Achieving negative emissions with BECCS (bioenergy with carbon capture and storage) in the power sector: New insights from the TIAM-FR (TIMES Integrated Assessment Model France) model

Energy 76 (2014) 967e975

Contents lists available at ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

Achieving negative emissions with BECCS (bioenergy with carbon capture and storage) in the power sector: New insights from the TIAM-FR (TIMES Integrated Assessment Model France) model Sandrine Selosse, Olivia Ricci*, 1 MINES ParisTech, Centre for Applied Mathematics, Rue Claude Daunesse, BP207, 06904 Sophia Antipolis, France

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 April 2013 Received in revised form 3 September 2014 Accepted 6 September 2014 Available online 29 September 2014

It seems increasingly likely that atmospheric greenhouse gas concentration will overshoot the recommended 450 ppm CO2 equivalent target. Therefore, it may become necessary to use BECCS (bioenergy with carbon capture and storage) technologies to remove CO2 from the atmosphere. This technique is gaining increasing attention as it offers the dual benefit of providing low-carbon energy products and leading to negative CO2 emissions. This study evaluates the possible deployment of BECCS in the power sector using the bottom-up multiregional optimization model TIAM-FR (TIMES Integrated Assessment Model France). Under two climate scenarios, a regional analysis is conducted to discuss where the technology will be developed. The impact of the unavailability of this technology on the structure of the electricity mix and the cost of the energy system completes the analysis. In line with literature, the results suggest that BECCS technology offers an environmentally and economically viable option to achieve stringent targets. The regional analysis shows that industrialized countries will develop CCS (carbon capture and storage) mainly on biomass power plants while CCS on fossil fuel power plants will be widely deployed in China. With a specific constraint on CCS diffusion, the share of renewables and nuclear energy becomes significant to meet the climate targets. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Biomass (CCS) carbon capture and storage TIAM-FR (TIMES Integrated Assessment Model France) Electricity Climate policies

1. Introduction In its Fourth Assessment Report, the IPCC (Intergovernemental panel on climate change) showed that in order to limit the longterm global temperature increase to 2
C above pre-industrial levels and avoid dangerous climate change consequences, a radiative forcing of below 3 W/m2 is required around the end of the century. However, it seems increasingly likely that we will overshoot this limit. The IEA (International Energy Agency), in its World Energy Outlook [1], announces that “the door to 2
C is closing”. Therefore, it may become necessary to develop technologies that capture emissions from the atmosphere (negative CO2 emissions technologies). By capturing CO2 from the air (directly or indirectly), CO2 emissions can be sequestered and the stock of atmospheric CO2 reduced to correct the overshoot. This technique could also be used to offset additional anthropogenic emissions from sectors where * Corresponding author. Tel.: þ33 4 97157086; fax: þ33 4 97157066. E-mail addresses: [email protected] (S. Selosse), olivia.ricci@ mines-paristech.fr (O. Ricci). 1
de La Re
union, CEMOI, 15 Avenue Rene
Cassin, Permanent address: Universite
union, France. BP 7151, 97715 St-Denis, La Re http://dx.doi.org/10.1016/j.energy.2014.09.014 0360-5442/© 2014 Elsevier Ltd. All rights reserved.

emissions reductions are difficult or uneconomical [2,3]. A range of negative emissions options have been identified, such as those that directly remove CO2 from the atmosphere, i.e. so-called direct aircapture technologies (artificial trees and Lime-Soda process), and those that remove emissions indirectly (augmented ocean disposal processes; biochar and BECCS (bioenergy with carbon capture and storage)) [4]. Given that the cost of direct air-capture technologies is still very uncertain and high [5,6], BECCS appears to be the negative CO2 emissions technology with the most immediate potential to reduce emissions. BECCS can be defined as a process in which CO2 originating from biomass is captured and stored in geological formations. Biomass absorbs CO2 from the atmosphere through photosynthesis and releases it during transformation or combustion. If the released CO2 could be captured and stored permanently in geological storage sites, this would result in a situation of negative CO2 emissions given sustainable biomass harvesting practices. According to [7], this option has the dual advantage of providing low-carbon energy products and allowing net removal of CO2 from the atmosphere. Many empirical studies show that the use of BECCS is increasingly significant to tackle strict stabilization targets

