CHEMICAL ENGINEERING TRANSACTIONS VOL. 56, 2017 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: Jiří Jaromír Klemeš, Peng Yen Liew, Wai Shin Ho, Jeng Shiun Lim Copyright © 2017, AIDIC Servizi S.r.l., ISBN 978-88-95608-47-1; ISSN 2283-9216 Synthesis of More Sustainable Total Site Andreja Nemet, Zdravko Kravanja* University of Maribor, Faculty of Chemistry and Chemical Engineering, Smetanova ulica 17, 2000 Maribor, Slovenia zdravko.kravanja@um.si Heat recovery between different streams of Total Site (TS) can significantly decrease external utility consumption. Decreased utility consumption leads to decreased impact on the environment resulting from lower GHG emissions produced by fuel. The optimal rate of heat recovery can be determined by establishing an appropriate trade-off between utility consumption and investment in heat transfer equipment. Therefore, it can be concluded that both economic and environmental pillars of sustainability encourage the goal of constructing TSs. However, the social pillar of sustainability should also be considered in order to ensure the overall sustainability of a TS. The social pillar is usually not included in the synthesis because different aspects of social performance are quite difficult to quantitatively determine. Safety is an important social aspect when considering TS, especially when TS complexes are located across areas with high population density where any failure is a potential source of events with catastrophic consequences. In order to construct a safe TS, one option is to perform risk assessment before the synthesis and forbid highly risky matches at the synthesis step or to perform the synthesis and the risk assessment simultaneously. In this study, risk assessment was performed simultaneously during the synthesis in order to design TSs with as low risks as are socially acceptable. The synthesis was performed in two steps. In Step 1 a globally optimal solution was obtained based on a simplified trade-off between investment and operating cost while simultaneously considering risk assessment using a mixed-integer linear programming (MILP) TransGen model, while in Step 2 a detailed synthesis considering risk assessment was performed on a reduced superstructure obtained in Step 1 with a mixed-integer nonlinear programming (MINLP) model called Total Site Synthesis model, which explicitly considers risk limits during optimization. The risk depends on the frequency of failures and the severity of the consequences. The former can be reduced by the selection of more suitable equipment and the latter by selecting indirect rather than direct heat transfer, selecting smaller sizes and safer operating conditions. The minimization of total annual costs (TAC) is a primary objective of this synthesis. There are significant differences in the results obtained when safety is not considered and when lower risk limits are set in order to obtain safer designs. It can be concluded that by performing TS synthesis using the proposed synthesis model, the inherent safety of the TS is significantly increased; however, this incurs economic expense. 1. Introduction Projections of energy consumption indicates that world energy consumption is increasing worldwide; even more alarming is the fact that consumption of fossil based fuels such as petroleum and other liquids, coal and natural gas is still increasing (EIA, 2016). Therefore, with current consumption trends, it is unrealistic to expect decrease in fossil fuel consumption, despite different international energy treaties such as the Kyoto Protocol (UNFCCC, 1998), or the latest attempt, the Paris Agreement, that has not yet been ratified by all Parties (UNFCCC, 2016). These are good initial attempts; however, achieving sustainable development is still in the initial stage. When considering global carbon emissions from fossil fuel burning, it can be seen that it is still increasing (C2ES, 2015). The IEA presented a Blue Map scenarios, where possible technologies for reducing CO2 emission were listed. In this prediction, the greatest reduction in CO2 emissions can be obtained by the end-use fuel and electricity efficiency (38 %), in the scenario, when CO2 emissions would be reduced to 14 Gt by the year 2050 (IEA, 2010). Heat Integration can significantly contribute to the CO2 emission reduction as a result of utility demand decrease. The savings can be even higher when TS synthesis is considered. However, DOI: 10.3303/CET1756004 Please cite this article as: Nemet A., Kravanja Z., 2017, Synthesis of more sustainable total site, Chemical Engineering Transactions, 56, 19- 24 DOI:10.3303/CET1756004 19 for successful implementation of TS, its heat exchanger network in the proposed design should be feasible and acceptable. One of the most important aspects of acceptability is the safety of the design obtained. Until now, a great deal of work has focused on enhancing the safety of already existing plans, estimating potential deviation events, and their consequences. However, the application of safety metrics as part of the design of many unit operation and chemical processes is still in an early stage of development (Roy et al., 2016). Moreover, existing metrics are mainly qualitative (Marhavilas et al., 2011). Jung et al. (2010) presented a methodology for optimizing the placement of hazardous processes and other facilities using mixed-integer nonlinear programming considering the risk map of a plant area. Kim et al. (2011) tested a qualitative index-based approach in the case of hydrogen infrastructures, evaluating different scenarios. Shariff et al. (2012) dealt with the identification of critical streams with high explosion potential in order to indicate critical points in networks via Process Stream Index (PSI). A similar study for toxic releases was later conducted by Shariff and Zaini (2013). Strictly focusing on safety in a Heat Exchanger Network, there has been some work conducted by Chan et al (2014) who presented a combined methodology of Stream Temperature vs. Enthaphy Plot (STEP) with risk assessment. Liu et al. (2015) assessed the risk in TS using a step-by-step procedure considering direct or indirect heat transfer. Vázquez-Román et al (2015) used a cause-effect analysis in mathematical programming approach to determine optimal layout of facilities considering toxic releases. Inchaurregui- Méndez et al. (2015) considered inherent safety when synthesizing Heat Exchanger Networks based on layout with allocation of hot/cold streams. Previously, we developed a sequential step-by-step approach for considering safety in a TS (Nemet et al, 2015). The methodology presented in that work cut off some heat transfer options with the highest heat integration potential as matches were assessed prior to synthesis based on maximal heat transfer. In this study, risk assessment is performed during the synthesis of TS via modelling the interactions between risk and selected Heat Exchanger Networks, leading to better optima. A two-step procedure was developed for obtaining TS design. In Step 1, a global solution is obtained based on simplified area, pipeline, operating cost and risk assessment using MILP model TransGen, while in Step 2 more detailed synthesis is performed using a MINLP Total Site synthesis model, based on the solution obtained in Step 1 that serves as the initialization and prescreening of the alternatives. 2. Methodology Obtaining a TS design is a complex problem, especially when synthesizing the HEN at both levels of integration simultaneously, within and between processes. Accounting for pressure/temperature drops, heat losses, and pipe design optimisation leads to a highly non-linear model, capable only of solving very small problems and obtaining poor locally optimal solutions. The two-step approach mentioned above was developed in order to obtain solutions closer to global optima and to solve larger problems. In Step 1 the TransGen model for TS area/energy targeting is thus used to identify promising alternatives to be developed further in the second step. Note that risk assessment is performed simultaneously with the targeting process. As the TransGen model is formulated as a MILP model, the solutions obtained are globally optimal; however, because of model simplifications, the trade-off obtained between operating cost and investment of heat exchangers and pipes is rough and should be improved in the second step based on a detailed MINLP synthesis model. As Step 2 is performed