Greenhouse Gas and Electricity Trading Simulation

October 1999 Ref : 1999-420-0013 ................................ ................................ ................................ ...................... Greenhouse Gas and Electricity Trading Simulation ................................ ................................ ................................ ...................... An exercise in trading carried out by the Electricity Industry in collaboration with the International Energy Agency and ParisBourseSBF SA. ................................ ................................ ................................ ...................... Greenhouse Gas and Electricity Trading Simulation ............................................................................................ This report has been drafted by Richard Baron, IEA and Raymond Cremades, ParisBourseSBF SA in collaboration with the UNIPEDE/EURELECTRIC Working Group on Climate Change Members of the Working Group Climate Change : J-P. Bourdier (FR), Chairman D. Baggs (GB), W. Barc (PL), J.E. Barroso (PT), R. Beising (DE), J-Y. Caneill (FR), R. de Lannoy (BE), G. Hovsenius (SE), P. Kubacka (SK), B. Meclot (FR), G. Montesano (IT), J. Rämö (FI), C. Rivero (ES), A. Sahar (IL), M. Schneeberger (AT), O. Simonsen (DK), J. Stehlic (CZ), S. Stibbe (NL), O. Simonsen (DK), O. Wilson (IE), T. Yamanaka (JP), J.F. Scowcroft (UNIPEDE/ URELECTRIC), A. Agrait (UNIPEDE/EURELECTRIC). E ............................................................................................ UNIPEDE/EURELECTRIC wishes to thank both the IEA and ParisBourseSBF SA for their invaluable assistance and participation in establishing and participating in the Simulation. TABLE OF CONTENTS EXECUTIVE SUMMARY GLOSSARY I. DESCRIPTION OF THE SIMULATION I.1. Introduction & History I.2. Participants : Virtual Companies I.3. Modalities I.4. CO2 and Electricity Trading II. MAIN STEPS OF THE SIMULATION 1 3 4 4 4 5 7 9 10 10 10 12 16 19 25 26 28 29 30 32 32 33 34 III. RESULTS III.1. Summary Results III.2. Emission Profiles III.3. Power Generation : Investment Patterns and Capacity Management III.4. Trading Activity : CO2 and Electricity III.5. Results : Compliance with CO2 objectives III.6. Lessons III.7. Limit of the Simulation IV. FURTHER DEVELOPMENTS APPENDICES PROFILES OF VIRTUAL COMPANIES IN THE YEAR 2000 DESCRIPTION AND ILLUSTRATION OF THE TRADING TOOL (EXCHANGE) 1. 2. 3. General Page : Market Summary Example of a Market : CO2 Emission Spot Information on Trading Activity Available to EachParticipant DEFINITIONS, PARAMETERS AND RULES OF THE SIMULATION 1. 2. 3. Objective of the Trading Simulation Description of Quiz Master/Advisor Development of the Trading Simulation 3.1. Simulation Periods 3.2. What will be traded andhow ? 3.3. Timing of the Game 3.4. Timing and Feedback from Players 3.5. Schedule of the Players Feedback 3.6. Feedback from the Quiz Master 35 35 35 35 35 36 38 39 42 43 Executive Summary: Lessons Learned In May 1999, 19 European electricity companies from 14 countries, members of the UNIPEDE/EURELECTRIC association, launched a simulation of CO2 and electricity trading. The purpose of the simulation was to explore the usefulness as well as the technical features of CO2 emissions trading in the context of an open international electricity market. The Climate Change Working Group of UNIPEDE/EURELECTRIC, in collaboration with the International Energy Agency and ParisBourseSBF SA, agreed on a set of rules for emission trading in the simulation, as well as on reporting guidelines on their activity during the simulation. A virtual stock exchange was provided by ParisBourse SBF SA, to enable CO2 and electricity trading. Over eight weeks, some 16 virtual power companies simulated growing electricity demand under a constraint on their individual CO2 emissions, and traded both CO2 and electricity. Each participating company gathered internal production and trading experts to decide on the most appropriate strategy. They also produced weekly reports indicating their electricity supply situation, their level of CO2 emissions, and their trading activity. IEA and ParisBourse SBF SA audited these reports on a weekly basis and shared public information with all participants. Key features that helped virtual companies towards compliance included: • the rapid emergence of a price signal for CO2 and electricity as a result of open trading; • the option to bank CO2 permits from one period to the next; and • the opportunity to trade CO2 after the end of each budget period, during so-called grace periods. The companies participating in the simulation quickly became accustomed to CO2 trading as a market mechanism. They developed analytical and decision-oriented tools to best manage their electricity production and their decisions on CO2; they acquired experience in developing cost-minimising strategies, which associated both electricity and CO2 trading to their investment choices in new capacity. Some virtual companies relied heavily on trading to comply with their CO2 objectives. Those companies for which the agreed CO2 targets were most stringent particularly benefited from being able to trade. Not all virtual companies eventually managed to comply with their emission objectives (two companies out of a total of sixteen). This was due in part to the design of the simulation and also because some virtual companies examined the implications of high-risk strategies that they would not necessarily have pursued if the targets had been real. All the other companies complied, four of which with a significant margin. There are some important lessons from the exercise in relation to meeting emission targets. In particular, it is clear that compliance with CO2 emission objectives is a matter of formulating the right investment strategy, which goes beyond emissions trading. The price signal given by the emission trading system is a crucial factor in elaborating such a strategy and each of the participants found trading beneficial. However, no policy tool can, in itself, guarantee that the environmental objectives will be met. 1 Participants to the simulation also noted that trading would be even more beneficial if it included other sectors. In all, this simulation demonstrated that emissions trading per se would not present any technical difficulty. Free and open trading can help to meet emission objectives by lowering compliance costs and by giving a strong signal, via the price of CO2 permits, on the economic implications of an emission objective. Banking and grace periods would also assist compliance strategies, without a negative impact on the environment. The role and design of a penalty for non-compliance, such as the one used in this simulation, should be carefully considered. Caveat This simulation was undertaken as an exploratory exercise. The rules for emissions trading used in this simulation (included in Appendix) did not intend to prejudge the possible design of an industry-based emissions trading regime. Rather, the agreed rules aimed to establish a sufficiently realistic framework to perform the simulation, but they should not be taken as formal recommendations. The reader may also be tempted to use price information in this report as an indication of the marginal cost of GHG reductions in the European electricity sector. The numbers reported in this report are unlikely to bear any resemblance with those that would be obtained in the real world for the following reasons: 1. The companies who participated in this game are virtual. Neither their size, nor their fuel-mix, correspond to the current structure of the European electricity market. Among other things, no trade in electricity is assumed in the initial year. 2. The price of traded CO2 units is entirely dependent on the level of the emissions constraint and on the sources covered by the trading regime. It is unlikely that an emissions trading regime would be limited to producers and distributors of electricity, as is the case in this simulation. Other sectors would probably be involved as well. 3. For the sake of simplicity, fuel prices were assumed to remain constant throughout the simulation, while the sharp increase in gas consumption would probably have an effect on price, hence on generation costs and mitigation costs. 4. Electricity trading was unhampered by transmission constraints, and essentially free in this simulation, regardless of geographical considerations. 2 Glossary Banking: A company which emits less than its emission objectives in a given commitment period and did not sell all its surplus CO2 permits can use them in the next commitment period. Commitment period: Period over which an emission objective is defined. Two commitment periods were simulated: 2005-2007 and 2008-2012. Also called budget period. CO2 permits: Traded CO2 units, which allow the buyer to emit more than its allowed emission objective over a given period, while the seller has to reduce its emissions by the same amount. Futures contract: Contract by which the seller agrees to make delivery of CO2 permits at a given price at the end of a designated period of time (here, either 2007 or 2012, the end years of the two commitment periods). Futures are exclusively traded through established exchanges. Grace period: Period after the end of each commitment period during which virtual companies can acquire CO2 permits generated in the period in order to comply with their emission objective. Spot market: Market on which CO2 permits are sold for immediate delivery (in this simulation, transactions on the CO2 spot market started in 2008, for acquisition of banked units over the previous budget period). 