European Scenarios of CO 2 Infrastructure Investment

Based on a review of the current state of the Carbon Capture, Transport and Storage (CCTS) technology, this paper analyzes the layout and costs of a potential CO 2 infrastructure in Europe at the horizon of 2050. We apply the mixed-integer model CCTS-Mod to compute a CCTS infrastructure network for Europe, examining the effects of different CO 2 price paths with different regional foci. Scenarios assuming low CO 2 certiﬁcate prices lead to hardly any CCTS development in Europe. The iron and steel sector starts deployment once the CO 2 certiﬁcate prices exceed 50 € /tCO 2 . The cement sector starts investing at a threshold of 75 € /tCO 2 , followed by the electricity sector when prices exceed 100 € /tCO 2 . The degree of CCTS deployment is found to be more sensitive to variable costs of CO 2 capture than to investment costs. Additional revenues generated from utilizing CO 2 for enhanced oil recovery (CO 2 -EOR) in the North Sea would lead to an earlier adoption of CCTS, independent of the CO 2 certificate price; this case may become especially relevant for the UK, Norway and the Netherlands. However, scattered CCTS deployment increases unit cost of transport and storage infrastructure by 30% or more.


1.INTRODUCTION
Carbon Capture, Transport, and Storage (CCTS) was originally seen as a central element for decarbonized electricity systems, worldwide (e.g.IEA, 2010).The International Energy Agency (IEA) consequently underlined its importance with a 20% contribution to achieving emission reduction goals and 40% cost increase for decarbonization in its absence (IEA, 2012).Even higher cost increases of 29-297% were estimated by the IPCC (2014) for reaching the 2ЊC target.Estimates for the European energy system projected 77 GW (IEA, 2012) to 108 GW (EC, 2011) of power generation capacity, respectively, to be equipped with CCTS and a CO 2 transport network of over 20,000 km by 2050 (JRC, 2011).The reality, however, is in stark contrast to these expectations, as documented in a special issue by Gale et al. (2015) commemorating the 10 th anniversary of the IPCC (2005) special report.Not a single full-scale CCTS project with long-term geological storage has yet been realized worldwide (GCCSI, 2014).At the same time, CO 2 transport infrastructure projects have been removed from the list of critical infrastructure projects of the EU (EC, 2013a).Furthermore, the London Protocol still prohibits the movement of CO 2 across marine borders for 6.The UK has introduced an Electricity Market Reform (The Parliament of Great Britain, 2013), where one of the four pillars builds on EPS benchmarked against gas-fired electricity generation,; similarly, the US (EPA, 2012, final rule pending for submission to the federal register since 05.08.2015) and Canada (Parliament of Canada, 2012) have introduced EPS for new electricity generation units.
annual amount of CO 2 emissions per installed unit of generation capacity and thereby the operation of new coal power plants without CO 2 capture. 6Using captured CO 2 for EOR purposes contradicts the idea of long-term geological storage but significantly improves the cost effectiveness of a CCTS project.Successful projects like Boundary Dam in Saskatchewan, Canada (in operation since October 2014), as well as the majority of upcoming projects in 2016-17 (e.g.Kemper County Energy Facility and Petra Nova Carbon Capture Project in the US) are associated with CO 2 -EOR.Little progress, however, is seen in the EU as only a few riparian states of the North Sea are capable of CO 2 -EOR projects.Nevertheless, the EU framework for climate and energy still aims at a commercial CCTS deployment by the middle of the next decade (EC, 2014).
In this paper we present a model analysis and interpretation of the potential role of CCTS to support the EU energy system transition to meet the emission reductions goals that are consistent with an international goal of staying below 2ЊC of global warming.Our hypothesis is that CCTScontrary to the dominant belief until recently-will at best be a niche technology applied in regions with highly conducive conditions, e.g.parts of the North Sea, but that due to its cost disadvantages and recent setbacks in many EU countries, will not contribute significantly to overall EU decarbonization.Moreover, the discrepancy between locations of CO 2 emissions and the availability of potential CO 2 storage sites call for regional cooperation, which is vital to the economies of scale associated with CO 2 transport infrastructure.Few papers currently address this issue (e.g.Geske et al., 2015) or try to find solutions to the resulting coordination game (e.g.Massol et al., 2015).We quantify scale economies of different CCTS network coverages.
Section 2 provides a non-technical description of our CCTS-Mod; a multi-period, scalable, mixed integer framework calculating beneficial investments in the CO 2 -chain (capture, transport, storage).Section 3 presents the outcome of the European-wide results.We find no role for CCTS in the 40% mitigation scenarios.In the 80% mitigation scenarios, some CO 2 -intensive industries might start to abate, followed by the energy sector at a high CO 2 price (above 100 €/tCO 2 ).We consider this scenario unlikely, because most of the countries involved have already given up CCTS as a mitigation option, e.g.Germany, Poland, France, and Belgium.Section 4 focuses on an alternative driver for CO 2 -abatement through CO 2 -EOR.We find that for North Sea riparian countries that have not given up on CO 2 capture, mainly the UK and Norway, the use of CO 2 -EOR might be an economical option, depending on oil prices and prices of CO 2 certificates.Once CO 2 -EOR resources are fully exploited, further CO 2 capture activity is solely incentivized by CO 2 certificate prices, which must cover at least the variable costs but also potential new investment costs.Also, the speed and extend of the deployment is highly dependent on assumptions for initial technology costs and learning effects.Section 5 concludes by analyzing the chances for a regional vs. Europeanwide CCTS application depending on the availability of CO 2 -EOR and other storage potentials.

