89112 LEVELING THE FIELD FOR RENEWABLES: MEXICO’S NEW POLICY FRAMEWORK FOR INCORPORATING EXTERNAL COSTS OF ELECTRICITY GENERATION April 2014 THE WORLD BANK Contents Abbreviations and Acronyms ....................................................................................................... vii Units of Measure .......................................................................................................................... viii Currency Equivalents ................................................................................................................... viii Executive Summary ..................................................................................................................... ix Chapter 1. Introduction .............................................................................................................. 1 Study Purpose and Objectives ........................................................................................................ 3 Structure of this Report ................................................................................................................... 3 Chapter 2. Mexico’s Policy Context for Incorporating Externalities ..................................... 5 Power Sector Overview .................................................................................................................. 5 Investment Outlook ......................................................................................................................... 7 Policy Mandate to Internalize External Costs and the New Methodology ................................... 10 Critical Steps and Potential Barriers to Implementation............................................................... 14 Chapter 3. Valuation of Externalities ...................................................................................... 17 Methodology ................................................................................................................................. 17 Review of Data Available in Mexico............................................................................................ 20 Meteorological Data and Regional Dispersion Domain ....................................................... 20 Exposure Response Functions (ERF) ................................................................................... 21 Economic Costs .................................................................................................................... 22 Externalities .......................................................................................................................... 24 Encompassing Climate-Change Externalities ....................................................................... 25 Key Recommendations ......................................................................................................... 27 Chapter 4. Internalizing Externalities ..................................................................................... 28 Including externalities in the Economic Evaluation Process ........................................................ 28 Project Financing .................................................................................................................. 28 Experience in Evaluating Emissions Externalities ............................................................... 31 Including externalities in Operations and Investment Decisions .................................................. 36 Model for economic dispatch................................................................................................ 40 Model for capacity expansion ............................................................................................... 45 Summary Remarks ........................................................................................................................ 60 ii Chapter 5. Conclusion ............................................................................................................... 61 Annex A. Method to Value Externalities for Mexico’s Electricity Generation .................... 68 Annex B. Investment Plan in the Power Sector in Mexico ..................................................... 73 Annex C. Modeling Framework and Methodology ................................................................ 76 Economic Modeling .............................................................................................................. 76 Integrated Assessment Modeling .......................................................................................... 77 Methodology Steps ....................................................................................................................... 78 Characterization of the Emissions Source ............................................................................ 78 Dispersion Analysis .............................................................................................................. 79 Risk Assessment and Impact Quantification ........................................................................ 83 Valuation of Externalities ..................................................................................................... 86 Annex D. Project Valuation with Environmental Externalities ............................................ 89 Case Study 1: Combined Cycle Technology ........................................................................ 89 Case Study 2: Geothermal Installation ................................................................................. 92 Case Study 3: Wind Farm ..................................................................................................... 96 Case Study 3: Hydroelectric facility ................................................................................... 100 References .................................................................................................................................. 103 Figures Figure 1. Technologies as Share of Total Installed Effective Capacity, 2013 ......................... 6 Figure 2. Technologies as Share of Total Electricity Generation, 2011 .................................. 6 Figure 3. Energy Requirements for Power Generation. 2012–26 ........................................... 9 Figure 4. SENER Methodology: Data Requirements ............................................................. 12 Figure 5. Electricity Sector Planning ....................................................................................... 12 Figure 6. Economic Dispatch of Electricity ............................................................................. 13 Figure 7. Cost Benefit Analysis of New Generation Investment ........................................... 13 Figure 8. Application to Independent Power Producers ........................................................ 14 Figure 9. Planning Process To Internalize Environmental Externalities ............................. 16 Figure 10. Four-Step Process To Estimate Externalities ....................................................... 18 Figure 11. Regional Domain Coverage of the WTM .............................................................. 20 Figure 12. Overview of Project Evaluation Process and Financial Schemes ....................... 28 iii Figure 13. Project Evaluation Approval Process, Highlighting Inclusion of Externalities. 30 Figure 14. Benefits from the Case Studies ............................................................................... 33 Figure 15. Environmental Benefits from Four Case Studies (Thousand $/MW installed) . 34 Figure 16. Emissions reductions from Four Case Studies (ton / MW ................................... 35 Figure 17. CO2 avoided emissions (ton / MW installed) ......................................................... 35 Figure 18. Economic Dispatch with Externalities (change in generation, GWh) ............... 42 Figure 19. Screening Curves (no externalities) ....................................................................... 46 Figure 20. Screening Curves (with externalities) .................................................................... 46 Figure 21. Fuel prices ($/MMBTU) ......................................................................................... 51 Figure 22. Baseline scenario. Generation and CO2 Emissions............................................... 51 Figure 23. Baseline scenario. Criteria pollutants .................................................................... 52 Figure 24. Baseline scenario. Total capacity ........................................................................... 53 Figure 25. Policy scenario. Sensitivity to externality cost ...................................................... 54 Figure 26. Combination of Policies: Emissions Targets and Externalities ............................ 55 Figure 27. Combination of Policies: Renewable Portfolio Standard and Externalities ....... 56 Figure 28. Emissions ................................................................................................................... 56 Figure 29. Costs of Expansion Plans ($).................................................................................... 58 Figure 30. Exposure Response Function .................................................................................. 85 Maps Map 1. New Planned Capacity for 2014–19............................................................................... 9 Map 2. Coverage Areas of Mexico’s National Electricity System ......................................... 31 Tables Table 1. Investment Requirements by Type of Investment and Technology, 2012–19 ......... 8 Table 2. Summary of models used in Mexico to Evaluate Externalities Investment ........... 19 Table 3. Increased Risk Ratios ................................................................................................. 22 Table 4. Unit Costs of Morbidity Impacts ............................................................................... 22 Table 5. Externality Costs for Power Plants ........................................................................... 24 Table 6. Externality Cost for Two Critical Zones ................................................................... 24 Table 7. Social Cost of CO2 in the US regulatory studies, 2015–50 ....................................... 26 Table 8. Project Evaluation Guidelines for PIDIREGAS ...................................................... 29 iv Table 9. Investment Projects (Million $ of 2009) .................................................................... 32 Table 10. Costs of generation including externalities ............................................................. 38 Table 11. Emissions Factors ...................................................................................................... 39 Table 12. ...................................................... 43 Table 13. ........................................................ 44 Table 14. Costs parameters for investment planning modeling exercise .............................. 49 Table 15. Load Blocks................................................................................................................ 50 Table 16. Availability Factors for Renewables........................................................................ 50 Table 17. Renewables potential in Mexico ................................................................................ 50 Table 18. Total Capacity expansion (MW) ............................................................................... 57 Table B.1. Investment requirements by activity 2012-2026. ................................................... 73 Table B.2. Investment requirements by type of investment and technology 2012-2019, ..... 75 Table C.1. External Costs of Energy Impacts by Category and Pollutant ........................... 84 Table D.1. Basic Parameters of Combined Cycle Project ...................................................... 89 Table D.2 Displaced Generation from Combined Cycle Project ............................................ 90 Table D.3. Total Avoided Emissions from Combined Cycle Project .................................... 91 Table D.4 Summary of BCA for Combined Cycle Project ..................................................... 91 Table D.5. Detailed Cost-Benefit Analysis for Combined Cycle Project ............................... 91 Table D.6. Basic Parameters of Geothermal Project .............................................................. 92 Table D.7 Income flows for Geothermal Project ..................................................................... 93 Table D.8. Total Avoided Emissions from Geothermal Project ............................................ 94 Table D.9. Displaced Generation from Geothermal Project ................................................... 94 Table D.10. Summary of BCA for Geothermal Project .......................................................... 95 Table D.11. Detailed Cost-Benefit Analysis for Geothermal Project ..................................... 95 Table D.12. Basic Parameters of Wind Project ....................................................................... 96 Table D.13. Displaced Generation from Wind Project ........................................................... 98 Table D.14. Total Avoided Emissions from Wind Project ..................................................... 99 Table D.15. Detailed Cost-Benefit Analysis for Wind Project ................................................ 99 Table D.16. Basic Parameters of Wind Project ..................................................................... 100 Table D.17. Displaced Generation from Hydro Project ........................................................ 100 v Table D.18. Total Avoided Emissions from Wind Project ................................................... 101 vi Abbreviations and Acronyms AERMOD AMS (American Meteorological Society)/EPA Regulatory Model CEM Continuous Emissions Monitoring CFE Federal Electricity Commission (Comisión Federal de Electricidad) COA Annual Operating Record (Cédula de Operación Anual) COI Cost of Illness COPAR Reference Costs and Parameters for the Formulation of Investment Projects in the Electricity Sector (Costos y Parámetros de Referencia para la Formulación de Proyectos de Inversión en el Sector Eléctrico) CRE Energy Regulatory Commission (Comisión Reguladora de Energía) ECLAC Economic Commission for Latin America and the Caribbean (Comisión Económica para América Latina y el Caribe [CEPAL]) EPA United States Environmental Protection Agency ERF Exposure Response Function INEGI National Institute of Statistics and Geography (Instituto Nacional de Estadística y Geografía) IPP Independent Power Producer IRR Increased Risk Ratio ISC3 Industrial Source Complex, version 3 LAERFTE Law for Harnessing Renewable Energy and Financing the Energy Transition (Ley para el Aprovechamiento de Energías Renovables y el Financiamiento de la Transición Energética) LGCC General Law on Climate Change (Ley General de Cambio Climático) LSPEE Electricity Public Service Law (Ley del Servicio Público de Energía Eléctrica) PIDIREGAS Investment Projects with Deferred Expenditure Registration (Proyectos de Inversión Diferida en el Registro del Gasto) PL Productivity Loss PM Particulate Matter POISE Electricity Sector Infrastructure Investment Program (Programa de Obras e Inversiones del Sector Eléctrico) SCC Social Cost of Carbon SEMARNAT Ministry of Environment and Natural Resources (Secretaría de Medio Ambiente y Recursos Naturales) SENER Ministry of Energy (Secretaría de Energía) SHCP Ministry of Finance and Public Credit (Secretaría de Hacienda y Crédito Público) TSP Total Suspended Particulates VoE Value of Externalities VSL Value of Statistical Life WTM Windrose Trajectory Model WTP Willingness To Pay YOLL Years of Life Lost vii Units of Measure cm centimeter cm3 cubic centimeter g gram g per kWh grams per kilowatt hour g per s grams per second kcal kilocalorie km kilometer km2 square kilometer kWh kilowatt hour m meter mbar millibar mol mole m per s meters per second MT metric ton MW megawatt MWh megawatt hour PJ Petajoule ppb parts per billion tCO2e tons of carbon dioxide equivalent TJ Terajoule TWh Terawatt hour Currency Equivalents Currency Unit = Peso (Mex$) 13.98 Mex$ = US$1 (December 2011) viii Executive Summary Mexico’s policies to achieve low-carbon growth and green development are remarkable. The country has started a number of efforts to develop adequate policy frameworks in several areas including the energy sector, transportation and industrial policies, and forestry and natural resources management. Mexico’s Climate Change Law and the National Strategy on Climate Change envision changing the upward trend of Mexico’s CO2 emissions towards a total decline of emissions of 30% by 2020, and 50% to 2050 from its business as usual. Achieving these ambitious policy goals is, however, challenging for the country and will require many distinct efforts to mainstream climate change in policy design. While institutional efforts are remarkable, implementation, enforcement, and careful economic analysis will be critical to effectively reduce emissions while allowing for sustainable development. Energy and climate policy are of course at the core of Mexico’s effort. Mexico has enacted laws for energy conservation and to accelerate the energy transition toward clean energy sources. The policy evaluated in this study is part of Mexico’s policy framework to promote renewable energy. While renewable energy has higher capital costs than some of the fossil fuels alternatives, they come along with several environmental benefits. Among these benefits are the reduced health impacts of air pollution and climate change. However, many of these benefits do not have a dollar value, and are therefore most of the times not taken into account in the decision-making process. These “unvalued” benefits not taken into account in production and consumption decisions are called externalities. Valuing these impacts is important, as it can help leveling the field for renewables. Adding these external costs to fossil fuel alternatives in principle could be beneficial to renewable energy competitiveness. The Ministry of Energy in Mexico, SENER, has issued a Methodology to incorporate external costs of electricity generation. While external costs are not privative of fossil energy, Mexico’s government has started this process by focusing on the external costs on health and climate change, two of the main impacts of energy use. This study analyzes the potential of this new policy to help Mexico in its energy and environmental goals. In particular, this policy establishes a new legal mandate to incorporate environmental externalities in investment projects evaluation, the electric system operation and capacity expansion planning. First, we present an overview of power generation in Mexico. The country is transitioning away from fuel-oil generation and investing heavily in natural gas fired power. Currently, 46% of the generation is powered by natural gas, 9% by coal, 24% by conventional thermal units and the rest by hydro and a small share of nuclear and geothermal energy. Electricity demand is expected to grow at a 4% rate on average. To meet growing demand, and replace old power plants, Mexico’s investments plan for the coming two decades accrues to ix US$109.7 billion. This economic and sectoral context highlights the need for policies that will promote clean energy choices. If the replacement and new infrastructure to meet growing demand were carbon intensive, it would be difficult to achieve stringent mitigation targets such as the ones proposed by Mexico. Chapter 2 describes the legal framework, as well as the proposed institutional process for internalizing environmental externalities. The Ministry of Energy and the Federal Commission of Electricity are responsible of internalizing externalities of electricity in the short, medium and long-term decision-making process. The implementation of the policy requires key inputs from the Ministry of Environment and Natural Resources (SEMARNAT). In particular, SEMARNAT needs to provide estimates for the value of pollution in $/MWh. Chapter 3 describes the methods used in Mexico to assess external costs, as well as the data availability. Mexico has conducted 2 main studies to assess externalities of power generation. The first study used simplified methods and found an average value of 0.5 US¢/kWh. The second study used improved methods and estimated a 3.6 and 2.7 US¢/kWh value in critical zones. Building on Mexico’s experience and on previous work of the World Bank, we present in Annex B a summary of the methodologies used and key recommendations to improve the methods. Chapter 4 evaluates the policy and its potential to promote renewable energy. Using the available information in Mexico, we assess the impact of the new policy in a) the evaluation of investment projects, b) the economic dispatch of power plants and c) capacity expansion planning. First, we describe the procedure to evaluate investment projects in Mexico, the critical role that socioeconomic evaluation has both in deciding whether the project takes place and in approving its financing, and the new role of externalities in the evaluation process. CFE and SENER will be adding external costs in project evaluation and also including any financial instruments, emissions taxes or credits, into the financial evaluation of projects. Secondly, we look at four case studies representative of real cases under evaluation in Mexico, where we analyze investments in wind, hydro, geothermal and combined cycle gas turbines. We find that including the external costs was marginally beneficial for the hydro and combined cycle projects, but play a more critical role in the geothermal and wind project. In particular, the wind project presented a positive net present value only after adding externalities. To assess operational and long-term investment decisions, we develop two models to illustrate the potential changes of adding externalities. Our models are only for concept demonstration, and are calibrated to one of the electric regions of Mexico (Central-West region). We demonstrate that adding externalities in short-term operational decisions could change the electric system dispatch. There are two important findings from our modeling of economic dispatch: a) more generation is shift to non-critical areas (as classified by the Mexican environmental standard that sets emissions limits to combustion equipment) considering the fact that environmental health effects are geographically differentiated and higher in critical areas and x b) adding CO2 at a $23 usd/ton did not change the merit order of the technologies analyzed in this particular system, and therefore did not change further the dispatch. The results are particular to this system configuration, and only illustrate some of the potential results of the policy implementation. These finding suggests that two critical important components of the policy are to distinguish external costs by region and also to further differentiate technology vintages. The policy currently allocates the same costs to technology categories (same external costs per MW for all combined cycle units, all fuel-oil units, etc.). It would be important to further differentiate by units to have a more realistic pricing and effects in the economic dispatch of the system considering differences in the efficiency of the units. Regarding the modeling of the capacity expansion decision, we ascertain that, at the current level of external costs estimates in the Mexican studies, the technologies chosen for capacity expansion do not change. This result is limited and specific to the system studied, but do show that under some circumstances, the externality costs will not be enough to trigger technological change. The results of the capacity expansion decision highlight the importance of improving the methods to value externalities and to select the social cost of carbon to consider an adequate price that reflects total costs to society. Our study suggests that a combination of policies could be needed to further incentivize low-carbon generation, such as externalities with emissions targets or renewable portfolio standards. Again, these modeling results are just an illustration of possible dynamics and do not intent to quantify or prescribed particular policies. Our conclusions serve the purpose to inform the externality policy design and elicit important system interactions that should be consider for improving this policy. At the end the report presents some recommendations and areas of improvement of SENER’s methodology. Some of the recommendations include i) the definition of a work plan for improving externalities valuation and publish official Guidelines, ii) evaluate approaches to estimate the social cost of carbon and evaluate trade-offs, iii) improve the valuation of health outcomes, iv) estimating the value of external costs of generation associated with ozone, v) building capacity for model building and maintenance and vi) aligning environmental regulation and economic instruments with the application of environmental externalities. xi The World Bank has assist Mexico in identifying mitigation options, and in evaluating policies with high mitigation potential in critical sectors. Certainly, one of these key sectors to reduce emissions is the decarbonization of the electricity system. Delivering low-carbon power systems require the implementation of new policies that adequately balance costs and benefits of mitigation. Mexico has moved forward advancing its legislation and developing specific instruments to trigger technological change in the power sector. We consider this policy to be of interest to other developing countries looking for policy options to balance costs and benefits of renewable energy. xii Chapter 1. Introduction Electricity presents a paradox when it comes to meeting the challenge of climate change, representing the largest and fastest-growing source of carbon dioxide (CO2) while holding the key to potential solutions for a more efficient, less carbon-intensive economy. Energy experts and economists agree that many proposed solutions to mitigating the effects of climate change are concentrated in the power sector. These include greater use of renewable-energy sources, increased end-use efficiency (ranging from that of fossil-fuel plants to household appliances), and nuclear energy and carbon capture and storage. In addition, the electrification of transport and heating and cooling systems might provide a way to move other sectors toward a low-carbon future. How such decarbonization—generation without net CO2 emissions—will occur hinges on policy decisions. A stated objective of Mexico’s National Energy Strategy for 2013–27 is diversifying and optimizing the nation’s electricity capacity, with the goal of internalizing environmental externalities while maintaining energy security. The Energy Reform of 2008 includes two key pieces of legislation designed to ease the country’s transition to a less fossil fuel–intensive economy: one law promoting the use of renewable energy and the other promoting energy efficiency. The former explicitly mandates the incorporation of externalities for energy planning and environmental regulation and incentive mechanisms for renewable energy. To enforce this law, Mexico’s Ministry of Energy (SENER) published the Methodology for the Valuation of Externalities of Electricity Generation in December 2012. Hereafter refer to as SENER’s methodology (to distinguish it from other methodologies referred to in the text). Externalities are costs or benefits from an economic activity not reflected fully in market prices. For example, pollution creates damages that affect third parties but that are not considered in the market prices of economic activities nor compensated. Economists consider this problem a market failure, because too much of the polluting activity takes place with respect to an optimum where all costs and benefits are taken into account. Internalizing externalities in this context is the process by which producers or consumers are regulated or incentivize to consider these social costs in their economic decisions and adjust their production and consumption accordingly. In the case of electricity, different technologies impose different costs to society. For instance, fossil fuel power plants emissions cause damages in human health due to the adverse effects of air pollution and can have other potential impacts in other economic activities such as fisheries, crops, materials and tourism, due to acidification of soils and water bodies, as well as landscape impacts. Also, CO2 emissions cause climate change externalities. External costs are not restrictive of fossil fuel energy. Big hydroelectric facilities need to create dams, which by 1 disrupting the natural flow of rivers could have negative impacts in the local and regional ecosystems. Wind and solar farms require large extensions of land, which could have negative impacts on the landscape. Also, wind farms could impact negatively some bird species. Internalizing externalities of electricity is the process by which these costs or benefits are added to the electricity prices, either by reflecting them as costs for producers or consumers. One of the problems to include externalities into decision-making is that very often these negative effects are not quantified nor valued. In general, there are no markets for these emissions or taxes that reflect the social damages of pollution. Estimating these values requires studies to estimate the damage costs. Current state of research shows that the impacts on health and climate change are critical components of the external costs. The Mexican government has decided to start the process of internalizing externalities limiting the analysis to health and climate change impacts. However, Mexico does envision adding more impacts as more information on other external costs become available in the country. Thus, the present study is limited to internalizing costs of health and CO2 emissions. We consider that a follow up study on methods to value other external costs would be valuable as Mexico expands the scope of its policy, and acknowledge that our study is limited in scope to these two types of impacts. In Annex C (table C.1), we show the external costs of energy impacts by category and pollutant that are more often discussed in the literature as identified by the ExternE study. Worldwide, various policy instruments have been designed to incentivize clean energy production; however, the mandate to incorporate environmental externalities into the planning and operation of a full power system is unique. Box 1.1 International Experience Internalizing Externalities in Electricity Generation The concept of internalizing externalities has always been at the core of environmental policy. Hence, many countries have developed approaches to value pollution and to consider the social and environmental costs of energy choices in policy design. While the idea of pricing pollution is generally accepted as a sound policy principle, placing a concrete dollar value on the damage of pollution is challenging. In particular, the methods to estimate non-market values of air pollution and climate change emissions from the power sector are data-intensive and require the use of models to track pollution, assess the damages and methods for economic valuation. Particularly for climate change, valuation techniques and models need to be improved to better assess the social cost of carbon. For this reason, most of the previous work on externalities valuation has been conducted primarily in developed economies. Both the European Union and the United States have used life cycle analysis to estimate externalities from electricity technologies. Extensive research has been conducted on the valuation side since the 1990s. In Europe, the ExternE project has estimated electricity externalities in different European countries (European Commission. 2005). In the US, the National Research Council updated the values of externalities from electricity generation and 2 examined them in the context of the so- called hidden costs of energy (National Research Council, 2010). Among the uses of damage costs are the assessment of new environmental standards and policies, project valuation and justification of green tax policies (EPA,1999; EIA- DOE, 1995; OECD, 2001a&b; Égert, B. ,2011). Along with the increasing attention to the climate problem, the interest on assessing environmental externalities of conventional pollution has spurred research in developing countries. Acknowledging the importance of co-benefits of climate policy-for example the reduce impact of conventional pollution- has become an area of opportunity to move forward climate action in the developing world. Because Mexico’s power-system structure allows for highly centralized planning, the country could set a precedent for how large state-owned enterprises can move forward with sound methods to incorporate environmental costs into their investment decisions and send a market signal to their private-sector partners, the independent power producers (IPPs), about the value of clean technologies. Many developing countries rely on IPPs to deliver their electricity, with various private-participation schemes. Such innovative legislation in a large developing country like Mexico could provide a real-world example of the benefits and challenges of such a policy. The policy we assess in this study, put forward by the Mexican Government, falls into the national mitigation policies that try to incorporate local, regional and global benefits in the decision-making process. Thus, while the concept is not new, its application in the power sector in Mexico and the proposed policy framework can provide a useful example for developing countries looking at the implementation of sustainability policies for the electric sector. Study Purpose and Objectives This study aims to evaluate the impact of the new methodology developed and used by SENER to internalize environmental externalities. Specific objectives are to identify (i) barriers to implementing the methodology and (ii) ways to improve it. The study builds on previous work on environmental externalities undertaken by SENER, as well as other ministries (e.g., Ministry of Finance and Public Credit [SHCP] and Ministry of Environment and Natural Resources [SEMARNAT]). It complements previous work undertaken by the World Bank’s Energy Sector Management Assistance Program (ESMAP), which has assisted Mexico in its efforts to design polices for low-carbon growth (e.g., World Bank 2010a; ESMAP 2010). SENER and CFE will apply this method to internalize externalities in their decisions regarding the investment and operation of the power system in Mexico. Structure of this Report This report is organized as follows. Chapter 2 describes the policy context and legal framework, 3 as well as the proposed institutional process for internalizing environmental externalities. Chapter 3 describes the models and methods used to assess external costs, as well as the data used in the estimates. Chapter 4 uses the method to evaluate policy implications, including incorporating the external costs of local and regional pollution and the social costs of carbon as a global benefit. Finally, chapter 5 offers conclusions and recommendations. 4 Chapter 2. Mexico’s Policy Context for Incorporating Externalities Power Sector Overview Mexico’s primary energy production depends heavily on hydrocarbons. According to the Ministry of Energy (SENER), total energy production in 2011 was 9,185 PJ, of which hydrocarbons accounted for 89 percent (SENER 2012a). Of that amount, oil and natural gas represented 73 percent and 26 percent, respectively. Renewable energy comprised 7 percent of primary energy production, of which biomass supplied 55 percent and hydropower and geothermal each accounting for 20 percent. Coal comprised 3.2 percent of total production, while nuclear energy accounted for 1.2 percent. A large oil exporting country, Mexico exported 3,137 PJ of its energy production in 2011, mainly in the form of crude oil. Energy imports, which are growing, accounted for 2,269 PJ, of which gasoline and natural gas represented 36 percent and 30 percent, respectively, followed by diesel and gasoil at 13 percent and coal at 8 percent. After accounting for losses and net trade, Mexico’s total primary energy supply was 8,399 PJ, with 1,049.6 PJ of that amount used for electricity generation. As of January 2013, Mexico’s electricity sector had a total installed effective capacity1 of 52,974.2 MW (SENER 2013), 76 percent of which was owned by the Federal Electricity Commission (CFE), the state-run electric utility, and the remainder by independent power producers (IPPs). In 2011, the country’s electricity system generated 2.59 TWh, 59.6 percent of which was generated by CFE and the remainder by the private sector. Currently, natural gas technologies (primarily combined cycle) account for 34 percent of total installed effective capacity, while conventional thermal and hydroelectric plants comprise 28 percent and 22 percent, respectively, followed by coal and dual coal/fuel oil at 5 percent each, nuclear at 3 percent, geothermal at 2 percent, and wind at 1 percent (Figure 1). In 2011, natural gas accounted for 46 percent of electricity generation, followed by conventional thermal (mainly fuel-oil powered) at 18 percent, hydro at 14 percent, and coal at 9 percent; nuclear and geothermal accounted for 4 percent and 3 percent, respectively (Figure 2). Altogether, non-fossil fuel technologies produced 21 percent of electricity in 2011. 1 Effective capacity refers to the total installed capacity minus programmed outages, such as those plants programmed for maintenance, unavailable due to the hydrologic cycle in the year, etc. 5 Figure 1. Technologies as Share of Total Installed Capacity, 2013 Geothermal Wind Nuclear 2% 1% Coal 3% 5% Dual coal/ Fuel oil 5% Natural gas 34% Hydro 22% Conventional thermal (fuel oil) 28% Total capacity: 52,974 Source: Authors, based on data from SENER 2013. Figure 2. Technologies as Share of Total Electricity Generation, 2011 Geothermal Wind Nuclear 3% 0.00004 4% Coal 9% Dual coal/ Fuel oil 4% Natural gas Hydro 47% 14% Conventional thermal (fuel oil) 19% Total generation: 2.59 TWh Source: Authors, based on data from SENER 2012b. 6 Investment Outlook CFE and SENER are in charge of planning future expansions of the electricity sector, considering both demand growth estimates and fuels and technology requirements. By law, SENER publishes a 15-year plan (perspective) for the power sector each year, outlining the government’s programs in electricity infrastructure investment. According to this plan, demand for electricity over the next decade will increase by 4 percent a year, reaching 480 TWh in generation by 2026. Meeting this demand will require 44,532 MW in expanded system capacity, of which 10,795 MW would come from projects not yet committed to any particular generator (private or public) or technology. For the 2012–26 period, the investment plan totals about US$109.7 billion2. Of this amount, 52 percent is for new generation facilities, 20 percent for the distribution network, 14 percent for transmission, 13 percent for maintenance, and 1 percent for other public investment (CFE 2012). Mexico has three main modes of capacity investment: public works, financed public works, and IPPs. Public works are paid directly from the government out of public budget, financed public works are included in the budget and approved by Congress, but will be financed and paid by services provided by the resulting infrastructure, and the IPPs are financed by the private sector. Of the total investment needs for 2012–19, 70 percent will come from the scheme for financed public works, 25 percent from IPPs, and 5 percent from the federal budget for public works (Table 1). Public works will account for 67 percent of new investment in transmission and 88 percent of investment in distribution. The government expects that 74 percent of IPP investments will be in natural gas combined cycle (CC) plants and the rest in wind farms. Investment in financed public works will be allocated to natural gas CC plants (42 percent), new hydro plants (23 percent), new wind farms (19 percent), and 16 percent (other technologies). Two-thirds of budgetary expenses will be allocated to maintenance and one-third to new hydro. (See Table 1 below, and Annex B for further details on investment requirements). The National Electricity System is the country’s largest fuel consumer and thus its main point source of emissions from polluting agents. It accounts for 80 percent of coal consumption, 77 percent of the heavy fuel oil (combustóleo) used (with approximately 4 percent sulfur content), and 35 percent of the natural gas consumed. The resulting air pollution from the electricity sector consists of 1.05 million tons of sulfur dioxide (SO2), 337,810 tons of nitrogen oxides (NOx), 68,950 tons of particular matter (PM10), 58,271 tons of fine particulate matter (PM2.5), 72,402 tons of carbon monoxide, 7691 tons of volatile organic compounds (VOC), 3,671 tons of ammonia (NH3) and 3,295 tons of black carbon and 84.5 million tons of carbon dioxide equivalent (CO2e) (SEMARNAT, 2013). The growth in emissions will depend on the selected technologies for system capacity expansion. 2 CFE figure is 1,533,359 million pesos of 2011. We convert to dollars using the official exchange rate in Mexico of 13.97 pesos per dollar of December 31, 2011. 7 Table 1. Investment Requirements by Type of Investment and Technology, 2012–19 (Millions of US$ (2011) Investment/technology type 2012 2013 2014 2015 2016 2017 2018 2019 Total Generation 3,457 3,099 3,426 2,861 2,583 3,315 2,614 2,658 24,013 Independent power producers 1,811 1,485 1,838 504 51 225 59 - 5,972 New combined cycle 712 1,128 1,753 482 51 225 59 - 4,410 New wind farms 1,099 356 85 21 - - - - 1,561 Financed public works 1,242 1,354 1,362 2,224 2,474 3,041 2,505 2,620 16,821 New hydro 115 49 501 671 757 684 504 558 3,838 New geothermal 8 101 123 7 86 75 30 65 496 New combined cycle 622 603 535 902 1,525 997 457 1,456 7,097 New wind farms 147 114 21 486 - 1,285 799 320 3,170 Maintenance and repower 150 237 182 157 31 1 - - 759 New clean generation 20 50 - - - - - - 70 New turbogas 179 199 - - 75 - - 221 675 Public works 404 260 226 133 58 50 50 38 1,220 Hydroelectricity 82 45 52 45 57 49 50 38 419 Maintenance 322 215 175 88 1 0 - - 801 Transmission 844 1,012 970 1,215 977 948 880 875 7,721 Financed public works 295 440 197 114 464 417 338 305 2,570 Transmission program 295 440 197 114 464 417 338 305 2,570 Public works 549 572 773 1,101 513 531 542 570 5,151 Transmission program 89 96 281 649 155 139 113 102 1,624 Transmission S T Y T 346 357 367 329 247 271 297 326 2,542 Modernizing CENACE 44 48 51 56 62 67 72 77 477 Modernizing central area 69 71 73 66 50 54 59 65 508 Distribution 1,885 1,969 1,590 1,415 1,455 1,523 1,535 1,520 12,892 Financed public works 410 388 114 11 182 228 224 233 1,789 Sub-transmission program 410 388 114 11 182 228 224 233 1,789 Public works 1,475 1,581 1,476 1,404 1,273 1,295 1,311 1,288 11,102 Sub-transmission program 242 371 272 199 61 76 75 78 1,373 Distribution program 440 431 433 440 444 454 476 451 3,569 Distribution program, central 354 332 329 320 317 319 315 307 2,594 Modernizing distribution 439 447 442 444 452 445 446 452 3,566 Maintenance 792 565 723 770 819 846 894 930 6,339 Other public works 33 34 35 36 37 38 39 41 294 Total 7,010 6,679 6,744 6,297 5,871 6,669 5,963 6,025 51,258 Source: CFE, Electricity Sector Infrastructure and Investment Program (POISE) 2012. Note: The exchange rate used was US$1 = Mex$13.9787 (closing rate, December 30, 2011) (www.banxico.org). Map 1. New Planned Capacity for 2014–19 MW Wind 2,108 Hydro 1,482 CC gas 9,397 IC 43 Turbogas 49 Geothermal 179 Source: SENER 2013. Figure 3. Energy Requirements for Power Generation. 2012–26 TJ per day Source: SENER 2012b. Expected electricity demand growth will require Mexico’s energy sector to expand natural gas use by 3.7 percent per year; pipeline distribution will originate from domestic and U.S. sources, as well as liquefied natural gas (LNG) import terminals. The country has limited coal reserves, located mainly in the north; thus, coal will be imported to meet power-plant demand in southern regions (Map 1). According to planning documents, the fraction of clean gas and clean coal shown in Figure 3 refers to gas and coal generation with carbon capture and storage. Clean generation could also include renewable sources, nuclear, and other technologies. 9 Policy Mandate to Internalize External Costs and the New Methodology Mexico’s National Energy Strategy to transition from heavy dependence on hydrocarbon-based electricity generation adheres to both environmental and energy-security objectives. Three statutes provide the legal basis for applying this methodology: the Law for Harnessing Renewable Energy and Financing the Energy Transition (LAERFTE), the Electricity Public Service Law (LSPEE), and the General Law on Climate Change (LGCC). The LAERFTE mandates the following: The Ministry of Energy, with the opinion of the Ministry of Treasury and Public Credit, the Ministry of Environment and Natural Resources, and the Ministry of Health, shall elaborate a methodology to value externalities related to electricity generation, coming from renewable and non-renewable sources at different scales, as well as the public policies to which this Law refers related to these externalities. Based on that methodology and public policies, the Ministry of Environment and Natural Resources will design the mechanisms for environmental regulation to harness renewable energy (LAERFTE, Article 10). In addition, the LSPEE mandates that For the public service of electricity, in the short and long term, electricity shall be produced from least-cost sources for the Federal Electricity Commission, considering the environmental externalities of each technology, and also providing optimal stability, quality, and security of the public service (LSPEE, Article 36bis). Furthermore, the LGCC states that The national mitigation policy for climate change should be instrumented gradually, promoting and enhancing national capabilities for mitigation and adaptation, as well as the adverse effects of climate change, and prioritizing the sectors with greater abatement potential, leaving at the end those of higher costs, in addition to taking care of the international agreements taken by Mexico. In the case of policies and activities that imply or transfer a cost to the private sector or society in general, and for which no international funding is available for implementation, these policies could be implemented in two phases, when there is an area of opportunity for the regulated entities: I. Enhancing national capabilities where policies shall be implemented on a voluntary basis, with the objective of strengthening capabilities of the regulated sectors, considering: …g. Analysis of the electricity generation sector, including external social and environmental costs, as well as the costs of emissions in the selection of sources for generating electricity (LGCC, Article 32). In order to reduce emissions, the Agencies of the Federal Government, the States, and the Municipalities, within the scope of their competences, shall promote policies and mitigation actions in the corresponding sectors, considering: …d. Including the costs of the environmental and social externalities, as well as the costs of emissions, in the selection of sources for electricity generation (LGCC, Article 34). 10 On December 14, 2012, Mexico published the Methodology for the Valuation of Externalities of Electricity Generation in Mexico, which substitutes for SENER’s earlier efforts to initiate accounting for electricity externalities using CO2 costs in new investment projects. The new methodology includes the impact of local and regional pollution, including SO2, NOx, and PM10. It also extends the requirement of accounting for externalities beyond the planning process to encompass system dispatch and evaluation of new investment projects. This allows policy makers to consider all available options, taking into account the cost of pollutants produced per kilowatt-hour generated. The method’s stated objective is “valuing the externalities associated with electricity generation by using different energy sources at different scales in the system” (section 1.3). In accordance with the legal framework, the method “will allow for evaluating the environmental impact of the development plans of electricity generation projects, which will impact the expansion plans of the system and the economic dispatch in the short run.” In the document’s “Scope and Objective” section, SENER establishes that only externalities that differ significantly from zero will be considered, emphasizing the need to include the external costs imposed on health. The document’s “Definitions” section includes terminology to operationalize the process. Externalities are defined as “positive or negative impacts associated with electricity generation, caused by the provision of a good or service, that impact or could impact a third party in Mexico locally, regionally, or globally over the short, medium, and long term; these externalities occur when the cost paid for a good or service differs from the total costs of damages or benefits in economic terms.” Other key terms and definitions include displaced energy, defined as energy substituted by the new project, incremental energy, meaning additional energy associated with the new project, and financial instruments, which refer to carbon credits and/or carbon taxes. The method describes the process that the Federal Government of Mexico is to follow for incorporating externalities. The four government entities with competencies in the process are SENER, SEMARNAT, CFE, and SHCP. SENER establishes the methodology, coordinates with the entities, and includes IPP emissions in the evaluation of externalities. SEMARNAT provides information on the external costs associated with the various pollutants per technology, conducts a financial analysis of the CO2 value in regional and international markets to provide the dollar- per-ton value to use in the valuation process, and facilitates use of annual emissions reports from the CFE and IPPs. CFE elaborates the reference costs and parameters for formulating investment projects (COPAR), including information on the cost of externalities (new procedure), and uses these to develop the electricity sector’s 15-year investment plan, known as the Electricity Sector Infrastructure Investment Program (POISE). CFE competency is to include externalities in system dispatch by adding external costs to the computation of total variable costs. Finally, externalities have to be included in the elaboration of investment projects that CFE submits, with SENER approval, for SHCP evaluation. Socioeconomic evaluation of the projects is to include all pollution costs, while financial evaluation is to include CO2 revenues from regional or international markets. SENER’s schematic representation, including the 11 responsible entities throughout the process, is highlighted below (Figures 4–8). Annex A provides a full translation of the method for those interested in the technical and legal details. Figure 4. SENER Methodology: Data Requirements SEMARNAT • Provide Value of Externalities (VoE) in pesos (Mex$) per ton pollutant for the various technologies (conduct studies and update them every three years). • Provide analysis of financial instruments regarding value of CO2 based on regional or international markets (updated each year). • Facilitate use of annual operating records (COAs) by SENER and CFE to integrate the electricity sector's emissions inventory. CFE • Prepare the electricity sector's emissions inventory. • Include in the COPAR emssions per megawatt hour for each technology and the electricity sector's emissions inventory. SENER • Obtain IPP emissions from SEMARNAT and provide CFE information for the emissions inventory. Source: Authors, with data from SENER’s Methodology (SENER, 2012c). Figure 5. Electricity Sector Planning SEMARNAT SENER CFE • Send VoE 80 days after • Send CFE • Include $ per MWh per publication of - VoE, pollutant for each methodology. - Financial cost, and technology in the • Send analysis of - IPP emissions. POISE, using emissions financial cost of CO2 to inventory and costs SENER (updated (COPAR) for $ per ton annually). of pollutant provided by SEMARNAT. • Compute “not compensate” externality. • Include cost of pollution as a variable cost in the planning process. Source: Authors, with data from SENER’s Methodology (SENER, 2012c). 