i
Master’s Thesis
Research Title:
Natural Gas as a Transition fuel: Does Natural Gas Help or Hinder a
Clean Energy Transition?
Personal Information Author: Cem Gürsan
Program: Business Analysis and Modeling Student Number: s1017951
Supervisors
1) Vincent de Gooyert 2) Rob van der Heijden
ii Index 1) Introduction 1 2) Research Objective 3 a. Research Questions 3 3) Theoretical Background 4 4) Methodology 7
a. Intended Data Collection Method 9
b. Access to Sources 9
c. The underpinning of the sample 9
d. Data Analysis 10
i. Literature Review 10
ii. Theory Exploration and Causal Loop Diagram 11
iii. Computational Model and Synthesis 11
5) Results 12
a. Energy Reliability 16
i. Positive Direct Effects 16
b. Crowd-out and Carbon lock-in 16
i. Uncertain Indirect Effects 16
c. Energy Costs 19
i. Positive Direct Effects 19
ii. Negative Indirect Effects 19
1. Transition Costs 19
2. Energy Demand 20
3. Electricity Generation Costs 21
d. Need for Clean Energy 21
i. Positive Direct Effects 21
ii. Negative or Uncertain Indirect Effects 22
1. CO2 Emissions 22
2. Other Emissions 24
3. Fossil To Gas 24
4. Other Environmental Effects 26
5. Combined CLD of natural gas’ direct and indirect effects 26
e. Computational Model 29
i. Boundary of the Computational Model 29
ii. The structure and equations of the Computational Model 30
iii. Analysis of Model Outcomes 32
1. Natural Gas might crowd-out renewables in the medium term. What about the future? 32
2. Sustainable Future is possible 33
3. Costs of Technologies: Good for the investors or Good for all? 34
4. Timing matters for a sustainable future 37
5. Renewables are not enough: CCS can help with already existing carbon-emitting sources 38 6. A counter-intuitive discovery: Cheap Natural Gas can obstruct sustainability goals 39
6) Discussions 41
a. Limitations and Further Research Implications 45
7) Conclusions 45
8) Research Ethics 50
9) Acknowledgements 50
10) References 51
iii
Table of Contents
Figure 1 Filtering method for literature Table 1 Table of Literature Categorization
Figure 2 Coding Tree - Natural Gas Helping or Hindering Clean Energy Transition Figure 3 Learning Effect CLD
Figure 4 Directed Policy Efforts CLD Figure 5 Crowd-out CLD
Figure 6 Rebound Effect CLD Figure 7 Direct Emissions CLD
Figure 8 Transition from Fossils to Gas CLD Figure 9 Shifting the burden system archetype Figure 10 Indirect Emissions CLD
Figure 11 Spillover Effect CLD Figure 12 Combined CLD Table 2 Bull's eye diagram
Figure 13 Natural gas crowds-out in the medium term.
Figure 14 Renewable Energy to Total Electricity Production Ratio Figure 15 World stays under 450 PPM levels
Figure 16 Subsidy Ratio Graph for Renewables for figure 16 Figure 17 Costs of technologies and Installed Capacities
Figure 18 Model Outcomes for different learning rate factors for Renewable energy Figure 19 Wind energy fare better than solar energy due to low costs
Figure 20 Starting the transition late Figure 21 Starting transition early
Figure 22 "Early Transition” & "Late Transition" on 2100 horizon Figure 23 CCS can help with a successful transition
Figure 24 Cheap Natural Gas might obstruct Clean Energy Transition Figure 25 Data used and Pathways analyzed within the model Figure 26 Scenario Plan
Figure 27 Renewable Capacity Sector Figure 28 Natural Gas Capacity Sector Figure 29 Fossil Capacity Sector Figure 30 Energy Discrepancy Sector Figure 31 Learning Rate Sector Figure 32 Emission Calculation Sector Figure 33 Need for Clean Energy Sector Figure 34 Cost Perception Sector
iv
List of Abbreviations
SD - System Dynamics
CLD - Causal Loop Diagram
CCS - Carbon Capture and Storage
SFD - Stock and Flow Diagram
Intergovernmental Panel on Climate Change - IPCC
International Energy Agency - IEA
Energy Information Administration - EIA
Parts per million - PPM
Tons of Coal Equivalent - TCE
British Columbia - B.C.
