CCHHEEMMIICCAALL EENNGGIINNEEEERRIINNGG TTRRAANNSSAACCTTIIOONNSS VOL. 45, 2015 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Sharifah Rafidah Wan Alwi, Jun Yow Yong, Xia Liu Copyright © 2015, AIDIC Servizi S.r.l., ISBN 978-88-95608-36-5; ISSN 2283-9216 DOI:10.3303/CET1545270 Please cite this article as: Tapia J.F.D., Lee J.-Y., Ooi R.E.H., Foo D.C.Y., Tan R.R., 2015, Design and scheduling of CO2 enhanced oil recovery with geological sequestration operations as a strip packing problem, Chemical Engineering Transactions, 45, 1615-1620 DOI:10.3303/CET1545270 1615 Design and Scheduling of CO2 Enhanced Oil Recovery with Geological Sequestration Operations as a Strip Packing Problem John F. D. Tapia a , Jui-Yuan Lee b , Raymond E. H. Ooi c , Dominic C. Y. Foo c , Raymond R. Tan a * a Chemical Engineering Department, De La Salle University, Manila, Philippines. b Department of Chemical Engineering and Biotechnology, National Taipei University of Technology, Taipei, Taiwan c Department of Chemical and Environmental Engineerin, University of Nottingham Malaysia Campus, Semenyih, Malaysia raymond.tan@dlsu.edu.ph Carbon capture and storage (CCS) is an important technology that enables the reduction of CO2 emissions while allowing the use of power plants running in fossil fuels. Coupled with enhanced oil recovery (EOR), it allows additional revenues to be realized by increased oil production through the injection of captured CO2 into depleted oil reservoirs. For a system with multiple reservoirs, it is necessary to determine the amount of CO2 to be supplied and the schedule of EOR operations to be conducted. This also requires the optimal design and scheduling of each operation to maximize profitability. In this study, a mixed integer linear program (MILP) was developed using the analogy of the strip packing problem. A case study is presented to illustrate the model. 1. Introduction Carbon capture and storage (CCS) is an important technology for reducing industrial carbon dioxide (CO2) emissions. It involves removing potential CO2 emissions from flue gas, transporting through pipelines and storing it into geological reservoirs for permanent storage (Davison et al, 2001). The most mature technique for permanently storing CO2 is the injection into depleted oil reservoirs for enhanced oil recovery (EOR), which could provide initial momentum for further large-scale CCS deployment. According to the current worldwide status of CCS, 25 % of the CCS operations in the world are in the operation stage, and 63% of them are applied for EOR operations (Global CCS Institute, 2012). Coupling CO2 sequestration with EOR could provide long term storage potential in a particular CO2 sink benefiting the environment; whilst serving as a business driver to oil companies on improving oil recovery.CCS operations are highly dependent on the 3 critical success factors on CO2 capture, CO2 transport and CO2 storage. Both CO2 capture and CO2 storage involves technology selection, R&D maturation and large scale deployment which is widely discussed today as compared to CO2 transport. CO2 transport revolves around transporting the captured CO2 to the storage sites, typically via pipeline networks.However, an effective pipeline network should be synthesized to minimize the cost of pipeline infrastructure, and to minimize the required CO2 supply to execute EOR operations. This would require the proper scheduling of each operation to minimize the annual CO2 requirement. Various models for CCS source-sink matching with sequestration scheduling have been developed using both discrete-time (Tan et al., 2013), continuous-time (Tan et al., 2012) and simplified continuous-time (Lee and Chen, 2012) approaches, while a unified approach considering both power generation losses and source-sink matching has been proposed (Lee et al., 2014). Addressing uncertainties to manage storage risks using fuzzy optimization (Tapia and Tan, 2014) and new information to modify the existing network using optimal revamp (Tapia and Tan, 2015) have been addressed in previous papers. Cormos and Simon (2013) developed a dynamic model for CO2 capture utilizing-calcium looping cycle. Also, several 1616 approaches have been proposed to optimize EOR operations; for example, Jahangiri and Zhang (2012) and Leach et al. (2011) both addressed the economic viability of operations. These studies, however, only address the economics of CCS alone, or CCS in combination with EOR, but lack consideration to