Available online at ScienceDirect. Energy Procedia 63 (2014 ) GHGT-12. M. Hossein Sahraei, L.A. Ricardez-Sandoval*

Transcription

1 Available online at ScienceDirect Energy Procedia 63 (2014 ) GHGT-12 Simultaneous design and control of the MEA absorption process of a CO 2 capture plant M. Hossein Sahraei, L.A. Ricardez-Sandoval* Department of Chemical Engineering, University of Waterloo, Waterloo, ON, Canada N2L 3G1 Abstract This paper presents the simultaneous design and control of the MEA absorption section of a post-combustion CO 2 capture plant. An optimization framework is proposed to minimize the process economics while complying with the process constraints in the transient domain using a decentralized control scheme. The optimal design obtained by the proposed methodology has been validated and compared against the traditional optimal steady-state design approach. The results indicate that the proposed design is dynamically feasible in the presence of oscillatory disturbances in the flue gas flowrate The The Authors. Authors. Published Published by Elsevier by Elsevier Ltd. This Ltd. is an open access article under the CC BY-NC-ND license ( Selection and peer-review under responsibility of GHGT. Peer-review under responsibility of the Organizing Committee of GHGT-12 Keywords: Design and control, MEA absorption process and CO 2 capture 1. Introduction Sustainability of coal-based power plants under environmental standards requires the reduction of carbon dioxide (CO 2 ) emissions through the implementation of technologies typically known as CO 2 capture and sequestration (CCS). Many current R&D activities have been focused on developing and improving the CO 2 capture process for fossil-fired power plants. The most feasible and commercial CCS strategy is post-combustion capture of CO 2 using Mono-Ethanol-Amine (MEA) absorption. In this process, the CO 2 contained in the flue gas stream is absorbed by the amine solution whereas the treated (vented) gas leaves at the top of the absorption tower. The benefits of this process are high capacity for CO 2 capture, fast reaction kinetics and inexpensive and abundance of the amino solvent, i.e., MEA. On the other hand, the main challenge facing this process is the potential substantial drop in the power plant s * Corresponding author: Tel: (+) x38667, Fax, (+) address: laricard@uwaterloo.ca The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ( Peer-review under responsibility of the Organizing Committee of GHGT-12 doi: /j.egypro

2 1602 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 ( 2014 ) efficiency due to the large amounts of energy required to regenerate the solvent that would otherwise be used to supply electricity to the grid. This aspect has limited the applicability of this technology to build and operate commercial-scale CO 2 capture plants for fossil-based power plants. Nomenclature CAP capital costs CC CO 2 removal CCS CO 2 capture and sequestration CO 2 carbon dioxide d disturbance DV dynamic variability F ss flue gas nominal flow rate (mol/s) CO 2 molar flowrate in the vented gas stream (mol/s) CO 2 molar flowrate in the flue gas stream (mol/s) MEA mono-ethanol-amine OP operating costs t time (s) x decision variables design variables control variables objective function The integration of a CO 2 capture unit to power plants has increased the complexity of designing a feasible CO 2 capture unit that satis es the CO 2 capture design goals at steady state and during the transient operation of the process in the presence of external disturbances mostly coming from the power plant, e.g. changes in the flue gas flowrate due to changes in the power plant s electricity demands. CO 2 capture plants have been traditionally designed based on process synthesis heuristics and steady-state calculations [1-3]. This design approach takes process controllability into account only after the process design variables have been specified or fixed from the steady-state calculations, e.g., the dimensions of an absorption column. This sequential design approach is often inadequate because the optimal steady-state design is inoperable at transient conditions since it cannot satisfy process specifications, i.e., the process design imposes a limitation on the process dynamics, and therefore the controllability, dynamic feasibility and flexibility of the process. An alternative approach that has been proposed for optimal design is to perform simultaneous design and control. In this approach, the process optimal design is obtained by simultaneously considering steady-state decisions and process dynamics in the analysis. Although the idea is attractive and has been widely accepted by the academia and the industry, the simultaneous design and control of a dynamic system is a challenging task since it involves trade-offs between the optimal steady-state operation of the plant and the dynamic operability of the plant under the effect of disturbances and process uncertainty. To date, a unified framework that addresses the simultaneous design and control under uncertainty is not currently available. Instead, several methodologies for integration of design and control have been proposed in the literature [4-13]. Recent comprehensive reviews on the current techniques and methods on integration of design and control are available [14], [15]. These methodologies have been applied to different industrial plants such as waste water treatment plant [16], Tennessee Eastman plant [17], and distillation processes [8], [18]. Despite these efforts, a study that takes into account the process dynamics while performing the optimal design of a CO 2 capture plant has not been reported. Thus, the aim of this work is to present the simultaneous design and control of the MEA absorption section of a post-combustion CO 2 capture plant in the presence of transient changes in the disturbances, i.e., the flue gas flow rate. The structure of this paper is as follows: section 2 presents the CO 2 capture plant model and the controller structure considered in this study. The proposed optimization framework for

3 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 (2014)

