Multiscale design and analysis of CO2 networks

Multiscale design and analysis of CO2 networks

International Journal of Greenhouse Gas Control 94 (2020) 102925 Contents lists available at ScienceDirect International Journal of Greenhouse Gas C...

2MB Sizes 0 Downloads 49 Views

International Journal of Greenhouse Gas Control 94 (2020) 102925

Contents lists available at ScienceDirect

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

Multiscale design and analysis of CO2 networks Ahmed Alhajaj

a,b,

T

b

*, Nilay Shah

a

Research and Innovation Centre on CO2 and H2, Department of Chemical Engineering, Khalifa University of Science and Technology, PO Box 127788, Abu Dhabi, United Arab Emirates Centre for Process Systems Engineering, Department of Chemical Engineering, Imperial College London, SW7 2AZ, UK

b

A R T I C LE I N FO

A B S T R A C T

Keywords: CCS CO2 transport Optimization Cost Supply-chain CO2 network

Carbon Capture and Storage (CCS) is an essential technology for CO2 emissions reductions, which will allow us to continue consuming fossil fuels in the short to medium term. In this work, we developed a multiscale modeling and optimization approach that links detailed models of the capture plant, compression train and pipelines with the CO2 supply-chain network model. This was used to find the cost-optimal CO2 network considering a case-study of meeting a national reduction target in the United Arab Emirates that supplies CO2 for EOR activities. The main decision variables were the optimal location and operating conditions of each CO2 capture and compression plant in addition to the topology and sizing of the pipelines while considering the whole-system behaviour. A key result of our study was that the cost-optimal degree of capture should be included as a degree of freedom in the design of CO2 networks and it is a function of several site-specific factors, including exhaust gas characteristics, proximity to transportation networks and adequate geological storage capacity. This conclusion serves to underscore the need to comprehend the science governing the physical behaviour at different scales and the importance of a whole-system analysis of potential CO2 networks.

1. Introduction

1.1. Previous work

Carbon capture and storage (CCS) is an essential technology for CO2 emissions reductions which will allow us to continue consuming fossil fuels in the short to medium term. Although the technologies involved in CCS have been proved independently, the remaining challenge facing governments and industries is to implement the whole system safely and at the minimum cost on a large scale that makes a material effect in global warming mitigation. This requires detailed analysis of the CO2 network system and understanding of several site-specific factors that must be considered when applying this technology on a scale that minimizes the environmental impact of global warming. For instance, the proximity of the sources to appropriate geological storage sites and the ability of the storage sites to accommodate these emissions at a certain rate per year and for the life-time of the project. The CO2 network design involves formulating the problem using different methods to identify appropriate values of some of the following decision variables: sources to be included in the network; degree of capture from each selected source; network topology including pipeline sizes, capacity and flow rate; sinks to be included in the network.

Previous work formulated the design of CO2 network problem using the following approaches: a simplified approach in which the potential CO2 networks were obtained by limiting the distance between attractive CO2 sources and sinks (Bradshaw et al., 2004; van Bergen et al., 2004) scenario based approaches in which different scenarios are manually assessed while using different common performance indicator (Brunsvold et al., 2011; Jakobsen and Brunsvold, 2011; Jakobsen et al., 2011; Jakobsen et al., 2008; Poyry Energy, 2007; Røkke et al., 2009; Forward, 2008) heuristic approaches in which the problem is formulated and solved using generic algorithms (Fimbres Weihs and Wiley, 2012; Fimbres Weihs et al., 2011; Krickenberger and Lubore, 1981; Turk et al., 1987) systematic approaches in which performance objective functions presented in mathematical form are used to find the optimal layout of the CO2 network (Akimoto et al., 2004; Alhajaj, 2008; Bakken and von Streng Velken, 2008; Boavida et al., 2011; Han and Lee, 2011a, b, 2012; Kemp and Sola Kasim, 2010; Klokk et al., 2010; Kuby et al., 2011a, b; Lee et al., 2012; Middleton and Bielicki, 2009; Middleton et al., 2012a, b, c; Strachan et al., 2011; van den Broek et al., 2009; van den Broek et al., 2010). The CO2 network design problem is best dealt with using a systems

⁎ Corresponding author at: Research and Innovation Centre on CO2 and H2, Department of Chemical Engineering, Khalifa University of Science and Technology, PO Box 127788, Abu Dhabi, United Arab Emirates. E-mail address: [email protected] (A. Alhajaj).

https://doi.org/10.1016/j.ijggc.2019.102925 Received 10 June 2019; Received in revised form 23 November 2019; Accepted 24 November 2019 1750-5836/ © 2019 Elsevier Ltd. All rights reserved.

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Fig. 1. Multiscale CCS modelling components.

attempted by two groups. The first group developed multiscale models for CO2 network design in which storage site models that range from the pore scale to the reservoir and site scales are coupled with network models (Middleton et al., 2012a, c; Keating et al., 2010). The second group focused on developing surrogate models for different CO2 capture technologies assuming a 90 % degree of capture and compressor output pressure of 150 bar, which was then integrated in the CO2 network model (Hasan et al., 2014, 2015). However, this fixed design degree of capture and compressor output pressure might not be the optimal choice when looking at the whole integrated CO2 network design, where multiple capture sites are available.

modeling and optimization approach. It is characterised by a set of mathematical equations representing performance objectives that assess alternative decisions of the system and a set of equality constraints that represent the behaviour of the system in addition to inequality constraints that limit the solutions to a feasible region only. The optimization-based model comprises parameters (e.g., CO2 reduction target) and decision variables that can be binary (e.g., establish capture plant or not) or continuous (e.g., amount of CO2 transported along a particular link). The solution obtained while minimizing the objective function (e.g., cost, environmental impact) and incorporating the constrained equations that limits the feasible region (e.g., capacity of pipelines) is the global optimal, if it is formulated as mixed integer linear programming (MILP) problem or other convex model. Thus, this makes optimisation-based network design ideal for decision makers. This type of model can be classified according to how uncertainties in the parameters are handled and the level of details incorporated, leading to: deterministic; stochastic; and multiscale models. The first two approaches were used in the literature to find the cost optimal CO2 network while utilizing single-scale models in which cost and physical models of many components were simplified or overlooked (Akimoto et al., 2004; Alhajaj, 2008; Bakken and von Streng Velken, 2008; Boavida et al., 2011; Han and Lee, 2011a, b, 2012; Kemp and Sola Kasim, 2010; Klokk et al., 2010; Kuby et al., 2011a, b; Lee et al., 2012; Middleton and Bielicki, 2009; Middleton et al., 2012a, b, c; Strachan et al., 2011; van den Broek et al., 2009, 2010). On the other hand, there have been limited attempts in developing multiscale modeling approaches in which the complex CCS system that exhibits behaviours across different length and time scales is described using a series of interacting scale-specific models. This ensures that the science governing the physical behaviour of the components is taken into consideration while making non-trivial strategic decisions. The previous attempts in developing multiscale model were mainly

