A techno-economic analysis of post-combustion CO2 capture and compression applied to a combined cycle gas turbine: Part II. Identifying the cost-optimal control and design variables

International Journal of Greenhouse Gas Control 52 (2016) 331–343

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control journal homepage: www.elsevier.com/locate/ijggc

A techno-economic analysis of post-combustion CO2 capture and compression applied to a combined cycle gas turbine: Part II. Identifying the cost-optimal control and design variables Ahmed Alhajaj a,c,∗ , Niall Mac Dowell b,c , Nilay Shah c a The Institute Centre for Energy, Department of Chemical and Environmental Engineering, Masdar Institute of Science and Technology, P.O. Box 54224, Abu Dhabi, United Arab Emirates b Centre for Environmental Policy, Imperial College London, South Kensington, London SW7 1NA, UK c Centre for Process Systems Engineering, Department of Chemical Engineering, Imperial College London, SW7 2AZ, UK

a r t i c l e

i n f o

Article history: Received 9 July 2015 Received in revised form 2 July 2016 Accepted 4 July 2016 Available online 25 July 2016 Keywords: CO2 capture Absorption model Optimization Cost Economics gPROMS

a b s t r a c t A detailed optimization-orientated model of monoethanolamine-based CO2 capture plant and compression train in which all the technical and economic assumptions are defined and/or optimized was developed and used to simultaneously determine the cost optimal control and design variables including feed fraction ratio at different degrees of capture (DOC), which represents the amount of CO2 removed, for plant designs that partially bypass the CO2 capture process so as to achieve low to moderate reductions of CO2 , but at lower overall cost. The effects of varying carbon prices on the levelized cost of CO2 capture and compression were also studied. The capture bypass option was observed to be the cost optimal choice for lower than 60% overall DOC. Carbon prices were observed to have a clear impact on the cost optimal DOC, with the cost-optimal DOC shifting from 70%–80% to 85%–90% at carbon prices of $4/tCO2 to $23/tCO2 respectively. The study highlighted that if a suitably high carbon price does not materialize through a market mechanism, appropriate policies need to be put in place to achieve decarbonisation targets. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction CO2 capture, transport and storage (CCS) is a promising near term option to mitigate CO2 emissions associated with fossil fuel combustion and usage. The main challenge associated with the large-scale deployment of CCS, however, is its cost, with CO2 capture plant and compression train accounting for the majority of the whole CCS chain cost (i.e., 80% for capture and compression, 10% for transportation and 10% for storage) (Metz et al., 2005). Thus, there exists a well-recognized imperative to minimize this cost. In the literature, capture cost estimates vary widely because of the different technical assumptions regarding power plant size, load factor, fuel properties and net efficiency in addition to the diverse economic and financial assumptions such as, discount rate, plant life, and fuel cost (Adams and Mac Dowell, 2016; EBTF, 2011; Mac Dowell and Shah, 2013; Metz et al., 2005; Rubin et al., 2007a,b). These cost estimates can be grouped into those developed by governmental bodies (DECC, 2011; Finkenrath, 2011; GCCSI, 2011;

∗ Corresponding author at: The Institute Centre for Energy, Department of Chemical and Environmental Engineering, Masdar Institute of Science and Technology, P.O. Box 54224, Abu Dhabi, United Arab Emirates. E-mail address: [email protected] (A. Alhajaj). http://dx.doi.org/10.1016/j.ijggc.2016.07.008 1750-5836/© 2016 Elsevier Ltd. All rights reserved.

Metz et al., 2005; NETL, 2010a; ZEP, 2011), private companies and licensors of commercial solvents (Iijima, 1998; Mariz, 1998; Scottish Power, 2011), and those arising from detailed capture plant models (Abu-Zahra et al., 2007a; EBTF, 2011; Mores et al., 2012; Rao and Rubin, 2002, 2006; Singh et al., 2003). The former estimates are generated by organizations that do not make public the details of their assumptions and models. The latter option, which is based on coupling economic and physical models of the CO2 capture plant, is the focus of this paper. The efforts in this area can be classified as: (1) a step-wise approach where the simulation of the capture plant and the calculation of cost are performed separately and (2) a simultaneous approach in which the economic model is integrated within the physics-based model of the CO2 capture plant. The following aspects are highlighted in the literature: (1) methodology of linking CO2 source plant model to the cost model; (2) assumptions of control or state variables and design variables (e.g., diameter, height of the column); (3) main variables and size factors used to obtain the purchased cost of the CO2 capture plant; (4) approach used to obtain the plant cost; (5) assumptions of the power plant size, life, and capacity factor; (6) inclusion of uncertainty and variability. The earliest step-wise approach was carried out by the Carnegie Mellon University (CMU) group who developed a power plant simulation model (the integrated environmental control model (IECM)

332

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

Nomenclature Variables CAPEX Capital expenditures CCS Carbon capture and storage CER Certified emission reduction Capacity factor CF CP Carbon price CRF Capacity recovery factor Carbon steel CS D Diameter (m) DCC Direct contact cooler DEC Direct equipment cost Degree of capture DOC DOF Degree of freedom EOR Enhanced oil recovery FE Equipment multiplying factor FI Instrument multiplying factor Fixed operation and maintenance FOM FCC Facility capital cost ($) Flue gas feed fraction ratio FFR GIS Geographical information system Horse power HP ISBL Inside battery limit Key performance indicators KPIs KOPs Key operating parameters LCCC Levelized capture and compression cost ($/tCO2 ) LMTD Log mean temperature difference ˙ m Mass flow rate (kg s−1 ) Operating expenditures OPEX P Pressure (Pa) PUI Process unit investment ($) Cooling or heating duty (W) Q S Molar specific entropy (J mol−1 ) Sequential quadratic programming SQP SS Stainless steel Temperature (K) T TC Total capital ($) Ton of CO2 tCO2 U Overall heat transfer coefficient (W m−2 K−1 ) VC Variable cost ($/tCO2 ) ˙ W Power of rotary equipment (W) Working capital ($) WC Greek symbols Lean loading of solvent at the inlet of the absorber  lean   x(CO2 ,in) /x(MEA,in) Subscripts/superscripts cap Captured Cut-off pressure of the multi-stage compression cut cw Cooling water E Equipment f Factors: materials; labor; right of way (ROW); miscellaneous Flue gas fg g Gas phase I Instrument Components: 1 = H2 O; 2 = MEA; 3 = CO2 ; 4 = N2 i in Inlet Lean solvent ls out Outlet Pump p Vented ven ∗ Outlet assuming isentropic path

for the US department of Energy) with an option to attach SO2 and NOx emissions control technologies (Rubin et al., 1997). This model was subsequently extended with a carbon control technology (i.e., MEA, ammonia, membranes and oxyfuel-based decarbonisation). It was represented by a regression model containing the main variables that allow the calculation of the capital cost and operating cost such as MEA make-up, energy requirement, and environmental emissions generated from the MEA degradation and the bottom of the reclaimer (Rao, 2002). This regression model was based on a number of simulations for the CO2 capture plant and a compression train with 4 compression stages developed on ProTreat and Aspen Plus respectively. The following main variables were obtained from the regression model which in turn are functions of amine lean loading, CO2 content, MEA concentration, flue gas temperature, and degree of CO2 capture: (1) liquid to gas flow rate; (2) heat supplied per unit of liquid (kJ/kmol); (3) compression power required per kg of CO2 . The main equipment purchased cost was obtained by applying the sixth-tenth power relationship to the data obtained from Fluor as the reference size (Rao, 2002). The following size factors were used to scale the cost: volume and temperature of flue gas for blowers, absorbers and DCC; volume of lean solvent for stripper, lean amine cooler, storage tank and rich lean cooler; volume of solvent and amount of steam for the reboiler. The total capital costs were obtained using the Energy Power Research Institute (EPRI) costing methodology, which estimate other cost components such as engineering as a fraction of the total equipment cost (EPRI, 1993). Rao and Rubin (2002) initially used the Integrated Environmental Control Model for Carbon Sequestration (IECM-CS) (i.e., the IECM model with carbon capture plant regression model) to explore the effect of the assumptions made for the reference plant (i.e., coal fired power plant) and the interactions with the existing pollution control strategy on the key operating parameters of the carbon capture plant. It was observed that the triangular distribution of amine lean loading had resulted in 95-percentile range of $43–72/tCO2 avoided. Rao and Rubin (2006) examined the effect of degree of capture between 70 and 90% on the energy penalty (i.e., electricity consumption and electricity loss of the power plant for steam derating) and the cost of CO2 avoided. The study was based on attaching a CO2 capture plant to coal fired power plants with sizes of 600 MW and 1000 MW. The flue gas bypass option was also examined in the study to lower the cost of the carbon capture plant. The benefits of using the flue gas bypass option at higher than 60% DOC were anticipated as a result of treating less flue gas and solvent into the system, consistent with the conclusions of our previous work (Mac Dowell and Shah, 2013), however in the presence of a carbon price sufficient to incentivise CO2 capture, this strategy is unlikely to be generally cost-optimal, with varying the degree of solvent regeneration in line with electricity prices expected to be a more profitable option (Mac Dowell and Shah, 2014b, 2015). The main driver to use the flue gas bypass option in Rao and Rubin (2006) work is to minimize the cost resulting from treating less flue gas and capturing the remaining flue gas at 95% DOC. This was driven by a combination of lower CAPEX associated with smaller equipment sizes and with similarly lower OPEX arising from reduced solvent flowrates and thus solvent regeneration costs. The assumption of fixed column height along with the absorber cost calculation method, which is based on the amount of flue gas treated, are the reasons for lower cost predicted in the earlier study. Further, it was found that the cost-optimal degree of capture depends on the size of the plant, and it is less than the assumed value of 90% degree of capture. The main limitations of the study were the assumptions of a fixed 12 m height column and an amine lean loading of 0.2 which might not be the optimal operating and design variables. Further, the assumption of having a maximum capacity for each train might affect the cost of CO2 avoided.