optimized scheduling of each oil reservoir. This aspect can be useful for minimizing the annual CO2 requirement for the entire planning horizon. This paper demonstrates that the scheduling and design of EOR operations can be represented as a strip packing problem (Riff et al., 2009), since its economic and temporal parameters can be represented in the components of strip packing. In the context of this study, the strip represents the CO2 source of definite planning horizon (strip width) and the packing objects represents the EOR operations with definite flow rate requirement (object height) and definite operation length (object width). The design and scheduling of EOR operations for multiple oil reservoirs has not yet been made. In this study, this problem has been addressed in this paper as a strip packing problem. The rest of the paper is organized as follows: Section 2 discusses the problem statement while Section 3 shows the optimization model based on the strip packing problem. Section 4 illustrates the model using a simple case study. Lastly, Section 5 shows the conclusions and future works. 2. Problem Statement The formal problem statement addressed in this paper is as follows:  The system consists of m depleted oil reservoirs and single CO2 source.  In this study, the CO2 source operates for T years. In strip-packing, the planning horizon scheduled for the CO2 source is considered the strip width, (PH) of the strip of infinite height. It is assumed that the CO2 source can supply all operations’ CO2 requirement, but the annual CO2 supply incurs costs for the design of the pipeline infrastructure.Note that the other cost elements related to CO2 capture (e.g membrane, amine unit & etc) and compression for transport is not considered as a cost variable in this work.  The injection to each reservoir yields an amount of oil equal to (OR)i. The oil recovered can be sold at a price denoted by (OP)I depending on its value. Each operation can start between the earliest time (Ti) e and the latest time (Ti) l . It has a fixed duration (object width) of (ΔT)i and requires a flow rate (object height) equal to fi.  The amount of CO2 stored in each reservoir is defined as Si of the CO2 injected. The amount of CO2 stored is credited by $C* per t CO2 stored.The amount of CO2 stored permanently is assumed to be traded in exchange for carbon credits (Roussanaly and Grimstad, 2014).  The cost of transporting and injecting CO2 to the reservoir is estimated using a linear cost function with a fixed cost component of (Cf)pp and a variable cost component of (Cv)pp. The variable cost component is proportional to the flow rate to the reservoir and the distance of the reservoir from the source.  For the pipeline infrastructure, a common pipeline is routed from the source up to a common point (typically an offshore platform/structure) and redistributed via individual pipelines branches from the common point to respective reservoirs. It is assumed that the common pipeline and the individual pipeline distances are predetermined based on the geographical and geological conditions, while the sizes are determined based on the flow rates of CO2 transported. 3. Optimization Model The objective function is to maximize the profit generated based on the total CO2 credits and the oil recovered: TOTALCOSTCARBCREDOILREVPROF max (1) Subject to:   i iiii fTOROPOILREV )]()()()[( (2)   i iii fTSCCARBCRED )]())()([( * (3) INDIVPIPECOMMONPIPETOTALCOST  (4) ))(()()( HDCvCfCOMMONPIPE pppp  (5) ))(()()(  i iipppp fDCvCfINDIVPIPE (6) where (OILREV), (CARBCRED) and (TOTALCOST) are the oil revenue generated, carbon credits obtained from storage and the total costs for infrastructures, respectively, in all the operations. The oil 1617 revenue generated is based on the oil yield, (OR)i and the oil value (OP) i for each reservoir, while the carbon credits is based on the sequestration parameter (S)i, which denotes how much CO2 is stored permanently with respect to how much CO2 is injected. The total cost in this model is based only on the pipeline infrastructure. The pipeline structure is based on two parts: the common pipeline (COMMONPIPE) and the individual pipelines (INDIVPIPE) for redistribution to the individual reservoirs. The cost function is assumed to be a linear function with a fixed cost and a variable cost proportional to the flow rate requirement and the pipeline length. Also the common pipeline is based on the overall height of the strip used. The design