1.2. Aim and novelty of this paper This aim of this paper is to develop a multiscale modeling and optimization approach that links detailed models of the capture plant, compression train and pipelines with the CO2 supply-chain network model. This can provide decision makers with a systematic tool that can help them find and analyse the cost optimal deployment of CCS infrastructure that meets different CO2 reduction mandates at the national or regional level. The novelty of this paper is based on adding dimensionalities in the multiscale network model through inclusion of design variables of the capture and compression plants such as degree of capture, bypass ratio and compressor output pressure that can be obtained through the optimization of interconnected CO2 network, allowing simultaneous process and system optimization. This was done by obtaining the optimal design variables for varying degree of capture attached to different sources and then use this macro variable with associated cost as input in the CO2 network design. The remainder of this paper is laid out as follows: we start by presenting a multiscale model of the CO2 network system. Then, we demonstrate the use of the model in finding the cost-optimal CO2 network 2

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

affected by the local and regional parameters. For example, the availability and temperature of cooling water which in turn affect details of the design such as compression path. Further, the corrosion rate when considering the type of material and solvent used is another important factor that is quantified when incorporated in details of the capture plant. Other global parameters are reflective of the economic conditions such as fuel, water and electricity price. The results obtained by the fine scale models are then passed to the macro scale models as a macro variable (e.g., degree of capture) embedded with a database, which summarise their behaviour in terms of a small number of key variables. These include energetic and economic performance of attaching amine-based post-combustion CO2 capture plants and compression trains to different CO2 sources, which is used in high-level network design.

considering four major CO2 sources and three potential oil recovery sites amenable to CO2 flooding in the UAE. After this, we examine the effect of having various CO2 reduction mandates in the overall cost of the network. Finally, we present concluding remarks.

2. Methodology for multiscale Design and anlysis of CO2 networks A layout of the multi-scale detailed carbon capture, transport and storage system is shown in Fig. 1. There are four macro components within the CCS system: CO2 sources at fixed locations, CO2 capture plants that need to be built, CO2 transportation links that need to be picked among different potential nodes and CO2 storage sites at fixed locations. The CO2 source plants (e.g., power plant, process plant) emit certain amounts of CO2 with varying compositions. The level of CO2 emissions changes in response to demand changes for services, products and utilities throughout the year. The CO2 is assumed to be transported to a storage site via a pipeline network. Different sections of the network are composed of pipelines with different diameters with a minimum and a maximum capacity. The CO2 can be stored in a permanent storage site such as depleted oil reservoir or it can be used for EOR projects. Each storage site has a certain capacity and injectivity that allow it to accommodate a certain amount of CO2 per year. The multiscale approach comprises a series of interacting models that are able to capture behaviours at specific length and time scales. The following scale models are incorporated (see Fig. 2):

2.1. Multiscale model development The following steps were taken to develop the integrated multi-scale optimization based model of CO2 capture, transport and storage networks: (1) Build an inventory of CO2 sources that include composition, flowrate and temperature of the flue gas in addition to CO2 source location (2) Build an inventory of CO2 storage and potential EOR sites including location, capacity and injectivity. (3) Develop economic-performance relationships for attaching MEA based CO2 capture plants and compression trains to each potential CO2 source. This is obtained by performing a number of optimizations using the detailed techno-economic model developed in gPROMS (Alhajaj et al., 2016). A database of the optimal operating, control and design variables in addition to the cost of capture and compression against degree of capture will be developed for these sources. (4) Obtain the optimal operating capacity and velocity for each pipeline diameter while considering the techno-economic interactions between the CO2 compression train and transportation system (Alhajaj and Shah, 2019). (5) Run the network model while using information obtained from steps (1)-(4) as inputs to find the optimal design of CO2 network:

(1) Molecular scale: a detailed thermodynamic model using SAFT-VR (i.e., molecular model based on Helmholtz free energy) approach is used (Galindo et al., 1998; Chapman et al., 1989, 1990; Gil-Villegas et al., 1997) to evaluate thermophysical properties and reaction equilibria of the MEA-based solvent and flue gas (Mac Dowell et al., 2010, 2011; Rodriguez et al., 2012) (2) Film scale: a two film theory is utilized to find the height required for the diffusion of CO2 through a packed column; (3) Unit scale: an equilibrium-based model comprising mass and energy transfer equations for the whole amine-based CO2 capture plant are used to analyse its performance; (4) Regional scale: a high-level supply chain network model is utilized to map the cost optimal deployment of the CO2 network at the national or global level.

Given In order to inform the decisions at the macro level (e.g., select a source or not, degree of capture, network topology), a number of optimization runs are performed for attaching a CO2 capture plant to each potential CO2 source while using detailed fine scale sub-process models (e.g., units of the capture plant, cost model). Each of these units is



○ Number, location, flue gas characteristics of potential CO2 sources

Fig. 2. Illustration of multiscale CCS modelling. 3

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

factor, life time of the plant

Table 1 Case scenario represents phase 1 of announced UAE reduction target of 5 million tons of CO2 per year. Source/ sink type

Name and location

Steel production plant

Emirates Steel Industry (S1)

Boiler based power plant

Taweela power company (S2)

NGCC power plant

Oil field Oil Field Oil Field

Emirates Aluminum (S3); (S4)

Bu Hasa (O1) Bab (O2) Asab (O3)

