PRES22_0231.docx DOI: 10.3303/CET2294174 Paper Received: 15 April 2022; Revised: 14 May 2022; Accepted: 23 May 2022 Please cite this article as: Oqbi M., Linke P., Al-Mohannadi D.M., Bishnu S., 2022, Synthesis of Water-Energy Network Considering Seasonality, Chemical Engineering Transactions, 94, 1045-1050 DOI:10.3303/CET2294174 CHEMICAL ENGINEERING TRANSACTIONS VOL. 94, 2022 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Yee Van Fan, Jiří J. Klemeš, Sandro Nižetić Copyright © 2022, AIDIC Servizi S.r.l. ISBN 978-88-95608-93-8; ISSN 2283-9216 Synthesis of Water-Energy Network Considering Seasonality Manar Oqbi*, Patrick Linke, Dhabia Al-Mohannadi, Sumit Bishnu Chemical Engineering Program, Texas A&M University at Qatar, Education City, Doha, Qatar manar.oqbi@tamu.edu With the increase in water and energy demands to satisfy industrial processes requirements and convert raw materials into value-added products, natural resources are experiencing depletion stress. One of the effective solutions to decrease freshwater and energy consumption and production in industrial cities is to employ water- energy integration. Due to increasingly strict environmental regulations, integration networks became essential. Water and carbon footprints are reduced significantly via water and energy integration networks. The performance of the integration networks is affected by seasonal changes. Previous work ignored seasonal fluctuations in water/energy supply and demand or mainly utilized multi-period planning to consider seasonal variations while designing integrations networks. It is important to consider seasonal variations to reflect the real performance of the network and avoid operation disturbances. One of the drawbacks of multiperiod planning is the resulting complicated integration model. Multiperiod planning may result in implementation difficulties due to constraints on the piping layout that hinder its applicability. This paper investigates and assesses seasonal changes’ impacts on different segments of the water-energy network using several tools. Based on seasonality assessment, a novel approach is proposed to design optimal water-energy integration network. The approach depends on designing the network units and utility system based on the maximum required capacity (i.e., peak conditions) to ensure that water/energy demands will be satisfied over the year. The water-energy network connectivity is determined based on average demand/supply while any water source-to-sink pipeline is designed based on maximum potential flowrate. The water network is designed based on the worst-case scenario of removal ratios to ensure the required water quality for each sink is satisfied for all connections over different seasons. A MINLP mathematical model was expanded to include the proposed approach. The objective function is to minimize the total annual cost (TAC) of the design. Finally, the framework was demonstrated by applying it to a case study which was solved using a stochastic programming tool to illustrate the applicability of the developed model. The results indicate that the optimal design of the water-energy network that considers seasonal changes in water/energy demands, and supplies can be achieved with the proposed method with a TAC of 78 MUSD/y without the need for multiperiod planning. The optimal treatment units selected in this case were one-stage and two-stage nanofiltration. 