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

In 2007, the IECM-CS model, which was originally integrated with a sub-critical coal fired power plant, was extended to include CCGT and Integrated Gasification Combined Cycle (IGCC) power plants (Rubin et al., 2007a). The effects of varying fuel price and composition on the cost of electricity with and without CO2 capture were investigated for three different power plants; sub-critical coal fired plant, IGCC and CCGT (Rubin et al., 2007a). The results revealed that although the cost of electricity (COE) with and without capture for CCGT is lower at gas prices lower than $4/GJ; the order was different at a gas price of $6/GJ at that time (Rubin et al., 2007a). They also stressed the importance of accounting reduction in cost of technologies occurring over time as a result of learning by doing, innovation and technological development (Rubin et al., 2007b). Application of a higher premium capital charge factor for riskier technologies such as IGCC was suggested (Rubin et al., 2007a). This eliminated the advantages of IGCC, making its cost similar to CCGT and coal-fired power plants with carbon capture. Uncertainty in capital cost and the effect of research and development in lowering important cost elements such as the reboiler duty were examined by Singh et al. (2003). Further, the effect of varying interest rates on the cost of CO2 avoided attached to a coal fired plant were also studied by Abu-Zahra et al. (2007a) and Klemeˇs et al. (2007). Singh et al. (2003), Abu-Zahra et al. (2007a) and European Benchmark Tasking Force (EBTF) (2011) used a step-wise approach in which the output of the simulation of CO2 capture model developed in Aspen plus was used directly to obtain the CO2 capture plant cost. Singh et al. (2003) assumed attaching a carbon capture plant to a 400 MW coal plant while using a supplementary gas turbine to supply the electricity and steam required to run the capture plant. They used the Icarus Process Evaluator within Aspen Plus to obtain the cost of the main equipment of the capture plant model and the compression system (i.e., absorption and regeneration columns, heat exchangers, tanks, pumps and compressor). The balance of the plant costs such as engineering, start-up costs and contractor fees were assumed to be the same as those obtained by Mariz (1998). The indirect cost and contingency were taken as percentages of the equipment cost. The effect of the control variables such as amine lean loading on the cost was not considered in this study. Abu-Zahra et al. (2007a), however, have examined the effect of amine lean loading, MEA weight, and stripper operating pressure and temperature on the cost per ton of CO2 avoided. Their work was based on a detailed model developed in Aspen plus (AbuZahra et al., 2007b). The results of the model were then used to size the columns while the purchased costs of the rest of the equipment were obtained using several references without disclosing details of the sizing factors used. The results of this study showed that the cost of CO2 avoided was constant and minimum when operating at an amine lean loading between 0.25–0.32 and 80–90% degree of capture. Operating the stripper at high pressure also minimized the cost of CO2 avoided. Similarly, EBTF (2011) used the cost of CO2 avoided as a metric to assess the performance of different capture technologies attached to different power plants. The costing methodology is based on a framework similar to Abu-Zahra et al. (2007a) in which sizing factors obtained from a simulation model were used to estimate the purchased equipment cost. This was acquired from vendors while the values used in calculating the balance cost of the plant were obtained from the discussion within EBTF group. The group was successful in developing common parameters and assumptions (e.g., fuel properties, blower efficiency, flue gas composition, discount rate) that form a guideline for benchmarking CO2 capture technologies (EBTF, 2011). There are relatively few contributions in the literature, which simultaneously find the cost optimal design and control variables of an integrated CO2 capture plant and compression train at the design stage. Mores et al. (2012) developed a detailed model of

333

the CO2 capture plant with compression and cost models in GAMS to simultaneously find the optimal design and operating conditions that minimize the total cost of CO2 captured and compressed while meeting different CO2 reduction targets. The purchased equipment cost was obtained using the following size factors: superficial area and packing volume for packing; area for heat exchangers; volumetric capacity for water storage tanks; horsepower for compressors and pumps. These variables were used to scale the purchased cost using the sixth-tenth economies of scale rule while the reference unit sizes were selected randomly and their respective costs were calculated from correlations of Henao (2006) and Seider et al. (2009). The balance cost of the plant was based on the assumptions made in Abu-Zahra et al.’s (2007a) work. The results obtained from attaching a CO2 capture plant to a theoretical plant with CO2 content similar to NGCC plant predicted a relatively large reboiler duty of 5.7 GJ/tCO2 in comparison to the 3.8–4.2 GJ/tCO2 commonly found in the literature in the case of CCS on coal fired power plant. This is in line with previous observations that decarbonised gas-fired power plants are more costly in terms of GJ/tCO2 recovered but they are relatively low cost in terms of $/MWh of low-carbon electricity generated which is the ultimate aim of applying CCS technology to power generation (Rubin et al., 2007a). The following limitations in current CO2 capture cost estimations using detailed models are summarized: assumptions of scaling factor based on the sixth-tenth power relationship used to estimate the purchased equipment cost, which is not applicable to all units at the same rate (e.g., see Ulrich and Vasudevan, 2009); inconsistency in size factors and cost elements used to obtain the equipment cost and the balance cost of the plant respectively; separation of cost model from the capture plant model, which limits the response-capability of the model to different assumptions and does not exploit the possibility of simultaneously varying all the important variables. 1.1. Objective of the paper The objective of this paper is to develop an optimizationorientated mathematical model of the CO2 capture plant and compression train in which a comprehensive costing approach is integrated and applied. This model is then used to simultaneously find the optimal control and design variables (e.g., amine lean loading, reboiler and stripper pressure, absorber height and diameter) that minimize the total levelized cost of CO2 capture and compression at different degrees of capture. This approach ensures that the design is undertaken for the optimum control variable values rather than assuming non-optimal control variable values. The remainder of the paper is presented as follows: it starts by formulating the optimization problem of the CO2 capture plant and compression train, followed by a description of the comprehensive economic model; then the cost optimal control and design variables for CO2 capture plant and compression train with flue gas bypass option while considering different carbon prices are determined and analysed. It concludes with a summary and analysis of the main results. 2. Methodology 2.1. Model development The engineering models of the CO2 capture process and compression train illustrated in Fig. 1 were developed in part I of this series (Alhajaj et al., 2016). The packed sections of the absorber and desorber are described using an equilibrium-stage model. This was selected in the interest of flexibility and speed of conver-

334

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

Fig. 1. Carbon capture plant with multi-stage compression process flow sheet modelled in gPROMS.

gence required for the simultaneous optimization of control and design variables. All thermophysical properties were calculated using the statistical associating fluid theory SAFT (Chapman et al., 1989, 1990; Rubin et al., 2007a) for potentials of variable range: SAFT-VR (Galindo et al., 1998; Gil-Villegas et al., 1997) in which the chemical reaction between the amine and the CO2 is explicitly described in the thermodynamic model (Llovell et al., 2012; Mac Dowell et al., 2010, 2011). The engineering model presented in part I of this series (Alhajaj et al., 2016) was extended with a detailed economic model in order to develop a tool capable of determining the cost optimal control and design variables in addition to performing further analysis to study the effects of carbon price and the usage of flue gas bypass option on the process costs.