Flow rate, temperature, pressure and composition / capacity

Comment

46,920 (Nm3/ hour) 98(oC) 1.01 (bar) 90.4 (mol.%) CO2 9.6 % (mol.%) H20 1,741,000 (Nm3/ hour) 150 (oC) 1.01 (bar) 8.55 (mol.%) CO2 17 (mol.%) H20 74.45 (mol.%) N2 1,800,000 (Nm3/ hour) 98 (oC) 1.01 (bar) 5.06 (mol.%) CO2 12 (mol.%) H20 82.94 (mol.%) N2 9.13 (Mt CO2 /year) 3.91 (Mt CO2 /year) 3.88 (Mt CO2 /year)

No capture required, just compression and dehydration unit

Decision variables



○ Which CO2 sources to select? ○ Which degree of capture and operating conditions should be selected for each capture plant? ○ Where to establish pipelines and at which size? ○ How to connect these pipelines in a network? ○ Which CO2 storage sites to exploit? Objective Function

An indicative 600 MW boiler based power plant data from Iijima (1998) was used



An indicative 400 MW CCGT plant data from Bailey and Feron (2005) was used

○ The objective function is the minimization of the total levelized cost (i.e., capital and operating cost) of the whole system (i.e., CO2 capture, compression, transport and storage).

The optimization is performed subject to a national or regional CO2 emissions reduction target. 2.1.1. Mathematical formulation of the CO2 network model The multiscale model was formulated as a MILP problem in which the complex system was described by a set of equality equations that describe the behaviour of the system, a set of inequality constraints that constrain the solutions in a feasible region and an objective function that the model minimizes in order to identify the optimal set of decisions that represent the deployment of the CO2 network. The mathematical formulation consists of elements namely indices, parameters, continuous variables, and binary variables, constraints and an objective function.

○ Database of economic performance (i.e., capital and operating cost) of CO2 capture and compression to 14 MPa (for different degrees of capture, in the form of data points) from all potential sources using the developed amine based post-combustion capture plant in gPROMS™ ○ CO2 reduction target or potential market for EOR ○ Number, location and capacity of each storage site ○ Financial and technical data such as discount rate, capacity

2.1.1.1. Nomenclature 2.1.1.1.1. Indicesi grid squares j grid squares such that i ≠ j k Degree of capture from each potential source (e.g., 25 %–99 %)

Fig. 3. Location of CO2 sources (S1, S2, S3) and sinks (O1, O2, O3) of case study. 4

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Table 2 The cost-optimal control variables obtained for the capture plant and compression train attached to CCGT and boiler based power plant. Amine lean loading

Tfg after cooling (oC)

Tls inlet to absorber (oC)

Reboiler and stripper pressure (Mpa)

ΔT Rich lean HE (oC)

ΔT Scrubber cooler (oC)

ΔT Scrubber cooler (oC)

0.28-0.29

50.0

45.0

0.202

20.0

15.0

15.0

TCCCk, j Total levelized cost of CO2 captured and compressed in grid square j at degree of capture k, $ t−1 UMF Uncertainty multiplying factor used to take into account the uncertainty in the pipeline cost (e.g., 1.25, 1.5) z k Number represent degree of capture for each k (e.g., 0.25,0.26,…. 0.99) 2.1.1.1.3. Continuous variables. bj Amount of CO2 sequestered in grid square j for EOR, t d−1 ckj Optimal amount of CO2 captured in grid square j at degree of capture k, t d−1 CCl Capital cost for each transportation mode l, $ FCCj Total levelized cost of establishing capture plant and compression train in grid square j, $ d−1 OCl Operating cost for each transportation mode l, $ Qlij CO2 flow rate between grid square i and grid square j by transportation mode l, t d−1 TC Total cost of the whole CCTS network, $ d−1 TDl Total distance of each transportation mode l, km TFCC total levelized cost of establishing all capture plants and compression trains in the network, $ d−1 TPCC Total pipeline capital cost, $ TPOC Total pipeline operating cost, $ d−1 2.1.1.1.4. Binary variables. Xlij 1 if transportation between grid square i and square j by mode l exists, 0 otherwise Ykj 1 if CO2 capture plant will exist in grid square j at degree of capture k, 0 otherwise

Table 3 Parameters and nominal variables used in the optimization and simulation of carbon capture plant and compression train. Pressure of the absorber (kPa) Pressure drops in packing (kPa/m) Water makeup (kg/ton CO2) Capacity factor (CF) Cost year MEA make-up (kg/ton CO2) Norton IMTP 50 mm dry packing area (a) (m2/m3) CO2 outlet pressure from compressor (kPa) CO2 product content from condenser (mol%) Compressor efficiency (ηisn , ηmec ) (%) Pumps efficiency (%)

101 0.2 0 0.7 2004 1.1 kga 120 14000b 90.4c (80, 95)d 75b

a Chapel et al. (1999) predicted total of 1.6 kg/ton CO2 of which 0.5 kg/ton CO2 is vaporized at the optimal capture rate as predicted by the model. This increased model flexibility to account for vaporization. b (Rao and Rubin, 2002). c (Iijima, 1998). d (Kvamsdal et al., 2007).

l Transportation modes l using pipes with different diameter (i.e., 6 inch, 8 inch, 10 inch, 12 inch, 16 inch, 20 inch, 24 inch, 30 in.) 2.1.1.1.2. Parameters. ajmax Maximum amount of CO2 emitted in grid square j, t d−1 ak, j Amount of CO2 captured in each grid square j at degree of capture k ak, j = z k ajmax max atotal Total amount of CO2 emitted from the CO2 sources max ajmax , total = ∑j aj bjmax Maximum amount of CO2 that can be sequestrated for EOR in grid square j, t d−1 or maximum injection capacity Capexl Pipeline capital cost charge factor for each transpiration mode l, $ km−1 Capexlcor Pipeline capital cost charge factor for each transpiration mode l corrected for uncertainty factor, $ km− Capexlcor = UMF Capexl CF Capacity factor CRF Capital recovery factor, (i.e., 15 %) dij Distance between grid i and grid j, km dij = 2 (dyj − dyi )2 + (dx j − dx i )2 dx j Horizontal distance relative to origin for each grid square j, km dyj Vertical distance relative to origin for each grid square j, km Opex Pipeline operating cost charge factor for each transpiration mode l (i.e., 2 % of Capexl ) , $ km−1 Qlmax Maximum optimal flow rate via transportation mode l, t d−1 Qlmin Minimum optimal flow rate via transportation mode l, t d−1 RT CO2 reduction target that need to be met, %

