A techno-economic case study of CO2 capture, transport and storage chain from a cement plant in Norway

To make carbon capture and storage (CCS) happen, clever strategies are needed for robust decision-making under uncertainties. This paper investigates various alternatives for the application of CCS at a cement plant in Norway. A matrix consisting of nine CCS chain designs was analysed in order to quantify and compare alternatives designed to reveal important facts and relationships that would provide a sound knowledge foundation for project prioritization and decision-making. An in-house techno-economic assessment methodology and tool, iCCS, was used to evaluate and compare the costs of different technologies and estimate their costcutting potential. In particular, the paper discusses the effects of the choice of capture technology and transport solution, the potential of EOR as a market maker, the effect of the quota-price incentive, and the need for a long-term governmental strategy for CCS, focusing specifically on the need for public support of CCS from a cement plant. Finally, the importance of considering potential socio-economic benefits in analyses of CCS project viability is highlighted and developed.


1.
Introduction There is currently a consensus about the role of carbon capture and storage (CCS) as a vital part of any greenhouse gas emission mitigation scenario [1] and the importance of CCS is widely acknowledged [2]. Furthermore, it has been shown that without CCS, reduction targets cannot be met and the costs of greenhouse gas emission mitigation measures would be more than doubled [3]. CCS technology is available and could be applied in the power generation sector as well as in industry. In spite of these positive developments with respect to global CCS attitude, the rate of CCS deployment has stalled, especially in Europe [2]. The deployment of planned CCS projects, in addition to the initiation of new ones, will require strong policy drivers.
Traditionally, the power generation industry was the most favoured candidate for CCS application. However, focus has recently also turned towards industrial sources of CO2. Cement and steel production are examples of industrial processes that will always produce CO2 as part of their production process. CO2 emissions from the cement industry represent 5% of global anthropogenic CO2 emissions [4]. In the case of cement production, CO2 is an unavoidable byproduct of the process and in order to significantly reduce the climate impact of cement production, CCS is unavoidable. Despite the moderate size of the CO2 stream resulting from a typical size cement plant, cement is often considered as an attractive candidate for CCS due to the rather high CO2 content of the flue gas, the stability of the stream over time and the availability of waste heat at the plant which could be utilized for CO2 capture. For these reasons, CCS from the cement industry has gained interest in Norway and throughout Europe. The focus on CCS from cement in Norway centres on the Norcem Brevik plant, that produces approximately 1.2 million tons of cement and 0.925 million tons of CO2 per year. The Zero Emissions Platform (ZEP) [5] concludes that the major reason hindering deployment of the CCS technologies is the lack of business case. In simple terms, a viable business is characterised by a cumulative income that exceeds its cumulative costs. Many actors will be involved in CCS unless one company owns and operates the whole CCS chain. This means that there is a significant counterparty risk associated with CCS that will need to be dealt with. Furthermore, several factors that are principally different from each other will affect the technical feasibility of the chain and its financial viability. Examples of such factors are commercial technology availability, logistics and transport infrastructure, business economics, global market economics, socio-economics, environmental sustainability, risk and safety, public acceptance, national and international policies and regulatory frameworks. It is obvious that the complex set of factors and the involvement of several actors are associated with many uncertainties. Indeed, uncertainty seems to be the largest obstacle to the development of a business case for CCS and therefore a key barrier for CCS deployment, because investors need to be confident of generating income of sufficient size and duration [6]. Clever and creative solutions for CCS business cases are urgently needed in order to encourage commercial deployment of the technology. In this paper, we investigate different alternative CCS chain designs suitable for application on a cement plant in Norway. We evaluate and compare the costs of different technologies and estimate their cost cutting potential. We also investigate the effects of various external parameters such as the CO2 quota price on the commercial viability of the chains. We further study the effect of enhanced oil recovery (EOR) on the overall economic viability of the CCS chain, in order to verify or disprove its proclaimed role as CCS business-makers [7]. We also identify specific needs for public support of CCS from a cement plant with particular focus on the socio-economic benefits.

2.
Case study design To investigate the issues mentioned above, a case study was designed. A matrix of different case alternatives and parameter variations was analysed with in-house techno-economic assessment tool iCCS developed within the BIGCCS research centre [8]. With respect to the chain design, focus is set on capture technology choice, type of transport, and type of storage. In addition to the analysis of various chain designs, several parameter sensitivity studies are performed, varying the process performance characteristics and costs of capture, the transport distance and investment models for transport infrastructure, as well as fluctuations in global parameters such as CO2 quota and oil price.