Greenhouse Gases- GHG
1 1) Introduction
In the aims of achieving a more efficient and cleaner world, governments, intergovernmental organizations and businesses have been motivated towards investing in clean energy transitions in the electric power industry (Gothenburg European Council, 2001). Securing affordable, uninterrupted, and environmentally sustainable electricity production has been one of the most emphasized topics of the world because of climate change, constant energy demand growth, fossil fuel’s variable energy prices, and hardship in sustained procurement of fossil fuels (IPCC, 2018). Natural gas is considered as a transition fuel. Transition fuel can be defined as “a next step in the choice of fuel but not the destination”. Natural gas is considered as the next step in our fuel choice because it has positive effects on clean energy transition by helping with renewable energy sources’ critical challenges. However, natural gas cannot be our destination fuel choice because it is still a fossil fuel and has negative effects that are counter to the sustainability and economic goals. The uncertainties and complexity of the issue divide the relevant stakeholders (policy-makers, the scientific community, NGOs, and even the populace) into rigid conflicting factions. One view on natural gas defends that natural gas can help with the current challenges of renewables (Smil, 2005; Colombo et al., 2016; Flavin, 2008), and the other view opposes this by warning us that natural gas might create a continued dependency to fossil fuels (Unruh, 2000; Baron, 2013; Bosman, Loorbach, Frantzeskaki, & Pistorius, 2014; Dupont, & Oberthür, 2012).
Both the supporters and opposers describe natural gas’ possible positive and negative effects on clean energy transition in their studies. However, the debate exists because their arguments seem to conflict with each other. This conflict in the literature might hurt the clean energy transition by delaying the convergence on what future policies should be. In order to produce robust future policies on clean energy transition, we must extend our current understanding of the debate first. Hence, this research will focus on the evaluation of the existing schools of thoughts on the issue by attentive literature review. Secondly, when working with theories, it is vital to have a relevant framework to investigate and synthesize different scientific and expert works done by different authors on a single platform. This single platform would enable us to merge different theories on the same frame with the same syntax.
2
Performing the research on a single platform which enables communication and linking between thoughts can help to extend our understanding, and thus provide a way to resolve the debate. Thus, this research intends to use a relevant framework for the theoretical exploration and synthesis in which different theories can be investigated to address the conflicts in the natural gas debate.
A critical challenge for the conflicts of using natural gas as a transition fuel comes from the interconnectedness between the indirect and direct effects. In addition, the complexity of this issue is found out to be caused by the separation of time in between our decisions and the system’s reactionary outcomes to our decisions (time delays) and the existence of circular causal relationships of critical factors affecting the system (feedback loops). Time delays make it harder for decision makers to accurately pinpoint which one of their decisions have created the desired or undesired effect on the system. Feedback loops can be defined as the following: “A system element affects the total system. Then, the system affects that system element in turn”. Likewise, indirect and direct effects of natural gas are causally related to each other, and thus creating these feedback effects. The existence of feedback loops in a system makes it harder for decision makers to determine what the whole effect of a decision would be to the system. The feedback loops have two types: Reinforcing and Balancing. Balancing feedback loops forces the “system to come back to the position that it started from” (de Gooyert, Rouwette, van Kranenburg, Freeman, & van Breen, 2016: 137). Contrarily, reinforcing feedback loops amplifies the system even with the smallest “intervention inflicted on them”. (de Gooyert et al., 2016: 137). When these dynamics are overlooked, it is natural for conflicts such as we have on natural gas to occur.
As discussed, the debate on natural gas should be analyzed in a dynamic framework. System Dynamics (SD) is a systemic way of thinking to handle dynamic problems where the complexity is caused by the interconnectedness of the elements, feedbacks and time delays (Meadows, 2008; Sterman, 2000). The System Dynamics modeling approach can help us build simulation models that we can use to understand this complexity with a systemic lens. The System Dynamics approach can be utilized for building theoretical simulation models to explore
3
and synthesize theories (de Gooyert, 2018) where this debate can be understood with all its dynamics.