1. Introduction The development of different sectors depends on utilizing water and energy. This includes using water and energy for residential, industrial, and auxiliary purposes. The extreme consumption of water and energy natural resources and the rapid increase in the global population lead to depletion danger. One of the main consumers is industrial activities. To mitigate this issue, various integration networks are designed and utilized. Seasonal variations need to be considered while designing water-energy integration networks to ensure continuous operation during different seasonal conditions over the year. Seasonal variability originating from different weather conditions has a direct impact on integration networks’ components such as cooling systems, treatment units, desalination units, and utility systems. It is required to cover this gap by analyzing and assessing the significance of the water-energy network seasonal variations and developing a systematic approach for finding the optimal design that can handle seasonal variations. Several studies focused on developing flexible integration networks that are feasible, and energy efficient over different time periods. Earlier work focused on water-energy networks without considering seasonality while other work addressed seasonality issues by considering multi-period planning. Multiperiod planning was used for synthesizing heat exchanger networks. An 1045 early work considered a MILP model that aims to find minimum utility requirements with the fewest number of units in each period (Floudas and Grossmann, 1986). Another study provided a MINLP model to design or retrofit heat exchanger network (HEN) that operates flexibly at different conditions (Papalexandri and Pistikopoulos, 1993). Kim and Han (2001) utilized heuristics and dynamic programming (DP) to develop a three-step approach for utility system short-term (days/weeks) multi- period planning. Another work focused on using total site analysis of industrial cluster after dividing the year into n-periods and identify the minimum and maximum energy supplies and demands (Bungener et al., 2015). Isafiade (2017) proposed a MINLP model to integrate renewables into the synthesis of HEN over different seasons of the operational year considering economic and environmental aspects. Hot utilities involved three levels of steam namely, HP, MP, and LP while cold utilities involved cold air and cooling water. A method for total site (TS) energy targeting considering daily (short-term), and seasonal (long-term) variations in energy supplies and demands was developed (Liew et al., 2018). The method utilized the total site energy targeting approach and the total site heat cascade to determine the short-term and long-term utility requirements. Multiperiod planning was utilized for designing water networks. Burgara-Montero et al. (2013) designed distributed treatment systems for industrial discharges into watersheds using a multi-objective MINLP model. The model considers minimizing treatment unit TAC and pollutants’ concentrations in the final destination. Bishnu et al. (2014) considered long-term planning for direct water reuse by two optimization models. The models provide the lowest cost design of the water network by minimizing the TAC and freshwater consumption. Arredondo-Ramírez et al. (2015) highlighted the optimal multi-period planning of agricultural water systems via multi-objective MINLP model that considers water collections, reuse, and distribution strategies. Another MINLP model considers the direct reuse of water and the regeneration of wastewater (Bishnu et al., 2017). The objective of the model is to minimize the TAC by long-term multi-period planning of water network considering the entire planning horizon. Another study explored water-energy nexus seasonality considering flowrates of water streams and electricity prices (Gaudard et al., 2018). Al-Mohannadi et al. (2020) proposed a MILP to evaluate CO2 reduction policies using a two-step multiperiod planning approach. Taheri (2021) developed a MINLP model. The model considers multiperiod planning of the water network in Eco-Industrial Park (EIP) targeting minimizing TAC. The model was