Scenario Analysis of Power Mix in Taiwan

Scenario Analysis of Power Mix in Taiwan Chia-Hao Liu, Ching-Han Deng, Chia-An Chang* and Fu-Kuang Ko Institute of Nuclear Energy Research Abstract Carbon emissions have become a critical issue concerned by energy sectors. It is essential to take multiple subjects or various factors into account when adjusting a proper energy portfolio. In order to find the proper energy portfolio form unbiased systematic results, all possible combinations of parameters should be included in the scenario analysis. This study analyzes three scenarios (Non-nuclear, Life extension, Nuclear in full) achieving the national carbon reduction target of Master Action Plan of Energy Conservation and Carbon Reduction through adjusting power mix and imposing carbon tax (400 USDs/ton-CO 2 ) and purchasing carbon credits. Moreover, we further discuss the possible condition which the carbon reduction target is achieved without purchasing carbon credits. There are important findings as below: In 2025, the Non-nuclear scenario levies carbon tax, 400 USDs/ton-CO 2, for a contribution to 8% carbon reduction. Under the condition of maintaining stable economic growth, sustained developments of low-carbon generation technologies are still necessary to achieve carbon reduction target. Purchasing plenty of carbon credits also makes contribution to carbon reduction being up to 54%. In order to achieve carbon reduction target without purchasing carbon credits, the electric demand will be drastically suppressed to almost zero growth. The economic impacts reflect high marginal carbon reduction cost, 5,600 USDs/ton-CO 2, as almost twice as the value via LNG in 2010. The result reveals that a large carbon reduction gap would exist in 2025. We suggests that Taiwan should learn lessons from Japan s practical experience to re-examine carbon-reduction targets and to consider other strategies, such as extension of existing nuclear power plants lives, curtailment of LNG s reserve and building coal-fired plants etc. --- keywords: Electricity mix, Carbon reduction, Energy modeling 1. Introduction Before examining Taiwan s Power Mix, this study analyzed trends of economic development, energy consumption and CO 2 emissions etc. Figure 1 presents the trend of Taiwan's GDP and CO 2 emissions from 1990 to 2013. Real GDP has grown up steadily over the years. There had been a decrease in CO 2 emissions since 2008 and CO 2 emissions never exceeded the highest record in 2007. In 2013, real GDP was 15,343,607 million NTD (at 2006 Prices) with an ~ 1 ~

annual growth rate of 2.09%; CO 2 emissions are 250,297 kton with an annual growth rate of 0.67%. (Bureau of Energy, 2014b); energy intensity is 7.46 LOE/kNTD. On the other hand, observed relationship between energy consumption and GDP indicates that energy intensity has decreased since 2002. In recent years, expansive negative decoupling energy consumption from GDP helps to promote carbon reduction. 16000 14000 12000 10000 8000 6000 4000 2000 GDP: Billion NTD 300 250 200 CO2:million ton 150 100 50 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 GDP 5317 5736 6169 6585 7084 7536 7954 8389 8680 9198 9731 9571 1007 1044 1109 1161 1224 1297 1307 1283 1421 1479 1498 CO2 109 118 126 135 143 150 158 171 181 190 209 213 221 231 239 245 252 256 245 232 248 254 249 0 Fig.1 Historical trends of GDPand CO 2 emissions Then we analyze the domestic energy system, the domestic demand side of energy consumption trends. In 2013, the percentage of total energy consumption in electricity is 48.6%. Oil and petroleum products are 38.9%. Coal and coal products are 8.7%. Natural gas is 3.2%. Renewable energy is 0.4%. The long-term domestic energy consumption trend is shown as Figure 2. Therefore, electricity is crucial to domestic energy consumption system. Long-term trend is shown as Figure 3. Domestic thermal power generation in 2013 is the dominant source of electricity in Taiwan. Coal-fired plants generated 121.27 TWh, accounting for 48.1% of total output, followed by 69.52 TWh (27.6%) from gas-fired plants, 41.64 TWh (16.5%) from nuclear power plants, 10.8 TWh (4.3%) from renewable energy, 5.94 TWh (2.4%) from oil-fired plants and 3.19 TWh (1.3%) from pumped-storage hydroelectric. (Bureau of Energy, 2014b) Moreover, Taiwan's energy system is highly dependent on importing energy for supporting demostic energy supply. The annual energy statistics from the Bureau of Energy indicated that, in 2013, Taiwan s extremely high dependence on imported energy and petroleum was 97.58% and 99.98%, repectively. (Bureau of Energy, 2014a) ~ 2 ~