2.1.1. Cost model The CO2 capture and compression cost model was incorporated into the model developed in part I of this series (Alhajaj et al., 2016). There are many approaches available to calculate the total cost of the plant, which are mainly based on obtaining the purchased equipment cost and then utilizing a factorisation approach to obtain the total cost of the plant. Purchased equipment cost was usually obtained using scaling approaches utilizing power relationships based on cost plots (Green and Perry, 2008; Peters et al., 2004; Ulrich and Vasudevan, 2004) in addition to using the Douglas (1988) approach, which is based on Guthrie’s (1969) simplified cost correlations. Douglas (1988) and Peters et al. (2004) factorisation approaches were mainly used in the literature to obtain the balance cost of the plant. The main challenge in applying these approaches is the difficulty in choosing fixed values for various factors contributing to the total cost of the plant. In this study, a standardised Chauvel et al. (1981) approach was used to calculate the total cost of the plant. This method is simple with fewer elements contributing to the total capital cost; thus eliminating errors accumulating with estimations of wide range of elements used in different methods (e.g., process pipelines, land, yard improvements, electrical). Other standardized approaches such as EPRI and IEA can also be used but it requires contacting vendors or contractors for cost estimations (Rubin et al., 2013). This cannot be applied to optimization models that demand continuous mathematical functions describing unit cost. The purchased equipment cost was obtained from the updated cost correlations outlined in Couper’s (2010) work. Further, a decoupling between instrument cost and equipment cost based on Cran’s (1981) approach was applied in this study to obtain

Table 1 Elements to calculate the total capital cost (Chauvel et al., 1981). Code

Capital cost element

Value

A B

Purchased equipment cost (Couper, 2010) Instrument cost (Cran, 1981)

Ei ∀i = 1· · ·n Ii ∀i = 1· · ·n

C

Direct equipment cost (DEC)

n 

i

D E F G H I J K L M N O P

Indirect equipment cost Inside Battery Limit Investment (ISBL) Off sites (OS) Process unit investment (PUI) Engineering Paid up royalties Process data book Facility capital cost (FCC) Initial charge of feed stocks Interest during construction Start up cost Total capital (TC) Working capital (WC)

Ei FE +

n 

Ii FI i

31% DEC C +D 31% DEC ISBL + OS 12% PUI 7% ISBL 265,000 US$ in 2004a PUI + H + I + J Amine feedstockb × Cost 7% FCC 1 month of operating cost FCC + L + M + N 1 month of operating cost

a

Average value scaled using Marshall and swift index. This is the liquid circulation rate in addition to amine hold up calculated from (Billet and Schultes, 1993). b

the direct equipment cost due to the fact that instrumentation cost depends mainly on the type of equipment used. There are variety of approaches used in the literature to report the cost of CO2 capture and compression (Metz et al., 2005; Rubin, 2012). The cost of CO2 avoided was used widely in the literature to compare the performance of different CO2 capture plant technologies coupled with power plants. However, a levelized cost of CO2 capture and compression was used in this work in order to avoid the sensitivity of assumptions regarding plant size, life and capacity and in order to apply it to different CO2 sources and CCS projects. Further, this way of reporting is also flexible to study the effect of carbon price for each ton of CO2 vented in addition to being expandable to account for the opportunity loss of utilizing steam and electricity instead of selling the electricity at the prevailing market price. The total levelized cost of CO2 capture and compression is based on calculating CAPEX and OPEX. 2.1.1.1. CO2 capture and compression CAPEX. The total capital cost (TC) of the capture plant and compression train was calculated using the Chauvel et al. (1981) approach as shown in Table 1. The first step was to calculate the purchased equipment cost (Couper,

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

335

Table 2 Elements to calculate the direct equipment cost (Couper, 2010; Cran, 1981). Equipment

Size factor

Direct cost multiplying factor

Instrument cost $ in 2004

Material of construction

Blower Columns Columns packing Compressor Condenser Instrument Reboiler and coolers Rich/lean HE Pump

HP Diameter, height and weight Volume HP Area

1.4 2.1

2500$ per stage 44,250$

SS

1.3 2.2 2.5 2.2 1.9 2.0

2500$ per stage 10,500$

CS/SS

9750$ 9750$ 2500$

CS/SS SS/SS SS

Area Area Head, HP and volumetric flow rate

2010), which depends on the grade of material, the operating conditions and the size factors representing key characteristics of the equipment (see Table 2). These size factors were obtained from simulations of the carbon capture and compression train model developed in this work. Further size factors such as weight of columns and heat exchanger areas were developed in the model. The second step was to obtain the direct equipment cost representing the manufacture and installation of primary equipment in addition to any secondary equipment in the site using equipment multiplying-factor listed in Table 2. Then, the cost of instrumentation was added separately using values and instrument multiplying factor listed in Table 2 because it does not vary proportionally with the size of the equipment. After which, indirect costs that cover transportation cost, site preparation for special equipment such as cranes, temporary buildings, and contingency to account for surprises during construction are added. Then, the offsite cost (i.e., storage and utilities) was calculated. The rest of the cost elements were obtained using the relationships that depend on the investment in units (see Table 1). 2.1.1.1.1. Weight of columns. The cost of the columns depends on the weight of the columns and the type of material. For our case, stainless steel was used for all the columns due to the high corrosion level normally associated with aqueous solutions of alkanolamines: a rate of 0.286 mm per year in the absorber and 4.5–8.5 mm per year in the stripper were reported while using carbon steel (Kittel et al., 2009); this is higher than the accepted level of 0.1 mm (Kittel et al., 2009). It is acknowledged that cladding stainless steel to carbon steel or concrete (e.g., Boundary Dam capture plant in Canada) can reduce the cost of columns, however, evaluating these particular details are beyond the scope of our current study, which primarily intends to present a methodological approach. The weight of the columns depends on the diameter, the height and the thickness. This in turn depends on the internal pressure, radius, material strength properties and wind speed. The thickness expressions obtained from Megyesey and Buthod (1986) and Richardson et al. (1999) in addition to diameter and packing height calculation obtained in part I of this series (Alhajaj et al., 2016) were incorporated in the cost model of the capture plant in gPROMSTM . The total heights of the absorber and stripper were obtained by applying a multiplying factor (i.e., 1.25) to the calculated packing height (Oexmann et al., 2008). This takes into account extra spacing needed for additional equipment in the columns such as distributors. 2.1.1.1.2. Heat exchanger area. The required area for transferring the required heat duties (Q) in the coolers, the rich lean heat exchanger, the condenser and the reboiler were obtained from Equation (1).

Area =

Q

⎛ U∗⎝



⎞ 

(Thot in −Tcold out )−(Thot out −Tcold in ) Ln

(Thot (Thot

in −Tcold out out −Tcold in

) )

(1)

Table 3 Overall heat transfer coefficients and type for the heat exchangers used within the capture plant and compression train (GPSA, 1987).

Condenser and after-coolers Lean amine cooler Reboiler Rich lean HE Water cooler

U (W m−2 K−1 ) (GPSA, 1987)

Type of HE

425 795 850 710 1070

Shell and tube Shell and tube Kettle Shell and tube Shell and tube

The overall heat transfer coefficient (U) depends on the type of medium and phases present in the heat exchanger and other dimensionless factors. A summary of the main overall heat transfer coefficients and types of heat exchanger used to calculate the capital cost in this study are outlined in Table 3. It is worth nothing that Equation (1) is accurate to within ±10%1 for estimation of reboiler and condenser areas. This discrepancy arises as the temperature profile is not linear along the path of vaporization and condensation. In fact, this simplified method is safe for systems that do not have serious fronts (i.e., the existence of lower boiling components that reduce the bubble point) or tails (i.e., the existence of higher boiler components that increase the dew point) (Kern, 1965). It is acknowledged that utilizing weighted LMTD, which divide the condensation and vaporization regions into zones (e.g., sensible and phase change region), increases the model prediction accuracy but it will adds complexity to the optimization of the current model. This study was based on initial size estimation assuming a linear temperature profile for condensation and vaporization. A more accurate area prediction will be obtained in the detailed heat exchanger design, which is out of scope of the current work. 2.1.1.2. CO2 capture and compression OPEX. Table 4 lists the main elements of the operating cost. The variable cost comprises utilities consumption and amine makeup cost. The fixed cost consists of maintenance, insurance, labour cost and overheads that cover the non-productive elements of the plant such as administration, workshops and office management (Chauvel et al., 1981). 2.1.1.3. Levelized capture and compression cost. The levelized capture and compression cost (LCCC) is calculated using Equation (2). LCCC =

(TC) CRF + FOM cap

˙ CO CF m

2



+ VC +

ven

˙ CO2 CPm cap

˙ CO m



(2)

2

The capital recovery factor (CRF) takes into account the depreciation of the plant and interest rates through its lifetime. A CRF of 0.15 was assumed for both the capture and compression train facilities (Rao and Rubin, 2002). FOM is the fixed operation and

⎠ 1 This error figure was obtained from discrepancies in areas obtained using LMTD and weighted LMTD considering number of cases in this study.

336

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

Table 4 Elements to calculate fixed and variable operating and maintenance cost (Chauvel et al., 1981). Code

Capital cost element

Value

OA OB OC OD OE OF OG OH OI OJ OK OL

Electricity Steam Cooling water Utilities (U) MEA make up Variable cost (VC) Labour Maintenance Taxes and Insurance Overhead Financing working capital Fixed operating & maintenance (FOM)

$0.04/kWh $1.4/GJa $0.02/m3 U = OA + OB + OC $1.2/kg VC = U + OE One operation engineer and four shift crews consisting of one foremen and two operatorsb ( $398,000 ) year 4%PUI 2%PUI 1%PUI 9%WC FOM = OG + OH + OI + OJ + OK

a A fixed steam price was assumed in this study. If steam is extracted from the power plant, the price will depend on the pressure, and the opportunity loss of transferring this steam into electricity at specific market price. b Iijima (1998).

maintenance cost including labor in $ year−1 , CF is the capaccap ˙ CO is the amount of CO2 captured and compressed ity factor, m 2

(tCO2 /year), VC is the variable cost (e.g., utilities) ($/tCO2 ), CP is the ˙ ven carbon price ($/tCO2 emitted) and m CO2 is the amount of CO2 vented (tCO2 /year) excluding any extra CO2 vented from the capture plant and compression train. 2.2. Optimization problem The optimization problem is described as follows: GIVEN • • • •

The flue gas flow rate, temperature and composition. The MEA concentration (e.g., 30 wt% MEA). DOC (50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 97%).2 Carbon price (CP). DECISION VARIABLES

• Determining the optimal control variables for the capture plant and compression train.

train. This can be formulated in a general form as shown in Equation (3) being subject to equality and inequality constraints shown in Equations (4) and (5) respectively. Min F (x)

(3)

Subject to : H (x) = 0

(4)

G (x) ≥ 0

(5)

x is the vector that contains the above mentioned control and design variables in addition to operating variables such as utilities consumption and amine make-up. Other economic variables that the vector contains are the economic factors such as finance and labour cost. F (x) is the levelized carbon capture and compression cost (LCCC) obtained in Equation (2). H (x) is the set of equality constraints represented by the modelling equations in the earlier work such as mass and energy balance in addition to the economic equations presented here. G (x) is the set of inequality constraints represented for the decision variables of the optimization problem as outlined in Equations (6)–(14). 0.2 < lean < 0.38 ◦

40 C < T flue

gas

◦

Amine lean loading. Flue gas temperature after cooling. Lean solvent temperature inlet to absorber. Reboiler and stripper pressure. Temperature difference in DCC and scrubber cooler in addition to after-coolers. – Temperature difference in rich and lean heat exchangers.