3 I. Description of the Simulation I.1. Introduction & History In March 1999, 19 European electricity companies, members of the UNIPEDE/EURELECTRIC association decided to engage in a practical emission and electricity trading simulation. The goals of this exercise were: • To try and simulate the effects of a CO2 constraint on power generation, when CO2 emissions and electricity can be traded; • To learn how electricity and emissions trading could be integrated in companies’activity; • To draw practical lessons for the design of emissions trading under the Kyoto Protocol of the United Nations Framework Convention on Climate Change. This greenhouse gas and electricity trading simulation (GETS) was supported by ParisBourseSBF SA (the Paris stock exchange). ParisBourseSBF SA designed and maintained an exchange and provided a trading tool that enabled each company to buy and sell CO2 and electricity in real time via internet. The International Energy Agency (IEA) advised UNIPEDE/EURELECTRIC on the design of the simulation, including the rules of the game, and the reporting obligations of virtual companies during the simulation. In addition, the IEA monitored the activity of the virtual companies to ensure realism from the power generation standpoint. I.2 Participants: Virtual Companies Participants to the simulation were officials of European electricity companies members of UNIPEDE/EURELECTRIC. However, they did not play on behalf of their companies, but via so-called virtual companies, which did not necessarily bear any resemblance with their parent company. Companies were free to choose any profile they wished for their virtual companies. Each virtual company was defined for the year 2000 by: - its level of electricity production; - its plant types and capacity; and - its level of CO2 emissions, to be used as the reference for its emission constraints. This information, shown in Appendix, was shared among all virtual companies prior to the beginning of the game. Table 1 summarises the main features of the virtual companies (VC) in 2000: CO2 emissions and power production. The third column indicates the CO2 intensity of power generation for each virtual company. 4 Table 1: Main features of virtual companies CO2 (MtCO2) 5.56 4.18 8 9.56 2.1 3.2 2.4 26 9.34 15 3.82 15.2 3.22 6.422 4.695 4.975 124.3 TWh CO2 intensity (MtCO2/TWh) 0.585 0.284 0.313 0.637 0.168 1.000 0.218 0.810 0.827 0.250 0.201 0.608 0.239 0.917 0.032 0.452 VC1 9.5 VC2 14.7 VC3 25.6 VC4 15.0 VC5 12.5 VC6 3.2 VC7 11.0 VC8 32.1 VC9 11.3 VC10 60.0 VC11 19.0 VC12 25.0 VC13 13.5 VC14 7.0 VC15 146.2 VC16 11.0 Total 416.6 Average 0.298 Comment: VC 15 is the largest power producer (146.2 TWh) and the least CO2-intensive (0.03 MtCO2 per TWh). VC 6 is the most CO2-intensive (1 MtCO2 per TWh). There is wide variety in both the volume of electricity production (from 3.2 –VC6– to 146.2 TWh –VC15– per year), and the CO2 intensity of generation, ranging from 0.03 to 1.0 MtCO2 / TWh. This reflects the diversity of fuel mixes chosen by virtual companies: in 2000, VC 6 relies exclusively on coal, while VC 15 produces more than 90% of its power from nonCO2 emitting sources: nuclear, hydro, biomass and wind. The diversity in production mixes also suggests that the marginal cost of CO2 emission reductions is likely to differ among virtual companies. I.3 Modalities I.3.a – Simulation period The simulation period lasted eight weeks, covering the 2000-2012 period. Each week represented either one or two years of activity. Virtual companies could trade electricity and CO2 on an exchange once a week for two hours. I.3.b - Electricity supply constraint First and foremost, virtual companies had to supply a fixed amount of electricity, based on their initial production level in 2000. At the beginning of each trading session, the IEA and ParisBourseSBF SA would announce its level of electricity demand to each virtual company. A reasonable margin for annual growth rates was agreed before the beginning of the game (0 to 2% per year). 5 It was also agreed that “accidents” would be introduced randomly in the course of the game. As a result, each virtual company faced a 4% surge in its electricity demand during one of the sessions. Electricity demand could either be met by domestic production or by electricity imports, through electricity trading. For the sake of simplicity, no constraint was applied on the transport of electricity. After each trading session, each virtual company had to report on how it had met its electricity demand. In particular, it filled in a report showing how exactly it had generated its power, and the corresponding CO2 emissions (see Table 1 in Appendix: Definitions, Parameters and Rules). I.3.b – Investment and financial constraints In order to obtain a simulation that would generate realistic results, some real-world constraints were imposed on the activity of the virtual companies, especially on their investment: new capacity could only come on line after a certain lead-time, specific to the chosen technology. For instance, a virtual company could only start operating a combined-cycle gas turbine if it had invested in the plant three “years” before; the capital cost and the minimum size for each technology was also set. This was meant to avoid price differences across regions that would only be the result of different assumptions on capital cost. - At the same time, no financial constraint was defined. It was decided at the outset that virtual companies would not be assessed on the basis of their financial gains at the end of the game, as the purpose was to learn about emission trading. As a result, the virtual companies presented a rich spectrum of different generating capacities, reflecting what was perceived as local constraints and resources, and not only cost considerations. I.3.c CO2 emission commitments Virtual companies were given limits on their CO2 emissions, based on their level in 2000, for two consecutive periods, 2005-2007 and 2008-2012: • a 2% reduction from 2000 emission levels during 2005-2007; • a 5% reduction during 2008-2012. These targets were expressed as averages over the commitment period, and not as annual targets, in a way that is similar to the Kyoto Protocol commitment period. The purpose of the first budget period was to set companies on track with their 2008-2012 commitments, along the lines of the “demonstrable progress” by 2005 mentioned in the Protocol. From a learning perspective, it was also desirable to give participants two opportunities to achieve an emission objective, each with a different time horizon. 6 Virtual companies could either reduce their emissions, or acquire CO2 emission permits from the market to cover any emissions over their objective. Banking, i.e., the possibility to carry over extra reductions from one period to the next, was allowed. A one-year grace period was introduced after each commitment period to allow non-complying participants to acquire CO2 emission units from participants with banked units. Failure to comply with either period’ commitment levels obliged the failing virtual company s to reduce its emissions even more in the next budget period, by an amount equal to its extra emissions in the previous budget period. This also triggered a financial penalty equal to 150% of the highest observed price. In the end, two kinds of financial penalties were tested in the simulation: • At the end of the first budget period, before the grace period, the penalty was set at 150% of the highest CO2 contract price observed between 2005 and 2007 (in this case, at Euro 37.5). Participants knew this level when they traded in the grace period. • For the second budget period, the game master announced that the penalty would be set after the end of the grace period, at 150% of the highest observed price. As we later explain, such design for the penalty explain why the price of CO2 contracts rose at the end of the second commitment period. Accordingly, the seller of CO2 units was made liable in case of non-compliance (so-called seller liability or seller beware). Buyers were able to use the acquired CO2 permits for compliance, regardless of the seller’ compliance. s I.4 CO2 and Electricity Trading I.4.a – The exchange Companies had the possibility to trade both CO2 and electricity on an organised exchange, through a dedicated web site provided by ParisBourseSBF SA. All transactions took place through the exchange. Trading occurred weekly during a two-hour session. Buyers and sellers posted their “bids” and “asks” and traded in real-time: trades were contracted when two parties agreed on a price for a given quantity.1 (see Appendix for details). All trades were conducted anonymously, a common practice in stock exchanges. Of course, ParisBourseSBF SA had full knowledge of the trading activity and recorded parties’ identities for each trade. I.4.b – Trading sessions The trading sessions were organised as shown in Table 2: 1 An alternative practice, not tested in this game, is the so-called price fixing. 