968

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

[2,8e17]. The availability of BECCS decreases the cost of meeting low stabilization targets. A comparison of the results of three energy models that include BECCS technologies [2] shows that atmospheric CO2 concentration can be reduced by 50e100 ppm for the same cost as when BECCS is used. In fact, negative emissions are essential to meet low concentration targets. As part of the EMF-22 (Energy Modeling Forum) study [9], ten integrated assessment models focused on the requirements of limiting radiative forcing from Kyoto gases to 2.6 W/m2 (equivalent to 450 ppmCO2) by 2100. Assuming that all countries participated in climate policy in 2012 and integrating the possibility of temporarily exceeding the target before 2100 (i.e. an “overshoot”), some models showed that the 2.6 W/m2 target was unattainable without the use of BECCS [9,18]. When a model was able to reach the 2.6 W/m2 target without BECCS, the cost was much higher than when BECCS was available [19]. The introduction of BECCS in integrated assessment models diverts the emission reduction pathway toward the long-term radiative forcing objective. It increases near-term flexibility in abatement timing in such a way that emissions reduction can occur in the second half of the century, and modest emissions reduction can be achieved earlier, decreasing the total discounted abatement cost [2,9,15,19]. However, the argument of postponing emissions reduction raises serious concerns, since relaxing action today could lead to a high overshoot in the concentration level with irreversible consequences on the climate [2]. Several sectors have been identified as appropriate targets for the BECCS option, such as the heat and pulp mill industries [20,21], the bio fuel sector [22e27] and the electricity sector [28e30]. This study focuses on the electricity sector, which is the main producer of energy-related CO2 emissions. CO2 capture and storage technologies have therefore been recognized as critical factors to decarbonize this sector [31]. Indeed, most support for CCS (carbon capture and storage) demonstration projects has focused on power applications [32]. We contribute to the growing body of literature on BECCS by introducing a wide variety of CCS technologies on coal, gas, cocombustion of coal, and biomass and biomass power plants in the multi-region model TIAM-FR (TIMES Integrated Assessment Model France). This bottom-up optimization model provides a technology-rich basis for estimating energy evolution and structural changes in the long-term. It depicts the energy system over the period 2005e2100 in such a way as to minimize the total discounted cost of the system under a number of environmental, technological and demand constraints. The aim of this paper is to compare the results of TIAM-FR to existing estimates and underline new insights with a detailed technological and regional approach. We analyze the deployment of BECCS technologies in the electricity sector up to 2100 with ambitious climate objectives. Our study involves making global and regional analyses in order to quantify BECCS potential in industrialized, fast-developing and developing countries. Moreover, the feasibility of BECCS as a negative emissions process technology is heavily dependent on the future development of carbon capture and storage technology. Due to substantial uncertainties regarding storage capacities, the availability of CO2 transport networks, social acceptability, legal issues, and adequate technology incentives [33], we also investigate the impact that exogenous constraints on CCS and BECCS availability will have on the global and regional electricity mix structure, substitution between CCS technologies, and the global and regional costs of the energy system. The paper is organized as follows: Section 2 describes the model, the main assumptions and the climate scenarios. Section 3 presents and discusses the results of the long-term modeling. The final section gives some concluding remarks.

2. TIAM-FR model and scenarios 2.1. TIAM-FR structure Analyses carried out in this paper are based on the TIAM-FR model developed by the MINES ParisTech Center for Applied Mathematics (see Refs. [31,34,35]). TIAM-FR is based on the TIAM (TIMES Integrated Assessment Model), a widely used, linear-programming TIMES family model developed under the IEA's ETSAP (Energy Technology Systems Analysis Program) [36and 37]. The TIAM model is a bottom-up optimization model that offers a technology-rich representation of the energy system compared to other energy system models [38]. TIAM-FR is geographically integrated in 15 global regions (Industrialized countries: AustraliaeNew Zealand (AUS); Canada (CAN), United-States of America (USA), Western Europe (EU-15, Iceland, Malta, Norway and Switzerland, WEU), Eastern Europe (EEU), Japan (JPN); Fast developing countries: India (IND), China (includes Hong Kong excludes Chinese Taipei, CHI); Developing countries: Africa (AFR), Central and South America (CSA), Middle-East (includes Turkey, MEA), Mexico (MEX), SouthKorea (SKO), Other developing Asian countries (includes Chinese Taipei and Pacific Islands, ODA (other developing Asian countries)), Former Soviet Union (include the Baltic states, FSU (Former Soviet Union)). The model is driven by a set of 42 demands for energy services in agriculture, the residential, commercial, industry, and transport sectors. Demands for energy services are exogenously specified. TIAM-FR is a linear-programming approach in which the technical optimum is computed by minimizing the discounted global system cost. For each region, it computes a total net present value of the stream of the total annual cost, discounted at 5%1 to the selected reference year 2005. The total annual cost includes investment and dismantling costs (capital cost) that are annualized using hurdle rates, annual fixed and variable operation and maintenance costs, costs incurred for exogenous imports and for domestic resource production minus revenues from exogenous exports, and delivery costs for commodities consumed by processes. These regional discounted costs are then aggregated into a single total cost which is the objective function to be minimized by the model while satisfying a number of technological and/or environmental constraints. The objective function is:

NPV ¼

R X

X  refyy *ANNcostðr; yÞ 1 þ dr;y

r¼1 y2years

where NPV is the net present value of the total cost; ANNcost (r,y) is the total annual cost in region r and year y; dr,y is the discount rate, refy is the reference year for discounting, years is the set of years and R the set of regions. Each step of the energy chain, from mining to final energy service demands, is identified in the model in terms of both economic and technical characteristics. Each primary energy form is extracted from multiple layers of either reserves (fossil) or resource potential (wind, hydro, geothermal, biomass, etc.), each with a potential and a specific unit cost. This constitutes a supply curve for each energy form. Some types of energy are endogenously traded between the 15 regions (coal, crude oil, refined petroleum products, natural gas, and liquefied natural gas). The costs of these energy forms are therefore endogenous. This is not the case for biomass. In the model, biomass is characterized by manifold sources e industrial waste, municipal waste, landfill gas, bioenergy crops, and solid

1 This discount rate is in line with literature on the cost of climate targets, particularly integrated models like the one used in this study [15,18,19,39].