on a reduced superstructure identified from the global solution obtained in Step 1, larger problems can now be solved in a rigorous way and the final solutions can thus be guided close to global optima. Note that risk assessment is performed simultaneously in the second step, too. In Step 1, the TransGen (Nemet et al, 2016) targeting model for the synthesis based on an extended transhipment model using temperature intervals is used. Although area calculation is simplified, it is still realistic enough in order to achieve a reasonable trade-off between investment and operating cost; moreover, the risk assessment is also based on the determined area. For this purpose, the heat transfers between process streams are explicitly determined as, qi,j,k,kk representing the heat transferred from hot stream i released at interval k to cold stream j consuming the heat at the same or lower temperature interval kk. The area is determined for each heat transfer from each interval k to the same or lower temperature interval kk. Note that besides indirect heat transfer between processes via intermediate utilities, direct heat transfers between processes are also allowed, assuming a transport of cold stream from one process to the hot streams of the other process and after returning to its original process. Heat losses are accounted for in both cases. It should be noted that risk assessment is determined for each heat exchanger. This is defined as a multiplication of the failure frequency and the severity of the consequences (Eq. 1). The failure frequency ffail can be determined based on historical data, while the severity of the consequences is determined based on the amount of the substance present in the heat exchangers and its properties and operating conditions. The amount of the substance is determined by multiplication of the area of the heat transfer Ahx and the density of 20 the substance divided by area density βhx of the heat exchanger (ratio between area and volume of the medium in the heat exchanger). The operating conditions and placement of heat exchangers are considered via factors f1-f3 and the limiting value Grisk for the considered risk type (toxicity, flammability, explosiveness) depending on the properties of the substance. 1 2 3 , , ⋅ ⋅ ⋅ ⋅ ⋅ = ⋅ hx sfail hx hx s s hx risk hx rx iskh f fA f R f G ρ β ∀s∈S, hx∈HX, risk∈RISK (1) It should be noted that the area in the first step is somewhat overestimated. Larger heat exchangers lead to higher risk; therefore, they reach the upper limit of individual risk limit earlier compared to the optimised area. Therefore, parallel heat exchangers between the same hot and cold streams are included as an option not to cut off matches with high heat integration potential and still acceptable risk. For this purpose, the number of parallel heat exchangers is selected via binary variable yp that is related to coefficient k phe presenting the number of parallel heat exchangers. =new phe phei , j i , j , phe phe A k A⋅ ∀i∈I, j∈J (2) max1/phe phei , j , phe pA k y A≤ ⋅ ⋅ ∀i∈I, j∈J, phe∈PHE (3) Another binary i , j , phey , presenting exactly the heat transfer between hot stream i, cold stream j and number of parallel heat exchangers phe is introduced in order to enable the risk assessment of pipe when direct heat exchange between process streams belonging to different processes is selected for a heat match. i , j , phe i , jy y≤ ∀i∈I, j∈J, phe∈PHE (4) i , j , phe py y≤ ∀i∈I, j∈J, phe∈PHE (5) + -1i , j , phe i , j py y y≥ ∀i∈I, j∈J, phe∈PHE (6) The risk for heat transfer between a pair of hot and cold process streams within a process is determined as the risk of substance of the hot process streams in a heat exchanger as well as the risk of medium in the cold process stream. For direct heat transfer between different processes, the risk assessment for a pipe should be added: 1 2 3 , , , cp , , , 2 600 yhx hx cpfail phes hx risk hp cp hp cp phe cp risk f f f R f L qm G ⋅ ⋅ = ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ∀s∈S, hx∈HX, risk∈RISK (7) Note that for the risk of a pipe, when utilising indirect heat transfer via steam, a zero value is used as steam is neither toxic, flammable nor explosive. The objective function is defined as total