The Model CCTS-Mod
For our numerical analysis, we use the "CCTS-Mod" (Oei et al., 2014).The model is a multi-period, scalable, mixed integer model coded in GAMS (General Algebraic Modeling Soft-7.The exogenous CO 2 certificate allows for no impact of CCTS investments on the certificate price.This limitation is necessary as such an analysis is currently only possible if technical restrictions-like non-linearities-are not considered (see e.g.Kanudia, et al., 2013).ware) and solved with a CPLEX solver.For each power plant or industrial facility covered in our input database (see section 2.2), an omniscient planner decides on whether to invest into a CCTS chain or to buy CO 2 certificates.The model decides in favor of CCTS whenever the net present value of CO 2 certificates required to cover emissions during the model horizon ( 2055) is higher than the net present value of all costs related to CCTS.
In this case, investments into a capture unit facing respective capital costs have to be made.It takes five years after the investment decision before the capture unit becomes operational.Whenever a facility is used to capture CO 2 , variable costs are induced.The capture rate is capped at 90%.CO 2 capture has to be balanced with CO 2 transport and storage.Again, respective infrastructure investments have to be made taking into account a construction period of five years.Capital costs for transport cover right of way (ROW) costs and other investment cost parameters.If a new pipeline is constructed along a route that is already developed, ROW costs do not apply.This ensures that transportation routes are bundled in corridors, which is consistent with practices for laying natural gas or crude oil pipelines.The construction of a pipeline is a binary decision with discrete pipeline diameters and associated throughput volumes.CO 2 storage is again subject to a five year construction period and has associated variable and capital costs.
A refined version of the model which is used for the model runs of this paper includes the option to use captured CO 2 for enhanced oil recovery.CO 2 -EOR storage is associated with additional investment and variable costs for equipment and operation, but generates revenue from oil recovered with each ton of CO 2 stored.A simplified decision path of the CCTS-Mod is illustrated in Figure 1.The model is based on literature on CCTS infrastructure models, developed by Middleton and Bielicki (2009) and Morbee et al. (2012).
The motivation behind the model is to provide an estimate on overall costs and needed infrastructure of a possible CCTS roll-out in Europe.Such figures are important from a planner's perspective and serve as bottom line for future cost estimates.The main drivers of the model are location and volumes of CO 2 emissions, storage capacities, investment and variable costs at each stage of the CCTS technology chain, and assumptions about future CO 2 certificate and oil prices.The development of the CO 2 price is thus exogenously given, independent from the CCTS deployment allowing for better comparison between the scenarios. 7Several uncertainties persist regarding the model: First, the cost-minimizing approach from a planner's perspective likely underestimates the actual costs of CCTS technology, as we assume perfect foresight, no market power by individual companies or sectors as well as a vertically integrated CCTS chain.Second, the model assumes the existence of certain technologies that have not been proven to work in practice on larger scales.The "cost" estimates for CO 2 capture and storage are especially uncertain, and most likely highly underestimated.The model also does not take into account the transaction costs of bringing the immature technology to implementation, to build infrastructure or to develop the storage sites.We also do not include costs associated with rising public opposition.However, total cost figures might be also overestimated, due to the limited utilization horizon of the infrastructure.

Mathematical Formulation
We define the objective function to be minimized as follows: The model is restricted by:

Pa Pa Pa
Equation 5 defines that a facility's CO 2 stream can be treated in two ways, or a mixture of them: CO 2 emissions can either be balanced with CO 2 certificates (zPa), or the CO 2 can be cycled through a capture system (xPa).
∑ ∑ ∑ ∑ Equation 6 specifies the physical balance condition, which states that all flows feeding into a node j must be discharged from the same node.
x ≤ (inv_x ) ∀P,a The capturing capacity of each producer P in period a is given in equation 7. Note that all terms in this inequality are decision variables, meaning that injection in period a is possible only if capacity was expanded prior to period a.
The capacity restriction of pipelines in Equation 8 works similarly to Equation 7.
∑ Sa Sb b Ͻ a Inequality 9 states that the annual injection rate of a storage facility S is limited to the sum of investments in annual injection capacity inv_y Sb from previous periods b.
Inequality 10 restricts the amount of CO 2 injected into reservoir S to its overall physical capacity.The multiplication by 5 resembles the amount of years per period a. Planing, licensing, and optimal routing of pipelines is ensured via Equation 11 where max_pipe is the maximum number of pipelines that can be built on a licensed route.
The model is formulated as Mixed-Integer-Linear-Program (MIP).It is solved with the General Algebraic Modeling System (GAMS) using the CPLEX solver.An optimality gap of 3% is applied in all scenarios.

