Algorithms: Empirical Results,” Evolutionary Computation, vol 8(2): 173-195 Zitzler, E., Laumanns, M., and Thiele, L (2001) SPEA2: Improving the strength Patero evolutionary algorithm Technical Report 103 Compuer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Zurich, Switzerland Zitzler, E and Thiele, L (1999) Multiobjective Evolutionary Algorithms: A Comparative Case Study and the Strength Pareto Approach, The Institute of Electrical and Electronics Engineers Transactions on Evolutionary Computation, vol 3(4): 257271 7-11 [...]... Chapter 2 Multi- objective optimisation modelling in water resources planning - a review 2.1 Introduction This chapter presents a critical review of the literature on multi- objective optimisation modelling in water resources planning Specifically, it deals with (i) the various aspects of water resources planning and the multi- criterial nature of problems concerning the planning and operation of multi- purpose,... significance of the research in terms of the need for optimising the operation of water resources systems and proposes a structured procedure for the development of a shared vision for the operation of water resources systems It also presents the aims of the study and describes the tasks undertaken to achieve these aims The second chapter presents a critical review of the literature on multi- objective optimisation. .. regulate the water resources within the WGWSS The constraints of the problem were specified both in terms of the formulation of the MOOP and also in terms of the real-world limitations of the WGWSS 1.3.2 Phase (2) - Development of O-S model 1.3.2.1 Setup of optimisation engine The setup of the optimisation engine was aimed at demonstrating the novelty of the structured multi- objective optimisation. .. operation of complex water resources systems can be applied to any water resources system 1.3.1 Phase (1) - Formulation of MOOP 1.3.1.1 Identification of major interests for water Much of the information required to identify the major interests for water in the WGWSS had already been collected as part of various recently completed planning studies A desktop study of this information was undertaken as part of. .. improve the sustainability of the water resources system Loucks and Gladwell (1999) argued that sustainable development can only succeed with sustainable water resources systems supporting that development In their review of the many definitions of sustainable development, they propose the following definition for the management of water resources systems: “Sustainable water resource systems are those designed... movement of water, and the hydro-climatic processes that affect the availability of water resources The structured multi- objective optimisation procedure achieves this greater level of realism through, both, a holistic approach of formulating the problem and the use of O-S modelling The problem formulation approach sets out a flexible basis on which to establish an overall goal for the water resources. .. remainder of this section provides further details of the three areas of study highlighted above Developing a thorough understanding of the major interests for water in water resources systems provides valuable insights into the type and extent of conflict that may exist between the different uses for water In the WGWSS for example, many of the 12 headworks storages having conflicting interests in terms of. .. this thesis, is the use of combined optimisation simulation (O-S) models given that optimisation methods can be directly linked with trusted simulation models (Labadie, 2004) Many of the interests for water that exist in water resources systems are conflicting and non-commensurable which can be generally reduced to multi- objective optimisation problems (MOOPs) in which all objectives are considered... observations of increases in global average air and ocean temperatures, widespread melting of snow and ice and rising global average sea level.” (IPCC, 2007, p2) 1.2 Aims of the study The aim of this project is to develop a structured procedure for the optimisation of operation of water resources systems considering climate change This procedure will take explicit account of:  competing objectives concerning... to Section 2.2 for details of this part of the study Many of the interests for water in water resources systems are conflicting and noncommensurable which can be generally reduced to multi- objective optimisation problems (MOOPs) Characteristically, these problems give rise to a set of optimal solutions referred to as Pareto-optimal solutions or the Pareto front, instead of a single optimal solution ... operation of water resources systems for the benefit of all stakeholders The aim of this study was to develop a structured multi-objective optimisation procedure for the optimisation of operation of water. .. optimisation of operation of complex water resources systems can be applied to any water resources system 1.3.1 Phase (1) - Formulation of MOOP 1.3.1.1 Identification of major interests for water. .. details of this part of the study 2-2 2.2 Water resources planning 2.2.1 Water resources systems As part of the process of finding optimal or compromise solutions, there are a number of challenges