Co-firing Coal with Biomass under Mandatory Obligation for Renewable Electricity: Implication for the Electricity Mix.
In the last few years, co-firing coal with biomass has become very popular in the European power sector, where firms have to comply with stringent policies to reduce C[O.sub.2] emissions and increase their output of renewable electricity. Co-firing provides short-term opportunities for increasing the share of renewable energy sources (RESs) and reducing C[O.sub.2] emissions in a very cost effective way through conventional technologies that are not subject to problems of intermittency and that do not require additional investments.
In addition to exemption from surrendering C[O.sub.2] allowances under the European Union Emission Trading Scheme (EU ETS) when burning biomass (equivalent to a zero emission factor), several European states have implemented arrangements to include co-firing in their support schemes for renewable electricity (e.g. Poland, UK, Denmark, Netherlands), which raised concerns about the consequences for the contribution from coal to the electricity mix (even through co-firing with biomass) and the resulting C[O.sub.2] emissions. As recently pointed out in debates on energy agreements in the Dutch parliament, it may seem strange that some coal plants are set to close down due to environmental regulation while the same units can receive subsidies when co-firing biomass. This raises questions about the actual incentives to invest in traditional RES technologies (e.g. wind, solar, dedicated biomass units) to meet European targets and the consequences for future energy mixes.
In the same way as for natural gas, co-firing has often been pointed to by practitioners from the energy industry as a transitional technology, which may limit the C[O.sub.2] emissions from existing power fleets in the short run, before deployment of a true energy transition in the long run, with a carbon-free power sector. However, whereas co-firing reduces the carbon intensity of coal plants (compared with the classical configuration when coal is the only input), it still generates C[O.sub.2] emissions. Hence, if coal is maintained and co-firing diverts investments away from traditional RES, it is to be expected that C[O.sub.2] emissions from electricity will be higher in the long run (compared with a more radical energy transition in which pure renewables dominate the fleet of power plants). Accordingly, any policy that promotes co-firing as a renewable may result in higher C[O.sub.2] emissions in the long run, if it gives incentive to use coal plants under co-firing instead of investing in the RES carbon-free technologies. In this context, this paper aims to investigate what would happen if co-firing, which is thought of as a transitory option by some (but also receives support from renewable schemes in some countries, see section 2), is continued and becomes a long-term solution. (1) We focus on European countries with simulations for France and Germany. However, beyond the case of Europe and other developed countries, we believe this paper may help to understand what could happen in developing and emerging countries if co-firing is used to develop RESs. Given the high share of coal in the generation mixes of these countries, this may be a big question in the future. This paper provides some insights that may help to anticipate this issue.
The question of the consequences of promoting co-firing as renewable electricity has attracted little attention in the economic literature. To date, to the best of our knowledge, the only contribution comes from Lintunen and Kangas (2010) who provides a theoretical model to analyze the effect of co-firing in a stylized and simplified power system. The authors consider profit maximizing power producers that optimize production and investment decisions so as to meet static power demand. Investments in renewables are modeled as demand for new wind turbines (the only renewable technology in the analysis) that can increase the production capacity with the aim of meeting the static power demand at the lowest cost. Results show that promoting co-firing as a renewable decreases investments in pure renewable technology, whereas the C[O.sub.2] intensity of electricity is not significantly impacted. However, although Lintunen and Kangas (2010) illustrate their results through a numerical application with parameters reflecting the Finnish power sector, they fail to provide comprehensive estimations of investment decisions over a complete power system management including elements such as a dynamic time horizon, the decommissioning of old capacities, or increasing demand for power. Notably, their modeling approach with a simplified power system cannot be used to investigate the consequences for long-term C[O.sub.2] emissions when the electricity mix is continuously modified by policies promoting co-firing against pure renewables and carbon-free technologies. Considering a more detailed treatment of power systems through simulations with a dynamic time horizon is likely to produce more significant effects when co-firing steadily displaces traditional RES technologies over time, resulting in a power plant fleet that is more carbon intensive in the end. This is the starting point of our analysis, which extends the previous contribution by Lintunen and Kangas (2010).
Compared with previous work, this paper uses a simulation approach to analyze the consequences for the electricity mix when co-firing is recognized as renewable electricity. We use the Green Electricity Simulate (GES) model, which is a simulation model for electricity designed to focus on biomass-based electricity and co-firing in European countries (Bertrand and Le Cadre, 2015). In order to assess the effect of promoting co-firing as a renewable option, we run the model with and without co-firing in the set of RES technologies that are accounted for to meet the RES targets. Our simulations rely on a detailed representation of the power system, which can be used to derive more general results taking into account elements such as a dynamic time horizon, the decommissioning of old capacities, rising demand for power or increasing renewable targets. This extends the study by Lintunen and Kangas (2010).
As an illustration, we provide simulations for France and Germany, which offer good cases of study for our analysis because they have large coal capacities (even in France where the coal capacity is not negligible in volume with respect to other European countries, see Table 3) and because no support scheme for co-firing has been implemented in these countries so far. Hence, France and Germany provide relevant counterfactuals with which to investigate the consequences of implementing such provisions that recognize co-firing as renewable electricity. The case of France is also interesting regarding the effect of nuclear reduction that may greatly impact the electricity mix in this country. Whereas French electricity has historically been highly dependent on nuclear power, a law on "energy transition" was passed in 2015, which aimed at reducing the share of nuclear power by 2025. (2) Although the full application remains uncertain, such a reduction of French nuclear power combined with the RES targets is likely to be offset by some RES power plants, to which co-firing may contribute if counted as a renewable. This is something of interest for our study. (3)
Results confirm that recognizing co-firing as an RES would jeopardize investments in traditional RESs, which would be largely ousted in favor of increased generation from existing coal power stations under co-firing plus some new investment in coal. The additional coal investments are more substantial in France because French coal capacities are lower than German capacities, thus limiting the scope for using existing coal plants to meet the RES targets through co-firing. The additional French coal capacities may reach 18 GW when the model is implemented with exogenous decommissioning of old nuclear power plants. Comparatively, the maximal additional coal capacity in Germany is close to 14 GW when co-firing is included in the set of RES, which corresponds to a progression of about 27% for coal in 2030 compared with the initial capacity, whereas the same progression is more than 243% in France when old nuclear power stations are decommissioned (107% when nuclear plants are prolonged), with almost 26 GW of coal accounting for 20% of the 2030 French capacity mix. Hence, including co-firing in RESs may more radically change the French capacity mix, in which coal may change status and become an important source of French electricity output.
Regarding C[O.sub.2] emissions, results indicate that recognizing co-firing as an RES generates sharp increases because of reduced traditional RESs (carbon-free) and more coal in electricity. This effect is more significant in Germany than in France due to its much greater coal capacities. Moreover, in the case of France, the magnitude of the carbon increase depends largely on the share of nuclear power, with fewer increases when old nuclear power stations are prolonged. Finally, we show that including co-firing in the set of RESs reduces the overall costs associated with managing the power system, because this allows compliance with the RES constraint through a conventional and low-cost option that does not require additional investments. When balancing this cost saving against the increased social cost from higher C[O.sub.2] emissions, results show that the cost saving may be dominated by the increased carbon cost with a high carbon valuation around 100 Euros per tC[O.sub.2]. An exception comes from France when the service life of ageing nuclear power stations is prolonged. In this case, the cost saving is very high and the increased C[O.sub.2] emissions are slight (because massive cheap and carbon-free nuclear power continues to be used for base-load generation) with the result that the cost saving always dominates the increased carbon cost.