European Data Set
Data was collected for the period between 2015 and 2055 8 and is based on a more detailed description of the cost data in Mendelevitch (2014).The scope of this study includes all members of the EU as well as Switzerland and Norway, and their respective Exclusive Economic Zones (EEZs).Data on location and emission volumes of refineries, steel and cement production facilities as well as coal-and gas-fired power plants is taken from a database developed earlier in Oei et al. (2014).
The database assumes an economic lifetime of 40 years for gas-fired and 50 years for coalfired power plants.Power generation facilities are supposed to be shut down without replacement after the economic lifetime is reached while industrial plants are assumed to be replaced by facilities with similar characteristics.It is unclear how emissions from industrial facilities and power plants will evolve in the future.In the electricity sector, a reduction of emissions is very likely due to rising shares of renewable energy sources.The model therefore includes projections for the closure of existing plants as well as the construction of new conventional power sources resulting in a reduction of overall CO 2 emissions from the electricity sector.In the industry sector a reduction of emissions due to efficiency improvements, increased recycling shares, alternative industrial processes as well as relocation of industries away from Europe are to be taken into account.A lack of source specific data for the industry including e.g.information regarding construction date or planned new sites and relocations, however, make such estimates very complicated.The model therefore assumes a continuous flat profile in the industry sector that overestimates the potential for CCTS in the industry.The model hereby assumes that old industrial facilities are replaced with new ones at the same location, not taking into account strategic relocation due to CCTS costs at other locations.This underestimates the application of CCTS in the industry sector.We assume this latter effect to be relatively small as the CO 2 capture costs are the biggest share of costs and independent of the source location.
The same database was used for location and capacities of potential storage in depleted hydrocarbon fields and saline aquifers.Data on location and volumes of CO 2 -EOR storage sites is taken from Mendelevitch (2014).Figure 2 illustrates the distribution of emission sources and their respective emission volumes for 2010 as well as the distribution of storage sites by type and their respective capacities.It depicts the fact that emission sources and storage sites are not equally spread across Europe.While the largest emission sources are located in the Rhine Area, the largest storage capacities can be found offshore in the UK and Norwegian EEZs. 9Denmark, UK and Norway are the only countries that have potential for CO 2 -EOR in their parts of the North Sea.Strong opposition in several European countries have formed against onshore CO 2 storage.All scenarios in this paper only include the option of storing the CO 2 in offshore fields.

Assumptions for All Scenarios
Two key parameters drive the results of our model runs: On the one hand CCTS deployment is triggered by the CO 2 certificate price path which governs the profitability of the CCTS  Source: Knopf et al. (2013).
technology in comparison to balancing CO 2 emissions with purchased CO 2 certificates.If in the long run, anticipated prices are higher than the costs of using the technology chain, then CCTS is employed.We use two possible price pathways generated by the PRIMES model (EC, 2013b) which represent the outcomes of two sets of scenarios for climate change mitigation policy up to 2050 (see Table 1).The 40% scenarios include the EU 2020 targets as well as a 40% greenhouse gas (GHG) reduction by 2050 compared to 1990.The 80% scenarios are more ambitious including an 80% GHG reduction by 2050.All scenarios do not allow for emission trading across macro regions, i.e. no trade between EU-ETS and RGGI or other schemes (but trade within macro regions, e.g.within the EU through a cap and trade system).They include moderate assumptions on efficiency gains and availability of nuclear and renewable energies (see Holz andvon Hirschausen, 2013 andKnopf et al., 2013 for a detailed description of the underlying assumptions).The availability of storage capacity is the second decisive parameter.France, Germany and Belgium, in particular have their storage resources located mostly in onshore saline aquifers and depleted hydrocarbon fields.However, onshore storage is associated with significantly higher complexity of regulation and a higher number of stakeholders involved.The Global CCS Institute has performed a comprehensive assessment of CO 2 storage readiness on a country level and come to the conclusion that Norway is the only European country currently ready for a wide-scale CO 2 storage deployment (GCCSI, 2015).Germany, the Netherlands and the UK are the only countries that are at least ranked advanced.The assessment reveals a high correlation between a country's ranking and the existence of an advanced hydrocarbon industry, and its dependence on fossil resources.Following long debates, onshore storage was excluded as a storage option in Germany (Schumann et al., 2014), 10 Denmark, 11 the UK and Netherlands (GCCSI, 2012).Analogous developments are conceivable for other countries, leaving offshore storage as the only remaining storage option in Europe.Accordingly, none of the Europe-based large-scale integrated CCTS projects listed in the Global CCS Institute database include onshore CO 2 storage (GCCSI, 2014).Therefore, in all presented scenarios, onshore storage capacity is not available, which reduces total available storage capacity from 94 GtCO 2 to 50 GtCO 2 in the European-wide scenarios and from 56 GtCO 2 to 42 GtCO 2 for the scenarios which focus on the North Sea region.As a consequence, France and Belgium lose most of their domestic storage potential.Despite a number of minor storage resources (1.2 GtCO 2 ) in saline aquifers in the German North Sea, the situation in Germany is similar.
The resulting scenarios shown in Table 2 differ in their respective CO 2 price path, the availability of storage potential (offshore with vs. without CO 2 -EOR) and geographical coverage (European-wide vs. the North Sea region vs. selected countries).Section 3 describes the European scenarios (EU_40% and EU_80%) while section 4 further analyzes regional scenarios (NorthSea_40%, NorthSea_80% and DNNU_80%).

EU_40% Scenario
CCTS starts being deployed from the year 2035 onwards when the CO 2 certificate prices pass the 40 €/tCO 2 threshold.Nevertheless only a very small annual amount of around one MtCO 2 is captured and stored in offshore hydrocarbon fields as well as saline aquifers.Hydrocarbon fields are the cheapest storage option when excluding CO 2 -EOR, but are not available in all locations.Four iron and steel factories in Norway and Estonia are the only emitters that invest in capture technology, benefiting from the lower variable and fixed costs assumed for this industry.The investing factories are located at the coast which leads to lower transport costs than for other industrial facilities.The overall costs sum up to €0.2 bn of investment costs and an additional €0.4 bn of variable costs until 2050.