Rewable Energy Electricity Natural Gas Oil and Petroleum Coal Fig. 2 Historical trends of energy comsumption Rewable Energy Nuclear Gas-fired Generation Oil-fired Generation Oil-fired Generation Coal-fired Generation Fig. 3 Historical trends of generation 1 On the other hand, developed and developing countries have explicit tendencies toward international carbon reductions and targets are clearly developed in recent years. U.S. s policy is that the quantity of carbon emissions in 2020 should be reduced to 83% the level of that in 2005. EU s policy is that the quantity of carbon emissions in 2020 and 2030 should be reduced to 80% and 60% the level of that in 2005, repectively. Japan s policy is that the quantity of carbon emissions in 2020 should be reduced to 75% the level of that in 1990. Indonisa s policy is that the quantity of carbon emissions in 2020 should be reduced to 80% the level of that in 2005. China s policy is that the intensity of carbon emissions in 2020 should be reduced to 55-60% the level of that in 2005. South Korea sets 2020 targets are reduced to 70% the level of baseline scenario in current year. Due to carbon dioxide 1 Renewable energy includes of hydro, wind, solar, biomass and waste. ~ 3 ~

emissions caused by extensive use of fossil fuels, there are potential threats ensuing in future sustainable developments, such as discussed in responsibility sharing of international carbon reduction, EU s Energy Using Products Directive (EuP), carbon tariffs and voluntary carbon labeling in U.S. These international issues should be seriously considered. Taiwan has faced multi-dimention pressures and impacts from power mix, environment changes, scarce energy resources, fluctuations of international energy prices, energy demand raise and global climate changes. Taiwan governorment has been taking energy and economic into account in order to persue energy safty of supply and demand. As mentioned, Taiwan governorment plays a predominant role who proposed a significant series of important polices related to energy savings and carbon reductions, such as Framework of Taiwan's Sustainable Energy Policy (2008), National Action Plan on Energy Conservation and GHGs emission Reduction (2010), Energy Development Guideline (2012), Master Action Plan of Energy Conservation and Carbon Reduction (2014) etc. Now, Taiwan aims to develop clean energy and willings to reduce nationwide CO 2 emission, so that total emission could return to its 2008 level between 2016 ~ 2020, and further reduced to the 2000 level in 2025. While Taiwan s power mix is transition now, researches of future power mix and carbon reductions are surely urgent and nessesary. This research applied the MARKAL-ED model to analyze Taiwan s power supply and demand, provide a feasible solution, aesign scenario analysis and find the path to achieve the reduction target. Finally, this research assesses effects based on different scenarios and assist Taiwan in developing power system roadmap for national target, energy saving and carbon reductions. 2. Methodology In this study, we applied MARKAL-ED (MARKet Allocation- Elastic Demand) model to analyze the entire energy system in Taiwan. Through the utilization of a systematic approach, it is possible to take into account the interaction between setors and to show the detailed results and impacts of different power mixes over a multi-period time horizon. MARKAL was developed within the Energy Technology Systems Analysis Programme (ETSAP) set up by International Energy Agency (IEA) in 1976 and currently used by 77 institutions in 37 countries around the world. MARKAL is a partial equilibrium linear programming model which provides a bottom-up view and a technology-rich basis for energy system simulation. The model tries to supply energy at minimum net total cost to fullfil the endogenous energy service demands; meanwhile, it also makes investment and operation decisions for energy-consumed equipment. The final solution is considered to be optimal under the objective function of cost minimization. MARKAL-ED is one of the MARKAL family modules. It concerns an ~ 4 ~