The remainder of the paper is organized as follows. Section 2 gives an overview of existing support schemes for renewable electricity in European countries that include provisions for co-firing. In section 3, we provide an overview of the economic literature on co-firing and a brief presentation of the methodology and data. Section 4 presents the results and discussions. Section 5 concludes.
2. CO-FIRING IN THE RENEWABLE ELECTRICITY SUPPORT SCHEMES OF EUROPEAN COUNTRIES: AN OVERVIEW
The option to co-fire biomass with coal has been implemented in numerous European coal-fired power stations. Major co-firing applications include large coal plants such as Ferrybridge (2000 MW, UK), Fiddler's Ferry (2000 MW, UK), Amer (1000 MW, Netherlands), Gelderland (630 MW, Netherlands), Ensted (620 MW, Denmark), and Lagisza (460 MW, Poland). (4)
The treatment of co-firing in support schemes for renewable electricity is highly heterogeneous among European countries. In general, in most cases, co-firing is not counted as an RES, and, as such, it is not subsidized. However, there are notable exceptions to this, with some countries that generate significant amounts of electricity from coal having included provision for co-firing in their support schemes. Table 1 provides an overview of treatments for coal plants under co-firing in support schemes from different European countries.
In the UK, banding has been introduced awarding different co-firing configurations at various rates of certificates. Whereas 1.5 Renewables Obligation Certificates (ROCs) are given for each MWh of electricity generated in dedicated biomass units, the ROC rate (ROC per [MWhe.sub.lec]) is less than one when co-firing is involved. The rate ranges from 0.3 to 0.9 depending on the percentage of biomass co-fired (Table 2).
The UK system used to be more generous regarding co-firing, with one ROC per [MWh.sub.elec] of co-fired electricity regardless of the configuration. In order to avoid excessive development of co-firing in the country, banding has been introduced so as to limit the level of subsidy. Nevertheless, even with the banding system, co-firing biomass with coal still tends to be more cost effective than investing in new dedicated biomass units.
In the Netherlands, another country with high co-firing, the SDE+ (Stimulering Duurzame Energie) auction subsidy-system for renewables provides producers of co-fired electricity with grants as for other RESs. The SDE+ was introduced in 2015 and basically the (sealed-bid) auction gives bonus payments to compensate for the difference between the market prices for electricity (which are based on fossil fuel sources) and the electricity cost from RESs. (5) The scheme works with multiple bidding phases (nine in 2015 and four in 2016), with a budget cap and a maximal premium for each technology and phase. Each bidder submits a (bid) premium (lower than the maximal) and a level of output. For each technology, the auction continues until the budget is reached. Bidders with the lowest bids are served first, and they receive the premium they bid. In the co-firing category, producers can bid for a maximum premium of 107 Euros/[MWh.sub.elec] for a period of eight years (Netherlands Enterprise Agency, 2016; AURES-Ecofys, 2016).
During the 2016 auctions, co-firing units were among the biggest winners. For example, in the first phase of July, several coal stations received around 1.5 billion Euros to co-fire up to 50 percent biomass for a total SDE+ budget of 8 billion Euros in 2016. However, whether these subsidies for co-firing will actually be implemented or not remains uncertain because of the Dutch government's plans to close all coal stations by 2020, which are still under debate. (6)
In the context of our paper, we choose to focus on France and Germany rather than directly considering those countries with RES supports given to co-firing. These two countries offer useful cases for our analysis because they have substantial coal capacities and no subsidy for co-firing, meaning they provide relevant counterfactuals with which to investigate the consequences of implementing such provisions for co-firing in RES support schemes.
Plainly Germany uses a far larger proportion of coal for electricity generation than France. However, even though coal makes up a rather small share of French electricity, the associated volumes are quite significant compared with other European countries in which coal is known as an important source of electricity (Table 3)
3. SIMULATION METHODOLOGY
3.1 The economics of co-firing: A literature overview
The economic literature about co-firing mainly relies on three approaches: Estimations of technical potential given by existing fleets in electricity (Berggren et al., 2008; Hansson et al., 2009; Bertrand et al., 2014), calculations for profitability of co-firing depending on representative market conditions (Bertrand et al., 2014; Xian et al., 2015; Mei and Wetzstein, 2017), and theoretical framework with stylized power system (Lintunen and Kangas, 2010). (7)
Among the papers that investigate the technical potential of co-firing (i.e. the question of how much biomass can be used in co-firing and electricity, and the related C[O.sub.2] abatements), Berggren et al. (2008) focuses on matching the potential biomass supply in Poland with estimated opportunities for biomass co-firing in the existing Polish coal plants. The authors also derive the C[O.sub.2] abatements from co-firing. Results indicate that about 4 Mt of C[O.sub.2] can be abated each year in Poland through biomass co-firing. Hansson et al. (2009) estimate the potential power generation from co-firing in the existing European coal plants. They report that co-firing can generate yearly electricity production from 50 to 90 [TWh.sub.elec] in the EU-27. However, as opposed to Berggren et al. (2008) for Poland, the authors do not provide comparisons of their results with the potential biomass supply, and they do not compute the C[O.sub.2] abatements from co-firing. Bertrand et al. (2014) extend these works by estimating the potential biomass demand from both co-firing and dedicated biomass power plants in the EU-27, and they compare the results with the potential biomass feedstocks in Europe. Moreover, the authors also compute C[O.sub.2] abatements associated with co-firing. Results show that co-firing offers the highest potential demand in electricity with up to 80 percent of the overall potential demand from power. Additionally, depending on the assumptions on co-firing and biomass availability in Europe, the EU-27 potential demand is found to account for 8 to 148 percent of the European potential supply. Regarding C[O.sub.2] emissions, results indicate that implementing co-firing in existing coal plants can produce high volumes of abatements, with a maximal potential estimated to be around 360 Mt of C[O.sub.2] per year in the EU-27.
A second strand of literature focuses on estimating the profitability of co-firing depending on market conditions. Bertrand et al. (2014) compute the biomass and C[O.sub.2] breakeven prices for co-firing in the European context with carbon pricing through the EU ETS. These values reflect the economic conditions that make co-firing profitable depending on prices for coal, biomass and C[O.sub.2]. Results indicate that co-firing is profitable with a biomass price that is in the range of 16 to 24 (25 to 35, respectively) Euros per [MWh.sub.prim] when the carbon price is 20 (50, respectively) Euros per tonne of C[O.sub.2]. Another contribution comes from Xian et al. (2015) and Mei and Wetzstein (2017), who analyze the competitiveness of co-firing for coal plants in Georgia, United States, over the period 2009-2014. Results demonstrate that co-firing is not a profitable option under the current US market conditions, with free carbon emissions and competition of low-cost shale gas. Hence, the government must implement incentive schemes if the development of co-firing is an objective. The timing of policy also matters, with early incentives needed to avoid adoption of alternative energies through shale gas. The authors conclude that a subsidy of 1.40 US Dollars per MMBTU on co-fired biomass or a tax of 1.50 US Dollars per MMBTU on coal would be needed to trigger the development of co-firing in the United States. Given an average cost of 0.12 US Dollars per k[Wh.sub.elec] for electricity in the United States, this would represent about 4 percent of the electricity cost. (8)
A last contribution to the economics of co-firing comes from Lintunen and Kangas (2010), and it provides a theoretical framework in which to analyze the effect of co-firing in a stylized and simplified power system. Their paper aims to assess the effect for investments in pure renewable electricity plants and C[O.sub.2] emissions when promoting co-firing as a renewable option. The authors consider profit maximizing power producers that optimize production and investment decisions so as to meet static power demand over different sub-periods reflecting load profile. The model considers exogenous hydro and nuclear power generation, so that the decisions satisfies only the residual demand that is left to other technologies. Investments in renewables rely on wind turbines only, and all other renewable technologies are excluded from the analysis. In this case, investment decisions are modeled as demand for new wind turbines that can increase the production capacity with the aim of meeting the static power demand at the lowest cost. The authors provide a numerical application with parameters reflecting the Finnish power sector. Results show that promoting co-firing as a renewable decreases investments in pure renewable technology, whereas the C[O.sub.2] intensity of electricity is barely impacted. Given the qualitative shape of results regarding investments in renewables, it is very likely that a more general treatment of the power sector (with a dynamic time horizon, the decommissioning of old capacities, rising demand for power, increasing renewable targets, etc.) may modify the impact for C[O.sub.2] emissions when co-firing is considered as a renewable. This is the starting point of our analysis with the GES model.
