International Journal of Sustainable Energy Planning and Management Vol. 18 2018 29 1Corresponding author - e-mail: kartono.sani@sbm-itb.ac.id and kartsani@yahoo.com International Journal of Sustainable Energy Planning and Management Vol. 18 2018 29–52 ABSTRACT Indonesia is facing tremendous challenges in bridging the broadening gap of demand and supply prior to developing a steady energy vision. This paper is to introduce system dynamics in solving the problematic national energy management, share an initial stage of the modelling and discuss the challenges. A literature review defines the preferred modelling, empirical data demonstrates the supply mix trends and used in developing the initial model to represent the past behaviour prior to be enhanced, upgraded and simulated towards the vision. The initial simulation runs succeed in imitating the historical trends and suggests that its engineering to the envisaged patterns may offer viable solution. The use of System Dynamics in Indonesia is unprecedented and the results are noteworthy in supporting the formulation of the national Energy Mix Vision. 1. Introduction Worldwide, a wide range of energy technologies exist, and each country has developed vastly different core competencies to generate their unique portfolios and sustainable energy vision. Many countries signed the Kyoto Protocol (1997), an international treaty whose critical features aim to prevent climate change, reduce greenhouse gas (GHG) emissions, and accelerate renewable energy use. The tradition of energy modelling had seemingly begun when the world was urged to develop energy system models for a sustainable supply and national energy security due to the 1970s energy crisis [1]. All countries have since been competing to develop their own unique energy portfolio to ensure their respective domestic energy supply. In the 1990s, the focus shifted toward the interactions between energy, the environment, and climate change issue, and various new features have then been developed as the existing models were updated and expanded. Indonesia has depended heavily on fossil fuels to maintain sustainable growth, despite having considerable energy resources, worked hard to maintain its declining domestic oil supply and to increase the amount of renewable energy resources in its national energy mix. For those reasons, in the Presidential Regulation No.5/2006 [2], a different mix was promoted, and a much higher share would be coming from renewables, while coal uses would be suppressed due to environmental issues, as its share is projected to be multiplied to substitute the severe shortage of oil (Figure 1). The U.S. Energy Information Administration’s (EIA) short-term outlook [3] shows a broadening gap for petroleum and other liquid supplies versus consumption in Indonesia, after more than a decade of being a net oil importer (Figure 2). An energy shortage is forecasted by 2022, as Indonesia’s crude production continues to fall as its domestic demand is climbing. Business Monitoring International Ltd. [4] projected that production will only Indonesia energy mix modelling using system dynamics Kartono Sani*, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto School of Business and Management, Institut Teknologi Bandung, Jakarta Campus, Graha Irama, 12th floor Jl. HR Rasuna Said Kav.1–2 Jakarta 12950, Indonesia Keywords: Comparative overview; Empirical data; System dynamic modelling; Initial model of indonesia today; Modelling challenges; URl: http://dx.doi.org/10.5278/ijsepm.2018.18.3 http://www.dx.doi.org/10.5278/ijsepm.2018.18.3 30 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics be about half of domestic consumption in 2020, and most believe that with the current trend, Indonesia may only produce oil for another 10 years. According to the EIA, Indonesia’s gas production will peak in 2018, after rising approximately 25 percent since 2005; production will then decline sharply [3]. Ramping up Indonesia’s per capita GDP requires a sustainable energy supply to maintain growth, while the fossil fuels heavily dependent energy supply portfolio caused a rapid increase in emissions. Empirical data on the primary energy supply (Figure 3 and 4) suggest that the current trends will not lead the country to the desired optimum energy mix. The over produced coal is a strong negative indicator in the country’s performance for the Kyoto Protocol’s Clean Development Mechanism. Likewise, the under-performed oil and gas production suggests weak capital stewardship in the upstream ventures. So does the ironic story of the clean geothermal resources for the country that is the majority shareholder of the world’s potential. Even though the new Presidential Regulation No.79/2014 [6] replaced the older one and regulated more ambitious energy mix goals for 2050, the trends remain opposite, and the country is facing tremendous challenges in energy management prior to developing a steady energy vision. This paper aims to introduce System Dynamics modelling as a preferential approach to portrait the past behavior state of Indonesia’s energy supply mix performance in the way to understand and develop a new model to support the formulation of the national Energy Mix Vision. It discusses the energy statistics in the background, an early stage of the modelling with a unique causal-loop diagram reflecting the country energy supply model today, an initial simulation run and the modelling associated challenges. 1.1. Scope and structure This research is began with the finding of the most suitable modelling approach to be used in developing new model to portray the performance of the energy supply mix implication in the past. It is then expected to be able to simulate the behavior of the supply mix in the Energy Mix 2011 (Hydro 2.1%, Geothermal 1.2%, Other 0.9%) RKEN Target Energy Mix 2025 RKEN Target Energy Mix 2050 NRE 4% Coal 26% Oli 50% Gas 20% NRE 23% Coal 30% Oli 25% Gas 22% NRE 31% Coal 25% Oli 20% Gas 24% 6.9 EJ (165 MTOE) 15.9 EJ (380 MTOE) 41.0 EJ (980 MTOE) Figure 1: The past and targeted energy mix of Indonesia [5] 2000 2002 2004 2006 2008 2010 2012 2014 G J /Y e a r Broadening Consumption Indonesia becomes net imported oil Source: U.S. Energy Information Administration, Short Term Energy Outlook, September 2015 Production 0 30 25 20 15 10 5 Figure 2: Petroleum and other liquids supply and consumption in Indonesia after more than a decade of being a net oil import [7] International Journal of Sustainable Energy Planning and Management Vol. 18 2018 31 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto more comprehensive models that incorporate a larger number of economic components, and new models that include the interactions between energy, society, economy, and environment as a major innovation, e.g., GEM- E3, T21 and MCM. And only 12 of which discuss a causal loop diagram or a stock and flow diagram in the presentations [1]. In their two-step comparative overview of models covering energy systems, [9] concluded that “the bottom-up accounting type of framework appears to be more appropriate for developing country contexts for their flexibility and limited skill requirement, they can capture rural-urban differences, traditional and modern energies and can account for non-monetary transactions, the models do not look for an optimal solution, can take non-price policies prevailing in developing countries enhances their suitability, their inability to analyze price-induced effects is the main weakness though; however, given the regulated nature of prices in many developed countries and incompleteness of markets, this weakness is not a major concern for modelling”. However we argue that this weakness would be fatalistic for developing countries because of their less regulated nature; thus, an alternative model is needed that can analyze the price- and many other exogenously-induced effects. Although the best non-simulation model can take non-price policies prevailing in developing countries and enhance their suitability, a good model should look for an optimal solution. Both weaknesses of the bottom-up accounting model can be covered by system dynamics modelling [1]. According to [9] the hybrid models come next and followed by the optimization and econometric models, the latest “use price-driver which play a limited role in developing countries and cannot capture informal sector or traditional energies adequately,” besides having “difficulties in capturing the technological diversity that require high skill levels.” So most of the essentially global models are not suitable for developing countries contexts and so inappropriate as the essential features are not explicitly covered, developed from different perspectives applying entirely those features common to developed countries and fail to capture specific needs of developing countries [1]. [10] argue “A simple distinction is often made between a bottom–up approach, which is more data intensive and more appropriate for detailed analysis of individual energy policies and a top–down approach, which has a more econometric approach and uses less future under certain selected terms and conditions towards the desired destination. Empirical data solicited from the corresponding energy departments is used to examine the past behavior trends (2000–2016), whereas the energy mix vision will be determined and engineered in defining the envisioned behavior modelling (2017-2050). The modeling is to be focused on energy supply system, confronts the dominated fossil energies against the new and renewable energies (NRE), their impact to environment and so development as well as the regulatory instruments. The oil and gas sectors are separated, so are coal and biomass, while the renewable sector is simply represented by geothermal, biofuels, hydropower and solar photovoltaic energy. The simulation programs Vensim PLE 7.1 is used in the building of the qualitative and quantitative modelling. The paper is structured as follows. Section 2 summarized the background literature which have been discussed in more detailed in the author’s previous publication on this topic [1], whereas sub-section 2.1 highlights the advantages of using System Dynamics as research approach. Section 3 outlines the approaches used in this study: (a) analyses the background statistics, (b) energy supply preferred model and its uniqueness, and (c) regulatory instruments and the supply mix dynamics. Section 4 discusses the initial system dynamics model of Indonesia today: (a) causal loop model, (b) stock and flow model, and (c) results of the initial simulation run. Finally, discussion and conclusion are given in Section 5. 