Since theoretical SD methodologies can help with understanding the dynamics behind the indirect and direct effects of natural gas as a transition fuel, utilizing SD is a fitting candidate for this research’s methodology. This research will use a theoretical simulation modeling approach to explore and synthesize earlier studies on direct and indirect effects of natural gas, investigate whether these earlier studies are internally consistent, and explore the implications of combining these earlier studies, and thus hopes to contribute to our understanding on the issue by revealing the conditions where natural gas might help or hinder the clean energy transition.
2) Research Objective
Renewables have many challenges as we try to pursue a clean energy transition. Using natural gas in the clean energy transition has direct positive effects which can help the renewables towards clean energy transition. However, using natural gas as a transition fuel might also have unintended indirect effects which can be contrary to sustainability goals. In order to overcome these challenges, this research’s objective is to contribute to the understanding of the dynamics between direct and indirect effects of investing in natural gas as a transition fuel in the electricity production industry with the help of system dynamics (SD) models and to contribute to the debate by analyzing the literature on the issue (Appendix 1), and thus, to provide a framework where different theories on the issue will be explored and synthesized to reveal the conditions where natural gas help or hinder clean energy transition.
a) Research Questions
1) What do the existing conflicting factions claim when it comes to the direct and indirect effects of natural gas as a transition fuel in terms of helping or hindering the energy transition?
2) Are these different lines of theories/thoughts internally consistent in their claims when the dynamics between direct and indirect effects of natural gas are taken into account?
3) Under what conditions can the intended transition fuel responsibility of natural gas be achieved without the unintended indirect effects?
4 3) Theoretical Background
Natural gas, due to having approximately half CO2 polluting effects vis-à-vis other
fossil fuels (Ahmed, & Cameron, 2014; Arent, Logan, Macknick, Boyd, Medlock, O'Sullivan, Edmonds, Clarke, Huntington, Heath, & Statwick, 2015; Colombo, El Harrak, & Sartori, 2016; Heath, & Statwick, 2015; Smil, 2005; Stephenson, Doukas, & Shaw, 2012), has been considered as a transition fuel before renewable energy’s technological viability can step up in producing secure and sustained energy (Smil, 2005; Colombo et al., 2016; Flavin, 2008). The challenges for renewable energy mentioned frequently in the literature are the intermittent energy production due to the variable nature of renewable sources, current inadequacy of energy storage options and high costs (Ahmed, & Cameron, 2014, Arent, et al., 2015; Baron 2013).
Renewables provide intermittent energy (Baron, 2013; van Foreest, 2010; van Foreest, 2011; Boersma, & Jordaan, 2017). Solar panels can only generate electricity when we have adequate solar source and windmills can only generate electricity when we have an adequate amount of wind. Since human beings cannot control these forces, we are dependent on the amount of energy that nature provides us when using these technologies. A solution to this problem might be in storing the energy. Currently, these energy storage options are not commercially viable (Colombo et al., 2016; Smil, 2015; Ahmed, & Cameron, 2014) or environmentally sustainable as a result of their chemical pollution (Larcher, & Tarascon, 2015). Another challenge for the renewables is their high costs. Since renewables are an emerging technology compared to other types of fuels, natural gas stands out as a transition fuel because of its economic viability (Colombo et al., 2016; Flavin, 2008; Van van Foreest, 2010; Smil, 2015; Boersma, & Jordaan, 2017) and less polluting effects compared to other fossils (Colombo et al., 2016; Flavin, 2008; Boersma, & Jordaan, 2017).