important economic concept: under perfect competition, the market equilibrium price and quantity can be achieved when maximizing the sum of producer and consumer surplus. In MARKAL-ED, supply options and energy service demands are both computed by the model. Each energy service demand changes as a fuction of its market price. Besides, the results from MARKAL-ED should be taken as cost-effective scenarios rather than predictions because of the perfect foresight assumption of the model. Various parameters used in MARKAL-ED were discussed in advance. 3. Assumptions Energy service demands are crtitcal exogenous inputs for MARKAL since the model is mainly driven by those demands. The model aims to reach equilibrium by getting a least-cost set of technology portfolio for energy system based on fulfilling the energy service. When estimating future energy service demands, some statistics such as GDP, population, household number, and energy prices are required. Following sections discuss in detail the parameters used to estimate the service demand. 3.1 Economic and Population Growth The historical economic growth rate used in this paper was based on data released by Directorate-General of Budget, Accounting, and Statistics, Executive Yuan, Taiwan. In 2014, the gross domestic product (GDP) growth rate is 2.98%. In this study, we eatimated that average economic growth rate during the period 2014 to 2025 will decline to around 2.93% due to halting construction of fourth nuclear power plant(4npp). Population statistics including the future year forcasting result in this paper was collected from Department of Household Registration. In 2014, total population is approximately 23,407 thousand. Owing to decreasing in birth rate, total population will start declining in 2019, and the population growth rate will become -0.092% / year during 2021 to 2025. Besides, the estimated household number increases over time but in a declining rate. Table 1 shows the values adopted in this paper. 3.2 Energy Price The long- term prediction of primary energy prices is crucial for the MARKAL-ED model since these prices are used to calculate energy service costs. The high dependence on import of primary energy in Taiwan makes domestic energy prices are largely rely on the global market. In this study, long run price projections released by credible international institutes such as International Energy Agency are used. Shown as Table 2, the primary energy price in 2025 includng crude oil (79.19 USD/barrel), LNG (14.13 USD/Mbtu), steam coal (78.63 ~ 5 ~

USD/tonne), and coking coal (170.75 USD/tonne). For crude oil, price forecasting is based on Annual Energy Outlook 2013 (AEO). For LNG, steam coal, and coking coal, price forecasting is based on World Energy Outlook 2013 (WEO). Table 1 Forecast values for Taiwan s socio-economic status Period Real GDP growth Household Population Houshold rate under Reference size growth growth rate growth rate Case Scenario rate (% / year) (% / year) (% / year) (% / year) 2011-2015 2.403 0.187 0.836-0.643 2016-2020 3.000 0.016 0.306-0.289 2021-2025 2.744-0.092 0.129-0.221 2011-2025 2.767 0.039 0.474-0.433 Source: estimated by this study Table 2 Forecast values for Taiwan s imported primary energy price Crude oil LNG Steam Coal Coking Coal (2001USD/barrel) (2001USD/Mbtu)(2001USD /tonne) (2001USD/tonne) 2010 55.65 9.21 74.47 140.31 2015 67.89 13.05 71.18 147.02 2020 73.78 13.74 75.81 161.26 2025 79.19 14.13 78.63 170.75 Source: estimated by this study 4. Scenario Design In this study, we focus on discussing how to allocate the power generation technology mix if the CO2 emission target has to be achieved. Based on Master Action Plan of Energy Conservation and Carbon Reduction, the targets on GHGs emission reduction are the amount of emissions in 2020 decrease to the amount (244Mt CO2) in 2005, and decrease the amount of carbon dioxide emissions in 2025 to the amount in 2000 (208Mt CO2). Three scenarios are designed and named as Non-nuclear, Life extension and Nuclear in full. The details and assumptions are presented in Table 4. Based on the above table, the main ways in Non-nuclear scenario to achieve the nation s CO 2 emission target include enlarging gas-fired generation, promoting renewable energy, and enforcing economic political actions. In Life extension scenario, the life extension of existing nuclear power plants provides one more solution to achieve the nation s CO 2 ~ 6 ~

