Design and operation of water desalination supply chain using mathematical modelling approach

This study presents the retrofit of water supply chain problem through multi period mixed integer linear program (MILP) model. The major strategic decisions include the determination of optimal new facility locations and capacity expansions of water desalination supply chain infrastructure assets which consists of water desalination plants, pipelines, and storage tanks, over a long time planning horizon. Other strategic decisions deal with the optimal selection of desalination technologies for existing and new desalination plants. In addition, the model provides decisions to define the pipeline network configuration for water transportation among several sites. The operation decisions are modelled to optimize the water production from desalination plants, energy consumption, brine disposal, as well as CO2 emissions. In addition, water transportation through the pipeline network and water storage at every site are optimized to satisfy water demand at every time period, and to minimize the net present value of the supply chain network. Finally, the proposed approach is solved for a case study to illustrate the application of the proposed mathematical programming model. The results show economic and environmental benefits if one considers full site integration and water production coordination in the supply chain network through the planning time horizon. © 2014 Elsevier B.V. All rights reserved.

This paper addresses the problem of retrofitting water desalination supply chain through multiperiod mixed integer linear programing model. The optimization model takes into account strategic and operational decisions through 15 years planning time horizon. Major strategic decisions looks for optimal desalination plant locations, desalination technology selection, desalination technology capacity expansion as well as pipeline and water storage capacity expansion. The operational decisions examine optimal operation of water desalination plants and water transportation through the pipeline network. In addition, the model gives an estimate of environmental impact from the desalination supply chain network over the planning time horizon. The proposed mathematical programming model is demonstrated on a case study which covers large geographical area. In general, the results show several advantages of economic and environmental impact when analyzing the supply chain problem over long-time horizon. The study also proposes several research extensions from the given study. These extensions include the design of water desalination supply chain under water demand uncertainty and the supply chain problem of water and power by cogeneration plants. Abbreviations GAMS MENA MINLP MSF NLP NPV RO Sets cs dt epis eplcs ncs npis nplcs pi PonitD s, s′ t General algebraic modelling system Middle East and North Africa Mixed integer nonlinear program Multi-stage flash distillation Nonlinear program Net present value Reverse osmosis Coastal sites Desalination technologies Existing pipeline set which link adjacent sites Existing desalination plant set Noncoastal sites New pipeline set which link adjacent sites New desalination plant set Set of all pipeline set Set of discrete points of the capacity for the desalination technologies Set of sites in the supply chain Time periods Parameters NEXP Number of technology expansion over the planning horizon CLT Construction lead time (years) WATERD Water demand at every time period t and every site s (m3) τ Storage time (days) α Water yield coefficient for a desalination technology β Energy yield coefficient for a desalination technology (kWh/m3) ε CO2 emission factor for a desalination technology (kg CO2−e/ kWh) χ Efficiency of water pump ρ Water density (kg/m3) g Standard gravity acceleration (m/s2) Hloss Head loss of water (m) δ Distance between adjacent sites (m) e Q Maximum water flowrate through a pipeline (m3/day) ξ Pipeline roughness constant d Pipeline diameter (m) COC Capital cost coefficient ($.day/m3) for desalination technology or a pipeline and ($/m3) for a storage tank OCCEn Operational cost for energy consumption ($/kWh) OCCCO2 Operational cost for emission ($/kg CO2−e) OCCBrine Operational cost for brine disposal ($.day/m3) OCCMain Operational cost for maintenance ($.day/m3) OCCLab Operational cost for labor ($.day/m3) OCCTrea Operational cost for pretreatment ($.day/m3) OCCMem Operational cost for membrane replacement ($.day/m3) Continuous variables CC Capital cost ($) CuCap Current capacity for technology or pipeline (m3/day), and for water tank (m3) EnC Energy consumption (kWh) ExpCap Capacity expansion (m 3 /day) for technology or pipeline (m3/day), and for water tank (m3) FCO2 Emission flowrate (kg CO2−e) NPV Net present value ($) OCBrine Operation cost for brine disposal ($) OCCO2 Operation cost for emission ($) OCEn Operation cost for energy consumption ($) OCLab Operation cost for labor ($) OCMain Operation cost for maintenance ($) OCMem Operation cost for membrane replacement ($) OCTrea Operation cost for intake pretreatment ($) Q Water flowrate (m3/day) SIndex Ranking number for the sites Binary variables yac Decision for capacity expansion availability at time period t. ynIns Decision for installing a desalination plant or a pipeline over the planning time T. ync Capacity expansion decision of a desalination technology in every time period t ypiop Decision for water product transportation direction through a new pipeline over the planning time T Superscript Brine Water brine from desalination technology Represent an emission property CO2 EN Represents energy consumption for a process In Represent intake flow of seawater Pipeline Designate a property for a pipeline Plant Designate a property for desalination plant Total Represent sum property for a flow UP Upper bound value Water Represent a property for water WaterTank Designate a property for water tank YD Yield of water from a desalination technology