– – – – –

• Determining the optimal value of key process design variables. – Area of heat exchangers (i.e., coolers, condensers, after-coolers and rich lean heat exchangers). – Rate power (i.e., compressor, blower, and pumps). – Volume and weight of packing columns (i.e., absorber, stripper, DCC, and scrubber). – Flue gas utilization factor (i.e., considering bypassing portion of the flue gas). 2.2.1. Objective function The objective function is to minimize the total levelized cost (i.e., CAPEX and OPEX) for the whole capture plant and compression

40 C < T lean

< 50 C

solvent

◦

45 C < T gout,after 1 < TDCC

cooler

1 < TScubber 1 < TRich

(6) ◦

< 45 C

cooler

(8) ◦

< 50 C

< 15

cooler

(7) ◦

< 15

(9) (10) (11)

< 20

(12)

Min Dcolumn < Dcolumn < 15

(13)

0.2 < FFR < 1

(14)

lean HE

The amine lean loading constraint (i.e., Equation (6)) covers the operating range considering technical performance using different KPIs (Alhajaj et al., 2016). The maximum operating temperature of the flue gas and the amine lean solvent loading were set to minimize the vaporization of the amine observed in the parametric study of part I (Alhajaj et al., 2016). The temperature differences for the heat exchangers are set within the feasible operating range. The minimum diameter of the column was obtained from an empirical correlation, which was based on capturing 90% of the flue gas (Chapel et al., 1999). 3. Case study

2 A lower than 70% DOC was considered in this study to examine the capability of the model in predicting the optimal flue gas utilization factor.

We perform a case study using an exhaust gas typical of a 400 MW CCGT power plan as an input to our model of the CO2

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343 Table 5 Case study of 400 MW CCGT input values in the optimization of carbon capture plant and compression train. CCGT flue gas flow rate (N m3 /h)

1,800,000a (357.1 kg/s)

◦

CCGT flue gas temperature ( C) CCGT flue gas pressure (kPa) CCGT flue gas molar H2 O composition (mol%) CCGT flue gas molar CO2 composition (mol%) CCGT flue gas molar N2 composition (mol%) CCGT flue gas molar O2 composition (mol%) Pressure of the absorber (kPa) Pressure drops in packing (kPa/m) Water makeup (kg/tCO2 ) Capacity factor (CF) Cost year MEA make-up (kg/tCO2 ) Norton IMTP 50 mm dry packing area (a) (m2 /m3 ) CO2 outlet pressure from compressor (kPa) CO2 product content from condenser (mol%) Compressor efficiency (isn , mec ) (%) Pump total efficiency (%) Reboiler steam inlet temperature (◦ C)

98 101 12 5 73 10b 101 0.2 0 0.7 2004c 1.1 kgd 120 14,000e 90.4f (80, 95)g 75e 140h

337

prices, which is reflective of current market prices (i.e., $0/tCO2 for no market price, $4/tCO2 for certified emission reduction (CER) and $23/tCO2 for Australia levy), on the levelized cost of CO2 captured and compressed while having a flue gas bypass option. 4. Results and discussion 4.1. Base-case study—no bypass of capture

a

Bailey and Feron (2005). From the prospective of thermodynamic modelling approach employed in this study, O2 was considered to behave like N2 in terms of VLE. As the main effect of O2 within CO2 capture application is the degradation of solvent, which is a phenomenon observed only via detailed dynamic modelling, it is mediated via solvent make up cost in this study. c The cost year was not escalated to year 2015 as the objective is to analyse system trade-offs. d Chapel et al. (1999) predicted total of 1.6 kg/tCO2 of which 0.5 kg/tCO2 is vaporized at the optimum capture rate as predicted by the model. This increased model flexibility to account for vaporization. e Rao and Rubin (2002). f Iijima (1998). g Kvamsdal et al. (2007). h Jordal et al. (2012). b

capture and compression process. The key input parameters are presented in Table 5 (Alhajaj et al., 2016). The optimization model formulated above was implemented in gPROMSTM , and solved to find the cost optimal control and design variables for different degrees of CO2 capture. The SRQPD solver in gPROMS that employs a sequential quadratic programming (SQP) solution of the nonlinear programming (NLP) problem was used to find the optimum control or state variables. For each degree of capture, 20 optimization runs were done in gPROMS using a random perturbation of the original initial guess; this multi-start method helps partially to circumvent the non-convexity of the optimisation problem, and approach a global solution to the optimisation problem. These optimum control variables are then used as input for the simulation of the capture plant and compression train in which the remaining design variables are obtained. In order to understand the tradeoffs present in the CO2 capture-compression system, the technical performance of this system is evaluated using the KPIs outlined in part I of this series (Alhajaj et al., 2016), which help to analyse system behaviour at the optimum control variables. Furthermore, the impact of optimal control and design variables on cost components and contribution of each individual sub-unit vs. DOC is investigated. This paper will examine two scenarios: The first is based on integrating a CO2 capture plant with a NGCC power plant with the assumption that there is no flue gas bypass option and no carbon price; the second case will explore the effects of carbon

4.1.1. Optimal control and design variables for varying degree of capture The optimization problem in gPROMSTM was solved to determine the optimal control and design variables that minimized the total levelized cost of the capture plant and compression train attached to CCGT power plant with no flue gas bypass option and no carbon price. The solution obtained from the model is the local optimum because it is based on a local rather than global non-linear programming solver. In order to find the optimal variables that can be closer to the global optimum, different initialisation values of the control variables were used. It was observed that the reported control and design variables listed in Table 6 are the same with varying degree of capture. Thus, a constrained optimization was applied to find the optimal amine lean loading and design variables at different degrees of capture, which is listed in Table 7 alongside the Key Performance Indicators (KPIs) (i.e., reboiler and cooling duty, amine slippage, volume of packing, solvent flow rate and ancillary power consumption) listed in Table 8. For each DOC, the values outlined in each line in Tables 7 and 8 correspond to an entirely new plant design. Table 6 highlights the fact that the diameter of the absorber and the stripper columns did not change with varying DOC because the same amount of flue gas was treated without considering the flue gas bypass option. The optimal flue gas and lean solvent temperature are observed to be set at the highest possible values based on the constraints. This is due to the increase of reaction rate constants and diffusivity, which decreased the absorber packing volume and hence cost (Alhajaj et al., 2016). The temperature-related compromise in mass transfer rates, reaction rates and thermo-physical solvent capacity is extensively discussed in our previous work, and is not repeated here (Mac Dowell and Shah, 2014a). The optimal temperature differences for the heat exchangers are chosen to be 14.4 ◦ C for DCC cooler, 20 ◦ C for rich-lean heat exchanger and 15 ◦ C for scrubber cooler. This results in smaller heat exchange areas and hence CAPEX, which outweighs the increased OPEX. It is found that the optimal operating stripper and reboiler pressure (presented in Table 6) is the highest available owing to the reduction in reboiler duty, height of the stripper column, cooling duty and compression power as outlined in part I of this series (Alhajaj et al., 2016). The reboiler duty and height of the stripper column decreased as a result of the consequent increase in operating temperature, which increased the chemical potential gradient in the stripper column resulting in less steam being generated to maintain the driving force. This outweighed the increase of energy required to vaporize the solvent at high pressure. The cooling duty decreased as a result of reduction in condenser cooling duty associated with less steam being generated in the reboiler. Table 7 illustrates that the optimal amine lean loading, which results from the trade-off between CAPEX (e.g., volume of pack-

Table 6 Optimum constrained control and design variables for the CO2 capture plant and compression train. Absorber, Scrubber and Stripper DCC diameter (m) diameter (m)

Tls inlet to absorber column and Tg exiting after-cooler (◦ C)

Flue gas temperature after cooling (◦ C)

Reboiler and T DCC cooler T Rich lean HE T Scrubber cooler stripper pressure (Mpa)

14.5

45.0

50.0

0.202

8

14.4

20.0

15.0

338

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

Table 7 Optimum amine lean loading and design variables for the CO2 capture plant and compression train. Degree of capture (%)