System characteristics
The Norcem Brevik cement plant emits a flue gas containing 17,8 %CO2,wet, with a total yearly emissions of 925 ktCO2/y [10]. In addition to the high CO2 concentration in the flue gas, the cement plant also presents the advantage of having excess heat available, which could be used for the CO2 capture unit. Finally, the plant is located by the fjord and already possesses a harbour facility to export the produced cement to customers by ship. Regarding possible CO2 storage locations, two options seem promising [11,12], as shown in Figure 1: • The Gassum formation, referred to as L1, which extends over large parts of the Skagerrak and south towards Denmark. While most aquifers and geological structures of the region have not been sufficiently qualified for CO2 storage, depths and reservoir characteristics enable the Gassum formation to emerge as the most attractive storage option in the area. The Gassum formation has been identified as the most promising option for storage of CO2 from point sources in eastern Norway. This storage site lies about 300 km from the Norcem cement plant and is therefore rather close; however, it is believed that this location would be suitable only for storage in an aquifer. • The Johansen formation in the northern part of the North Sea, referred to as L2, is a water-filled reservoir located 500-600 meters deep, under which a porous reservoir produces oil and gas from the Troll field. While there has been an extensive debate on most suitable CO2 storage sites between Johansen formation and the Utsira formation (south of Sleipner), especially for CO2 emissions from Mongstad and Karstø, the Johansen formation is felt to be the most costefficient solution. This option is around 730 km from the Norcem plant and offers opportunities for both CO2 storage only in an aquifer or combined with an Enhanced Oil Recovery project.

Chain design alternatives
Nine alternative CCS chain designs are analysed in order to quantify and compare their potential benefits and costs. The complete case matrix is summarized in Table 1.
In the chain A, considered as base case, CO2 emissions from the cement plant is captured using a MEA-based process. After the capture unit, the CO2 is liquefied and transported via ship to a saline aquifer at the L1 storage location. In the base case as well as in chains B to H, a partscale capture is considered where only 385 ktCO2/y (~42% of the plant emissions) are captured from the cement plant. This part-scale capture is applied in order to use the 31 MWth of waste heat available from the cement plant for the CO2 regeneration and limit the need for supplementary steam production from a natural gas boiler and associated CO2 emissions. As CO2 capture is a major component of the CCS chain costs, the first set of alternatives (chains B to D) investigates the potential benefits of a CO2 capture systems based on advanced solvent or polymeric membranes. In chain B, the capture technology is a solvent-based process, just as in chain A, but the reboiler steam duty associated with the solvent regeneration is assumed to be 2.1 GJ/tCO2 which is 33% lower than in the base case -chain A, assuming use of a novel less energy intensive solvent 1 . The next two alternatives, chains C and D, are based on membrane-based capture. Two possible types of membranes are considered. The Polaris polymeric membrane developed by MTR [13] is believed to be at an advanced stage of development and testing [14] and is referred to as the reference membrane in several studies [15]. The Fixed Site Carrier (FSC) membrane was developed by NTNU [16] and has been tested at Norcem Brevik cement plant [17]. The set of chain options E and F investigate the potential of alternative transport technologies and transport distances. A stand-alone CO2 pipeline transport and a shared pipeline transport leading to a shared saline aquifer for CO2 storage are investigated in respectively cases E and F 2 . The chain alternatives G and H investigate the potential benefits of CO2 EOR. In chain G, most of the CO2 is used for EOR production over 25 years, while the excess CO2 is stored in a nearby aquifer. In chain H, the CO2 is used for EOR production over the first 10 years while the excess CO2 and the CO2 that is to be stored after 10 years is stored in a nearby aquifer. Due to the small volume of CO2 considered, the EOR and saline aquifer storages are assumed to be shared in both G and H cases 3 . Finally, as the base case and the alternative chains B to H are based on partial-scale capture, chain I is used to consider the implications of capturing at full-scale from the cement plant (i.e. 85% of plant yearly emissions).

Technical assessment
The following sections describe technical elements of the individual capture, transport and storage alternatives.

MEA-based CO 2 capture [18, 19]
In the MEA-based process, as shown in Figure 2, the flue gas is fed to the absorber after being cooled and pumped, using blowers to overcome the pressure drops in the columns. In the absorber, the flue gas comes into contact with a MEA-based solution containing 30%wt of MEA. After absorption, the CO2 is recovered at the bottom of the column, chemically bound to the solvent, while the flue gas passes through a wash section to balance water and recover solvent carried out as droplets or vapour. The "CO2 -rich" solvent is removed from the bottom of the absorber, pumped and enters a hot-cold heat exchanger to be preheated (to 120°C) by the regenerated lean solvent, before entering the top of the stripper. Significant amounts of heat are required at the stripper reboiler to break the chemical bond between CO2 and the solvent, and maintain regeneration conditions in the column. The vapourised water from the top of the column is recovered in the condenser and fed back to the column, while the purified CO2 is sent through the conditioning process to reach the requirements for pipeline transport. The "lean" solvent recovered at the bottom of the column is pumped back to the top of the absorber through the hot-cold heat exchanger and a cooler is used to reach a lower solvent temperature that enhances the absorption process. It is worth noting that 31 MWth of waste heat is available at the cement plant and can be used to generate a significant proportion of the steam required by the stripper. Therefore it is assumed that stripper steam is produced from the available waste heat in priority, while a natural gasfired boiler for additional steam requirement. A summary of the solvent-based capture characteristics is presented in section 7.1.