EU_80% Scenario
The increase of the CO 2 price in the EU_80% scenario is higher than in the EU_40% scenario.The price increases gradually until a stronger rise kicks in in 2030, resulting in its final value of 270 €/tCO 2 in 2050.CCTS deployment starts once the CO 2 price exceeds 40 €/tCO 2 which happens in the year 2030 due to the steep path increase.The first investments into the CCTS technology are seen in the previous years (2020)(2021)(2022)(2023)(2024)(2025).The iron and steel sector is-similar to previous modeling runs in Oei et al. (2014)-again the first mover until some cement works start capturing CO 2 from 2035 onwards (see Figure 3).At that point a certificate price of 75 €/tCO 2 is reached and a total of 300 MtCO 2 is annually stored in offshore hydrocarbon fields and saline aquifers.CCTS becomes economical for power plants and refineries as soon as the price exceeds 100 €/tCO 2 in the year 2040.Still rising prices above 180 €/tCO 2 in 2045 lead to additional economic incentives for more distant power plants to invest in further CCTS deployment.Annual captured emissions sum up to more than one billion t CO 2 from 2040 to 2045.These emissions are then transported via a pipeline network of 44,800 km to different storage locations.Total captured emissions start decreasing after 2045 due to the phase-out of several older power plants.12.2 GtCO 2 is stored in offshore storage sites until 2050; 55% of these emissions originate from industrial sources.
The capturing costs have the highest share (60-70%) in variable as well as fixed costs of the CCTS chain.The infrastructure costs of storage comprise around 30% of the total investment costs, but have relatively small share of total variable costs of 10%.Transport costs depend very much on the location of each facility and range around 10-15% in variable and fixed costs.This step of the CCTS technology chain is also the driver making CCTS a more beneficial option for a Data specification used for coal-fired power plants in Mendelevitch (2014).b EC (2013); based on emission factor 0.9 tCO 2 /MWh, load factor 86%, reference power plant 2100€/kW overnight capital costs.
12. The given costs only include the additional variable and fixed costs for equipping a power plant or industrial facility with a capturing unit compared to a facility without a capturing unit.
facilities closer to potential storage sites.This can be clearly seen as the first movers are mostly located near the North Sea where the highest offshore storage potential can be found.The overall investment costs until 2050 exceed €300 bn with an additional €730 bn of variable costs.

Sensitivity to Investment and Variable Costs
Many cost studies of the CCTS technology chain name the capture stage as most cost intensive for both investment and variable costs (e.g.The Crown Estate et al., 2013).To assess the sensitivity of the resulting CCTS infrastructure to these cost parameters, we simulated four additional scenarios: Two where we double the capital costs (Inv_200%) and variable costs (Var_200%) respectively, one with double capital and variable costs (Inv&Var_200%), and one with variable and capital costs both increased by 50% (Inv&Var_150%). 12Table 3 provides the input values for the sensitivity analysis and reference values from CCTS-Mod and the PRIMES model of the European Commission (EC, 2013b) for comparison.The capital costs used for the base run are 25-30% below the input values in the PRIMES model.For variable costs no values for comparison were available.The results are calculated using a discount rate of 5%.In general, higher discount rates give lower weight to costs that occur further in the future.Investment cost are not annualized in CCTS-Mod but are incurred as a total five year before the infrastructure can be used.By contrast, variable costs occur on an annual basis.Therefore their relative importance investment cost increases with higher discount rates.Additional sensitivity analysis with 2.5% and 7.5% discount rate, however, confirm the relative importance of variable versus investment costs in all runs even though absolute importance diminishes with higher discount rates and increases with lower discount rates.
In all sensitivity runs the increase in costs leads to a significant delay in initial deployment of CCTS technology.Figure 4 (left side) shows that while in the base run CCTS is first introduced in 2030, in the Inv_200% and Inv&Var_150% scenario the technology is first used in 2035, and only in 2040 in the other two scenarios.The figure also illustrates the sensitivity of total costs and length of the pipeline network until 2050.For all sensitivity runs cost figures are 5-25% higher than in the base case, showing an increasing sensitivity over the model horizon due to the accumulation of higher variable costs.Figures on CO 2 capture, storage and pipeline network are lower  for the sensitivity runs than for the base case, with the gap narrowing between 2040 and 2050 (see Figure 4 right side).For the two scenarios Inv_200% and Inv&Var_150% the overall impact on key results like capture, and storage amounts and length of pipeline infrastructure is at most 10% or less.By contrast, doubling the variable capture costs has a strong impact on the length of the pipeline network with a decrease of over 35% compared to the base case.The future development of a CCTS infrastructure is therefore more sensitive to its variable costs than its investment costs.However, the deployment of the CCTS technology as a whole is not very sensitive to even drastic increases in capture costs, given high CO 2 certificate prices in the end of the modeling horizon (270 €/tCO 2 ) and the lack of alternative technologies, as both prevailing in this modeling framework.