2. Literature review Some tabulation structured using energy selection parameters developed on energy portfolio management [8] reveals that none of the articles reviewed discuss the energy supply mix vision is about Indonesia and discuss the energy supply mix quality [1]. The article review was subsequently focused on energy supply system modelling and concentrated on a theoretical justification of why system dynamics is needed to answer the main research questions. Of the 35 articles reviewed on energy system modelling, some papers are categorized as comparative overviews, while the rest discuss the applications of specific system dynamic models for certain energy case studies. They include Bottom-up Optimization/ Accounting Models, Hybrid Models, Electricity System Models, Causal-Descriptive (System Dynamics Models) or Correlation (Top-Down Econometric). There are also 32 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics for the development of renewable energies and a sustainable economy in general, the environmental issue and its impact on the financial statements is no longer an ecological question. For those reasons, [14] “investment in renewables can be at risk, depending on the continued existence of financial incentives.” Policy makers have to either prolong financial incentives to renewables (in spite of the recognized maturity), capacity payments to dispatchable power generation, or by any other design change to provide adequate signals for existing and new generation capacity. 2.1. System dynamics as a suitable energy modeling approach System dynamics is generally understood as an approach to understanding the non-linear behavior of complex systems over time using stocks, flows, internal feedback loops, table functions and time delays. It is a mathematical modeling technique to frame, understand, and discuss the complex issues and problems. In this methodology, a problem or a system is qualitatively represented as a causal loop diagram, a simple map of a system with all its constituent components and their interactions. Two types of feedback loops are the positive reinforcement labeled as R and the negative reinforcement (B or balancing). A causal loop diagram is then transformed to a stock and flow diagram to perform a more detailed quantitative analysis. A stock is the term for any entity that accumulates or depletes over time, a flow is the rate of change in a stock, while time delay is a shift in the effect of an input on an output dynamic response. Thus, System Dynamics modeling is most suitable for modeling in the energy realm because it is: • reliable for large, complex subsystems, nonlinear and dynamic problems. [15,16,17,18,19] emphasize that such problems can only be properly represented or solved by a reliable model of the system dynamics approach. The present models have considered the factors that influence the oil and gas exploration/exploitation industry, along with their effects on the system. A system dynamics model with numerous interconnected variables forming loops is particularly suitable and can be used with relative ease and convenience by following the methodology without sacrificing the basic character of the problem. • can replicate world patterns. [20] “A System Dynamics model that uses energy as the describing medium of all socioeconomic activity has been technology explicit data.” [5] finally suggest “despite the distinction being widespread, both categories are not mutually exclusive, there also exists a ‘hybrid’ class where the two approaches are combined; one of the main contributions of the hybrid approach is the detection of missing information and dynamics that simple top– down or bottom-up models cannot detect on their own.” Based on the literature review, the discussion comes to a conclusion that non-simulation models are not suitable for this research since the models “inadequately capture the developing country characteristics, the level of data requirement and the theoretical underpinning of these models, as well as their inability to capture specific developing country features,” [97] concluded. Subsequently [1] emphasize when the best model of these conventional approaches is applied to developing country like Indonesia, the problems are increased. Thus, it is certain that a very large, complex and dynamic energy portfolio management with feedback of many subsystems in a non-linear fashion cannot be simply represented or easily solved by known or extended mathematical optimization models. Nevertheless [11] presents “two key questions should have been at the top of policy makers’ agenda: (a) can the Government develop a national energy system that will provide security and jobs and also leave a heritage of clean air, clean water, and pristine wilderness areas for the children and grandchildren? (b) can the nation reduce carbon dioxide emissions, which threaten to destabilize the global climate, by developing a truly balanced portfolio of clean energy solutions that would allow to also having economic growth?" This is a very idealistic view that an energy supply system may be built and maintained at zero cost to the environment and is contradicted considering the arguments of [12] that “if the rest of energies are likely to come on-stream fast enough to offset conventional oil decline, what would be the new scenarios of greenhouse gases if this would happen.” He continued “the results show that even strongly optimistic rhythms of substitution have a hard time to continue the growing demand of energy that characterizes today’s pattern.” And concludes that if the present relationship between energy and the economy is maintained these results lead to a long economic stagnation period in the most optimistic scenarios. [12] Whereas ironically [13] emphasizes that International accounting standards do not differentiate between low and carbon intensive investment and do not take into account climate risks beforehand.” This is a critical issue International Journal of Sustainable Energy Planning and Management Vol. 18 2018 33 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto • widely used and has abundant achievements. [17] “System Dynamics is widely used in the study of sustainable development and has plentiful research achievements from macro- perspective but few studies in the microcosmic project systems. Studies that take a look at the complete picture, paying attention both to the economical, the geological and the technological aspects are not frequent in the literature.” In addition, [22] concluded that “the challenge of doing business in development of biofuel in developing countries, such as Indonesia, is that the market suffers from a lack of information, infrastructure and institutions. With inadequate assessment and a poorly equipped infrastructure (local scale policy, market, science and technology and public acceptance), any initiative for a large-scale introduction of biofuel will be premature. Assessing the present governmental policy may help to identify the barriers and at a later stage to find the solutions to ease the penetration of biofuel into existing energy systems.” While [23] highlight the problems associated with public policy using traditional non- simulation approaches, which have several characteristics that impede resolution. Further from most recent publications [24] concludes that “Large shares of renewable energy sources are decreasing energy prices in spot markets due to the merit order effect. This is good news for the consumer welfare,” Meanwhile in crowdfunding platforms for renewable energy investments [25] argues that renewable energy- crowdfunding activity thrives on stable long-term policy support schemes for small and medium scale projects, as well as on comprehensive financial regulation that exempts crowdfunding from traditional financial service regulatory obligations.” Interestingly, comparing the evolution of financial regulation of crowdfunding, it is concluded that “a loose financial regulatory framework leads to a range of business models and financial instruments, while a more specific framework tends to reduce RES-crowdfunding to one business model/one instrument,” [25]. However, back to the merit-order effect, it is argued that in fact the value estimated for the financial incentives is often lower than the merit-order effect. 3. Research approaches To conduct this research, at the initial phase, a model is built to delineate the existing system as baseline to demonstrate the strengths and weaknesses of the existing possible to replicate the world patterns as the result of the interplay of the industries that extract and refine primary energy and compete to provide the driving force of socioeconomic society.” • can lead to develop effective policies to achieve sustainability and be used to analyze possible threats and design optimal adaptation strategies. [19] explain that from a policymaker perspective, a System Dynamics model of the automotive sector can lead to the development of effective policies, while from an energy company’s such a model could be used for the analysis of a highly volatile and market that is always on the edge of starting a new major transition. “The model presented can serve both purposes, and the results obtained show how a similar instrument can really make the difference in highly dynamic sectors with ongoing major transitions.” [19]. • a good method for a holistic approach, “since it enables the integration of several aspects and is designed to take into account all sorts of feedbacks among those aspects.” [13] • a valuable tool to forecast supply and consumption. [12] argue that the simulation results allow the model to forecast fuels supply and consumption that making it “a valid alternative to the most known and used forecasting techniques in the context of energy-economics.” This model “can be used to investigate the factors influencing the long-term supply and demand of energy and to determine the nature of system behavior as well as examining the effectiveness of various policies in softening the transition from self-sufficiency to energy import- dependence in the long term.” [12]. • a sophisticated modelling method, which is needed for the supply side. [21] state that “the wide array of current and proposed production technologies, each with different costs and benefits and different greenhouse gas emissions levels, makes electricity production the most complicated element of the supply-side.” and so needs the sophisticated System Dynamics modeling. • useful for gaining insight into the underlying behavior [16], “System Dynamics modeling is useful for understanding the underlying behavior of complex systems over time, taking into account time delays and feedback loops” and 34 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics resolution, formulation of which will subsequently be examined and fine-tuned through Focused Group Discussions. Under this scenario, the concerned parties will have a better understanding of the concept and so a higher sense of belonging and will therefore buy in and be interested to participate with a stronger sense of urgency to implement the envisioned energy mix goals. The model may subsequently be improved to obtain the ultimate energy mix vision through further feedback from energy industry practitioners [1]. In those regards, the following discusses the background statistics, the regulatory instruments, and the energy supply preferred system dynamics models for this study as well as the uniqueness of the new model. 