Natural gas can be used with renewable energy sources in a synergetic manner to overcome these difficulties (Ahmed, & Cameron, 2014; Arent et al., 2015). According to some experts, a shift from other fossil fuels to natural gas is treated as an environmental friendly transition as a result of natural gas’ less polluting effects (Smil, 2005; Baron, 2013; BP Energy Economics, 2018; Colombo et al., 2016; Flavin, 2008; Shell Scenarios, 2018); however, there is still a fiery debate on natural gas as a transition fuel because of the environmental uncertainties
5
and future repercussions of investing in natural gas (Arent et al., 2015; Baron 2013; Elliston, MacGill, & Diesendorf, 2014; Stephenson et al., 2012). If this transition is not well handled, investing in natural gas without a future policy strategy may lead to fossil lock-in (Unruh, 2000; Baron, 2013; Bosman, Loorbach,Dupont, & Oberthür, 2012; Frantzeskaki, & Pistorius, 2014). In other words, investing in natural gas might lead to self-perpetuating decisions and crowd-out renewable energy technologies or zero-carbon technologies.
According to the IPCC’s 2018 report (IPCC, 2018), implementing appropriate climate policies in a timely manner is gaining importance every day; thus, it is vital to figure out whether investments in natural gas as a transition fuel are done at the expense of other more sustainable technologies such as renewables or to the advantage of leaving more polluting fossil fuels such as coal and oil in the ground.
Uncertainties on natural gas’ indirect effects have led to the current debate on the topic. Thus it makes it worthwhile to investigate whether investing in natural gas for its beneficial possible direct effects on clean energy transition might lead to unintended indirect effects. Before continuing, it is important to define what direct and indirect effects mean.
Effects have causes. We take decisions because we have a cause which we would like to address. Then, these decisions’ effect provides an immediate response from the system. These immediate responses are called direct effects (Miller, 2006). However, these effects we have created also become causes of other effects (Miller, 2006). Indirect effects represent the response of the system to the direct effects (Miller, 2006). Direct effects can be classified as local, immediate and foreseeable. Whereas indirect effects are global, delayed, and most of the time unforeseen. In dynamic systems such as clean energy transition, it is possible to see what direct effects will be; however, it is not that simple to grasp how the indirect effects will play out because of the interconnectedness of the system, time delays, and feedback loops.
Studies on natural gas as a transition fuel mention the utilization in the medium term to enable a transition to renewables in the long term (Baron, 2013; Dupont, & Oberthür, 2012). Natural gas seems like a viable transition fuel because, when applied, its direct effects could solve many of the challenges of renewable energy such as energy variability and high costs. However, in the long-term use of enabling natural gas as a transition fuel might have
6
environmental repercussions (Stephenson et al. 2012) and create a lock-in to fossil fuels as an indirect effect (Baron, 2013; Dupont, & Oberthür, 2012; Unruh, 2000; Frantzeskaki, & Pistorius, 2014).
There have been several SD studies on the electricity production industry and clean energy transition topic. For instance, Moxnes (1990) focuses on substitute fuel prices and consumer biases towards substitution fuels. By using simulation models to calculate energy prices, Moxnes (1990) has discovered that there was a bias towards coal possibly due to employment reasons and reduction on import dependence. Another important finding was that natural gas has a “negative premium” compared to oil (Moxnes, 1990: 62). This study shows interesting results; however, it was done in 1990. It is also relevant now to check if this bias towards coal still exists after the increased urgency of climate change. Ford (1997) has made an extensive literature review on SD and electricity production industry. His aim was to collect relevant papers on electricity production industry to show that the feedback effect exists in electricity production systems and it is fitting to use SD models in the electricity production industry. Next, Tan, Anderson Jr., Dyer, and Parker (2010) explore the use of decision trees and the SD model by using an illustrative example in alternative energy production. The authors in this research (Tan et al., 2010) are aiming to create a method to transform the SD model into a decision tree. Tan et al. (2010) mention that the electricity production industry is highly uncertain. Also, in order to come up with viable policies, SD needs to be implemented with stochasticity. This effort is achieved by using SD models with Monte Carlo simulations to embed stochasticity on the electricity production industry. Moreover, Tan et al. (2010) discovered that the value of alternative energy technology projects is largely determined by fossil fuel prices and capital equipment costs. Lastly, de Gooyert et al. (2016) focus on policy resistance in sustainability transitions with a case on Dutch energy transition. De Gooyert et al. (2016) suggest that SD can be used as a complementary tool to overcome policy resistance which is a frequent problem in sustainability transitions. In addition to handling policy resistance with SD, de Gooyert et al. (2016) also focuses on the Dutch clean energy transition case at hand. Although the research’s case is on a national level, it has relevant insights for the transition dynamics at large.