applied for a three-period problem considering fluctuations in flowrate due to new plant construction within the EIP. The author stated that the obtained solution by GAMS might not be globally optimal due to model complexity and reformulation- linearization approach was used to alleviate the problem complexity. A recent study by Zhou et al. (2021) proposed MINLP model to design water network considering seasonal changes in water demands and supplies from multiple water resources. The model aims to minimize the TAC. The fluctuations in water flowrates over different periods were considered using multiperiod planning. Two technologies were considered for desalination which are ion exchange (IX) and reverse osmosis (RO). The results indicate that desalination technology selection depends on water price. So far, most studies either ignored seasonal variations of water/energy supply and demand within industrial park networks or tackled seasonality issues by multiperiod planning. This gap needs to be addressed. It is essential to consider seasonal fluctuations in water and energy supply and demand which will reflect the real water-energy system and help maintain continuous operation. Multi-period planning may lead to complicated models which might be difficult to be solved and obtain globally optimal solutions. Implementation difficulties may hinder the applicability of the proposed models via muti-period planning due to constraints on the piping system affected by the geographical region. Figure 1 demonstrates the design of this study which focuses on three main elements. First, the study explored seasonality impacts on different components of the water-energy integration networks considering the water-energy nexus. Second, the significance of the observed variations was assessed. Third, based on performed assessment, a novel approach capable of handling seasonal fluctuations in water and energy demands within water-energy integration networks was proposed. Figure 1: Illustration of study design 1046 2. Mathematical Model A mathematical optimization model is formulated based on minimizing the total annual cost (TAC) of the water- energy network considering seasonality. A mixed-integer nonlinear programming model (MINLP) was developed earlier by Alnouri et al. (2015) to design optimal water integration networks. The author illustrated a representation of an industrial city and used direct and indirect integration to integrate water sources and sinks while minimizing the total annual cost as the objective function. The proposed method indicates effective freshwater savings and wastewater minimization. The model was expanded by Fouladi et al. (2017) to consider the water-energy nexus. The model was extended and modified for the purpose of this study. The focus of this work is to find the optimal design of the water-energy network considering seasonal changes. The proposed approach depends on designing the water-energy connectivity based on average supplies and demands over the year, while considering minimum removal ratios of treatment units and maximum units’ capacities. The network involves treatment units, cooling systems, and the waste-heat to power unit. The objective function is to minimize capital and operating cost including the cost of treatment, desalination, cooling, piping, wastewater discharges, and waste heat to power cost as indicated by Eq(1). 