3.1. Background statistics This empirical data [5] is presented as comparison to the modelling outcomes of the past behavior and used as baseline in engineering the model to the envisaged future. The total of supply of primary energy in barrel oil equivalent (BOE) shown on figure 3 appears to have a steady increase from approximately one billion BOE in 2000 to approximately 1.6 billion BOE in 2015. When the Presidential regulation No.5/2006 was issued [2], the total was still less than 1.2 billion BOE. In terms of volume, the graphs also indicate that each energy technology shows significant increases, except in the case of biomass, which appears to remain flat. This apparently corresponds to the steady ramping up of Indonesia’s population, per capita GDP, as well as the intensity of the final energy consumption per capita. system based on the system behavior today, the trends and the potential pitfalls to which it is heading toward, its impacts to sustainability of the national energy supply and its capability in leading the country to a securer energy supply system. Along with the statistics in the background, the System Dynamics modelling are structured and assembled based on the past constrains and situations. The developed simulation model is then processed and run sequentially to obtain the outcomes that a) closely represent the past realities or statistics and b) would have represented the past if all the variables representing best practices were well implemented. So the objective of this initial modelling is to delineate the past situation that led to the current problematic behavior, and take lessons learned to be used in developing new model that can lead to the desired behavior. This is with respect to the research question how system dynamic modelling will be able to simplify the complex and dynamic realm and be reliable as an alternative model in the development of the Energy Mix Vision. In the next phase of this research, the model will be upgraded and engineered to accommodate more key variables or feedback loops that potentially constrain the future and be directed to the desired energy journey and destination. The following proposition may eventually be developed: through studying the structure, statistics and policies of the past energy supply system, it is expected that the historical barriers may be identified, mapped and subsequently used in developing alternative models. The model is in turn will be used as basis in the formulation of a new energy mix vision and policy frameworks of P ri m a ry e n e rg y su p p ly [T W h ] 3,000 2,500 2,000 1,500 1,000 500 20 00 Year Coal excl. Exported Crude Oil & Product Natural Gas & Product Hydro Power Geothermal Biomass Biofuel Total 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 *) 20 14 ** ) 20 15 0 Figure 3: Indonesia supply of all primary energy in barrel oil equivalent [5] International Journal of Sustainable Energy Planning and Management Vol. 18 2018 35 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto when using traditional non-simulation approaches into account. And marking the uniqueness of the new model, new variables are developed from the common characteristic impediments [19] as new criteria as shown in Table 1. In this case, the Ghaffarzadegan’s factors will affect the delay of policy designing and communication and the delay of policy implementation which are along with Government ability & capability will in turn cause delay in power development for additional capacity of NRE (Figures 5 and 6). The delay in policy designing and communication is contributed by under estimated policy maker (gf1), scapegoat-minded perspective (gf2) and political link & lobbying (gf3), while the delay in policy implementation is composed by policy acceptance (gf4) and rate of trials and errors (gf5). See Appendices A and B. The new approach focuses not on discrete decisions but on the potential impediments and the policy structure underlying the decisions, and emphasizes a continuous view that strives to look beyond events to see the dynamic patterns underlying them as partly represented by the additional criteria [1]. So it would be a combination of the different energy supply models, with all those endogenous and exogenous variables to be selectively integrated into the model. Some novelties of the new model may be claimed by the following: 1) the complex realm of energy supply system is modelled using system dynamics, an unprecedented modelling effort for Indonesia to help policymakers to gain insight the system and subsequently use it in policy making consideration, 2) the new modelling includes new variables developed from the common characteristic impediments in public policy Those assumptions will soon be changed when the graphs are presented in terms of share percentages of each technology (Figue 4). At the time that the Presidential Regulation No. 5/2006 was issued, the national energy mix was oil 39.24%, gas 17.51%, coal 16.72%, biomass 23.51%, while all renewables contributed only 3.02% (hydropower 2.06%, geothermal 0.95% and biofuel only 0.01%). The question is, what have been happening since the issuance of the presidential regulation and the national energy mix goals were regulated? The shares of oil and biomass have continuously been declining, gas remains relatively stable, while coal has agressively been ramping up, maintaining the previous robust trends, whereas renewables remain at a crawl. It is probably for those reasons that the Presidential Regulation No. 79/2014 [6] has subsequently been issued. The gaps between those envisaged by the energy mix vision 2025 (oil 25%, gas 22%, coal 30% and New and Renewables 23%) and reality remained unbridged after almost a decade, as the role of oil remained dominant at 38.35%, gas is unchanged at 17.03%, coal increased to 22.21% and biomass remained high at 18.86%, while New and Renewables were stagnant at only 3.56%. There appears no clue with the current trends that the energy journey is heading to the desired destination. 3.2. System dynamics model of preference The new model will be focused on the energy supply system dynamics that take the unique factors of the developing archipelagic country and the common characteristic impediments of public policy development S h a re [ % ] 50 45 40 35 30 25 20 15 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 Year 2008 2009 2010 2011 2012 2013*) 2014**) 2015 The issuance of Presidentail Decree No. 5/2006 The issuance of Presidentail Decree No. 79/2014 Oil Coal Gas Hydropower Geothermel Biomass Biofuel 41.74 42.42 42.32 40.37 43.52 42.32 39.24 38.5 38.08 37.28 37.57 38.91 37.32 36.79 39.41 38.35 28.8 25.83 25.25 24.05 23.75 23.15 23.51 22.37 22.03 21.71 19.03 18.64 19.04 18.51 19.49 18.86 16.54 16.53 17.65 18.05 16.39 16.39 16.77 14.92 18.7 19.3 19.84 17.23 16.43 16.33 17.06 17.03 9.42 11.44 11.48 14.58 13.24 14.89 17.51 20.97 17.8 18.18 22 23.92 24.56 20.21 22.21 2.54 2.82 2.34 2.03 2.13 2.32 2.06 2.31 2.3 2.2 3.1 2.06 2.04 2.58 2.4 2.15 0.96 0.96 0.96 0.92 0.97 0.94 0.95 0.93 1.06 1.26 1.08 1 0.96 0.92 1.02 1 0 0 0 0 0 0 0.01 0.02 0.03 0.06 0.1 0.15 0.29 0.32 0.42 0 Figure 4: Indonesia supply of all primary energy in percentage [5] 36 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics with the many ministerial policies and regulations have since been prevailing, it does not seem obvious that the legal instruments have succeeded in influencing the past trends of any technology or energy resource. The most distinctive is the case of renewable technologies, although in volume (Figure 3) they have shown significant increases, their shares in percentage are almost flat or even decrease (Figure 4). Only coal’s supply demonstrates a progressive increase and has seemingly substituted the steady drop of oil supply and the stagnant growth of the gas share. Biomass continuously declined from 23.5% in 2006 to 18.86% in 2015, possibly suggesting more development when using traditional non-simulation approaches as shown on Table 1. 3.3. Regulatory instruments and the energy supply mix dynamics An overview to the corresponding ministerial regulatory instruments that were issued during the same period as governmental efforts to manage the energy mix performance nationally and to influence the ongoing trends is performed and finds that at least 76 relevant ministry regulations have been enacted since the issuance of the Presidential Regulation No.5/2006. Interestingly, Table 1: New variables developed from the five characteristics that impede resolution in public policy development and their relations to the policy making process (Modified after Sani, K. et al, 2017) Tradition Non- Policy Making Simulation Approach System Dynamics (Simulation Approach) Process Characteristic Impedences to resolution in public policy development (Ghaffarzad egan et al, Proposed New Commu- Implemen 2007) Criteria New Variables Designing nication tation Over-confident Under-Estimate Confidence level of policy makers  policymakers Policymaker Complexity & difficulty of the policy challenges  Potential delay & degree of uncertainty of policy making  Policymakers' assumptions, model of thinking & strategies  Need to have an Scapegoat Minded Potential of undesirable events to endogenous perspective occur from the policy  Potential self-serving bias  Ability to learn from the environment  The tendency to looking for scapegoat  Need to persuade different Political Link and General agreement among diverse stakeholders Lobbying stakeholders  Potential merits of policy to  Broader public consensus behind the policy  Effective means to inform & persuade stakeholders  Policy resistance from the Policy Acceptance Public resistance against the policy  environment Delay between policy action & result  Immediate impact of policy to the industry  Immediate benefits brought by the policy  Need to experiment & the Rate of Traols and Cost of the policy experimentation  cost of experimenting Errors Attempt to improve future performance  Implement, observe and adjust policy  Experimentation attitude  International Journal of Sustainable Energy Planning and Management Vol. 18 2018 37 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto 4.1. Causal-loop model As shown on figure 5, the following three major causal loops were demonstrated in this qualitative model of the past that resulted in the behavior: 1. Reinforcing Loop A (Government Policy - Fossil Energy Dominated System Economics - Power Demand). 