2.1.1.2. Constraints 2.1.1.2.1. Logical constraints. The total amount of CO2 sequestrated in each sink j for EOR should not exceed the maximum injectivity per day:

bj ≤ bjmax ∀ j

(1)

To ensure that only one capture plant with specific degree of capture can be built at each prospective CO2 source, we impose

ckj = akj Ykj

(2)

∑ Ykj ≤ 1

(3)

k

2.1.1.2.2. Material balance constraints. To ensure that CO2 captured or transported within each grid square must be transported out of the grid square or injected for EOR in the grid square, we have

bj =

∑ Qlij − ∑ Qlji + ∑ Ckj l, i

l, i

∀ j (4)

k

Table 4 Optimal operating pipeline variables including capacity and cost for each nominal pipeline size. Pipeline diameter

Capacity (Mt CO2/year)

Velocity (m/s)

Input pressure (MPa)

Compressibility (Z)

Pressure drop (Pa/m)

Cost (k$/ km)

168 mm 219 mm 273 mm 324 mm 406 mm 507 mm 610 mm 762 mm

0.4–0.79 0.8–1.5 1.51–2.2 2.21–3.0 3.01–4.67 4.68–7.02 7.03–11.04 11.05–17.01

0.3–1.4 0.8–1.5 1–1.4 1–1.4 1.0–1.6 1.0–1.5 1.1–1.7 1.1–1.7

8.8–14 10–13.5 10.2–11.9 10.0–11.1 9.8–11.3 9.5–10.5 9.4–10.5 9.2–10.0

0.19–0.24 0.20–0.24 0.20–0.22 0.20–0.21 0.20–0.21 0.19–0.20 0.19–0.20 0.19–0.20

1.9–54 13–49 15.6–32.8 13.7–24.9 12.9–27.2 8.7–19.2 7.6–18.4 6–13.8

146 186 228 268 311 388 469 592

(6 in) (8 in) (10 in) (12 in) (16 in) (20 in) (24 in) (30 in)

5

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Fig. 4. Prospective network while meeting 25 % CO2 reduction target: 219 mm (8 in. pipeline transferring 2316 (tCO2/day) between S2 and S1; 273 mm (10 in. pipeline transferring 4464 (tCO2/day) between S1 and O2.

Fig. 5. Prospective network while meeting 50 % CO2 reduction target: 168 mm (6 in. pipeline transferring 1116 (tCO2/day) between S3 and S2; 324 mm (12 in. pipeline transferring 6730 (tCO2/day) between S2 and S1; 406 mm (16 in. pipeline transferring 8922 (tCO2/day) between S1 and O2.

6

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Fig. 6. Prospective network while meeting 75 % CO2 reduction target: 219 mm (8 in. pipeline transferring 2619 (tCO2/day) between S4 and S3; 273 mm (10 in. pipeline transferring 5282 (tCO2/day) between S3 and S2; 406 mm (16 in. pipeline transferring 11,176 (tCO2/day) between S2 and S1; 507 mm (20 in. pipeline transferring 13,368 (tCO2/day) between S1 and O2; 219 mm (8 in. pipeline transferring 4110 (tCO2/day) between O2 and O3.

We ensure that the flow between two grid squares can only exist in one direction with the optimal Nominal Pipeline Size (NPS) diameter by:

Qlmax Xlij ≤ Qlij ≤ Qlmin Xlij ∀ l, i, j: i ≠ j

(5)

Xlij + Xlji ≤ 1 ∀ i, j: i ≠ j

(6)

k, j

(8)

Qlij ≥ 0 ∀ l, i, j: i ≠ j

(9)

bj ≥ 0 ∀ j

(10)

2.1.1.3. Objective function. In order to find the cost-optimal CO2 infrastructure, the objective function here is to minimize the total levelized cost (i.e., capital and operating cost) incurred with establishing and operating CO2 capture plants and pipelines, and subject to the constraints outlined earlier. The total cost assumed in this study is the gate delivery cost to the potential EOR site. Any additional cost associated with pressurizing the gas further or modifying the existing storage sites infrastructure is not included in this study. 2.1.1.3.1. Capture plant and compression train levelized cost. The total daily levelized cost of establishing capture plant and compression train in grid square j is shown in Equation (Fimbres Weihs et al., 2011). The total facilities cost associated with building all the capture plants and compression trains for the CCS network is shown in Equation (Krickenberger and Lubore, 1981).

The cost-optimal operating capacity for each transportation mode (i.e., pipeline with different NPS diameter) was obtained by performing a number of simulations of the techno-economic model developed in our work (Alhajaj and Shah, 2019) while assuming a minimum delivery pressure of 8.7 MPa, a maximum operating pressure of 14 MPa, a distance of 100 km and a pipeline cost data based on a regression model obtained for the south-eastern region of the USA (McCoy and Rubin, 2008). The levelized cost of transportation were used to find the optimal operating capacities, which considers the trade-offs in CAPEX and OPEX between larger pipeline dimeters with lower pressure drops and smaller pipeline dimeters with larger pressure drops and compression power. 2.1.1.2.3. Reduction target objective constraint. To meet a reduction target of capturing certain amount of CO2 emitted per year from all CO2 sources considered in the study, we have

∑ ckj ≥ RT

ckj ≥ 0 ∀ j

FCCj =

∑ TCCCkj ckj

(7)

TFCC =

∑ FCCj j

2.1.1.2.4. Non-negative constraints. All decision variables must be non-negative:

∀ j (11)

j

ajmax , total

(12)

2.1.1.3.2. Pipelines capital cost. The capital cost of establishing 7

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Fig. 7. Prospective network while meeting 95 % CO2 reduction target: 219 mm (8 in. pipeline transferring 3950 (tCO2/day) between S4 and S3; 324 mm (12 in. pipeline transferring 7900 (tCO2/day) between S3 and S2; 507 mm (20 in. pipeline transferring 14,778 (tCO2/day) between S2 and S1; 507 mm (20 in. pipeline transferring 16,948 (tCO2/day) between S1 and O1. Table 5 The cost-optimal capture rate for each selected source and injectivity in each potential reservoir. CO2 reduction target

S1 (DOC %)

S2 (DOC %)

S3 (DOC %)

S4 (DOC %)

O1 (Mt CO2/ year)

O2 (Mt CO2/ year)

O3 (Mt CO2/ year)

25 50 75 95

98 100 100 99

33a 80 84 98

0 26b 62 92

0 0 61 92

0 0 1.5 6.19

1.63 3.26 3.38 0

0 0 0 0

% % % %

a This was obtained by bypassing 53 % of the flue gas while capturing the reaming flue gas at 70 % DOC. b This was obtained by bypassing 58 % of the flue gas while capturing the reaming flue gas at 60 % DOC.