Compared with the aforementioned literature, the GES model is an electricity simulation model that considers a detailed representation of power system and co-firing based on actual data rather than representative situations. (9) This can be used to investigate the effects associated with co-firing from a more accurate perspective, when considering possible evolutions of policy and other energy-related decisions over time. Hence, the GES model offers a flexible tool for analyzing the effect of co-firing in combination with policy scenarios and other real problems that apply to the power sector.
3.2 Model description
GES is a dynamic simulation model that is designed to investigate questions related to biomass-based electricity in European countries, with a special focus on biomass co-firing in coal plants. The model minimizes the overall cost of electricity (generation and investment), over the 2010-2030 time interval with a range of economic, technical, and legal constraints: capacity (generation [less than or equal to] available capacity), market clearing for electricity, share of RES in power generation, physical constraints associated with co-firing (loss in efficiency of coal plants and percentage of biomass that can be co-fired depending on the resource quality), etc. In this work, we use the French and the German modules from the 1.0 version (Bertrand and Le Cadre, 2015). Extensive documentation on the model is provided as supplementary materials (see Appendix A).
For each year in the considered time interval, the model determines the power generation mix (based on a merit order logic) and investment decisions so as to meet electricity demand at the least cost. It computes the optimal dispatch of generating capacities into intra-annual hourly time slices with unequal power demand. This reflects different load levels associated with more or less electricity demand.
The modeling framework can also be used to investigate the consequences of modifications in generating capacities through investments in new power stations and decisions regarding decommissioning or prolongation of old units that have exceeded their theoretical lifetime. (10) Hence, the structure of the fleet is made flexible, allowing any change in the electricity mix in favor of biomass to be analyzed with a degree of flexibility that depends on relative prices and technological and legal aspects. Figure 1 provides an overview of the model framework.
Investments mainly rely on comparisons between the LLCOEs of different power technologies, which evolve through time with the price trends for fuels and carbon. However, as a simplification, we assume that the cost and technical parameters (see section 3.3) remain constant throughout the time interval (but the cost parameters are still subject to discounting). As a consequence, the modeling does not consider the effect of technological progress, which may reduce the cost of investing in renewable technologies compared with others as time goes on. (11) Technically, such investigations may be implemented by applying learning rates to the cost parameters. In this case, a well-sound analysis would need to run a sensibility analysis on the learning rate coefficients, which are themselves subject to uncertainties and are likely to evolve over time. Since our focus is on the prospective analysis of policies around the treatment of co-firing, we choose not to investigate such evolutions of cost parameters, which would multiply the number of scenarios to consider. Hence, we neglect the effect of technological progress in order to keep the prospective analysis of policies more compact so as to make the results more tractable. Predictably, a likely consequence of dropping this assumption is that renewable technologies will gain in profitability thanks to technological progress, which would increase investments in these technologies, ceteris paribus, and reduce the magnitude of the crowding-out effect (replacement of traditional RESs by co-firing) with lesser induced effects. This would not invalidate our analysis, but just diminish the magnitude of the results.
3.3 Data and model calibration
The dataset for the power system is based on a literature review providing representative values for cost and technical parameters associated with different power technologies of varying vintages: efficiency rates of power plants, load-factors, fixed and variable operation and maintenance costs, refurbishment costs, decommissioning costs, theoretical lifetimes (depending on whether stations have been prolonged or not), etc. (12)
In order to derive realistic projections, the model has been calibrated to actual market data for the base year (i.e. 2010). We focused on reproducing the observed yearly generation by fuel through iterative adjustments of availability and marginal costs so as to best replicate the French and German power generation mix as given by (RTE, 2011) and Eurelectric (2011). Such model calibration is a standard exercise in simulation, which is necessary to avoid simulation results departing too much from actual data. In particular, as pointed out in previous studies, simulations relying on unadjusted models are likely to generate errors in estimations derived from uncorrected power generation. For instance, estimating C[O.sub.2] emissions based on (simulated) uncorrected power generation can lead to significant bias in abatement estimates due to divergences in the utilization of power technologies with varying carbon intensity compared with real world responses under similar conditions (Delarue et al., 2010; Weigt et al., 2013; Solier, 2014).
Coal (bituminous), gas, oil, and carbon prices are based on the Current Policy Scenario (CPS) from IEA (2012) and other fuel prices are derived from the literature review. In all cases, the model considers price trends that are indexed on the Average Annual Growth Rates (AAGR) from the IEA-CPS scenario as well as other projections (from different references) reflecting specific evolutions in other fuel industries (uranium, lignite, solid biomass, biogas, bio-liquids, and mixed grade waste). (13)
The annual electricity demand is obtained from the 2010 ENTSO-E values to which we apply the AAGR from the IEA-CPS scenario to compute projections over the time interval. (14) The resulting yearly demands are then disaggregated on hourly levels, using weighting coefficients reflecting intra-annual time slices of varying length and power load (Bertrand and Le Cadre, 2015).
Regarding the installed capacities for power plants, the model uses data from the World Electric Power Plants (WEPP) data base by Platts, which provides a global inventory of electric power stations with information such as location, year of commissioning, size, etc. In the case of Germany, the data has been completed with a listing of planned nuclear decommissioning to account for the 2011 decision by the German government to shut down all the country's nuclear power plants by 2022 (Appendix B). This allows us to include exogenous reduction of nuclear capacity in the data for the model in line with the German nuclear phase-out plan. In order to investigate the effect of reductions in French nuclear capacity (in line with the French nuclear strategy enacted by France's energy transition law of 2015), we have included an additional constraint in the model that proscribes prolongation of out-of-lifetime nuclear power stations. This is equivalent to exogenous decommissioning of old nuclear power plants.