Absorber Height (m)

Amine lean loading

Blower power (kW)

Compressor Compressor Condenser power Area (m2 ) after(kW) cooler area (m2 )

DCC Height (m)

Lean amine cooler area (m2 )

Reboiler Area

Rich lean HE area (m2 )

Scrubber Height (m)

Stripper Height (m)

50 55 60 65 70 75 80 85 90 95 97

13.1 14.6 16.4 18.3 20.5 23.0 25.9 29.5 34.3 41.4 42.8

0.283 0.283 0.283 0.283 0.283 0.283 0.282 0.282 0.282 0.285 0.290

1770 1984 2217 2474 2764 3095 3483 3962 4590 5530 5712

892 981 1070 1159 1249 1338 1427 1516 1605 1695 1730

8 8 8 8 8 8 8 8 8 8 8

2838 3122 3406 3690 3973 4257 4526 4808 5091 5453 5897

7278 8005 8738 9473 10193 10927 11735 12469 13202 13766 13879

7972 8769 9566 10362 11159 11955 12717 13512 14307 15280 15837

2.0 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.1 3.0

15.2 15.6 15.9 16.3 16.6 16.9 17.2 17.5 17.7 18.1 18.6

8894 9784 10673 11562 12452 13341 14231 15120 16010 16899 17255

1689 1858 2028 2197 2366 2535 2733 2903 2955 3074 3103

Table 8 Key performance indicators for the optimum operation of the CO2 capture plant and compression train. Degree of capture (%)

Amine Slippage (kg/tCO2 )

Ancillary power consumption (kWh/tCO2 )

Cooling duty (MJ/tCO2 )

Reboiler duty (MJ/tCO2 )

Solvent flow rate (m3 /tCO2 )

Volume of packing (m3 /tCO2 h−1 )

50 55 60 65 70 75 80 85 90 95 97

0.50 0.49 0.49 0.49 0.50 0.52 0.55 0.60 0.70 0.91 1.05

123.5 123.8 124.3 124.9 125.6 126.6 127.9 129.6 132.0 136.1 136.6

5523.5 5422.3 5338.5 5268.3 5209.1 5158.8 5126.0 5090.4 5061.9 5002.0 4965.2

4473.9 4474.0 4474.1 4474.2 4474.3 4474.3 4484.2 4484.2 4484.1 4443.2 4415.6

24.0 24.0 24.0 24.0 24.0 24.0 23.9 23.9 23.9 24.3 25.1

26.2 26.1 26.2 26.5 27.0 27.8 28.8 30.3 32.6 36.6 37.0

ing required in the absorber column) and OPEX (e.g., the amount of steam required in the reboiler), changes slightly with higher degree of capture. The optimal design variables listed in Table 7 vary linearly with DOC except for the absorber height and the blower power, which grow exponentially at higher than 60% DOC. Table 8 shows that optimal amine slippage, ancillary power consumption and volume of packing have similar behaviour with different DOC. Their values decrease slightly with higher DOC as a result of economies of scale. Then, the effect of economies of scale vanishes with higher DOC because of the difficulty associated with the process. Amine slippage increases at very high DOC as a result of flue gas exiting a taller absorber column at higher temperature, which in turn vaporized the amine. The ancillary power consumption increases with higher DOC due to dramatic increase of volume of packing, which led to higher power consumption mainly from the blower associated with the increase in pressure drops in taller columns. Table 8 also illustrates that cooling duty decreases with higher DOC as a result of distributing cooling duty consumed on cooling fixed amount of flue gas over increased amount of CO2 being captured. There is a slight change on the reboiler duty with varying DOC as it is highly linked with optimum amine lean loading, which was maintained at a value of 0.28. The reboiler duty obtained in this study (i.e., 4.5 GJ/tCO2 ) is higher than Fluor Econamine process (i.e., 4.2 GJ/tCO2 ) in which MEA based CO2 capture plant is attached to coal-fired power plant (Chapel et al., 1999). This higher rate is due to cost-optimal prediction of lower amine lean loading and higher temperature difference in rich lean heat exchanger, which resulted in a rich solvent entering the stripper column at a lower temperature. A similar reboiler duty (i.e., 4.2 GJ/tCO2 ) can be obtained by operating the capture plant at an amine lean loading of 0.31 and a rich lean heat exchanger T of 10 (Alhajaj et al., 2016). The discrepancy in reboiler duty obtained in this study in comparison to the ones reported in the literature

is due to advancement in process configuration (Amrollahi et al., 2012; Knudsen et al., 2011; NETL, 2010b; Reddy et al., 2003) and utilization of different thermodynamic models (Amrollahi et al., 2011; EBTF, 2011; Fernandez et al., 2014; Mac Dowell et al., 2013; Manzolini et al., 2015; Mores et al., 2012). In fact, more than 10% variation in reboiler duty was reported when using two different thermodynamic models (Darde et al., 2012). The current study highlights the impact of optimizing the carbon capture block on the performance of CCGT power plant. The results of the optimization model tend to minimize both CAPEX and energy consumption, which will lead to a reduction in the energy penalty associated with extra fuel needed to maintain electricity production at the nameplate capacity level.

4.1.2. Relationship between CAPEX, OPEX and LCCC vs DOC The optimum operating control variables outlined in Table 6 in addition to amine lean loading and design variables listed in Table 7 resulted in CAPEX, OPEX and LCCC of the capture plant and compression as shown in Fig. 2. It illustrates that the path for the capital cost and operating cost are similar at the optimum operating conditions. Initially, there is a gradual decrease in CAPEX, OPEX, and LCCC with higher degree of capture (i.e., 50%–55% DOC) as a result of economies of scale. Then, there is a shallow minimum between 55%–80% DOC arising from the simultaneous optimization of both design and control variables. After this, the cost of capture increases gradually with higher than 80% DOC as shown in Fig. 2. This figure also highlights that CAPEX has a major weighting factor in selecting the cost-optimal design and control variables. CAPEX is highly linked with the Capacity Factor (CF), which is assumed to be 0.7 in the current study. A number of optimizations were performed considering a CF of 0.5 and 0.9 to examine their effects on the results of the model. A higher CF (i.e., 0.9) reduced the CAPEX weighting value, which resulted in a higher amine lean

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

339

Table 9 Distribution of the ISBL cost between the carbon capture plant and compression train. DOC (%)

Absorber Blower (%) (%)

Compressor system (%)

CondenserDCC (%) system (%)

Lean amine cooler (%)

Reboiler Rich & lean (%) pumps (%)

Rich lean HE (%)

Scrubber system (%)

Stripper system (%)

50 55 60 65 70 75 80 85 90 95 97

28.8 28.9 28.9 29.8 30.8 32.0 33.3 35.0 37.2 40.4 40.7

23.9 23.5 23.3 23.0 22.6 22.2 21.7 21.1 20.3 19.1 18.9

1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.8 1.7 1.5

3.2 3.2 3.2 3.2 3.2 3.2 3.1 3.1 3.0 2.9 3.1

13.3 13.4 13.5 13.5 13.5 13.5 13.4 13.2 12.8 12.1 11.9

10.6 10.6 10.7 10.7 10.7 10.7 10.5 10.4 10.1 9.7 9.8

2.4 2.5 2.5 2.5 2.4 2.4 2.3 2.2 2.2 2.0 1.8

9.0 9.1 9.1 8.6 8.2 7.7 7.3 6.8 6.4 5.9 5.9

3.8 3.7 3.7 3.7 3.8 3.8 3.9 4.0 4.1 4.3 4.3

2.9 2.9 2.9 2.7 2.6 2.4 2.2 2.1 2.0 1.8 1.8

0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.2 0.2 0.2

Table 10 Distribution of the variable cost including amine make-up for different DOC of the CO2 capture plant and compression train. DOC (%)

Amine make-up (%)

Blower (%)

Compressor system (%)

Condenser (%)

DCC system (%)

Lean amine cooler (%)

Reboiler (%)

Rich & lean pumps (%)

Scrubber system (%)

50 55 60 65 70 75 80 85 90 95 97

14.5 14.4 14.4 14.4 14.5 14.6 14.8 15.2 15.9 17.3 18.4

4.8 4.9 5.0 5.2 5.3 5.6 5.8 6.2 6.7 7.5 7.5

26.0 26.1 26.1 26.1 26.1 26.0 25.9 25.7 25.3 24.8 24.5

3.7 3.7 3.7 3.7 3.7 3.7 3.8 3.7 3.7 3.4 3.1

3.8 3.5 3.2 3.0 2.8 2.7 2.5 2.4 2.3 2.1 2.1

5.7 5.7 5.7 5.7 5.7 5.7 5.7 5.6 5.6 5.5 5.9

37.8 37.9 38.0 38.0 37.9 37.8 37.7 37.4 36.9 35.7 35.2

0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.8

3.0 3.0 3.0 3.1 3.1 3.0 3.0 3.0 3.0 2.9 2.6

Fig. 2. The profile of unit cost (i.e., OPEX, CAPEX and LCCC) against different DOC.

loading (i.e., 0.289). This reflects in an optimal design that slightly favours reduction in reboiler duty over CAPEX. A lower CF (i.e., 0.5) increased the CAPEX weighting value, which fixed the results at their previous values. 4.1.2.1. Distribution of units CAPEX, OPEX and utilities cost vs DOC. The CAPEX distribution is presented here using the Inside Battery Limit (ISBL) investment while the OPEX values are presented through variable cost and the utilities consumptions (i.e., steam, cooling water, electricity and amine make up). Table 9 lists the contribution of each unit in the total CAPEX for different CO2 removal targets. The most expensive units in this process are as follows: absorber; compressor train; reboiler; rich lean heat exchanger; stripper. This order does not change with different DOC.