Fig. 8. Levelized cost of establishing CCS network that meet reduction targets at 25 %, 50 %, 75 % and 95 % out of the 4 CO2 sources announced in Abu Dhabi for phase 1 (Nader, 2009).

Table 6 Optimal design variables of capture plant and compression train attached to selected CO2 sources. Reduction target (%)

Sources

Absorber & DCC diameter (m)

Absorber height (m)

Blower power (kW)

Compressor after-cooler area (m2)

Compressor power (kW)

Condenser area (m2)

DCC height (m)

Lean amine cooler area (m2)

Reboiler area (m2)

Rich lean HE area (m2)

Stripper Diameter (m)

Stripper Height (m)

25

S1 S2 S1 S2 S3 S1 S2 S3 S4 S1 S2 S3 S4

— 10.6 — 15.4 9.5 — 15.4 14.5 14.5 — 15.4 14.5 14.5

— 16.3 — 20.6 16.4 — 22.9 17.1 16.8 — 35.6 36.5 36.5

— 1155 — 3096 950 — 3428 2309 2272 — 5302 4883 4883

822 962 835 2331 4934 835 2447 1103 1090 829 2857 1638 1638

9815 9594 9974 23246 4572 9974 24406 11002 10871 9895 28497 16333 16333

— 1625 — 3797 9348 — 3987 2090 2065 — 4803 3103 3103

— 10.3 — 10.3 7.9 — 10.3 7.9 7.9 — 10.4 8.1 8.1

— 2871 — 7030 1459 — 7385 3511 3469 — 9702 5212 5212

— 1470 — 3535 748 — 3711 1800 1778 — 4448 2672 2672

— 8168 — 19955 4097 — 20946 9860 9742 — 23428 14636 14636

— 7.1 — 10.3 5.3 — 10.3 8.1 8.1 — 10.3 8.1 8.1

— 16.4 — 17.1 15.9 — 17.3 16 16 — 18.3 17.9 17.9

50

75

95

8

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

pipelines given the uncertainty in reported cost data is obtained by first calculating the total distance being built of each mode as shown in Equation (Turk et al., 1987) then, the capital cost of establishing specific pipelines are obtained as in Equation (Akimoto et al., 2004) finally, the total capital cost incurred for building all the pipelines in the network is calculated in Equation (Alhajaj, 2008).

TDl =

portion of CO2 potential demand for EOR projects. This was calculated for three major oil fields (i.e., Asab, Bab, Bu Hasa) assuming a 17 % incremental in oil production while employing CO2 in miscible floods considering that 1 ton of CO2 is capable of producing 4 barrels of oils per day on average (Beecy and Kuuskraa, 2005). 3. Results and discussion

∑ dij Xlji (13)

i, j

CCl = Capexlcor TDl ∀ l TPCC =

∑ CCl l

The cost optimal operational and design variables including capture bypass ratio at varying degree of capture (DOC) (i.e., 25 %–99 %) were obtained by performing a number of optimizations of the 30 wt% MEAbased CO2 capture plant and compression train techno-economic model (Alhajaj et al., 2016) attached to all the sources in the case study. These optimal operational variables listed in Table 2 along with the data listed in Table 3 were used to simulate the capture plant compression train in order to obtain the levelized cost relationships against degree of capture which in turn is used as a meta-model input in the network model. The optimal design variables listed in Table 6 along with operating variables listed in Table 2 obtained at each degree of capture help us in minimizing the effects of uncertainties in these variables on the CCS network design. Other uncertainties such as the effects of fuel and steam price were not considered in this study. These can be looked at in more detail while developing the model to be probabilistic rather than deterministic, which can consider the optimal design under uncertainty. The cost-optimal operating capacity and velocity for each nominal pipeline size (NPS) are listed in Table 4. In order to account for the uncertainty in this reported data, compared to published work in the literature in which double the cost was reported (van der Zwaan et al., 2011), an uncertainty multiplying factor that represents the following increments in pipeline capital cost (i.e., +25 %, +50 %, +75 %, +100 %) was included. An operating cost of ($3100/km/year) was used in the study (Heddle et al., 2003). These results were then fed into the CO2 network model in which the decision variables are picked from sets of potential variables that should meet the above-mentioned constraints and the objective function. This was formulated as a mixed integer linear programming (MILP) problem and solved in the GAMS environment using the CPLEX solver. The results of running our multiscale model using 4 national reduction targets (i.e., 25 %, 50 %, 75 %, 95 %) that minimized the total levelized cost of the system were obtained including: the prospective network topologies shown in Figs. 4–7; the rate of capture and potential exploitation for EOR from selected sources and sinks shown in Table 5. It was observed that the uncertainty in the pipeline construction cost did not change the optimal results obtained for the entire chosen CO2 reduction targets. The results reveal the effects of having different CO2 reduction targets on the cost-optimal CO2 network topology. At a 25 % CO2 reduction target, a hub was deployed near at Al-Marfraq area in which the steel production plant was exploited completely in addition to CO2 coming from the boiler based power plant (i.e., the second highest CO2 content source) being exploited proportionally and then transported to the closest sink (i.e., Bab oil field). At a 50 % CO2 reduction target, a hub at Taweelah region was developed to transport the CO2 through a 324 mm (12 in. trunk line to the steel production plant in which a 406 mm (16 in. trunk line was built to transport the CO2 to the closest oil reservoir in Bab oil field. At a 75 % reduction target, a new hub was created at Bab oil field reaching its maximum utilization capacity in order to transport the excess CO2 to the closest oil field (i.e., Bu Hasa) using 219 mm (8 in. pipeline. At a 95 % reduction target, a hub was created at Al-Mafraq to transport all the CO2 to the sink with the highest utilization capacity using 507 mm (20 in) trunk line. The results highlight that Bu-Hasa was not exploited in this study due to the availability of a good utilization capacity in oil fields with a good proximity from the CO2 network. It is worth noting that ADNOC recently built a 218 mm (8 in. pipeline to transfer CO2 from steel plant to Bab oil field, which might not be the optimal choice for future expansions.