2. Balancing Loop B (Government Policy - Supply and Demand Gap - Fossil Dependence System). 3. Balancing Loop C (the Energy Vicious Circle or Supply and Demand Gap Oil Gas Coal Biomass - NRE). Reinforcing Loop A, in which all relationships are positive, is typical of the past system characterized by the fossil energies dominated regime in which various incentives for fossil energies development were regulated and maintained. Therefore, the Government had since been growing a fossil energy-based economic system for economic growth with the impact on greenhouse effects, besides GDP and then power consumption/demands and dependence of fossil power system. Loop A was supported by or resulted in Balancing Loop B in which, besides the positive relationships, some are negative, by which the Government policies primarily encouraged and were dependent upon imported refined oil in countering the supply and demand gap that in turn increased the country’s dependence on fossil energies. Next at the very core of the system dynamics model is Balancing Loop C, or the Energy Vicious urbanization as people from the villages crowded the major cities and metropolises, or the aggressive expansion of cities in Indonesia. 4. Results - Initial system dynamics model of Indonesia today According to the Handbook of Energy and Economic Statistics of Indonesia 2016, from 2000 to 2015, the population of Indonesia has increased from 205,843,000 to 255,462,000. This was followed by GDP growth from 1,390 to 3,042 trillion rupiah, the Primary Energy Supply increases from 726,687,000 to 1,332,242,000 BOE, and Primary Energy Supply per Capita from 3.53 to 5.22 BOE/capita. This will be used as the basis in developing the system dynamics model of the past. In this model, the core is surrounded by at least six major Causal Loop Feedbacks that are tapped into it, i.e., the Oil sector, Gas sector, Coal sector, Biomass, New & Renewable sector, and the Regulator. The oil and gas sectors are usually treated as one sector in the upstream during the exploration, drilling and exploitation, and separated as they are transported and enter the mid- downstream industry. The coal sector is usually divided into two subsectors, the strip-mined coals and underground- mined coals from which the national coal productions are derived. In a much smaller scale, the renewable sector is traditionally supported by the subsectors of geothermal, biofuels, hydropower and solar photovoltaic energy. Reinforcing A Balancing B CBalancing Population GDP LNG exported Economic Growth Greenhouse effects issues Fossil Energies Base Economic System Depreciation of fossils Energy system Delay of Policy Desigining & Communication Power Consumption Government ability & capability Ghaffarzadegan’s factors Delay of Policy Implementation Additional power capacity Power capacity needed Dependency4 Dependency1 Dependency5 Power Generation by NRE Total Power availble Energy Vicious CircleUnder-developed NRE CO2 emission from fossils power generation Dependence on fossil power system Domestic supply and demand gap Coal Exported Crude Exported Power Generation by Coal Dependency2 Dependency3 Refund Oil Imported Public investment Power Demand Power Shortage PLN Tariff Barriers Robust Fossil Energies Development Limited NRE power system Traditional NRE power system Various Incentives for Fossil Fuels Investment Power Generation By Gas Government Policy in favor of Fossil Energies Power Generation by Hydro Power Generation by Biomass Power Generation by Oli Power Generation by Geothermal Oil Import Cartel - + + + + + + + + - - + + + + + + + - + + + + - - - + + + - + + + + + + + + + + + Figure 5: Causal loop diagram for energy portfolio management in indonesia today 38 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics and market optimization, energy production efficiency, and energy diversity.” Thus, given the present conditions of the energy portfolio management formulated above, the key variables deriving the current system are identified, and their possible relations and interconnections are analyzed, the initial principal causal-loop diagram is subsequently generated to represent the system’s view of the present conditions. After the model structure is determined, assessed, and defined at the very core of the system’s thinking, the potential feedback loops of various subsystems with the many variables that potentially influence and shape the system’s behavior will then be progressively expanded and included. Various incentive for fossil fuels investment, PLN tariff barriers, oil import cartel of Reinforcing Loop A are less relevant, whereas expansion of the power generations by various renewables is a must to get rid of the energy vicious circle. As more renewables can substitute the fossil energies the Balancing Loop B will be weakening while a new Balancing Loop associated with the growing renewable energies to grow, on which a new energy system are relied on. 4.2. Stock and flow model In this quantitative model, six major and five minor stocks are developed, the major levels consist of Population, Power Generation by Oil, Gas, Coal, Biomass and Oil Generation by Imported, while the minor ones are composed of Power Generations by biofuel, hydro, geothermal and conceptually solar photo- voltage and nuclear. GWh is used as the unit of power (electric or non-electric) produced and Gw is used for the unit of power capacity. While Government ability and capability is set at approximately 0.4 to fulfill the additional power plant capacity needed, this is based on the Government budget readiness in building new power plants that according to [7], required an investment of US 2 million/GW. Subsequently, power electric growth per capita and non-electricity growth per capita are defined as RAMP to represent the demand growth of electricity per capita of 930 KWh/person, and a growth of electricity of 8.6% per year. The power capacity needed is derived from the domestic supply and demand gap multiplied by factor 1.3, first used by PLN to cope with the demand, as the power capacity has to be maintained higher than the growing demand. Appendix A describes all the key variables, units and equations used in the quantitative modelling resulted in Circle, as introduced by [1], who developed this initial model based on early observations “of the way the government has since been handling the energy supply portfolio management to cope with the continuously increasing energy demand in supporting sustainable economic growth, while the energy resources of preference, the top-priority oils are not always sufficient to meet the vital requirements.” Except for several or a couple of developed countries who have made breakthrough policies, like Germany and France [1], national energy management has been dragged down by the classical global energy practices, so that its characteristics, trends, and patterns are very similar in many countries and may be described as follows: • The system treats the oil sector as a top priority in meeting the national energy demand regardless of the localities of the resources, its impact on environment, and the volatility of its prices. “This is seemingly due to the well-established oil-based energy utility systems in most sectors throughout the world.” • Therefore, the gas sector resides in the second level despite being more environmentally friendly and more stable, having longer-term pricing, and recently having more discoveries made and new reserves booked. • The coal sector, whose reserves are abundant and easy to explore and extract is in the next level. It holds a larger portion in the supply system to meet the electricity demand, and substitutes the shortage of oil despite its environmentally unfriendly technology characteristics. • Last is the New and Renewables sector, which has indeed been treated as the last resort in energy supply portfolio management. This sector has been left behind in terms of exploration, exploitation and utilities despite their abundance and environmental friendliness, “This position is seemingly due to the old premise that the technologies are expensive, for the long-term externalities costs of their competitors are not taken into account, and thus their ability to be a primary energy supply is deteriorated by their intermittence and scattered characteristics.” • “The current national energy policy and strategy have not taking into account the externality costs of the energy technologies, the localities of the energy resources concerning modes of transport International Journal of Sustainable Energy Planning and Management Vol. 18 2018 39 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto The second scenario still occurs during the past time frame (2000–2015) but considers if the energy mix policy and green energy policy were fully committed, dependency on imported oil was minimized, although government ability and capacity remained the same, public investment in NRE was very high as the government improved the business process resulting in insignificant time delays, which attracted investors. The energy portfolio performance would have been totally different. In the quantitative model (Figure 6), at the very core is the total domestic power available, in which the Balancing Loop C is controlled by many stocks (power generation by oil, gas, coal, biomass and most importantly the renewable energies) and