annual cost, where the additional investment for Heat Integration is considered: ( ) ( ) ( ) ( )CU op CU HU op HU HU op HU HU op HU + + annualhp cp icup Ccond hp cp icup Ccond Q t c Q tTAC c Q t c Q t c f I= ⋅ ⋅ + ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ + ⋅    (8) 3. Case study 3.1 Input data Different types of heat exchangers are considered during optimisation of the heat exchanger network of a TS (Table 1). The case study consisted of two processes, P1 and P2. Process P1 includes six hot streams and two cold streams, process P2 consists of two hot and four cold process streams. The input data for these processes is presented in Table 2. It should be noted that additional data were required in order to perform risk assessment. Four different intermediate utilities were available for indirect heat transfer between processes (Table 3). The temperatures of intermediate utilities are fixed during the optimisation in Step 1 (Tfix,Step 1), while in Step 2 they are considered as optimisation variables within given upper and lower temperature bounds as presented in Table 3. 21 Table 1: Input data for different heat exchanger types Type of HE Amax/m2 TLO /°C TUP /°C cf/ k€ cv / k€/m2 β / (m2/m3) ffail/y-1 Double pipe 200 -100 600 46 2.742 80 0.009929 Plate and frame 1,200 -25 250 129.8 0.347 80 0.010908 Fixed plate shell and tube 1,000 -200 850 121.4 0.193 720 0.009029 Shell and tube with U-tubes 1,000 -200 850 100.9 0.304 1,300 0.009929 Evaporator 1,000 -10 600 174.4 0.919 720 0.009929 Condenser 1,000 -10 600 105.6 0.272 720 0.009929 Table 2: Input data for process streams Process Stream Tin /°C Tout/°C FC/ (kW/(m2.K) h/ (kW/(m2.K)) LC50 (rat, inh, 1h)/ (mg/m3) Qexpl (kJ/kg) flammable P1 H1P1 445 303 9 0.33 500 500 Yes H2P1 398 25 10 0.36 900 0 Yes H3P1 436 297 19 0.35 300 100 Yes H4P1 389 32 8 0.42 100 5 Yes H5P1 451 300 24 0.38 800 10 Yes H6P1 401 44 11 0.44 5000 5 Yes C1P1 40 80 45 0.40 500 0 Yes C2P1 50 60 15 0.36 600 500 Yes P2 H1P2 148 90 16 0.33 100 5 Yes H2P2 127 75 12 0.4 50 10 Yes C1P2 100 248 45 0.43 700 5 Yes C2P2 76 325 10 0.38 500 0 Yes C3P2 100 236 25 0.45 600 500 Yes C4P2 80 318 15 0.38 100 100 Yes Table 1: Intermediate utility properties Intermediate utility TLO /°C TUP /°C Tfix,Step 1 / °C h /(kW/(m 2 °C)) Low pressure steam 120 148 136 10,000 Medium pressure steam 148 208 180 10,000 High pressure steam 208 252 230 11,000 Ultra-high pressure steam 252 275 280 11,000 3.2 Solutions 3.2.1 Without risk assessment First, a solution without risk assessment was obtained that served as a reference from the safety point of view. The optimal scheme of TS is shown in Figure 1. The HEN for the TS consisted of 20 HEs, of which 12 are fixed plate shell and tube, 6 double pipe and 2 are shell and tube with U-tubes type. Note that only direct heat transfer without intermediate utilities was selected. The TAC was 2,659 k€/y with 315 kW consumption of hot utility and 564 kW of cold utility. It should be noted that this scheme enables a 98.26 % reduction of GHG emissions as a result of Heat Integration within and between the processes. The risk recalculated after optimization was 0.048 y-1 for toxicity, 0.002 y-1 for flammability and 4.44 x 10-7 y-1 for explosiveness. 3.2.2 Simultaneous risk assessment In the second case, the risk assessment was performed simultaneously by setting an upper bound on the overall risk at one-half of initial risk in order to obtain twice safer design than in the first case. The safer TS is shown in Figure 2. The safer HEN of TS now consists of an increased number of HEs – 33, of which 11 are fixed plate shell and tube, 6 are plate and frame, 7 are double pipe, 5 are shell and tube with U-tubes, two evaporators and two condensers. Note that besides direct, indirect heat transfer via medium pressure steam (MPS) was also selected with a 70 % recycle of condensate and preheating of fresh water and condensate streams. The TAC was increased to 3,268 k€ as a result of more than double the hot utility consumption, now 851 kW, and cold utility consumption of 754 kW. The reduction of GHG emissions caused by Heat Integration is now slightly worse than the first case, a somewhat smaller 95.29 %. 