(14) (15)

2.1.1.3.3. Pipeline operating cost. Pipeline operating costs considering maintenance depend mainly on the length of the pipeline, as shown in Equation (Bakken and von Streng Velken, 2008), regardless of the pipe diameter. Other costs associated with compressions power to 14 MPa is accounted for in TFCC. It was also assumed that no further pressurization required as the distance between sources and sinks within 100 km. The current study assumes operating high-pressure pipelines from all sources. It is acknowledged that considering operating pressure as a decision variable in the optimization of CCS network will improve the results at the expense of increasing the complexity of the model being non-linear. The daily total operating cost for maintenance is outlined in Equation (Boavida et al., 2011).

OCl = Opexl TD ∀ l

TPOC =

∑ OCl l

(16) (17)

2.1.1.3.4. Overall objective function. The objective function is to minimize the total daily cost incurred in establishing and operating the CCS infrastructure as shown in Equation (Han and Lee, 2011a). Any generic high-performance MILP solver can be used to find the optimal decision variables that optimise the objective function in addition to satisfying the equality and inequality constraints obtained from equations (Bradshaw et al., 2004; van Bergen et al., 2004; Brunsvold et al., 2011; Jakobsen and Brunsvold, 2011; Jakobsen et al., 2011, 2008; Poyry Energy, 2007; Røkke et al., 2009; Forward, 2008; Fimbres Weihs and Wiley, 2012; Fimbres Weihs et al., 2011; Krickenberger and Lubore, 1981; Turk et al., 1987; Akimoto et al., 2004; Alhajaj, 2008; Bakken and von Streng Velken, 2008; Boavida et al., 2011; Han and Lee, 2011a).

min TC =

CRF (TPCC ) + TFCC + TPOC 365 CF

(18)

2.2. Case study The feasibility of multiscale approach was demonstrated within the region of the United Arab Emirates (UAE). As this region has the sixth highest carbon emissions per capita in the world, it has an ambitious primary target to reduce Abu Dhabi’s carbon footprint by a third through implementing large scale CO2 networks for Enhanced Oil Recovery (EOR) (Nader, 2009). The key drivers are the proximity of large CO2 sources and reservoirs and the benefit of releasing natural gas currently being used for EOR. The phase 1 CO2 reduction mandate announced in Abu Dhabi was to capture 5 million tons of CO2 from three different sources namely boiler flue gas from Taweelah power company (TAWEELA), combined cycle gas turbine (CCGT) flue gas from Emirates Aluminium plant (EMAL) and a pure stream of CO2 from Emirates Steel Industry (ESI), shown in Table 1 and Fig. 3 (Nader, 2009). The objective here is to find the cost-optimal deployment of the CO2 network while capturing different amounts of CO2 emitted from all the three CO2 sources (i.e., 25 %, 50 %, 75 %, 95 %) and satisfying a 9

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Another important observation of the results shown in Table 5 is that the cost-optimal degree of capture from the selected CO2 sources depends on site-specific factors in addition to the techno-economic interconnectedness between CCS components. Thus, the deployment of a compression plant at the steel plant with the highest degree of capture (i.e., 100 %) is expected as it has a pure stream of CO2 that only needs to be compressed and dried for transportation. However, a higher CO2 content in the boiler fired power plant compared to the CCGT does not mean that this source needs to be exploited completely as can be seen in the results in Table 5. These results highlight the complexity of deploying an interconnected network to meet a specific reduction target. It also emphasizes the need to explicitly account for the effect of optimal degree of CO2 capture from different sources as this is a site specific factor that is highly interconnected with the availability and the cost of pipeline networks and the specific cost of capture at each site. The detailed design variables listed in Table 6 highlight the usefulness of using a multiscale modelling approach as a design tool to find the optimal sizes of the major units of the CO2 capture plant and compression train including cost-optimal bypass ratio and DOC for the remaining flue gas. Further, optimal operating or control variables such as amine lean loading need to be sustained to ensure cost-optimal operation of the CO2 network. The flexibility in the level of detail outlined in the multi-scale modelling approach allows it to be expandable in that it can be modified to account for uncertainty and in that it can be developed to a dynamic model, which can be used to assess the operability and controllability of the system. Fig. 8 presents the total levelized cost of CO2 capture, compression and transport to a geological storage site that has a specific CO2 demand for EOR while assuming four reduction targets. The cost of capture and compression increases gradually with higher reduction targets because more expensive sources need to be connected to the system. This increases the total cost of the network as the cost of capture and compression contributes to more than 90 % of total cost. The cost of transportation, however, decreases with larger CO2 reduction targets as a result of aggregating CO2 flows into larger diameter pipelines and hence exploiting economies of scale. It is worth noting that the total cost of the CO2 network shown in Fig. 8 is lower than the CO2 capture and compression cost from CCGT plant alone. This is due to the low cost associated with capture and compression from the steel plant, which lowered the overall CO2 network cost for this case study.

of capturing CO2 from increasingly more expensive sources, which outweigh the benefits gained from aggregating CO2 flows into larger pipelines; the uncertainty in reported pipeline construction costs did not affect the optimal layout of the network because the lion’s share of the total CCS cost is associated with the capture plant and compression train. Further, the study emphasizes the need to explicitly account for the effect of degree of capture, and site-specific factors such as flue gas characteristics in planning future CO2 reduction targets. Higher CO2 content does not necessarily indicate that the source needs to be exploited completely. Other factors such as the location of the CO2 source plant in the path of a CO2 network is another important factor that may be overlooked by policy makers. The results of this study indicated optimal capture rates lower than the ones obtained by taking into account the economies of the capture plant alone. This is due to a higher marginal cost of transporting extra CO2 when capturing at higher level, which requires a larger diameter pipeline. Thus, it competes with benefits of capturing CO2 at the optimal rate considering the single CO2 source alone. This conclusion underscores the importance of a whole-system analysis of potential CO2 networks, and serves to highlight that plant level details should be used to inform the design of CO2 networks.