4 1604 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 ( 2014 ) Component Object Model (COM) technology. This capability allows implementing control and optimization techniques to the actual non-linear process model thereby avoiding any process model approximations. 3. Simultaneous design and control formulation In this section, the mathematical formulation of the simultaneous design and control methodology is presented. The optimization framework used in this methodology explicitly accounts for the dynamic variability of process and its effect on the process economics. The objective function is de ned as the addition of the annualized capital costs (CAP), the operating costs (OP) and the dynamic variability costs (DV). The capital costs refer to the xed annualized expenses of purchasing and installing equipment and units in the absorption process. Estimates for the equipment s cost and utility cost have been obtained from empirical correlations available in [25] and [26], respectively. The annual operating costs (OP) refer to the cost of the utilities used in the daily operation of the plant, which mainly involves the costs incurred for solvent regeneration. In order to consider the steam costs (steam is needed to regenerate the solvent in the stripping section of a post-combustion CO 2 capture plant), a sensitivity analysis has been performed for the complete CO 2 capture process to determine a non-monotonic correlation factor between the solvent s flowrate and the amount of steam required for the regeneration process; this factor states that larger solvent flowrates are needed when the plant s steam consumption increases. Both the capital (CAP) and operating (OP) costs are estimated using steady-state conditions. On the other hand, the dynamic variability costs (DV) aim to measure the process variability in economic terms due to sudden or sustained uctuations in the process. In this study, the tax of carbon ($30 per tonne of CO 2 captured [27]), has been considered in the plant s DV costs to capture the variability in the CO 2 removal due to transient changes in the process operating conditions. The disturbance d(t) affecting the process was assumed to follow a specific time-dependent trajectory, i.e., a sinusoidal disturbance signal around a nominal flue gas flowrate ( =4.3 mol/s) and with an amplitude that is 20% above and below ; the disturbance s frequency was set to 2h. The present analysis assumed that the amount of CO 2 capture removal (%CC) is to be maintained at 90%; a constraint on the minimum allowed CO 2 removal (85%) in presence of disturbances is also enforced in the present optimization problem. This constraint ensures that the CO 2 removal needs to be at least at 85% under transient changes in the system. As shown in Equation (1), %CC is de ned as the amount of CO 2 captured at any time t per the total amount of CO 2 entering the plant in the flue-gas flowrate stream. where and represent the CO 2 molar flowrate in the vented and the flue gas streams, respectively. Based on the above developments, the simultaneous design and control methodology presented in this work can be mathematically formulated as follows: s.t. MEA absorption column s model Multi-loop PI control scheme (2) where the decision variables (x) for this problem includes process design variables ( ) and the PI controllers tuning parameters ( ), i.e., the controllers gain and time integrals. As the absorption column s diameter increases, the average velocity decreases which reduces chemical absorption of CO 2 into the lean solvent; however; as the column s height increases, the contact time of the liquid and gas phases increases which results in higher CO 2 removal. Based on such design complexity and controllability of the system, a sensitivity analysis was performed to define the feasible limit for the decision variables considered for simultaneous design and control. In the optimization framework, the absorption column s diameter and length, the set point for the liquid level of the sump

5 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 (2014)

6 1606 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 ( 2014 )

7 M. Hossein Sahraei and L.A. Ricardez-Sandoval / Energy Procedia 63 ( 2014 ) [9] Mohideen MJ, Perkins JD, Pistikopoulos EN. Optimal design of dynamic systems under uncertainty. AIChE J. 1996b;42: [10] Kookos IK, Perkins JD. An algorithm for simultaneous process design and control. Ind Eng Chem Res 2001;40: [11] Bahri PA, Bandoni JA, Romagnoli JA. Integrated flexibility and controllability analysis in design of chemical processes. AICHE J 1997;43: [12] Hamid MKA, Sin G, Gani R. Integration of process design and controller design for chemical processes using model-based methodology. Comp Chem Eng 2010;34: [13] Ricardez Sandoval LA, Budman HM, Douglas PL. Simultaneous design and control of processes under uncertainty: A robust modelling approach J Process Control 2008;18: [14] Ricardez-Sandoval LA, Budman HM, Douglas PL. Integration of design and control for chemical processes: A review of the literature. Annual Rev Control 2009;33: [15] Sharifzadeh M. Integration of process design and control: a review. Chem Eng Res Design 2013;91: [16] Bahakim SS, Ricardez-Sandoval LA. Simultaneous design and MPC-based control for dynamic system sunder uncertainty: a stochastic approach. Comp Chem Eng 2014;63: [17] Ricardez-Sandoval LA, Budman HM, Douglas PL. Simultaneous design and control of chemical processes with application to the Tennessee Eastman process. J Process Control 2009;90: [18] Sakizlis V, Perkins JD, Pistikopoulos EN. Recent advances in optimization-based simultaneous process and control design. Comp Chem Eng 2004;28: [19] Dugas ER. Pilot plant study of carbon dioxide capture by aqueous monoethanolamine. MSE Thesis, The University of Texas at Austin, [20] Aroonwilas A, Tontiwachwuthikul P, Chakma A. Effects of operating and design parameters on CO 2 absorption in columns with structured packings. Sep Purif Technol 2001;24(3): [21] Versteeg GF, Van Dijck LAJ., Van Swaaij WPM. On the kinetics between CO 2 and alkanolamines both in aqueous and non-aqueous solutions, an overview. Chem Eng Commun 1996;144: [23] Nittaya T, Douglas PL, Croiset E, Ricardez-Sandoval LA. Dynamic modeling and evaluation of an industrial-scale CO 2 capture plant using monoethanolamine absorption processes. Ind Eng Chem Res 2014;53(28): [24] Nittaya T, Douglas PL, Croiset E, Ricardez-Sandoval LA. Dynamic modelling and control of MEA absorption processes for CO 2 capture from power plants. Fuel 2014;116: [25] Guthrie KM. Process plant estimating evaluation and control. Craftsman Book Co [26] Nuchitprasittichai A, Cremaschi S. Sensitivity of amine-based CO 2 capture cost: The influences of CO 2 concentration in flue gas and utility cost fluctuations. Int J Greenhouse Gas Control 2013;13: [27] Lee M. Building a fair and effective carbon tax to meet BC's greenhouse gas targets. Canadian centre for policy alternatives, 2012.