Membrane based CO 2 capture [20-22]
A numerical version of the Attainable Region Approach developed and described in detail by SINTEF Energy Research [20][21][22] is used here. In practice, the numerical model optimises the membrane-based capture process, considering configurations of up to three stages, as shown in Figure 3, with the objective of minimising the overall investment and operating costs of the process. Based on the membrane properties and the system conditions being considered, the numerical model first generates the attainable region diagrams in order to select the ranges of stage feed and permeate purities relevant for each stage of the different multi-stage membrane process configurations. Based on the selected ranges of permeate purity, the cost-optimal designs of the one-, two-and three-stage configurations are identified and compared in order to identify the overall cost-optimal membrane configuration and design. Once the cost-optimal design has been set, the actual operating conditions (feed pressure, permeate pressure and area) are calculated from the targeted stage purity using the membrane model. A summary of the membrane-based capture characteristics is presented in section 7.1.  The design methodology considers a membrane model for binary components, after Saltonstall [23], and a membrane unit in cross-flow configuration with plug flow on the feed side and no mixing with the bulk stream on the permeate side. It is worth noting that this approach and the graphical solution generated are used to evaluate simple multi-stage configurations without advanced process features, such as retentate recycles or retentate heating before expansion.

CO 2 transport pipeline [24, 25]
To reach the pressure of 200 bar desired at the inlet of an offshore pipeline 4 , conditioning before pipeline transport is needed. This consists of compression stages and pumping, combined with the removal of unwanted components (dehydration) 5 . In order to assess the conditioning characteristics, simulations were performed under Aspen HYSYS ® using the Peng-Robinson thermodynamic property package. The process was modelled into four compression stages with intercooling followed by a pumping stage to reach 200 bar and with the characteristics given in Table 10. Before the offshore pipeline, a flexible riser transports the CO2 from the shore to the seabed. Depending on the water depth, the liquid head provides a 10 bar safety margin for the pressure drops in the injection network. Due to prohibitive subsea pumping costs, no reboosting is considered along the offshore pipeline, and the pressure drop must therefore be limited in order to maintain the outlet pressure above 60 bar [26], as shown in Figure 4. Here, 20 pipeline diameters ranging from 6 5/8" to 44" are considered. The offshore pipeline chain has different characteristics depending on the diameter: pressure drops, costs, etc. The pipeline designs are based on the minimum wall thickness required [27] and according to the American Petroleum Institute (API) specification 5L standard [28] with the characteristics shown in Table 10. The pressure drop is calculated using the Fanning equation, and does not take the effects of potential elevations into account [29]. A summary of the pipeline conditioning and transport chain design characteristics is presented in section 7.2.

CO 2 shipping [24, 25]
To obtain liquid CO2 at 6.5 bar and -50 °C, conditioning before shipping transport is required, and this consists of compression stages followed by a liquefaction process using ammonia cooling cycles, combined with the removal of unwanted components (dehydration) 6 . In order to assess the characteristics of the conditioning process, simulations were performed using Aspen HYSYS ® using the Peng-Robinson thermodynamic property package. The process was modelled as three compression stages followed by ammonia cooling, liquefaction by expansion and recycling of the remaining gaseous part of the stream [30] and with the characteristics given in Table 11. After liquefaction, cryogenic buffer storage facilities are required, as shipping involves batch export, while liquefaction and injection are continuous processes, as shown in Figure 5. Depending on the size of the vessel involved, the shipping chain has different characteristics: number of ships in the fleet, buffer storage capacity, fuel consumption, costs, etc. Here three vessel sizes are considered with the associated characteristics given in Table 12. It is here assumed that the volume of the buffer storage before export is equal to the ship's cargo volume [26]. For each ship size, the fuel consumption is assumed to be proportional to the distance and transport volume, and estimates are based on the figures of Roussanaly et al. [31]. At the offshore field, a cryogenic buffer storage is also required to match the batch exports involved in shipping and the injection, which is a continuous process. It is here assumed that the buffer will be provided by a vessel with the same cargo volume as the others in the fleet [32]. The on-board reconditioning of CO2 consists of repumping to 60 bar with heating, to ensure a temperature after reconditioning above 0°C. Even if frigories have an economic value on an industrial site, there is here almost no integration possibility 7 , and reconditioning investments and operating costs must therefore also be included in the shipping chain values. Simulations are performed using Aspen HYSYS ® to assess the characteristics of the on-ship reconditioning process. The electricity for on-ship reconditioning is produced by burning shipping fuel, with a conversion factor of 12,029 kWh/tfuel [33]. Finally, a flexible riser transports the CO2 from the ship to the seabed. Depending on the water depth, the liquid head will provide a 10 bar safety margin for pressure drops in the offloading hose and injection network. A summary of the shipping conditioning and transport chain design characteristics is presented in section 7.3.