7 Table 2: Schedule of trading sessions Trading date 27 May 1 June 8 June 15 June 22 June 29 June 6 July 13 July Simulated period 2001 – 2002 2003 – 2004 2005 – 2006 2007 2008 2009 – 2010 2011 – 2012 2013 Markets (commitment periods) Electricity, CO2 (1&2) Electricity, CO2 (1&2) First budget period Electricity, CO2 (1&2) Electricity, CO2 (1&2) Grace period CO2 (1) Second budget period Electricity, CO2 (2) Electricity, CO2 (2) Electricity, CO2 (2) Grace period CO2 (2) I.4.c – Trading contracts Electricity was traded on a spot market only, whereas CO2 was traded for both the first and the second budget period right from 2001-2002. In technical terms, CO2 contracts for the first and second budget period were “futures” contracts, with 2007 and 2012 maturity dates. Starting 2008, a ‘ spot’ market for CO2 was introduced, on which units banked from the first budget period were traded. 8 II. Main Steps of the Simulation A set of rules and guidelines, given in appendix, were sent to the participants, describing their reporting requirements and the various constraints described above. Each week (corresponding to a one-year or two-year period), the simulation unfolded as follows: • Announcement of electricity supply assumptions to all virtual companies; • Communication of investment decisions by virtual companies, compiled and communicated to all virtual companies; • A two-hour trading session for CO2 and electricity; • Confirmation of trades to each virtual companies by the exchange; • Report on electricity supply, electricity trade, CO2 emissions and CO2 trades sent by virtual companies for audit; • Audit of electricity and CO2 results for each company. 9 III. Results III.1 Summary Results • Fourteen companies complied with their global emission objectives (first and second budget period combined) at the end of the second budget period. Only twelve companies were in compliance after the first budget period. • In the end, two companies did not meet their emission objectives, in spite of large acquisitions of CO2 permits from the market. • Over 60 million tonnes of CO2 and 30 TWh were traded in the course of the simulation. • Traded electricity prices increased over the period, reflecting the cost of CO2 reductions (as shown by the price of emission permits). III.2 Emission Profiles III.2.a First commitment period: general over-compliance Table 3 provides emission profiles of the virtual companies compared with their objectives, which indicates probable compliance strategies considered by the virtual companies. For most companies (10 out of 16), emissions for the period 2005-2007 are below their assigned emission objectives. Supposedly, the remaining six companies intended to comply through the acquisition of CO2 emission permits on the market. The total level of emissions was below the assigned emission objective (by 3.152 million tonnes of CO2): in theory, all six companies with emissions above their objectives could therefore comply through trading. 10 Table 3: Emissions and objectives for 2005-2007 Virtual company Annual 2005-2007 2005-2007 Emissions in Objective (*) Emissions (*) 2000 (*) (**) 5 558 4 180 8 000 9 560 2 100 3 200 2 400 23 745 9 340 15 000 3 820 15 226 3 273 6 422 4 695 4 975 121 494 16 341 12 289 23 520 28 106 6 174 9 408 7 056 69 810 27 460 44 100 11 231 44 764 9 623 18 881 13 803 14 627 357 192 14 531 15 010 22 758 22 016 6 558 6 615 856 62 400 24 000 41 399 12 360 52 377 9 912 18 567 30 824 13 856 354 040 Objective – Emissions VC1 VC2 VC3 VC4 VC5 VC6 VC7 VC8 VC9 VC10 VC11 VC12 VC13 VC14 VC15 VC16 TOTAL 1 810 -2 721 762 6 090 -384 2 793 6 200 7 410 3 460 2 701 -1 129 -7 613 -289 314 -17 021 771 3 152 (*) Units : 1 000 tonnes of CO2 (**) 2005-2007 objective : 98 % of 2000 over three years. III.2.b Second commitment period: emissions above objectives The emission objective for the second budget period was set at 95% of 2000 emission levels. These objectives were augmented by the tonnes which virtual companies had banked in the first budget period, or reduced by the emissions above targets, for those companies which did not comply. Table 4 therefore presents the emission objectives for 2008-2012 with full account taken of compliance with the 2005-2007 objectives (including trading). Over the 2008-2012 period, five virtual companies had emitted CO2 above their allocated objectives (see last column in Table 4), indicating that they hoped to acquire CO2 emission permits from the market to achieve compliance. However, while the first budget period showed general over-compliance, emissions were above the objectives in the second budget period (see total), indicating that some companies would eventually be out of compliance. 11 Table 4: Emissions and objectives for 2008-2012 2008-2012 Objective Objective (*) (**) + Banking (B) – Debt (D) 26 401 19 855 38 000 45 410 9 975 15 200 11 400 112 789 44 424 71 250 18 145 72 322 15 549 30 505 22 301 23 631 577 156 26 160 17 634 38 362 46 500 10 029 17 993 14 100 118 044 44 483 72 376 18 145 69 524 15 549 30 514 16 693 24 002 580 108 2008-2012 Emissions (*) Objective (+B–D) – Emissions Virtual company VC1 VC2 VC3 VC4 VC5 VC6 VC7 VC8 VC9 VC10 VC11 VC12 VC13 VC14 VC15 VC16 TOTAL 25 129 15 730 38 102 44 878 14 137 12 283 11 448 104 500 40 684 72 670 14 795 70 792 16 152 26 871 77 792 22 773 608 735 1 031 1 904 260 1 622 -4 108 5 710 2 652 13 544 3 799 -294 3 350 -1 268 -603 3 643 -61 099 1 229 -28 627 (*) Units : 1 000 tonnes of CO2 (**) 2008-2012 objective: -5% from 2000 levels III.3 Power Generation: investment patterns and capacity management Over the simulation period, electricity demand grew by 24%. For those virtual companies which complied with their CO2 emission objectives, this represented a 26% reduction in the CO2 intensity of their power generation. This could not be achieved without significant investment in new, less CO2 emitting, power plants. This is reflected in Figure 1, which illustrates the coming on-line of new capacity over time for the whole simulation. The vast majority of new capacity installed during the period was based on natural gas, either through a coal-to-gas conversion, or through the installation of co-generation and combined cycle gas turbines. The last two options, when substituting for coal, reduced CO2 emissions more cheaply than a simple coal-to-gas conversion, which delivered almost no improvement in plant efficiency. As a result, shown in Figure 2, there was a dramatic decline in coal-based power generation. 12 Figure 1 Additions to generating capacity 4000 Coal 3500 Coal to gas Co-gen gas CCGT Biomass Wind 3000 2500 MW 2000 1500 1000 500 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Smaller investments were recorded in wind and biomass-based generation, even though they represented large percentage increases during the period (95% and 23% respectively). One company invested in a coal plant that came on line in 2010, as it was expected that electricity prices may rise faster than CO2 emission permits. This did not turn out to be the case in the simulation, at least not to the point where such investment generated significant profits. As shown, no new capacity was installed after 2011. This feature of the simulation deserves a note of caution. No emission objectives were specified beyond 2012; there was therefore no incentive to invest in less CO2 intensive capacity beyond that date. In fact, further reductions beyond 2012 may have affected the choice of technology earlier. As we will see below, the absence of a next commitment period also affected trading and compliance strategies. The investment behaviour observed in the simulation largely reflected (1) the constraint on emissions and (2) the assumptions on the capital cost of various technologies adopted for this game.2 In the real world, capital cost may vary for a given technology from one country to the next, as well as the cost of generated power, e.g. for wind or biomass. Figure 2 gives a more complete picture of capacity changes over the simulation, showing the significant increase in gas capacity, and reduction in coal. 2 See Appendix: Definitions, Parameters and Rules of the Simulation. 13 Figure 2 Changes in installed capacity (2001-2012) 30,000 25,000 2001 2012 20,000 (in MW) 15,000 10,000 5,000 - Pe at Co al O il Ga s Hy dr o Other means of CO2 abatement were available to participants, in addition to straight investment in less CO2-intensive capacity. Those virtual companies who relied on hydro and nuclear resources often increased the use of available resources, as these were not necessarily used at their full potential in the base year; indeed, no new investment was made in these technologies in the course of the simulation. As for other low-CO2 emitting sources (natural gas, biomass and wind), virtual companies increased the number of hours of operation, as shown in Figure 3. When combined with the results on new capacity, this figure shows clearly how virtual companies managed their installed capacity to minimise CO2 reductions: the usage of gas, biomass, wind and nuclear was increased over the simulation. While the number of hours of use of peat and oil plants also increased, the total capacity of these plant types was reduced, as shown in figure 2: in all, these technologies contributed a lower share of total power supply in 2012 than in 2001. The small operation time of coal plants in 2012 is also consistent with the CO2 objective applied to the virtual companies. 