4. Conclusions

Akimoto, K., Kotsubo, H., Asami, T., Li, X., Uno, M., Tomoda, T., et al., 2004. Evaluation of carbon dioxide sequestration in Japan with a mathematical model. Energy 29 (9–10), 1537–1549. Alhajaj, A., 2008. Design and Analysis of CO2 Networks. Imperial College London, UK MSc Thesis. Alhajaj, A., Shah, N., 2019. Optimal Design of CO2 Transport System Under Uncertain Purity and Temperature. In submission. . 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 II. Identifying the cost-optimal control and design variables. Int. J. Greenh. Gas Control. 52, 331–343. Bailey, D.W., Feron, P.H.M., 2005. Capture post-combustion. Oil Gas Sci. and Technol. – Rev. IFP 60 (3), 461–474. Bakken, B.H., von Streng Velken, I., 2008. Linear models for optimization of infrastructure for CO2 capture and storage. IEEE Trans. Energy Convers. 23 (3). Beecy, D.J., Kuuskraa, V.A., et al., 2005. Basin strategies for linking CO2 enhanced oil recovery and storage of CO2 emissions. In: Rubin, E.S., Keith, D.W., Gilboy, C.F., Wilson, M., Morris, T., Gale, J. (Eds.), Greenhouse Gas Control Technologies 7. Elsevier Science Ltd, Oxford, pp. 351–359. Boavida, D., Carneiro, J.F., Ramírez, A., Martinez, R., Czernichowski-Lauriol, I., Tosato, G., et al., 2011. Integrated infrastructure for CO2 transport and storage in the west Mediterranean. Energy Procedia 4 (0), 2440–2447. Bradshaw, J., Allinson, G., Bradshaw, B.E., Nguyen, V., Rigg, A.J., Spencer, L., et al., 2004. Australia’s CO2 geological storage potential and matching of emission sources to potential sinks. Energy 29 (9–10), 1623–1631. Brunsvold, A., Jakobsen, J.P., Husebye, J., Kalinin, A., 2011. Case studies on CO2 transport infrastructure: optimization of pipeline network, effect of ownership, and political incentives. Energy Procedia 4, 3024–3031. Chapel, D., Ernest, J., Mariz, C.L., 1999. Recovery of CO2 From Flue Gases: Commercial

CRediT authorship contribution statement Ahmed Alhajaj: Methodology, Software, Writing - original draft, Formal analysis. Nilay Shah: Conceptualization, Supervision, Writing review & editing. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements We acknowledge financial support for this work from Masdar Institute, Khalifa University of Science and Technology under project RC2-2019-007. References

A detailed multiscale model that describes a complex CCS system using a series of interacting scale specific models has been developed to design and analyse the optimal deployment of a CO2 network. It takes into account the limits that physical material represented by a solvent poses in the performance of the capture plant and in the system as a whole. Thus, the optimal network design obtained for the whole system incorporates the potential interactions between models at different scales. This goal was achieved by analysing and identifying the key operating parameters of the system and then setting them at their optimal conditions while generating the data (e.g., levelized cost of capture and compression) needed for high-level analysis. This was also used to obtain the optimal operating, control and design variables including flue gas bypass ratio, degree of capture and unit sizes. The model was successfully demonstrated for a case study in the UAE that includes 4 major CO2 sources and three potential oil recovery sites amenable to CO2 flooding. Although the results obtained are specific to the case study, a number of general observations can be summarized including: the most promising CO2 sources for CCS implementation are the high CO2 content sources within a reasonable distance to sinks; the deployment of large scale CCS networks that meet high CO2 reduction mandate increases the CCS levelized cost as a result 10

International Journal of Greenhouse Gas Control 94 (2020) 102925

A. Alhajaj and N. Shah

Kvamsdal, H.M., Jordal, K., Bolland, O., 2007. A quantitative comparison of gas turbine cycles with capture. Energy 32 (1), 10–24. Lee, J.-U., Han, J.-H., Lee, I.-B., 2012. A multiobjective optimization approach for ccs infrastructure considering cost and environmental impact. Ind. Eng. Chem. Res. 51 (43), 14145–14157. 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 (4), 1883–1899. Mac Dowell, N., Pereira, F.E., Llovell, F., Blas, F.J., Adjiman, C.S., Jackson, G., et al., 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 (25), 8155–8168. McCoy, S.T., Rubin, E.S., 2008. An engineering-economic model of pipeline transport of CO2 with application to carbon capture and storage. Int. J. Greenh. Gas Control. 2 (2), 219–229. Middleton, R.S., Bielicki, J.M., 2009. A scalable infrastructure model for carbon capture and storage: SimCCS. Energy Policy 37 (3), 1052–1060. Middleton, R.S., Keating, G.N., Stauffer, P.H., Jordan, A.B., Viswanathan, H.S., Kang, Q.J., et al., 2012a. The cross-scale science of CO2 capture and storage: from pore scale to regional scale. Energy Environ. Sci. 5 (6), 7328–7345. Middleton, R.S., Kuby, M.J., Bielicki, J.M., 2012b. Generating candidate networks for optimization: the CO2 capture and storage optimization problem. Comput. Environ. Urban Syst. 36 (1), 18–29. Middleton, R.S., Keating, G.N., Viswanathan, H.S., Stauffer, P.H., Pawar, R.J., 2012c. Effects of geologic reservoir uncertainty on CO2 transport and storage infrastructure. Int. J. Greenh. Gas Control. 8 (0), 132–142. Nader, S., 2009. Paths to a low-carbon economy—the Masdar example. Energy Procedia 1 (1), 3951–3958. Poyry Energy, 2007. Analysis of Carbon Capture and Storage Cost-supply Curves for the UK. 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 (20), 4467–4475. Rodriguez, J., Mac Dowell, N., Llovell, F., Adjiman, C.S., Jackson, G., Galindo, A., 2012. Modelling the fluid phase behaviour of aqueous mixtures of multifunctional alkanolamines and carbon dioxide using transferable parameters with the SAFT-VR approach. Mol. Phys. 110 (11-12), 1325–1348. Røkke, P.E., Jakobsen, J.P., Tangen, G., Mølnvik, M.J., 2009. ECCO–European value chain for CO2. Energy Procedia 1 (1), 3893–3899. Strachan, N., Hoefnagels, R., Ramírez, A., van den Broek, M., Fidje, A., Espegren, K., et al., 2011. CCS in the North Sea region: a comparison on the cost-effectiveness of storing CO2 in the Utsira formation at regional and national scales. Int. J. Greenh. Gas Control. 5 (6), 1517–1532. Turk, G.A., Cobb, T.B., Jankowski, D.J., Wolsky, A.M., Sparrow, F.T., 1987. CO2 Transport: a new application of the assignment problem. Energy 12 (2), 123–130. van Bergen, F., Gale, J., Damen, K.J., Wildenborg, A.F.B., 2004. Worldwide selection of early opportunities for CO2-enhanced oil recovery and CO2-enhanced coal bed methane production. Energy 29 (9–10), 1611–1621. van den Broek, M., Brederode, E., Ramírez, A., Kramers, L., van der Kuip, M., Wildenborg, T., et al., 2009. An integrated GIS-MARKAL toolbox for designing a CO2 infrastructure network in the Netherlands. Energy Procedia 1 (1), 4071–4078. van den Broek, M., Ramírez, A., Groenenberg, H., Neele, F., Viebahn, P., Turkenburg, W., et al., 2010. Feasibility of storing CO2 in the Utsira formation as part of a long term Dutch CCS strategy: an evaluation based on a GIS/MARKAL toolbox. Int. J. Greenh. Gas Control. 4 (2), 351–366. van der Zwaan, B.C.C., Schoots, K., Rivera-Tinoco, R., Verbong, G.P.J., 2011. The cost of pipelining climate change mitigation: an overview of the economics of CH4, CO2 and H2 transportation. Appl. Energy 88 (11), 3821–3831.