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

biomass resources e and the fact that it is not traded between regions. The maximum amount of available biomass for each region is determined exogenously. The global potential is estimated at 234 EJ per year in 2050 (72 EJ come from bioenergy crops, 72 EJ from solid biomass resources and the rest from industrial waste, municipal waste and landfill gas). In literature, biomass potential varies greatly given the different assumptions on land use, yield development, food consumption and other sustainability criteria, such as water scarcity and loss of biodiversity. This potential varies from 100 EJ to 300 EJ per year by 2050 [40,41]. Through its integrated climate module, the model makes it possible to analyze atmospheric GHG (greenhouse gas emissions) concentrations, radiative forcing and temperature changes. It explicitly models CO2, CH4 and N2O emissions from all anthropic sources (energy, industry, land, agriculture and waste) related to energy consumption and the processes themselves. The climate module is inspired by the Nordhaus and Boyer model [42]. It consists of three sets of equations (concentration, radiative forcing and temperature increase), for more details see Ref. [39]. The first set of equations calculates the atmospheric concentrations of the three gases (CO2, CH4 and N2O) using recursive dynamic equations. The second set of equations calculates the atmospheric radiative forcing of these three gases from the atmospheric concentrations. Emissions of some Kyoto gases (CFCs (chlorofluocarbons), HFCs (hydrofluorocarbons), SF6 (sulfur hexafluoride) are not explicitly modeled but a special forcing exogenous term is added into the equation. The third set accepts as input the total radiative forcing and recursively calculates the annual change in mean global temperature. The climate module of the model has been compared with more sophisticated climate models and found to be quite accurate within the range of emissions considered [39]. 2.2. CCS and BECCS in the electricity sector TIAM-FR integrates several carbon capture and sequestration technologies on fossil or bioenergy resources. The purpose of the capture process is to obtain a concentrated stream of almost pure CO2 at high pressure. There are three modes of capturing CO2 from fossil fuels (coal, oil, natural gas) in the model: 1) a postcombustion mode using a variety of processes such as reactive absorption or membranes, 2) a pre-combustion mode with conversion of fuel-chemical energy into H2, followed by simultaneous low cost carbon separation and 3) an oxy-combustion mode characterized by the low cost of CO2 separation, but necessitating a supply of O2. For bio plants and co-firing plants (co-combustion of 20% biomass and 80% fossil fuel), two capture technologies are retained: pre-combustion capture for the biomass gasification process, and post-combustion capture for the biomass direct combustion process. CCS technologies are assumed to be available from 2020. For each technology, economic parameters must be completed, such as capital costs incurred from investing and dismantling processes, operation and maintenance costs etc. The investment costs (in $2005) of an exclusively biomass combustion plant using post-combustion capture technology, and an exclusively biomass gasification plant with pre-combustion capture technology are respectively around 2125 USD/kWe and 2425 USD/ kWe. Co-combustion of biomass and coal plants with CCS are less expensive, with investment costs of between 1700 and 1800 USD/ kWe. CO2 can be stored in depleted oil and gas fields, deep saline aquifers and deep unmineable coal seams. Theoretical storage capacities are indicated by region with a distinction between onshore and offshore for oil and gas fields. Cumulated global storage capacities are 14.8 Gt of CO2 of which 12.6 Gt of CO2 can be stored in deep saline aquifers. Storage costs depend on the capacity of the reservoir and the number of injection wells. Based on the costs of

969

the Zero Emissions Platform (the European Technology Platform for Zero Emission Fossil Fuel Power Plants, ZEP (zero emissions platform)), the assumed investment costs are: 3 USD/tCO2 for onshore depleted oil and gas fields and 7 USD/tCO2 for offshore depleted oil and gas fields, 5 USD/tCO2 for aquifers and 5 USD to 10 USD/tCO2 for coal seams. Regarding carbon transportation, investment costs represent about 90% of the total cost of transport, and a cost of between 3 USD and 10 USD/tCO2 is assumed in the model. In order to evaluate the role of BECCS in long-term climate scenarios, we make different assumptions on the stringency of the environmental policy. 2.3. Climate and technology availability scenarios The two climate-related targets explored in this study are defined in terms of radiative forcing (i.e., the impact on the Earth's radiative balance) caused by Kyoto gases (CO2, N2O, CH4), HFCs), PFCs (prefluorocarbons) and SF6). The two targets are 2.6 W/m2 and 3.7 W/m2 in 2100. These targets were also used in the EMF-22 (Energy Modeling Forum) study [9], which will facilitate a comparison of the results. We also assume that all regions participate in emissions reduction. In the EMF-22 first-best case, mitigation is undertaken where it is the least costly, and the marginal abatement costs are equalized across regions. [31] have shown that this case is idealized and unrealistic, as countries made diverse emission reduction commitments to the UNFCCC (United Nations Framework Convention on Climate Change). However, it provides a useful benchmark for analyzing global and regional abatement costs and it is in line with EMF-22 published studies. In TIAM-FR model, it is possible to overshoot the long-term target at any time during the century as long as the target is met in 2100. This overshoot option gives greater near-term flexibility to meet targets. The climate scenarios are:  BAU scenario: A baseline scenario with no emission constraint is calculated first. This BAU (Business As Usual) scenario outlines some key patterns in the evolution of the energy system, and serves as the starting point for the analysis. The BAU scenario is then compared to the emissions reduction policy scenarios to assess the implications of carbon constraints on the evolution of the electricity system and formulate policy recommendations.  RF_2p6 scenario: This scenario consists in limiting radiative forcing from Kyoto gases to 2.6 W/m2 by 2100. This objective is compatible with the UNFCCC consensual 2e2.4
C objective (as specified by IPCC).  RF_3p7 scenario: This scenario limits radiative forcing from Kyoto gases to 3.7 W/m2 by 2100. 3. Results and discussion: assessing CCS and BECCS deployment in the electricity sector We also evaluate the impact of the availability of CCS, BECCS and co-firing technologies on the electricity mix and on CCS technology substitution in the most stringent scenario (RF_2p6). The technology scenarios are the following:  RF_2p6_NoCCS: CCS seems to be a promising technology to reduce CO2 emissions but a significant number of uncertainties and key aspects need to be addressed to scale up this technology. Therefore, we consider the case where CCS is not deployed on any type of power plant throughout the time horizon.  RF_2p6_NoBECCS: We assume that CCS technology on full biomass power plants is not available. It is only developed on coal, gas, and coal and biomass co-combustion power plants.