The DOC effects on different units can be summarized as follows. The absorber and blower contributions to the capital cost increase with higher DOC as a result of the dramatic increase in the height of the absorber column that outweighed the benefits of economies of scale and resulted in higher pressure drops. The compressor train, stripper and DCC system contributions to CAPEX decrease with a higher degree of capture as a result of economies of scale. The degree of capture has a slight effect on the CAPEX of the rest of the units. Table 10 lists the contribution of the above units in addition to amine make-up to the variable cost of the CO2 capture plant and compression train. The largest contributors to this cost in order are as follows: reboiler (heat); compressor (work—Electric-driven in our case); amine make-up (material). This order does not change with varying the degree of capture. In fact, aside from the blower and DCC system, the contributions of the other unit operations do not change with DOC. The blower contribution to variable cost increases with a higher DOC as a result of the extra power requirements to overcome the pressure drop in the system. The relative decrease in the variable cost of the DCC system arises primarily from the increased cost associated with the significant increase in amine make-up costs arising from the increasing DOC. Table 11 lists the distribution of utilities cost against varying DOC. The order of their contributions in the total utility cost from higher to lower are as follows: steam; electricity; cooling water; MEA make-up. This order does not change with varying DOC. Although the steam contribution to the utility cost decreases at high degree of capture, it is steady between 50%–85% DOC. The cooling water consumption decreases with higher DOC as a result of economies of scale. Electricity consumption increases with higher DOC because of the higher solvent rate and higher mass of CO2 that need to be compressed. MEA make-up increases at both high and low DOC.

340

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

Table 11 Distribution of utilities cost contributions for varying DOC of the CO2 capture plant and compression train. DOC (%)

Cooling water (%)

Electricity (%)

MEA make-up (%)

Steam (%)

50 55 60 65 70 75 80 85 90 95 97

17.9 17.7 17.5 17.3 17.1 16.9 16.7 16.5 16.2 15.7 15.4

29.8 30.0 30.1 30.3 30.4 30.6 30.7 30.9 31.0 31.3 31.1

14.5 14.4 14.4 14.4 14.5 14.6 14.8 15.2 15.9 17.3 18.4

37.8 37.9 38.0 38.0 37.9 37.8 37.7 37.4 36.9 35.7 35.2

Table 12 Optimum design variables for CO2 capture plant and compression train with flue gas bypass option.

Fig. 3. Effects of flue gas bypass option on optimal solvent flow rate and reboiler duty required for varying DOC.

CO2 removal target (%)

Feed Fraction Ratio (FFR) Absorber, DCC, and Scrubber diameter (m) Stripper diameter (m) Degree of capture (%) Amine lean loading

50

55

60–97

0.84 13.3 7.4 59 0.283

0.91 13.8 7.7 61 0.283

1.0 14.5 8.0 60–97 0.282–0.29

4.2. Effects of flue gas bypass option The optimization problem described above with an additional flue gas bypass option and carbon price of $0, $4 and $23/tCO2 vented was solved in gPROMS. It was observed that similar control variables listed in Table 6 are optimal for varying DOC and carbon price except for the diameters of the columns. Thus, the previously described optimization problem was used to find the optimal amine lean loading, flue gas feed fraction ratio (FFR) representing the percentage allowance of flue gas into the system and diameters of the columns listed in Table 12. The optimal FFR listed in Table 12 reveal that the flue gas bypass option is the cost optimal choice for lower than 60% overall DOC, in agreement with our previous work (Mac Dowell and Shah, 2013). The optimal degree of capture for the remaining flue gas listed in Table 12 is close to the optimal rate (i.e., 60% DOC) that has the lowest volume of packing and blower power consumption per ton of CO2 captured. In the literature, however, the flue gas bypass option was considered to be the optimal choice at higher than 60% overall DOC as a result of treating less flue gas and hence lower solvent circulation rate into the system (Rao and Rubin, 2006). This in turn reduced the reboiler duty, which decreased OPEX in addition to reducing CAPEX resulted from anticipated smaller absorber column. However, our results shown in Fig. 3 indicate that the reboiler duty and liquid circulation rate per ton of CO2 captured against degree of capture are constant and do not change with the flue gas bypass option. This is in agreement with the earlier study of Sanpasertparnich et al. (2010). In fact, the solvent circulation rate per ton of CO2 captured is observed to be linked to the optimal amine lean loading and the amount of CO2 captured, which were similar at varying flue gas bypass ratio (see Fig. 3). 4.3. Effects of carbon price The control and design variables listed in Table 12, which were similar at different carbon prices, were used to obtain the levelized cost against DOC for different carbon prices shown in Fig. 4.

Fig. 4. Effects of flue gas bypass option and carbon price on the total cost of CO2 capture and compression.

Fig. 4 shows the effects of the flue gas bypass option and different carbon prices (i.e., $0, $4 and $23/tCO2 vented) on the levelized cost of CO2 capture and compression from CCGT power plant. The results reflect the benefits of bypassing the flue gas while meeting lower than 60% CO2 reduction targets because of a lower volume of packing required and hence lower CAPEX. The carbon price has a clear impact on the cost optimal DOC: at $0/tCO2 , there is a shallow minimum between 55%–80% DOC arising from the simultaneous optimization of both design and operating parameters; at $4/tCO2 , there is a shallow minimum between 70%–80% DOC; at $23/tCO2 , there is a shallow minimum between 85%–90% DOC. Thus, the cost optimal DOC will shift to more than 90% DOC at higher carbon prices which is in agreement with results obtained in our previous study (Mac Dowell and Shah, 2013). Fig. 4 also highlights that a carbon price scheme which ramps up slowly over time might, in the first instance, encourage a partial capture solution, with the subsequent expansion of the capture process in line with increasing carbon prices. This could have the effect of reducing the disincentive normally associated with the substantial up-front capital cost required to deploy a CO2 capture process capable of capturing 90% of the CO2 arising from the power plant or industrial facility. However, a practical engineering challenge associated with this approach would be ensuring that the partial capture scenarios were suitable for subsequent expansion and that sufficient space and cooling capacity would be subsequently available. It is also shown in Fig. 4 that the lowest cost of CO2 capture and compression is $62.6/tCO2 which is much higher than the current carbon price of $23/tCO2 vented assumed in this study. Thus, it will be cheaper to pay the costs associated with venting CO2 emissions than investing in CCS. At the right carbon price (i.e., higher than $63/tCO2 ), which covers the cost of CO2 capture and compression,

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

there will be a motivation to capture the CO2 at higher than 95% DOC. This is potentially an important message to policy makers; if a suitably high carbon price does not materialize through a market mechanism, appropriate policies may need to be put in place to achieve decarbonisation targets. Alternatively, stringent emissions performance standards (EPS) will be required to incentivize high rates of decarbonisation (i.e., >90%). The policy makers need also to take into account the right scheme that will encourage the widespread adoption of carbon capture and storage should it be deemed a necessary part of the energy mix (which is likely given its ability to provide both base load and flexible generation). 4.4. Comparison between optimal design and control variables found in this study and values reported in literature The optimal design and control variables obtained in this study were compared to the results obtained by Mores et al. (2012) in which half of the flue gas was treated and compressed to a lower pressure (i.e., 8600 kPa compared to 14,000 kPa used in this study). Thus, higher values for most of the design variables listed in Table 13 were expected. There is an agreement in the design and most of the control variables of the absorber column (i.e., flue gas and lean solvent temperature inlet to absorber column) obtained in both studies. A higher optimal amine lean loading was predicted in the current study (i.e., 0.28 compared to 0.24 in previous study). This has resulted in a lower reboiler duty obtained in this study (i.e., 4484 MJ/tCO2 ) than those reported in the previous study (i.e., 5751 MJ/tCO2 ). In fact, a reboiler duty of 5700 MJ/tCO2 corresponds to operating the capture plant at the amine lean loading of 0.24 obtained in part I of this series (Alhajaj et al., 2016). A relatively high blower power duty was obtained in the earlier study due to a higher pressure drops predicted in the earlier study (i.e., 0.46 kPa/m compared to 0.2 kPa assumed in the current study). The compression power duty obtained in this study is inline with CO2 being treated and compressed to a higher pressure in this study as a result of design factors considering the availability of cooling water temperature in hot countries (Alhajaj et al., 2016). There are discrepancies in the optimal design variables obtained for heat exchangers due to the different assumptions made about the overall heat transfer coefficients, steam inlet temperature, cooling water inlet and exit temperature and reported heat loads. The prediction of lower condenser duty in this study is partially attributed to the assumption of 90% purity of CO2 exiting the condenser in this study based on industrial practice (Iijima, 1998), as more water was condensed in the first stages of compression. Further, a lower reboiler duty predicted in the current study resulted in less steam being generated and much of that steam is being condensed in a taller stripper column obtained in the current study. A relatively high amine lean cooler duty was predicted in the current study as a result of solvent exiting the rich lean HE at high temperature. This is due to a prediction of higher optimal temperature difference in the rich lean HE in the current study. 5. Conclusions A detailed techno-economic model of a CO2 capture and compression train was proposed and implemented in gPROMS. An optimization-based study was carried out in order to simultaneously design and find the cost optimal amine lean loading, flue gas after cooling and lean solvent temperature, reboiler and stripper pressure, temperature differences in heat exchangers and flue gas feed fraction ratio (FFR) for different DOC at the design stage. There was consistency on most of the above-mentioned variables for a