12 Figure 6. Economic Dispatch of Electricity SEMARNAT SENER CFE • Send VoE 80 days • Send CFE • Use VoE in after publication - VoE, computing total of methodology. - Financial cost, variable costs of • Send analysis of and the various financial cost of - IPP emissions. technologies. CO2 to SENER • Exclude any (updated externality cost annually). unrelated to unit operations. • Use total variable costs that include the cost of electricity in the economic dispatch of the system. Source: Authors, with data from SENER’s Methodology (SENER, 2012c). Figure 7. Cost Benefit Analysis of New Generation Investment CFE SENER SHCP • Include externalities in • Review projects. • Conduct the socioeconomic • Include in sectoral socioeconomic evaluation of budget. evaluation of the investment projects. • Present to SHCP. project presented by • Estimate externalities CFE/SENER. with and without the • Conduct financial project, considering the evaluation of the difference in energy project presented by displaced and CFE/SENER. incremental energy. • In the socioeconomic • Follow additional evaluation, value principles. impacts per ton of • Include all pollutant using data environmental provided by SENER and externalities in SEMARNAT. socioeconomic • In the financial evaluation of projects. evaluation, inlcude CO2 • Include revenue of CO2 financial income. markets in financial evaluation of projects. Source: Authors, with data from SENER’s Methodology (SENER, 2012c). 13 Figure 8. Application to Independent Power Producers SENER CFE • Request information on • Include IPP emissions in the electricity sector's operational emissions of emissions inventory. IPP projects from SEMARNAT each year. • Send emissions information to CFE. Source: Authors, with data from SENER’s Methodology (SENER, 2012c). Critical Steps and Potential Barriers to Implementation As the above description of the methodology implies, critical steps, outlined below, are required for its implementation:  SEMARNAT must provide updated values for the costs of externalities from the various technologies for electricity generation. The method requires updated analysis every three years. The values estimated will eventually determine the overall impact of the policy.  CFE must prepare the electricity sector’s emissions inventory with IPP information. This will require extensive collaboration with SEMARNAT and SENER to access environmental performance information from the annual operating records (COAs) of the various facilities and IPPs.  CFE must include these costs in the investment projects submitted to SENER and SHCP for approval.  SHCP and SENER are required to evaluate investment projects in a timely manner, and CFE’s overall planning process must be aligned with their evaluation.  Policy timelines are short; the method requires that external costs be included in the 2014 planning process, which leaves little time for coordination. To identify potential barriers to implementing these critical steps, this study conducted interviews with officials from SENER, SEMARNAT, CFE, and SHCP. During the interviews, officials expressed their concerns about the models and methods required to estimate the value of pollution, inter-institutional coordination, specifics on the legal wording of the method, and budgetary constraints to conducting the needed studies. Some stakeholders expressed their concern about the potential overall impact of the policy on electricity tariffs. 14 Clearly, the first critical step in moving forward with policy implementation is ensuring that studies to estimate the valuation of pollution in the country are based on sound methods. SEMARNAT has made substantial progress in this area. Two available studies, currently in the process of being updated, include VoE estimates in dollars per ton of SO2, NOx, and PM10 for pollution produced by selected power plants in the country, including critical areas. CFE, SENER, and SEMARNAT have various databases available, based on their work in analyzing emissions of the electric power sector. Also, detailed power-plant data from a tri-national collaboration project of the North American Free Trade Agreement (NAFTA) Commission for Environmental Cooperation is available and open to the public. However, no single official dataset is available explicitly for evaluating externalities. Setting up a standardized procedure is critical for estimating the sector’s emissions inventory and providing access to the stakeholders involved. Such a method requires ongoing annual updating, which could present a bottleneck if detailed COA information is required. Given the need for annual updating of the emissions inventory, simplified estimation methods, such as emissions factors, might be preferable to COA information, which entails a lengthy process. Evaluating IPP emissions using COA report data could also present a bottleneck since explicit authorization may be required and data use limits could apply. Since 2006, CFE, SENER and SHCP have been using the externalities studies of SEMARNAT to conduct cost-benefit analyses of socioeconomic projects, particularly to evaluate the potential benefits of renewable energy and more recently the displacement of older power plants with more efficient fossil-fuel generation. This means that substantial expertise has been developed in the planning and investment units of these entities. However, implementation of this policy would expand the use of externalities in project evaluation, and some discussion would be needed to address issues of adding the external costs to new fossil-fuel generation projects, particularly the initial ones evaluated. Incorporating externalities into the COPAR and POISE are new procedures for CFE, and time would be required for implementation. However, previous experience in using externalities for project evaluation could facilitate the process. If CFE is provided data on the VoE, then the additional costs, like all other variable costs, could be included in the process. SENER and SEMARNAT would have to provide CFE the VoE numbers in a timely manner, and internal coordination of CFE areas would be required to ensure timely delivery of the emissions inventory to the planning area. Ideally, this should be a standardized internal process, conducted annually, to better equip CFE areas to comply with their new responsibilities (Figure 9). 15 Figure 9. Planning Process To Internalize Environmental Externalities Source: CFE 2005 modified by Authors to include new steps for externalities. Note: Boxes shaded in blue represent the current planning process, while green ones indicate the added processes from the new mandate to incorporate environmental externalities. The method aims to have a long-term planning impact by incorporating externalities into COPAR, while the respective medium- and short-term impacts are expected through analysis of concrete infrastructure programs and its addition to the economic dispatch of the system. Given the importance of VoE to this process, the next chapter presents the information required for these studies and the current data available in Mexico to identify potential bottlenecks, as well as areas for improving VoE computation methods. 16 Chapter 3. Valuation of Externalities Energy production and consumption are essential for economic development and therefore the goal of energy policies is to supply fuels and electricity to meet growing demand while minimizing costs. However, these policy objectives face constraints related to the impact of energy-polluting emissions on human health and the absorptive capacity of the environment more generally. Energy production and consumption cause local, regional, and global externalities, mainly associated with the use of fossil fuels. At local and regional levels, a main concern is the health-risk exposure of vulnerable populations to criteria pollutants (those that are toxic and/or that can have a negative impact in human or ecosystems’ health), which can result in acute and chronic respiratory and cardiovascular diseases and related illnesses. At the regional level, a major environmental concern is the effect of sulfates and nitrates from acid rain on ecosystems and agricultural crops, as well as material damage. On a global scale, the foremost challenge is greenhouse gas (GHG) emissions. Evaluating the new methodology for Mexico’s electricity production matrix requires estimating the value of such externalities. Building on earlier studies of the Mexican government (SEMARNAT and ECLAC 2004 & 2007) and previous World Bank work to assist the Mexican government in identifying methods to evaluate externalities (World Bank, 2010b), this chapter describes the methods and data that could serve as the basis for policy implementation. Methodology Mexico’s studies on external costs of energy have followed the methodology Impacts Pathway, developed by the European Union for the research project Externalities of Energy (European Commission. 2005). To quantify and monetize external costs, this methodology requires four consecutive steps that follow the route of pollutants from the emissions starting point to the endpoint where damages are caused (Figure 10). E The first step is characterization of the emissions source, including a detailed analysis of the emissions inventory and facilities (e.g., release of pollutants, stack heights, and temperature and velocity of flows). The second step is dispersion analysis, which estimates the pollution concentration in ambient air and deposition in ecosystems. The third step involves risk assessment and quantification of impacts, applying exposure-response functions to estimate units of damage (e.g., number of cases of bronchitis or restricted activity days). Finally, the fourth step is valuation of externalities, which considers the valuation of market and non-market goods; 17 economic valuation techniques are applied to non-market goods to estimate society’s willingness to pay (WTP) in order to avoid external costs (e.g., hedonic wages and contingent valuation).3 Figure 10. Four-Step Process To Estimate Externalities 1. Characterization of 2. Dispersion 3. Risk Assessment and 4. Valuation of the Emissions Source Analysis Impact Quantification Externalities Dose Impact - Urban or rural - Meteorological conditions - Exposure-response functions -Valuation techniques - Chimneys - Models -Damage cost - Emissions parameters Source: CEPAL-SEMARNAT 2004. The methodology relies on modeling techniques to quantify the environmental damage, and data from emissions, epidemiology studies and economic costs. Annex B Modeling Framework and Methodology provides detailed descriptions of the underlining economic modeling framework and of each step in the valuation process. Each of the methodology steps mentioned in Figure 10 is described fully in the Annex, citing relevant literature in each of the methods. It also describes the specific models used in Mexico, identifying the main developments and areas of opportunity to improve the methods used. Table 2 provides a summary of the models used in Mexico, as well as a summary of main data requirements. While the description of the steps involved in analyzing externalities is very important, plenty of literature exists regarding these methods, therefore interested readers should look for details in the Annex and on the literature cited on it. Following we focus rather on describing the available data in Mexico, as well as the externalities estimates of the Mexican government. 3 The methodology update for the ExternE project provides a description of the full methodology, uncertainties, and new developments (European Commission 2005). 18 Table 2. Summary of models used in Mexico to Evaluate Externalities Investment Model Developer Description Application in Mexico SIMPACTS International Atomic The Simplified Approach for Estimating Impacts of Electricity Generation (SIMPACTS) is a model This model was used by SEMARNAT and ECLAC to Energy Agency (IAEA) developed by the IAEA that provides simplified methods to evaluate external costs. evaluate 11 big utilities in Mexico, comprising the majority of emissions from fossil fuels. Designed for the use of developing countries with limited data availability and low capacity to run more sophisticated models, the SIMPACTS model provides a first approximation to externalities. In The results of this study were used in the cost- one hand the data requirements are less intensive, facilitating its use. On the other hand, the benefit analysis of new emissions standards for underlining dispersion and valuation methods are limited compared to state-of-the art models used in power plants. Also, the results are used for the developed countries. evaluation of investment projects in the power sector (SEMARNAT and ECLAC, 2004) SEIA UN Economic The System for the Evaluation of the Environmental Impact of Energy Installations model was This model was applied by SEMARNAT and ECLAC to Commission for Latin developed by ECLAC, following the methodology of the European Commission’s model EcoSense. It is two installations in critical regions in the country. America and the calibrated to Mexico and the Caribbean region. (SEMARNAT and ECLAC, 2007) Caribbean, ECLAC Compared to SIMPACTS this model has a stronger representation of the dispersion model, particularly The results of this study are currently used for the in the regional domain. Also, it is more flexible to evaluate pollution dispersion of the local domain, as evaluation of investment projects in the power it allows the modeler to use outputs from more complex dispersion models in the analysis sector ISC-LT USEPA Industrial Source Complex-Long Term. This is one of the underlining models for pollution dispersion 2 power plants were evaluating using this model, analysis in SIMPACTS. Evaluates pollution dispersion in the local domain using annual statistics of resulting in the lowest costs of externalities meteorological data. (SEMARNAT and ECLAC, 2004) ISC-ST USEPA Industrial Source Complex- Short Term. This is one of the underlining models for pollution dispersion This model was used in 9 power plants in Mexico analysis in SIMPACTS. It uses hourly meteorological data. (SEMARNAT and ECLAC, 2004) AERMOD USEPA AMS/EPA Regulatory Model is the model currently use in the US for the evaluation of pollution This model was used to analyze two power plants in dispersion of industrial sources in the local domain Mexico (SEMARNAT and ECLAC, 2007) WTM European Commission- The Windrose Trajectory Model was developed by the EU and was the underling pollution dispersion This model was used to analyze two power plants in ExternE project model in the EcoSense model used in Europe to evaluate externalities. It uses detailed regional data, Mexico (SEMARNAT and ECLAC, 2007) however it simplifies pollution dispersion by considering only one mixing layer in the atmosphere. CALPUFF TRC This model is used for assessing long-range transport of pollutants. It is the regulatory model in the US This model was used by INE to evaluate one utility to analyze pollution transport in regional domains in Mexico (INE, 2003 and 2006) Source: Authors. 19 Review of Data Available in Mexico The data available in Mexico for valuing externalities was first compiled in the above-mentioned studies by SEMARNAT and the Economic Commission for Latin America and the Caribbean (ECLAC). In order to internalize external costs, SEMARNAT is in the process of updating these studies. This section presents the externalities figures currently available in Mexico and presents updated data for some of the critical parameters of the valuation. Meteorological Data and Regional Dispersion Domain The National Meteorological Service provides local meteorological data required for numerous variables, ranging from mean temperature, wind direction and velocity, precipitation, and solar radiation to mixture height, cloud cover, and land use. This information is available and is updated hourly for the network of 184 stations in the country. CFE also gathers meteorological data in many of its power plants. The data for the National Meteorological Service and from CFE can be used for local-dispersion modeling. For the analysis of regional dispersion, information is required for a large geographic area. This data normally comes from international datasets of re-analysis data4. ECLAC calibrated the Windrose Trajectory Model (WTM) for an area covering Mexico, Central America, the Caribbean islands, and Florida (Figure 11) using reanalysis data of 5 years . Figure 11. Regional Domain Coverage of the Windrose Trajectory Model Source: ECLAC. has data on precipitation and wind patterns, emissions, and population for the entire grid at a 55 km2 resolution. Overall, there are some 400,000 pieces of data obtained from processing about 158 4 Re-analysis datasets are constructed with information coming from monitoring stations around the globe, and with the use of models and supercomputers to reproduce large-scale phenomenon affecting meteorological variables. 20 million pieces of data from various information sources. The ECLAC regional office makes the dataset available to all countries in the region for research purposes. The dataset is described in the subsections below. Wind 925 mbar (approximately 800 m above sea level). Precipitation Available data on precipitation in the WTM comes from the Global Precipitation Climatology Centre (GPCC) and the Global Precipitation Climatology Project (GPCP), Version 2 Combined Precipitation Data Set. The GPCC source is a free-access dataset with quality-controlled data for some 7,000 monitoring stations; it has global data from 1986–present, at resolutions of 2.5 and 1. The GPCP dataset contains remote sensing data in addition to measurements, and has a 2.5 resolution. Emissions The WTM requires that each cell in the domain contain emissions data on SO2, NOx, and NH3 in order to estimate the background concentrations needed to run the program’s chemical transformation modules. SEMARNAT and ECLAC updated data for Mexico’s energy sector emissions as part of the study (SEMARNAT and ECLAC 2007). Energy-sector emissions, from CFE and Pemex (Petróleos Mexicanos), were updated for the 2005 reference year. Emissions from the rest of the grid vary. ECLAC’s database has used various emissions datasets. For SO2 and NOx emissions, the Emissions Database for Global Atmospheric Research (EDGAR) was used,i while data on NH3 emissions used the Ammonia International Emission Inventoryii and the Global Emissions Inventory Activity (GEIA).iii Datasets on NH3 emissions, which are quite difficult to capture at a global scale, used 1990 and 1995 data, with respective resolutions of 0.5 and 1 . The NH3 background level is critical to estimating nitrates and sulfate reactions. The WTM requires background concentrations of ozone (O3) and hydroxyl radicals (OH); for the region, the ECLAC team recommended the following default values: 40 ppb (O3) and 1.3 million mol per cm3 (OH). Exposure Response Functions (ERF) The available data for ERF was compiled originally by the World Bank in a meta-analysis of epidemiological studies in Mexico, in the context of health impacts valuation for Mexico City. McKinley et al. (2003) and SEMARNAT and ECLAC (2004, 2007) have use the values for increased risk of exposure coming from the WB meta-analysis updating information for baseline health statistics. Table 3 presents the IRRs and baseline information, with original sources. 21 Table 3. Increased Risk Ratios 3 Percent per μg per m Risk Cases Population Endpoint group PM10 per year share Bronchitis Adults 0.360 0.00700 0.41 Restricted activity days Adults 0.774 6.46000 0.54 Hospital admissions, respiratory All 0.139 0.00260 1 Hospital admissions, cardiovascular > 65 0.060 0.00210 .005 Emergency visits All 0.311 .03200 1 Asthma attacks Asthmatics 0.774 .12440 0.056 Chronic mortality > 30 0.384 .09600 0.41 Sources: CEPAL-SEMARNAT 2004 Economic Costs Mexico’s National Institute of Public Health has conducted research to value the impacts of smoking in respiratory diseases, which may serve as a proxy for estimating air pollution damages. The main reference for morbidity impacts is a detailed study conducted jointly by the National Institute for Public Health and the National Institute of Ecology, with technical cooperation from the USEPA, to value the health impacts as co-benefits in the metropolitan area of Mexico City (McKinley et al. 2003). 5 Other studies in Mexico have used data coming from international studies, adjusting costs using the differences in purchasing power parity. Table 4 presents a summary of the unit costs for morbidity and mortality used in Mexico for estimating externalities. Table 4. Unit Costs of Morbidity Impacts, US$ 2005 Cost of Productivity Willingness Impact Illness Loss To Pay Total Chronic bronchitis 18,941 127 80,188 99,256 Hospital admission Respiratory 2,336 145 188 2 669 Cardiovascular 11,621 73 188 11,882 Emergency visit 288 91 97 476 Asthma attack 338 18 87 443 Restricted activity days - 18 20 38 Sources: CEPAL-SEMARNAT (2004) 5 All unit costs were adjusted to 2005 dollars. 22 Value of Statistical Life No doubt, value of statistical life (VSL) is the most controversial figure used to analyze external costs. As will be shown in the sensitivity analysis, total external costs are highly sensitive to this value. Typically, external costs are dominated by those associated with mortality and thus results of the sensitivity analysis are expected. In general, the approach to value chronic mortality, using the VSL per case or converting each case to the associated YOLL, is critical for externalities valuation. To date, only one study has been conducted in Mexico to estimate the VSL. Because of the importance of the VSL to estimating the external costs of this research, we briefly describe the methods used in that study in this section. Conducted by Hammitt and Ibarrarán (2006), the study used compensating wage differentials to estimate marginal rates of substitution between income and occupational-injury risks in Mexico City. The study surveyed some 600 workers to identify their perceived risks of both fatal and non-fatal occupational injury. The survey was supplemented by actuarial-risk estimates to prove for consistency. Hedonic-wage regressions were estimated by regressing the logarithm of the hourly wage on occupational risk and human-capital variables (e.g., age, schooling, work experience, and sex). The mean age was 33 years, and mean education was 8 years. Experience with current employer averaged 6 years and average monthly wage was Mex$3,500 (US$350). Estimates of the value per statistical life were in a range of US$235,000– $325,000 (2002 figures); these values were much smaller than corresponding estimates for higher-income countries, at US$4–9 million, but were compatible to the few estimates previously made for lower-income countries such as India (Hammitt and Ibarrarán, 2006). Since mortality impacts are the main component of external costs, it is important to conduct some sensitivity analysis, particularly since few studies are available for this important figure in the estimate. Transferring the VSL from U.S. studies would give a total value of US$560,000 (2000 figure), and chronic mortality would be estimated at US$13,327 per YOLL. By comparison, transferring the VSL from EU studies would result in a value of $17,485 per YOLL. Thus, the Value of Statistical Life is in the lower side of the range of values that would usually be used in this type of studies, in the case that there were no reference studies in the country. The discount rate used was 3 percent, a commonly accepted discounting parameter for health policy. 23 Externalities As explained, the values of externalities are a result of several processes that transport pollution from its origin until it has an impact. The total impact is a function of the total emissions, but also of the meteorology affecting pollution transport and the population density in the areas where this pollution is emitted. Mexico has a standard to define that defines the emissions limits for combustion equipment (NOM-085). This regulation defines critical and non-critical areas. Critical areas are those very densely populated such as the metropolitan areas of Mexico City, Guadalajara and Monterrey, or that are critical due to the high levels of pollution emitted by industry which can result in violations of the ambient levels of pollution that are consider adequate to protect human health or the environment. The available externality estimates for power plants in all the country and for power plants in some critical zones are presented in Tables 5 and 6, respectively. Estimates for power plants in all the country are based on simplified methodologies that aimed at having a first approximation of the external costs, while those for critical zones utilize more complex models. Models using the simplified method estimated lower externalities costs as a result of more conservative assumptions. Table 5. Externality Costs for Power Plants US$/ton Region PM10 SO2 NOx US¢/kWh Tuxpan 219 397 235 0.70 Manzanillo 189 328 193 0.56 Tula 1,157 359 199 0.66 Petacalco 121 244 144 0.41 Río Escondido 121 93 319 0.26 Salamanca 991 360 200 0.71 Altamira 715 395 227 0.83 Mazatlan 396 278 193 0.59 Puerto Libertad 7 121 72 0.23 Samalayuca 63 132 78 0.20 Rosarito 502 28 619 0.12 Source: SEMARNAT-ECLAC 2004. Table 6. Externality Cost for Two Critical Zones Total external cost (million US¢/ Region US$, 2004) kWh Tula 402.5 3.63 Salamanca 85.9 2.70 Source: SEMARNAT-ECLAC 2007. 