14 Bi om as s Nu cle ar W ind Figure 3 Changes in plant use (2001-2012) 8000 2001 2012 7000 Hours (8765 hours in a year) 6000 5000 4000 3000 2000 1000 0 O il Pe at So lid Ga s Hy dr o Participants felt that they could not uniquely base their production behaviour on capital cost agreed for the simulation and that needed to respect some constraints that they would face in the real world. For instance, one company decommissioned a nuclear power plant in the last budget period. Another closed its nuclear capacity during several months to comply with national regulations, which were not specified in the simulation rules. Another participant relying on hydro resources introduced climatic variations. In brief, virtual companies combined new investment in less CO2-emitting capacity (gas, but also renewable sources) with increased use of available capacity in order to meet their emission constraints while supplying increasing amounts of electricity over the period. For practical reasons, demand-side management, and other energy efficiency measures were not introduced as options to reduce emissions in the simulation. This would have required much more detailed assumptions about the potential and cost of such options on a companyby-company basis. Bi om as s 15 Nu cle ar W ind III.4 Trading Activity: CO2 and electricity III.4.a CO2 trading: a reflection of underlying supply and demand Out of the 16 virtual companies, 15 traded CO2 actively. Figure 4 summarises CO2 trading during the different sessions of the simulation, including the so-called grace periods (2008 and 2013); the black line indicates the price range recorded in the session (highest and lowest prices), while squares show closing prices. The volumes traded were rather minimal in the first two trading sessions (2001 to 2004), with only a few futures contracts for the first budget period (2005-2007). A few points deserve attention: • CO2 prices before the first commitment period started at a low 6 Euro per tonne, as participants were “testing the waters” and looking for low-cost permits. Much higher prices were recorded during the first commitment period; • Prices increased, especially during the grace period (2008), largely the result of much excess demand, as some participants rushed to acquire permits to cover their emissions (recall that four VC were in non-compliance at the end of the first budget period); • The rapidly rising price observed during the second grace period (125 Euro per tonne) was primarily caused by excess demand on the market, since overall, the system’ emissions s were 28 million tons above the objective. The way the penalty would be applied also influenced prices: the penalty would be equal to 150% of the highest observed price in the grace period: any CO2 permit would therefore be cheaper than paying a penalty for noncompliance. The penalty regime adopted for this game did not put a cap on the price, quite the contrary. • The lower trading volume in the grace period, compared to 2011-2012, can be explained by the fact that there would be only limited supply in any grace period, as permits are generated in the previous period and no new permits can be generated for trading in the grace period. 16 Figure 4 CO2 Trading Activity 25000 140 120 20000 100 (Euro / tCO2) 15000 (1000 tCO2) 80 10000 60 40 5000 20 0 2001-02 2003-04 2005-06 2007 Grace period-1 2008 2009-10 2011-12 Grace period-2 0 Vertical lines: price range in the trading session (right-hand scale) Squares: closing price (right-hand scale) Bars: traded volume (left-hand scale) III.4.b Electricity trading: As shown in Figure 5, electricity trading started early in the simulation, as companies were immediately required to meet their electricity demand and found opportunities to do so more efficiently through trading. The drop in trading in the 2005-2007 period is probably due to more attention paid to the CO2 market, as both commodities were traded simultaneously. Overall, traded volumes only accounted for about 0.5% of total electricity demand over the simulation period. This may appear small, but is actually significant, as all virtual companies started from a situation in 2000 where they met their electricity demand entirely without imports from other countries. In the real world, electricity is already traded among companies in different countries (roughly 3.5% of total electricity in OECD countries). 17 Figure 5 Electricity trading 10000 9000 60 8000 7000 6000 5000 4000 3000 2000 10 1000 0 2001-02 2003-04 2005-06 2007 2008 2009-10 2011-12 0 30 50 70 1000 MWh 40 20 While the market tended to slow down between 2005 and 2008, the last two trading sessions witnessed active electricity trading. Virtual companies used electricity trade as a possible tool for meeting their CO2 goals, or did some arbitrage between CO2 and electricity. The next section expands on this point. An interesting feature of electricity trading in the simulation is the rise in traded electricity prices, from an average of 38 Euro to 51 Euro per megawatt-hour between the first and last trading periods. This increase in the price of electricity is in fact a very close reflection of the average carbon content of electricity generated in that period.3 3 On average, one MWh of electricity released 0.26 tonnes of CO2. At an average price of 48 Euros per tonne of CO2 in that period, this should represent an add-on of 12.5 Euros to the price of electricity. The observed electricity prices increased by 13.4 Euros: which shows that the electricity market internalised fully the cost of CO2 emissions. 18 Euro/MWh III.5 Results: compliance with CO2 objectives III.5.a – First commitment period The compliance results and trading activity are summarised in Table 5. In brief: • Twelve companies were in compliance with their emission objectives in the first commitment period, while four did not comply. • Compliance was mostly achieved through internal emission reductions. This generated a large potential for banked CO2 permits, available for trading in the second budget period. • Three virtual companies acquired CO2 permits from the market to meet their first period commitment. Table 5: Trading and compliance for the first period (2005-2007) CO2 trading activity (*) Compliance (+) 2005-2007 2005-2007 (with the grace period) Non compliance (-) Objective (*) Emissions (*) (**) Purchase Sell 16 341 12 289 23 520 28 106 6 174 9 408 7 056 69 810 27 460 44 100 11 231 44 764 9 623 18 881 13 803 14 627 357 192 14 531 15 010 22 758 22 016 6 558 6 615 856 62 400 24 000 41 399 12 360 52 377 9 912 18 567 30 824 13 856 354 040 500 -400 -5 000 438 -3 500 -2 155 -3 400 -3 550 -2 050 -241 -2 221 362 1 090 54 2 793 2 700 5 255 59 1 126 0 -2 798 0 9 -5 608 371 2 952 Virtual company VC1 VC2 VC3 VC4 VC5 VC6 VC7 VC8 VC9 VC10 VC11 VC12 VC13 VC14 VC15 VC16 TOTAL 1 975 1 130 4 815 300 11 413 -11 -305 -400 20 571 -20 771 (***) (*) Units : 1 000 tonnes de CO2 (**) 2005-2007 objective : 98 % of 2000 (***) The discrepancy between total purchases (20,571) and sales (20,771) is explained by the intervention of a virtual trader, who acquired 200 CO2 contracts in the first budget period. 19 Diverse factors explain why four companies did not comply in the first period, which may or may not be found in the real world, but bring some interesting lessons. 1. One company over-emitted by a large amount (17 million tonnes of CO2). In spite of large permit purchases in 2006-2007 and in the grace period (a total of 11 million tonnes), it did not manage to make up for the difference. Another company was in the same situation. 2. Another company stopped acquiring CO2 permits during the grace period as it saw the price level go beyond the penalty. This resulted from a misunderstanding about how the penalty would operate: the environmental debt was not cancelled by the penalty, but had to be offset in the next budget period. The penalty did not, in this experiment, represent a cap on the cost of compliance. 3. One company inadvertently sold 2008-2012 units on the spot market. These units were therefore subtracted from its 2005-2007 budget, and brought the company in noncompliance. Overall, the system was in line with the emission goal, with banked units offsetting extra emissions by non-complying companies. This suggests that companies in non-compliance could have purchased the necessary permits from the market. In fact, two companies tried but did not manage to acquire all the necessary units before the market closed. In the real world, more continuous trading, and longer trading sessions would probably accommodate this problem. Importantly, trading helped companies unable to comply to minimise the cost associated with non-compliance: it decreased the environmental debt carried over the next budget period. This proved to be an appropriate decision, as CO2 permit prices increased in the second commitment period. In that respect, the above table shows that the quantities of CO2 permits traded by virtual companies were not equal. This is explained by the participation of a trader in the simulation. The trader acquired 200,000 tonnes in the first period, which were sold in the second commitment period. The role of the trader, in the end, was not significant, as it decided to restrict its activity in order to respect the financial constraints that would be imposed upon its activity in the real world. 