341

Table 13 Comparison between optimal design and control variables obtained in this study and values reported in literature. This study

Mores et al. (2012)

Flue gas stream Molar flow rate (mol/s) CO2 composition (mol%) DOC (%)

22,000 5 85

10,000 4.22 85

Absorber Diameter (m) Height (m) Volume (m3 /tCO2 h−1 ) Amine lean loading Flue gas temperature after cooling (◦ C) Tls inlet to absorber column (◦ C)

14.5 29.5 32 0.282 50 45

12.13 19.9 40.5 0.241 50 45.4

Blower Power duty (kWh/tCO2 ) Pressure drops (kPa/m)

26.1 0.2

56.1 0.46

Compressor Power duty (kWh/tCO2 ) CO2 outlet pressure (kPa)

99.4 14000

96.7 8600

Condenser Area (m) Duty (MJ/tCO2 ) LMTD U (kW m−2 K−1 )

2903 1307 44.7 0.425

4031 3259.1 39.8 0.320

Lean amine cooler Area (m) Duty (MJ/tCO2 ) LMTD U (kW m−2 K−1 )

4808 1974.1 21.8 0.795

1302 1694.6 20.4 1.005

Reboiler Area (m) Duty (MJ/tCO2 ) LMTD U (kW m−2 K−1 ) Reboiler Pressure (kPa)

12468.9 4484.2 17.9 0.850 0.202

2687.6 5751 24.8 1.360 0.202

Rich lean heat exchanger Area (m) Duty (MJ/tCO2 ) LMTD U (kW m−2 K−1 )

13512 4436 19.5 0.710

15046 6974 9.6 0.7608

Stripper Diameter (m) Height (m) Volume (m3 /tCO2 h−1 )

8 17.5 5.8

4.59 5.49 0.0005

range of DOC. The cost optimal amine lean loading has an average value of 0.28, which is the balance between the OPEX represented in the reboiler duty and the CAPEX represented by the height of the columns. Operating the capture plant at flue gas temperature (after cooling) of 50 ◦ C and lean solvent temperature of 45 ◦ C minimized the total levelized cost because of the reduction of the height of the columns associated with enhanced reaction rate and solubility. It was also found that operating the reboiler and stripper at 0.2 MPa was the cost optimal solution as a result of the reduction of stripper height and enhanced driving force in the stripper column. The optimal temperature difference for the rich lean heat exchangers was observed to be 20 ◦ C. Similarly, the temperature difference for the DCC cooler, and the scrubber cooler were 14.4 ◦ C and 15 ◦ C respectively. The flue gas bypass option was observed to be the cost optimal approach for lower than 60% overall DOC because capturing the remaining flue gas at higher than 60% DOC results in an increase in volume of packing and hence higher blower power. The rankings of the importance of the different elements contributing to CAPEX, OPEX and utilities remained the same when DOC was varied. The rankings for contribution to CAPEX were as follows: absorber; compressor train; stripper; rich lean heat

342

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343

exchanger. The rankings for contribution to OPEX were as follows: Reboiler; compressor; amine make-ups. The ranking of their contributions in the total utility cost from higher to lower are as follows: steam; electricity; cooling water; MEA make-up. It was also observed that the carbon price has a clear impact on the cost optimal DOC: at $0/tCO2 , there is a shallow minimum between 55%–80% DOC; at $4/tCO2 , there is a shallow minimum between 70%–80% DOC; at $23/tCO2 , there is a shallow minimum between 85%–90% DOC. The assumed carbon prices are lower than the minimum cost of CO2 capture and compression cost (i.e., $63/tCO2 ). Thus, it would be cheaper to vent the CO2 rather than investing in CCS with current carbon prices. At a sufficiently high carbon price (i.e., higher than $60/tCO2 ), which covers the cost of CO2 capture and compression, there will be sufficient incentive to capture the CO2 at higher than 95% DOC. Acknowledgement The authors thank Masdar Institute and the Engineering and Physical Sciences Research Council (EPSRC) of the UK (MESMERISECCS, Grant number: EP/M001369/1) for funding required to carry out this work. References Abu-Zahra, M.R.M., Niederer, J.P.M., Feron, P.H.M., Versteeg, G.F., 2007a. CO2 capture from power plants: part II. A parametric study of the economical performance based on mono-ethanolamine. Int. J. Greenh. Gas Control 1, 135–142. Abu-Zahra, M.R.M., Schneiders, L.H.J., Niederer, J.P.M., Feron, P.H.M., Versteeg, G.F., 2007b. CO2 capture from power plants: part I: A parametric study of the technical performance based on monoethanolamine. Int. J. Greenh. Gas Control 1, 37–46. Adams, T., Mac Dowell, N., 2016. Off-design point modelling of a 420 MW CCGT power plant integrated with an amine-based post-combustion CO2 capture and compression process. Appl. Energy 178, 681–702. Alhajaj, A., Mac Dowell, N., Shah, N., 2016. A techno-economic analysis of post-combustion CO2 capture and compression applied to a combined cycle gas turbine: part I. A parametric study of the key technical performance indicators. Int. J. Greenh. Gas Control 44, 26–41. Amrollahi, Z., Ertesvåg, I.S., Bolland, O., 2011. Thermodynamic analysis on post-combustion CO2 capture of natural-gas-fired power plant. Int. J. Greenh. Gas Control 5, 422–426. Amrollahi, Z., Ystad, P.A.M., Ertesvåg, I.S., Bolland, O., 2012. Optimized process configurations of post-combustion CO2 capture for natural-gas-fired power plant–power plant efficiency analysis. Int. J. Greenh. Gas Control 8, 1–11. Bailey, D.W., Feron, P.H.M., 2005. Capture post-combustion. Oil Gas Sci. Technol.: Rev. IFP 60, 461–474. Billet, R., Schultes, M., 1993. A physical model for the prediction of liquid hold-up in two-phase countercurrent columns. Chem. Eng. Technol. 16, 370–375. Chapel, D., Ernest, J., Mariz, C.L., 1999. Recovery of CO2 from flue gases: commercial trends. Presented at the Canadian Society of Chemical Engineering Annual Meeting. Chapman, W.G., Gubbins, K.E., Jackson, G., Radosz, M., 1989. SAFT: equation-of-state solution model for associating fluids. Fluid Phase Equilib. 52, 31–38. Chapman, W.G., Gubbins, K.E., Jackson, G., Radosz, M., 1990. New reference equation of state for associating liquids. Ind. Eng. Chem. Res. 29, 1709–1721. Chauvel, A., Miller, R., Miller, E.B., 1981. Manual of Economic Analysis of Chemical Processes: Feasibility Studies in Refinery and Petrochemical Processes. McGraw-Hill. Couper, J.R., 2010. Chemical Process Equipment: Selection and Design, Rev. 2nd ed. Elsevier Butterworth-Heinemann, Amsterdam, London. Cran, J., 1981. Improved factored method gives better preliminary cost estimates. Chem. Eng. 88, 65–79. DECC, 2011. Electricity generation cost model-2011 update revision, London. Darde, V., Thomsen, K., van Well, W.J.M., Bonalumi, D., Valenti, G., Macchi, E., 2012. Comparison of two electrolyte models for the carbon capture with aqueous ammonia. Int. J. Greenh. Gas Control 8, 61–72. Douglas, J.M., 1988. Conceptual Design of Chemical Processes. McGraw-Hill, New York. EBTF, 2011. ED 4.9 European best practice guidelines for assessment of CO2 capture technologies. EPRI, 1993. TAGTM Techanical Assessment Guide—Electercity supply, Palo Alto, CA. Fernandez, E.S., Goetheer, E., Manzolini, G., Macchi, E., Rezvani, S., Vlugt, T., 2014. Thermodynamic assessment of amine based CO2 capture technologies in power plants based on European Benchmarking Task force methodology. Fuel 129, 318–329.