𝑇𝑇𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝑜𝑜𝑎𝑎𝑎𝑎𝑎𝑎𝑜𝑜𝑜𝑜 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 = 𝑇𝑇𝑇𝑇𝑇𝑇𝑜𝑜𝑜𝑜𝑇𝑇𝑇𝑇𝑎𝑎𝑜𝑜 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 + 𝐷𝐷𝑇𝑇𝑐𝑐𝑜𝑜𝑜𝑜𝐷𝐷𝑎𝑎𝑜𝑜𝑜𝑜𝐷𝐷𝑜𝑜𝑎𝑎 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 + 𝐶𝐶𝑜𝑜𝑜𝑜𝑜𝑜𝐷𝐷𝑎𝑎𝐶𝐶 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 + 𝑃𝑃𝐷𝐷𝑃𝑃𝐷𝐷𝑎𝑎𝐶𝐶 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 + 𝑤𝑤𝑜𝑜𝑐𝑐𝑜𝑜𝑇𝑇𝑤𝑤𝑜𝑜𝑜𝑜𝑇𝑇𝑇𝑇 𝑑𝑑𝐷𝐷𝑐𝑐𝑐𝑐ℎ𝑜𝑜𝑇𝑇𝐶𝐶𝑇𝑇𝑐𝑐 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 + 𝑤𝑤𝑜𝑜𝑐𝑐𝑜𝑜𝑇𝑇 ℎ𝑇𝑇𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜 𝑃𝑃𝑜𝑜𝑤𝑤𝑇𝑇𝑇𝑇 𝑐𝑐𝑜𝑜𝑐𝑐𝑜𝑜 (1) The objective function subject to equality and inequality constraints including water balances of sources and sinks as indicated by Eq(2) and Eq(3) respectively. � � 𝑀𝑀𝑖𝑖𝑖𝑖,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 + � � 𝑇𝑇𝑖𝑖𝑖𝑖,𝑟𝑟𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 + � � 𝑇𝑇𝑖𝑖𝑖𝑖,𝑠𝑠𝑠𝑠 𝑎𝑎𝑎𝑎𝑎𝑎 𝑠𝑠∈𝑇𝑇𝑠𝑠∈𝑆𝑆 + 𝐷𝐷𝑖𝑖𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑊𝑊𝑖𝑖𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑗𝑗∈𝑆𝑆𝑆𝑆𝑝𝑝𝑖𝑖∈𝑃𝑃 (2) � � 𝑀𝑀𝑖𝑖𝑖𝑖,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 + � � 𝑇𝑇𝑟𝑟𝑖𝑖,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 + � � 𝑇𝑇𝑠𝑠𝑠𝑠.𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑠𝑠∈𝑇𝑇𝑠𝑠∈𝑆𝑆 + � � 𝑇𝑇𝑚𝑚𝑖𝑖,𝑗𝑗𝑖𝑖 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 𝑚𝑚∈𝑅𝑅𝑖𝑖∈𝑃𝑃 + � � 𝑇𝑇𝑛𝑛𝑛𝑛.𝑗𝑗𝑖𝑖 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 𝑛𝑛∈𝐾𝐾𝑛𝑛∈𝑆𝑆 + � 𝐹𝐹𝑙𝑙,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑙𝑙 ∈𝐿𝐿 𝑖𝑖∈𝑆𝑆𝑆𝑆𝑝𝑝𝑖𝑖∈𝑃𝑃 = 𝐺𝐺𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 (3) The minimum removal ratios of treatment units were considered as shown via Eq(4). 𝑥𝑥𝑐𝑐,𝑟𝑟𝑖𝑖 𝑇𝑇,𝑚𝑚𝑎𝑎𝑚𝑚 = 𝑥𝑥𝑐𝑐,𝑟𝑟𝑖𝑖 𝑖𝑖𝑛𝑛,𝑚𝑚𝑎𝑎𝑚𝑚 �1 − 𝑅𝑅𝑅𝑅𝑐𝑐,𝑟𝑟𝑖𝑖𝑚𝑚𝑖𝑖𝑛𝑛� (4) Energy sources such as the waste heat to power unit and energy sinks including cooling systems, treatment units, are subject to inequality constraints as indicated in Eq(5) and Eq(6) respectively. Eq(7) indicates the total cost of central and decentral treatment units. � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑟𝑟𝑖𝑖 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑠𝑠𝑠𝑠 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 𝑠𝑠∈𝑇𝑇𝑠𝑠∈𝑆𝑆 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑚𝑚𝑖𝑖 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 𝑚𝑚∈𝑀𝑀𝑖𝑖∈𝑃𝑃 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑛𝑛𝑛𝑛 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 𝑛𝑛∈𝐾𝐾𝑛𝑛∈𝑆𝑆 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝐶𝐶𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 𝑗𝑗′∈𝑆𝑆𝑆𝑆𝑝𝑝 ′𝑖𝑖∈𝑃𝑃 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝑂𝑂𝐶𝐶𝑆𝑆𝑂𝑂,𝑎𝑎𝑎𝑎𝑎𝑎 𝑗𝑗′∈𝑆𝑆𝑆𝑆𝑝𝑝 ′𝑖𝑖∈𝑃𝑃 + � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝐴𝐴𝐶𝐶,𝑎𝑎𝑎𝑎𝑎𝑎 𝑗𝑗′∈𝑆𝑆𝑆𝑆𝑝𝑝 ′𝑖𝑖∈𝑃𝑃 ≤ � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑖𝑖′∈𝑆𝑆𝑆𝑆𝑝𝑝 ′𝑖𝑖∈𝑃𝑃 (5) � � 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑟𝑟𝑖𝑖 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 ≤ � � 𝑃𝑃𝑊𝑊𝑟𝑟𝑖𝑖 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 (6) 𝐶𝐶𝑇𝑇 = 𝐾𝐾𝐹𝐹 (� ��𝑇𝑇𝑟𝑟𝑖𝑖𝑚𝑚𝑎𝑎𝑚𝑚� 𝛼𝛼 𝐶𝐶𝑟𝑟𝑖𝑖𝐶𝐶𝐶𝐶 𝑟𝑟∈𝑅𝑅 𝑖𝑖∈𝑃𝑃 + � �(𝑇𝑇𝑠𝑠𝑠𝑠𝑚𝑚𝑎𝑎𝑚𝑚)𝛼𝛼𝐶𝐶𝑠𝑠𝑠𝑠 𝐶𝐶𝐶𝐶 𝑠𝑠∈𝑇𝑇 𝑠𝑠∈𝑆𝑆 ) + 𝐻𝐻𝑦𝑦 ( � �𝑇𝑇𝑟𝑟𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝐶𝐶𝑟𝑟𝑖𝑖𝑂𝑂𝐶𝐶 𝑟𝑟∈𝑅𝑅𝑖𝑖∈𝑃𝑃 + ��𝑇𝑇𝑠𝑠𝑠𝑠 𝑎𝑎𝑎𝑎𝑎𝑎 𝐶𝐶𝑠𝑠𝑠𝑠 𝑂𝑂𝐶𝐶 ) 𝑠𝑠∈𝑇𝑇𝑠𝑠∈𝑆𝑆 (7) 3. Case Study Three processes are involved in the case study which are ammonia, methanol, and Gas-to-Liquid (GTL). Process-related data including basic power load and minimum cooling requirement are shown in Table 1. Process water demands and supplies are represented in Table 2. In total, the case study involves nine water sources including freshwater, and nine water sinks including discharge with different flowrates and qualities. Three cooling systems are considered which are air-coolers, once-through cooling seawater (OCSW), and cooling towers. It is assumed that desalination units use Reverse Osmosis (RO) technology to produce freshwater. This scenario demonstrates the case of designing the water-energy network connections based on annual average values, minimum removal ratios of the treatment units, and maximum units’ capacities to consider seasonal variations of water/energy elements. It is worth mentioning that the symbols P1, P2, and P3 stand