including the various energy exported as well as imported, each of which possess internal feedback loop that behaves independently in non-linear fashions over time. The difference between the energy supply mix targets and the total domestic power available result in the domestic supply and demand gap. Under the current system, the gap is short-termly solved by the variable of additional of imported oil and energy strategy and policies in favor of the fossil energies, that is combined poor energy mix and uncommitted green energy policies. Resistance on the Balancing Loop B leads to the problem of system dependent on fossil energies. The domestic supply- demand gap should be managed through committing the various energy strategy and policies in long terms, by which the shares or additional capacities of NRE and fossil energies are regulated to produce an increasingly improving environmentally friendly energy supply balances. Besides the variables of government ability and public investment that determine the additional power capacity, the Reinforcing Loop A involves time delays that lead to the delay of policy implementation and delay of power development. Both delays are accumulation of the Ghaffarzadegan’s factors or variables. This positive reinforcement feedback loop is to determine the balance of additional capacity of NRE against additional capacity of fossil producers, the energy mix that in turn dictates the CO2 emissions and defines independency of the energy supply system. The simulation results are quite encouraging. Compared to the past statistics (black, dotted line), the old model during the past time (scenario 1 old past, blue) obviously shows similar trends, both graphs resemble one another, and the model simulation moderately succeeded in delineating the historical data. Meanwhile, the behavior of Indonesia Today, while Appendix B shows the initial value of the variables for the initial simulation. The delay time is usually delaying time for construction of a power plant, in this case, because PLN needs to plan, design and budget approximately 2 years before obtaining permits from the government. In this modelling, the delay time is separated into delay time1, which is associated with the delay of power development, whereas delay time2 is associated with policy implementation. Delay time1 and delay time2 both correspond to common characteristic impediment in public policy designing, communication and implementation [23]. Next is the capacity factor (cf), which is the ability of the power plant to supply energy in a year, the ratio of an actual electrical energy output over a given period of time to the maximum possible electrical energy output over the same amount of time. Their values are very dependent upon the technology and thus differs for different kinds of power plants. [23] uses the average values from PLN 2016, but an average of various sources is used here. All the initial value of stocks (for the year of 2000) are quoted from [5]. Meanwhile, it is important to note that the Energy Mix Policy here is prefixed with the adjective “poor” and is defined as IF THEN ELSE (energy mix <0.25, 0.12, 0.12) to suggest that in the past, although the policy had already been issued early with the Government Regulation No.5/2006, in practice the derivatives regulations are still in favor of fossil fuels. Similarly, the Green Energy Policy has the adjective “uncommitted,” as it defined at a higher threshold of CO2 emissions under the function of IF THEN ELSE (CO2 emission of fossil power generation >50, 0, 0), which is still in favor of fossil fuels. 4.3. Initial simulation run In this initial experimentation, two scenarios were developed, the simulation period is set for 15 years, i.e., from 2000 to 2015, in accordance with the data available to delineate the past. Figure 6 shows the flow and stock diagram, and the scenarios are summarized in table 2 as follows: The first scenario represents the past behaviour, although the energy mix policy and green energy policy were regulated earlier, following the Government Regulation No.5/2006 commitment to implement the policies were still low, dependency on imported oil was robust and ruled by the oil import cartel, the government’s ability and capacity to grow energy capacity was low, while public investment in NRE was almost zero, and significant time delays occurred as the government struggled to overcome the bureaucracy. 40 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics P ow er ge ne ra tio n by so la r c s po w er P ow er ge ne ra tio n by hy dr o po w er P ow er ge ne ra tio n by ge ot he rm al P op ul at io n ad di tio na l e ne rg y fro m s ol ar c s pp ad di tio na l e ne rg y fro m h yd ro p p ad di tio na l e ne rg y fro m g eo th er m al p p fra ct io n of so la r c s fra ct io n of hy dr o fra ct io n of ge ot he rm al p p N R E to ta l p ow er ca pa ci ty bi rt hs bi rt h ra te ad di tio na l o f so la r p p ad di tio na l o f hy dr o pp ad di tio na l o f ge ot he rm al p p cf o f s ol ar cf o f h yd ro cf o f ge ot he rm al de at hs de at h ra te P ow er ge ne ra tio n by bi of ue l ad di tio na l e ne rg y fro m b io fu el ed p p cf o f b io fu el ad di tio na l o f bi of ue l p p fra ct io n of bi of ue l P ow er ge ne ra tio n by nu cl ea r ad di tio na l e ne rg y fro m n uc le ar p ow er cf o f n uc le ar po w er ad di tio na l o f nu cl ea r p p fra ct io n of nu cl ea r p p P ow er ge ne ra tio n by o il P ow er ge ne ra tio n by g as P ow er ge ne ra tio n by c oa l do m es tic s up pl y an d de m an d ga p en er gy s up pl y m ix ta rg et s en er gy s tr at eg y an d po lic ie s cf o f g as su pp ly cf o f c oa l su pp ly ad di tio na l o f o il pr od uc er s ad di tio na l o f g as pr od uc er s ad di tio na l o f c oa l pr od uc er s to ta l p ow er de m an d el ec tr ic ity g ro w th pe r c ap ita no n el ec tr ic ity gr ow th p er c ap ita to ta l d om es tic po w er a va ila bl e fo ss il fu el s to ta l ca pa ci ty po w er c ap ac ity ne ed ed ad di tio na l p ow er ca pa ci ty ca pa ci ty co nv er si on go ve rn m en t a bi lit y & c ap ab ilit y pu bl ic in ve st m en t de la y of p ow er de ve lo pm en t de la y tim e1 ad di tio na l c ap ac ity of fo ss il pr od uc er s C O 2 em is si on o f fo ss il po w er ge ne ra tio n em is si on b y co al em is si on b y ga s em is si on b y oi l fa ct or em is si on o il fa ct or em is si on g as fa ct or e m is si on co al gr ee n en er gy po lic y en er gy m ix po lic y en er gy m ix ad di tio na l e ne rg y fro m c oa l ad di tio na l e ne rg y fro m g as ad di tio na l e ne rg y fro m o il cf o f o il su pp ly to ta l e le ct ric ity de m an d to ta l n on e le ct ric po w er d em an d to ta l c oa l ex po rt ed to ta l b io fu el ex po rt ed to ta l L N G ex po rt ed to ta l c ru de ex po rt ed fra ct io n of bi of ue l ex po rt ed fra ct io n of c oa l ex po rt ed fra ct io n of c ru de ex po rt ed de la y of p ol ic y im pl em en ta tio n de la y tim e2 gf 1 gf 2 gf 3 gf 4 gf 5 P ow er ge ne ra tio n by bi om as s ad di tio na l e ne rg y fro m b io m as s pp cf o f b io m as s ad di tio na l o f bi om as s pr od uc er s em is si on b y bi om as s fa ct or e m is si on bi om as s < en er gy m ix > O il ge ne ra tio n by im po rt ed ad di tio na l e ne rg y fro m im po rt ed o il cf o f i m po rt ed oi l ad di tio na l o f im po rt ed o il fra ct io n of im po rt ed o il ad di tio na l ca pa ci ty o f N R E < ad di tio na l ca pa ci ty o f N R E > G ha ffa rz ad eg an 's fa ct or s de pe nd en cy fr ac tio n of o il pr od uc er s fr ac tio n of g as pr od uc er s fr ac tio n of c oa l pr od uc er s fr ac tio n of bi om as s < de pe nd en cy > O il su pp ly ch an ge Ti m e to o il su pp ly c ha ng e H ei gh t t o oi l su pp ly c ha ng e O il im po rt ed ch an ge Ti m e to o il im po rt c ha ng e H ei gh t t o oi l im po rt c ha ng e G as s up pl y ch an ge Ti m e to g as su pp ly c ha ng e H ei gh t t o ga s su pp ly c ha ng e C oa l s up pl y ch an ge B io m as s su pp ly ch an ge B io fu el s up pl y ch an ge G eo th er m al su pp ly c ha ng e Ti m e to g eo th er m al su pp ly c ha ng e H ei gh t t o ge ot he rm al su pp ly c ha ng e H yd ro po w er su pp ly c ha ng e So la r p v su pp ly ch an ge LN G e xp or te d fra ct io n ch an ge fra ct io n of L N G e xp or te d LN G e xp or te d fra ct io n A vg F ig ur e 6: S to ck a nd f lo w d ia gr am f or t he s tu dy Table 2: Initial scenarios of the simulation (2000–2015) Green Fraction of Government Public Energy Mix Energy Oil ability & Investment Delay Delay Scenario Policy Policy Imported capacity on NRE time1 time2 Scenario 1 Old (energy mix Past <0.25, 0.12, 0.12) (CO2 emissions >50, 0, 0) 35% 40% 10% 2 2 Scenario 2 New (energy mix Past <0.75, 0.3, 0.3) (CO2 emissions >10, 0, 0) 15% 40% 60% 0.5 0.5 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 41 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto of the Banyu Urip oil field in 2008 is moderately delineated. An integration error tests that held by means of repetitiously cutting the time step in half and running the model from original time step of one year to only 0.03125 year when the results are no longer sensitive to the choice of time step. [26] argues “The integration error test should be the first simulation test you carry out, since failure here renders all model results meaningless.” Subsequently each decision rule on Table 2 in the model are examined for extreme condition tests by simulation and asked whether the output of the rules are feasible and reasonable even when each input to the equation takes on their maximum and minimum values (Table 3). if the two scenarios are compared, it is obvious that under the second scenario (Scenario 2 new past, green, bold line) great improvement could have been achieved in all sectors of the national energy realm if the second scenario was well implemented. Figure 7 demonstrates that under the Scenario 2, although the total capacity of fossil fuels still increase significantly, the total power capacity of NRE increased drastically (green) from the previously almost flat curve (blue) under the old energy regime (Figure 8). These changes can be also seen in the oil sector (Figure 9–11), whereby imported oil (Figure 11) was drastically forced to decline as the domestic crude production (Figure 9) and crude exported (Figure 10) naturally declined under both scenarios. The presence 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 F o ss il e n e rg ie s [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Figure 7: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total capacity of fossil energies by means of exporting table time down data from the simulation run into the same excel spreadsheet with the empirical data 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 1,600 1,400 1,200 1000 800 600 400 200 0 N e w & R e n e w a b le s [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year The issuance of PPNo. 5/2006 on Energy Mix Policy Figure 8: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total capacity of new & renewable energies by means of the same as in Figure 7 42 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics maximum, but not power generation by coal and its export that approach zero. Under this scenario, only oil imported drop to almost zero, oil and gas production and In the extreme high case production of NRE goes to maximum capacity (includes biofuel, geothermal, hydro and solar powers), total domestic of power available also 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 1000 900 800 700 600 500 400 300 200 100 C ru d e p ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year BanyuUrip Oil onstream (2008) Figure 9: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total capacity of crude production by means of the same as in Figure 7 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 450 400 350 300 250 200 150 100 50 0 E xp o rt e d C ru d e [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Figure 10: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total exported crude by means of the same as in Figure 7 400 350 300 250 200 150 100 50 Im p o rt e d O il [T W h ] Statistics Scenario 2 new past Scenario 1 old past 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 Year Figure 11: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total imported oil by means of the same as in Figure 7 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 43 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto Back to the simulation, the same trends occur in the gas sector (Figure 12–13), the performance of power generation by gas shows slightly decline under both scenarios, which is in turn also affect the total of LNG exported, despite more LNG exported under Scenario 2. Nevertheless, for the time period, this simple initial simulation poorly reveals the incoming of Tangguh gas field (Figure 12) with the associated Tangguh LNG biomass go flat at their initial values. Meanwhile in the extreme low case, production of NRE immediately drop to zero, total domestic of power available also go to zero, so does power generation by coal and its export. Under this scenario, only oil imported steady high, while oil, gas and biomass remains flat at their initial values. The extreme condition tests suggest that the model is fine, no implausible behavior generated and no flaws uncovered. Table 3: Extreme condition scenarios of the simulation (2000–2015) Green Fraction of Government Public Energy Mix Energy Oil ability & Investment Delay Delay Scenario Policy Policy Imported capacity on NRE time1 time2 Scenario (energy mix (CO2 Extreme <1, 1, 1) emissions High >0, 0, 0) 0% 100% 100% 0.01 0.01 Scenario (energy mix (CO2 Extreme <0, 0, 0) emissions Low >100, 100, 100) 100% 0% 0% 100 100 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 1350 1300 1250 1200 1150 1100 1050 1000 G a s p ro d u ct io n [T W h ] Scenario 2 new pastStatistics Scenario 1 old past Year Tangguh gas onstream (2009) Figure 12: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total gas production by means of the same as in Figure 7 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 330 310 290 270 250 230 210 190 170 150 L N G e xp o rt e d [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Tangguh gas onstream (2009) Figure 13: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total LNG exported by means of the same in Figure 7 44 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics this is followed by the total coal exported which demonstrates the same trends. The simulation however missed some deflections on the statistics of both graphs that mark the beginning of global coal prices drop in 2012 since feedback loops for coal prices fluctuates as well as the global market have not been taken into account. The biomass, the power generation by this old, traditional and rural fueling system shows an obvious drop in Scenario 2 from previously inclined to increase under scenario 1 and statistically (Figure 16). Otherwise, during the same period of time, the NRE, including power generation by biofuel (Figure 17), power generation by geothermal (Figure 18) and power generation by hydro (Figure 19), are strongly leveraged and demonstrate drastic increases under Scenario 2 (after being suppressed under Scenario 1) thanks to the government policies that regulated various incentives in exported on stream in 2009 (Figure 13), as the stock fluctuates cannot be captured by the curves of both scenarios. This can be understood as the onstream of new oil and gas fields or green fields and the shutdown of brown fields as separated feedback loops have not been included in the quantitative model. Similarly, this initial simulation model has also missed the complexity of Indonesia’s annual imported refined oil (Figure 11), as it is believed that more non- linear feedback loops are involved in its system dynamics. This may include oil and gas prices fluctuates, available supply from spot markets, domestic oil demands as well as the oil production. In the coal sector (Figures 14 –15), the power generation by coal keeps increasing under both scenarios in line with the existing energy mix goal to manage the share of coal to ramping up from 26% in 2011 to 30% in 2025 while the contribution of oil and gas decreases, and 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 *) 20 14 ** ) 20 15 8000 7000 6000 5000 4000 3000 2000 1000 0 C o a l P ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Figure 14: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total coal productionby means of the same as in Figure 7 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 *) 20 14 ** ) 20 15 6000 5000 4000 3000 2000 1000 0 C o a l e xp o rt e d [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Figure 15: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total coal exported by means of the same as in Figure 7 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 45 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto Building a more appropriate, long-termly viable energy system is needed to ensure a sustainable energy supply. The literature review revealed that a limited number of publications are available on this topic, and identified that the energy supply quality system with the unique factors of developing country like Indonesia constitutes a research gap, to which the work was subsequently directed to find a more reliable modeling. Subsequently, assessing the present government policy helped identify the barriers, and at a later stage, find the solutions to ease the penetration of renewable energies into existing energy systems. System dynamics modeling is recognized as an excellent methodology with strong advantages for such holistic approaches in energy supply management, all of which makes it a valid alternative to the most well-known favor of NRE investment and took firm side with the environmentally friendly energies. 5. Discussion The statistics, the causal loop diagram and the discussion in chapter 4.4 reveal that the challenges and opportunities in the energy portfolio management in the country are real and quite complicated. The current supply system, which is exposed openly to free market in striving to meet uncontrollable and disintegrated sectorial demands, is obviously problematic. The current energy mix vision with the supporting policy instruments issued to date have not succeeded in taking the energy journey of Indonesia to the desired destination. This is partly due to the absence of a clear model of the ongoing system and a traceable origin of the current energy mix vision. 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 600 550 500 450 400 350 B io m a ss P ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year Figure 16: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total biomass power production by means of the same as in Figure 7 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 *) 20 14 ** ) 20 15 1,200 1,000 800 600 400 200 0 B io fu e l P ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year The issuance of PPNo. 5/2006 on Energy Mix Policy Figure 17: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total biofuel power production by means of the same as in Figure 7 46 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics involving the government policies that treated fossil energies as preference and to have caused strong dependence on it. This situation was driven by the major loop A, which also begins with the government policies that regulated various incentives in favor of fossil energy investment, PLN tariff barriers, oil import cartel that led to robust fossil energies development and otherwise hindering the growth of NRE power systems. This fossils-based economic system is in turn resulting in environmentally unfriendly economic growth, with the associated green-house gas issues and has impeded the country in building its sustainable and self-sufficient power system, even though it is geologically blessed many types of renewable energy potential. The initial simulation of the stock and flow model is quite encouraging in that, through the 2 scenarios developed, and used forecasting techniques in the context of energy economics and confirmed the superiority of system dynamics modeling compared to the traditional non- simulation approaches. So, this modelling approach itself and its focus on the unprecedented energy supply quality mix of an archipelagic country and in the inclusion of a series of new variables discussed early make this research are interesting and rather unique. The principle causal-loop diagram illustrated in Figure 5 demonstrates the complexity of the energy supply realm, the balancing loop C at the core of the system has apparently been dragging the entire causal loop feedbacks tapped into it and involved in the problematic energy circle. This old circle has traditionally been maintained by the supporting balancing loop B 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 