7
To summarize the SD literature review, the topics that the literature is interested in are relevant and useful to the study at hand: using SD in the electricity production industry. However, there is also a gap in the SD literature. Currently, there exists no work that is solely focusing on natural gas as a transition fuel debate by building a simulation model to explore and synthesize theories. This research intends to fill that gap because it can contribute valuable insights to the current debate on natural gas.
4) Methodology
Since this research intends to synthesize and explore the implications of combining various views on natural gas in the existing academic literature, I will take a theoretical quantitative discovery SD modeling approach to answer the research questions (de Gooyert, 2018; Axelrod, 2003; Sastry, 1997).
Theoretical models aim to provide a framework where the ongoing scientific debate can be seen with clarity and experimented upon (de Gooyert, & Größler, 2019). Including too many aspects in a theoretical model might overshadow the real purpose by preventing a clear and convincing contribution to a scientific debate (de Gooyert, & Größler, 2019). Hence, in theoretical SD studies, “only critical aspects of the system” which actually can help us to understand the existing theories are chosen as the boundary of the system (de Gooyert, & Größler, 2019: 581).
To answer the research question 1, I will do a literature review on the issue of “natural gas as a transition fuel” in order to capture the academic community’s leading thoughts. I intend to categorize the debate on natural gas as a transition fuel by summarizing the existing literature and different schools of thoughts on the subject. I will categorize the natural gas’ helping and hindering effects into groups and subgroups according to the document’s topic of interest and their claims. This research has two aims in this categorization. Firstly, I will use the categorization to create a snapshot view of the debate to contribute to the understanding of the issue. Secondly, this categorization effort will provide the foundation to build the computational model. The outcome of this analysis will be presented in a table where natural gas’ direct and indirect effects can be observed under different themes.
8
For research question 2, I will combine these themes from literature in a single framework where these themes’ counteracting effects can be seen clearly. In doing this, I intend to check the internal consistency of these different lines by showing how direct and indirect effects are working against each other. To achieve this, we need a dynamic framework where counteracting direct and indirect effects of natural gas can be seen in relation to each other. The internal consistency analysis is not intended for building confidence for or invalidating the existing views and thoughts, rather it will be used as a critical method to investigate the dynamics of these thoughts on natural gas. Causal Loop Diagrams (CLD) under the umbrella of SD can represent how different variables of a system are interconnected. Direct and indirect effects of natural gas can be shown in a single CLD, especially the feedback loops. Reinforcing feedback loops are presented with an “R” letter and balancing feedback loops are presented with a “B” letter in the CLDs. Using this frame will provide me to depict the emerging themes in relation to each other with all its dynamics.
“Causal loop diagrams are visual representations of dynamic influences and inter-relationships that exist among a collection of variables”.(Spector, Christensen, Sioutine, & McCormack, 2001: 536).
Furthermore, CLDs can be a suitable approach to be used as a preliminary step to build simulation models (Spector et al., 2001). In addition, CLDs are appropriate for “complex domains supported by system dynamics modeling and simulation techniques” (Spector et al., 2001: 536). Therefore, using CLDs will also help me to start with the preliminary work for building the computational model.