Finkenrath, M., 2011. Cost and Performance of Carbon Dioxide Capture from Power Generation. IEA, Paris. GCCSI, 2011. Economic Assessment of Carbon Capture and Storage Technologies. Global CCS Institute, Canberra ACT, Australia (2011 update). GPSA, 1987. Engineering Data Book, 10th ed. Gas Processors Suppliers Association, Tulsa. Galindo, A., Davies, L.A., Gil-Villegas, A., Jackson, G., 1998. The thermodynamics of mixtures and the corresponding mixing rules in the SAFT-VR approach for potentials of variable range. Mol. Phys. 93, 241–252. Gil-Villegas, A., Galindo, A., Whitehead, P.J., Mills, S.J., Jackson, G., Burgess, A.N., 1997. Statistical associating fluid theory for chain molecules with attractive potentials of variable range. J. Chem. Phys. 106, 4168–4186. Green, D., Perry, R., 2008. Perry’s Chemical Engineers’ Handbook, 8th ed. McGraw-Hill, New York. Guthrie, K.M., 1969. Data and techniques for preliminary capital cost estimating. Chem. Eng. 76, 114–142. Henao, C.A., 2006. Simulación y evaluación de procesos químicos, 1st ed. Universidad Pontificia Bloivariana, Medellín. Iijima, M., 1998. A feasible new flue gas CO2 recovery technology for enhanced oil recovery. In: S.o.P. Engineers (Ed.), SPE/DOE Improved Oil Recovery Symposium. Society of Petroleum Engineers, Tulsa, pp. 333–342. Jordal, K., Ystad, P.A.M., Anantharaman, R., Chikukwa, A., Bolland, O., 2012. Design-point and part-load considerations for natural gas combined cycle plants with post combustion capture. Int. J. Greenh. Gas Control 11, 271–282. Kern, D.Q., 1965. Process Heat Transfer. Mcgraw-Hill. Kittel, J., Idem, R., Gelowitz, D., Tontiwachwuthikul, P., Parrain, G., Bonneau, A., 2009. Corrosion in MEA units for CO2 capture: pilot plant studies. Energy Procedia 1, 791–797. Klemeˇs, J., Bulatov, I., Cockerill, T., 2007. Techno-economic modelling and cost functions of CO2 capture processes. Comput. Chem. Eng. 31, 445–455. Knudsen, J.N., Andersen, J., Jensen, J.N., Biede, O., 2011. Results from test campaigns at the 1 t/h CO2 post combustion capture pilot-plant in Esbjerg under the EU FP7 CESAR project. 1st Post Combustion Capture Conference, 17–19. Kvamsdal, H.M., Jordal, K., Bolland, O., 2007. A quantitative comparison of gas turbine cycles with capture. Energy 32, 10–24. Llovell, F., Mac Dowell, N., Blas, F.J., Galindo, A., Jackson, G., 2012. Application of the SAFT-VR density functional theory to the prediction of the interfacial properties of mixtures of relevance to reservoir engineering. Fluid Phase Equilib. 336, 137–150. Mac Dowell, N., Shah, N., 2013. Identification of the cost-optimal degree of CO2 capture: an optimisation study using dynamic process models. Int. J. Greenh. Gas Control 13, 44–58. Mac Dowell, N., Shah, N., 2014a. Dynamic modelling and analysis of a coal-fired power plant integrated with a novel split-flow configuration post-combustion CO2 capture process. Int. J. Greenh. Gas Control 27, 103–119. Mac Dowell, N., Shah, N., 2014b. Optimisation of post-combustion CO2 capture for flexible operation. Energy Procedia 63, 1525–1535. Mac Dowell, N., Shah, N., 2015. The multi-period optimisation of an amine-based CO2 capture process integrated with a super-critical coal-fired power station for flexible operation. Comput. Chem. Eng. 74, 169–183. Mac Dowell, N., Llovell, F., Adjiman, C.S., Jackson, G., Galindo, A., 2010. Modeling the fluid phase behavior of carbon dioxide in aqueous solutions of monoethanolamine using transferable parameters with the SAFT-VR approach. Ind. Eng. Chem. Res. 49, 1883–1899. Mac Dowell, N., Pereira, F.E., Llovell, F., Blas, F.J., Adjiman, C.S., Jackson, G., Galindo, A., 2011. Transferable SAFT-VR models for the calculation of the fluid phase equilibria in reactive mixtures of carbon dioxide, water, and n-alkylamines in the context of carbon capture. J. Phys. Chem. B 115, 8155–8168. Mac Dowell, N., Samsatli, N.J., Shah, N., 2013. Dynamic modelling and analysis of an amine-based post-combustion CO2 capture absorption column. Int. J. Greenh. Gas Control 12, 247–258. Manzolini, G., Sanchez Fernandez, E., Rezvani, S., Macchi, E., Goetheer, E.L.V., Vlugt, T.J.H., 2015. Economic assessment of novel amine based CO2 capture technologies integrated in power plants based on European Benchmarking Task Force methodology. Appl. Energy 138, 546–558. Mariz, C.L., 1998. Carbon dioxide recovery: large scale design trends. J. Can. Petrol. Technol. 37, 42–47. Megyesey, E.F., Buthod, P., 1986. Pressure Vessel Handbook, 7th ed. Pressure Vessel Handbook Publishing, Tulsa. Metz, B., Davidson, O., de Coninck, H.C., Loos, M., Meyer, L.A., 2005. IPCC Special Report on Carbon Dioxide Capture and Storage. Cambridge University Press for the Intergovernmental Panel on Climate Change, Cambridge. Mores, P., Rodríguez, N., Scenna, N., Mussati, S., 2012. CO2 capture in power plants: minimization of the investment and operating cost of the post-combustion process using MEA aqueous solution. Int. J. Greenh. Gas Control 10, 148–163. NETL, 2010. Cost and performance baseline for fossil energy plants, Rev. 2, Report DOE/NETL-2010/1397, Nov, Pittsburgh, PA. NETL, 2010b. Cost and Performance Baseline for Fossil Energy Plants Volume 1: Bituminous Coal and Natural Gas to Electricity. National Energy Technology Laboratory, Washington, DC, USA. Oexmann, J., Hensel, C., Kather, A., 2008. Post-combustion CO2 -capture from coal-fired power plants: preliminary evaluation of an integrated chemical absorption process with piperazine-promoted potassium carbonate. Int. J. Greenh. Gas Control 2, 539–552. Peters, M.S., Timmerhaus, K.D., West, R.E., 2004. Plant Design and Economics for Chemical Engineers, 5th ed. McGraw-Hill, Boston.

A. Alhajaj et al. / International Journal of Greenhouse Gas Control 52 (2016) 331–343 Rao, A.B., Rubin, E.S., 2002. A technical, economic, and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. Environ. Sci. Technol. 36, 4467–4475. Rao, A.B., Rubin, E.S., 2006. Identifying cost-effective CO2 control levels for amine-based CO2 capture systems. Ind. Eng. Chem. Res. 45, 2421–2429. Rao, A.B., 2002. Details of a technical, economic and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. IECM Documentation. Reddy, S., Scherffius, J., Freguia, S., Roberts, C., 2003. Fluor’s econamine FG PlusSM technology. In: Proceedings of the Second Annual Conference on Carbon Sequestration, Citeseer, pp. 5–8. Richardson, J.F., Harker, J.H., Backhurst, J.R., 1999. Coulson & Richardson’s Chemical Engineering, 3rd ed. Butterworth/Heinemann, Oxford. Rubin, E.S., Kalagnanam, J.R., Frey, H.C., Berkenpas, M.B., 1997. Integrated environmental control modeling of coal-fired power systems. J. Air Waste Manag. Assoc. 47, 1180–1188. Rubin, E.S., Chen, C., Rao, A.B., 2007a. Cost and performance of fossil fuel power plants with CO2 capture and storage. Energy Policy 35, 4444–4454. Rubin, E.S., Yeh, S., Antes, M., Berkenpas, M., Davison, J., 2007b. Use of experience curves to estimate the future cost of power plants with CO2 capture. Int. J. Greenh. Gas Control 1, 188–197. Rubin, E.S., Booras, G., Davison, J., Ekstrom, C., Matuszewski, M., Short, C., 2013. Toward a common method of cost Estimation for CO2 capture and storage at fossil fuel fower plants.

343

Rubin, E.S., 2012. Understanding the pitfalls of CCS cost estimates. Int. J. Greenh. Gas Control 10, 181–190. Sanpasertparnich, T., Idem, R., Bolea, I., deMontigny, D., Tontiwachwuthikul, P., 2010. Integration of post-combustion capture and storage into a pulverized coal-fired power plant. Int. J. Greenh. Gas Control 4, 499–510. Scottish Power, 2011. CCS project costs, UK carbon capture and storage demonstration competition FEED close out report, UK. Seider, W.D., Seader, J.D., Lewin, D.R., 2009. Product and Process Design Principles: Synthesis, Analysis and Evaluation, 3rd ed. John Wiley & Sons. Singh, D., Croiset, E., Douglas, P.L., Douglas, M.A., 2003. Techno-economic study of CO2 capture from an existing coal-fired power plant: MEA scrubbing vs. O2 /CO2 recycle combustion. Energy Convers. Manag. 44, 3073–3091. Ulrich, G.D., Vasudevan, P.T., 2004. Chemical Engineering Process Design and Economics: A Practical Guide, 2nd ed. Process Publishing. Ulrich, G.D., Vasudevan, P.T., 2009. Capital costs quickly calculated: estimating capital costs early can prevent unnecessary expenditures on dead-end projects. Chem. Eng., 46–52. ZEP, 2011. The cost of CO2 capture, transport and storage: post-demonstration CCS in the EU, Brussels.