for ammonia, methanol, and GTL plants respectively. The symbols “S1, S2, S3”, “D”, “I”, and FW represent 1047 numbered water sources, water sinks, irrigation demand and, freshwater respectively. The symbols 1S and 2S stand for one-stage and two-stage treatment unit respectively while TR refers to the treatment unit. Different water sources and sinks vary in both flowrates and quality and the network is designed accordingly as shown in Table 3 and Table 4. The minimum cooling requirement Qmin represents the heat that should be removed by any selected cooling system. Case study data were obtained from an earlier work by Fouladi et al. (2017). Table 1: Process data Table 2: Process water supply and demand 4. Results and Discussion The formulated case study was solved using a stochastic optimization tool. The obtained results for source-to- sink water allocation are presented in Table 3. The allocation is based on average supply and demand of different sources/sinks and sink permissible pollutant concentrations. Freshwater connections from desalination plant is used to satisfy the demands of some sinks in the three processes as these sinks require freshwater. Some water from GTL process is directly discharged. Wastewater from industrial park sources can be either directly allocated into sinks or discharged based on source and sink water quality and requirements or it can be allocated into treatment units and then utilized into sinks or discharged. In Table 3 and Table 4, 0 t/d indicates that a connection between the source on the same raw and the sink of the same column does not exist. Table 4 shows wastewater allocation from different sources into treatment units. Results indicate that 14,171 t/d of wastewater from source 1 in process 3 are treated in a two-stage nanofiltration unit. Table 5 represents treated water allocation from treatment units into different sinks or discharges. Decentral one-stage and two-stage nano- filtration membrane units are selected for wastewater treatment as required. Results show that treatment of wastewater from some sources is required in some cases prior to wastewater discharge to abide by environmental regulations. Table 3: Source-Sink water allocation Treatment units and cooling system selection were made based on minimizing the total annual cost (TAC) which is the objective function of the formulated problem. The obtained results show that the optimal water network design utilizes air coolers as the cheapest cooling option compared to cooling towers and once-through cooling seawater (OCSW). Waste heat is converted into power via the WHP unit while the remaining waste heat will be rejected through air coolers. Table 6 represents power allocation in (MW) into different considered processes. The TAC for the optimal design of the water-energy network considering seasonal fluctuations in water and Plant Basic Power load (MW) Qmin cooling (MW) Ammonia 111 750 Methanol 162 409 GTL 287 1,961 Plant Process Supply (m3/d) Process demand (m3/d) Ammonia 599 2,571 Methanol 896 1,912 GTL 16,795 7,115 Sink Source P1D1 (t/d) P1D2 (t/d) P2D1 (t/d) P2D2 (t/d) P3D2 (t/d) P1I1 (t/d) P2I1 (t/d) P3I1 (t/d) Discharge (t/d) P1S1 0 0 0 0 0 40 0 0 5 P1S2 0 0 0 0 0 0 0 0 0 P1S3 0 0 0 0 0 0 0 0 0 P2S1 0 0 0 0 0 0 139 0 0 P2S2 0 0 0 0 0 0 0 0 0 P2S3 0 0 0 0 0 0 0 0 0 P3S1 0 0 0 0 0 0 0 0 2,477 P3S2 0 0 0 0 0 0 0 0 0 FW 0 840 0 500 0 163 0 0 0 1048 energy supply and demand using the proposed method and without the need for multiperiod planning is (78 MUSD/y). One-stage and two-stage nanofiltration units were selected as the optimal treatment options. Table 4: Source-Treatment unit wastewater allocation Table 5: Treatment unit-Sink water allocation Table 6: Allocation of generated power (MW) Process Ammonia Methanol GTL Ammonia 34.55 0 7.08 Methanol 12.46 3.46 