CO 2 storage
The evaluations of CO2 storage are performed using the CO2 storage module of the iCCS value chain tool [7,8] based on the ZEP report on CO2 storage of CO2 storage in depleted oil and gas fields (DOGF) and deep saline aquifers (SA) [34] extended to include CO2-Ehanced Oil Recovery (EOR) storage [7]. The module includes four offshore storage pre-set cases depending on the type of storage and the well legacy for DOGF (reuse or new wells) as shown in Table 2.
Here, field characteristics of the ZEP medium cost scenario 8 are considered, while other parameters such as well depth, economic and climate data can be modified. In the EOR case, a medium EOR scenario that takes an additional oil recovery of 7% into account [7], is included.

Cost evaluation
This study attempts to include costs expected to be representative of a demonstration projects to be built in a near future even if the technology is not yet fully mature and demonstrated on the commercial scale level. Such estimates do not reflect the expected benefits of technological learning and therefore lead to costs higher than normally estimated in the literature which do not adequately take into account the greater costs that typically occur in the early stages of commercialisation [35].
Estimating cost data that are representative of demonstration projects has been, and remains, a challenging task. In this work, the increased investment costs associated with demonstration projects are performed in accordance with the National Energy Technology Laboratory (NETL) cost estimation guidelines [36,37].

3.2.1
Investment cost Various investment cost estimation methods are used: a general method for the CO2 capture processes and a more specific one for CO2 transport and storage. Investment and operating costs are given in 2014 prices. Instances where cost data not available directly in 2014 prices are be updated using the Chemical Engineering Plant Cost Index (CEPCI) [38] for investment costs while the utilities costs are corrected considering a yearly inflation of 2% [39].

CO 2 capture processes
A factor estimation method is used to estimate the investment costs of the process equipment. In the approach, the direct costs of equipment in suitably selected materials are evaluated using the Aspen Process Economic Analyzer ® , based on simulations based on the solvent process flow diagram or the numerical membrane model. It is worth noting that as the membrane numerical model requires a cost model to optimise the membrane design, the direct cost function for each of the equipment in the membrane process has been regressed using the Aspen Process Economic Analyzer ® and are presented in section 7.4. The investment cost of a given item of equipment is then calculated by multiplying the component's specific direct cost by an overall indirect cost factor specific to include the indirect cost 9 , the demonstration cost, the location cost and retrofit cost as shown in Table 3. It is important to note that MEA and membrane-based processes account for different levels of project contingencies due to differences in levels of maturity between the two technologies. The total investment cost of the capture process is then determined by summing the estimated investment cost for all components within defined system boundaries (Equation 1).

CO 2 transport
For pipeline transport, the pipeline investment costs are determined assuming a CAPEX for offshore pipeline of 75,000 €2010/inch/km 10 based on the EU FP7 CO2Europipe project [43]. This cost, adapted to a North-West European concept, is based on a maximum operating pressure of 200 bar for offshore transport. For the shipping transport, the process equipment costs are evaluated following the power plant and CO2 capture cost methodology, while the ships' investment costs are determined directly, using the total investment cost per ship [31], which is a function of its effective capacity, as shown in Table 4. For both pipeline and ship transport, the cost of the pipeline riser is based on a reference cost of 8.5 M€2008 11 for a 1MtCO2/y [44] and scaled up, assuming costs linear to the diameter 12 .

CO 2 storage
The CO2 storage costs consist of six components, which include all of the phases in the lifetime of the CO2 storage project: 1) Pre-FID, 2) Platform, 3) Injection wells, 4) Operating, 5) Monitoring, Measurement and Verification (MMV) 6) Close-down. Based on the ZEP cost methodology [34] combined with data from Holt et al. [46] to represent the costs and benefits associated with the CO2 EOR production, the CO2 storage module evaluates the detailed costs of the six components and provides the unit storage costs, including investment and operating costs. The unitary storage costs of offshore storage are shown in Table 5 for the medium cost scenarios and the four types of offshore storage. For the base case scenario, the cost of an offshore DOGF for the medium cost scenario is considered.