emission target. Besides, the Nuclear in full scenario has the other solution which it assumes that No.1 and 2 reactors in NPP4 will return to operate in 2020 and 2025 respectively. Finally, the characteristics of three scenarios are listed in Table 5. Table 4 Scenario settings CO 2 emission target To achieve CO 2 emission target in 2020 &2025 Scenario Non-nuclear Life extension Nuclear in full Energy service demand The demand is estimated under the medium GDP growth and low population growth. Nuclear To halt 4NPP construction;1npp and 2NPP decommission as scheduled To halt 4NPP construction;to keep the 1NPP and 2NPP in operation before 2025 4NPP resume to work;to keep the 1NPP and 2NPP in operation before 2025 Coal-fired Following the setting of Taipower Long-Term Power Development Program before 2020 CCS Without commercial operation before 2025 Annual LNG import Renewable energy Economic political actions Maximum supply 1 Full scale promotion of renewable energy To levy carbon tax (400 NTD/ton) 2 and purchasing carbon credit Note:1. Bureau of Energy(2013)announces that the maximum supply of annual LNG import is 14 million tons in 2013, 15 million tons in 2019, and 20 million tons in 2024. 2. Carbon tax is reported by Ministry of Finance (2011). ~ 7 ~

Table 5 The characteristics of three scenarios Scenario Advantage Disadvantage Non-nuclear Life extension Nuclear in full 1. No dispute on nuclear power generation. 2. The pressure of carbon reduction is mitigated by technology development with economic actions. 1. Low power generation cost. 2. Low CO 2 emissions. 3. Low carbon credit. 1. The dependence to fossil fuel is declined. 2. The most possible scenario to reach the CO 2 emission target. 1. High power generation cost. 2. People and industries may take against levying carbon tax. 3. More carbon credit have to be traded. 1. The existing nuclear power plants keep in operation, the dispute is unavoidable. 1. 4NPP resume to work against the current energy policy. Public opinion fights back strongly. 5. Results and Discussion Table 6 and Figure 8 show the results of 2010-2025 power generation capacity of each scenario. In Non-nuclear scenario, the existing nuclear power plants decommission as scheduled, and 4NPP is mothballed. Furthermore, new coal-fired plants are limited by following Taipower s program. Gas-fired plant is certainly the main power generation source, and its capacity is up to 42% of the entire electricity system in 2025. Besides, under the promotion from government agency, the total capacity of renewable energy such as onshore wind and solar power also grows from 5% in 2010 to 15% in 2025. In Life extension scenario, the development of most power generation technologies have similar trends with that in the Non-nuclear scenario. Gas-fired plant is also the main electricity generation source in 2025, and its capacity is up to 41% of the entire electricity system, and then the capacity of renewable energy in Life extension scenario is the same as in Non-nuclear scenario. However, the life extension of existing nuclear power plants results in the percentage of its capacity still maintain in 8% of the entire electricity system. In Nuclear in full scenario, the capacities of gas-fired plants and renewable energy power plants are 41% and 15% of the entire electricity system in 2025. In addition, the capacity of nuclear power plants is up to 7,844 MW due to 4NPP resumes to operate since 2020. The dependence of using coal-fired plant as the option of base load power generation decreases significantly and the capacity is below to 20% of the entire electricity system. ~ 8 ~

Table 6 The capacity in each scenario (Unit: MW) coal-fired gas-fired oil-fired nuclear pump storage renewable energy co-generation Total 2010 Non-nuclear 11,897 15,194 3,625 5,144 2,602 2,475 7,943 48,882 Life extension 11,897 15,194 3,625 5,144 2,602 2,475 7,943 48,882 Nuclear in full 11,897 15,194 3,625 5,144 2,602 2,475 7,943 48,882 2015 Non-nuclear 10,697 16,615 2,710 5,144 2,602 3,727 6,125 47,620 Life extension 10,697 16,615 2,710 5,144 2,602 3,727 6,125 47,620 Nuclear in full 10,697 16,615 2,710 5,144 2,602 3,727 6,125 47,620 2020 Non-nuclear 14,697 18,541 1,431 3,872 2,602 5,656 4,748 51,547 Life extension 14,588 18,658 1,431 5,144 2,602 5,656 4,748 52,826 Nuclear in full 12,869 18,665 1,431 6,494 2,602 5,656 4,748 52,464 2025 Non-nuclear 17,126 24,684 831 951 2,602 8,904 3,739 58,837 Life extension 13,588 24,767 831 5,144 2,602 8,904 3,739 59,575 Nuclear in full 11,869 24,821 831 7,844 2,602 8,904 3,174 60,045 Figure 8 The capacity in each scenario The net electricity generation and its trends are presented in Table 7 and Figure 9. The results show that the net electricity generation of each scenario is among 322.4 to 330.8 TWh with a growing trend from 2010 to 2025. Comparing to the statistical data in the base year, the net electricity generation increases almost 40% in 2025. It is easy to perceive the contrast ~ 9 ~