3.4 RES OBLIGATIONS AND CO-FIRING
In order to investigate the question of how co-firing may impact the electricity mix if it is recognized as an RES, we run simulations with and without co-firing in the set of RES technologies that are accounted for to meet the RES targets. As a simplification, we assume that only the biomass part from the primary energy in coal plants is accounted for as an RES. Hence, we run the model by considering either equation (1a) or (2a), depending on whether co-firing is included or not in the set of RESs:
[summation over (u[member of]URES)] [P.sup.i.sub.t,u][greater than or equal to] [[tau].sup.i.sub.2020]x([summation over (u[member of]U)][P.sup.i.sub.t,u]), (1a)
[summation over (u[member of]URES)][P.sup.i.sub.t,u]+[summation over (u[member of]UC)][summation over (b[member of]FSB)]([[eta].sup.cf.sub.u,b][F.sup.i.sub.t,u,b])[greater than or equal to][[tau].sup.i.sub.2020]x([summation over (u[member of]U)][P.sup.i.sub.t,u]), (2a)
where [P.sup.i.sub.t,u] stands for power generation from unit u in country i [member of] [France, Germany] during year t, [for all]t [member of] [2020,..., 2030]. [[tau].sup.i.sub.2020] is the 2020 RES target of country i (percentage of RESs in overall power generation, see Table 4). U is the set of all power technologies, and FSB represents the set of all solid biomass fuels of varying quality. UC and URES stand for the sets of coal and RES units, with UC [subset] U and URES [subset] U. In (2a), when co-firing is counted as a RES, [F.sup.i.sub.t,u,b] represents the quantity of solid biomass b that is included in coal plants u [member of] UC under co-firing. [[eta].sup.cf.sub.u,b] is the reduced efficiency rate of coal plants u [member of] UC under co-firing (cf) due to loss in combustion efficiency with biomass (increased moisture content and presence of air). In this case, [[eta].sup.cf.sub.u,b] < [[eta].sup.nocf.sub.u], where [[eta].sup.nocf.sub.u] is the efficiency rate of coal plants under the classical configuration when coal is the only input."
In order to consider the 2030 targets, in addition to those of 2020, we add (1b) to (1a) or (2b) to (2a):
[summation over (u[member of]URES)] [P.sup.i.sub.2030,u] [greater than or equal to] [[tau].sup.i.sub.2030] x ([summation over (u[member of]U)][P.sup.i.sub.2030,u]), (1b)
[summation over (u[member of]URES)] [P.sup.i.sub.2030,u] [greater than or equal to] [[tau].sup.i.sub.2030]x([summation over (u[member of]U)][P.sup.i.sub.2030,u][F.sup.i.sub.2030,u,b]) [greater than or equal to] [[tau].sup.i.sub.2030]x([summation over (u[member of]U)][P.sup.i.sub.2030,u]). (2b)
4. RESULTS AND DISCUSSIONS
4.1 Implications for the electricity mix
Results confirm that recognizing co-firing as an RES may greatly modify the electricity mix, whatever the country. (16) We observe an increased contribution from coal when co-firing is counted as an RES (co-firing in RES) compared with when it is not (co-firing out RES). Figures 3, 4, and 5 indicate that when co-firing is included in the set of RES technologies, the RES capacity remains constant so investments in traditional RESs disappear compared with the situation in which co-firing is considered a non-renewable option.
Increased coal-based generation is more significant in Germany due to its greater coal capacity (Figure 6). There are also some new investments in coal when co-firing is included in the set of RESs (Appendix C). Even though the existing German coal capacity is already very high, it appears that it is not large enough to offset the reduced investments in traditional RESs to meet the RES targets through co-firing. The new coal investments vanish when co-firing is excluded from the set of RESs. In France, existing coal capacities are lower than in Germany (approx. 7.5 GW for the French initial coal capacities against 51.2 GW in Germany), on the one hand, but the RES targets are less significant, on the other hand (Table 4). This translates into two counteracting effects for the need to invest in new coal stations, with French coal capacities that are too small to allow substantial co-firing to meet the RES obligations, but RES targets that are also lower than in Germany (which reduces the need for coal stations to co-fire biomass). The actual effect also depends on the share of nuclear electricity and the resulting need for conventional capacities, such as coal, to fill the nuclear power gap. Overall, when the prolongation of out-of-lifetime nuclear plants is not allowed and co-firing is included in the set of RESs, the increased coal contribution is maximal (Figures 6), which translates into more investments in new coal stations (Appendix C). In this case, the additional French coal capacities may reach up to 18 GW, whereas these investments disappear when old nuclear power stations are maintained in service and co-firing is excluded from the set of RESs. Comparatively, the maximal additional coal capacity in Germany is close to 14 GW when co-firing is included in the set of RESs. In this case, the 2030 German capacity mix exhibits a progression of about 27% for coal compared with initial capacity (with 65 GW of coal in 2030, accounting for 48% of the capacity mix), whereas the same progression is more than 243% in France when old nuclear power stations are decommissioned, with almost 26 GW of coal accounting for 20% of the 2030 capacity mix (compared with 5% when old nuclear power stations are kept on and co-firing is excluded from RESs). (17) That is, although including co-firing in RESs would merely make German electricity still more dependent on coal, it might more radically modify the French capacity mix, in which coal may change status and become an important source of French electricity.
Figures 3, 4, and 5 show that when co-firing is omitted from the set of RESs, investments in traditional RESs (to meet the mandatory obligations) mainly benefit biogas, wind, and dedicated biomass. First, investments in biogas and dedicated biomass appear to be an interesting option because they are competitive RES technologies that are not subject to the same drawbacks as other RESs with low availability. In the case of wind, the drawback of low availability is outweighed by a low investment cost with zero marginal cost, so that it remains a competitive option. (18) Second, investing in biogas and dedicated biomass meets the need for new conventional generation capacities with German nuclear phasing-out, exogenous decommissioning of France's old nuclear power stations, and substantial endogenous decommissioning of out-of-lifetime German combined-cycle units (Figure 10 in Appendix D). (19) Biogas and dedicated biomass offer interesting characteristics in this context, because they are RES technologies with high availability as conventional units.
In the case of Germany, the rapid decline in conventional capacities with nuclear phasing-out and decommissioning of old combined-cycle units as of 2012 is creating an early need for new dispatchable units from the very beginning of the time horizon. This, combined with higher German RES targets, favors more investments in dedicated biomass than in France. Because the model considers an upper limit for new investments that can be implemented during a year in each technology, new capacities have to be directed more towards dedicated biomass in Germany, early in the time horizon (when considering the RESs with high availability, dedicated biomass is the second best technology after biogas at the beginning of the time horizon, see Figure 11 in Appendix D), once the investment potential for biogas has been exhausted. (20)
4.2 Implications for C[O.sub.2] emissions and electricity cost
All the results above indicate that, if co-firing is included in support schemes for renewable electricity, coal would crowd-out traditional RESs, not only with increased generation from existing coal plants but also with additional investments in coal that would be substituted for wind, dedicated biomass, biogas, and other traditional RESs. This may raise political and economic issues in the long run among populations concerned about tackling climate change effects and reducing the share of polluting fossil fuels in the energy mix.
Figure 7 shows that recognizing co-firing as an RES generates sharp increases of C[O.sub.2] emissions due to reduced traditional RESs (carbon-free) and more coal in the electricity mix. As illustrated in Figure 6, including co-firing in the set of RESs produces a much larger increase in coal-based generation in the case of Germany due to its much greater coal capacities. This translates into a more significant increase of C[O.sub.2] emissions in Germany than in France (Figure 7). Although coal plants are mainly used under co-firing in this case, substituting coal with reduced emissions for carbon-free RESs inevitably increases C[O.sub.2] emissions. In France, the effect on C[O.sub.2] emissions depends largely on the share of nuclear power in electricity. When it is not allowed to keep out-of-lifetime nuclear plants in service and when co-firing is included in the set of RESs, the large increased contribution from coal (Figures 6), which is substituted for carbon-free RESs and nuclear power, causes a very significant increase in C[O.sub.2] emissions (Figure 7).