of the EOR infrastructure is based on these constraints while the scheduling will strongly be based on the strip packing approach. The positioning variables in the strip packing approach are shown as follows: )(])()[(5.0 ',','' iiiiiiii QPPHxxTT  iiiii  ,',, (7) )1(])()[(5.0 ',','' iiiiiiii QPPHxxTT  iiiii  ,',, (8) )1(][5.0 ',','' iiiiiiii QPMyyff  iiiii  ,',, (9) )2(][5.0 ',','' iiiiiiii QPMyyff  iiiii  ,',, (10) For Eqs(7-10), the x-axis is represented by the length of EOR operation while the y-axis is the flow rate requirement. The coordinates shown in the model are the coordinates of the center of the boxes in the strip. Pi,i’ and Qi,i’ are binary variables that determines the position of the object i with respect to object i’; Pi,i’ denotes which axis are taking in consideration while Qi,i’ denotes whether the object is ahead or behind the other. These equations, based on Castro and Grossmann’s (2012) continuous strip packing approach, ensure that no rectangles overlap. In this system, the strip packing method allows the determination of the maximum flow rate requirement for the common pipeline, which is only possible by using the strip packing approach. The boundaries for xi and yi are based on scheduling constraints in which for every object, its rightmost side should be located on a specific boundary: ])[(5.0 ii Tx  i (11) ])[(5.0 ii TPHx  i (12) )(])[(5.0 e iii TTx  i (13) )(])[(5.0 l iii TTx  i (14) ii fHy 5.0 '  i (15) }1,0{ ',  ii P ', ii (16) }1,0{ ',  ii Q ', ii (17) Eqs (11)-(14) denotes the range at which xi can be placed, implying that the operation can only start between the earliest specified time (Ti) e and the latest specified time (Ti) l . The model is tested using a case study in Lingo 14.0 in a PC with 2.40 GHz processor and 4.00 Gb RAM. In this case, the time required to obtain the optimal solution is negligible. 4. Case Study Four reservoirs are assumed to be connected to a source which operates at 30 y in the planning horizon (strip width). Table 1 shows the data for the oil reservoirs while Table 2 shows the parameters required to assess the economics of the entire EOR system. Table 1: Reservoir Data for the Case Study Sink 1 Sink 2 Sink 3 Sink 4 Earliest Start of Operation, (Ti) e , y 0 0 0 0 Latest Start of Operation, (Ti) l , y 5 7 15 10 Flow Rate Requirement, fi 2.4 3.0 1.5 2.5 Oil Recovery, (OR)i, million bbls/Mt 10 4 8 5 Sequestration Parameter, Si 0.95 0.85 0.90 0.95 Operating Life for EOR, (ΔT)i, y 10 15 20 10 Distance from the common point (km) 100 200 150 150 1618 Table 2: Cost Data for Case Study Solving the model yields the optimal solution shown in Figures 1 and 2. Figure 1 illustrates the optimal schedule as the strip packing, and Figure 2 the pipeline network with a common pipeline and four individual pipelines. The flow rates for each operation are also indicated. Figure 1: Strip Packing Solution for the Case Study Figure 2: Pipeline Design and Operation Schedule for the Case Study The total profit generated is equal to US$ 57.4 billion, 97 % of which comes from the oil revenue. As shown, Operations 3 and 4 should start after 10 years, while Operations 1 and 2 at the beginning of the Value Length of Planning Horizon, y 30 Oil Price, $/barrel 70 Carbon Credit, M$/Mt 23 Fixed cost, million$ 20 Variable Cost, million$/Mt-km 0.01 Common Pipeline Distance 150 1619 planning horizon. With this schedule, the CO2 flow rate from the source has a maximum of 7 Mt/y during 10-15 y, and is delivered using a 150 km common pipeline. As shown in Figure 2, the flow rate supplied by the source is 5.4 Mt/y for the first 10 y, 7 Mt/y up to 15 y, 4 Mt/y up to 20 y and 1.5 Mt/y up to 30 y. This shows that the capacity used for the common pipeline ranges from 21 % up to 100 %. If all operations overlap, the maximum total CO2 flow rate required is equal to 9.4 Mt/y, which is 34 % greater than the minimum flow rate required (7 Mt/y). For the common pipeline design, the minimum flow rate that would occur is 1.5 Mt/y when only Operation 3 is active in the later stage of the planning horizon. The strip packing approach for this problem allows the minimization of the capacity required for the common pipeline by scheduling the different operations. Results from the strip packing approach can be translated into critical design decisions such as selecting the optimized pipeline size, compressor capacity or even deciding on the operations philosophy (timing of operation, well shut-in & etc) in EOR management. Table 