24 The 2004 study of SEMARNAT and ECLAC estimated externalities for power plants located throughout Mexico, including main facilities in critical areas; while their 2007 study estimated externality costs for only two critical zones (Tula and Salamanca), using various models to account for interactions with oil-refining facilities in the respective areas. The 2007 study estimated externalities in the higher portion of the confidence range of the 2004 study, which used simplified methods. The average national cost for externalities was US$0.50 per kWh, with cost per ton of SO2, NOx, and PM10 averaging $489, $332, and $873, respectively. Currently, SEMARNAT is updating these studies using the simplified methodology of the International Atomic Energy Agency (IAEA), with the SIMPACTS model. Encompassing Climate-Change Externalities Analysis of climate-change externalities requires a separate in-depth, integrated-assessment modeling exercise detailing damages and mitigation costs. Since climate change is a global impact, the methods require a global analysis encompassing not only the electricity or energy sector, but also the impacts derived from other sectors of the economy, along with other social and ecosystem impacts. There are many problems of using cost of mitigation and price of carbon in current markets as a proxy for the social cost of carbon (SCC) (See World Bank 2010b). Currently, SENER proposes using a method that considers an annual review of the regional or international carbon markets to set a value on the carbon externality. The main problem with this approach is that the current value of carbon credits is a poor estimate of the real marginal damage to the economy and other unvalued Earth-system impacts that accrue from CO2 emissions. Integrated assessment modeling has been used to estimate climate impacts, and extensive research has been devoted to estimating the social value of 1 avoided ton of emissions. For example, the Dynamic Integrated Climate Economy (DICE) model of Nordhaus (2008), the first integrated assessment model focused on the climate issue, provides mitigation paths to climate change, estimating an optimal carbon price or tax of US$27 per MT (2005 figure). By contrast, the British government’s Stern Review (Stern 2006), which estimated climate-change externalities using the Policy Analysis of the Greenhouse Effect (PAGE) model, resulted in a carbon cost of $320 per MT. Such divergence in modeling results stem from different model assumptions, particularly with regard to the discount rate used for inter-temporal valuation and the slope of the mitigation costs and damage curves (Mendelsohn 2007; Dasgupta 2006). The Intergovernmental Panel on Climate Change (IPCC) literature review revealed a median carbon cost of $45 per MT; however, the ranges were quite large. The IPCC review showed that stabilization scenarios at about 550 ppm CO2e by 2100 imply carbon prices rising to 20–80 US$ per tCO2e by 2030 and to 30–155 US$ per tCO2e by 2050. For the same stabilization level, studies since the Third Assessment Report that account for induced technological change lower these price ranges to 5–65 US$ per tCO2e in 2030 and 15–130 US$ per tCO2e in 2050. In its second study of externalities, SEMARNAT used a value of $29; 25 however, the costs of climate change were not incorporated into the final estimates and were only used for comparison with total health impacts. In many countries, pricing carbon has been a long-standing policy goal. The major policy exercise is perhaps the European Union allowance trading system derived from the EU commitment to the Kyoto Protocol. The Clean Development Mechanism (CDM) carbon market is also a reference of policy implemented; still in its infancy, the policy instrument is fragmented as CO2 prices fall short of estimating the SCC. Despite not having been a signatory to the United Nations Framework Convention on Climate Change (UNFCCC), the U.S. administration has made some efforts to estimate the SCC for regulatory purposes. For example, the EPA used it to enforce new Corporate Average Fuel Economy (CAFE) standards to increase vehicle fuel efficiency. Table 7 provides averages for several discount rates, as well as an overall average since the expert panel formed to analyze methods for computing the SCC did not agree on a single rate, given the open discussion in economics about adequate value for discounted future social benefits. Table 7. Social Cost of CO2 in the US regulatory studies, 2015–50 US$ 2011 Discount rate (statistic) 5 percent 3 percent 2.5 percent 3 percent Year (average) (average) (average) (95th percentile) 2015 $12 $39 $61 $116 2020 $13 $46 $68 $137 2025 $15 $50 $74 $153 2030 $17 $55 $80 $170 2035 $20 $60 $85 $187 2040 $22 $65 $92 $204 2045 $26 $70 $98 $220 2050 $28 $76 $104 $235 Source: EPA (http://www.epa.gov/climatechange/EPAactivities/economics/scc.html). Note: The values for the social cost of carbon are dollar-year and emissions-year specific. Studies conducted by the World Bank have typically used a value of US$23 per tCO2. For our 2006 reference year, power-plant emissions were 84,513.50 million tCO2. Using the $23 reference estimate for a ton of carbon yields US$1.943 billion in total costs associated with climate change. Using an uncertainty interval of $5 per ton of carbon (conservative estimate) to $320 per ton of carbon (high estimate of the Stern Report) yields a range of $433 million to $7.375 billion in external GHG emissions costs for Mexico’s power sector. 26 Key Recommendations Improving these estimates will require efforts on a number of fronts. These include improving emissions inventories by increasing the frequency of measurements in major power plants, along with increasing the meteorological monitoring of power plants and making the findings public for research use. In addition, models should be calibrated for estimating ozone. Furthermore, research is needed to improve the ERF, particularly for chronic mortality, and to estimate the VSL by contingent valuation, particularly for environmental improvements. Finally, evaluating climate-change policy requires state-of-the-art global models with regional detail in order to account for the impacts of various policies at regional and global levels. 27 Chapter 4. Internalizing Externalities SENER has established processes for internalizing the external costs of investment projects into short- and long-term planning, economic dispatch of the system, and the cost-benefit analysis stage of project submission. This chapter analyzes these processes in order to better understand the feasibility of using the methodology and identify potential barriers to implementation. The sections that follow describe the ongoing policy process and, based on the experience of four case studies, how investment decisions have been affected by including external costs. The four case studies refer to the analysis of four investment projects in different electricity technologies: a) a natural gas combined cycle, b) a geothermal power plant, c) wind farm and h) hydro power plant. To evaluate the use of externalities in the economic dispatch of the system and long-term expansion we run two simulation models. Including externalities in the Economic Evaluation Process Currently, the long-term investment planning process for the electricity sector is the 15-year POISE, described in chapter 2 (Figure 9). Once the POISE has been decided, the specifics of each project are analyzed in detail. Each year, CFE prepares and SENER authorizes projects in accordance with the overall strategic energy-sector plans and programs defined by the POISE. Externalities are used to include some of the environmental benefits of renewable energy projects in the socioeconomic evaluation of projects. SHCP, which has oversight responsibility for the sector, approves the investment projects under a particular financial scheme. Project Financing About 70 percent of projects are financed through the Investment Projects with Deferred Expenditure Registration (PIDIREGAS) scheme, while 5 percent are public works projects— mainly hydroelectric and system maintenance projects—which are funded directly through federal budget allocations (Figure 12). Figure 12. Overview of Project Evaluation Process and Financial Schemes PIDIREGAS Financing Submission to Authorization requirements SHCP POISE Public works Budget Authorization END Source: Authors, with data from SHCP. PIDIREGAS is a long-term financing scheme for infrastructure development used by the Mexican government to supplement investments in the oil and electricity sector. It works on a deferred financing schedule. To qualify for execution under PIDIREGAS, a project must generate enough resources from the sale of goods and services to cover the incurred financial obligations. Projects are paid with revenues generated during their operation and require the signature of a long-term contract for energy sales (or associated good or service). Under the scheme, the federal government assumes the investment risk since PEMEX or CFE sign the contract as guarantee, while private investors recover their investment within an agreed time frame. Federal guidelines for projects under the PIDIREGA scheme cover comprehensive socioeconomic and financial-performance evaluations (Table 8). SENER’s methodology establishes that externalities of local, regional, and global pollution should be used in the socioeconomic evaluation, whose objective is to decide whether the project should take place. The underlying concept is that the government can maximize social welfare by taking the overall stream of costs and benefits into account, considering market values and externalities. Table 8. Project Evaluation Guidelines for PIDIREGAS Project Evaluation type category Socioeconomic Financial Verify that project generates Decide whether project should sufficient net income each year to Objective take place. cover debt and payments. Investment requirements Financial costs and capacity Costs payments Savings in total costs, reduced Income from energy sales, minus Benefits failure rates and external costs production costs. Source: SHCP 2008. The objective of the financial evaluation is to ensure that the project can generate enough income to cover debt and payments. To demonstrate this payment capacity for each year of operation, a special requirement covering both financial costs and capacity payments is included. SENER’s methodology indicates that any financial instrument (i.e., either carbon credits or taxes) should be considered in a project’s financial evaluation as revenue or expenses associated with the particular technologies (Figure 13). Since Mexico has started applying carbon taxes to some fossil fuels, these should be reflected in the cost of new projects. 29 Figure 13. Project Evaluation Approval Process, Highlighting Inclusion of Externalities Investment plan Economic Include externalities of Modify plan evaluation SO2, NOX, PM, and GHG Benefits and costs PIDIREGAS CFE OWN RESOURCES Define financial scheme Include financial instruments associated with carbon markets Financial (revenue) or regulations (taxes) evaluation NO YES Profitable ? Document Document evaluation evaluation Request final decision End Source: Authors, with data from SHCP. Note: Green arrows indicate modifications derived from SENER’s new methodology. 30 Experience in Evaluating Emissions Externalities To internalize externalities, new investment projects must submit as part of their evaluation package a Study on Avoided Emissions. CFE’s Sub-directorate for Programming provides annual data for all projects on expected displaced generation and resulting avoided or added emissions. To estimate displaced generation, CFE runs its systems operation model twice—the first time without the new generation project (i.e., reference case) and the second time with it. Defining avoided emissions is not a straightforward process since net total emissions depends not only on the new technologies evaluated but also on the location of the power plant and the resulting generation it could displace. For example, Mexico’s National Electricity System covers five areas (Map 2), with diverse generation mixes and various transmission interconnection capacities between regions. If a new project displaces generation originating from an older conventional steam turbine powered by fuel oil, the displaced emissions will be higher than if the new project were to displace hydro or combined cycle (CC) technologies. A system-wide analysis to evaluate generation displaced by the new projects is appropriate, and is the procedure already followed by CFE. Map 2. Coverage Areas of Mexico’s National Electricity System Source: SENER 2008. All projects follow the same procedure to estimate displaced energy, incremental energy, and resulting avoided or added emissions. The next step is to estimate their value of externalities and add this value to the total benefits normally considered in Mexico for the Cost Benefit Analysis. 31 CFE has included environmental externalities in the economic evaluation stage of some of the projects presented for authorization to SHCP. We present four case studies showing how investment decisions have been affected by including external costs. We use information provided by SENER on recently evaluated projects. The case studies include geothermal, wind, hydro and combined cycle technologies. For reasons of confidentiality, actual information from these projects is scaled so that it is merely representative of any projects using similar technologies in the areas where these projects were proposed. Annex D presents the details of the calculation of displaced energy and the Benefit Cost Analysis (BCA) for each of the case studies. Table 9 shows a summary of the main figures of the BCA. Table 9. Investment Projects (Million $ of 2009) Geothermal Wind Hydro Combined Cycle Capacity (MW) 88 395 1,107 650 Region Southwest Southeast Southeast Northwest Present Value (PV) Investment 69 624 883 529 PV Environmental benefits 23 116 125 30 PV Benefits from reduced production costs 124 516 890 627 PV Benefits from Reduced NSE 10 0 669 138 SUM benefits 158 633 1,684 795 B/C ratio 2.3 1.0 1.9 1.5 Source: Authors with data from CFE The projects differ in total investment and benefits given their different size and the regions they are located. Figure 14 shows a breakdown of the projects’ benefits. For all projects the main benefit is a net reduction in production costs for the overall system. The reason is that all the projects are displacing more expensive generation from old units. The second component of benefits accrued to the projects is system security. System security is evaluated considering the minimum reserve requirements to guarantee system reliability and security. It considers a reserve margin to meet coincident peak demand and an operating reserve criterion for load following purposes. These criteria are established for the different regions and give CFE the portfolio of plants that would be optimal for the system expansion. The project accounts from benefits for the amount of firm capacity contributed to reach these margins in a particular region. In this case, the hydro project, given its size and firm power capacity, credits a big share of its total benefits from this service, followed by the combined cycle project. In contrast, the wind project has zero benefits in this category and the geothermal project only accrued 4% of its total benefit from increased system security. Environmental benefits from reduced emissions accounted for 18% of total benefits for the wind case study, 10% for geothermal, 7% for hydro and 4% for the combined cycle project. It is important to highlight that without accounting for environmental externalities the wind project would not have taken place. 32 Figure 14. Benefits from the Case Studies Reduced production costs Security (NSE) Environmental 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 CC Geothermal Wind Hydro Source: Authors. We compare the four cases in terms of the total environmental benefits and emission reductions per MW installed. The wind project reported the highest environmental benefits in dollars per MW installed, followed by the geothermal project. The hydro project reported half of the environmental benefits per MW installed than the wind project. The combined cycle project has the lowest environmental benefit per MW (See Figure 15). It is important to highlight that the overall ranking is dependent on the VoE considered for each of the impacts, which differs by region. Some power plants are closer to highly populated areas, which increase the total health costs. It is important to consider too that the cost associated with CO2 emissions was different for the different projects. The reason for this difference in the price of CO2 is the fluctuating behavior of the market of CO2, and the different assumptions that CFE and SHCP consider to estimate the social cost of carbon at the time the project were evaluated6. Figures 19 and 20 show the breakdown of the total emissions avoided per MW installed for each of the case studies. We can see that even though the emissions reduced by the MW of the hydro plant are higher the total 6 For the wind and geothermal project a value of 18.35 $/ton CO 2 was used, for the hydro project 15$/tonCO2 and for the combined cycle project the value was 8.5 $/ton CO 2. 33 environmental benefit accounted for the project was lower given the lower cost of the emissions it avoided. Figure 15. Environmental Benefits from Four Case Studies (Thousand $/MW installed) Hydro Wind Geothermal CC 0 50 100 150 200 250 300 350 Source: Authors. 34 Figure 16. Emissions reductions from Four Case Studies (ton / MW Hydro Wind PST NOx Geothermal SO2 CC 0 2 4 6 8 10 12 Source: Authors with data from CFE Figure 17. CO2 avoided emissions (ton / MW installed) Hydro Wind Geothermal CC 0 200 400 600 800 1000 1200 1400 Source: Authors with data from CFE. 35 Including externalities in Operations and Investment Decisions     36   37 Table 10. Costs of generation including externalities Capacity Costs in USD/MWh Technology Units Total Externalities health Gross Net Investment Fuel O&M Water Externality CO2 Non-critical zone Critical Zone Thermal conventional 2 350.0 332.2 25.8 87.6 6.6 1.5 5.0 34.7 19.0 121.4 2 160.0 149.2 36.4 93.7 9.3 1.8 5.0 34.7 19.0 141.1 2 80.0 74.5 46.0 100.4 11.7 2.0 5.0 34.7 19.0 160.1 Turbogas aeroturbine gas 1 42.2 40.9 93.8 52.8 15.7 0.5 3.5 12.0 162.3 1 103.7 100.3 86.8 50.4 14.6 0.5 3.5 12.0 151.8 Turbogas industrial 1 84.7 83.1 68.1 66.0 11.4 5.0 34.7 16.6 145.6 1F 182.1 178.2 58.7 57.9 9.9 5.0 34.7 16.6 126.4 1G 262.7 256.6 53.6 55.4 9.0 5.0 34.7 16.6 118.0 1H 273.9 266.2 54.6 54.2 9.2 5.0 34.7 16.6 118.0 Turbogas aeroturbine diesel 1 39.9 39.3 90.2 147.1 15.1 5.0 34.7 16.6 252.3 Combined cycle gas 1Ax1 109.0 105.7 19.1 42.5 6.6 0.0 0.5 3.5 12.0 68.1 1Fx1 281.9 273.3 15.2 38.2 5.2 0.5 3.5 12.0 58.7 2Fx1 567.0 549.6 14.7 38.0 5.1 0.5 3.5 12.0 57.7 3Fx1 851.6 825.7 14.4 37.9 5.0 0.5 3.5 12.0 57.3 1Gx1 393.7 380.7 14.0 37.5 4.8 0.5 3.5 12.0 56.3 1Hx1 789.4 763.7 13.8 37.4 4.8 0.5 3.5 12.0 56.0 2Hx1 405.7 391.3 14.0 36.9 4.8 0.5 3.5 12.0 55.7 813.6 785.0 13.8 36.8 4.8 55.3 Internal Combustion 1 44.0 42.2 46.4 80.6 18.0 0.0 5.0 34.7 145.1 3 3.6 3.3 66.4 99.8 7.7 0.1 5.0 34.7 174.0 Coal 2 350.0 331.1 33.8 28.8 8.6 2.6 26.0 25.0 71.2 Coal supercritical 1 700.0 671.4 30.3 27.0 7.7 2.6 26.0 25.0 65.1 Technology Units Capacity Costs in USD/MWh Total Coal supercritical with scrubber 1 700.0 668.6 31.4 26.9 8.0 25.0 66.4 Nuclear 1 1400.0 1351.0 72.1 5.4 13.9 1.0 91.3 Geothermal 0.0 Cerro Prieto 4 27.0 25.0 36.2 70.3 12.5 0.1 119.0 Los Azufres 4 26.6 25.0 30.7 30.0 9.9 0.1 70.5 Hydroelectric 2 375.0 373.1 112.3 0.7 10.4 123.4 3 45.0 44.8 100.6 9.8 9.4 119.7 2 8.6 8.5 45.4 8.6 4.2 58.3 Wind 0.0 Class 6 67 1.5 1.5 77.4 9.2 86.6 Class 7 67 1.5 1.5 67.7 8.0 75.8 PV 1 60.0 59.9 182.1 7.6 2.0 189.8 Table 11. Emissions Factors Emissions kg/MWh Emissions CO SO No PM Fuel oil 822.4 3.0 1.3 Coal 1083.0 4.5 1.3 1.0 Natural gas 524.0 0.2 0.6 Hydropower 15.0 Nuclear 65.0 Solar 106.0 Wind 21.0 Biomass 1403.0 4.2 3.3 39 Model for economic dispatch 40 41 0.00 0.20 0.40 0.60 0.80 1.00 1.20 -0.80 -0.60 -0.40 -0.20 Tula Tula Tula Tula Tula Salamanca Source: Authors. Salamanca Salamanca Salamanca Valle de México Valle de México Valle de México El Sauz El Sauz Critical Regions El Sauz 42 El Sauz El Sauz El Sauz El Sauz Tuxpan Tuxpan Non-critical regions Tuxpan Tuxpan Tuxpan Tuxpan Altamira Altamira Altamira Altamira Poza Rica Figure 18. Economic Dispatch with Externalities (change in generation, GWh) Poza Rica Poza Rica Table 12. Unit Unit Power plant Type Capacity Power plant Type Capacity ID ID 43 Table 13. Case 1 Case 2 Regions Regions Units Power plant Fuel Difference Baseline Externalities increasing decreasing 1 Tula (Francisco Pérez Ríos) Fuel oil 1.93 1.39 -0.53 -5 2 1.99 1.39 -0.61 3 2.24 1.58 -0.66 4 2.06 1.51 -0.55 5 1.74 1.3 -0.45 6 Salamanca Fuel oil 1.03 0.74 -0.29 7 1.02 0.74 -0.28 8 1.72 1.32 -0.4 9 1.63 1.25 -0.38 10 Valle de México Fuel oil 0.98 0.7 -0.28 11 0.94 0.66 -0.28 12 0.94 0.66 -0.29 13 El Sauz Natural Gas 0.32 0.46 0.14 1.39 14 CC 0.33 0.45 0.13 15 0.32 0.45 0.14 16 0.46 0.6 0.14 17 0.74 1.07 0.33 18 0.83 1.13 0.3 19 0.89 1.12 0.23 20 Valle de Mexico Natural gas 2.63 2.63 - 21 CC 0.73 0.73 - 22 0.73 0.73 - 23 0.73 0.73 - 24 Tula Natural gas 0.6 0.6 - 25 CC 0.6 0.6 - 26 0.88 0.88 - 27 0.63 0.63 - 28 0.63 0.63 - 29 0.94 0.94 - 30 Las Cruces diesel 0 0 - 31 TG 0 0 - 32 Sn Lorenzo diesel 0 0 - 33 TG 0.04 0.04 - 34 0.04 0.04 - 35 Tuxpan Fuel oil 2.04 3.06 1.02 4.86 36 2.32 3.06 0.75 37 2.47 3.06 0.6 38 2.34 3.06 0.72 44 Case 1 Case 2 Regions Regions Units Power plant Fuel Difference Baseline Externalities increasing decreasing 39 2.19 3.06 0.88 40 2.18 3.06 0.89 41 Altamira Fuel oil 0.98 0.68 -0.3 -1.55 42 0.98 0.69 -0.29 43 1.64 1.15 -0.49 44 1.61 1.15 -0.47 45 Poza Rica Fuel oil 0.24 0.34 0.1 0.3 46 0.24 0.34 0.1 47 0.25 0.34 0.09 48 Dos Bocas Natural Gas 0.55 0.55 - 49 CC 0.55 0.55 - 50 0.55 0.55 - 51 0.55 0.55 - 52 0.88 0.88 - 53 0.88 0.88 - 54 Tuxpan Turbogas 0 0 - 55 Los Humeros Geothermal 0.04 0.04 - 56 0.04 0.04 - 57 0.04 0.04 - 58 0.04 0.04 - 59 0.04 0.04 - 60 0.04 0.04 - 61 0.04 0.04 - 62 0.04 0.04 - Total displaced generation 6.55 -6.55 Source: Authors. Model for capacity expansion 45 Figure 19. Screening Curves (no externalities) 1000 900 800 700 Fuel oil 600 Gas turbine $/kW-year CC 500 Coal 400 Nuclear 300 Geothermal Wind 200 Solar 100 0 10% 12% 15% 17% 20% 22% 24% 27% 29% 32% 34% 37% 39% 42% 44% 46% 49% 51% 54% 56% 59% 61% 64% 66% 68% 71% 73% 76% 78% 81% 83% 86% 88% 90% 93% 95% 98% 0% 2% 5% 7% Capacity factor Source: Authors. Figure 20. Screening Curves (with externalities) 1200 1000 800 Fuel oil Gas turbine $/kW-year CC 600 Coal Nuclear 400 Geothermal Wind Solar 200 0 0% 3% 5% 8% 10% 13% 15% 18% 20% 23% 25% 28% 30% 33% 35% 38% 40% 43% 45% 48% 50% 53% 55% 58% 60% 63% 65% 68% 70% 73% 75% 78% 80% 83% 85% 88% 90% 93% 95% 98% Capacity factor Source: Authors. 46 Objective Function: 47 Where, Indices d demand block t year g generator technology p pollutant Parameters variable costs for generation type g variable costs for generation type g capital costs for generation type g Hg heat rate for generation type g Fg fuel price for generation type g Eg CO2 Emissions rate for generation type g Sg unit size for generation type g r discount rate k demand growth rate RR(t) accumulated return rate factor KK(t) accumulated demand growth factor LIFE capital lifetime WACC weighted average cost of capital CRF capital recovery factor Ld demand in block d Dd duration of block d A(d,g) Availability of solar/wind in block d Variables Z Objective Function: discounted sum of total system costs TC(t) total cost in year t FC(t) fixed cost in year t VC(t) variable cost in year t CC(t) capital (overnight) cost in year t N(t,g) Number of installed units of technology g in year t P(t,d,g) Power output from tech g in year t, block d 48 EM(t,p) CO2 ,SO2, NOX, PM emissions in year t The above is a standard formulation of the capacity expansion problem with the additions of external costs to Variable Costs (Eq. 4), following SENER’s Methodology for internalizing externalities. The model minimize total discounted value of system costs, Z, considering total fixed costs in each period, FC(t), variable costs, VC(t), and capital costs, CC(t) (See Eq. 2). The model decides the number of units to install, N, in each period t by technology type g. Total variable costs, VC, include fuel costs, health externalities and CO2 externalities (See Eq.4). Capital costs consider a Capital Recovery Factor, CRF (See Eq. 5). The usual balance constraint, that supply needs to meet demand each block load d and period p, is imposed. Following SENER’s electricity demand growth estimate, we consider demand grows at a rate of 3.4% per year, k, (See Eq. 8). Total system capacity, including conventional and renewable units, must be greater than power demanded each load block, d (See Eq. 6 and 7). Emissions from each of the pollutants are estimated using emissions factors (See Eq. 10). Table 14. Costs parameters for investment planning modeling exercise Overnight Health Technology Unit Size (MW) Capital Costs Fuel O&M CO2 Externality externalities ($/kW) ($/MWh) ($/MWh) Thermal (fuel 332 1203 87.6 6.6 5.0 19 oil) TG 100 778 50.4 14.6 5.0 12 CC 105 997 42.5 6.6 0.5 17 Coal 330 1630 28.8 8.6 2.6 25 Nuclear 1350 3475 5.4 13.9 1 Geothermal* 25 3500 0.0 12.5 Wind 1.5 1866 9.2 Solar PV 60 2985 7.6 2 Source: Authors with data from (CFE 2012). Note: * CFE includes the investment of the geothermal field in fuel costs, we increased the capital costs and left fuel cost zero to better reflect the total investment problem/dispatch problem with externalities. (Original overnight cost was 1981 $/kW) 49 Table 15. Load Blocks Load Block Duration (hrs) Power (MW) 1 500 9382 2 2050 6412 3 6210 4940 Source: SENER (2012). Note: We assume 6%, 23%, 71% peak, shoulder and base load respectively. Table 16. Availability Factors for Renewables Load Block Wind Solar 1 0.3 0.43 2 0.24 0.36 3 0.3 0.24 Source: Authors. Table 17. Renewables potential in Mexico Potential (MW) Energy Probed Estimate Possible Geothermal 111 6874 10950 Minihidro 656 3304 Wind 3105 50000 Solar 434 5000000 Biomass 118 3640 Source: SENER. National Inventory of Renewable Energy. SENER http://iner.energia.gob.mx/publica/version2.0/ We run 2 scenarios: baseline (no externalities) and policy scenario with externalities. Following we present the main results for each scenario and highlight key implications for Mex ico’s policy. Finally, we run two sensitivity scenarios to the externalities price and to different renewable energy potential. 