970

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

 RF_2p6_NoBECCSCF: Finally, we assume that CCS is only developed on coal and gas power plants. BECCS and CCS for co-firing technologies are not available. 3.1. Environmental impact of the climate policies The CO2 emissions total will increase significantly in the BAU scenario, that is, from about 25 Gt in 2005 to 65 Gt by 2100. Driven by these increased emissions, CO2 atmospheric concentration will rise significantly over time and reach 628 ppm in 2100, which corresponds to a radiative forcing from Kyoto gases of about 5.4 W/m2 during the same period. Under scenario RF_2p6 global CO2 emissions decrease by 50% in 2050 and 84% in 2100 compared to the model's reference year of 2005. It allows the atmospheric CO2 concentration to stabilize at 424 ppm by 2050 and 402 ppm by the end of the century (Fig. 1). If we compare the metrics of atmospheric CO2 concentration with the EMF-22 results, we are in line with the participating models. They show that a radiative forcing of 2.6 W/m2 corresponds to an atmospheric CO2 concentration of between 355 and 415 ppm. Under RF_3p7, CO2 emissions increase by 75.6% from 2005 to 2050 and reach their highest level in 2050 (43.9 Gt of CO2). Then emissions decrease in the second part of the century and reach 15.7 Gt of CO2 in 2100, a reduction of 37.2% compared to 2005 (Fig. 1). The atmospheric concentration pursues its growth until 2080 and then slows down to reach 493 ppm in 2100. When CCS technologies are not available more emission reductions are achieved in the first half of the century to meet the 2p6 target. In both climate policy scenarios, cumulative emissions in the electricity sector over the course of the century are reduced by 85% compared to the reference scenario. In scenario RF_3p7 emissions rise from 2005 to 2050 and then decrease sharply, while in scenario RF_2p6, emissions decline from 2010 and then level off in the second part of the century. 3.2. Impact of climate policies on the electricity mix and CCS deployment In this subsection, we analyze the impact of the two climate policies on the structure of the global and regional electricity mix and on the deployment of CCS and BECCS technologies.  Evolution of the electricity mix

Fig. 2 describes the global electricity mix evolution from 2020 to 2100 in view of the two climate scenarios investigated. In the BAU scenario, electricity generation increases from 17,934 TWh in 2005 to 35,334 TWh in 2050 and 57,505 TWh in 2100. The implementation of a carbon policy induces a change in the structure of the electricity mix and the wide deployment of carbon capture and storage technologies from 2030 in RF_2p6, and from 2070 in the less stringent scenario RF_3p7. In RF_2p6, in 2050, electricity from coal, gas and biomass plants with carbon capture and storage represents 31% of the mix, with nuclear and renewables accounting for 21% and 32% respectively. In 2050, 99% of biomass power plants, 97% of coal power plants, and 82% of gas power plants are equipped with CCS technology. This explains the low contribution of coal and gas without CCS in Fig. 2. This result is consistent with literature [13,31]. In 2100, there is significant deployment of renewables and nuclear energy. Electricity generated from nuclear and renewables multiplies by respectively 6 and 5 between 2020 and 2100. In RF_3p7, in 2050, coal remains the principal energy used in electricity production. It represents 59% of the mix, nuclear accounts for 20% and renewables for 8%. In 2100, 20% of electricity is produced with CCS, 35% with nuclear and 33% with renewables. We also observe that as the climate constraint becomes more severe, electricity production increases in 2100. Two factors come into play: the first is that producing more electricity is a way of capturing and storing more CO2 especially when CCS is applied to biomass power plants [39], the second is that the model adopts more electric vehicles in the transport sector in the RF_2p6 scenario in 2100. The extra cost in power generation gets balanced by the reduction in fuel usage in the transport sector resulting to cost diminishing. The regional analysis shows that fast developing countries (China and India) rely heavily on coal in BAU. In both climate scenarios, they develop largely renewable energies and CCS on coal because of its low cost in these regions, while developing countries favor nuclear energy. Industrialized countries turn to renewable energies, nuclear power and CCS technologies.  CCS and BECCS deployment The following graph focuses on the CCS technologies developed at global level under the two climate policies.

Fig. 1. Resulting atmospheric concentration (ppm CO2) and CO2 emissions (Gt CO2).

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

971

80000 70000 60000

RENEWABLES BIOMASS

50000

HYDRO

40000

NUCLEAR 30000

OIL AND GAS COAL

20000

ELC WITH CCS

10000 0 2020

2050

2100

2020

BAU

2050

2100

RF_2p6

2020

2050

2100

RF_3p7

Fig. 2. World electricity production (TWh).