20 III.5.b – Second commitment period Table 6: Trading and compliance for the second period (2008-2012) Trading activity Virtual 2008-2012 2008-2012 Banking (+) Compliance (+) company Objectives Emissions (*) Debt (-) Non Purchase Sell (*) (**) compliance (-) (*) (*) VC1 26 401 25 129 -241 206 -815 422 VC2 19 855 15 730 -2 221 2 995 -4 800 99 VC3 38 000 38 102 362 1 900 -2 159 1 VC4 45 410 44 878 1 090 1 500 -3 090 33 VC5 9 975 14 137 54 3 640 -468 VC6 15 200 12 283 2 793 5 710 VC7 11 400 11 448 2 700 -2 650 2 VC8 112 789 104 500 5 255 -13 399 145 VC9 44 424 40 684 59 -3 740 60 VC10 VC11 VC12 VC13 VC14 VC15 VC16 TOTAL 71 250 18 145 72 322 15 549 30 505 22 301 23 631 577 156 72 670 14 795 70 792 16 152 26 871 77 792 22 773 608 735 1 126 -2 798 9 -5 608 371 2 952 25 161 -750 41 191 -40 991 (***) 360 4 629 800 -40 -3 350 -2 400 -197 -3 601 26 0 961 0 41 -35 937 479 -28 427 (*) Units : 1 000 tonnes of CO2 (**) 2008-2012 objective : -5% from 2000 levels over five years (***) The discrepancy between total purchases (41,191) and sales (40,991) reflects the sales of 200 units by the virtual trader, acquired in the first budget period. • Three of the four companies in non-compliance in the first budget period complied during the second period, account taken of their excess CO2 emissions in the first period. • Three companies complied through acquisition of CO2 permits. Other companies complied internally and supplied CO2 permits on the market. • Two companies did not comply at the end of the second commitment period. At the end of the second commitment period, the overall emission objective was not met, as the emissions of one virtual company were much higher than extra CO2 permits supplied by others (see text box below). As a whole, emissions were 4.3% above emission objectives. However, there was an improvement in the number of virtual companies in compliance (only two did not comply), with some showing significant over-compliance. As a number of companies had banked CO2 permits from the previous commitment period, they were able either to emit more in the second period, or to sell CO2 permits at a higher price than during the first period. Banking clearly helped companies in their effort to achieve compliance at minimum cost. 21 VC11 Early acquisition of permits, investments, and transfers VC11 is one of the three virtual companies that acquired CO2 permits to comply with its emission objective in the first commitment period. It invested in wind (70 MW) and a combined-cycle gas turbine to further reduce the CO2 intensity of its generation. The company also exported electricity to other companies in that period. All of the company’ purchases of CO2 permits were done prior to the s 2008 grace period. Another CCGT was launched in 2005, which came on line in 2008. This resulted in a sharp reduction in the company’ emissions over the second budget period (more than 20% below its 2000 emission s levels in 2012). The reductions below its objective were sold on the market during the same period. In the case of VC11, the cost of compliance was 3.1% lower than it would have been without the possibility to trade emission reductions. One company did not comply by some significant amount, which could not be covered by supply of permits in the second budget period (see box below). The other company in noncompliance was unable to access the trading site during the grace period, and therefore unable to acquire the CO2 permits necessary to cover emissions above its objective. As this company had anticipated excess supply of CO2 permits and excess demand for electricity, it had invested in a coal-fired power plant, which drove its emissions up. VC 15: What caused non-compliance Over the two commitment periods, VC 15 emitted 60 million tonnes of CO2 above its emission goal. In spite of active purchases of CO2 permits (11 million tonnes in the first budget period and 25 million in the second), it missed its objective by some 35 million tonnes. VC 15 was by far the largest virtual company, with more than 30% of total electricity demand in the simulation. At the same time, its CO2 intensity (emissions per unit of electricity produced) was the lowest. It did not choose to increase its reliance on hydro and nuclear, and invested exclusively in natural gas (co-generation) to meet its increase in electricity demand. Given its low fossil fuel base in 2000, emissions rose rapidly to more than twice their 2000 level by 2007. By 2012, emissions were almost four times their original level. VC 15 started acquiring CO2 permits in 2007 (the last year of the first commitment period). It therefore had relatively little influence on other participants’ investment decisions. It is also noteworthy that very few “futures” contracts for the second commitment period were concluded during the simulation. Without a clear idea of the level of demand in the next budget period, companies with a potential for further CO2 reductions did not have the proper incentive to invest in less CO2-emitting power supply. One can also question whether in the real world the price of CO2 permits alone could drive investment decisions in other companies. While the ability to trade did not bring VC 15 in compliance, it helped it reduce the gap between its emissions and its objective. 22 In the end, four categories of participants can be identified, vis-à-vis their CO2 emission objectives, and whether and how they were met: • Internal reductions and banking/trading for revenues;4 • Internal reductions and banking for compliance in the next period; • Acquisition of permits for compliance; and • Non-compliance. III.5.c Combining CO2 and electricity strategies Virtual companies developed strategies to meet both their internal electricity demand and CO2 emission goals, in which they incorporated the possibility to trade both “commodities”. These strategies needed to consider the following aspects of their business: • Current structure of power production and possibilities to reduce CO2 emissions within that structure (e.g. through increased availability of nuclear power, fuel switching from oil to gas, increased reliance on hydro-power); • New investment required to meet increased electricity demand and CO2 objectives in the two budget periods, and the corresponding marginal cost of reduction; • The possibility to bank extra CO2 emission reductions from one period to the next period; • Alternative to in-house emission reductions: purchasing CO2 permits; • Alternative to domestic electricity supply: electricity imports; • The prospects of over-achieving CO2 emission goals to sell CO2 permits on the market; and • The potential cost of non-compliance with CO2 emission objectives, versus the cost of complying with the emission goals. Most of these issues could only be resolved with information on other participants’ generation and CO2 reduction costs. The existence of a market on which both CO2 and electricity were traded generated the appropriate price information to guide decisions. Indeed, the virtual companies quickly developed tools to help their decision-making on the above elements, in light of the electricity and CO2 prices. 4 In the absence of a constraint beyond 2012, many companies with excess CO2 permits found it profitable to sell these permits during the grace period. The existence of a constraint beyond 2012 would of course change their trading behaviour. 23 CO2 and electricity strategy: VC 13 Characteristics of the virtual company: majority of non-CO2 emitting capacity (nuclear and hydro) with some flexibility to expand supply from hydro capacity. Other capacity: coal, dual oil/gas. Main strategy features: The company’ starting point was the need to take measures in order to meet s CO2 objectives. It considered: • The construction of new, less emitting equipment • The purchase of electricity (to reduce internal production and CO2 emissions) • The purchase CO2 permits, and • Non-compliance (discarded as an option, given the uncertain cost of such a strategy). First commitment period: High electricity prices triggered the decision to operate a new combinedcycle gas turbine in 2006. Considering rising CO2 prices, first anticipated to decrease, VC 13 acquired the necessary CO2 and electricity in order to meet its emission goals. At the margin, imported electricity was cheaper than generation, given the cost of acquiring permits to cover the associated CO2 emissions. Second commitment period: VC13 decided to acquire CO2 permits as opposed to importing electricity in the first years. The surge in CO2 prices at the end of the period encouraged the company to buy electricity and to sell their extra CO2 permits, at a net benefit. In that sense also, trading helped reduce overall compliance cost. VC 13 spent E. 34 million to acquire CO2 permits (a total of 1,1 million tonnes to cover emissions that were 3% above the target), and E. 70 million to import electricity (representing 1% of its total demand over the period). In absence of a CO2 constraint, it would have produced electricity internally with its thermal (coal) power plants, at a cost of E. 42 million. 