50 Figure 21. Fuel prices ($/MMBTU) Fuel oil Gas Coal Uranium 12 10 8 6 4 2 0 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 Source: CFE(2012) Results: Baseline scenario Figure 22. Baseline scenario. Generation and CO2 Emissions (GWh ; ton) 140,000 450,000 400,000 120,000 350,000 100,000 300,000 Geothermal 80,000 250,000 Combined Cycle GWh 60,000 200,000 Gas Turbine 150,000 Fuel oil 40,000 100,000 CO2 20,000 50,000 0 0 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 Source: Authors. 51 Figure 23. Baseline scenario. Criteria pollutants (thousand ton) SO2 emissions NOX PM 700 600 500 mil ton 400 300 200 100 0 Source: Authors. 7 Those that constitute a health effect concern. 52 Figure 24. Baseline scenario. Total capacity (MW) Geothermal Conventional thermal Combined Cycle-gas Open Gas Turbine 30000 25000 20000 MW 15000 10000 5000 0 Source: Authors. Results: Policy scenario 53 Figure 25. Policy scenario. Sensitivity to externality cost Generation (GWh) Fuel oil GT CC Nuclear Geothermal 140000 120000 100000 80000 GWh 60000 40000 20000 0 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 Source: Authors. 54 Figure 26. Combination of Policies: Quantity Targets and Externalities Generation (GWh) Fuel oil GT CC Nuclear Geothermal Wind Solar 140000 120000 100000 80000 GWh 60000 40000 20000 0 2014 2016 2018 2020 20222024 2026 2028 2030 2032 2034 2036 2038 2040 2042 Source: Authors. 55 Figure 27. Combination of Policies: Renewable Portfolio Standard and Externalities Generation (GWh) Fuel oil GT CC Geothermal Wind Solar 140000 120000 100000 80000 GWh 60000 40000 20000 0 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 Source: Authors. Figure 28. Emissions (103 ton) CO2 NOX 10000 600 9000 500 8000 7000 thousand ton thousand ton 400 6000 Baseline Baseline 5000 300 4000 Cap+ext Cap+ext 3000 200 RPS+ext RPS+ext 2000 100 1000 0 0 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2018 2028 2038 2014 2016 2020 2022 2024 2026 2030 2032 2034 2036 2040 2042 SO2 PM 700 250 600 200 500 thousand ton thousand ton 150 Baseline 400 Baseline 300 CAP+ext 100 CAP+ext 200 RPS+ext RPS+ext 50 100 0 0 2018 2028 2038 2014 2016 2020 2022 2024 2026 2030 2032 2034 2036 2040 2042 2018 2028 2038 2014 2016 2020 2022 2024 2026 2030 2032 2034 2036 2040 2042 Source: Authors. Note: CO2 emissions refer to total cumulative emissions 56 Table 18. Total Capacity expansion (MW) Externalities Technology Baseline Cap+Externalities RPS+Externalities high GT 4116 1302 0 2940 CC 14580 14175 10530 5670 Nuclear 0 1350 2700 0 Geothermal 0 2450 2475 2475 Wind 0 0 2000 9562 Solar 0 0 44040 11100 Source: Authors. 57 Figure 29. Costs of Expansion Plans ($) RPS+Externalities Externalities_high Cap+Externalities Externalities only Baseline 0 2000 4000 6000 8000 10000 12000 14000 million $ Source: Authors. Sensitivity analysis to renewables learning rates and to the discount rate We investigate the sensitivity to critical parameters in the investment decision. First, one important component determining the policy impact has to do with the cost of renewable energy technologies considered. In our baseline scenario, we consider the costs for capital investment for wind and solar of 1866 and 2985 $/kW, as used by CFE for planning purposes. However, it is well known that costs for renewables have been decreasing over the last years, with important cost reductions achieved as global installed capacity increases. We introduce an exogenous decreasing trend for capital cost reductions for wind and solar. We use two scenarios for the learning curves of renewables. First, we consider a conservative estimate following the World Energy Outlook modeling assumptions that considers that future learning rates for the period between 2010 to 2035 of 5% for geothermal, 5% for wind onshore, and 17% for solar PV at large scale. We also use a more optimistic scenario for cost reductions of renewables, following the estimates of the International Energy Agency (IEA) Technology Roadmaps (IEA, 2013 and 2010). In the case of on-shore wind, the IEA considers that capital costs could decrease 25% by 2050. For solar utility scale photovoltaic systems, the IEA expects that large-scale utility system prices can drop to $1800/ kW by 2020 and $800/kW by 2050. These estimates consider high investments on research and development, a working assumption of the policy scenarios of the technology roadmaps. To explore illustrative results of these trends, we impose exogenously a decreasing cost of capital for the renewable energy technologies to match these targets. 58 Also, we conducted sensitivity to the discount rate. A 12% discount rate is high and penalizes capital-intensive technologies. We use a 7% discount rate instead, as a reference number typically used for renewable energy public investment projects in the US. We found that the carbon and health externalities values needed to trigger changes in the investment decision decreases significantly in all the scenarios, considering both learning and lower discount rate, particularly on the high learning curve scenario. However, the carbon and health externalities at the value currently consider by the Mexican government were not enough to change the investment decisions. We analyze the combination of externalities and a 35% renewable portfolio + externalities and found, as expected, that considering future technology cost reductions and lower discount rates significantly reduce the policy cost. Figure 30 shows the difference in costs from the renewable portfolio standard policy, with high discount rate and no technological learning, and the sensitivity analysis with lower discount rate and cost reduction estimates. This policy simulation also outperformed the cap and higher externalities case, adding to the importance of considering future technology trends and lower discount rates. Interestingly, we find that capital cost reductions in the upper range of years of the evaluating period did not decrease the costs of the program as much. Figure 30. Sensitivity analysis Policy costs, $ 14000 12000 10000 million $ 8000 6000 4000 2000 0 High learning rate and Low learning rate and No learning and Discount Discount rate =7% Discount rate =7% rate=12% Source: authors. 59 Summary Remarks In this chapter we laid out the new process to internalize externalities in project valuation, and we analyze Mexico’s experience using four case studies. Adding externalities was beneficial for the new power plant investment projects. In particular, we found that accounting for externalities can promote investment in renewable energy projects that would have otherwise not taken place. By developing two models to analyze the dispatch and capacity expansion decisions under the new policy, we found out that a) adding externalities affect the short-term dispatch decision switching generation from critical to non-critical areas and b) accounting for 60 Chapter 5. Conclusion The main objective of this paper was to evaluate the impact of the new methodology to internalize environmental externalities in Mexico. Also, throughout this study we had two sub- objectives: find barriers that prevent the methodology to be implemented and identify areas of improvement of the methodology. To conclude, we present conclusions and recommendations. Evaluation of the methodology This study shows that the methodology published by SENER on December of 2012 to include local, regional and global pollution can be implemented in Mexico. We show that Mexico has already experience using externalities in the evaluation of investment projects, and that this method has proven useful in incentivizing investment in cleaner technologies. While these actions show the determination of the country to incorporate externalities in decision-making, critical steps remain for the internalization of externalities in the economic dispatch and in the long-term planning of the system, as established by SENER’s methodology. In this study we show that the methodology can be applied in the economic dispatch and capacity expansion decisions, and that its implementation could change the selection of units operating and technologies to invest in. While some approximations were needed, we believe enough information has been developed in the country to move forward with the implementation of this methodology. As shown in the four cases examined and on the modeling results, the costs of electricity generation rise and therefore further discussion of how to account for these new costs is needed in Mexico. If Mexico moves forward with this implementation, it could set an example of how big state-own enterprises could use sound methodologies to incorporate environmental costs in their investment decisions, as well as sending market signals regarding the value of clean technologies for their partners in the private sector, the independent power producers. Many developing countries rely in this type of companies to deliver their electricity, with different schemes of private participation, and therefore, this innovative legislation in a developing country the size of Mexico could provide a real world example of the benefits and challenges of such a policy. While having an economy-wide carbon tax or trading scheme for the overall economy would be preferable, this approach is a promising policy to internalize environmental costs in power generation. 61 Barriers that prevent implementation Promoting investment in clean technologies including renewable sources of energy and carbon- reducing technologies is everything but a simple methodological exercise. It has much more to do with policies and regulation driving the financing of the extra investments needed to secure the environment benefits / mitigate externalities. To internalize externalities, stimulating an active dialogue among the stakeholders seems as important as ensuring quality methodological work. For this reason, the main barriers of implementation identified in this exercise are in the arena of stakeholders’ interaction and ensuring adequate financing of incremental costs. As highlighted in the document new responsibilities for different stakeholders arise for the implementation of this methodology, and adequate stakeholders interaction is needed to prevent barriers halting implementation. The complexities involved in the implementation along with the multiple agencies that need to be involved in the application of the methodology can constitute administrative barriers slowing down the policy process. Some of the concerns expressed by government officials and the public during the consultancy process include: 1. Stakeholders disclose concerns regarding the overall impact that this policy could have in electricity tariffs, which was not a topic addressed in SENER’s methodology. 2. CFE needs to include the VoE in the investment projects presented for approval to SENER and SHCP. As shown in this analysis, the incorporation of these values could accelerate the replacement of inefficient equipment using fuel oil, and coal, potentially affecting CFE finances that owns and operates directly this infrastructure. 3. SHCP and SENER need to align long-term planning and investment projects evaluation in order to make sure that they are in consent, as differences could result in suboptimal decisions, in this case particularly as it refers to environmental costs. 4. SEMARNAT needs continuous assistance to provide updated values for the costs of externalities from different technologies for electricity generation. The methodology established by SENER requires updated analysis every 3 years. This is a critical step for implementation, and the values estimated will evidently determine the overall impact of the policy. For this reason, SEMARNAT will need resources to build capacity on this matter. The time needed to update these values and the uncertainties around the estimates were mentioned as potential barriers. 62 Recommendations and areas of improvement of the methodology A detailed description of areas of improvement of the methodology to estimate the VoE is presented in each of the method steps in Appendix C and in the previous report (World Bank 2009). Here we highlight the ones we consider could have the more dramatic impact and therefore should have special attention during the implementation and future actions: a) Define a work plan for improving externalities valuation and publish official Guidelines This work plan could include the development of guidelines for each of the different steps of the methodology, stating in each one, the models and information that should be used for regulatory purposes. The variety of methods to estimate externalities could lead to different results and therefore it is important to standardized and set transparent guidelines so that the authorities and the power producers know what are the models and methods to be applied in the process. This could include the elaboration of a special Guidance on Treatment of the Economic Value of a Statistical Life, a Guidance on the Use of the Social Cost of Carbon and the definition of regulatory models to be used for pollution dispersion. For example, in the US, the Office of Management and Budget has primary responsibility for coordinating and reviewing regulatory analyses across federal agencies and sets guidelines for conducting cost-benefit analysis including non-market values. Different agencies follow these guidelines and further define the values needed in their regulations. For instance, the US Environmental Protection Agency set guidelines and define the regulatory models to be used in preparing economic analysis of environmental regulations and policies. These procedures are published as official documents and are available to the public online. Models to be used for air pollutants dispersion are made available online as well. This favors transparency and allows anyone interested in overviewing the process to have an accurate way to understand/reproduce the analysis. This work will favor not only the process of internalizing externalities in the electric sector, but also the comparison between different investment options in environmental protection and climate change mitigation. It also serves the purpose of facilitating further development and improvement of methods by the academic community. In Mexico, the SHCP could provide this executive guidance, and SEMARNAT and SENER could further define in their respective areas values to assess environmental costs, publish the guidelines and procedures, and update them as information becomes available. b) Cost of Carbon The methodology mandates SEMARNAT to annually send to SENER a proposal of CO2 prices to use. Therefore, SEMARNAT needs to evaluate approaches to estimate the social cost of carbon and evaluate trade-offs. SENER however establishes that this analysis should be based on regional or international markets. We believe this could harm the application of the methodology, since current prices in the carbon market do not reflect the real social cost of carbon. For this reason, we recommend to establish a shadow price that could serve the purpose 63 of valuing carbon more reasonably. In previous studies of the World Bank a value of $23 usd/ton of CO2 has been used. For the financial evaluation of the projects, a market value is adequate, but not for the socio-economic evaluation or system planning. As discussed in Chapter 3, recent studies have updated the value of the social cost of carbon in the US used for policy analysis to $39 dollars per metric ton of CO2 (using a 3% discount rate) and the international literature as assessed by the IPCC Third Assessment Report shows a range for the implicit price of carbon from 5 to 65 US$/tCO2eq in 2030 and 15 to130 US$/tCO2eq in 2050. c) Improve the valuation of health outcomes: Several recommendations were derived on this specific issue, since health outcomes constitute the main component of external costs. Currently, health externalities in Mexico fall into the low range of externalities found in developing countries. The measures to improve this valuation could include:  Conduct more studies with different methodologies to estimate the Value of Statistical Life (VSL) in Mexico, including WTP studies. The external costs are typically dominated by costs associated to mortality and therefore the approach to value chronic mortality is a key number in the valuation. Currently, there is only one study in Mexico with an estimate of US$235,000 a value much smaller than corresponding estimates for higher-income countries, which are in the range of US$4-9 million. In the US, for instance, several studies are used to recommend a figure for regulatory purposes (the USEPA recommends a central estimate of $7.4 million ($2006) for all regulatory analysis of environmental regulation).  Conduct a valuation study of medical health costs in different levels of service in Mexican institutions, this will allow to better capture de costs of morbidity in different areas of the country  The Unit of Economic Analysis of the Ministry of Health in Mexico, recommended the following regarding the use of health and population statistics: o Use of the new information systems for incidence data. Considering that these systems capture only the public sector hospitals, and therefore could be underestimating the total cases they also recommended the use of different statistics and weighting other institutions for a better approximation (Hospitals of Health Assistance (SSA), Hospitals for the workers of the public sector (ISSTE), private hospitals, etc.) o To value productivity loss, they recommend the use of the value for lost salaries of the Circular from the Mexican Institute of Social Security (IMSS) instead of the Household National Income and Expenditure Survey’s minimum wage. The 64 use of the minimum wage currently used by SEMARNAT will undervalue the total impact o Use the codes of the Ministry of Health system (CIE 2008) to be precise on the diseases that are reported, as well as on the incidence rates o For population, use CONAPO mortality statistics since it is important to smooth the curves used, a treatment INEGI data does not have o Consider the use of the World Bank Methodology (World Bank Group, 2006), for risks factor and Burden of Disease Framework, which is used by the Health Ministry to evaluate health policies impact d) Clarification of the definition of externalities to be considered: Different definitions of externalities in the Mexican Laws (LGCC, LAERFTE and LSPEE) are used, which could be misleading in the implementation process. For example, sometimes the Laws refer only to environmental impacts, and sometimes they require the internalization of social and cultural externalities as well. Clarifying the definitions and scope is important. It is recommended to prioritize the externalities and include in the working plan for valuation different phases so that information can be use as studies to value other impacts become available (but do not prevent the use of available information on already valued impacts). In the future, other externalities other than human health impacts could be included. In particular, damages to sensitive ecosystems or relevant resources to local communities (fisheries, tourism, etc.) e) Valuation of Ozone. Estimating the value of external costs of generation associated with ozone has not been done and could be an important reference for the VoE, particularly given that the expansion of the electric system is mainly driven by the use of natural gas which NOX emissions (ozone precursor) are important (more important than SO2 or PM, the main health problem for fuel oil or coal power generation) f) Discount rate, technological learning and indexing the externalities As shown in our modeling exercise, using a lower discount rate and considering expected technological learning from renewables is critical to reduce policy costs. Also, SENER’s methodology indicates that the externality studies need to be conducted every two years. In addition to updating the studies, for planning purposes, we recommend to allow external costs to follow a growing pattern following at least the inflation rate. Otherwise, our modeling results show that the costs of externalities being flat contribute less and less in the long-term capacity 65 expansion problem. In practice, the values should reflect the welfare losses which would increase both as the economy increases and also as income increases (it is well documented that as income increases the willingness to pay for environmental quality increases). g) Valuation of other environmental externalities We recommend conducting studies that can inform on the external costs of other technologies, including hydropower and other renewables. Some of this external costs studies will need to develop frameworks to value ecosystem services. For this, we recommend building on the Wealth Accounting and Valuation of Ecosystem Services project (WAVES) an initiative supported by the World Bank to enhance green accounting and considering the value of the environment in policy design. In particular, the methodologies used to value biodiversity losses and disruption/loss of natural ecosystems can be useful to develop cases for impacts of energy infrastructure projects in Mexico. h) Valuation of Energy Security While Mexico emphasizes the need to include energy security constraints in energy planning, there are no methodologies to incorporate this important consideration. Incorporating externalities to diversify the energy matrix of the country can contribute to energy security, and therefore some methodologies could be developed to consider this important relationship. i) Standardization of emissions calculation and use of information technology to have an open dataset of emissions The goal of this standardization should be to have an annual update of the emissions inventory to use in this exercise available to all stakeholders, preferable online and accessible timely for the implementation. In the methodology published, there are several steps dedicated to information flow. We recommend having a practical approach of how these emissions should be computed and using methods that will allow comparability within years and technologies, as well as, opportune data. Using information from the COAs in this regard could add unnecessary time constraints. In addition, since these emissions will have a value, it is important that all power plants and stakeholders have access to the numbers use for transparency purposes. For this, an open online dataset could be built.Also we recommend having guidelines for estimating NO x emissions that do not use the “maximum” limit established in the bidding or tender process or the standard NOM-085 environmental maximum limit. As explained, this could be imprecise and make the valuation process incomparable between different bidding processes and/or send incorrect signals to minimize NOX emissions. Given that the system expansion is planned with natural gas technologies, correctly assessing the impact of NOx should be prioritize. 66 j) Private sector participation After Energy Reform in Mexico, and the expected increased participation of the private sector in electricity markets, we recommend to clarify how this methodology will apply to new private generators. While there is a mention of IPPs in the methodology published, it is not clear what adjustments should be made to consider the emissions of all electricity generators. k) Conducting uncertainty analysis to the costs of externalities Given the uncertainty ranges reported in the valuation studies for health impacts and the size of the ranges reported in the literature for the social cost of carbon, conducting uncertainty analysis in the application of this policy is recommended. In particular, using models for decision making under uncertainty could be useful such as stochastic. In the case of carbon prices, sensitivity analysis could be useful as well, considering different levels of carbon prices corresponding to different discount rates. l) Building capacity for model building and maintenance. As explained in chapter 3 and 4, several models are required and some of the models used in Mexico could be improved. A research program and technical assistance to gradually improve the estimates of externalities is a desirable next step to improve the methodology. Both models for estimate pollution dispersion and models to value costs should be built, parameterized to the conditions in Mexico, and maintain so that the data and assumptions represent as accurately as possible local conditions. m) Aligning environmental regulation and economic instruments Finally, it is recommended that the application of environmental externalities be aligned to the overall environmental regulatory framework applied to the sector. For instance, the sector has to comply with several command-and-control instruments such as: emissions standards, environmental impact assessment, and local air quality programs. In addition, other programs such as the Special Program on Climate Change and the Special Program on Renewable Energy could require specific emissions reductions/technology adoption from the sector. While this has not yet been discussed in the context of the application of the externalities methodology in Mexico, it is recommended to align instruments so that they work together, reinforce each other and do not impose excess burden on the power sector. As shown in Chapter 3, the combination of policies is critical to incentivize investment in clean technologies. In sum, this study showed that SENER’s methodology could be applied, and accelerate adoption of cleaner technologies. It identifies potential barriers of implementation, areas of improvement of the methodology, and derived recommendations to move forward with the policy implementation. The study is timely also because it responds to the objectives of the National Energy Strategy 2013-2027, which was recently published by the Federal Government of Mexico. 