CCS on coal and biomass power plants becomes a competitive option in 2030 in RF_2p6 and in 2070 in RF_3p7. CCS on gas power plants enters the market in 2040 in RF_2p6 and only in 2090 in RF_3p7 Fig. 3. The results show the role CCS plays in the electricity mix. In scenario RF_2p6, between 2030 and 2100, the average annual growth rate of electricity produced with a CCS technology is 3%. In 2100, 55% of CCS technologies are applied to fossil resource power plants and 45% to solid and crop biomass power plants. 4.8 Gt of CO2 are captured and stored in geological formations in 2050 and 12 Gt in 2100. In scenario RF_3p7, electricity produced with CCS reaches 12,244 TWh in 2100 and 8 Gt are captured per year in the same period. 61% of technologies are applied to coal power plants and 39% to biomass power plants. We analyze CCS deployment in industrialized, fast developing and developing countries. In industrialized countries, in scenario RF_2p6, CCS enters the market in 2030 on both conventional pulverized coal power plants and solid biomass direct combustion. Electricity from plants with carbon capture and storage increases by 20% per year between 2030 and 2050. It peaks in 2070 and then decreases over time (it is mainly the share of coal and gas þ CCS that decreases) (Fig. 4). Given the progressively stringent climate obligations, the low but

not nil CO2 emissions rate of coal and gas þ CCS is compensated by the increase of BECCS and other totally carbon-free technologies, such as nuclear. In 2100, 1.9 Gt and 1.7 Gt of CO2 are captured and stored per year respectively in RF_2p6 and RF_3p. In industrialized countries, CCS is mainly applied to biomass power plants after 2050 to satisfy both environmental constraints. In RF_2p6, 51% of the deployed CCS is applied to bio plants in 2050; this share increases and reaches 86% in 2100. In Europe (WEU þ EEU) and the USA, CCS on bio plants represents 35% and 55% respectively of total CCS deployment in 2050, and more than 80% in 2100. Australia, Canada and Japan only develop CCS on bio plants to meet the global objective. In RF_3p7, in 2100, CCS is also essentially applied to bio plants. In fast-developing countries, the CCS power share of electricity production is multiplied by almost 3 from 2030 to 2050 and by 1.7 from 2050 to 2100 in RF_2p6 (Fig. 4). In RF_2p6, 54% of CCS is applied to fossil power plants in 2050 and 66% in 2100. In the less stringent scenario, CCS is deployed at 72% on coal power plants. CCS is mainly developed on coal power plants in China because of its low cost and abundance, whereas India relies more on BECCS than on fossil CCS. About 5 Gt of CO2 are captured and stored in India and China in 2100. In developing countries, in 2050, half of the CCS deployed is applied to bio plants, and in 2100 CCS is more developed on fossil

9000 8000 7000

COAL. Conventional Pulverized Coal+Oxyfueling

6000

BIO. Crop Direct Combustion + postcombustion

5000 4000

BIO. Solid Biomass Direct Combustion + postcombustion

3000

GAS. NGCC+postcombustion

2000

GAS. NGCC+Oxyfueling

1000 0 2050

2100 RF_2p6

2100 RF_3p7

Fig. 3. World electricity production from CCS technologies (TWh).

972

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

Fig. 4. Electricity production from CCS technologies (TWh) in industrialized, fast developing and developing countries.

resources in developing countries due to the limited potential of biomass (Fig. 4). However, differences exist between countries. Countries that are well endowed with biomass only develop BECCS. This is the case for Africa, South America and Central America. However, countries with abundant gas mainly develop CCS. In the Middle East and in the former Soviet Union, CCS is mainly used on gas-fired power plants by 2050 and coal plants in 2100. Those regional results concur with the study [31]. Developing countries store 3 Gt of CO2 in geological formations by the end of the century in the most stringent scenario. The next subsection evaluates the impact of the availability of CCS, BECCS and co-firing technologies on the electricity mix and on CCS technology substitution in the most stringent scenario (RF_2p6). 3.3. A technology availability analysis in a climate-constrained economy Figs. 5 and 6 show the structure of the electricity mix for 2050 and 2100 and the deployment of CCS under the technology scenarios for the climate scenario (RF_2p6). When CCS technologies are not available (RF_2p6_NoCCS) the environmental objective is achieved by the rapid deployment of renewable energies.

In 2050, the fact that biomass-CCS technology is not available does not change the structure of the electricity mix. In RF_2p6 and RF_2p6_NoBECCS, the contribution of CCS is about 40% of the electricity mix. In RF_2p6, we can see that CCS technology is equally applied on biomass and fossil power plants (Fig. 6). When BECCS is not available there is a switch from biomass plants to mainly cofiring plants (Fig. 6). However, in 2100, comparing RF_2p6 and RF_2p6_NoBECCS, the CCS share in the power mix increases. Electricity produced from CCS rises from 15,062 TWh in RF_2p6 to 22,169 TWh in RF_2p6_NoBECCS. In this last scenario, CCS is applied to pulverized coal and air-blown IGCC (integrated gasification combined cycle) co-firing plants (Fig. 6). When neither CCS on biomass nor CCS on co-firing power plants are available (RF_2p6_NoBECCSCF), the CCS share in the mix, compared to its share in scenario RF_2p6, decreases slightly in 2050 (11%) and considerably in 2100 (76%). In 2100, 15,060 TWh of electricity is produced with a CCS technology in RF_2p6 compared to only 3550 TWh in RF_2p6_NoBECCSCF. This decrease is mainly compensated by a large increase of renewable energies. Why does CCS on coal power plants not act as a substitute for BECCS and cofiring technologies? CCS on coal is penalized under a stringent emissions reduction scenario because it does not allow a zero rate of CO2 emissions compared to totally carbon-free technologies, such as renewable.