From a more policy-oriented point of view, the increased C[O.sub.2] emissions when recognizing co-firing as an RES should be balanced against the associated cost saving in the electricity sector, which may reduce the cost of policies to achieve objectives for renewable electricity. In order to bring the cost savings out, Figure 8 depicts the overall annual costs associated with managing the power system (generation, investments, prolongations, provisions, etc.) so as to meet electricity demand at the lowest cost. Unsurprisingly, Figure 8 shows that including co-firing in the set of RESs reduces the overall electricity cost in all the situations considered, because this means the RES constraint can be complied with through a conventional and low-cost option, which does not require additional investments for coal plants from existing capacities. For France, the highest cost reduction associated with recognizing co-firing as an RES occurs when the out-of-lifetime nuclear power stations are kept on. In this case, the nuclear plants continue to generate base-load electricity because they are the cheapest conventional technology. The increased coal generation (under co-firing) is essentially located in higher load levels, where it competes with technologies that are less cost effective than nuclear power. Hence, co-firing can reduce the cost of complying with the RES constraint without increasing the cost in base-load because nuclear power is still predominant in this generation segment. (21) By contrast, when prolongation of nuclear power is not allowed, increased coal generation is mainly substituted for nuclear plants, which entails a substantial cost increase in base-load generation even if the cost of complying with the RES constraint is reduced.
A more comprehensive comparison should consider the increased carbon cost for society when recognizing co-firing as an RES. On the one hand, any cost saving due to including co-firing in the set of RESs may reduce the cost of policies to attain objectives for renewable electricity. On the other hand, if this also entails a rise in C[O.sub.2] emissions, one should consider the associated increase in the carbon cost so as to evaluate the actual benefit for society. In order to run this comparison, we evaluate the increased carbon cost (based on increased emissions corresponding to the difference between values associated with co-firing in and out RES in Figure 7) using a series of valuations for C[O.sub.2] emissions reflecting different assumptions about the Social Cost of Carbon (SCC). Meanwhile, the carbon cost that is paid by the power sector is still included in the overall electricity cost, and it relies on the price data for C[O.sub.2] presented in section 3.2.
Nordhaus (2017) provides values for the SCC of 2030 that reflect the emission path with current policies depending on different discount rates. The SCC is in a range of 30 to 165 US Dollars of 2010, which approximately equates to 20 to 130 Euros. (22) Accordingly, we consider the following valuations for estimating the increased carbon cost for society: 20, 30, 100, and 130 Euros per tC[O.sub.2]. The computed carbon costs are compared with the overall cost savings in electricity, which corresponds to the difference between values associated with co-firing in and out RES in Figure 8. Results are presented in Figure 9.
As illustrated in Figure 9, the cost saving from including co-firing in RESs dominates the increased carbon cost when the SCC is low (20 and 30 Euros per tC[O.sub.2]), whereas the opposite occurs with higher SCC (100 and 130 Euros per tC[O.sub.2]). An exception is found for France when the out-of-lifetime nuclear power stations are prolonged. In this case, the cost saving is very high and the increased C[O.sub.2] emissions are slight (see discussions above) with the result that the cost saving invariably outweighs the increased carbon cost, whatever the SCC.
This paper explores the effect of recognizing co-firing coal with biomass as renewable electricity so as to meet the RES mandatory requirement. We provide simulations for the French and German electricity mix with investigations into the consequences for cost savings in the power sector and C[O.sub.2] emissions. We focus on France and Germany because they have substantial coal capacities and no support scheme for co-firing has been implemented in these countries so far. Hence, they are suitable cases for our analysis. The case of France is also useful for analyzing the effect of reduction in French nuclear power in combination with the RES targets. The resulting gap in the fleet of power plants is likely to be offset by some RES power plants, among which co-firing can contribute if counted as a renewable. This is something of interest for our study.
The question of consequences when promoting co-firing as renewable electricity has attracted little attention in the economic literature. To date, to the best of our knowledge, the only contribution comes from Lintunen and Kangas (2010), who provide a theoretical model to analyze the effect of co-firing in a stylized and simplified power system. Compared with this previous work, our paper uses a simulation approach to analyze the consequences for the electricity mix when co-firing is recognized as renewable electricity. We use the Green Electricity Simulate (GES) model, which is a simulation model for electricity designed to focus on biomass-based electricity and co-firing in European countries (Bertrand and Le Cadre, 2015). In order to assess the effect of promoting co-firing as a renewable option, we run the model with and without co-firing in the set of RES technologies that are accounted for to meet the RES targets. Our simulations rely on a detailed representation of the power system, which can be used to derive more general results taking into account elements such as a dynamic time horizon, the decommissioning of old capacities, rising demand for power or increasing renewable targets. This extends the study by Lintunen and Kangas (2010).
Results indicate that, if co-firing is recognized as an RES, coal would crowd-out traditional RESs not only with increased generation from existing coal plants but also with additional investments in coal that would be substituted for wind, dedicated biomass, biogas, and other traditional RESs. We find that the additional investments in coal may be more significant in France than in Germany because current French coal capacities are smaller than German capacities, limiting the possibility of using existing coal plants to meet the RES targets through co-firing. The additional coal capacities may attain a maximum of 18 GW in France (when the model is implemented with exogenous decommissioning of old nuclear power stations) against 14 GW in Germany. This corresponds to adding 27% of coal capacity in German electricity by 2030, whereas the same progression is more than 243% in France when old nuclear power stations are decommissioned (107% when the life of nuclear power plants is extended).
The analysis of C[O.sub.2] emissions reveals sharp increases when co-firing is recognized as an RES. Indeed, substituting coal for carbon-free RESs inevitably increases C[O.sub.2] emissions even if the emissions from coal are reduced through co-firing. The rise is more significant in Germany due to its greater coal capacities. In France, the magnitude of increased emissions depends largely on the share of nuclear electricity, with smaller increases when old nuclear power stations are kept in service. Finally, we find that including co-firing in the set of RESs reduces the overall costs associated with managing the power system because this allows compliance with the RES constraint through a conventional and low-cost option that does not require additional investments. We also offset the cost saving for the power sector against the increased social cost from higher C[O.sub.2] emissions in order to provide a more comprehensive evaluation of the actual benefit for society. Results show that the cost saving is dominated by the increased carbon cost for society if the carbon valuation is high (around 100 Euros per tC[O.sub.2], which is not an unusual value in studies evaluating the SCC), except in France when old nuclear power stations are prolonged (in this case, the cost saving is very high and the increased C[O.sub.2] emissions are slight, because coal competes higher in the merit order and base-load continues to be generated by massive cheap and carbon-free nuclear power).
Overall, our paper raises questions about the incentives to invest in traditional RESs if co-firing is recognized as a renewable. The consequences may be detrimental for the future energy mixes in European countries, with more coal (even if implemented under co-firing), fewer renewables, and resulting higher C[O.sub.2] emissions. The cost arising from adapting electricity to climate change is an important issue with populations that are increasingly concerned by this issue. As illustrated in the US presidential campaign, policy makers can also face complicated trade-offs between climate concerns and employment from the coal industry. (23) In this context, co-firing can be a useful option in the short run, but it can be risky in the long run if it jeopardizes a deeper transition towards more renewables and less carbon in energy. More generally, any policy that promotes co-firing against traditional renewables may result in higher C[O.sub.2] emissions in the long run, if it provides incentive to use coal plants under co-firing instead of investing in pure renewables. Whereas co-firing reduces the carbon intensity of coal plants, it still generates C[O.sub.2] emissions. Hence, if co-firing steadily displaces investments in traditional RESs over time, one may expect the C[O.sub.2] emissions from electricity to be higher in the long run (compared with a more far-reaching energy transition in which pure renewables dominate the fleet of power plants). This is something policy makers should remember when considering whether it is expedient to include provisions for co-firing in the support schemes for renewable electricity.
This work is part of the Agriculture and Forestry research program by the Climate Economics Chair, Paris-Dauphine University. The author want to thank the Climate Economics Chair for financial support.