67 Annex A. Method to Value Externalities for Mexico’s Electricity Generation 1. Scope and Objectives 1.1. This document establishes the methodology that the electricity Supplier should follow to value the externalities associated with electricity generation in Mexico that impact the economy, society, the environment, and health, considering the different technologies and energy sources. 1.2. The Supplier for this Methodology refers to Article 2, section XVI of the Code of Regulations of the Electricity Public Service Law (LSPEE). 1.3. The Ministry of Energy developed this Methodology based on Article 10 of the Law for Harnessing Renewable Energy and Financing the Energy Transition (LAERFTE), Article 16 of the LSPEE, and Articles 32 and 34 of the General Law on Climate Change (LGCC) in order to accomplish the objective of valuing the externalities associated with electricity generation, by using various energy sources at different scales in the system. 1.4. The valuation of environmental and social externalities associated with electricity generation included in this Methodology will allow for evaluating their impact on the development plans of electricity generation projects, which will impact the expansion plans of the system and the short- run economic dispatch of the power plants. 1.5. To implement this Methodology, consideration should only be given to the positive and negative impacts that can be estimated through proven and accepted methods based on technical, economic, and environmental criteria and whose values differ significantly from zero. In accordance with the Intergovernmental Panel on Climate Change (IPCC), health and climate change are the only impacts covered by these criteria. However, the application of this Methodology does not prevent consideration of other impacts that could be estimated with these criteria in the future. 2. Definitions 2.1. Displaced energy. Energy that will be substituted by the new generation project under consideration. The Supplier shall indicate the distribution of locally or regionally displaced energy. 2.2. Electricity for public-service use. Energy derived from Supplier-owned power plants, independent power producers, or imports from the Supplier through long-term contracts. Generally, this energy includes all generation sources considered in the Electricity Sector Infrastructure Investment Program (POISE). Exceptions to this definition are the occasional Supplier purchases of energy from auto-generators, co-generators, small producers, importers, and electric utilities in other countries. 2.3. Incremental energy. The energy generated by the new project under consideration. 2.4. Externalities. The positive or negative short-, medium-, and long-term impacts, whether actual or potential, associated with electricity generation caused by the provision of a local, regional, or global good or service on a third party in Mexico. Externalities occur when the cost paid for a good or service differs from the total costs of damages or benefits in economic terms. 2.5. Environmental and social externalities. These externalities are those associated with electricity generation with positive or negative environmental and social impacts, generated from the production or consumption of a good or service for which there is no compensation or payment of costs or benefits from the good or service. Externalities that manifest environmental, social, or health impacts are not excluded. 2.6. Electricity Emissions Inventory. A study conducted by the Supplier, based on yearly operational parameters, including the fuels used by its power plants, independent power producers (IPPs), and small producers, that shows the emissions of pollutants associated with the various fuels used to meet the demand for electricity for public-service use. The inventory should include emissions from hydropower, geothermal, and other non-fossil fuel energy sources. In addition, the inventory should include the emissions of other operators when such information is available. 2.7. Financial instruments associated with the emissions of pollutants. The mechanisms that allow an entity to receive an income or pay a financial cost as a result of the emissions of pollutants or actions to reduce those emissions. These refer to carbon certificates, which create a financial income, or a carbon tax promulgated by law, which could generate a financial cost. The financial costs of these instruments could differ from the value of the impacts created by the produced or prevented pollution associated with the financial instruments. 2.8. Methodology. Valuing Externalities Associated with Electricity Generation in Mexico. 2.9. MWh. Megawatt hour. 2.10. System. National Electricity System. 2.11. Supplier. Federal Commission for Electricity (CFE), as referred to in Article 2, section XVI of the LSPEE Code of Regulations. 2.12. Pollutants. Nitrogen dioxide; sulfur dioxide; total suspended particulates (TSP); and greenhouse gases (GHGs), which include carbon dioxide, methane, hydrofluorocarbons (HFCs), and sulfur hexafluoride. 3. Data Estimation for the Methodology 3.1. The methodology requires the following data: 69 3.2. To quantify pollutant emissions from the various technologies considered in electricity system planning, the Supplier will annually include the quantity of pollutant emissions per megawatt hour for each technology in the COPAR document, “Reference Costs and Parameters for the Formulation of Investment Projects in the Electricity Sector.” These numbers will be estimated, considering the emissions inventory of the electricity-sector power plants operated during the previous year. For technologies not in use in the country, the Supplier should use the numbers provided by the equipment manufacturer or those recognized by international or national institutions as the reference values. 3.3. For pollutant impact valuation, the Ministry of Environment and Natural Resources (SEMARNAT), in coordination with the Ministry of Energy (SENER) and other responsible federal agencies, will develop studies to measure the local and/or regional short-, medium-, and long-term impacts (in pesos per ton of emissions) associated with electricity generation on Mexico’s economy, society, environment, and health. These studies should emphasize the valuation of environmental and social impacts, using official information sources, and should be updated at least every three years. 3.4. For the valuation of financial instruments associated with pollutant emissions, SEMARNAT will send a document to SENER with the expected short-, medium-, and long-term mitigation values, based on regional or international markets where Mexico’s electricity projects are eligible or the fiscal regimes in which they are applied. This study should be updated annually. 3.5. The Supplier will develop the Electricity Emissions Inventory based on the annual operational parameters and fuels used for generating electricity for public-service use. 3.6. For projects operated by IPPs, SENER will obtain the emissions from the annual operational reports that they present to SEMARNAT. 4. Supplier Application of the Methodology 4.1. Electricity Sector Planning 4.1.1. The POISE shall incorporate the Environmental and Social Externalities according to Article 36bis of the LSPEE and Article 34 of the LGCC. 4.1.2. For the new power plants, the Supplier shall use the unitary values in tons per megawatt hour for each pollutant and technology. In addition, the Supplier shall use the values of the Electricity Emissions Inventory for operational power plants. 4.1.3. SEMARNAT shall give the supplier, through SENER, the unit values of the environmental and social impacts per region and the unit values of the financial instruments associated with pollutant emissions in pesos per ton of pollutant. Those values will result from the studies that SEMARNAT shall coordinate for this purpose. In cases where no values are available, the Supplier will present a proposal to SENER to resolve them. 70 4.1.4. The Supplier will compute the environmental and social externality per unit of pollutant as a proportion of the environmental and social impacts that are not currently compensated or redistributed. 4.1.5. For each pollutant, the Supplier shall multiply the unit values from section 3.2 (in tons per megawatt hour) by the unit values of the environmental and social externality from section 3.3 (in pesos per ton), the result of which shall be added to the variable costs of generation for each technology and fuel for the elaboration of POISE. 4.2. Dispatch of Electricity-Sector Units for Public Service Use 4.2.1. The environmental and social externalities shall be incorporated into the dispatch of the electricity-sector units that will be used as a variable cost excluding the externalities that are not a function of the total energy generated (for example, the emissions caused by the reservoirs of hydroelectric plants). 4.2.2. The Supplier will use the unitary values of the Electricity Emissions Inventory in tons per megawatt hour for each pollutant and power plant in operation. In the case of new power plants, the Supplier will use the reference values of COPAR for each particular case. 4.2.3. SEMARNAT shall give the Supplier, through SENER, the unit values of the environmental and social impacts per region and those of the financial instruments associated with pollutant emissions in pesos per ton of pollutant. Those values will result from the studies that SEMARNAT shall coordinate for this purpose. In cases where no values are available, the Supplier will present a proposal to SENER to resolve them. 4.2.4. The Supplier will compute the environmental and social externality, per unit of pollutant, as a proportion of the environmental and social impacts that are not currently compensated or redistributed. 4.2.5. For each pollutant, the Supplier shall multiply the unit values from section 3.2 (in tons per megawatt hour) by the unit values of the environmental and social externality from section 3.3 (in pesos per ton), the result of which shall be added to the variable costs of generation. The result should be the total variable cost used in the assignment and dispatch of units. As a result, the short-term total cost will include the environmental and social externalities. 4.2.6. According to Article 36bis of LSPEE (definition of economic dispatch) and Article 34, section Id of LGCC (selection of electricity generation sources), the Supplier shall use the least cost electricity, considering its associated environmental and social externalities, and it should provide optimal stability, quality, and security. 4.3. Cost-Benefit Analysis of Investment Projects for Electricity Generation 4.3.1. The environmental and social externalities shall be incorporated into the socio-economic evaluation of the program or investment projects considered by the Supplier. 71 4.3.2. In the socioeconomic evaluation, the following aspects shall be included: • The quantification and valuation of environmental and social externalities associated with electricity generation, with local, regional, and global impact caused by the program or project in the short, medium, and long term. • The quantification of the benefits or costs of environmental and social externalities associated with electricity generation by the new program or investment project shall be calculated as the difference in energy displaced (situation without the project) and incremental energy (situation with the project) for each environmental and social externality. • For the valuation of those impacts, the values per ton of pollutant, defined in sections 3.3. and 3.4 of this Methodology, shall be considered. • For income derived from carbon markets (foreign currency coming into the country), the quantification of greenhouse gases shall be done according to the additionality principle established by the Kyoto Protocol or in any other specific International Agreement. The price per ton of carbon shall be quantified accordingly to section 3.4 of this Methodology. 4.3.3. In the financial evaluation, the additional revenue derived from reduction in greenhouse gas emissions shall be considered in agreement with the Kyoto Protocol or any other specific International Agreement. The price per ton of carbon avoided shall be quantified as specified in section 3.4 of this Methodology. 5. Application of the Methodology for the IPP 5.1. Each year, SENER will request from SEMARNAT information on pollutant emissions for each IPP with projects in operation, as specified in section 3.6. This information shall be sent to the Supplier for incorporation into the Electricity Emissions Inventory. 72 Annex B. Investment Plan in the Power Sector in Mexico Table B.1. Investment requirements by activity 2012-2026. Million 2011 US dollars 2012 2013 2014 2015 2016 2017 2018 2019 Generation 3,457 3,099 3,426 2,861 2,583 3,315 2,614 2,658 Transmission 844 1,012 970 1,215 977 948 880 875 Distribution 1,885 1,969 1,590 1,415 1,455 1,523 1,535 1,520 Maintenance 792 565 723 770 819 846 894 930 - - - - - - - - Subtotal 6,977 6,645 6,708 6,261 5,834 6,631 5,924 5,984 - - - - - - - - Other - - - - - - - - Public 33 34 35 36 37 38 39 41 Investment - - - - - - - - - - - - - - - - Total 7,010 6,679 6,744 6,297 5,871 6,669 5,963 6,025 Table B.1. Investment requirements 2012-2026. Million 2011 US dollars 2020 2021 2022 2023 2024 2025 2026 Total Generation 3,472 4,713 5,863 6,408 7,275 4,500 707 56,953 Transmission 980 1,054 1,306 1,166 1,200 1,124 963 15,514 Distribution 1,460 1,238 1,256 1,279 1,270 1,284 1,302 21,981 Maintenance 991 1,069 1,102 1,189 1,256 1,330 1,356 14,630 - - - - - - - - Subtotal 6,903 8,074 9,527 10,042 11,001 8,238 4,328 109,077 - - - - - - - - Other - - - - - - - - Public 42 43 44 46 47 49 50 615 Investment - - - - - - - - - - - - - - - - Total 6,945 8,117 9,571 10,088 11,048 8,287 4,378 109,693 Source: CFE, POISE 2012* Exchanged rate used 13.9787 according to closing rate of December 30, 2011. www.banxico.org.mx (tipo de cambio para solventar obligaciones denominadas en moneda extranjera pagaderas en la República Mexicana publicado en el DOF. 74 Table B.2. Investment requirements by type of investment and technology 2012-2019, Millions of usd of 2011 2012 2013 2014 2015 2016 2017 2018 2019 TOTAL Generation 3,457 3,099 3,426 2,861 2,583 3,315 2,614 2,658 24,013 Independent power producers 1,811 1,485 1,838 504 51 225 59 - 5,972 New combined Cycle 712 1,128 1,753 482 51 225 59 - 4,410 New wind farms 1,099 356 85 21 - - - - 1,561 Financed public works 1,242 1,354 1,362 2,224 2,474 3,041 2,505 2,620 16,821 New Hydro 115 49 501 671 757 684 504 558 3,838 New Geothermal 8 101 123 7 86 75 30 65 496 New Combined Cycle 622 603 535 902 1,525 997 457 1,456 7,097 New Wind Farms 147 114 21 486 - 1,285 799 320 3,170 Maintenance and repower 150 237 182 157 31 1 - - 759 New Clean generation 20 50 - - - - - - 70 New turbogas 179 199 - - 75 - - 221 675 Public works 404 260 226 133 58 50 50 38 1,220 Hydroelectricity 82 45 52 45 57 49 50 38 419 Maintenance 322 215 175 88 1 0 - - 801 Transmission 844 1,012 970 1,215 977 948 880 875 7,721 Financed public works 295 440 197 114 464 417 338 305 2,570 Transmission Program 295 440 197 114 464 417 338 305 2,570 Public works 549 572 773 1,101 513 531 542 570 5,151 Transmission program 89 96 281 649 155 139 113 102 1,624 Transmission S T Y T 346 357 367 329 247 271 297 326 2,542 Modernizing CENACE 44 48 51 56 62 67 72 77 477 Modernizing Central area 69 71 73 66 50 54 59 65 508 Distribution 1,885 1,969 1,590 1,415 1,455 1,523 1,535 1,520 12,892 Financed public works 410 388 114 11 182 228 224 233 1,789 Subtransmission program 410 388 114 11 182 228 224 233 1,789 Public works 1,475 1,581 1,476 1,404 1,273 1,295 1,311 1,288 11,102 Subtransmission pgram 242 371 272 199 61 76 75 78 1,373 Distribution Program 440 431 433 440 444 454 476 451 3,569 Distribution Program central 354 332 329 320 317 319 315 307 2,594 Modernization of distribution 439 447 442 444 452 445 446 452 3,566 Manteinance 792 565 723 770 819 846 894 930 6,339 Other public works 33 34 35 36 37 38 39 41 294 Total 7,010 6,679 6,744 6,297 5,871 6,669 5,963 6,025 51,258 Source: CFE, POISE 2012* Exchanged rate used 13.9787 Annex C. Modeling Framework and Methodology While the theoretical grounding of externalities valuation is mainly in the field of welfare economics and the theory of externalities, input information is also required from other fields in order to assess environmental costs and benefits. For example, valuation of externalities (VoE) requires input data from point sources (e.g., emissions inventories and flow characteristics), modeling of pollutant dispersion, health and environmental risk evaluation, and economic valuation. To make these assessments, integrated models are commonly used; these models use interdisciplinary approaches based on knowledge of (i) the physical processes that generate and disperse pollutants and (ii) economics for providing policy makers information. Economic Modeling The economic model for analyzing the problem can be traced back to the classic model of Baumol and Oates, which maximizes consumer utility in the presence of an externality (Muller and Mendelsohn 2007). In this model, the utility function depends on the good demanded and the disutility of pollution. Firms (K) produce goods and emit pollutants (e). The emissions are dispersed in the environment and reach consumers (J). Emissions from firm k results in a concentration (zjk) to which consumer j is exposed. We wish to solve the problem such that making one consumer better off does not make any other worse off. This is expressed as follows: max ( 1 , 2 , 3 ,..., , ) * . . ( , )£0 ( = 1,..., ), å £å ( = 1,..., ), [1] =1 =1 ³ 0, ³ 0, ³ 0 "( , , ) ³ ( ¹ *) where U(X,z) is the utility function of individual j, zj is total exposure to pollution of consumer j (sum of all concentrations generated by emissions from all firms [sum of k = 1 to K]),8 xij is the amount of good (i) consumed by individual j, Fk(Y,ek) is the production function of firm k, yik is the amount of good (i) - produced/used by firm k, ek is emissions of pollutant e released by firm k, and is the initial level of utility of all other consumers. 8 For each consumer, the disutility or damage is the result of aggregate exposure to pollution, regardless of the source ( k). The Lagrangian is expressed by the following equation: - æ ö = ål ( - ) - åm çå + åv ç -å ÷÷ [2] =1 =1 =1 è =1 =1 ø The Langrarian multipliers are , , and . Therefore, differentiating with respect to ek, the optimization conditions are as follows: ¶ æ ¶ öæ ¶ ö æ ¶ ö [3] = åç l ÷ç ÷ - çm ÷ ¶ =1 è ¶ øè ¶ ø è ¶ ø Marginal damage (first right-hand term) is the sum across all consumers in society (J) because it is a public good due to an emission of ek. As Muller and Mendelsohn (2007) point out, although the economic model sums for all consumers, spatial differentiation will be important because the concentration decreases rapidly with distance from the source and because consumers care only about total concentration, regardless of the original emitting source. Marginal cost (second right-hand term) shows the resources needed to reduce one unit of e, given the market output of firm k. The optimal condition equates marginal costs to marginal damages; the condition shows the efficiency criterion for society, considering the externalities of pollution. As described by Muller and Mendelsohn (2007), modeling of multiple pollutants requires an extension of the Baumol and Oates model that not only requires summing for different pollutants but also demands a more complex treatment of the synergies and chemical transformation of the pollutants once released into the environment. We consider that consumers are exposed to a list of pollutants el (l = 1,…, L) that differ from the original list of emissions (M) generated by firms. In this case, a single emission can result in increased concentrations of several pollutants (zlj). The marginal damages are expressed at follows: æ ¶ öæ ¶ ö ( ) = ååç l ÷ç ÷ [4] =1 =1 è ¶ øè¶ ø Information on the treatment of secondary pollutants will be detailed in the dispersion analysis. From an economic perspective, it is important to highlight these complexities, which can change the optimal emission levels of pollutants with higher associated risks, as both primary and as secondary species. Integrated Assessment Modeling As shown above, externalities are a function of a number of variables, including emissions level, resulting concentrations (in a spatial basis), population distribution, and disutility assigned to the 77 damages. Valuation of external costs thus requires information on the physical nature of air pollution, health impacts, and social value assigned to the damages. The United States and Europe have used integrated assessment modeling to evaluate air- pollution damages for decades. For example, the United States Environmental Protection Agency (EPA) used such techniques to conduct cost-benefit analyses to amend such environmental policies as the Clean Air Act of 1990 (EPA 1999). The ExternE project, the more comprehensive evaluation of energy externalities in the European Union (European Commission 2005), has estimated the external costs of electricity across a broad range of technologies for various country cases.9 The main results, involving multi-year efforts of several European scientific research centers, show that energy externalities (per kilowatt hour) are 2–10¢ for coal; 5–11¢ for oil; 2–5¢ for biomass; and less than 1¢ for hydro, PV, and wind. Ongoing efforts to improve the integrated assessment method and policy applications include the AirClim project (Alcamo et al. 2002), which analyzes linkages between air pollution and climate change in Europe; the RAINS project for acid-rain control (Schöpp et al. 2005); the Integrated Assessment for Fine Particles in Europe, conducted by the International Institute for Applied Systems Analysis (IIASA) (Amann et al. 2001); and local-scale development and application of the OSCAR Air Quality Modeling System (Sokhi et al. 2008). Relevant studies in China and Cuba ratify the need to evaluate externalities in the developing world, showing that the impacts of air pollution are a serious concern with a high social-welfare cost (Jiang et al. 2008; Wei et al. 2009; Aden and Sinton 2006; Turtós Carbonell et al. 2007). Methodology Steps Characterization of the Emissions Source To characterize the emissions source, data must be compiled on the location of the point source, emissions inventory, stacks dimensions, and emissions parameters (e.g., temperature and velocity of the pollutant flow). Locating the point source involves using a Geographic Information System (GIS),10 as well as establishing whether the source is urban or rural (population density within 50 km of the source is a commonly used criterion). Establishing the urban or rural source is also important for dispersion modeling since cities have a potential heat-island effect and because some dispersion models work better in rural areas, where infrastructure friction is less of an issue. The approaches for analyzing emissions inventories are generally classified as top-down or bottom-up.iv A top-down approach uses sectoral data to estimate emissions based on emissions factors that consider fuel characteristics, while a bottom-up approach uses a detailed analysis of current 9 Figures of the ExternE project’s national implementation phase and the damages a ssessed are available online (http://www.externe.info/). 10 GIS was used to process the information. 