Operating costs
The operating costs are split into fixed and variable operating costs.

Fixed operating costs
The fixed operating cost depends on the investment cost and covers maintenance, insurance, and labour. The annual fixed operating cost is set at 6% of process units CAPEX [47]. The annual pipeline fixed operating costs are assumed to be a fixed yearly cost per kilometre, independent of the pipeline diameter and equal to 7,000 €2010/km/y [43]. The annual fixed operating cost per ship is a constant function of the ship size [31,45], as shown in Table 4.

Variable operating costs and revenues
The variable operating costs of the power plant and CCS chain are a function of the amount of CO2 captured, and cover consumption of utilities: electricity cost, steam, cooling water, MEA make-up, vessel fuel, and harbour fees. The annual variable operating costs are estimated using the utilities consumptions estimated by the iCCS tool and the utility costs given in Table 6.
However, additional operating costs can also be expected in the first years of operation, especially for a demonstration project due to learning and training time, inefficiency, and so on. Utility consumption is therefore assumed to be 15% higher than the basis during the first three years of operation [22]. As a principal purpose of this study is to calculate the cost of capturing, transporting and storing the CO2 for each case, no CO2 tax is taken into consideration.
In cases in which CO2 EOR is a possibility, the revenues associated with the oil production are included in the valuation of the CCS chain. As shown previously by Roussanaly and Grimstad [7], oil price to be considered in the valuation of the project is dependent on the cost and production profile of alternative methods to CO2 injection for EOR; two oil valuations 15 are therefore taken into consideration. The first one, equal to 60 €/barrel [49,50] 16 , is representative of cases in which alternative EOR methods rather than CO2 are not viable options. The second one, equal to 31€/barrel [7], is representative of cases in which chemical EOR is a viable and competitive method.

Greenhouse Gas assessment
To evaluate the chain cost per ton of CO2 avoided in a consistent way, it is necessary to include the carbon footprint of the chain. Here, only direct greenhouse gases emissions associated with the consumption of electricity, steam and shipping fuel are considered, as shown in Table 7.
The GHG emissions are converted into CO2 equivalents (CO2e) according to the IPCC guidelines [51] and summed. This sum indicates the potential climate effect and is often referred to as the global warming potential (GWP).  [52] 3.11 kg CO2e /kg oil 14 Low electricity cost representative of the Norcem Brevik cement plant. 15 The oil valuation considered is the oil barrel FOB (Free On Board) price in Rotterdam minus costs that are not included in the EOR production costs, such as normal production costs as well as transport from the oil platform (in the North Sea) to Rotterdam. 16 Corresponding to the average price of oil 90$ 2013 /barrel minus a production cost of 12$/barrel, which includes only the operating expenses associated with the oil production, as the investment part is normally covered by the normal production and the primary water injection recovery.

Key Performance Indicators for comparison of the options
The CO2 avoided cost [53] is here used as Key Performance Indicator (KPI) in order to compare the different chains. The CO2 avoided cost (€/tCO2,avoided) approximates the average discounted CO2 tax or quota over the duration of the project that would be required as income to match the net present value of additional capital and operating costs due to the CCS infrastructure. It is equal to the annualised costs divided by the annualised amount of CO2 avoided, as shown in equation (2). The annualised amount of CO2 avoided is defined as the amount of CO2 captured minus the direct emissions associated with the CCS infrastructure. The CO2 avoided cost is calculated assuming a real discount rate of 8% 17 , 7400 annual operating hours and an economic lifetime of 25 years [40]. In addition, investment costs consider that construction is shared over a three-year construction period [40].

Overall results
This section provides a brief summary of the results of the techno-economic analysis of the various chain alternatives, while specific issues arising through the different case variations are discussed in the following section. The CO2 avoided cost of the nine considered chain alternatives are summarised in Figure 6 while the detailed cost breakdown is presented in section 0. The results show that the full-chain cost of the base case is estimated to be around 120 €/tCO2,avoided, while the alternative chains lead to costs that lie between 100 and 150 €/tCO2,avoided. Although these values may appear high compared to conventional figures available in the literature, it is important to remember that these numbers include additional costs representative of demonstration projects and the maturity of the current technology, the remote location of the plant, retrofit projects, and small CO2 volumes (approximately 400 ktCO2/y). In the base case, CO2 capture and conditioning represent around half of the CO2 avoided cost, while transport and storage respectively represent 25 and 20% of the cost. These proportions vary from case to case, depending on the characteristics of the chain considered. These nine chains are used to discuss four topic of relevance to CCS deployment and the case under consideration: • The potential costs and benefits from capture technology improvement and secondgeneration technologies through comparison of the chains "improved amine", "high permeance membrane", and "high selectivity membrane" with the base case; • Opportunities for reduction of the transport cost through the comparison of the chains "stand-alone pipeline" and "shared pipeline" with the base case; • The potential value creation benefit of combining the CCS chain with CO2 EOR through comparison of the chains "EOR 25y" and "EOR 10y"; • The potential benefit of having a CCS project portfolio rather than a "full-scale" chain.