7.07 GTL 24.65 31.46 36.59 5. Conclusions This work provides a method for designing optimal water-energy networks considering seasonality to cover this gap in the literature as previous work either ignored seasonal changes or employed multiperiod planning which results in complex models that make finding the globally optimal solution difficult. It started by collecting data, analyzing seasonal changes of the network elements, and assessing the significance of the seasonal fluctuations. According to the assessment results, a novel approach was proposed to design a water-energy network that can handle seasonal variations efficiently. A MINLP model was proposed to optimize the design of the water-energy network considering seasonality. A case study consists of three processes; ammonia, methanol and GTL was considered to demonstrate the proposed approach and model. Three different options were considered for cooling purposes which are air coolers, cooling towers and once-through cooling seawater. The obtained results indicate the applicability of the proposed model which enable designing the water-energy network considering seasonality. Air coolers were selected for cooling purposes while one-stage and two-stage nanofiltration were utilized for wastewater treatment. The results show direct allocation of wastewater and indirect allocation of treated water. The power generated utilizing waste heat was allocated into power sinks. The TAC of the network is 78 MUSD/y considering seasonality without the need for multi-period planning. Nomenclature 𝐶𝐶𝑇𝑇 – Total central and decentral treatment cost, $/y 𝐶𝐶𝑟𝑟𝑖𝑖𝐶𝐶𝐶𝐶 – CAPEX parameter decentral treatment, $/kg 𝐶𝐶𝑠𝑠𝑠𝑠 𝐶𝐶𝐶𝐶 – CAPEX parameter central treatment t, $/kg 𝐶𝐶𝑟𝑟𝑖𝑖𝑂𝑂𝐶𝐶 – OPEX parameter decentral treatment, $/kg 𝐶𝐶𝑠𝑠𝑠𝑠 𝑂𝑂𝐶𝐶– OPEX parameter of central treatment, $/kg - 𝐷𝐷𝑖𝑖𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Average water discharge from source i, kg/h 𝐹𝐹𝑙𝑙,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Average freshwater flowrate, kg/h Sink Source 1S-TR-P1 (t/d) 2S-TR-P1 (t/d) 1S-TR-P2 (t/d) 2S-TR-P2 (t/d) 1S-TR-P3 (t/d) 2S-TR-P3 (t/d) P1S1 0 0 0 0 0 40 P1S2 154 0 0 0 0 0 P1S3 0 400 0 0 0 0 P2S1 0 0 142 0 0 0 P2S2 0 0 115 0 0 0 P2S3 0 0 0 500 0 0 P3S1 0 0 0 0 0 14,171 P3S2 0 0 0 0 0 147 Sink Source P1D1 (t/d) P1D2 (t/d) P2D1 (t/d) P2D2 (t/d) P3D2 (t/d) P1I1 (t/d) P2I1 (t/d) P3I1 (t/d) Discharge (t/d) 1S-TR-P1 0 0 0 0 0 0 0 0 139 2S-TR-P1 0 0 0 0 0 0 0 0 324 1S-TR-P2 0 0 0 0 0 0 0 74 158 2S-TR-P2 0 0 0 0 0 0 0 0 405 1S-TR-P3 0 0 0 0 0 0 0 0 0 2S-TR-P3 2,571 0 1,912 0 7,115 0 0 0 0 1049 𝐻𝐻𝑦𝑦 – Operating hours per year, h/y 𝐾𝐾𝐹𝐹 – Treatment cost annualizing factor, y-1 𝑀𝑀𝑖𝑖𝑖𝑖,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Average flow from source I to sink j, kg/h 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝐶𝐶𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎 – Power to cooling tower, kW 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝑂𝑂𝐶𝐶𝑆𝑆𝑂𝑂,𝑎𝑎𝑎𝑎𝑎𝑎 – Power to OCSW, kW 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑗𝑗′𝑖𝑖 𝐴𝐴𝐶𝐶,𝑎𝑎𝑎𝑎𝑎𝑎 – Power to air cooler, kW 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Power from source i', kW 𝑅𝑅𝑅𝑅𝑐𝑐,𝑟𝑟𝑖𝑖𝑚𝑚𝑖𝑖𝑛𝑛- Minimum removal ratio, (1) 𝑇𝑇𝑖𝑖𝑖𝑖,𝑟𝑟𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Average flow from to treatment r, kg/h 𝑇𝑇𝑖𝑖𝑖𝑖,𝑠𝑠𝑠𝑠 𝑎𝑎𝑎𝑎𝑎𝑎– Average flow from to treatment s, kg/h 𝑇𝑇𝑟𝑟𝑖𝑖,𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Treatment unit r flow to sink j, kg/h 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑟𝑟𝑖𝑖 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎– Power from source i' to decentral treatment, kW 𝑇𝑇𝑠𝑠𝑠𝑠.