Summary of the European-wide Scenarios
Table 4 summarizes the results of the different Europe-wide scenarios.A summary of all scenario results can be found in the Appendix.In the EU_40% scenario only four iron and steel factories in Norway and Estonia invest in the capture technology as they profit from the industry's low variable and fixed costs.These facilities additionally benefit from their ideal location close to storage sites in the North Sea, minimizing costs associated with CO 2 transport.CCTS cannot be considered as an abatement option for power plants if CO 2 prices barely rise above 50 €/tCO 2 .Sensitivity analysis shows that the future development of a CCTS infrastructure is more sensitive to its variable costs than its investment costs.
The EU_80% scenario arrives at CO 2 certificate prices around 250 €/tCO 2 in the year 2050.Under this assumption, investing in the CCTS technology is cost-efficient for all emitters, with industry still being the first mover.However, from today's perspective, these modeling results seem unrealistic.Even under the assumption of one omniscient planner, a CO 2 pipeline network of at 13.This is partly due to rising public opposition (not in my backyard effect-NIMBY) as well as different national interests (e.g.France focusing on nuclear energy, Germany on the other hand on renewable energy sources).least 45,000 km covering great parts of Europe would be needed.Overall system costs, including costs of carbon capture, transport and storage over the entire model horizon, sum up to €800-1,000 bn.The construction of such a huge new infrastructure network is highly dependent on public acceptance, especially in densely populated regions like Europe (Gough et al., 2014).Considering the number of different parties, technology stages, insecurities regarding CO 2 prices, learning rates and further policy measures, one comes to the conclusion that the necessary infrastructure and investment costs would be several times higher.This questions the fact whether CCTS may be able to fulfill its role as a decarbonization technology within Europe.The following section 4 therefore focuses on regional CCTS deployment around the North Sea only.

REGIONAL FOCUS: CO 2 -ENHANCED OIL RECOVERY OPTIONS IN THE NORTH SEA AND THE ROLE OF REGIONAL COOPERATION
The planned demonstration projects with the highest chance of realization are all close to the North Sea and aim for offshore storage with additional profit generated from CO 2 -EOR (GCCSI, 2014).The following scenarios depicted in sections 4.4 and 4.5 assess the implications of CO 2 -EOR for the development of a CCTS infrastructure with a focus on the North Sea Region.Several of these countries however, such as Germany and France, are unlikely to take part in any future CCTS deployment. 13Different national strategies towards implementation of CCTS, instead of a joint European energy strategy, thus seem most likely at the moment.Section 4 therefore includes a regionally focused analysis of four European countries where a joint CCTS and CO 2 -EOR deployment is most likely: Denmark, the Netherlands, Norway, and the UK (DNNU).One interesting aspect analyzed in this section is whether the employment of CO 2 -EOR by a limited number of countries increases costs due to a lack of economies of scale during the use of CO 2 -EOR and later.The assumed price paths are the same as in the previous scenarios.

The Role of CO 2 Reuse for CCTS
CO 2 -EOR is the most mature CO 2 reuse technology and has been practiced since the 1980s in the USA and Canada (cf.GCCSI, 2011).The application of other technologies that are in the commercialization phase like Bauxite residue carbonation and using CO 2 in methanol production is very site specific and requires favorable local conditions.The use of CO 2 in enhanced coal bed methane recovery, as a working fluid in enhanced geothermal systems, as feedstock in polymer processing, and for algae cultivation are all technologies that need to be further developed and proven in real world pilot or demonstration scale applications.The global market for CO 2 reuse across all technologies has a volume of approximately 80 Mt per year, which is equivalent to the annual emissions of the four biggest lignite power plants in Germany.CO 2 -EOR in the USA and Canada account for the biggest share with 50 Mt per year.80% of the CO 2 is supplied from natural CO 2 sources at a price in the order of 15-19 US$/tCO 2 .In total, anthropogenic CO 2 emissions can only be offset to a few percent from current and potential future demand for CO 2 reuse.Although reuse has very limited potential it can generate modest revenues for a selection of near term CCTS projects.Its impact to global CO 2 abatement, however, depends on the application as e.g.CO 2 -EOR and using CO 2 in methanol production have no positive climate effect due to the latter burning of the product (Gale et al., 2015).
IEA and UNIDO (2011) give a similar assessment of the role of CO 2 -EOR for the development of the CCTS technology appraising it as an important way to add value to a CCTS operation.The IEA ( 2012) is analyzing the role of this technology.It acknowledges that CO 2 -EOR not only offers a way to partly offset the costs of demonstrating CO 2 capture but also to drive the evolution of CO 2 transportation infrastructure, and incorporates opportunities for learning about certain aspects of CO 2 storage in some regions.Several studies have looked into the economics of CO 2 -EOR on a regional and national scale: e.g. the application of the technology in the UK Central North Sea/Outer Moray Firth region (Kemp and Kasim, 2013;Scottish Centre for Carbon Storage, 2009) and the Norwegian continental shelf (Klokk et al., 2010), and have found substantial potential for the combination of the two technologies and associated benefits.

CO 2 -EOR Resources in the North Sea
The analysis of the role of CO 2 -EOR for the development of a CCTS infrastructure requires a comprehensive estimation of the potential for CO 2 -EOR in the North Sea region.Mendelevitch (2014) performed an intensive literature review and presents own estimates to compile a consistent database of CO 2 -EOR potentials in the North Sea region.Data availability diverges significantly between the various countries of the North Sea Region.Therefore, different approaches have been chosen for each country.CO 2 injection potentials are considered as the net amount of CO 2 that can be stored during the CO 2 -EOR process and includes a constant recycling ratio of 40% following Gozalpour et al. (2005).
For the UK Mendelevitch (2014) finds 54 candidate fields with an estimated net injection potential ranging between 2 and 89 MtCO 2 (Forties field).Total UK potential sums up to 572 MtCO 2 which corresponds to 1733 Mbbl additional oil recovery potential.For Norway the author identifies seven candidate fields with an estimated net injection potential ranging between 4 and 130 MtCO 2 (Ekofisk field).Total storage potential in Norwegian oil fields in the North Sea add up to 314 MtCO 2 which corresponds to an additional oil recovery potential of 951 Mbbl.For Denmark the study finds 14 candidate fields with an estimated net injection potential ranging between 3 and 88 MtCO 2 (Dan field).Total storage potential in Danish oil fields sums up to 348 MtCO 2, which corresponds to an additional oil recovery potential of 1054 Mbbl.Other riparian countries of the North Sea do not exhibit substantial oil resources and are therefore not included in the analysis.