Fig. 5. World electricity mix in TWh.

S. Selosse, O. Ricci / Energy 76 (2014) 967e975 25000

973 COAL.Conventional Pulverized Coal+oxyfueling COAL.BIO.Pulverized Coal.CoFiring+postcombustion

20000

COAL.BIO.Air Blown IGCC.Co-Firing + precombustion

15000

GAS.NGCC+postcombustion 10000 GAS.NGCC+oxyfueling 5000 BIO.Sld Biomass Direct Combustion+postcombustion 0 2050

2100 RF_2p6

2050

2100

RF_2p6_NoBECCS

2050

2100

RF_2p6_NoBECCSCF

BIO.Crop Direct Combustion+postcombution

Fig. 6. World electricity production from CCS technologies (TWh).

The co-firing option appears to be a good solution to meet the climate target when CCS on bio plants is not available, notably IGCC power plants. IGCC systems are one of the most efficient clean-coal power production technologies. They have the advantage of processing many kinds of feedstock, such as coal, coke, residual oil, biomass and municipal waste. In principle, IGCC technology is the cheapest option for CCS [43]. However, IGCC plants are more expensive and less reliable than supercritical conventional pulverized coal power plants (in the model about 15% more expensive). There is no consensus on which option will cost the least in the future. In the model, the cost of the plants with capture is fairly similar for both technologies. In Fig. 6, we can see that NGCC (natural gas combined cycle) power plants with CCS also play a role in CCS deployment. NGCC is a mature technology. It was first introduced in the 1990s. Since then, the technology has made progress with cooling and materials development. As a result, efficiency has now reached 60% on a lower heating value basis, compared to standard gas-fired power plants in 2003, when the average efficiency was around 42%. NGCC emits less than half as much CO2 as coal-fired power plants per unit of electricity, making it an interesting option when co-firing and BECCS technologies are not available (RF_2p6_NoBECCSCF). The constraint on the availability of negative emissions technologies affects the evolution of the electricity mix structure. What, then, is its impact on the carbon price and on the total cost of the energy system? 3.4. Economic implications: carbon price and abatement cost The price of the carbon needed to induce the changes described above rises rapidly in all scenarios. The most rapid increase occurs in the RF_2p6_NoCCS from 12 USD/tCO2 in 2005 to 28 USD/tCO2 in 2020, and around 150 USD/tCO2 in 2050. Table 1 shows the carbon price for the different scenarios. To reach the overall objective of 2.6 W/m2, the carbon price needs to be almost twice as high in 2100 when CCS is not available.

When BECCS is not available (RF_2p6_NoBECCSCF), it needs to be about 1.5 times higher. Fig. 7 presents the cost of the different stabilization and technological scenarios. It is expressed as the net present value cost of additional mitigation expenditure compared to the BAU scenario (2005e2100) discounted at 5% in trillions (1012) of USD2005. The additional global discounted cost over the period 2005e2100 (in absolute terms, compared to BAU) is USD 6234 billion in RF_2p6 and USD 1018 billion in RF_3p7. In relative terms, this represents an increase of 2.9% in RF_2p6 and 0.5% in RF_3p7. In both climate scenarios, fast developing countries' abatement costs are very high compared with other regions because the mitigation potential at relatively modest costs is greater in those regions. In the most stringent climatic scenario (RF_2p6), compared to the BAU, the discounted cost of the system increases by 12% for China, 8% for India and only 2.5% for Western Europe, 1.8% for the USA and 2.4% for Japan. Clearly, the option of using BECCS reduces the cost of meeting the RF_2p6 stabilization target. We can expect that the lower the stabilization target, the more significant the contribution made by BECCS will be in reducing costs as negative emissions become critical. In line with literature, we show that the availability of BECCS also has an impact on the least cost emission reduction pathway towards the long-term concentration objective [2,9,15]. The availability of BECCS in RF_2p6 increases time flexibility and postpones emission reduction until after 2030. The CO2 emission peak is around 2030, after which stringent abatement occurs between 2030 and 2050 (scenario RF_2p6), whereas when BECCS is unavailable, CO2 emissions steadily decrease from 2010. Conversely, in the RF_3p7 scenario, the global discounted cost of the energy system for achieving the carbon target does not change much, whatever the CCS technology availability. Note however a higher additional cost in case of the total unavailability of CCS. Low-carbon transition tends to be based more on other clean technologies. At regional level, when CCS is not available and the climate policy is ambitious (RF_2p6_NoCCS), the discounted cost increases

Table 1 Carbon prices under different scenarios (USD/tCO2).

2005 2020 2040 2050 2060 2080 2100

RF_2p6

RF_2p6 NoBECCS

RF_2p6 NoBECCSCF

RF_2p6 NoCCS

RF_3p7

RF_3p7 NoBECCS

RF_3p7 NoBECCSCF

RF_3p7 NoCCS

7 16 48 85 153 584 689

10 22 68 121 216 820 993

10 24 75 132 236 897 1085

12 28 87 153 273 1041 1244

1 2 7 12 21 75 111

1 2 7 13 23 83 126

1 2 8 13 24 86 130

1 3 8 14 26 92 145

974

S. Selosse, O. Ricci / Energy 76 (2014) 967e975

10 9 8 7 6

BECCS+CCS+CF

5

NoBECCS

4

NoBECCSCF

3

NoCCS

2 1 0 RF_2p6

RF_3p7

Global

RF_2p6

RF_3p7

Industrialized

RF_2p6

RF_3p7

Fast developing

RF_2p6

RF_3p7

Developing

Fig. 7. Additional discounted costs compare to the BAU (trillions USD2005).