78 technology and flow measurements at each emissions point. Clearly, the latter approach provides a more thorough analysis, yet requires information from each unit and process at a level not always available. Thus, depending on the information provided by each power plant, the emissions inventory represents more or less accurately the annual flow of pollutants. For example, available technology allows for instantaneous emissions tracking using continuous emissions monitoring (CEM) equipment, which reports hourly on the pollutant concentrations and quantities emitted. With the help of information technology, such systems can report facilities’ emissions at each point in time. This technology has been widely used in the U.S. since it was first required under Title IV of the Clean Air Act Amendments of 1990, which established the market-based Acid Rain Program.v To date, power plants in Mexico have not used CEM, and inventories are usually done using top-down approaches; more recently, however, CFE and SEMARNAT have been working jointly to improve and combine methods. Emissions-inventory techniques have important implications for valuing externalities since emissions constitute the main input for estimating damage. However, the uncertainties of various techniques are well understood and key parameters are known (e.g., measurement of content of S). In this study, available inventories are generally fine for estimating total damage since externalities are calculated yearly using an annual estimate. Emissions considered in the damage-cost function include sulfur dioxide (SO2), nitrogen dioxide (NO2), particulate matter (PM10 and PM2.5), and sulfate and nitrogen aerosols. Dispersion Analysis Atmospheric dispersion modeling, the mathematical simulation of how air pollutants disperse in the ambient atmosphere, is used to predict the downwind concentration of air pollutants emitted from such sources as industrial plants and vehicular traffic. Clearly, the real trajectory that pollutants follow is a function of complex processes and specific conditions at the moment of emission. Thus, models have an explanatory power, but cannot be expected to reproduce exactly all of the factors affecting concentrations. One should keep in mind that model estimates and actual concentrations may differ. At the same time, models do a good job of estimating the direction and potential level of pollution, which is why they are widely used globally to set regulatory emissions-control strategies. In terms of evaluating externalities, concentration estimates allow for assessing the potential impacts of a source, given its location and meteorological conditions. The most common dispersion model follows the Gaussian dispersion equation, which assumes that the pollutant has a normal probability distribution. Gaussian models are most often used to predict the dispersion of continuous, buoyant air-pollution plumes. They may also be used to predict the dispersion of non-continuous air-pollution plumes, known as “puff models.” The general Gaussian equation is expressed as follows (Barratt 2001): 79 [1] where C is the concentration of emissions in grams per cubic meter (g per m3) at any receptor, Q is the source pollutant emission rate, in grams per second (g per s); u is the horizontal wind velocity along the plume centerline, in meters per second (m per s); f is the crosswind dispersion parameter; σy is the horizontal standard deviation of the emission distribution, in meters (m); σz is the vertical standard deviation of the emission distribution, in meters (m); and g is the vertical dispersion parameter, where g1 is the vertical dispersion with no reflections, g2 is the vertical dispersion for reflection from the ground, and g3 is the vertical dispersion for reflection from an inversion aloft. Gaussian models reproduce with high certainty the dispersion of pollutants within the local domain of the emissions source. Regional dispersion in the 50–1,000 km range is better reproduced using models that incorporate the transport of pollutants and their transformation into secondary pollutants in the atmosphere (Jacobson 1999). Regional dispersion models have two distinct types of flow-field specifications: Lagrangian and Eulerian. The Lagrangian dispersion model, which mathematically follows pollution-plume parcels as they move in the atmosphere, models the motion of the parcels as a random walk process. It then calculates the air pollution dispersion by computing the statistics of the trajectories of a large number of the pollution-plume parcels. The model uses a moving frame of reference as the parcels move from their initial location. The Eulerian model focuses on specific locations in the space through which the fluid flows. Like the Lagrangian model, it tracks the movement of a large number of pollution-plume parcels as they move from their initial location; however, its frame of references is a fixed three-dimensional Cartesian grid (Barratt 2001). The Lagrangian equations combine the principles of conservation of momentum with conservation of energy. One might think of the main Lagrangian equation in the following general terms. In the model’s specifications, all fluid parcels are defined by a vector field a, with a time- independent for each fluid parcel. Often, a is chosen as the position of the parcels at some initial time t0. In the Lagrangian description, the flow velocity v(a,t) is related to the position X(a,t) of the fluid parcels (Hadlock 1988): where v is the flow velocity, x is the position of the fluid, t is time, and u and v are related through , where a is the vector field. The Lagrangian and Eulerian specifications of the kinematics and dynamics of the flow field are related by the substantial derivative (also called the Lagrangian derivative), specified in the following equation: 80 The total rate of change of some vector function F as the fluid parcels moves through a flow field described by its Eulerian specification u is equal to the sum of the local rate of change and convective rate of change of F. A more detailed analysis of Lagrangian mechanisms and their equations can be found in Barratt (2001) and Jacobson (1999). Dispersion Models Used by SEMARNAT The first SEMARNAT study, in collaboration with the United Nations Commission for Latin America and the Caribbean, applied simplified methods developed by the International Atomic Energy Agency (IAEA), under the SIMPACTS model, to conduct a national assessment of environmental externalities. The Industrial Source Complex, version 3 (ISC-ST3), was the underlying dispersion model for evaluating local externalities (within 50 km from specific power plants). The SIMPACTS model approximates estimated regional impacts, and relies on deposition patterns and equations derived from case studies in Europe rather than detailed dispersion modeling. The second SEMARNAT study used the AERMOD model for local domain dispersion in specific critical zones and the Windrose Trajectory Model (WTM) to analyze regional domain dispersion. The three dispersion models are briefly described below.11 Industrial Source Complex, version 3 (ISC-ST3). The ISCST3 model is a steady-state Gaussian plume model, which can be used to assess pollutant concentrations from a wide variety of sources associated with an industrial complex. This model can account for the following: settling and dry deposition of particles; downwash; point, area, line, and volume sources; plume rise as a function of downwind distance; separation of point sources; and limited terrain adjustment. It was the regulatory model in the US until 2005, when SEMARNAT studies were conducted. ISC3 operates in both long- term and short-term modes. It runs in two versions: short-term and long-term. In general, the model accepts simple meteorological data, which makes the implementation easier. The short-term version will require only hourly average wind patterns. The long-term version does not require detailed hourly data, and instead can approximate using average annual meteorological information. Consequently, the resulting concentrations of the long-term version of ISC3 are only annual and cannot hourly estimates. AMS/EPA Regulatory Model (AERMOD). The United States Environmental Protection Agency (EPA), in collaboration with the American Meteorological Society (AMS), developed AERMOD, the regulatory model now used in the United States. Detailed information on this model can be found in the EPA’s Technology Transfer Network Support Center for Regulatory Atmospheric Modeling (EPA 2009). AERMOD is used to estimate concentrations of criteria pollutants within 50 km of the emissions source. The model incorporates planetary boundary layer theory; turbulence structure 11 It is beyond the scope of this study to describe the three models in detail; however, it is useful to understand the main computer models involved in estimating the concentration used in the damage function. 81 and scaling concepts, including treatment of both surface and elevated sources; and both simple and complex terrain. To acquire a more complete understanding of dispersion patterns in densely populated areas, AERMOD was used for emissions sources in only two critical zones, as defined in the 1994 SEMARNAT Regulation NOM-085.12 AERMOD consists of three sub-models: AERMOD Mapping (AERMAP) preprocessor, AERMOD Meteorological (AERMET) preprocessor, and AERMOD. AERMAP incorporates complex terrain information using digital elevation data of the U.S. Geological Survey. The digital elevation model (DEM) for Mexico was used, along with land-cover, geo-referenced grids for the study areas. Land-use classes required by AERMAP were specified, using 2000 data from the National Institute of Statistics and Geography (INEGI),13 as well as the EPA’s user’s guide for AERMAP. AERMET requires in-situ meteorological data from a reference station and upper-air information.vi For sites where information was unavailable, simpler modeling techniques were used. The preprocessor requires land- use information to estimate albedo and surface stiffness. Land-use cover data of the country was used to select appropriate parameters, referring to the EPA’s AERMET user’s guide. Windrose Trajectory Model (WTM). The WTM was developed by Alfred Trukenmüller under the ExternE project to estimate energy externalities in the European Union (Trukenmüller and Friedrich 1995). WTM is a Lagrangian, climatologic, receptor-oriented model with a windrose trajectory of 24 sectors (15º each). The mixing layer has a height of 800 m. The dispersion is modeled using a long-term process description of emissions, dispersion, physical and chemical transformations, and deposition. Each cell is computed as the average of the 24 trajectories that arrive at the same cell, weighted by the wind frequencies in each sector. Each trajectory is followed by 96 hours before arriving at the end cell. The process considers all sources that affect a single receptor; this is key to capturing reactions derived from secondary pollutants, such as sulfates and nitrates. The WTM requires the following information:  Forty-nine values representing probabilities of wind trajectories in 24 possible directions of 15º each and 24 values of mean wind speed (m per s) in each direction and a mean value of wind speed for each cell.  Average annual precipitation.  Low and high emissions of NOx and SO2 (high emissions are from stacks of at least 100 m). 12 SEMARNAT Official Regulation NOM-085 (1994) on Atmospheric Contamination indicates that, for fixed fossil-fuel sources (solid, liquid, gas, or any combination thereof), maximum allowable levels of emissions should be established for smoke, total suspended particulates, SO2, and NOx; requirements and conditions should be set for operating units heated indirectly by combustion, while maximum allowable SO2 emission levels should be set for units heated directly by combustion. 13 Shadings in digital elevation models are at a scale of 1: 250,000. 82  Emissions of ammonia (NH3).  Elevation of each cell (center).  Total population. Model Limitations The model used in the simplified methodology is no longer considered state of the art in the U.S., having been superseded by AERMOD in 2005, at least for the regulatory analysis. AERMOD is better than ISC3 Short Term (ST), particularly for simulating plume growth rates and interaction of the mixing layer. It also includes improved algorithms for simulating downwash. Both AERMOD and WTM consider one structure for the vertical layer of the atmosphere that extends from the ground level to the planetary boundary layer. As mentioned previously, the WTM is a climatologic model, meaning that all data is annual, with cells of 55 x 55 km. The meteorological data is obtained from large international datasets. AERMOD does not consider the formation of secondary species; thus, it is not possible to use this model to evaluate the impact of sulfates and nitrates in the local domain. Regional models tend to overestimate secondary pollutants in the vicinity of the emissions source. WTM is limited in that it provides information only at selected heights and does not allow for modeling ozone and other such pollutants, which require a photochemical module. While AERMOD is the preferred model for local pollution analysis, other models are preferred for simulating regional pollution dispersion. Using the simplified methodology, SEMARNAT was able to study 13 power plants, which at the time represented 77 percent of the power sector’s total SO2 emissions and 67 percent of its NOX and PM emissions. Only two critical areas were studied using the detailed dispersion models for local and regional impact assessment. Risk Assessment and Impact Quantification To demonstrate the damage relationships between pollution concentrations in the environment and expected impacts, the integrated assessment method relies on the existing available health and environmental impact data from epidemiologists and environmental health experts. After reviewing the existing literature on health and ecosystems impacts, the ExternE project provided an updated summary on the potential environmental impacts from energy generation (European Commission 2005) (Table C.1). 83 Table C.1. External Costs of Energy Impacts by Category and Pollutant Impact category Pollutant/burden Effect Human health Mortality PM10, PM2.5, SO2, O3 Reduction in life expectancy due to short- and long-term exposure Heavy metal (HM), benzene, benzo-[a]-pyrene, 1,3-butadiene, Reduction in life expectancy due to short and long time exposure diesel particles, radionuclides Accident risk Fatality risk from traffic and workplace accidents Noise Reduction in life expectancy due to long-term exposure Morbidity PM10, PM2.5, O3, SO2 Respiratory hospital admissions PM10, PM2.5, O3 Restricted activity days PM10, PM2.5, CO Congestive heart failure Benzene, benzo-[a]-pyrene, 1,3- Cancer risk (non-fatal) butadiene, diesel particles, Osteroporosis, ataxia, and renal dysfunction radionuclides, heavy metal (HM) Cerebrovascular hospital admissions, chronic bronchitis, chronic cough in children, cough PM10, PM2.5 in asthmatics, and lower respiratory symptoms Mercury Loss of IQ in children O3 Asthma attacks and symptom days Noise Myocardial infarction, angina pectoris, hypertension, and sleep disturbance Accident risk Risk of injuries from traffic and workplace accidents Building materials Aging of galvanized steel, limestone, mortar, sandstone, paint, rendering, and zinc for SO2, acid deposition utilitarian buildings Combustion particles Soiling of buildings Crops NOx, SO2 Yield change for wheat, barley, rye, oats, potato, and sugar beet O3 Yield change for wheat, barley, rye, oats, potato, rice, tobacco, and sunflower seed Acid deposition Increased need for liming N, S deposition Fertilizing effects Global warming Worldwide effects of temperature change and sea-level rise on mortality, morbidity, CO2, CH4, N2O coastal areas, agriculture, energy demand, and economy Amenity losses Noise Amenity losses due to noise exposure Ecosystems and land-use change Acid deposition, nitrogen Acidity and eutrophication and PDF of species deposition, SO2, NOx, NH3 Source: ExternE project 2005. Note: PM10 and PM 2.5 refer to particulate matter with respective aerodynamic diameters of < 10 µm and < 2.5 µm, including secondary particles (sulphate and nitrate aerosols). 84 One key element of the damage-cost function approach is the Exposure Response Function (ERF) (Figure 30). Health damages, which constitute more than 90 percent of the total impact of environmental pollution,vii have been estimated in several international studies (European Commission 2005; Muller and Mendelsohn 2007; EPA 1999). Because of the dispersion models used, it is also necessary to focus on the impacts of PM10, PM2.5, SO2, NOx, sulfates, and nitrates. Unfortunately, the models implemented for Mexico to date have not been able to estimate ozone; thus, this pollutant must be omitted from the analysis for the time being. It should be highlighted that including the impacts of ozone would certainly increase external costs. Figure 30. Exposure Response Function Source: European Commission 2005. The ERF relates the ambient concentration of the pollutant to the resulting impact on a receptor or at-risk subgroup (e.g., adults, children or crops). For health impacts, ERFs are straight lines with no threshold; that is, the line passes through the origin at zero concentration. The health impact is estimated using the following equation:14 Ii = j SERFi * Cj * POBj [1] where Ii is the impact in receptor I (number of cases), SERFi is the slope of the ERF (person- per m3), ∆Cj, is the incremental concentration in location j ( 3 ), and POBj is the at-risk population in location j (number of persons). The slope of the ERF is estimated as follows: 14 J. Spadaro, Airpacts Input Data: Exposure Response Function, 2001. 85 SERF = IRR * baseline = IRR* incidence * fPOP [2] where IRR (increased risk ratio) is the percentage change in the rate of occurrence of a particular disease in the at-risk population relative to its nominal rate of occurrence per unit change in ambient concentration (percent change per μg per m3), baseline is the number of cases per year per person, including the incidence rate (annual cases per receptor [e.g., adult or child]) and fPOP (fraction [percent] of the at-risk population). The following assumptions in the ExternE 2005 update were used to evaluate the impact of sulfates and nitrates:  Toxicity of sulfates = toxicity of PM10 (SERF sulfates = SERF PM10),  Toxicity of nitrates = ½ of PM10 (SERF nitrates = 0.5 SERF PM10). Valuation of Externalities To estimate the total cost of damage in dollars, the associated impacts are monetized. Externalities, by definition, are the costs or benefits not internalized in market prices, resulting from incomplete information on external costs, among other reasons. External costs can be associated with market or non-market values. For example, air pollution can require asthmatics to increase their purchase of medications and health services. These costs are market prices that can be estimated by applying commonly-known costing techniques. Since the costs are not taken into account by the pollution sources, they remain external costs. External costs can also have non-market values. For example, at- risk adult populations with long-term exposure to PM10 may have decreased life expectancy. Valuation techniques are needed to capture the value society is willing to incur in order to protect the population from such exposure and its risks. Overview and Current Status Extensive research has been conducted, particularly in developed countries, to capture these values. Valuing the environment is an inherently complex endeavor, requiring decision-making that might be questioned on ethical grounds. Economists usually acknowledge the limitations of these techniques to fully capture the value of the environment, and try to limit the scope of the valuation studies to the economic value by applying techniques that can provide shadow prices (i.e., those from other market goods that serve as a proxy for the price of the non-market good or service) or conducting surveys on consumers’ willingness to pay (WTP) for a specific environmental good or service (i.e., assessing the consumer surplus). The underlying assumption in these calculations is that people are able to determine the trade-off between goods that will leave their welfare level unchanged. Clearly, there are other non- use values from a clean environment, including both existence value and pure non-use value, as well as bequest value, option value, and value arising from paternalistic altruism (Foster 1997). Much remaining work is needed to improve valuation techniques and incorporate these values into a comprehensive policy evaluation procedure. Cost-benefit analysis should be part of an iterative 86 process that allows policy makers to consider all available information at a given point in time and improve decision-making as more data becomes available. The political process should use the best economic information available on the costs of policy, including external costs, as well as non- monetary, non-use and stewardship values, which are non-marketable commodities currently not captured in externality studies. Once the impact has been estimated, external costs are estimated by multiplying each case by its unitary cost, expressed as the following equation: Di = Ii · UCi [1] where Di is the damage cost of impact I, Ii is health impact i estimated in the previous step (cases per year), UCi is the unit cost of impact i (dollars per year [e.g., cost of an asthma attack or emergency hospital visit]). Monetizing Impacts This section briefly describes the main cost components included in the externalities presented. For this study, Mexican institutions used McKinley et al. (2003) as the main reference on cost valuation techniques, which they applied to health impacts in the country, using the following definitions:  Cost of Illness (COI). COI studies measure the economic burden of disease. The COI for morbidity outcomes can be understood as direct costs of an illness and can include expenditures on medication, doctor visits, hospitalization, laboratory tests, and human resources. This metric aims to include all incurred costs for an illness or medical attention. The costs can be paid directly by the ill individual, public or private insurance, and/or general taxation (McKinley et al. 2003).  Productivity Loss (PL). PL uses the difference in output (production) due to illnesses as the basis of valuing costs. This value can also be interpreted as the cost or value of the time an individual loses from being hospitalized, bedridden, or suffering premature mortality. PL can also be calculated for environmental contingency episodes, when industries must temporarily close their operations due to air pollution levels. To estimate PL, days lost per case are multiplied by the average daily wage. It should be noted that the estimated number of days is usually limited to days spent in the hospital. It is reasonable to assume that PL would be increased as a result of individuals missing more work days beyond their hospital stay. PL from premature mortality can be estimated using X amount of days lost. For example, McKinley et al. (2003) uses 260 days lost per premature mortality, applying the Years of Life Lost (YOLL) approach; however, in this study, we use the Value of Statistical Life (VSL) approach.  Willingness To Pay (WTP). WTP can be determined from contingent valuation and compensating wage studies, which in theory should account for the full cost of disease to an individual, including pain and suffering. Contingent valuation studies rely on an individual’s stated preferences through the use of surveys, in which individuals are asked how much they 87 would be willing to pay to reduce their risk of mortality or morbidity. Hedonic wage studies rely on revealed preferences through analysis of labor market data on wages and risk levels for certain jobs. Econometric models are used to determine the amount individuals are compensated for added risk in the workplace. These methods for estimating individual preferences rely on the concept of consumer sovereignty and people’s ability to make rational trade-offs. Traditionally, mortality impact valuation has been done using one of three approaches: (i) contingent valuation, (ii) hedonic wages, or (iii) human capital. In principle, contingent valuation would be better applied to environmental impacts since, ideally, individuals would be asked to value a particular situation considering environmental degradation. The hedonic wages approach is usually conducted using shadow prices derived from differentials in compensation for certain jobs where mortality risk is higher. The human capital approach considers the wealth an individual would have earned during the disability period or that was lost due to premature mortality. The first two approaches derive a VSL that equals the value of avoiding the risk of one fatal accident or premature mortality due to pollution exposure. Air pollution has both acute and chronic effects on mortality. The former, resulting from short- term exposure, can be statistically observed within days; while the latter, resulting from prolonged exposure (e.g., lifetime city dwellers), may only be apparent in the long term. In the case of chronic or premature mortality, an individual loses years of life due to contamination. The ExternE approach values only those YOLL and not the complete VSL per case. The reasoning is that air pollution does not kill you instantly and thus only the proportional part of YOLL should be considered in the analysis of externalities. Using the European population data, the YOLL for chronic mortality are assumed to be 10 years, while acute mortality is assumed to be 0.5 years of life. The equation for estimating the YOLL from the VSL is as follows: æ ö = ç ç1 å (1 + )- ÷ ÷ [1] è = ø where, a is the age of the cohort, aPi is the probability that the cohort of age a survives to age i (I is the average life expectancy of the population in the specific age group), Tl is the maximum life expectancy, and dr is the discount rate. Furthermore, 70 =å [2] =1 (1 + ) -1 where YOLLi is years of life lost in year i due to pollution exposure and YOLLtot is the total number of years lost in the population over a 70-year period. 88 Annex D. Project Valuation with Environmental Externalities Case Study 1: Combined Cycle Technology Mexico’s National Electricity System expects increased power demand in the Northwest to average 3.9 percent a year over the 2010–25 period. In accordance with long-term planning, CFE estimates that an additional 650 MW will be needed to meet this demand and maintain adequate levels of reserve margin. To create this new capacity, a combined cycle (CC) technology project was submitted, which included a study for avoided emissions (Table D.1). Table D.1. Basic Parameters of Combined Cycle Project Parameter Value Unit Capacity 655.7 MW Peak hours 3.32 h Failure rate 0 % Days for Maintenance 20 day Capacity Factor 70 % hr/yr Own consumption 3 % Power plant Cost 595.11 Million dollars Related infrastructure 6 Million dollars Levelized marginal cost 62 usd/MWh Marginal cost of capacity 158 usd/MWh Source: Data scaled from CFE sample project. With this data, generation of the plant’s fuel input was estimated. CFE’s planning department estimated the displaced generation, identifying specific units in the system. Analysis of economic dispatch of the system determined the avoided emissions per plant resulting from the new facility’s displacement of fuel-oil generation units (See Table D.2). Total emissions avoided are presented shown in Table D.3. 