24 III.6 Lessons III.6.a Trading as a tool for compliance Most companies traded CO2 actively, even though most of them complied through internal reductions. For these participants, trading provided the opportunity to manage their extra CO2 permits as an asset, either to be used as banking from one period to the next, or to generate revenues, in order to minimise their cost of meeting the CO2 emission objective. The simulation clearly showed that trading could help participants to best manage their CO2 emission objective together with their core activity. Through the price signal generated by the market, participants would be in position to decide whether it is more economical to reduce emissions internally, or to rely on the acquisition of permits in order to comply. However, another striking lesson from the simulation is that trading does not guarantee compliance by all participating sources. In a sector like power generation, if investments are not made early enough, the supply of CO2 permits may be fairly inelastic. This is illustrated by the surge of the CO2 permit price during the last grace period. It is only through the formulation of expectations on future production needs and emission levels that a robust compliance strategy can be established. Trading is only but one element in that picture. Still, emissions trading enabled companies that were unable to comply in the end to still offset an important share of their surplus emissions. Overall compliance would have been greatly hampered without the ability to acquire CO2 units from other participants. It is worth noting that neither of the two virtual companies in non-compliance had sold CO2 permits, a risk that is sometimes envisioned in the literature on emissions trading. III.6.b Market design Banking Virtual companies relied a lot on the possibility to bank emission permits generated in the first budget period, either to comply in the next period, or to sell for revenues, when market prices were favourable. In a sector like the power sector, the size of investment in new production is largely dependent on the chosen technology:5 investing in a new 300 MW combined-cycle gas turbine may deliver more low-emission generation than what the company needs to comply with its CO2 objective. Banking made it possible to benefit from these additional reductions, on top of the possibility to trade them immediately. 5 In the simulation, a minimum size was set for new investment, on a technology-by-technology basis (see rules in Appendix) 25 Grace Period The grace period enabled virtual companies to buy or sell CO2 permits after the end of the commitment period, that is, after they had delivered electricity to the market and measured the corresponding emissions. The grace period helps handle the uncertainty related to normal business operations, which may have unexpected effects on emissions. The grace period does not enable participants to issue new CO2 permits, as only those permits generated in the previous commitment period can be traded. In other words, the grace period does not change overall system compliance, but allows those in need of CO2 permits to restore compliance, provided they find other participants willing to sell. Penalty The penalty adopted for this simulation was designed as a strong incentive for participants to comply (penalty set at 150% of the highest observed price + a carry-over of the environmental debt in the next budget period). As compliance by all companies was expected, the modalities of use of penalty revenues were not envisioned. In a real-world situation, penalties could be used by to reduce emissions outside the system, or to acquire CO2 permits from the market. Such mechanism would reduce the environmental damage caused by non-compliance in the simulation. Participants noted the risk associated with such a penalty. Because it was based on a single price (the highest observed price), it could be manipulated by a participant through a single transaction at a very high price, which would not be costly for that participant, but could artificially endanger the financial situation of companies in non-compliance. Liability For the sake of simplicity, it was agreed that all traded CO2 permits were valid for compliance. The seller is entirely liable for having sold CO2 permits if its emissions are above its objective. Other liability rules (such as buyer liability) would have required more precise tracking of trade, with the generation of serial numbers to identify the original seller of permits. This was beyond the scope of this simulation. III.7 Limits of the simulation The electricity and CO2 trading simulation achieved a relatively high degree of realism. However, a few critical points are worth highlighting, as they indicate the limits of such an exercise. For these and other reasons mentioned earlier, investment strategies, electricity prices and CO2 permit prices would probably be different from what the simulation has shown. • Primary energy prices were assumed to remain constant. A sharp increase in demand for gas could drive gas prices up, while the decline in coal demand would bring prices down. This would change the economics of power generation and alter production and CO2 mitigation costs. • The simulation did not extend beyond the power sector. A trading system would probably seek to include as many different sectors as practically feasible, in order to maximise the efficiency gains among participants. 26 • Project-based activities (based on joint implementation or clean development mechanism) were not included, as it was unclear how one could control their validity in the simulation. • Electricity was transported at no cost, among companies that could be operating in fairly distant geographic areas. Transmission costs would need to be included to reflect this point. • Electricity was only traded on spot markets, in order to focus the simulation on CO2 strategies. The inclusion of a futures market for electricity would probably affect the strategies of participants, by allowing them to hedge against certain risks. • Revenues from penalties for non-compliance with CO2 objectives were not recycled in the CO2 permit market. A mechanism should be envisioned to use the revenues, if such penalties were agreed. • As Parties (governments) remain responsible for meeting greenhouse gas emission objectives, governments may wish to retain some authority on the quantity of permits that are transferred to other Parties. No control of this kind was simulated here. • The emission objectives for the simulation were defined until 2012, and not beyond. Virtual companies had little incentives to put together longer-term generation strategies, or to bank CO2 permits for future use. • A uniform CO2 emission constraint was applied to all virtual companies, regardless of their initial situations in terms of CO2 intensity of generation or the respective emission objectives of different countries. In reality, CO2 objectives might very well differ across companies. Some governments may also choose to auction emission permits, and not “grandfather” them, as was done in this simulation. 27 IV. Further Developments This electricity and CO2 trading simulation was limited to a single sector: the power generation industry. The lessons drawn from this exercise are therefore specific to the electricity sector. The introduction of other emission sources in a simulation, and most of all, in a real-world emission trading system under the Kyoto Protocol, would probably deliver different results in terms of supply and demand, price levels, and production strategies. In particular, energyintensive industries may adopt new energy technologies in response to a Kyoto emission target. This would be likely to affect the power sector, as well as other energy supply industries. A simulation involving companies beyond the power sector could be useful in that respect. The Working Group “Climate Change” of UNIPEDE/EURELECTRIC envisions the participation of other industries in a second round of trading simulations. 28 Appendices Profiles of Virtual Companies in the year 2000 Description and Illustration of the Trading Tool (exchange) Definitions, Parameters and Rules of the Simulation 29 Profiles of Virtual Companies in 2000 Virtual Company - Power plants in 2000 Coal CCGT “ “ “ “ “ “ “ “ Wind power 520 MW 340 MW 335 MW 225 MW 250 MW 230 MW 100 MW Installed capacity Generation in 2000 CO2 Emissions VC 1 2 000 MW 9.5 TWh 5.56 Mt VC 2 - Nuclear Coal Hydro run of river Hydro reservoir Wind turbines 1000 MW 3 x 350 MW 2 x 100 MW 2 x 100 MW 50 MW 2 500 MW 14.7 TWh 4.18 Mt VC 3 - Nuclear Coal Fuel oil / gas Hydro run-of-river Hydro reservoir 1 000 MW 4 x 550 MW 3 x 350 MW 6 x 200 MW 4 x 200 MW 6 250 MW 25.6 TWH 8 MT CO2 VC 4 - Coal CCGT Hydro run-of-river 1 500 MW 600 MW 2 x 150 MW 4 x 150 MW 3 000 MW 15 TWH 9.6 Mt CO2 VC 5 - Nuclear Fossil fuel Hydro CHP Biomass 700 MW 500 MW 4 x 100 MW 2 x 200 MW 2 000 MW 12.5 TWH 2.1 Mt CO2 VC 6 - Coal 550 MW 330 MW 880 MW 3,2 TWH 3.2 Mt CO2 VC 7 - Nuclear Fossil Hydro 1 000 MW 600 MW 600 MW 2 200 MW 11 TWH 2.4 Mt CO2 VC 8 Fossil CCGT Hydro 8 x 500 MW 4 x 375 MW 80 MW 5 580 MW 32,1 TWH 26 Mt CO2 30 Virtual Company VC 9 - Power plants in 2000 Lignite Fuel oil Gas turbine Hydro Nuclear MW Coal fired Gas fired 1 200 M W 200 MW 320 MW 480 MW 6 x 1 000 Installed capacity Generation in 2000 CO2 Emissions 2 200 MW 11.312 TWh 9.3 Mt CO2 VC 10 10 000 MW 4 x 600 MW 4 x 400 MW 60 TWh 15 Mt CO2 VC 11 - Nuclear 1 000 MW Coal 550 MW Fuel oil / gas 3 x 350 MW Run-of-river 5 x 200 MW Hydro reservoir 4 x 250 MW Hydo-pumped-storage 700 MW 5 300 MW 19 TWh 3.8 Mt CO2 VC 12 - Coal CCGT Oil Peat Hydro 1 000 MW 3 x 350 MW 4 x 250 MW 2 x 120 MW 500 MW 3 790 MW 25 TWh 15.2 Mt CO2 VC 13 - Nuclear Coal fired Multifuel oil/gas Hydro 1 000 MW 1 050 MW 350 MW 3 x 250 MW 3 150 MW 13,5 TWh 3.2 Mt CO2 VC 14 - Hard coal fired Lignite Hydropower 1 000 MW 600 MW 50 MW 1 650 MW 7 TWh 6.4 Mt CO2 VC 15 - Nuclear MW Hydro power Wind power CHP coal CHP oil CHP biomass 10 x 1 000 16 200 MW 200 MW 5x135 MW 10x135 MW 5 x 135 MW 29 100 MW 146.2 TWh 4.7 Mt CO2 VC 16 - Fossil fired Combined cycle Hydro-reservoir Hydro-run of river 2 x 300 MW 3 x 330 MW 2 x 315 MW 2 x 80 MW 2 460 MW 11 