Alexander, P., D. Moran, M.D.A. Rounsevell, and P. Smith (2013). "Modelling the perennial energy crop market: the role of spatial diffusion." Journal of the Royal Society Interface 10(88). https://doi.org/10.1098/rsif.2013.0656.
Argus (2016). Argus biomass markets. Issues 16-014.
AURES-Ecofys (2016). Auctions for Renewable Energy Support in the Netherlands: Instruments and lessons learnt.
Berggren, M., E. Ljunggren, and F. Johnsson (2008). "Biomass co-firing potentials for electricity generation in Poland - Matching supply and co-firing opportunities." Biomass and Bioenergy 32(9): 865-879. https://doi.org/10.1016/j.biom-bioe.2007.12.017.
Bertrand, V., B. Dequiedt, and E. Le Cadre (2014). "Biomass for Electricity in the EU-27: Potential Demand, C[O.sub.2] Abatements and Breakeven Prices for Co-firing." Energy Policy 73: 631-644. https://doi.org/10.1016/j.enpol.2014.06.007.
Bertrand, V. and E. Le Cadre (2015). "Simulating the use of biomass in electricity with the Green Electricty Simulate model: An application to the French power generation." CEC Working Paper, 2015-3.
Bubholz, M. and R. Nowakowski (2010). Mapping of subsidy systems and future consumption of biomass. Svensk Fjarrvarme, Technical Report.
BMWi (2015). The Energy of the Future--Fourth Energy Transition Monitoring Report. BMWi (German Federal Ministry for Economic Affairs and Energy), Technical Report.
Calvin, K., J. Edmonds, B. Bond-Lamberty, L. Clarke, S.H. Kim, P. Kyle, S.J. Smith, A. Thomson, and M. Wise (2009). "Limiting climate change to 450 ppm C[O.sub.2] equivalent in the 21st century." Energy Economics 31 (Supplement 2): S107-S120. https://doi.org/10.1016/j.eneco.2009.06.006.
CGDD (2016). Les energies renouvelables en France en 2015. CGDD (Commissariat General au Developpement Durable), French Ministry of Environment and Energy, Technical Report.
Clarke, L., J. Edmonds, V. Krey, R. Richels, S. Rose, and M. Tavoni (2009). "International climate policy architectures: overview of the EMF 22 International scenarios." Energy Economics 31(Supplement 2): S64-S81. https://doi.org/10.1016/j.eneco.2009.10.003.
Delarue, E.D., A.D. Ellernan, and W.D. D'haeseleer (2010). "Short-term C[O.sub.2] abatement in the European power sector: 2005-2006." Climate Change Economics 1 (2): 113-133. https://doi.org/10.1142/S2010007810000108.
De Perthuis, C. and B. Solier (2018). La transition energetique face au tempo de I'horloge climatique. CEC, Information & Debats no 56.
Entsoe (2016). Yearly statistics and adequacy Retrospect 2015.
Eurelectric (2011). Power Statistics and Trends 2011.
Hansson, J., G. Berndes, F. Johnsson, and J. Kjarstad (2009). "Co-firing biomass with coal for electricity generation--An assessment of the potential in EU27." Energy Policy 37(4): 1444-1455. https://doi.org/10.1016/j.enpol.2008.12.007. IEA (2012). World Energy Outlook 2012.
Jakob, M., G. Luderer, J. Steckel, M. Tavoni, and S. Monjon (2012). "Time to act now? Assessing the costs of delaying climate measures and benefits of early action." Climatic Change 114: 79-99. https://doi.org/10.1007/sl0584-011-0128-3.
Kannan, R. and H. Turton (2013). "A Long-Term Electricity Dispatch Model with the TIMES Framework." Environmental Modeling and Assessment 28(3): 325-343. https://doi.org/10.1007/s10666-012-9346-y.
Linden, M., M. Maekela, and J. Uusivuori (2013). "Fuel Input Substitution under Tradable Carbon Permits System: Evidence from Finnish Energy Plants 2005-2008." The Energy Journal 34(2): 145-166. http://dx.doi.org/10.5547/01956522.214.171.124.
Lintunen, J. and H.L. Kangas (2010). "The case of co-firing: The market level effects of subsidizing biomass co-combustion." Energy Economics 32(3): 694-701. https://doi.org/10.1016/j.eneco.2009.11.004.
Luderer, G., V. Bosetti, M. Jakob, M. Leimbach, J. Steckel, H. Waisman, and O. Edenhofer (2012). "The economics of decarbonizing the energy system-results and insights from the RECIPE model intercomparison." Climatic Change 114: 9-37. https://doi.org/10.1007/s10584-011-0105-x.
Luderer, G., C. Bertram, K. Calvin, E. De Cian, and E. Kriegler (2016). "Implications of weak near-term climate policies on long-term mitigation pathways." Climatic Change 136: 127-140. https://doi.org/10.1007/s10584-013-0899-9.
Mei, B. and M. Wetzstein (2017). "Burning wood pellets for US electricity generation? A regime switching analysis." Energy Economics 65: 434-441. https://doi.org/10.1016/j.eneco.2017.05.025.
Netherlands Enterprise Agency (2016). SDE+ 2016--Instructions on how to apply for a subsidy for the production of renewable energy.
Nordhaus, W.D. (2017). "Revisiting the social cost of carbon." Proceedings of the National Academy of Sciences of the United States of America 114(7): 1518-1523. https://doi.org/10.1073/pnas.1609244114.
Rentizelas, A.A., A.I. Tolis, and I.P. Tatsiopoulos (2012). "Investment planning in electricity production under C[O.sub.2] price uncertainty." International Journal of Production Economics 140(2): 622-629. https://doi.org/10.1016/j.ijpe.2010.11.002.
RTE (2011). Statistiques Production Consommation Echanges 2011. Reseau de Transport d'Electricite (RTE).
Sands, R.D., K. Schumacher, and H. Forster (2014). "U.S. C[O.sub.2] Mitigation in a Global Context: Welfare, Trade and Land Use." The Energy Journal 35(SI1): 181-197. http://dx.doi.org/10.5547/01956574.35.SI1.10.
Santisirisomboon, J., B. Limmeechokcha, and S. Chungpaibulpatana (2001). "Impacts of biomass power generation and C[O.sub.2] taxation on electricity generation expansion planning and environmental emissions." Energy Policy 29(12): 975-985. https://doi.org/l0.1016/S0301-4215(01)00028-3.
Solier, B. (2014). "An economic and ex post analysis of the impacts of the carbon price on the European power sector." PhD thesis, Paris-Dauphine University.
Strauss, W. (2017). Trump Wants to Save the Coal Industry. We Offer a Plan that can Help. FutureMetrics.
Weigt, H., A.D. Ellerman, and E. Delarue (2013). "C[O.sub.2] abatement from renewables in the German electricity sector: Does a C[O.sub.2] price help?" Energy Economics 40(1): 149-158. https://doi.org/10.1016/j.eneco.2013.09.013.
Xian, H., G. Colson, B. Mei, and M.E. Wetzstein (2015). "Co-firing coal with wood pellets for U.S. electricity generation: a real options analysis." Energy Policy 81: 106-116. https://doi.org/10.1016/j.enpol.2015.02.026.
Extensive documentation on the model is available in supplementary materials submitted with this paper. This includes the GAMS codes, the data files, appendix on treatments for the data, and the complete mathematical formulation of the model. More documentation is available on request.