Cost-reduction potential from capture technology improvement and secondgeneration technologies
CO2 capture is a major cost component of the CCS chain, and the improvement of existing technologies and development of 2 nd and 3 rd generation of capture technologies has been focal points of research and development throughout the past 20 years [54][55][56]. Here, the potential of both of these cost-reduction approaches have been evaluated through three chain alternatives: 1) solvent-based capture with an improved amine decreasing the steam requirement by 33% compared to MEA 2) membrane-based capture based on the high-permeance developed by MTR [13] 3) membrane-based capture based on the high-selectivity membrane developed by NTNU [16]. The results presented in Figure 7 show that these three options have the potential to significantly reduce the cost of CO2 capture from the cement plant. The solvent process based on improved amine with a steam requirement of 2.1 GJ/tCO2 has the potential to reduce costs by 40% compared to the base case. This significant cost reduction is primarily due to the cost savings directly associated with reduced steam consumption. However, the reduction in the CO2 emissions from steam consumption, which impacts both the capture costs and the overall chain costs, account for almost one fourth of the cost reduction. In the case of the membrane-based capture processes, the potential cost reduction might be also 40% compared to the base case. This is mainly due to the low cost of electricity at the cement plant. However, there are rather significant uncertainties associated with the estimated costs of the membrane module, as the literature often assumes a unitary cost of 40 €2014/m 2 irrespective of the properties of the membrane, and does not take the initial development cost into account [22]. Therefore subcases with membrane module cost of 80 €2014/m 2 are also presented in Figure 7 for comparison. In these cases, the cost advantage of the membrane-based processes is reduced to 30% compared to the MEA process of the base case. Both the improved amine-based and the membrane-based technology show a strong potential for reducing the cost of CO2 capture from the cement factory. Membranes for post-combustion CO2 separation are not yet commercially available and therefore represent a risk that is still too high in a short-term perspective, although it has great potential in the long run. On the other  hand, even though few improved solvents are commercially available, it is important to note that their performances are lower than the improved amine case considered here, and further development is also required to reach the full potential of improved amine-based capture. Figure 7: CO2 captured costs for the base case and the capture technology alternatives

Opportunities for reduction in transport cost
Two alternatives are possible to transport CO2 to an offshore location: pipeline and shipping. Offshore pipeline transport has often been regarded as the most advantageous, due to its costeffectiveness for large volumes and as it has the advantage of being a continuous process with opportunities for automation, but it requires large initial investments. On the other hand, CO2 shipping is less capital-intensive and more flexible. However, it is less cost-effective for short distances and large volumes. While transport is often regarded as a minor cost component of the CCS chain, higher costs than expected are incurred in this case, due to the small volumes involved as shown in Figure 8. Therefore CO2 shipping transport considered in the base case is compared to pipeline transport, both on a stand-alone basis and on a common shared infrastructure basis. Comparison of the stand-alone pipeline transport with the shipping base case indicates an increase of 45% in the cost of CO2 conditioning and transport. Indeed, even if the conditioning costs are lower in the case of pipeline transport [31], pipeline CO2 transport is an extremely cost-inefficient method for transport of small volumes over long distances, as shown in Figure 9 and as documented in the literature [25,27]. However when considering a common pipeline infrastructure shared with two other actors of similar size to the cement plant base case, the costs of transport and conditioning allocated to the cement plant are reduced by 50% compared to the stand-alone case, and thus lead to conditioning and transport costs 30% lower than in the base case (i.e using shipping). These results show that a common pipeline infrastructure and, even more, a pipeline network have the potential to significantly lower the cost of CCS for CO2 sources and sinks located in the Skagerrak region due to the small volumes involved. In theory, such an infrastructure could be beneficial for the Norcem case and could be possible as two other high-potential cases for Norwegian full-chain CCS demonstration (Yara's fertilizer production facility and Klemetsrud waste management and energy recovery plant) are located less than 200 km from the Norcem site. However, there are several significant obstacles to this solution, as it would require joint interest in such an infrastructure, guaranties from several actors, deployment coordination, high investment and therefore investment risk. Without a strong involvement and national financial support, the deployment of such an infrastructure would be difficult to put in place [32,57] and CO2 transport via shipping seems to be a more realistic solution. In addition, shipping transport will become even more cost-effective than pipeline transport in the case of transport to CO2 storage sites in the North Sea.