TWh 4.9 Mt CO2 31 Description and Illustration of the Trading Tool (exchange) 1. General Page: Market Summary Note: The above prices are indicated for illustrative purposes only. They do not reflect a particular trading session. This is the opening page of the trading tool, which summarized the situation of all contracts available to participants: • spot electricity; • spot for CO2 emissions (which was only open in 2008, for trading permits banked in the first budget period); • CO2 emissions for delivery in 2007 (only open from 2001 to 2007); • CO2 emissions for delivery in 2012 (open throughout the simulation). For each contract, the following information is available : • Price à Price of the last trade • Var à Difference between the last trade and the settlement price of the previous trading session • High à Highest price of the trading session • Low à Lowest price of the trading session • Tradevolume à [Exchange volume] x [price] • Cash à Currency at your deposal for the trade • Stock à Valorization of the portfolio in real time 32 This page allows participants to record instantaneously an order on a specific contract. For example, a record about a quantity of 100 “CO2 emission spot” contracts at 43 Euros per tonne of CO2. The trader must simply select “buy” or “sell” to send the order on the market. If the participant wants to obtain more information (“bid” and “ask” prices, the volume and the number of participant behind each price) before recording an order, he or she must select a contract in order to access a page dedicated to that contract, as shown in the next section. 2. Example of a Market : CO2 EMISSION SPOT Legend: • “NB” • “ QTY” • “LIMIT” • “ BUY” • “SELL” Number of counterparts who have an interest on the indicated price Number of contracts available on the market Price limit that the participant want to sell or buy at Bid price - Participant can immediately sell at this price Offer price – Participant can immediately buy at this price A participant who want to record an order on the trading tool have to fill up each field and, at the end, to select “buy” or “sell”. 33 3. Information on Trading Activity Available to Each Participant The window “portfolio” gives information (price, volume and valorization) on transactions which have been executed, and their value. Here the purchase of 500 contracts on CO2 for delivery in 2012 (times 1000 tonnes, the size of the contract), at 42 Euros per tonne, for a total value of 21 000 times 1000 Euros. The window “order book” indicate : • “Valid order” orders which have not been executed yet • “Historic” history of the orders which have been executed in full • “RMNQ” remaining orders which are not fully executed 34 Definitions, Parameters and Rules of the Simulation 1- Objective of the Trading Simulation The purpose consists in simulating the possible operation of an international electricity and CO2 trading mechanism. The goal is to advance thinking on what could be the future functioning and rules of a greenhouse gas emission market. Because this game is initiated by the climate change working group of UNIPEDE/EURELECTRIC, it rests on the experience and specifics of the power sector. As the purpose of the trading simulation is to gather experience in electricity and emissions trading, the financial performance of each player will not be assessed. The IEA and ParisBourseSBF SA will produce a report on the simulation in collaboration with the working group on climate change of UNIPEDE/EURELECTRIC. 2- Description of Quiz Master / Advisor Quiz master : ParisBourse SBF SA Role of the quiz master : - Organization (technical support internet site, monitor and recordation of trades… ) - Provider of the trading simulation assumption (demand of electricity) and diffusion of information - Control of trade - Writing the report Adviser : International Energy Agency Role of the adviser : - Control the validity of trades (coherence between production and CO2 emissions) - Control of compliance with objectives - Lessons of the simulation 3- Development of the Trading Simulation 3-1 Simulation Periods 8 periods, including 2 grace periods, will be simulated: à 2000 – 2002 ( 3 years) à 2003 – 2004 ( 2 years) à 2005 – 2006 (2 years) à 2007 (1 year) à 2008 + grace period for the commitment period 2005-2007 à 2009 – 2010 ( 2 years) à 2011 – 2012 ( 2 years) à grace period for the commitment period 2008-2012 35 Two commitments period will be simulated: à 2005 – 2007 target = - 2% from 2000 level (average over 3 years) à 2008 – 2012 target = - 5% from 2000 level (average over 5 years) The grace periods will provide an opportunity for players who have not complied with their objectives at the end of the period (2007 or 2012) to buy permits from others who have banked extra reductions. The overall assessment of compliance will be done after the grace periods are over. A penalty will be applied to all emissions above the objective at the end of each budget period, after the grace period has expired. The penalty will be set at 150% of the highest executed price observed during each period. 3-2 What will be traded and how ? Players will trade both electricity on a spot market and CO2 permits on futures and spot markets. They will rely on an electronic system to trade both commodities and manage their portfolio. Further information on the trading internet site will be forwarded to participants prior to the beginning of the trading simulation. Electricity Contract size: 1 000 MWh Trading price: Euros by MWh Each player must balance supply and demand for every period and can do so either through its own production or through trading. This reflects the usual international electricity market place. In order to avoid a situation where a virtual company stops producing electricity and imports everything from another company we impose an arbitrary limit on electricity imports: imports should not represent more than 7% of total demand for each company and each period. Illustration of an electricity trade Each virtual company is obliged to meet its electricity demand each year, but can do so through its own production or through imports. Similarly, each company can sell electricity to another company. This trading activity will take place on the trading simulation site as the CO2 trading. In 2003: Player 1 wants to sell 3000 MWh (corresponding to 3 contracts) at 35 Euro per MWh. It records this offer on the trading system. If Player 2 is willing to acquire electricity at 35 Euro per MWh, it will record an order to buy at that price and a corresponding quantity. The trade will then be executed. If Player 2 wanted to acquire more than 3 contracts, the remainder will still be on the trading system after the transaction is executed. Counterparts have to agree on a single price. The system itself will not adjust prices to match supply and demand. 36 Notes: - There will be no future transactions on electricity. - We assume no default on the transactions of electricity CO2 Contract size: 1 000 tons of CO2 (not tons of carbon, this is important for pricing) Trading price: Euros per ton of CO2 (no decimal points) Different maturity dates will be available on the CO2 trading market: - One future with a 2007 maturity. This will correspond to emission reductions beyond the player’ objective achieved over the 2005-2007 period. The future is s the commitment between two counterparts to deliver (or receive) certified emission reductions in 2007. - One future with a 2012 maturity. This future is the commitment between two counterparts to deliver (receive) certified emission reductions in 2012. - A spot market available only from 2008 on, for trading of CO2 units banked in 2005-2007. This spot market will continue provided there are banked units from the first commitment period available for sale in the second commitment period. Illustration of a CO2 trade In 2002: Player 1 wants to acquire 10 000 tons of CO2 (corresponding to 10 contracts) to be used over the 2005-2007 period. It puts an order corresponding to a buy of 10 contracts with a 2007 maturity on the electronic trading system to realise this transaction at 85 Euro per ton. Player 2 wants to sell 8 000 tons of CO2 (corresponding to 8 contracts) which it hopes to reduce beyond its target between 2005 and 2007. It puts an order corresponding to the sale of 8 contracts with a 2007-maturity on the system. Seeing that an interest to buy 10 contracts of CO2 at 85 Euro per ton is on the trading system, player 2 records its interest to sell 8 contracts at this price. The electronic trading system matches the two orders. Player 2 has sold 8 contracts (8 000 tons of CO2) at 85 Euro per ton. Player 1 has bought 8 contracts at 85 Euro per ton. Player 1 still needs 2 contracts, and its interest is still on the trading system at 85 Euro per ton. Trading principles (electricity or CO2) This example shows that the matching is made through the price. There will be no transactions if the buyer’ price is under the seller’ price. Players will need to s s modify their orders until they find a common price. In order to facilitate the development of the liquidity on each contract, players will be enable to realize arbitrage operations. For instance, a player may decide to buy more units than it needs if it expects to be able to sell them at a higher price later on (and vice versa). As stated above the financial performance of virtual companies will not be assessed as this is not the purpose of this game. - 37 