𝑗𝑗𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 – Treatment unit s flow to sink j, kg/h 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑠𝑠𝑠𝑠 𝑇𝑇,𝑎𝑎𝑎𝑎𝑎𝑎– Power from source i' to central treatment, kW 𝑇𝑇𝑚𝑚𝑖𝑖,𝑗𝑗𝑖𝑖 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 – Desalinated water flow to sink j, kg/h 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑚𝑚𝑖𝑖 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 – Power from i' to decentral desalination, kW 𝑊𝑊𝑖𝑖𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎– Average water supply from source i, kg/h 𝑃𝑃𝑊𝑊𝑖𝑖′𝑖𝑖,𝑛𝑛𝑛𝑛 𝐷𝐷𝐷𝐷𝑠𝑠,𝑎𝑎𝑎𝑎𝑎𝑎 – Power from i' to central desalination, kW 𝑥𝑥𝑐𝑐,𝑟𝑟𝑖𝑖 𝑇𝑇,𝑚𝑚𝑎𝑎𝑚𝑚– Maximum outlet concentration, ppm References Al-Mohannadi D.M., Kwak G., Linke P., 2020. Identification of optimal transitions towards climate footprint reduction targets using a linear multi-period carbon integration approach. Computers and Chemical Engineering, 140, 106907. Alnouri S.Y., Linke P., El-Halwagi M., 2015. A synthesis approach for industrial city water reuse networks considering central and distributed treatment systems. Journal of Cleaner Production, 89, 231–250. Arredondo-Ramírez K., Rubio-Castro E., Nápoles-Rivera F., Ponce-Ortega J.M., Serna-González M., El- Halwagi M.M., 2015. Optimal design of agricultural water systems with multiperiod collection, storage, and distribution. Agricultural Water Management, 152, 161–172. Bishnu S., Linke P., Alnouri S.Y., El-Halwagi M., 2017. Multi-Period Water Network Synthesis for Eco Industrial Parks considering Regeneration and Reuse. Chemical Product and Process Modelling, 12(3), 20160049. Bishnu S., Linke P., Alnouri S.Y., El-Halwagi M., 2014. Multiperiod planning of optimal industrial city direct water reuse networks. Industrial and Engineering Chemistry Research, 53(21), 8844–8865. Bungener S., Hackl R., Van Eetvelde G., Harvey S., Marechal F., 2015. Multi-period analysis of heat integration measures in industrial clusters, Energy, 93, 220–234. Burgara-Montero O., Ponce-Ortega J., Serna-González M., El-Halwagi M., 2013. Incorporation of the seasonal variations in the optimal treatment of industrial effluents discharged to watersheds. Industrial and Engineering Chemistry Research, 52(14), 5145–5160. Floudas C.A., Grossmann I.E., 1986. Synthesis of flexible heat exchanger networks for multiperiod operation. Computers and Chemical Engineering,10(2), 153–168. Fouladi J., Linke P., El-Halwagi M., Parsaei H., 2017. A Systematic Approach For Designing Industrial Park Integration Networks Across The Water-Energy Nexus. MSc Dissertation, Texas A&M university, Doha, Qatar. Gaudard L., Avanzi F., De Michele C., 2018, Seasonal aspects of the energy-water nexus: The case of a run- of-the-river hydropower plant. Applied Energy, 210, 604–612. Isafiade A., Short M., Bogataj M., Kravanja Z., 2017. Integrating renewables into multi-period heat exchanger network synthesis considering economics and environmental impact. Computers and Chemical Engineering, 99, 51–65. Kim J.H., Han C., 2001. Short-Term Multiperiod Optimal Planning of Utility Systems Using Heuristics and Dynamic Programming., Industrial & Engineering Chemistry Research, 40(8), 1928–1938. Liew P.Y., Wan Alwi S.R, Ho W.S., Manan Z.A., Varbanov P.S., Klemeš J.J., 2018. Multi-period energy targeting for Total Site and Locally Integrated Energy Sectors with cascade Pinch Analysis. Energy, 155, 370–380. Papalexandri K.P., Pistikopoulos E.N., 1993. A Multiperiod MINLP Model for Improving the Flexibility of Heat Exchanger Networks., Computers and Chemical Engineering, 17, S111–S116. Taheri S., 2021. A Multi-Period Water Network Planning for Industrial Park; Impact of design period on Park’s Flexibility, DOI: 10.48550/arXiv.2108.01047. Zhou W., Iqbal K., Lv X., Deng C., 2021. Optimal design and operation of multi-period water supply network with multiple water sources, Processes, 9(12), 2143. 1050 PRES22_0357.pdf Synthesis of Water-Energy Network Considering Seasonality