To answer research question 3, I will embed these thoughts in SD model where I can explore and synthesize them with the purpose of contributing to the debate. In order to achieve this, the research strategy is primarily building a “conceptual virtual laboratory” (de Gooyert, 2018) by developing dynamic theories with the use of simulation modeling (Davis, Eisenhart, & Bingham, 2007). Similar to how Sastry (1997) discusses a specific use of the simulation modeling; in the conceptual virtual laboratories, I will use the simulation model to discover the assumptions and outcomes of different theories raised by different authors on a single specific topic (Sastry, 1997; Repenning, 2002). As de Gooyert (2018) states, one of the perks of simulation (in the conceptual laboratory) lies in the “exploration” and “synthesis” of different
9
should demonstrate the “causal hypotheses embodied in written theories of scientific endeavor and test whether they can generate the dynamics as those authors see them” (Sterman, & Wittenberg, 1999: 110).
a) Intended data collection method
Using theoretical models can be considered as a fitting approach for “deriving new insights from established variables and relationships” instead of “reporting new data, demonstrating the existence of a new variable, or testing the strength of a specific linkage between two variables” (Repenning, 2002: 110). Thus, theoretical models, due to their intention of laying the issue on the table, use existing studies as their empirical data (de Gooyert, & Größler, 2019). Data for creating the theoretical model will be based on secondary data:
documents of relevant academic literature, scenario analysis done by academics and experts,
policy and status reports of intergovernmental organizations, and future forecasts on electricity production by industry and climate modeling pioneers.
b) Access to sources
There are three platforms where I will search for relevant academic or expert books and academic papers. Radboud University Library Repository, Google Scholar, and climate & energy industries’ online platforms will be used to find relevant documents. I have used these three platforms to reach scientific journals, industry reports, scenario analysis, and simulation modeling literature on the topic (Appendix 1 & 2). I will search the documents on Google Scholar and then access to these documents via Radboud University Library Repository and Google Scholar. Industry reports and data will be accessed via their online platforms.
c) The underpinning of the sample
When the keywords “natural gas”, “clean energy” and “transition fuel” are searched in Google Scholar, 435 academic papers could be found (23.01.2018). This research included the relevant documents that could push the research to achieve its objective and the rest of the documents were filtered out. The filtering method is shown as a flow chart in Figure 1. The detailed filtering method is explained in Appendix 3.
In the end, there were 56 academic works left and these relevant works were recorded in Appendix 1 & 2. The researcher will use these academic works to build this research’s
10
theoretical model. There are two types of selected documents: topic-related documents and modeling-scenario analysis related documents. Topic related documents refer to the sources which focus on the issue of natural gas as a transition fuel. These documents will be the source for the different lines of thoughts that this research will investigate. They can be found in
Appendix 1. Modeling-scenario analysis related document refer to the sources which
implement various analysis methods such as simulation modeling, scenario analysis, and decision analysis on the topic of natural gas, electricity production, and clean energy transition. These documents will be the source for building the conceptual virtual laboratory. They can be found in Appendix 2. By including these two aspects, I intend to cover a sufficient area for reaching the research objective.
d) Data Analysis
i) Literature review
In order to formalize the research and theoretical model, I will use coding methods in the literature. There are existing themes under the subject of natural gas as a transition fuel which this research would like to address. Thus, a method to evaluate the existent thoughts and theories is needed. Deductive coding can help us “structure and format patterns” and it gives us
“comparative methods to evaluate the applicability of patterns” (Schadewitz, & Jachna, 2007: 2). Thus, this research will implement a deductive coding (top-down) approach. Reoccurring and debated effects of natural gas in the clean energy transition literature were selected as this
11
research’s themes. Coding scheme will evolve throughout the research progress to reflect the new themes as they emerge. These themes will be presented in a coding tree.
Coding tree will be used for categorizing the different factions in the natural gas debate and finding out the causal relations to build the theoretical model (Turner, Kim, & Andersen, 2014). After the literature review is done, I will be able to categorize different factions in the natural gas as a transition fuel debate.I will answer research question 1 (RQ1) by the qualitative analysis of the documents and present the results in a table which shows how different authors are categorized under emerging themes on natural gas.
ii) Theory Exploration and Causal Loop Diagrams
The coding tree will provide emerging themes in the literature. The table which answers the research question 1 will be composed of positive/uncertain/negative and direct/indirect effects of natural gas. This table and the corresponding quotes of different authors will provide the necessary causal relations from the literature to build the CLDs (Spector et al., 2001; Turner, Kim, & Andersen, 2014).