between the more net electricity generation in the Life extension scenario and less in the Non-nuclear scenario, and it indicates that life extension of existing nuclear power plants results in the mitigation on the reducing electricity demand. Table 7 The net electricity generation in each scenario (unit: TWh) 2010 2015 2020 2025 Non-nuclear 233.5 256.8 281.7 322.4 Life extension 233.5 256.8 290.2 326.9 Nuclear in full 233.5 256.8 287.5 330.8 Figure 9 The net electricity generation in each scenario The result in Figure 10 shows that the amounts of CO 2 emission in each scenario at 2020 and 2025 are higher than the national CO 2 emission target of the Master Plan in Taiwan. The breach of CO 2 emission target is especially significant in 2025, and it indicates that the emission target of the Master Plan could not be achieved if the agency only considers adjusting the ratio of low-carbon electricity generation with levying 400NTD/ton carbon tax. Trading the carbon credit is necessary to raise the amount of CO 2 emission permissions. In addition, from the view point of the comparison between scenarios, the amount of CO 2 emissions could be reduced at least 10 and 15 million tons in 2020 and 2025 individually, if the 4NPP resumes to operate since 2020. ~ 10 ~

Figure 10 The CO 2 emissions in each scenario This research sets two carbon-credit prices as references. The lower one refers to CER (Certified Emission Reduction) average price set as 3.53 NTD/ton-CO 2 in late May 2014; the higher one refers to Taipower Company (TPC) andgreenhouse Gas Reduction Act set as 789 NTD/ton-CO 2. Table 8 summarizes all scenarios breach of carbon reduction and carbon credit cost. The breaches of carbon reduction are 8.3-18.3 and 47.6-84.4 million tons in 2020 and 2025, respectively. Based on resent carbon credit price, carbon credits take 29.3-64.6 and 168-297.9 million NTD in 2020 and 2025, respectively. The expenditures are 6,549-14,439 and 37,556-66,592 million NTD when the carbon credit price rises to 789 NTD/ton-CO 2. 2020 2025 Table 8 CO2 emission, breach of carbon reduction and carbon-credit price in 2 nd stage Note 1 : 3.53 NTD/ton-CO 2 Note 2 : 789 NTD/ton-CO 2 Non-nuclear Life extension Nuclear in full Emission target CO 2 emission (Mton) 263.5 262.9 252.3 244 breach of CO 2 reduction (Mton) 18.3 17.7 8.3 low carbon credit cost 1 (MNTD) 64.6 62.5 29.3 high carbon credit cost 2 (MNTD) 14,439 13,965 6,549 CO 2 emission (Mton) 293.8 271.2 255.6 208 breach of CO 2 reduction (Mton) 84.4 61.8 47.6 low carbon credit cost 1 (MNTD) 297.9 218.2 168 high carbon credit cost 2 (MNTD) 66,592 48,760 37,556 ~ 11 ~