Important trading information All future trades commit the seller to deliver the agreed reduction units to the buyer. Even if it is not in a position to do so, the quiz master will deduct the agreed units from the seller’ s account. This brings the seller into non-compliance: its emissions are above its emission rights. Because the defaulting trade brings the seller into non-compliance, the seller will pay the penalty on all the units it could not deliver (reminder: the penalty is equal to 150% of the highest executed price observed in the period). The buyer will not be considered in default and we will accept to add the units bought regarding the emission target. These extra-emissions will also be deducted from its emission rights in the next budget period (i.e. subtracted from the objective). There is therefore a financial and a physical penalty for selling units which are not available. Important trading information All transactions must be conducted through the trading site, without prior agreements and conversations between the counterparts. No other transaction than those executed on the site will be recorded. Players without access to the trading simulation site Players who are not able to access the game site can use the quiz master as a broker sending him trading orders before 10:00 AM of each trading session. At the end of the day the quiz master will return the results of his negotiation to the players. It is important that the trading orders of those who cannot access the site be forwarded by email to the quiz master. The quiz master will forward information, via email, on market prices to the players without access to the site. This will enable these players to change their orders if they wish, or if their previous orders have not been executed. 3-3 Timing of the Game In order to propose a less constraining trading session, players will have one week to finalize their decision, sell / buy electricity or sell / buy CO2 emission units, from the moment they receive the electricity growth assumption. Trades will take place each Tuesday from May 25 until July 13 between 10:00 and 12:00, Paris time. 38 Quiz master: Electricity growth assumption 2001 - 2002 2005-2006 2007 2008 +GP (*) 2011-2012 2009-2010 2003-2004 Calendar 0518 0525 0601 0608 0615 0622 0629 0706 0713 2003-2004 2001 - 2002 2005-2006 2007 2008 +GP (*) 2009-2010 GP(*) 2011-2012 (*) : Grace Period Trading days Illustration for 2003-2004 • The quiz master communicates the level of electricity demand growth on Tuesday, May 25. • Participants can trade electricity and CO2 emission on June 1. 3-4 Timing and Feedback from Players Once the quiz master has given growth assumption, players have one week to finalize their decision before using the trading system to exchange electricity or CO2 emission units. This trading system will be available on the simulation site each Tuesday from 10:00 AM to 12:00 AM (technical information will be available on the simulation site). After each trading session, players have to return the following tables (Cf. the schedule of the players feedback) to the quiz master on Wednesday: 39 Note: This table would either contain annual or biannual information. â Table-1 Type of plant I Capacity (MW) Gas old Gas new (CCGT or Cogeneration) Solid fuel old Type: Solid fuel new Type: Fuel oil – old Fuel oil – new Peat – old Peat – new Hydro-large Hydro-small Biomass Wind Nuclear II Production (MWh) III Efficiency (%)*** IV Fuel consumption (MWh) V Fuel consumption (GJ)* VI CO2 emissions (tCO2)** 0 0 0 0 0 ---- Total * 1MWh = 3.6 GJ. V = IV * 3.6 ** Use the coefficients indicated in the table below: *** In the case of co-generation, the efficiency is computed as the ratio of electrical output to primary energy. CO2 coefficients Fuel Anthracite – Bituminous coal – Sub-bituminous coal Lignite Oil – (gas/diesel oil) Natural gas Peat Biomass t CO2 / GJ 0.0983 0.0946 0.0961 0.1012 0.0741 0.0561 0.1060 0 â Indicate volume of traded electricity (global volume) as well as production, to allow for a reconciliation of annual production, imports and exports of electricity. Demand must always be equal to production plus imports minus exports. Table-2 Production (TWh) Import of electricity Export of electricity Production + import Electricity demand - export The price of traded electricity should also be given to the quiz master, which will keep it confidential. â Traded CO2 emission units (characteristics of each deal: which volume, maturity and price, also confidential) 40 Table-3 Maturity Purchase price (Euro per tCO2) Purchases (in tCO2) Sales price (Euro per tCO2) Sales (in tCO2) Permits for 2005 – 2007 Permits for 2008 – 2012 Spot for 2005 - 2007 in 2008 * Spot for 2008 - 2012 in 2013 ** * 2008 in this case represents the grace period during which final adjustments in CO2 trading positions are made for the first budget period. This box can only be filled in 2008. ** Same as above for 2013 â After the grace period of each commitment period (2005-2007 and 2008-2012), players must return an emission report to the quiz master: Table-4 : Period 2005-2007 Year I Objective* 2005 2006 – 2007 Grace period (2008) Total ------- II Bought (***) III Sold (***) IV Objective + B–S ------- V Emissions IV – V** Compliance? (yes if >= 0) ------- --- 3 * (98% * 2000 level) * Indicative for interim years, as compliance is only measured at the end of the period, for the whole period. ** Only the total matters, as there is no obligation to match emissions with permits until the end of the budget period. *** Permit for 2005-2007 Table-5 : Period 2008-2012 Year I Objective* 2008 2009 – 2010 2011 – 2012 Grace period (2013) Total --------5 * (95% * 2000 level) II Bought (*) III Sold (*) IV Objective + B–S --------- V Emissions IV – V** Compliance? (yes if >= 0) --------- (*) Permit for 2008-2012 â Players must inform the quiz master about their investment decisions at the time they are made. These investments must respond to the following constraints The quiz master and advisor will check on the realism of the investment behavior. Players are expected to undertake investments in a “reasonable” fashion. For this reason, we do not require that companies indicate the age of current capacity. Capacity additions can be rejected by the quiz master if they seem exaggerated, e.g. if a company were to renew its total capacity over the simulation period. 41 - Time necessary to bring new production on line - Coal à 5 years - Gas plant à 3 years - Nuclear à 10 years - Large hydro à 8 years - Small hydro à 3 years - Wind à 1 year - Biomass à 3 years - Fuel switching à 1 year - Co-generation à 3 years Investment cost data Plant type CCGT Coal Lignite Oil Co-generation coal Co-generation gas Gas conversion of coal Hydro-large (>100MW) Hydro-small Wind Biomass Nuclear Euro / kW 460 1 290 1 310 1 080 ? ? 46 1 650 1 300 1 150 1 540 1 750 Minimum size (MW) 300 500 500 500 200 200 Not applicable 1 000 20 10 50 1 000 Project life (years) 20 20 20 20 ? ? 10 40 40 20 20 40 - â The players are supposed to respect usual re-fuelling and other technical constraints when deciding on the operations of the different plants in order to reflect as much as possible the real operation of the different types of equipment. â For the sake of simplicity, the primary energy prices will be kept constant over the whole duration of the trading session. We are aware that these prices may currently differ from utility to utility. In order to assure coherence, the players are requested to forward their primary cost information for the game to the quiz master, who will keep this information confidential. 3-5 Schedule of the Players Feedback Calendar Simulated periods Table to return to the quiz master Table-1 + table-2 (*) + table-3 (*) Table-1+ table-2 (*) + table-3 (*) Table-1+ table-2 (*) + table-3 (*) Table-1 + table-2 (*) + table-3 (*) + table-4 (*) Table-1+ table-3 (*) + table-4 (*) Table-1 + table-3 (*) + table-5 (*) Table-1 + table-3 (*) + table 5 (*) Table-3 (*) and table-5 25th of May 2000-2002 1st of June 2003-2004 8th of June 2005-2006 15th of June 2007 22th of June 2008 + Grace period 2005-2007 29th of June 2009-2010 6th of July 2011-2012 13th of July Grace period 2008-2012 (*) Table to report only if trades have taken place. 42 3-6 Feedback from the Quiz Master Each Tuesday, the quiz master communicates to each player (e.g. each virtual company) the overall electricity growth assumption for the trading period that will take place the following Tuesday: • The “standard” growth will range between 0% and 2% annually; • There will be one “accident” per virtual company during one of the two budget periods (either 2005-2007 or 2008-2012). For that year, the company will be subject to an exceptionally high growth in its demand (4%).1 The quiz master will also announce the coming on line of new equipment as well as its type, but not its precise efficiency, supposed to remain confidential. Note that this efficiency should be forwarded to the quiz master. Each Friday the quiz master will diffuse the results of each session: The global volume of the electricity traded; Traded quantities and counterparts (who sold to whom, and how much); Cumulative emissions for the period and corresponding objectives for each participant, including trades. 1 This is an alternative to a proposal that was made to have the quiz master introduce climate accidents (low precipitation, high number of degree-days, etc.), which would have been more difficult to model. 43