Trends. Canadian Society of Chemical Engineers. Chapman, W.G., Gubbins, K.E., Jackson, G., Radosz, M., 1989. SAFT: Equation-of-state solution model for associating fluids. Fluid Phase Equilib. 52 (0), 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 (8), 1709–1721. Fimbres Weihs, G.A., Wiley, D.E., 2012. Steady-state design of CO2 pipeline networks for minimal cost per tonne of CO2 avoided. Int. J. Greenh. Gas Control. 8 (0), 150–168. Fimbres Weihs, G.A., Wiley, D.E., Ho, M., 2011. Steady-state optimisation of CCS pipeline networks for cases with multiple emission sources and injection sites: south-east Queensland case study. Energy Procedia 4 (0), 2748–2755. Forward, Yorkshire, 2008. A Carbon Capture and Storage Network for Yorkshire and Humber. Leeds. . 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 (2), 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 (10), 4168–4186. Han, J.-H., Lee, I.-B., 2012. Multiperiod stochastic optimization model for carbon capture and storage infrastructure under uncertainty in CO2 emissions, product prices, and operating costs. Ind. Eng. Chem. Res. 51 (35), 11445–11457. Han, J.-H., Lee, I.-B., 2011a. Development of a scalable and comprehensive infrastructure model for carbon dioxide utilization and disposal. Ind. Eng. Chem. Res. 50 (10), 6297–6315. Han, J.-H., Lee, I.-B., 2011b. Two-stage stochastic programming model for planning CO2 utilization and disposal infrastructure considering the uncertainty in the CO2 emission. Ind. Eng. Chem. Res. 50 (23), 13435–13443. Hasan, M.M.F., Boukouvala, F., First, E.L., Floudas, C.A., 2014. Nationwide, regional, and statewide CO2 capture, utilization, and sequestration supply chain network optimization. Ind. Eng. Chem. Res. 53 (18), 7489–7506. Hasan, M.M.F., First, E.L., Boukouvala, F., Floudas, C.A., 2015. A multi-scale framework for CO2 capture, utilization, and sequestration: CCUS and CCU. Comput. Chem. Eng. 81, 2–21. Heddle, G., Herzog, H., Klett, M., 2003. The Economics of CO2 Storage. Laboratory for Energy and the Environment, Cambridge, MA Report No.: No. LFEE 2003-003 RP Contract No.: No. LFEE 2003-003 RP. Iijima, M. (Ed.), 1998. A Feasible New Flue Gas CO2 Recovery Technology for Enhanced Oil Recovery. SPE/DOE Improved Oil Recovery Symposium. Jakobsen, J.P., Brunsvold, A., 2011. Development of scenarios for carbon capture and storage ECCO - European value chain for CO2. Energy Procedia 4 (0), 2677–2684. Jakobsen, J.P., Brunsvold, A., Husebye, J., Hognes, E.S., Myhrvold, T., Friis-Hansen, P., et al., 2011. Comprehensive assessment of CCS chains–Consistent and transparent methodology. Energy Procedia 4 (0), 2377–2384. Jakobsen, J.P., Tangen, G., Nordbø, Ø, Mølnvik, M.J., 2008. Methodology for CO2 chain analysis. Int. J. Greenh. Gas Control. 2 (4), 439–447. Keating, G.N., Middleton, R.S., Stauffer, P.H., Viswanathan, H.S., Letellier, B.C., Pasqualini, D., et al., 2010. Mesoscale carbon sequestration site screening and CCS infrastructure analysis†. Environ. Sci. Technol. 45 (1), 215–222. Kemp, A.G., Sola Kasim, A., 2010. A futuristic least-cost optimisation model of CO2 transportation and storage in the UK/UK Continental Shelf. Energy Policy 38 (7), 3652–3667. Klokk, Ø, Schreiner, P.F., Pagès-Bernaus, A., Tomasgard, A., 2010. Optimizing a CO2 value chain for the norwegian continental shelf. Energy Policy 38 (11), 6604–6614. Krickenberger, K.R., Lubore, S.H., 1981. Coal gasification: a source of carbon dioxide for enhanced oil recovery? Environ. Sci. Technol. 15 (12), 1418–1424. Kuby, M.J., Bielicki, J.M., Middleton, R.S., 2011a. Optimal spatial deployment of CO2 capture and storage given a price on carbon. Int. Reg. Sci. Rev. 34 (3), 285–305. Kuby, M.J., Middleton, R.S., Bielicki, J.M., 2011b. Analysis of cost savings from networking pipelines in CCS infrastructure systems. Energy Procedia 4 (0), 2808–2815.

11