15 20 13 *) 20 14 ** ) 200 180 160 140 120 100 80 60 40 20 0H yd ro p o w e r P ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year The issuance of PPNo. 5/2006 on Energy Mix Policy Figure 19: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total hydropower production by means of the same as in Figure 7 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 *) 20 14 ** ) 350 300 250 200 150 100 50 0G e o th e rm a l P ro d u ct io n [T W h ] Statistics Scenario 2 new past Scenario 1 old past Year The issuance of PPNo. 5/2006 on Energy Mix Policy Figure 18: Comparing statistics [5] with simulation results (Scenario 1 and 2) for total geothermal power production by means of the same as in Figure 7 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 47 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto old past) obviously shows similar trends, both graphs resemble one another, and the model simulation moderately succeeds in delineating the historical data. However, it is recognized that this simple initial simulation failed to capture some deflection points shown by historical data trends that is believed to corresponding to more non-linear feedback loops involved in the energy supply system. Meanwhile, if the two scenarios are compared, it is obvious that under Scenario 2 new past, great improvement could have been achieved in all sectors of the national energy realm if the second scenario was well implemented. This model is potentially reliable as an alternative approach for policy design and in the efforts to build a new and more viable Energy Supply Mix Vision. Future work recommendation: Engineering of the model to the envisaged patters and subsequently examined and fine-tuned through focus group discussions may offer a more viable solution. Ideally, the energy demands and economics are explicitly included to get more holistic understanding and representative behaviors. Meanwhile new energies such as tidal turbines, second and third generation biofuels, solarpanel positioning robots, photovoltaic transparent glass, space-based solar power, micro-nuclear reactors, and thorium reactor are more relevant in the future to include as the race and challenges for energy supply increase and the technology evolve rapidly. Acknowledgements This research was supported by School of Business and Management, Institut Teknologi Bandung and Research and Technology Center of Pertamina (PERSERO). References [1] K. Sani, M. Siallagan, U. S. Putro, and K. Mangkusubroto, “Policy Development for the Energy Mix in Indonesia Using System Dynamics. Global Journal Business Social Science Review 2017,” 2017. [Online]. Available: http://gatrenterprise. com/GATRJournals/vol5_2017_issue3.html. [2] “Presidential Regulation on National Energy Policy No. 5,” 2006. [Online]. Available: http://www.iea.org/publications/ freepublications/publication/Indonesia2008.pdf. [3] “U.S. Energy Information Administration,” 2015. [Online]. Available: https://www.csis.org/events/us-eias-annual-energy- outlook-2015. it is succeeded in building an initial model structure that is able to simplify the past complex energy supply dynamics and to imitate the historical trends, except for some incidental fluctuations. While comparison of the two scenarios demonstrates that under the better second scenario, great improvement could have been achieved in all sectors of the national energy realm should it have been well implemented in the past. Both are important as basis to understand the model structures and non- linear behaviors of the complex energy supply system over the time frame for future policy design. How interactions of the different variables of the internal feedback loops, table functions and time delays have jointly mathematically been shaping the different behaviors of the complex energy issues and problems. It is obvious that among the key variables that cause unproductive energy supply system behavior include poor energy mix and uncommitted green energy policies, dominant oil imported, low public participation in NRE investment, and long delays in policy designing, communication and implementation. By gaining insight the model behaviors, policy instruments may be designed so as to lead the energy supply system to behave productively controllably. This model is for those reasons potentially reliable as an alternative model for building a new Energy Supply Mix Vision if the main challenge facing the modelling may be resolved. It includes investigating the factors influencing the long-terms supply and demand of energy as key variables that potentially dictate the energy industry in the future, to gain insight into their uncertainties and to simply formulate them into the model. 6. Conclusion It is finally concluded that the qualitative model identifies the presence of a “vicious circle” in the energy supply system in Indonesia, in which oil sector (and other fossil energies) are prioritized in meeting the national energy demand and treat new and renewables sector as the last resort in the energy portfolio management. By identifying and mapping these historical barriers, a better alternative models may be developed. The quantitative model succeeds in building model structures and gain insight the non-linear behaviors of the complex energy supply system over the set timeframe as basis for future policy design in Indonesia. The initial simulation runs are quite encouraging in that it succeeds in developing, comparing and contrasting two scenarios represent the old fashion system as is (scenario 1 old past) and new fashion model (scenario 2 new past) as if all selected best practices were implemented. Compared to the past statistics, the results of the simulation runs are quite promising, the old past model (Scenario 1 http://gatrenterprise.com/GATRJournals/vol5_2017_issue3.html http://gatrenterprise.com/GATRJournals/vol5_2017_issue3.html http://www.iea.org/publications/freepublications/publication/Indonesia2008.pdf http://www.iea.org/publications/freepublications/publication/Indonesia2008.pdf 48 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics [17] J. Zhou and Y. J. Liu, “The Method and Index of Sustainability Assessment of Infra structure Projects Based on System Dynamics in China,” Journal of Industrial Engineering and Management, 2015. [Online]. Available: http://dx.doi. org/10.3926/jiem.1496. [18] S. Armenia, F. Baldoni, D. Falsini, and E. Taibi, “A System Dynamics Energy Model for a Sustainable Transportation System,” “Sapienza” University of Rome, CATTID, Rome (Italy), 2006. [Online]. Available: https://www.researchgate.net/ publi cation/257920548. [19] K. C. Chi, W. J. Nuttak, and D. M. Reiner, “Dynamics of the UK natural gas industry: System dynamics modelling and long- term energy policy analysis,” Technological Forecasting & Social Change, 2008. [20] P. S. Bodger and D. G. May, “A System Dynamics Energy Model of New Zealand,” Technological Forecasting and Social Change, 1992. [Online]. Available: https://www.sciencedirect. com/science/article/pii/004016259290018O. [21] E. G. R. Davies and S. P. Simonovic, “Energy Sector for the Integrated System Dynamics Model for Analyzing Behaviour of the Social-Economic-Climatic Model,” ISSN, 2009. [Online]. Available: https://ir.lib.uwo.ca/wrrr/26/. [22] J. Jupesta, Y. Harayama, and G. Parayil, “Sustainable business model for biofuel industries in Indonesia,” Sustainability Accounting, Management and Policy Journal, Emerald Group Publishing Limited, 2011. [Online]. Available: http://dx.doi. org/10.1108/20408021111185394. [23] N. Ghaffarzadegan, J. Lyneis, and G. P. Richardson, “How Small System Dynamics Models Can Help the Public Policy Process,” Rockefeller College of Public Affairs and Policy, University at Albany, SUNY , Sloan School of Management, [24] I. Soares, P. Ferreira, and P. A. Ostergaard, “Energy markets, financing and accounting —international conference on energy & environment special issue from 2017,” International Journal of Sustainable Energy Planning and Management Vol. 15, 2018. [Online]. Available: dx.doi.org/10.5278/ijsepm.2018.15.1. [25] W. De Broeck, “Crowdfunding platforms for renewable energy investments: an overview of best practices in the EU,” International Journal of Sustainable Energy Planning and Management Vol. 15, 2018. [Online]. Available: dx.doi. org/10.5278/ijsepm.2018.15.2. [26] J. D. Sterman, “Business dynamics: system thinking and modeling for a complex world,” Boston, MA Irwin McGraw- Hill, 2000. [Online]. Available: https://www.amazon.com/ BusinessDynamicsSystems-T. [4] “Business Monitor International Ltd. Indonesia Electric Mix from 2014 to 2023.” [Online]. Available: https://store. bmiresearch.com/indonesia-power-report.html. [5] “Handbook of Energy & Economic Statistics of Indonesia 2016,” ISSN 2528-3464, 2016. [6] “National Energy Policy (Government Regulation No. 79),” 2014. [Online]. Available: https://www.iea.org/ policiesandmeasures/pams/indonesia/name-140164-en.php. [7] S. Yokohama, “Improving Renewable Energy Technology Diffusion in Indonesia Using System Dynamics,” 2015. [8] K. J. Skipka and L. Theodore, “Energy resources: availability, management, and environmental impacts,” CRC Press, Tailor & Francis Group, 2014. [9] S. C. Bhattacharyya and G. R. Timilsina, “A review of energy system models. International Journal of Energy Sector Management,” 2010. [Online]. Available: https://doi.org/ 10.1108/17506221011092742. [10] F. Rogan, C. J. Cahill, H. E. Daly, D. Dineen, J. P. Deane, C. Heaps, M. Welsch, M. Howells, M. Bazilian, and B. P. O. Gallachoir, “LEAPs and Bounds - an Energy Demand and Constraint Optimised Model of the Irish Energy System. Energy Efficiency, Springer,” 2013. [Online]. Available: http:// doi.org/10.1007/s12053-013-9231-9. [11] A. M. Bassi, “A context-inclusive approach to support energy policy formulation and evaluation,” Springer-Verlag, 2010. [12] C. de Castro, J. M. Luis, and M. Margarita, “World Energy- economy scenarios with system Dynamics modeling,” University of Valladolid, Spain. [Online]. Available: https://www. crisisenergetica.org/ficheros/deCastroScenariosASPO7.pdf. [13] S. Rigot and S. Demaria, “Potential impediments to long-term and low-carbon investment: the international accounting standards at stake,” International Journal of Sustainable Energy Planning and Management Vol. 15, 2018. [Online]. Available: dx.doi.org/10.5278/ijsepm.2018.15.3. [14] N. C. Figueiredoa and P. P. da Silva, “The price of wind power generation in Iberia and the merit-order