Costs and Revenue of CO 2 -EOR
To assess the economics of a prospective CO 2 -EOR infrastructure correctly, it is crucial to accurately estimate the costs associated with it.Mendelevitch (2014) draws on various case studies on CO 2 -EOR projects in the North Sea to develop an inventory of the main investment and operating costs components (see Table 5).
Based on the cost components mentioned above investment costs add up to 103.9 €/tCO 2 stored per year and operating costs add up to 36.8 €/tCO 2 stored.Without costs for CO 2 import the costs for oil supply from CO 2 -EOR in the North Sea Region are in the range of €12-17 per bbl incremental oil (depending on site specific CO 2 utilization rates), which is consistent with estimates from OECD and IEA (2008), giving a range of US$40-80 per bbl (including costs of CO 2 supply) for long-term oil supply from CO 2 -EOR.a Variable costs of CO 2 storage include operational costs (OPEX) of oil production (see Table 5).Source: Mendelevitch (2014).
Expectations about the development of the crude oil price determine the attractiveness of CO 2 -EOR operations.The price not only has to cover investment and variable costs of incremental oil production but also has to refinance the capture and the transport of the CO 2 .DOE/IEA (2012) present a compilation of different oil price projections for the Western Texas Intermediate (WTI) crude oil price for the period up to 2035.The chosen medium oil price path represents an average of the price projections while the lower price path marks their lower bound.To provide a rough estimate of the profitability of combining CCTS with CO 2 -EOR, Table 6 compares cost and revenue items for a generic example.The sales price of additionally produced crude oil and the assumed CO 2 certificate price (as negative opportunity costs) of the respective year constitute the potential revenue side.On the costs side, investment and variable costs for each of the stages of the CCTS technology chain are included.Even for the high first-of-a-kind investment costs assumed for 2015 A CO 2 utilization rate of 0.33 tCO 2 /bbl (Mendelevitch, 2014) and an exchange rate of 1.25$/€ is being used.Additional capture costs for a coal-fired power plant equipped with post-combustion capture are calculated including a 5% discount rate and 30 years of operating life.The transport costs are estimated by assuming a 500 km long pipeline, with a lifetime of 30 years and a 5% discount rate.CO 2 -EOR equipment is expected to have a much shorter operating life of 10 years and the same discount rate of 5%.and 2020 the combination of the two technologies yields considerable profit of 100 €/tCO 2 and higher.The two most critical assumptions are the "bbl crude oil per tCO 2 injected" conversion rate and assumptions on the future development of oil prices.Until now, CO 2 -EOR operations have only been performed onshore.Employing the technology in the North Sea is associated with additional technological and therefore also financial risk which is not taken into account in this calculation. 14

Regional Scenario: NorthSea_40% Scenario with CO 2 -EOR Option
The NorthSea_40% scenario assumes the same CO 2 price path as the EU_40% scenario (see Table 1).Scenario results show that the use of CCTS is still most economical for the industrial sector, particularly iron and steel making plants.These facilities invest in a CCTS infrastructure from 2015 to 2020 in order to gain profits from additionally recovered oil from CO 2 -EOR from 2025 onward.After the exhaustion of most of the CO 2 -EOR fields in 2035, new storage sites in saline aquifers and depleted hydrocarbon fields closer to the shore are being used (see Figure 5 for the CO 2 flows in 2050).In this scenario, a total of 2.5 bn tCO 2 is stored until 2050 with annual storage volumes of around 100 MtCO 2 .The required CO 2 transport network spans approximately 15,000 km.The scenario indicates that the CO 2 -EOR technology could lead to additional early economic incentives for the construction of a CCTS infrastructure.Existing infrastructure can be used after the exploitation of the CO 2 -EOR potential in the North Sea as soon as the CO 2 price is high enough.In the case of CO 2 prices remaining at around 50 €/tCO 2 (as seen in the EU_40% scenario), it is still only economical for several industrial facilities such as steel or cement.The

Regional Scenario: NorthSea_80% Scenario with CO 2 -EOR Option
The NorthSea_80% scenario assumes the same CO 2 price path as in the EU_80% scenario (see Table 1).Until 2035-the point when the CO 2 -EOR operation stops due to depletion-results of the NorthSea_80% scenario are very similar to those of the respective NorthSea_40% scenario.From 2020 onwards an average of 100 MtCO 2 is transported each year from steel and cement facilities into CO 2 -EOR operations in the North Sea.Once CO 2 -EOR resources are depleted, further CO 2 capture activity is solely incentivized by the CO 2 certificate price, which has to cover at least the variable costs as well as potential new investment costs.New storage in non-CO 2 -EOR locations is being developed close to the shore and close to already existing transport routes.From 2035 onwards, with prices exceeding 75 €/tCO 2 , additional more distant industrial facilities start running their capturing units.Similar to the results from the respective EU_80% scenario without the CO 2 -EOR option, power plants only start capturing their CO 2 from 2040 onward.The network required to accomplish the CO 2 transport spans 27,000 km connecting the sources to the North Sea storage sites (see Figure 6).The investment costs sum up to €190 bn and there are an additional €540 bn variable costs over the whole time period until 2050 (see Figure 7).Revenues from selling additionally recovered crude oil sum up to €300 bn, similar to the results in the NorthSea_40% scenario.However, in contrast to the NorthSea_40% scenario, the high CO 2 price in this scenario creates enough incentive to pursue CCTS even after the depletion of CO 2 -EOR resources and eventually leads to full deployment of the technology in the modeled sectors.
Note that the total amount of CO 2 captured is lower than in the EU_80% scenarios without the CO 2 -EOR option because this analysis focuses only on the riparian countries of the North Sea.However, like in the EU_80% scenario, all examined industrial facilities and power plants start using the CCTS technology at some time; with industry still holding the higher share of total stored emissions over time.