CO 2 EOR an opportunity for CCS?
CO2 capture, transport and storage in connection with Enhanced Oil Recovery (EOR) is often considered to be a promising way to ensure cost-efficient avoidance of CO2 emissions to atmosphere [1]. The potential cost benefits of CO2 storage associated with EOR are evaluated and compared to the base case. In the chains associated with EOR, the storage considered is located in the North Sea (Storage L2)  Pipeline annual flow (MtCO 2 /y) CO2 transport cost Pipeline external diameter (beyond testing), it is assumed that the CO2 EOR storage is used for two other sources of the same size as the emissions stored from the cement plant. The costs of the two transport and storage cases in which the EOR storage can be used for respectively 25 and 10 years, while a saline aquifer is used for the excess CO2, are assessed and compared in Figure 10. The results of the assessment show that significant cost reductions can be achieved if CCS projects can be coupled with CO2 EOR storage under the oil prices considered. Indeed, the costs of CO2 transport and storage can be reduced by 35 and 28% respectively in the cases of 25 and 10 years of EOR storage. However, it is important to note that the viability of CO2 EOR storage is highly dependent on the oil value considered. Indeed, as shown by Roussanaly and Grimstad [7], the oil value considered for the valuation of a CCS project with CO2 EOR storage can be significantly lower than the oil price depending not only on production costs but also on the performances of alternative EOR methods for the oil field in question. As the oil value considered here takes in account both these issues, it is important to consider the influence of the oil price on the valuation of the CO2 transport and storage cost of CCS chain with EOR considered. Therefore two additional sets of sub-cases with oil values of 66 and 14 €/bbl [58,59] 18 are presented in Figure 11.
The results show that, in both EOR cases, an oil value for the CCS project of 66 €/bbl would lead to halving the CO2 transport and storage cost compared to the reference oil value. In this case, the CO2 EOR part of the chain can fully overcome the storage costs and even halve the transport cost. However, at an oil value of 14 €/bbl, the transport and storage costs increase to 53 €/tCO2,avoided with both EOR chains. In this case, which reflects closely the low oil prices at the end of 2015 and beginning of 2016, CO2 EOR storage leads to a very limited improvement in transport and storage costs compared to the base case in which CO2 is only stored in a saline aquifer in the Skagerrak region. These results confirm that CO2 EOR has significant potential for reduction of the cost of CCS; however both the length of time period during which CO2 can be injected in the EOR reservoir and the oil value have significant impacts on the profitability of this storage alternative.

Quota price and CCS impact on the cost of cement
The need to develop an economic and financial framework to discourage high carbon options and reduce uncertainty is often highlighted by actors involved in enabling and implementing low carbon options [60]. This issue is of special importance here due to the high CO2 avoidance cost obtained in all of the cases considered and more generally when considering the impact of CO2 capture from a cement plant. This means that a high CO2 tax/quota or a high level of public financial support is required to compensate for the additional costs associated with CO2 capture, transport and storage. Indeed, as shown in Figure 12, cement is a rather cheap material to produce (around 50€/tcement 19 ) with a large climate impact (0.66 tCO2/tcement) [61]. In practice, the inclusion of the CO2 capture and conditioning cost in the cement production cost leads to an increase of 70% in the cost of cement production, in the base case, which would have a significant impact on the competitiveness of the cement produced at the Norcem plant. This increase, which does not take transport and storage costs into account, would exceed the producer's margin rendering the cement production with CCS unprofitable at the current market price [62] unless it is compensated for. If CCS is to be implemented at Norcem, public financial support will therefore be required to overcome the additional cost associated with the implementation of CO2 capture and conditioning at the cement plant. This financial support required will depend on the capture technology employed, ranging from 40 €/tCO2,avoided with the improved amine and membrane based capture to 60 €/tCO2,avoided with the conventional MEA process.