for all regions. The scenario RF_2p6_NoBECCS mainly impacts countries that develop BECCS as a priority, such as Africa, Australia and Japan, as well as India which relies heavily on BECCS in its electricity production. 4. Conclusion In this paper we focused on BECCS technology, which has been acknowledged as an interesting negative emission option for achieving major CO2 emissions reductions. We therefore evaluated the role of power generation from bio plants with CCS in achieving a low climate target by 2100 using the optimization model TIAMFR. We conducted a regional analysis to understand where the technology will be developed. We also studied what impact the unavailability of this technology would have on the structure of the electricity mix and the cost of the system. Under a stringent climate control target, the model is favorable to the widespread deployment of CCS technologies in the power sector from 2030. 30% of the electricity generated in 2050 comes from plants equipped with CCS technology. At a global level, 50% of the CCS deployed is associated with bio plants and 50% with fossil plants, with a preponderance of coal power plants. The regional analysis suggests that industrialized countries will develop CCS mainly at their bio plants, whereas in fast-developing countries, principally China, CCS will be applied to coal power plants. Developing countries will take different approaches. Some will only develop BECCS, for instance: Africa, South and Central America and Japan, whereas the Middle East and the former Soviet Union, which possess abundant gas resources, will develop CCS at their gas power plants. It is important to keep in mind that the possible contribution of BECCS depends heavily on the potential and societal acceptance of bioenergy on one hand, and the deployment of capture and storage technologies on the other. Although there may be significant potential for this technology, uncertainties and concerns remain regarding technology development, carbon-negative life cycle assessment, food security, and biodiversity [44]. Therefore, this study also looked at what happens when BECCS is not available in the long run. The results show that with a specific constraint on CCS diffusion, the share of renewables and nuclear energy becomes significant to meet the climate target. Moreover, co-firing options tend to be a good substitute for CCS at bio plants when this last option in not available. But more importantly, when co-firing technologies and BECCS are not available, the share of CCS in the electricity mix significantly decreases in 2100. CCS on fossil fuel does not

compensate for the absence of carbon-negative technologies as it is not a carbon-neutral technology. In line with literature, our scenarios reveal that while a broad range of mitigation technologies are needed to attain low concentration targets, the availability of BECCS enhances the possibility of decreasing the cost of meeting those targets. The emissions reduction pathway strongly depends on technology assumptions. If negative emissions are available, less abatement takes place in the short term, and more aggressive action occurs later. Acknowledgment This research was supported by the Chair Modeling for sustainable development, coordinated by MINES ParisTech, Ecole des Ponts ParisTech, AgroParisTech, and ParisTech, and supported by ADEME, EDF, RENAULT, SCHNEIDER ELECTRIC and TOTAL. References [1] IEA. World energy outlook. Paris: IEA/OECD; 2011. [2] Azar C, Lindgren K, Obersteiner M, Riahi K, Van Vuuren D, Michel K. The feasibility of low CO2 concentration targets and the role of bio-energy with carbon capture and storage (BECCS). Clim Change 2010;100:195e202. [3] Viebahn P, Daniel V, Samuel H. Integrated assessment of carbon capture and storage (CCS) in the German power sector and comparison with the deployment of renewable energies. Appl Energy 2012;97:238e48. [4] Mcglashan RN, Workman MHW, Caldecott B, Shah N. Negative emissions technologies. Grantham Institute for Climate Change briefing paper No8, Imperial College London; 2012. [5] Keith DW. Why capture CO2 from the atmosphere? Science 2009;325(5948): 1654e5. [6] American Physical Society, editor. Direct air capture of CO2 with chemicals: a technology assessment for the APS panel on public affairs. Washington DC: American Physical Society; 2011. accessible at: www.aps.org/policy/reports/ assessments/upload/dac2011.pdf. €llersten K, Moreira J, Nilsson S, et al. [7] Obersteiner M, Azar C, Kauppi P, Mo Managing climate risk. Science 2001;294(5543):786e7. [8] Fisher BS, Nakicenovic N, Alfsen K, Morlot JC, de la Chesnaye F, Hourcade J, et al. Issues related to mitigation in the long term context. In: Metz B, Davidson OR, Bosch PR, Dave R, Meyer LA, editors. Climate change: mitigation. Contribution of working group III to the fourth assessment report of the intergovernmental panel on climate change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press; 2007. [9] Clarke L, Edmonds J, Krey V, Richels R, Rose S, Tavoni M. International climate policy architectures: overview of the EMF 22 international scenarios. Energy Econ 2009;31(Suppl. 2):S64e81. €llersten K. Carbon capture and storage from [10] Azar C, Lindgren K, Larson E, Mo fossil fuels and biomass costs and potential role in stabilizing the atmosphere. Clim Change 2006;74:47e79. [11] Edenhofer O, Knopf B, Barker T, Baumstark L, Bellevrat E, Chateau B, et al. The economics of low stabilization: model comparison of mitigation strategies and costs. Energy J 2010;31:11e48. [12] Katofsky R, Stanberry M, Frantzis L. Achieving climate stabilization in an insecure world: does renewable energy hold the key?. Report International Energy Agency e Renewable Energy Technology Development; 2010.