Regional Scenario: DNNU_80% Scenario Focusing on CO 2 -EOR in DK, NL, NO and UK
Against the background of a lack of industry and policy commitment in Germany, France, Belgium and Sweden, we examine an additional scenario where only Denmark, the Netherlands, Norway and the UK have the possibility to use the CCTS technology.In contrast to the other European countries, these four have a higher potential to use captured CO 2 to generate additional revenue in the domestic oil industry, or at least support the application of CCTS in the industry sector (like in the Netherlands).Moreover, UK and Norway are still the only two signatories to the amended London Protocol to allow transnational CO 2 transport for offshore storage (GCCSI, 2014), and these four are among the most advanced countries ready for large-scale CO 2 storage operation (GCCSI, 2015).Our goal is to compare these results to the results of the other scenarios and to examine to what extent CCTS deployment is reduced due to a lack of economies of scale.
Similar to the previous scenarios, the use of CCTS is mainly economical for the industrial sector, particularly iron and steel making plants.In the DNNU_80% scenario, facilities invest in a CCTS infrastructure from 2015 to 2020 in order to gain profits from additionally recovered oil from   7. Values for 2030 also include investments for non-CO 2 -EOR induced CO 2 transport and storage, as investments the model features a 5 year construction period before infrastructure can be used.
CO 2 -EOR from 2025 onward.Around 100 MtCO 2 is stored annually until the full exhaustion of the CO 2 -EOR resources, 10 to 15 years after the beginning of the operation (with a concentration in the first ten years).From 2035 onwards, additional storage sites in saline aquifers and depleted hydrocarbon fields closer to the shore are used by industrial facilities already equipped with CO 2 capture.With CO 2 prices exceeding 75 €/tCO 2 in the DNNU_80% scenario, additional, more distant industrial facilities start investing in capture units.Power plants only start using the CCTS chain from 2040 onwards, similar to the outcome of previous scenarios without the CO 2 -EOR option.
For the period of the CO 2 -EOR boom (2025-2035), the results of the DNNU_80% scenario on length of the pipeline network are similar to those of the NorthSea scenarios.While distances to deliver CO 2 up to the shore are shorter on average, CO 2 from the UK takes especially long routes offshore to arrive at CO 2 storage sites with CO 2 -EOR options (see Figure 8).The overall installed pipeline network in 2030 covers over 11,000 km (10,200 in the NorthSea scenarios).Similarly, the values for average investment in CO 2 transport and CO 2 storage per MtCO 2 per year during the initial phase in 2025 do not change for the DNNU scenario (cf.Table 7). 15Due to a similar de- ployment of the technology, no economies-of-scale effect between the NorthSea_80% scenario in 4.5 and the DNNU_80% scenario can be observed during this period.However, the DNNU_80% scenario exhibits a shift in CO 2 -EOR utilization.We find that UK CO 2 -EOR storage potential used by France and Belgium in the other scenarios is now intensively used to store domestic CO 2 from UK (increase of 46 MtCO 2 per year for the period from 2025 to 2040 in the UK).The same effect but to a smaller extent (9 MtCO 2 per year) can be observed in Norway.Danish oilfields that stored CO 2 from Germany in the other scenarios, now increasingly receive CO 2 from the Netherlands (increase of 27 MtCO 2 captured per year in the Netherlands in the period from 2025 to 2040).At the same time, capture activity in Denmark does not change significantly.After the CO 2 -EOR boom, the storage volumes for the four countries do not differ between the different scenarios.A clear economies-of-scale effect can be observed for the post-CO 2 -EOR period.In 2040, average investment costs in both CO 2 transport and storage infrastructure are much higher for the DNNU scenario compared to the NorthSea scenarios.CO 2 storage costs increase by more than 30% in 2040 while transport costs even double (cf.Table 7).The constructed transport network is much smaller than in the NorthSea_80% scenario (13,600 km compared to 26,800 km) which is due to a smaller observed area and the lack of economies of scale.Table 8 summarizes the key results of the NorthSea and DNNU scenarios.A summary of all scenario results can be found in the Table 9. Due to their regional focus, volumes of CO 2 stored and required transportation distances in these scenarios are likely to be shorter than in the European-wide scenarios of Section 3.