Table 5: German nuclear phase-out plan, based on World Nuclear Association (www.world-nuclear.org) and WEPP data. City Unit Year Shutdown Year MW Commissioning Biblis BIBLIS A 2011 1974 1225 (68643) Biblis BIBLIS B 2011 1976 1300 (68643) Brunsbuttel BRUNSBUTTEL 1 2011 1977 806 (25541) Essenbach ISAR 1 2011 1979 912 (84051) Geestacht KRUMMEL 1 2011 1984 1402 (21502) Neckarwestheim NECKAR 1 2011 1976 840 (74382) Philippsburg PHILIPPSBURG 1 2011 1980 926 (76661) Stadland UNTERWESER 1 2011 1978 1410 (26935) Grafenrheinfeld GRAFENRHEINFELD 1 2015 1982 1345 (97506) Gundremmingen GUNDREMMINGEN B 2017 1984 1344 (89355) Philippsburg PHILIPPSBURG 2 2019 1985 1458 (76661) Brokdorf BROKDORF 1 2021 1986 1480 (25576) Emmerthal GROHNDE 1 2021 1985 1430 (31860) Gundremmingen GUNDREMMINGEN C 2021 1985 1344 (89355) Lingen EMS (LINGEN) 1 2022 1988 1400 (49811) Essenbach ISAR 2 2022 1988 1488 (84051) Neckarwestheim NECKAR 2 2022 1989 1400 (74382) Table 6: Decommissioning of German nuclear units based on the nuclear phase-out plan (Table 5). Year Per year decommissioning Cumulated decommissioning 2011 8821 8821 2012 0 8821 2013 0 8821 2014 0 8821 2015 1345 10166 2016 0 10166 2017 1344 11510 2018 0 11510 2019 1458 12968 2020 0 12968 2021 4254 17222 2022 4288 21510
Table 7: Main results for coal-based electricity with the 2020 and 2030 RES targets. Germany 2015 2020 2030 Co-firing Co-firing Co-firing Co-firing Co-firing in RES out RES in RES out RES in RES Yearly Power 316 267.1 350.1 237.6 385 Generation ([TWh.sub.elec] per year) Total Installed 64.2 51.2 65.1 51.2 65.1 Capacities (GW) Cumulated New 13 -- 13.9 -- 13.9 Capacities (GW) Germany 2030 Co-firing out RES Yearly Power 175.3 Generation ([TWh.sub.elec] per year) Total Installed 51.2 Capacities (GW) Cumulated New -- Capacities (GW) France--Nuclear Reduction 2015 2020 2030 Co-firing Co-firing Co-firing Co-firing Co-firing in RES out RES in RES out RES in RES Yearly Power 83.6 39 99.8 22.1 193.4 Generation ([TWh.sub.elec] per year) Total Installed 24.5 12.8 25.7 12.8 25.7 Capacities (GW) Cumulated New 17 5.3 18.2 5.3 18.2 Capacities (GW) France--Nuclear Reduction 2030 Co-firing out RES Yearly Power 58.3 Generation ([TWh.sub.elec] per year) Total Installed 12.8 Capacities (GW) Cumulated New 5.3 Capacities (GW) France--Nuclear Prolongation 2015 2020 2030 Co-firing Co-firing Co-firing Co-firing Co-firing in RES out RES in RES out RES in RES Yearly Power 55.8 18.5 65.3 14.6 78.5 Generation ([TWh.sub.elec] per year) Total Installed 15.5 7.5 15.5 7.5 15.5 Capacities (GW) Cumulated New 8 -- 8 -- 8 Capacities (GW) France--Nuclear Prolongation 2030 Co-firing out RES Yearly Power 10.5 Generation ([TWh.sub.elec] per year) Total Installed 7.5 Capacities (GW) Cumulated New -- Capacities (GW) Figure 11: Levelized lifetime cost of electricity (LLCOE) computed for different RES technologies (Biogas-ST = Biogas Steam Turbine ; Biogas-CC = Biogas Combined Cycle ; Biomass-ST = Dedicated biomass Steam Turbine). For each technology, the value in bracket reflects the availability factor. In the case of biomass, AR stands for Agricultural Residues and WP for Wood Pellets. LLCOE 2015 LLCOE 2020 LLCOE 2025 BiogasST (70%) BiogasST (70%) BiogasST (70%) BiogasCC (70%) BiogasCC (70%) BiogasCC (70%) BiomassST-AR (68%) BiomassST AR (68%) BiomassST-AR (68%) BiomassST-WP (68%) BiomassST WP (68%) BiomassST WP (68%) Wind (30%) Wind (30%) Wind (30%) Solar (23%) Solar (23%) Solar (23%) Note: Table made from bar graph.
Vincent Bertrand (*)
(*) CRESE EA3190, Univ. Bourgogne Franche-Comte, F-25000 Besancon, France. Climate Economics Chair, Univ. Paris Dauphine, Paris, France. E-mail: firstname.lastname@example.org.
(1.) It is worth mentioning here the literature that is increasingly questioning the risk associated with "weak near-term" solutions for climate (such as substituting natural gas or co-firing for dirtier fossil based power stations), which are seen as delayed mitigation actions that may continue more than expected and may increase the cost of meeting the long-term mitigation targets or even make them impossible to reach (Clarke et al., 2009; Calvin et al., 2009, Jakob et al., 2012; Luderer et al., 2012; Luderer et al., 2016; De Perthuis and Solier, 2018). See Luderer et al. (2016) for a detailed literature review.
(2.) hoi no 2015-992 du 17 aoitt 2015 relative a la transition energetique pour la croissance verte (www.legifrance.gouv. fr).
(3.) In France, the electricity mix is largely dominated by nuclear power, which represents more than 50% of installed capacity, and around 75% of power generation (76% in 2015, according to RTE, Statistiques Production Consommation Echanges 2015). In comparison, nuclear power accounts for about 5% of all installed capacity in Germany (ENTSO-E, 2016), and it generated 14.2% of German electricity in 2015 (Gross electricity production in Germany from 2014 to 2016, www.destatis.de).
(4.) The Drax power station (UK) is known as the world biggest biomass-based power station with 1220 MW of 100% biomass generation capacity, i.e. two of the six Drax units (conversion of a third unit has been recently decided, which will increase the biomass capacity to 1880 MW). Such a conversion project would not be considered as co-firing because it burns biomass only, and, in the UK, it is entitled to receive a more generous subsidy treatment than co-firing (Table 2).
(5.) For readers familiar with the earlier application of emission trading in the UK (the so-called UK ETS), the design is similar, with participants bidding for premiums that cover increased costs associated with efforts (increased RES generation with the SDE+ and carbon abatements in the case of the UK ETS).
(7). One may also mention here the contributions by Linden et al. (2013), and Sands et al. (2014), which consider co-firing among other options to investigate the substitution between fossil and non-fossil fuels in the Finnish power system (Linden et al., 2013), and the effect of bio-electricity with other power technologies for the carbon balance between mitigation and land use (Sands et al., 2014). However, since co-firing is not the focus of these papers (but rather a secondary question that is not analyzed alone), we do not discuss them in detail in this literature review.
(8.) Another recent paper by Strauss (2017) estimates that the US government has to compensate the generators of co-fired electricity with a subsidy of about 0.007 US Dollars per k[Wh.sub.elec] so as to render co-firing competitive.