S. Selosse, O. Ricci / Energy 76 (2014) 967e975 [13] Luckow P, Wise M, Dooley JJ, Kim S. Large-scale utilization of biomass energy and carbon dioxide capture and storage in the transport and electricity sectors under stringent CO2 concentration limit scenario. Int J Greenh Gas Control 2010;4:865e77. [14] van Vuuren D, Den Elzen M, Lucas P, Eickhout B, Strengers B, Ruijven B, et al. Stabilizing greenhouse gas concentrations at low levels: an assessment of reduction strategies and costs. Clim Change 2007;81:119e59. [15] van Vuuren DP, Stehfest E, den Elzen MG, van Vliet J, Isaac M. Exploring IMAGE model scenarios that keep greenhouse gas radiative forcing below 3 W/m2 in 2100. Energy Econ 2010;32(5):1105e20. [16] van den Broek M, Veenendaal V, Koutstaal P, Turkenburg W, Faaij A. Impact of international climate policies on CO2 capture and storage deployment: illustrated in the Dutch energy system. Energy Policy 2011;39:2000e19. [17] Lemoine DM, Fuss S, Szolgayova J, Obersteiner M, Kammen DM. The influence of negative emission technologies and technology policies on the optimal climate mitigation portfolio. Clim Change 2012;113:141e62. [18] van Vliet J, den Elzen MGJ, van Vuuren DP. Meeting radiative forcing targets under delayed participation. Energy Econ 2009;31:S152e62. [19] Krey V, Riahi K. Implications of delayed participation and technology failure for the feasibility, costs, and likelihood of staying below temperature targetsdGreenhouse gas mitigation scenarios for the 21st century. Energy Econ 2009;31:S94e106. [20] Hektor E, Berntsson. Future CO2 removal from pulp mills e process integration consequences. Energy Convers Manag 2007;48:3025e33. €llersten K, Gao L, Yan J. CO2 capture in pulp and paper mills: CO2 balances [21] Mo and preliminary cost assessment. Mitig Adapt Strat Glob Change 2006;11: 1129e50. €llersten K, Yan J, Moreira JR. Potential market niches for biomass energy [22] Mo with CO2 capture and storage e opportunities for energy supply with negative CO2 emissions. Biomass Bioenergy 2003;25:273e85. [23] Kheshgi HS, Prince RC. Sequestration of fermentation CO2 from ethanol production. Energy 2005;30:1865e71. [24] Mathews JA. Carbon-negative biofuels. Energy Policy 2008;36:940e5. [25] Lindfeldt E, Westermark MO. System study of carbon dioxide (CO2) capture in bio-based motor fuel production. Energy 2008;33:352e61. [26] Lindfeldt E, Westermark MO. Biofuel production with CCS as a strategy for creating a CO2-neutral road transport sector. Energy Procedia 2009;1: 4111e8.

975

[27] Laude A, Ricci O, Bureau G, Royer-Adnot J, Fabbri A. CO2 capture and storage from a bioethanol plant : carbon and energy footprint and economic assessment. Int J Greenh Gas Control 2011;5:1220e31. [28] Carpentieri M, Corti A, Lombardi L. Life cycle assessment (LCA) of an integrated biomass gasification combined cycle (IBGCC) with CO2 removal. Energy Convers Manag 2005;46:1790e808. [29] Rhodes JS, Keith DW. Engineering economic analysis of biomass IGCC with carbon capture and storage. Biomass Bioenergy 2005;29(6):440e50. [30] Uddin SK, Barreto L. Biomass-fired cogeneration systems with CO2 capture and storage. Renew Energy 2007;32:1006e19. [31] Ricci O, Selosse S. Global and regional potential for bioelectricity with carbon capture and storage. Energy Policy 2013;52:689e98. [32] IEA. Energy technology perspectives. Paris: IEA/OECD; 2012. [33] Herzog HJ. Scaling up carbon dioxide capture and storage: from megatons to gigatons. Energy Econ 2011;33:597e604. [34] Dubreuil A, Assoumou E, Bouckaert S, Selosse S, Maïzi N. Water modeling in an energy optimization framework e the water-scarce middle east context. Appl Energy 2013;101:268e79. [35] Ricci O, Selosse S. A cost analysis of the copenhagen emission reduction pledges. Econ Bull 2013;33(1):764e71. [36] Loulou R, Labriet. ETSAP-TIAM: the TIMES integrated assessment model (part I: model structure). Comput Manag Sci 2008;5:7e40. [37] Loulou R. ETSAP-TIAM: the TIMES integrated assessment model. Part II: mathematical formulation. Comput Manag Sci 2008;5:41e66. [38] Ystanes Føyn TH, Karlsson K, Balyk O, Grohnheit PE. A global renewable energy system: a modelling exercise in ETSAP/TIAM. Appl Energy 2011;88:526e34. [39] Loulou R, Labriet M, Kanudia A. Deterministic and stochastic analysis of alternative climate targets under differentiated cooperation regimes. Energy Econ 2009;31:S131e43. [40] van Vuuren D, van Vliet J, Stehfest E. Future bio-energy potential under various natural constraints. Energy Policy 2009;37:4220e30. [41] IPCC. Renewable energy sources and climate change mitigation, special report. Chapter 2. Cambridge: Cambridge University Press; 2011. [42] Nordhaus WD, Boyer J. Roll the dice again: economic models of global warming. Yale University; 1999. manuscript edition. [43] IEA-ETSAP. CO2 capture & storage. Technology brief E14. October; 2010. [44] van Vuuren DP, Bellevrat E, Kitous A, Isaac M. Bio-energy use and low stabilization scenarios. Energy J 2010;31:193e221.