CONCLUSION: THE IMPORTANCE OF CO 2 -EOR FOR A EUROPEAN CCTS ROLL-OUT
In this paper we present scenario analyses and interpretation on the potential role of CCTS to support the EU energy system transition to meet emission reductions goals that are consistent with the international goal of staying below 2ЊC of global warming.The assumptions of the moderate scenarios include a CO 2 price of 50 €/tCO 2 in 2050 which triggers hardly any CCTS development in Europe.Additional revenues from applying CO 2 enhanced oil recovery (CO 2 -EOR) in the North Sea lead to an earlier adoption of CCTS starting in 2025 independent from the CO 2 certificate price.The lifespan of most CO 2 -EOR operations is expected to be around ten years.It is followed by conventional CO 2 storage in nearby depleted hydrocarbon fields and saline aquifers if the CO 2 certificate price exceeds the sector-specific thresholds to cover variable costs of carbon capture.Generally, the use of CO 2 for EOR projects is criticized by environmental organizations, as the overall CO 2 mitigation effect is negative considering the CO 2 content of the additional extracted oil.More stringent climate scenarios aim at an 80% GHG reduction until 2050.The resulting CO 2 price of 270 €/tCO 2 in 2050 pushes all EU-ETS industry and energy sectors to use CCTS at some point.It is, however, the iron and steel sector that starts deployment as soon as the CO 2 certificate price rises above 50 €/tCO 2 in 2030.The cement sector follows some years later at a threshold of around 75 €/tCO 2 .It is only with CO 2 certificate prices exceeding 100 €/tCO 2 that CCTS becomes lucrative for the electricity sector.Sensitivity analysis shows that the future development of a CCTS infrastructure is more sensitive to its variable costs than its investment costs.As European usage of onshore storage sites is unlikely due to high public resistance, transport distances increase.The resulting CO 2 transport network required to connect emission sources and storage sites across Europe would comprise of up to 45,000 km of pipeline and store up to 1,000 MtCO 2 per year.
Taking into account the difficulties to establish CCTS in the EU, there are only a handful of countries that still consider building CCTS in the medium term.A 20% CCTS penetration rate in the European power sector as calculated in the DNNU_80% scenario in 2050 thus seems more realistic.Concentrating on Denmark, the Netherlands, Norway and the UK, this scenario shows an increased utilization of CCTS-EOR especially in the UK and the Netherlands.However, a lack of economies of scale leads to increasing average costs, once the CO 2 -EOR-fields have been exploited: CO 2 storage costs increase by more than 30% in 2040 while transport costs even double.
A critical point of our analysis is that the employed model CCTS-Mod is purely costdriven and does not include any specific bound on the CCTS penetration.The model assumes a cost-minimizing player with perfect foresight and therefore tends to overestimate the potential for CCTS.Additional legal, political, and environmental issues with respect to transboundary CO 2 transport as well as CO 2 storage and liability issues are not included in the model.The model analysis highlights that the international distribution of CO 2 -EOR and non-EOR storage sites leaves room for significant international conflicts of interest and the need for coordination between the North Sea riparian countries.Some countries, e.g.France or Germany, only have limited (offshore) CO 2 -storage capacities.A non-coordinated European CCTS utilization does not profit from the mentioned economies of scale resulting in higher overall system costs.Real costs are expected to be higher and come with a lower deployment of CCTS in the future.On the other hand, already the existence of a CO 2 infrastructure might impact investment decisions in various sectors of the energy system (spatially and with respect to the selection of generation technology) that are ne-glected in this modeling approach.Considering the large number of different players and technologies, the insecurities regarding CO 2 prices, learning rates, legal issues, public resistance and further policy measures strongly question whether CCTS may be able to fulfill its role as a major bridging technology for the decarbonization of Europe.
CO 2 -EOR is the driver behind all global CCTS projects that will become operational in the near future or have already started operation (e.g.Boundary Dam, Canada).The underlying regulatory frameworks and support schemes can primarily be regarded as a cross-subsidization of the petroleum industry, while progressing the CCTS technology is of secondary interest.This is underpinned by observations in the Gulf States, the USA and Canada, where the legislative framework for CO 2 -EOR with CO 2 recycling is established, while the framework for long-term storage (which would be the primary goal of CCTS) is underdeveloped.The same is true for Europe, where the emergence of a regionally focused network around the North Sea, including only a few riparian countries using offshore CO 2 storage with CO 2 -EOR, is the most likely option.The mirage of a Pan-European network for CCTS in the EU-ETS industry and energy sectors, as envisioned in some long-term scenario projections, seems out of reach at present due to a combination of a lack of financial incentives as well as too little political and public support for CCTS as a mitigation technology.Further research, however, is needed to evaluate the effects of the newest European reforms (e.g. the reform of the EU Emissions Trading System ETS) as well as national regulations (e.g.emissions performance standards (EPS) and contract for differences (CfD) in the UK) on the development of CCTS.

Figure 1 :
Figure 1: Decision Tree of the Model CCTS-Mod with the Option of CO 2 -EOR

Figure 2 :
Figure 2: Distribution of Potential CO 2 Storage Sites (left) and CO 2 Source (right) by Type and Volume in the Data Set

Figure 3 :
Figure 3: Captured CO 2 Emissions by Source and Storage Type over Time in the EU_80% Scenario

Figure 4 :
Figure 4: Sensitivity of Captured Amounts over the Model Horizon (left side), and Total Costs and Length of the Pipeline Network in 2050 (right side)

Figure 5 :
Figure 5: CO 2 Flows in the NorthSea_40% Scenario in 2050 after CO 2 -EOR-fields are Exploited

Figure 6 :
Figure 6: CO 2 Flows in the NorthSea_80% Scenario in the Year 2050 after CO 2 -EOR Fields are Exploited

Figure 7 :
Figure 7: Cost Distribution over the Whole Timespan in the NorthSea_80% Scenario in €bn

Figure 8 :
Figure 8: CO 2 Flows in the DNNU_80% Scenario in 2025 using the CO 2 -EOR-option (left) and in 2050 after CO 2 -EOR-fields are Exploited (right)

Table 7 : Average Investment Costs in CO 2 Transport and CO 2 Storage per MtCO 2 per Year, Comparing the NorthSea_80% and DNNU_80% Scenarios
To assess economies of scale for the CO 2 -EOR boom period one has to compare 2025 values from Table