Socioeconomic aspects
While full-scale CCS is often referred as the goal for demonstration and implementation of the CCS technology due to potential economies of scale, it might not be the most suitable target for all cases of demonstration or implementation. Indeed, as shown in Figure 13, in the case discussed here, a full-scale implementation might reduce the CO2 avoided cost by only 6% 19 The cement cost used here is a generic European value and is not specific to the Norcem cement plant. compared to the base case in which partial capture is considered. Even if the whole chain benefits from the economies of scales associated with the higher CCS capacity, the CO2 capture cost increases significantly in the full-scale case. While a large proportion of the steam required by the capture unit can be produced at a low cost from the cement plant's waste heat in the base case, the capture cost increases in the full-scale case as the additional steam required will be produced by a natural gas boiler at full cost. This significantly offsets the economies of scale in the full-scale case.
This limited reduction in the CO2 avoided cost, while the overall CCS cost increases from 450 to 830 M€, underlines the fact that full-scale implementation based on the solvent technology would not be suitable in this case, unless it was supported by significant financial incentives (CO2 tax or quota, or public-sector financial support). If financial support from the government is available above the support required for the base case, support for further demonstration of other capture technologies, such as membrane, at the Norcem, site would be preferable to fullscale solvent based implementation. Alternatively, potential national CCS financial support available above the cost of Norcem base case, could also be used to support some of the other high potential cases for Norwegian full-chain CCS demonstration: the Klemetsrud waste management and energy recovery plant near Oslo, Yara's fertiliser production facility near Porsgrunn, and the Longyearbyen coal-fired power plant on Svalbard. In both cases, these potential "additional" available financial supports would help to demonstrate additional CCS technologies than in the Norcem base case and further qualify CCS as a safe and cost-effective option for reducing CO2 emissions from industry and the electricity generation sector.

Conclusions
To make carbon capture and storage happen, clever strategies are needed for robust decision making under uncertainties. The iCCS methodology and tool were developed in order to reveal important facts and relationships that would provide a sound knowledge foundation for project prioritization and decision-making. The methodology and tool were applied to the NORCEM Brevik cement plant case study in order to assess and understand the following issues: effects of the choice of capture technology, the potential of EOR as a market maker, effect of the quota- Capture Conditioning Transport Storage Overall CCS cost price incentive, and the need for a long-term governmental strategy for CCS. The conclusions drawn, based on the results obtained for the set of nine alternative chains analysed, can be summarized as follows: • The assumed reduced energy demand for solvent regeneration by developing improved solvents has the potential to reduce the capture cost by 40%. However, there may be thermodynamic constraints that will not allow further significant energy reduction. • Membranes appear to have great potential for capture cost reduction in the same range, but are currently associated with significant risks due to the lack of maturity and uncertainties regarding membrane properties and costs. • EOR can improve the economics of the whole CCS chain but the cost versus revenues balance is very sensitive to the oil price and depends on the availability, efficiency, and cost of alternative EOR technologies. • The CO2 quota price level required to cover the cost of the CO2 capture (not including transport and storage) implementation on the cement plant analysed here is around 60 €/tCO2. A more stringent CO2 quota policy is required. However, by itself, this is unlikely to be a sufficient measure to kick-start the implementation of CCS in the cement production industry. Additional incentives will also be required to ensure the viability of CCS from cement production: • Public investment in shared infrastructure and storage.
• Rewarding mechanisms for first movers, CO2 EOR, or other initiatives. • The development of a long-term strategic vision for CCS financing based on socioeconomic aspects and potential benefits might introduce new perspectives. • Innovative business models on the crossroad between the public and private sectors will have to be developed in order to help close the cost gap and align commercial and societal interests across the CCS chain.

CO 2 capture technologies
The characteristics of the solvent-based captures for chains A and B are presented in Table 8, while the characteristics of the membrane-based captures for chains C and D are presented in Table 9.      20 In order to take into account the total length of the pipeline (including t-junctions, terrain factors, etc.), its length is assumed to be 10% longer than the transport distance. 21 As the pipeline is mainly located at the bottom of the sea where the water is at 4°C, the CO 2 will rapidly cool and will be more dense than at ambient temperature. 22 Assuming mooring/loading/departure and mooring/unloading/departure durations of 12 h each. 23 Equivalent to 25.9 km/h. 24  A direct costs function for carbon steel equipment has been regressed for the membrane-based capture process using the Aspen Process Economic Analyzer ® (see Table 13), based on simulations performed using the numerical membrane model. However, due to their specificity, the CO2 membrane module and framework costs are estimated in a different way. The membrane module is estimated on the basis of the 50 $2010/m 2 cost adopted by Zhai and Rubin [15]. The membrane framework is based on the cost function suggested by van der Sluijs et al. [63] for the framework of the membrane separation system in an ammonia plant of DSM, and modified by Roussanaly et al. [21] to take the influence of the design pressure of the module into account, as shown in equation 1 and Table 14 25 .
Direct cost membrane framework = �    [63] 25 It is worth noting that a limit of 25,000 m 2 of membrane area per module is used in order to avoid having unrealistically large modules.