89 Table D.2 Displaced Generation from Combined Cycle Project Net Plant Net Plant Net Net Plant ID Technology Technology Technology Plant ID Technology Generation ID Generation ID Generation Generation 1 CC 978 23 CT 696 45 CC 100 67 CT 1992 2 GT 43 24 GT 81 46 CT 2 68 CC -489 3 GT 62 25 CC 27 47 Coal 701 69 CT 1575 4 GT 31 26 GT -264 48 CC 120 70 CT 1066 5 GT 51 27 GT -2 49 CT -8 71 CC 4500 6 GT 32 28 GT 0 50 CC 193 72 GT 182 7 CT 37 29 CT -85 51 CT 37 73 CT 412 8 CC 102 30 CC 34 52 GT 16 74 GT -1 9 GT 31 31 CC 47 53 GT 26 75 GT 268 10 GT 27 32 CC -9 54 GT 13 76 GT 9 11 GT 34 33 CC 56 55 GT 5 77 CC 7 12 GT 1 34 CT 128 56 CC 221 78 CC 97 13 CC 456 35 CT 7 57 CT 14 79 GT 1 14 CT 5 36 CC 23 58 CT 119 80 CT 187 15 GT -31 37 CT 386 59 CT 38 81 CT -27 16 CT 1141 38 CC 80 60 CT 33 82 CT 1205 17 CC 491 39 CC 14 61 CC -13 83 GT 49 18 CT 26812 40 GT 2 62 GT 12 84 CC -90839 19 CT 125 41 CT 194 63 CC 178 85 CT 37929 20 CT 17 42 CC 11 64 CC 4 86 GT 36 21 CT 361 43 CC 115 65 GT 439 87 CT 5810 22 GT 1 44 GT 37 66 GT 25 Source: Data scaled from CFE sample project. 87 plants were impacted by the new project (Actual Plant ID has been replace with a number) Note: Technology refers to combined cycle (CC), gas turbine (GT), conventional thermal generation (CT), coal power plant (coal). Table D.3. Total Avoided Emissions from Combined Cycle Project Emissions reduced by the project (ton) SO2 469 NOx 2268 PST 8 CO2 202745 Source: CFE sample project, using scaled data The cost benefit analysis summary is shown in Table D.4. The details of the calculation are shown in Table D.5. Table D.4 Summary of BCA for Combined Cycle Project Summary BCS PV Investment 529452 PV Environmental benefits 29586 PV reduced production costs 627429 PV Reduced NSE 137905 SUM benefits 794920 NPV 265468 BC ratio 1.5 Discount rate 12% Source: CFE sample project, using scaled data Table D.5. Detailed Cost-Benefit Analysis for Combined Cycle Project Benefits Year Investment Cost Security of Power Environmental Savings in production Supply (reduced Total benefits externalities costs APC NSE) 2014 56114 2015 382292 2016 234826 2017 11715 4167 62990 8795 75951 2018 39870 81566 13297 134732 2019 4211 82018 13083 99312 2020 3479 82358 12801 98638 2021 3674 82524 12804 99002 2022 3840 83378 12917 100135 2023 4621 85338 13033 102992 2024 6042 90878 13150 110070 2025 3752 77046 13267 94065 2026 5314 75938 13208 94459 2027 5480 73452 27737 151488 91 Benefits Year Investment Cost Security of Power Environmental Savings in production Supply (reduced Total benefits externalities costs APC NSE) 2028 2924 73452 17258 93634 2029 2924 73452 29381 105757 2030 2924 73452 29381 105757 2031 2924 73452 29381 105757 2032 2924 73452 29381 105757 2033 2924 73452 29381 105757 2034 2924 73452 29381 105757 2035 2924 73452 29381 105757 2036 2924 73452 29381 105757 2037 2924 73452 29381 105757 2038 2924 73452 29381 105757 2039 2924 73452 29381 105757 2040 2924 73452 29381 105757 2041 2924 73452 29381 105757 2042 2924 73452 29381 105757 2043 2924 73452 29381 105757 2044 2924 73452 29381 105757 2045 2924 73452 29381 105757 2046 2924 73452 29381 105757 2047 2924 73452 29381 105757 PV 529452 29586 627429 137905 794920 Source: CFE sample project, using scaled data Case Study 2: Geothermal Installation In Mexico’s Southwest, the National Electricity System expects power demand to grow by 3.9 percent a year until 2024 (Map 3). The geothermal project, which will add 50 MW of average gross annual capacity, is part of SENER’s Special Program for Renewable Energy, whose primary objective is diversification of power generation sources (Table D.6). Table D.6. Basic Parameters of Geothermal Project Parameter Value Unit Total capacity 87.5 MW Fuel Earth steam Capacity factor 86.5 % hr/yr Own uses 6.54 % Years 30 Years Fix O&M 13.93 $ / MWh Variable O&M 0.04 $ / MWh Levelized marginal 75 $ / MWh cost Source: Data scaled from CFE sample project. The socioeconomic evaluation of the project included an analysis of avoided emissions and a financial evaluation, which included revenue from carbon credits, valued at US$9.939 per tCO2 for 8 years of the facility’s 30 years of operation (Table D.7). The analysis of displaced generation and incremental energy is presented in Table D.9. 92 Table D.7 Income flows for Geothermal Project Energy + Income for Income for Income Carbon Year Generation Marginal Cost Capacity Total income Energy Capacity Credits income GWh $ /MWh Million $ 2014 494 65 32 16 48 3 51 2015 659 67 45 16 60 4 64 2016 659 68 45 16 61 4 64 2017 659 73 48 16 64 4 67 2018 659 73 49 16 64 4 68 2019 659 76 50 16 66 4 69 2020 659 77 51 16 66 4 70 2021 659 77 51 16 67 0 67 2022 659 76 50 16 66 0 66 2023 659 77 51 16 66 0 66 2024 659 77 51 16 67 0 67 2025 659 78 51 16 67 0 67 2026 659 78 52 16 67 0 67 2027 659 79 53 16 68 0 68 2028 659 80 53 16 69 0 69 2029 659 81 54 16 69 0 69 2030 659 82 54 16 70 0 70 2031 659 82 54 16 70 0 70 2032 659 82 54 16 70 0 70 2033 659 82 54 16 70 0 70 2034 659 82 54 16 70 0 70 2035 659 82 54 16 70 0 70 2036 659 82 54 16 70 0 70 2037 659 82 54 16 70 0 70 2038 659 82 54 16 70 0 70 2039 659 82 54 16 70 0 70 2040 659 82 54 16 70 0 70 2041 659 82 54 16 70 0 70 2042 659 82 54 16 70 0 70 2043 659 82 54 16 70 0 70 2044 165 82 14 16 29 0 29 NPV 352 114 466 13 479 Source: Data scaled from CFE sample project. 93 Economic dispatch modeling with and without the project shows that avoided emissions from the new geothermal facility over the 2014–43 period would average 846 tons for SO2, 255 tons for NOx, 58 tons for TSP, and 177,499 tons for CO2 (Table D.8). Economic evaluation of the project includes the externality cost, using SEMARNAT values, and CO2 costs valued at US$18.35 per ton. Most of the project benefits would consist of energy savings, while the remainder would accrue to the environment and system security. Table D.8. Total Avoided Emissions from Geothermal Project Emissions reduced by the project (ton) SO2 44,392 NOx 13,410 PST 3,029 CO2 9,318,691 Source: Data scaled from CFE sample project. Table D.9. Displaced Generation from Geothermal Project Net Net Net Plant ID Technology Plant ID Technology Plant ID Technology Generation Generation Generation 1 CC -4 23 CT 767 45 GT 11 2 CC 33 24 CT 278 46 GT -19903 3 CT 68 25 CC 266 47 CT 93 4 CC 109 26 CC 54 48 GT 5938 5 CT 634 27 CT 0 49 CC 936 6 CC 343 28 CT 142 50 GT 58 7 CT 2172 29 CT -19 51 CT 5 8 CT 7 30 CT 63 52 CT 1439 9 CC 196 31 CT 37 53 CT 84 10 CT 186 32 CT 9 54 CT 142 11 GT 4 33 CC 310 55 CT 376 12 GT 0 34 CT 2 56 CT 63 13 GT 7 35 CT 77 57 GT 553 14 GT 33 36 CT 352 58 GT 219 15 GT 16 37 CT 324 59 GT 86 16 CC 89 38 GT 32 60 CT 376 17 CC 105 39 CC 175 61 CC 221 18 CC 21 40 CT 116 62 CT 1115 19 CC 96 41 CT 21 63 CT 289 20 CT 23 42 CC 569 64 CT -201 94 Net Net Net Plant ID Technology Plant ID Technology Plant ID Technology Generation Generation Generation 21 CT 446 43 GT 12 65 CT 19 22 CT 287 44 CT 604 Source: Data scaled from CFE sample project. 65 plants were impacted by the new project (Actual Plant ID has been replace with a number) Note: Technology refers to combined cycle (CC), gas turbine (GT), conventional thermal generation (CT). Total environmental benefits were estimated at US$23 million, production costs savings at $124 million, and non-served energy at about $10 million. Total investment costs were estimated at about $69 million. The benefit-to-cost ratio without externalities was 2.35. Table D.10. Summary of BCA for Geothermal Project Summary BCS PV Investment 69,383 PV Environmental benefits 23,666 PV reduced production costs 124,295 PV Reduced NSE 10,650 SUM benefits 227,994 NPV 158,611 BC ratio 2.3 Discount rate 12% Source: CFE sample project, using scaled data Table D.11. Detailed Cost-Benefit Analysis for Geothermal Project Benefits Security of Investment Savings in Power Year Environmental Total Cost production Supply externalities benefits costs APC (reduced NSE) 2014 1877 2015 37146 2016 51499 2017 2265 5792 9347 2018 5888 13422 2019 4737 14590 2020 4819 16589 1981 2021 3914 16916 1941 2022 3898 17393 1864 2023 3093 16349 2087 2024 810 15194 2338 2025 745 15272 2182 2026 659 15639 1955 2027 576 15326 1189 2028 572 17165 2452 95 Benefits Security of Investment Savings in Power Year Environmental Total Cost production Supply externalities benefits costs APC (reduced NSE) 2029 947 19225 2030 458 17226 2031 382 17226 2032 346 17226 2033 309 17226 2034 276 17226 2035 246 17226 2036 220 17226 2037 196 17226 2038 175 17226 2039 157 17226 2040 140 17226 2041 125 17226 2042 111 17226 2043 99 17226 2044 89 17226 2045 79 17226 2046 71 17226 2047 63 17226 PV 69383 23666 124295 10650 158611 Source: CFE sample project, using scaled data Case Study 3: Wind Farm Mexico’s National Electricity System estimates demand growth for the Southeast region at 3.5 percent a year over the planning period. The 395 MW wind farm project for this region is part of the SENER’s long-term energy-diversification strategy. The project investment includes the cost of 303 wind turbines, with 1,065.92 GWh in annual production, and assumes a 40 percent capacity factor (Table D.12). Table D.12. Basic Parameters of Wind Project Parameter Value Unit Total capacity 395.2 MW Fuel Wind Capacity factor 40 % hr/yr Years 20 Years Fix O&M 33,662 $ / MWh Variable O&M 1.13 $ / MWh Levelized marginal 70.13 $ / MWh cost Source: CFE sample project, using scaled data 96 The estimated external benefits from reducing CO2 emissions are based on the avoided fossil- fuel emissions that would otherwise be produced without the displaced generation. The needed carbon-credit incentive is US¢12.52 per kWh over the useful life of the central plant facility, corresponding to US$18.35 per tCO2 avoided. The project will generate 1,065.92 GWh of energy per year on average, allowing for the capture of $18.510 million in revenue from carbon credits. In current value, these incentives represent a total of $82.1 million. The total benefits from incorporating production cost savings and carbon credits (without taking other environmental externalities into account) have a current value equal to the required project investment.15 The resulting annual avoided emissions for major pollutants average 3,483 tons for SO2, 1,275 tons for NOx, 191 tons for TSP, and 607,426 tons for CO2 (Table D.14).16 The externality benefits from avoided environmental costs equal US$7.2 million a year17. 15 No World Bank incentives are considered. 16 The external costs by zone and pollutant average about US¢0.604 per kWh (2009 figure). 17 It is important to stress the need for carbón-credit incentives, as well as avoided environmental cost, to ensure project viability. 97 Table D.13. Displaced Generation from Wind Project Net Net Net Net Plant ID Technology Plant ID Technology Plant ID Technology Plant ID Technology Generation Generation Generation Generation 1 CC -3075 23 CC 445 45 CT 2141 71 CT 16 2 GT -1 25 CT -1 46 CT -182 72 GT 1 4 CC -148 26 CC 313 47 CT 1348 73 CT -31 5 CC -714 27 GT -3 48 GT 1 74 CC 1438 6 CC 781 28 CT 823 49 CC 334 75 CC 1756 7 CC -39 29 CT 663 51 GT -326 76 CT -14 8 CT 48 30 CT 98 52 GT -12 77 CT -173 9 CC 589 31 CT 486 53 GT 8 78 CT 189 10 GT 476 32 CT 7 54 GT 252 79 CT 1 11 CC 177 33 CC 1515 56 GT -3 80 CT 796 12 GT 456 34 CT 1937 59 GT 1 81 CI -257 13 CT 159 35 CC 679 60 GT -469 82 CT -449 14 CT 205 36 CT 151 61 CT -407 83 CC -18 15 CT 1282 37 CI 1338 62 CT 12 84 CT 621 16 CT -23 38 CC 47 63 CC -267 85 CC 5 17 CT -27 39 CT 1438 64 CC 394 86 CC 293 18 CT 2939 40 CC 1530 65 CC 152 87 CC 1 19 CT -400 41 CT 213 66 COAL 528 88 GT 3 20 CT -122 42 CT 4606 67 COAL -209 89 GT 1 21 CT 1456 43 GT 7 69 CT 976 90 GT 523 22 CT 5 44 CT -1973 70 CI -198 Source: Data scaled from CFE sample project. 87 plants were impacted by the new project (Actual Plant ID has been replace with a number) Note: Technology refers to combined cycle (CC), gas turbine (GT), conventional thermal generation (CT), coal power plant (coal). 98 Table D.14. Total Avoided Emissions from Wind Project Emissions reduced by the project (ton) SO2 4,528 NOx 1,658 PST 248 CO2 789,654 Source: CFE sample project, using scaled data Table D.15. Detailed Cost-Benefit Analysis for Wind Project Benefits Security Investment Savings in of Power Year Environmental Total Cost production Supply externalities benefits costs APC (reduced NSE) 2014 97463.6 2015 541552.7 2016 148270.2 2017 18047.9 43395.3 0 61443.2 2018 24063 57905.9 0 81968.9 2019 24063 81424.2 0 105487.2 2020 24063 72434.7 0 96497.7 2021 24063 65994.5 0 90057.5 2022 13015.6 94351.4 0 107367 2023 9334 69642.3 0 78976.3 2024 9334 98014.8 0 107348.8 2025 9334 59813 0 69147 2026 9334 90764.7 0 100098.7 2027 9334 106850.9 0 116184.9 2028 9334 76220.3 0 85554.3 2029 9334 76220.3 0 85554.3 2030 9334 76220.3 0 85554.3 2031 9334 76220.3 0 85554.3 2032 9334 76220.3 0 85554.3 2033 9334 76220.3 0 85554.3 PV $624,279.37 116044.5 516831.9 0 632876.4 Source: CFE sample project, using scaled data 99 Case Study 3: Hydroelectric facility Mexico’s National Electricity System estimates demand growth for the Southeast region at 3.5 percent a year over the planning period (Map 2). The project capacity of 1107 MW requires an investment of 800 million usd; it is expected to generate annually 1370 GWh, 900 of which are expected to serve as peak capacity. The basic information of the project is shown in Table D.16. Table D.16. Basic Parameters of Wind Project Parameter Value Unit Total capacity 1107 MW Energy source Water Capacity factor 17% % hr/yr Years 50 Years Water specific 2.58 m3 / year consumption USD 2010 /1000 Water payment 0.2958 m3 Source: CFE sample project, using scaled data The marginal cost of energy considered in the evaluation was 74.74 dollars per MWh. For the financial assessment of the project a $9.9 per ton/CO2 were considered. The analysis of displaced generation is presented in Table D.17. The total emissions reductions economic dispatch simulation were 6520 ton SO2, 1779 ton of NOx, 431 ton of particulate matter and 860,490 ton of CO2 (See Table D.18). The summary of the BCA is shown in Table 18 and the details of the calculation in Table 19. Table D.17. Displaced Generation from Hydro Project Net Net Net Plant ID Technology Plant ID Technology Plant ID Technology Generation Generation Generation 1 CC 82 24 CC 490 47 CT -207 2 GT 50 25 CT 218 48 CT 5370 3 GT 6 26 GT 4314 49 CT 834 4 GT 82 27 CT 101 50 CC 133 5 CT 369 28 CC 155 51 CT 289 6 GT 87 29 CT 6 52 CC 21830 7 GT 177 30 CT 216 53 CC 228 8 GT 87 31 CT 1892 54 CT 14 9 GT 228 32 CT -125 55 CT 12 100 Net Net Net Plant ID Technology Plant ID Technology Plant ID Technology Generation Generation Generation 10 GT 123 33 CT 400 56 CT -629 11 TC 3467 34 CT -1 57 GT 59 12 CC 887 35 CT 293 58 GT 1301 13 CT 10 36 CT 2 59 CT 934 14 CT 23 37 CC 232 60 CT 32 15 CT 33 38 CT 1049 61 CT 172 16 CC 2517 39 CC 44 62 GT 550 17 CC 2919 40 GT 2684 63 CT 6328 18 CT 276 41 CT 176 64 CT -239 19 CC 188 42 CT 4043 65 CT 1496 20 CT 266 43 CT 10000 66 GT 5525 21 CT 336 44 CT 100 67 CT 727 22 CT 1034 45 CT 476 68 GT 240 23 CC -595 46 CT 435 69 HYDRO -84396 Source: Data scaled from CFE sample project. 87 plants were impacted by the new project (Actual Plant ID has been replace with a number) Note: Technology refers to combined cycle (CC), gas turbine (GT), conventional thermal generation (CT), coal power plant (coal). Table D.18. Total Avoided Emissions from Wind Project Emissions reduced by the project (ton) SO2 6520 NOx 1779 PST 431 CO2 860,490 Source: CFE sample project, using scaled data Table D.19. Detailed Cost-Benefit Analysis for Hydro Project Benefits Investment Savings in Security of Year Environmental Cost production Power Supply Total benefits externalities costs APC (reduced NSE) 2014 4954 2015 53478 48387 48387 2016 179017 588 588 2017 375868 -31094 -31094 2018 469586 -9223 -9223 2019 308974 -9223 -9223 2020 103506 22753 49215 71967 2021 10257 127908 99824 237989 2022 10441 126544 145397 282382 101 Benefits Investment Savings in Security of Year Environmental Cost production Power Supply Total benefits externalities costs APC (reduced NSE) 2023 18852 134473 140607 293933 2024 21365 127931 135975 285271 2025 59205 114530 98655 272390 2026 43397 115392 99765 258555 2027 25939 111207 109504 246651 2028 27254 125925 72253 225432 2029 27502 141035 61489 230026 2030 26203 130530 26181 182913 2031 26553 124391 35943 186887 2032 58832 124391 31781 215004 2033 43628 124391 31781 199800 2034 18498 124391 31781 174670 2035 19567 124391 31781 175739 2036 20145 124391 31781 176317 2037 20138 124391 31781 176309 2038 42915 124391 31781 199087 2039 31874 124391 31781 188046 2040 36773 124391 31781 192945 2041 36773 124391 31781 192945 PV 883,160 125306 889841 669332 969330 Source: CFE sample project, using scaled data 102 References Aden, N. T., and J. E. Sinton. 2006. “Environmental Implications of Energy Policy in China.” Environmental Politics 15: 248–70. Alcamo, J., P. Mayerhofer, R. Guardans, T. Van Harmelen, J. Van Minnen, J. Onigkeit, M. Posch, and B. de Vries. 2002. “An Integrated Assessment of Regional Air Pollution and Climate Change in Europe: Findings of the Air-Clim Project.” Environmental Science and Policy 5(4): 257–72. Amann, M., M. Johansson, A. Lükewille, W. Schöpp, H. ApSimon, R. Warren, T. Gonzales, L. Tarrason, S. Tsyro. 2001. “An Integrated Assessment Model for Fine Particulate Matter in Europe.” Water, Air, and Soil Pollution 130: 223–8. Barratt, R. 2001. Atmospheric Dispersion Modelling: An Introduction to Practical Applications. London and Sterling, VA: Earthscan. CFE (Federal Electricity Commission). 2012. Programa de Obras e Inversiones del Sector Eléctrico 2012-2026. Mexico City: CFE. ———.2005. “Planificación del Sector Eléctrico Nacional” PowerPoint presentation. In Seminario sobre Evaluación de Proyectos en el Sector Eléctrico. Morelia, Michoacán: CFE, SHCP, LyFC. Dasgupta, Partha. 2006. Comments on the Stern Review’s Economics of Climate Change. Cambridge, UK: University of Cambridge. Égert, B. (2011), “France's Environmental Policies: Internalizing Global and Local Externalities”, OECD Economics Department Working Papers, No. 859, Paris: OECD Publishing. EIA-DOE, 1995. Electricity Generation and Environmental Externalities: Case Studies. 1995. Washington, DC: Energy Information Administration-US Department of Energy. EPA (United States Environmental Protection Agency). 1999. The Benefits and Costs of the Clean Air Act: 1990–2010. EPA Report to Congress. Washington, DC: EPA. ———. 2009. “Technology Transfer Network Support Center for Regulatory Atmospheric Modeling” (http://www.epa.gov/scram001/dispersion_prefrec.htm - aermod). ESMAP (Energy Sector Management Assistance Program). 2010. Johnson, T., C. Alatorre, Z. Romo, F. Liu. Low-Carbon Development in Mexico. Washington, DC: World Bank. European Commission. 2005. ExternE. Externalities of Energy: Methodology Update 2005. Luxembourg: Office for Official Publications of the European Communities. 103 Foster, J. 1997. Valuing Nature?: Economics, Ethics, and Environment. London: Routledge. Galindo, L. M. 2010. The Economics of Climate Change in Mexico. Mexico City: SEMARNAT and SHCP. Hadlock, C. 1988. Mathematical Modeling in the Environment. Washington, DC: Mathematical Association of America. Hammitt, J. K., and M. E. Ibarrarán. 2006. “The Economic Value of Fatal and Non-Fatal Occuptational Risks in Mexico City Using Actuarial- and Perceived-Risk Estimates.” Health Economic Letters 15: 1329–35. INE (National Institute of Ecology). 2006. Inventario Nacional de Emisiones de México, 1999. Mexico City: SEMARNAT-INE. INE (National Institute of Ecology). 2003. Evaluación del impacto de las emisiones de la termoeléctrica de Tuxpan en lacalidad del aire de la región, usando el sistema de modelado de dispersion CALMET-CALPUFF. Mexico City: INE INE (National Institute of Ecology). 2006. Introducción a la evaluación de los impactos de las termoeléctricas de México. Mexico City: INE INEGI (National Institute of Statistics and Geography). 2000. Marco Geoestadístico Municipal 2000. Mexico City: INEGI. ———. 2005. “Salario Mínimo General Promedio de los Estados Unidos Mexicanos: Comisión Nacional de Salarios Mínimos, y Distribución Porcentual de los Hogares por Sexo del Jefe y Rangos de Salarios Mínimos Equivalentes, 1992 a 2004.” Mexico City: INEGI. ———. 2006. Población Total por Grupos Quinquenales de Edad según Sexo, 1950 a 2005. Mexico City: INEGI. International Energy Agency. 2013. Technology Roadmap. Wind Energy. IEA/OECD. Paris: 2013. _______. 2012. World Energy Outlook 2011. Paris: 2012 _______. 2011. Technology Roadmap. Geothermal Heat and Power. Paris:2011 _______. 2010. Technology Roadmap. Solar Photovoltaic Energy. IEA/OECD. Paris: 2010. Jacobson, M. 1999. Fundamentals of Atmospheric Modeling. Cambridge and New York: Cambridge University Press. Jiang, Z. Y., J. P. Cheng, S. L. Liu, and Z. Q. Pan. 2008. “The External Cost of Coal Power Chain in China.” Meitan Xuebao/Journal of the China Coal Society 33: 1325–28. McKinley, G., M. Zuk, M. Hojer, M. Avalos, I. Gonzáles, M. Hernández, R. Iniestro, I. Laguna, M. A. Martínez, P. Osnaya, L. M. Reynales, R. Valdés, and J. Martínez. 2003. The 104 Local Benefits of Global Air Pollution Control in Mexico City. Mexico City: National Institute of Ecology (INE) and National Institute of Public Health (INSP). Mendelsohn, Robert. 2007. “A Critique of the Stern Report”. Regulation 29(4): 42–6. Moreno, A., and J. Urbina. 2008. Social Impact of Climate Change in Mexico. Mexico City: National Institute of Ecology (INE). Muller, N. Z., and R. Mendelsohn. 2007. “Measuring the Damages of Air Pollution in the United States.” Journal of Environmental Economics and Management 54(1): 1–14. National Research Council. Hidden Costs of Energy: Unpriced Consequences of Energy Production and Use. Washington, DC: The National Academies Press, 2010 Nordhaus, W. 2008. A Question of Balance: Weighing the Options on Global Warming Policies. New Haven, CT: Yale University Press. OECD (2001a). Externalities and Energy Policy: The Life Cycle Analysis Approach. Paris: Nuclear Energy Agency and OECD OECD. (2001b). Environmentally Related Taxes in OECD Countries. Paris: OECD Schöpp, W., Z. Klimont, R. Suutari, and J. Cofala. 2005. “Uncertainty Analysis of Emission Estimates in the RAINS Integrated Assessment Model.” Environmental Science and Policy 8(6): 601–13. SEMARNAT (2013). Inventario Nacional de Emisiones 2008. Sistema Nacional de Emisiones a la Atmósfera. http://sinea.semarnat.gob.mx/sinae.php? SEMARNAT. 2011. “Norma Oficial Mexicana NOM-085-SEMARNAT-2011, Contaminación atmosférica-Niveles máximos permisibles de emisión de los equipos de combustión de calentamiento indirecto y su medición”. In Diario Oficial de la Federación, 2 de febrero de 2011. Mexico City: SEGOB. SEMARNAT and ECLAC (Ministry of Environment and Natural Resources and UN Economic Commission for Latin America and the Caribbean). 2004. Evaluación de las externalidades ambientales de la generación termoeléctrica en México. Mexico City: CEPAL. ———. 2007. Evaluación de las externalidades ambientales del sector energía en las zonas críticas de Tula y Salamanca. Mexico City: CEPAL. SENER (Ministry of Energy). 2006. “Emisiones del Sector Eléctrico.” Mexico City: SENER. ———. 2008. “Prospectiva del Sector Eléctrico 2008–2017.” Mexico City: SENER. ———. 2012a. “National Energy Balance 2011.” Mexico City: SENER. ———. 2012b. “Prospectiva del Sector Eléctrico 2012–2016.” Mexico City: SENER. 105 ———. 2012c. “Metodología para valorar externalidades asociadas con la generación de electricidad en México.” In Diario Oficial de la Federación, 14 de diciembre de 2012. Mexico City: SEGOB. ———. 2013. “Estadísticas e Indicadores del Sector Eléctrico: Capacidad Efectiva de Generación” (http://www.energia.gob.mx/webSener/portal/Default.aspx?id=1430), accessed May 24. Sokhi, R. S., H. Mao, S. T. G. Srimath, S. Fan, N. Kitwiroon, L. Luhana, J. Kukkonen, M. Haakana, A. Karppinen, K. Dick Van Den Hout, P. Boulter, I. S. McCrae, S. Larssen, K. I. Gjerstad, R. San José, J. Bartzis, P. Neofytou, P. Van Den Breemer, S. Neville, A. Kousa, B. M. Cortes, and I. Myrtveit. 2008. “An Integrated Multi-Model Approach for Air Quality Assessment: Development and Evaluation of the OSCAR Air Quality Assessment System.” Environmental Modelling and Software 23: 268–81. Stern, Nicholas. 2006. Stern Review on the Economics of Climate Change. London: HM Treasury. Trukenmüller, A., C. Klamt y W. Krewitt. (2001), Revision of WTM Parameters for EcoSense China, Stuttgart, Sttutgart University, IER Turtós Carbonell, L., E. Meneses Ruiz, M. Sánchez Gacita, J. Rivero Oliva, and N. Díaz Rivero. 2007. “Assessment of the Impacts on Health Due to the Emissions of Cuban Power Plants That Use Fossil Fuel Oils with High Content of Sulfur: Estimation of External Costs.” Atmospheric Environment 41: 2202–13. Wei, J., D. Zhao, R. Jia, and D. Marinova. 2009. “Environmental Damage Costs from Airborne Pollution in the Major Cities in China.” International Journal of Environmental and Sustainable Development 8: 190–207. World Bank. 2002. Improving Air Quality in Metropolitan Mexico City, An Economic Valuation. Policy Research Working Paper 2785. Washington, D.C: The World Bank. ———. 2009. World Development Indicators. Washington, DC: World Bank. ———. 2010b. Estudio de Alternativas Metodológicas para la Evaluación de Externalidades. Mexico City: World Bank. 106 Notes i This dataset contains global data, but the quality varies depending on the region and country; it can be consulted at http://arch.rivm.nl/env/int/coredata/edgar.html. ii This dataset can be consulted at http://arch.rivm.nl/databases/nh3. iii This dataset can be consulted at http://www.rivm.nl/geia/data/Ammonia. iv A good compilation of emissions inventories techniques is available from the United States Environmental Protection Agency Technology Transfer Network at http://www.epa.gov/ttn/ chief/eiinformation.html. v As part of establishing an emissions trading mechanism for SO2, this program created an economic value for SO2 emissions, thereby making accurate measurement a critical issue. Arguably, these expensive technologies are needed in all cases; clearly, they provide better information for estimating pollution damage, among other uses. vi The model runs in three separate stages combined in STAGE1N2.exe and STAGE3.exe applications. vii Clearly, further research is needed to more fully understand (i) ERFs or an equivalent method for assessing ecosystem damage and (ii) valuation of environmental damages. It is the authors’ opinion that these are two nascent research areas; thus, the major impacts are on human health, which is a critical area that has been prioritized. 107