Taking the Non-nuclear scenario in Fig. 11, major medium of carbon reduction is power mix adjusting (65%) assisted in carbon credits (25%) in 2020 and change to buy carbon credits (54%) in 2025. Carbon credits penetration of carbon reduction rises to double amount (50%) and power-mix adjustment due to power mix adjusting only reduce 38% national CO 2 emissions. In order to achieve 2025 s carbon reduction target of Master Action Plan of Energy Conservation and Carbon Reduction, expanding usage of LNG and renewable energy without nuclear will need to buy large more amount of carbon credits. Relatively, Life extension scenario has more than 50% carbon reduction due to power mix adjusting in 2020 and 2025. In 2025, it s necessary to buy extra carbon credits equal to 39% national CO 2 emissions. Fig.11 penetrations of carbon tax, carbon credit and power mix adjusting Comparison with two scenarios, Non-nuclear and Life extension, indicates that extension existing nuclear plants is able to sharply increase the penetration of power mix adjust, which can decrease demands, dependence and expenditure of buying carbon credits. If carbon credit price grows up to 789 NTD/ton-CO 2 in the future, the price gap between two scenarios, Non-nuclear and Life extension, will reach to 17.83 billion NTD and donate significant benefits to cost-savings of carbon credits. Otherwise, comparison with two scenarios, Life extension and Nuclear in full, indicates that Nuclear in full in 2020 will promote penetration of power mix adjusting to 77%. Carbon credit expenditure will so sharply decrease that only 8% and 27% national CO 2 emissions are needed in 2020 and 2025, respectively. As discussed, even Taiwan power mix set as Life extension based on 60% more ~ 12 ~

low-carbon generation, like gas-fired, nuclear and renewable energy in 2025, this research estimated national carbon emissions would be so much higher than the target of Master Action Plan of Energy Conservation and Carbon Reductio that abroad buying large amount of carbon credits is certainly needed. However, in order to emphasis that purchasing carbon credit is not the only solution to reaching the nation s emission reduction target, this study also evaluate the impacts on electricity generation and emission reduction cost without purchasing carbon credit in Life extension scenario. The result indicates that significantly reducing electricity demand and high-carbon generators outputs and introducing low carbon or high-effeciency technologies are major means undertaken. Significantly reducing electricity demand (nearly zero growth in electricity demand) might cause economic impact. On the other hand, usage of low carbon or high-effeciency technologies will increase electricity generation cost. The marginal abatement cost in 2025 will be 5,600 NTD/ ton, twice the amount compared to the abatement cost which relies only on excessive usage of nature gas. It is suggested that the government have to figure out coping strategies basing on the consideration of trade-off relationships embedded in each strategy. 6. Conclusions and Suggestions In response to the global GHGs mitigation trend, Taiwan government has set ambitious CO 2 reduction target. The result indicates that adjusting power generation mix including excessive usage of nature gas, development of renewable energy could make low-carbon electricity account for more than 60% in 2025 (Life extension scenario); however, compared with the government s CO2 reduction target, the gap would still exist. It is necessary to incorporate economic political actions such as purchasing carbon credit into consideration. For example, in Non-nuclear scenario, the attribution of purchasing carbon credit account for 54% the amount of carbon reduction in this study, and levying carbon tax (400NTD/ton) accounts only for 8%. Reducing electricity demand (nearly zero growth in electricity demand) might be a solution to achieve the emission reduction target, but it will cause economic impacts as well. The abatement cost is another important factor. The marginal abatement cost of Non-nuclear scenario will increase to 5,600 NTD/ ton in 2025, twice the amount compared to the abatement cost which relies only on excessive usage of nature gas. It is suggested that the government have to figure out coping strategies basing on the consideration of trade-off relationships embedded in each option. Besides, the usage of nuclear power is a critical issue in Taiwan. In this study, we examed different levels of usage of nuclear power corresponding to the three scenarios. Finally, according to the three major energy policies, application of renewable energy has been a core ~ 13 ~

manner implemented in Taiwan. However, the result shows that even if the development of renewable energy meets the government s anticipation, emission reduction gap will still exist. It is necessary to further evaluate the possibilities and effects of enlarging renewable policy objectives, keeping the 1NPP and 2NPP in operation, making 4NPP resume to work, or purchasing carbon credit. After all, the government should adopt rolling management, and adjustments and changes can be made to energy policies over time. References 1. Energy Statistical Handbook (Bureau of Energy, 2014a) 2. Statistics of CO 2 Emission form Fuel Combustion in Tiawan. (Bureau of Energy, 2014b) 3. Framework of Taiwan's Sustainable Energy Policy (Excutive Yuan, 2014) 4. Master Action Plan of Energy Conservation and Carbon Reduction (Ministry of Economic Affairs, 2008) 5. Energy Development Guideline (Ministry of Economic Affairs, 2012) ~ 14 ~