3 summarizes the economic performance of the EOR system Table 3: Economic Result for EOR Case Study Cost Parameter Value (M$) Profit 57,396 Oil Revenue 54,950 Carbon Credits 2571.40 Costs 30.50 (common pipeline) 94.40 5. Conclusions and Future Work A strip packing problem-based optimization model for the design and scheduling of EOR operations with geological sequestration has been developed. The model aims to minimize the CO2 supplied by the CO2 source to meet the flow rate requirements of EOR operations. The model also aims to maximize the total profit based on CO2 credits, oil recovery and the costs for the pipeline infrastructure with a common pipeline and individual pipelines for each reservoir. The strip packing approach is utilized to schedule EOR operations to minimize the total CO2 flow rate from the source. Future work includes the extension of the current model to consider uncertainties in the parameters such as oil and carbon prices, oil yield and storage capacity and to consider multiple CO2 sources with various profit allocation schemes. 6. Acknowledgement J.F.D. Tapia would like to acknowledge the Department of Science and Technology through its Engineering Research and Development for Technology (DOST-ERDT) program and the Commission on Higher Education through the Philippine Higher Education Research Network (CHED-PHERNet) for financial support. References Castro P.M., Grossmann I.E., 2012, From time representation in scheduling to the solution of strip packing problems, Comput. Chem. Eng., 44, 45-57. Cormos A.M., Simon A., 2013, Dynamic modelling of CO2 capture by calcium-looping cycle, Chemical Engineering Transactions, 35, 421–426. Davison J., Freund P., Smith A., 2001, Putting carbon back into the ground, International Energy Agency Greenhouse Gas R&D Programme, Cheltenham, UK. Global CCS Institute, 2012, The Global Status of CCS 2012, Global CCS Institute, Canberra, Australia. Jahangiri H.R., Zhang D., 2012, Ensemble based co-optimization of carbon dioxide sequestration and enhanced oil recovery, Int. J. Greenh. Gas Control, 8, 22-33. Leach A., Mason C.F., Veld K.V.T., 2011, Co-optimization of enhanced oil recovery and carbon sequestration, Resource and Energy Economics, 33, 893-912. Lee J.Y., Chen C.L., 2012, Comments on “Continuous-Time Optimization Model for Source-Sink Matching in Carbon Capture and Storage Systems”, Ind. Eng. Chem. Res., 51, 11590–11591. Lee J.-Y., Tan R.R., Chen C.-L., 2014, A unified model for the deployment of carbon capture and storage, Applied Energy, 121, 140-148. 1620 Tan R.R., Aviso K.B., Bandyopadhyay S., Ng D.K.S., 2012, Continuous-Time Optimization Model for Source–Sink Matching in Carbon Capture and Storage Systems, Industrial and Engineering Chemistry Research, 51, 10015-10020. Tan R.R., Aviso K.B., Bandyopadhyay S., Ng D.K.S., 2013, Optimal source-sink matching in carbon capture and storage systems with time, injection rate and capacity constraints, Environ. Prog. Sust. Energ., 32, 411-416. Tapia J.F.D., Tan R.R., 2014, Fuzzy optimization of multi-period carbon capture and storage systems with parametric uncertainties, Process Saf. Environ., 92, 545-554. Tapia J.F.D., Tan R.R., 2015, Optimal revamp of multi-region carbon capture and storage (CCS) systems by two-step linear optimization, Energ. Syst., 6(2), 269-289. Riff M.C., Bonnaire X., Neveu B., 2009, A revision of recent approaches for two-dimensional strip-packing problems, Eng. Appl. Artif. Intel., 22(4-5), 833-837. Roussanaly S., Grimstad A.-A. 2014, The economic value of CO2 for EOR applications, Energy Procedia, 63, 7836–7843. http://www.scopus.com/record/display.url?eid=2-s2.0-67349157662&origin=resultslist&sort=plf-f&src=s&st1=A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems&sid=0589AF557E528218BE87E93BA75B42E3.WeLimyRvBMk2ky9SFKc8Q%3a20&sot=b&sdt=b&sl=89&s=TITLE-ABS-KEY%28A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems%29&relpos=0&relpos=0&citeCnt=13&searchTerm=TITLE-ABS-KEY%28A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems%29 http://www.scopus.com/record/display.url?eid=2-s2.0-67349157662&origin=resultslist&sort=plf-f&src=s&st1=A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems&sid=0589AF557E528218BE87E93BA75B42E3.WeLimyRvBMk2ky9SFKc8Q%3a20&sot=b&sdt=b&sl=89&s=TITLE-ABS-KEY%28A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems%29&relpos=0&relpos=0&citeCnt=13&searchTerm=TITLE-ABS-KEY%28A+revision+of+recent+approaches+for+two-dimensional+strip-packing+problems%29