effect,” International Journal of Sustainable Energy Planning and Management Vol. 15, 2018. [Online]. Available: dx.doi.org/10.5278/ijsepm.2018.15.4. [15] S. Chowdhury and K. C. Sahu, “A System Dynamics Model for the Indian Oil and Gas Exploration/Exploitation Industry. Technological Forecasting and Social Change,” Elsevier Science Publishing Co. Inc., 1992. [Online]. Available: https://www. sciencedirect.com/science/article/pii/0040162592900733. [16] H. Flynn, D. Breger, A. Belden, A. Bier, C. Laurent, N. Andrews, and W. Rickerson, “System Dynamics Modeling of the Massachusetts SREC Market. Sustainability,” ISSN, 2010. [Online]. Available: http://doi.org/10.3390/su2092746. https://store.bmiresearch.com/indonesia-power-report.html https://store.bmiresearch.com/indonesia-power-report.html https://www.iea.org/policiesandmeasures/pams/indonesia/name-140164-en.php https://www.iea.org/policiesandmeasures/pams/indonesia/name-140164-en.php https://doi.org/10.1108/17506221011092742 https://doi.org/10.1108/17506221011092742 https://www.crisisenergetica.org/ficheros/deCastroScenariosASPO7.pdf https://www.crisisenergetica.org/ficheros/deCastroScenariosASPO7.pdf http://dx.doi.org/10.5278/ijsepm.2018.15.3 http://dx.doi.org/10.5278/ijsepm.2018.15.4 https://www.sciencedirect.com/science/article/pii/0040162592900733 https://www.sciencedirect.com/science/article/pii/0040162592900733 http://dx.doi.org/10.3926/jiem.1496 http://dx.doi.org/10.3926/jiem.1496 https://www.researchgate.net/publication/257920548 https://www.researchgate.net/publication/257920548 https://www.sciencedirect.com/science/article/pii/004016259290018O https://www.sciencedirect.com/science/article/pii/004016259290018O http://dx.doi.org/10.1108/20408021111185394 http://dx.doi.org/10.1108/20408021111185394 http://dx.doi.org/10.5278/ijsepm.2018.15.1 http://dx.doi.org/10.5278/ijsepm.2018.15.2 http://dx.doi.org/10.5278/ijsepm.2018.15.2 https://www.amazon.com/BusinessDynamicsSystems-T https://www.amazon.com/BusinessDynamicsSystems-T International Journal of Sustainable Energy Planning and Management Vol. 18 2018 49 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto Table of key variables, units and functions in the initial quantitative modelling of indonesia today Variable Unit Function population person birth - death birth person/year population * birth rate death person/year population * death rate electricity growth per capita KWh/person 930 + RAMP(7.998.e-05, 2000, 215) total electricity demand GWh electricity growth per capita * population total demand for non-electric power GWh non-electric power growth per capita * population total energy demand GWh total electricity demand + total demand for non-electric power fossil fuel total capacity GWh power generation by oil + power generation by gas + power generation by coal NRE total capacity GWh power generation by biofuel + power generation by geothermal + power generation by hydro + power generation by solar pv + power generation by nuclear dependency Dmnl fossil fuel total capacity/ (fossil fuel total capacity - NRE total capacity) energy mix Dmnl NRE total capacity/ (fossil fuel total capacity + NRE total capacity) total power generated GWh fossil fuel total capacity + NRE total capacity total domestic power available GWh (NRE total power capacity-total biofuel exported)+ (power generation by coal-total coal exported)+ (power generation by gas-total LNG exported)+(power generation by oil-total crude exported)+oil generation by imported+power generation by biomass total power exported GWh total crude exported + total LNG exported + total coal exported + total biofuel exported total crude exported GWh fraction of crude exported * power generation by oil total LNG exported GWh fraction of LNG exported * power generation by gas total coal exported GWh fraction of coal exported * power generation by coal total biofuel exported GWh fraction of biofuel exported * power generation by biofuel oil generation by imported GWh additional energy from imported oil = cf of imported oil*additional of imported oil additional of imported oil GWh fraction of imported oil*domestic supply and demand gap oil imported change Dmnl STEP (Height to oil import change, Time to oil import change) domestic supply and demand gap GWh total energy demand - total domestic power available power capacity needed GW capacity conversion * domestic supply and demand gap *1.3 additional power capacity GW power capacity needed * (government ability + public investment) delay of power development GW DELAY1(additional power capacity, delay time1) delay time1 Year gf4 + gf5** delay of policy implementation Year DELAY3(energy strategy and policies, delay time2) delay time2 Year gf1 + gf2 + gf3** additional capacity of fossil producers GW (delay of power development - additional capacity of additional capacity of NRE GW (energy mix policy + green energy policy) * delay of poor energy mix policy Dmnl IF THEN ELSE (energy mix <0.25, 0.12, 0.12) uncommitted green energy policy Dmnl IF THEN ELSE (CO2 emission of fossil electric generating > 50, 0, 0) APPENDIX A (Continued) 50 International Journal of Sustainable Energy Planning and Management Vol. 18 2018 Indonesia energy mix modelling using system dynamics Table of key variables, units and functions in the initial quantitative modelling of indonesia today (Continued) additional of oil producers GW (fraction of oil producers*additional capacity of fossil producers)*(1+Oil supply change) oil supply change Dmnl STEP (Height to oil supply change, Time to oil supply change) additional of gas producers GW fraction of gas producers * additional capacity of fossil producers gas supply change Dmnl STEP (Height to gas supply change, Time to gas supply change) fraction of LNG exported Dmnl LNG exported fraction Avg*(1+LNG exported fraction change) LNG exported fraction change RAMP(–0.01, 2000, 2015) additional of coal producers GW fraction of coal producers * additional capacity of fossil producers coal supply change Dmnl RAMP (0.392, 2009, 2015) additional of biomass producers GW (fraction of biomass*additional capacity of fossil producers)* (1+Biomass supply change) biomass supply change Dmnl RAMP (0.8, 2009, 2015) additional of biofuel pp GW (fraction of biofuel * additional capacity of NRE)*(1+biofuel supply change) biofuel supply change Dmnl RAMP (3.5, 2005, 2015) additional of geothermal pp GW (fraction of geothermal * additional capacity of NRE)*(1+geothermal supply change) geothermal supply change Dmnl STEP (Height to geothermal supply change, Time to geothermal supply change) additional of hydro pp GW (fraction of hydro * additional capacity of NRE)*(1+hydropower supply change) hydropower supply change Dmnl RAMP (1.5, 2005, 2015) additional of solar pv pp GW (fraction of solar pv * additional capacity of NRE)*(1+solar pv supply change) solar pv supply change RAMP (1.5, 2005, 2015) additional of nuclear pp GW fraction of nuclear * additional capacity of NRE additional energy from oil GWh/year additional of oil producers * cf of oil supply additional energy from gas GWh/year additional of gas producers * cf of gas supply additional energy from coal GWh/year additional of coal producers * cf of coal supply additional energy from biofuel pp GWh/year additional of biofuel pp * cf of biofuel additional energy from geothermal pp GWh/year additional of geothermal pp * cf of geothermal additional energy from hydro pp GWh/year additional of hydro pp * cf of hydro additional energy from solar cs pp GWh/year additional of solar cs pp * cf of solar additional energy from nuclear power GWh/year additional of nuclear pp * cf of nuclear power CO2 emission of fossil power generation mmtonnes emission by coal + emission by oil + emission by gas emission by coal mmtonnes power generation by coal * factor emission coal emission by oil mmtonnes power generation by oil * factor emission oil emission by gas mmtonnes power generation by gas * factor emission gas ** = contributors of delay time variables corresponding to Ghaffarzadegan's characteristic impediment in policy designing, communication and implementation gf1 = "under estimated policymakers" variable in policy designing gf2 = "scapegoat minded perspective" variable in policy designing gf3 = "political link & lobbying" variable in policy designing and policy communication gf4 = "policy acceptance" variable in policy implementation gf5 = "rate of trials and errors" variable in policy implementation International Journal of Sustainable Energy Planning and Management Vol. 18 2018 51 Kartono Sani, Manahan Siallagan, Utomo Sarjono Putro and Kuntoro Mangkusubroto Table of Initial Values of Variables for Simulation Modelling of Indonesia’s Past Energy (2000–2015) Variable Unite Initial Value population person 205,840,000 birth rate 1/year 16.72/1000 death rate 1/year 6.73/1000 capacity conversion GW/GWh 1/(365*24) government ability Dmnl 0.4 public investment Dmnl 0.6 gf1+gf2+gf3** Year 0.5+0.2+0.3 gf4+gf5** Year 0.4+0.6 fraction of oil producers Dmnl 581 Fraction of oil exported Dmnl 0.43 height to oil supply change Dmnl -0.5 time to oil supply change Year 2008 fraction of gas producers Dmnl 193 height to gas supply change Dmnl -1.1 time to gas supply change Year 2008 LNG exported fraction avg Dmnl 0.27 fraction of coal producers Dmnl 75 Fraction of coal exported Dmnl 0.793 fraction of biomass Dmnl 50 fraction of biofuel Dmnl 0.1 (0.3) fraction of geothermal Dmnl 0.032 height to geothermal supply change Dmnl 4 time to geothermal supply change Year 2006 fraction of hydro Dmnl 0.248 fraction of solar pv Dmnl 0.001 (0.131) fraction of nuclear Dmnl 0.0001 cf of oil supply (GWh/GW*year) -0.525 cf of gas supply (GWh/GW*year) -0.674 cf of coal supply (GWh/GW*year) 62 cf of biomass (GWh/GW*year) 5.25 cf of biofuel (GWh/GW*year) 0.1 cf of geothermal (GWh/GW*year) 0.032 (0.32) cf of hydro (GWh/GW*year) 0.248 cf of solar pv (GWh/GW*year) 0.001 cf of nuclear (GWh/GW*year) 0.0001 factor emission coal mmtonnes/GWh 318.37/106 factor emission oil mmtonnes/GWh 249.65/106 factor emission gas mmtonnes/GWh 181.08/106 power generation by oil GWh 879,426 power generation by gas GWh 1,167,010 power generation by coal GWh 624,025 power generation by biomass GWh 437,233 power generation by biofuel GWh 0 power generation by hydro GWh 42,908 power generation by geothermal GWh 16,308 power generation by solar pv GWh 0 power generation by nuclear GWh 0 oil generation by imported GWh*Year 133,599 height to oil import change Dmnl 1.1 time to oil import change Year 2005 APPENDIX B