22 Figure 1: Optimal Total Site when no risk assessment is performed Figure 2: Twice safer Total Site obtained with upper bound on the overall risk set at one-half of initial risk. 23 4. Conclusion A two-step procedure for TS synthesis with embeded safety analysis has been developed. The solutions obtained indicate that safety improvements can be obtained at the cost of economic expense. A further observation is that safer designs may exhibit somewhat larger GHG emissions as the level of Heat Integration can be lower. In future studies, the synthesis of TS, besides setting limits on the overall risk, will be performed also by setting limits on individual units (heat exchangers and pipes), because it may also significantly affect the final design. Also, a composed objective, including both economic and safety aspects directly into the objective function, is planned. This will enable an optimal TS design by performing the synthesis in a single optimization, rather than sequentially by executing Pareto solutions at different risk limits. More appropriate trade-offs between utility consumption, investment and safety can be obtained in this way. The methodology presented here will be extended to other process subsystems as well as to overall process systems. Acknowledgments The authors acknowledge financial support from the Slovenian Research Agency (programs P2-0032, project L2-7633 and scheme “Spodbujanje mladih doktorjev znanosti“ contract number: 1547/FKKT-2015). References C2ES (Centre for Climate and Energy Solutions), 2015, Historical global CO2 emissions, , accessed 8.10.2016 Chan I., Wan Alwi S.R., Hassim M.H., Manan Z.A., Klemeš J.J., 2014, Heat Exchanger Network Design Considering Inherent Safety, Energy Procedia 61, 2469-2473. Energy Information Agency – EIA, 2016, International Energy Outlook 2016, accessed 10.11.2016 Inchaurregui-Méndez J.A., Vázquez-Román R., Ponce-Ortega J.M., Sam Mannan M., 2015, A Heat Exchanger Networks Synthesis Approach Based on Inherent Safety, Journal of Chemical Engineering Research Updates 2, 22-29. IEA (International Energy Agency), 2010, Energy Technology Perspectives, , accessed 15.11.2016 Jung S., Ng D, Laird C.D., Mannan M.S., 2010, A new approach for facility siting using mapping risks on a plant grid area and optimization, Journal of Loss Prevention in the Process Industries 23, 821-830. Kim J., Lee Y., Moon I., 2011, An index-based risk assessment model for hydrogen infrastructure, International Journal of Hydrogen Energy 36, 6387-6398. Liu X., Klemeš J.J., Varbanov P.S., Qian Y., Yang S., 2015, Safety issues consideration for direct and indirect heat transfer on total sites, Chemical Engineering Transactions 45, 151-156. Marhavilas P.K., Koulouriotis D., Gemeni V., 2011, Risk analysis and assessment methodologies in the work sites: On a review, classification and comparative study of the scientific literature of the period 2000-2009, Journal of Loss Prevention in the Process Industries 24, 477-523. Nemet A., Čuček L., Kravanja Z., 2016, Procedure for the simultaneous synthesis of heat exchanger networks at process and total site level, Chemical Engineering Transactions 52, 1057-1062. Nemet A., Lee K., Klemeš J.J., Varbanov P.S., Moon I., Kravanja Z., 2015, Safety analysis embedded in total site synthesis, Chemical Engineering Transactions 45, 121-126. Roy N., Eljack F., Jiménez-Gutiérrez A., Zhang B., Thiruvenkataswamy P., El-Halwagi M., Sam Mannan M, 2016, A review of safety indices for process design, 2016,Current Opinion in Chemical Engineering 14, 42- 48. Shariff A.M., Leong C.T., Zaini D., 2012, Using process stream index (PSI) to assess inherent safety level during preliminary design stage, Safety Science 50 (4), 1098–1103 Shariff A.M., Zaini D., (2013) Inherent risk assessment methodology in preliminary design stage: A case study for toxic release, Journal of Loss Prevention and the Process Industries 26 (4), 605-613. United Nations Framework Convention on Climate Change, 1998, Kyoto Protocol, , accessed 10.10.2016 United Nations Framework Convention on Climate Change, 2015, , accessed 12.10.2016 Vázquez-Román R., Inchaurregue-Méndez J.A., Sam Mannan M., 2015, A grid-based facilities allocation with safety and optimal heat exchanger network synthesis, Computers and Chemical Engineering 80, 92-100. 24