(9.) Electricity simulation models or electricity models refer to models that simulate power generation and investment decisions in electricity, based on actual data rather than representative situations as in theoretical approaches. Many electricity models have been developed to simulate the impact of different scenarios for energy and environmental policy. Among those contributions, some have focused on questions around biomass-based electricity, but they neglect biomass co-firing in coal plants (Santisirisomboon et al., 2001; Rentizelas et al., 2012). To the best of our knowledge, the GES model is the only electricity model that is primarily designed to analyze questions related to co-firing. See Kannan and Turton (2013) and Rentizelas et al. (2012) for a literature review on electricity models.
(10.) At the beginning of each year, the model identifies which are the out-of-lifetime power plants (i.e. age > theoretical lifetime). Once the set of out-of-lifetime power plants has been identified, the model implements calculations for each unit in this set, so it can be determined whether it is a profitable option to refurbish and extend the life of those units, or whether it is cheaper to decommission them and consider new investments. The calculation relies on comparing the Levelized Lifetime Costs of Electricity (LLCOE) associated with new or prolonged units. In the case of coal plants, this calculation can be implemented by taking into account the ability to co-fire coal with biomass or not. See supplementary materials from Appendix A for the complete mathematical formulation of the model.
(11.) In general, investment costs for renewables have been considerably reduced during the last few years, and this process is likely to continue in the future. This may impact our analysis, and this would merit closer investigation in further studies. Nevertheless, even if the cost to invest in renewables is drastically reduced, this would still be possible to get support from renewable schemes with co-firing, without implementing any investment. Hence, it is still possible to replace traditional RESs (which require investment) by co-firing (which needs no investment) in order to obtain subsidies.
(12.) See Appendix A.
(13.) See Appendix A.
(14.) See Power Statistics on www.entsoe.eu.
(15.) The model considers different types of solid biomass with varying quality: agricultural residues (AR), wood chips (WC), wood pellets (WP), and torrefied biomass pellets (TOP). The higher the quality (AR quality < WC quality < WP quality < TOP quality), the higher the percentage of biomass (that can be included in coal plants). Moreover, for a given percentage of biomass, the actual reduction in the efficiency rate depends on the type of biomass, based on a loss coefficient that increases when the biomass quality is reduced. Hence, for a given percentage of biomass, [[eta].sup.cf.sub.u,AR] < [[eta].sup.cf.sub.u,WC] < [[eta].sup.cf.sub.u,WP] < [[eta].sup.cf.sub.u,TOP]. See supplementary materials from Appendix A.
(16.) To save space, we only report the results associated with implementation of the model with both the 2020 and the 2030 constraints for the RES targets (i.e. (1a) with (1b) or (2a) with (2b)). Alternative settings do not qualitatively modify results. Additional results are available upon request.
(17.) When old nuclear power stations are kept on and co-firing is in the set of RES, the French coal capacity increases by 107% in 2030 compared with initial capacity, with about 15.5 GW of coal accounting for 12% of the 2030 capacity mix. Here again a surge occurs.
(18.) The competitiveness of biogas, wind, and dedicated biomass is illustrated by the levelized lifetime cost of electricity (LLCOE) in Appendix D.
(19.) It appears that prolonging old combined cycle (gas or oil) is not a profitable option because investing in new fashion units is not very costly (e.g. at half of the cost of investing in a comparable new coal plant), and it provides a greater increase in the efficiency rate than competing technologies.
(20.) Setting such per technology maximal amounts for yearly investments is a common assumption in simulation models for electricity (e.g. Rentizelas et al., 2012; Kannan and Turton, 2013). This reflects real-world constraints and avoids unrealistic situations in which power generation would rely on a single or very few technologies due to massive investments.
(21.) For the same reasons, recognizing co-firing as an RES increases C[O.sub.2] emissions more when nuclear power plants are decommissioned. In this case, coal under co-firing is substituted for nuclear power in base-load, thereby emitting more C[O.sub.2] than when nuclear plants are kept on to generate base-load and co-firing is implemented for higher load-levels (Figure 7).
(22.) We used a representative EUR/USD exchange rate of 2010 (from ECB) to convert the values.
(23.) In the case of the United States, Strauss (2017) proposes a strategy based on a policy that supports co-firing in US coal power plants, for the Trump administration to save the coal industry while creating new jobs for the forestry industry in states where pulp and paper mills are closing due to decline in the demand for paper. Additionally, C[O.sub.2] emissions from the US coal plants would be reduced, which is seen as a by-product of job protection. The author concludes that the strategy is "a win-win-win for the coal sector, the forest products sector, and for the environment".
Table 1: Treatment for co-firing in support schemes of European countries (Bubholz and Nowakowski, 2010). Country Subsidy for co-firing--Euros Support system /[MWh.sub.elec] in 2010 Austria 63 (a) Feed-in-Tariff Belgium 0 None Denmark 20 (b) Feed-in-Premium Estonia 0 None Finland 0 None France 0 None Germany 0 None Italy 0 None Latvia 0 None Lithuania 0 None Norway 0 None Poland 64 Green Certificates Spain 20 (c) Feed-in-Premium Sweden 28 (d) Green Certificates Netherlands 61 Feed-in-Tariff United 25 (e) Green Certificates Kingdom (a) Maximal value. Reductions can be applied depending on the biomass material (up to 50% for lowest quality). (b) A subsidy is given for each tonne biomass that is burned (depending on local agreements), in addition to certificates. (c) Reference value. The actual premium is calculated based on the plant data (e.g. energy output, investment cost, biomass material). (d) Only the biomass part can receive certificates. (e) Value with 0.5 certificates per MWhelec (the applied rate of certificates depends on the percentage of biomass in the coal plant). Table 2: Cost of generating electricity with biomass in the UK under ROC banding (Argus, 2016; Alexander et al., 2013). ROC rate ROC value Electricity Cost (Euros (Euros /[MWh.sub.elec]) (a) /[MWh.sub.elec]) (b) Dedicated biomass 1.5 78.60 117.41 Conversion--100% 1 52.40 91.21 biomass Co-firing--More 0.9 47.16 85.97 85% biomass than Co-firing--50 to 0.6 31.44 70.25 85% biomass Co-firing--Up to 0.3 15.72 54.53 50% biomass (a) Based on the ROC value of May 2016 (52.40 Euros). (b) Cost associated with 34% efficiency power stations, and market prices (coal, EUA, wood pellets) of May 2016. Table 3: Coal in the 2010 European electricity (Eurelectric, 2011). Germany Poland UK Denmark France Netherlands Coal power 55 547 34 305 28 068 9 272 8 153 5 641 capacity (a) (29%) (86%) (28%) (42%) (6%) (14%) Coal power 262.4 154 102.9 27 19.8 27.1 generation (38%) (87%) (26%) (42%) (4%) (15%) Greece Belgium Coal power 4 744 1 156 capacity (a) (29%) (6%) Coal power 27.5 6.2 generation (49%) (6%) (a) Countries are ranked from left to right by increasing coal capacities. Table 4: 2020 and 2030 RES targets for power generation in France and Germany (BMWi, 2015; CGDD, 2015). The values are expressed in percentage of the 2020/2030 overall power generation. 2020 France [[tau].sup.France.sub.2020] = 27% Germany [[tau].sup.Germany.sub.2020] = 35% 2030 France [[tau].sup.France.sub.2030] = 40% Germany [[tau].sup.Germany.sub.2030] = 50%
|Printer friendly Cite/link Email Feedback|
|Publication:||The Energy Journal|
|Date:||Jul 1, 2019|
|Previous Article:||The Spatial Deployment of Renewable Energy Based on China's Coal-heavy Generation Mix and Inter-regional Transmission Grid.|
|Next Article:||Can China's Energy Intensity Constraint Policy Promote Total Factor Energy Efficiency? Evidence from the Industrial Sector.|