Every investigated theme with the CLD method will be presented individually first. This simple presentation will enable us to see a single effect clearly. When all the investigated effects are presented, an overview CLD will be built by combining them. I will answer research question 2 by the individual and combined CLDs which will enable us to perceive the dynamicity of the clean energy transition in an effective fashion, and thus reveal counteracting effects that are challenging the consistency of our arguments.
iii) Computational Model and Synthesis
I will use the table (RQ1) and CLDs (RQ2) when building the conceptual virtual laboratory (de Gooyert, 2018; Sastry, 1997). In order to keep the theoretical computational model simple to enhance its effectiveness, a handful of behaviors from research question 2 will be selected for the simulation. To answer research question 3, I will explore and synthesize existing theories to discover viable paths for a clean energy transition. Sensitive parameters of the simulation model will be chosen for further scenario and pathway analysis. By using 2 scenarios (early and late) with a range of different exogenous parameters, the pathways for sustainable and unsustainable futures will be investigated. Finding the similarities or the differences between
12
these conditions and their outcomes will provide the framework to answer how the natural gas might be utilized as a transition fuel in theory.
The answer to research question 3 (RQ3) will be achieved by exploring and presenting possible model outcomes. The model will be tested by structure confirmation, parameter confirmation and extreme condition tests (Barlas, 1996; Sterman, 2000).
5) Results
As it can be seen clearly in Table 1, many authors mention the positive direct effects of the natural gas for the clean energy transition. Natural gas can immediately help renewable energy technologies with their challenges under many themes. Scientists seem to be converging on the fact that natural gas can help with the transition with its direct effects.
However, it is also clear in Table 1 that when it comes to the indirect effects, the issue gets complicated. Indirect effects of natural gas under the need for clean energy, energy costs, crowd-out effect, and carbon lock-in themes shows some positive, but mostly negative and uncertain effects. Many authors mention natural gas’ both positive direct and uncertain (or negative) indirect effects in their studies (Arent et al., 2015; Ahmed, & Cameron, 2014; van Foreest, 2010; van Foreest 2011; Stephenson et al., 2012; Baron, 2013; Smil, 2015; Boersma, & Jordaan, 2017; Colombo et al., 2016). Issue’s complexity makes it hard to distinguish which dynamics will be dominant (positive direct or uncertain/negative indirect effects) when we are managing the clean energy transition.
The coding tree in Figure 2 has been used to categorize the thoughts and views in the literature about natural gas in the clean energy transition. There are 5 recurring themes in the literature including both direct and indirect effects of natural gas: Energy Reliability, Need for Clean Energy, Energy Costs, Crowd-out Effect, Carbon lock-in. In order to understand these themes, it is important to put them into context. This frame will also provide us with the emerging topics and categories on the issue to answer research question 1.
Energy Reliability theme focuses on the provision of energy in a reliable manner. There are 3 sub-categories under the energy reliability theme: intermittency, flexibility, and energy capacity. “Intermittency” theme refers to renewable energy sources’ intermittent energy production. Renewable energy sources can only produce energy when they have a resource
13
14
15
availability (wind for windmills and sun for solar panels etc.). Since we cannot control these forces, renewable energy sources can only provide variable energy outputs. Natural gas energy plants can be leveled to an energy demand if wanted. “Flexibility” theme refers to the ability to increase and decrease energy production at will. Renewable energy plants are dependent on renewable resources. They can only flex down if the energy demand is low, but not the other way around. Whereas natural gas plant has short on-off cycles and can address the intermittency and flexibility issue in a short time interval. “Energy Capacity” theme refers to the ability to provide sufficient energy according to the energy demand. Renewable energy plants lack this ability because of the challenges in energy storage and intermittency. Since the energy storage technologies are not commercially available for renewable energy plants, unused renewable energy goes to waste and cannot be used in the peak demand time. Natural gas can help with peak demands if renewable energy plants are not able to provide sufficient energy.
Need for Clean Energy theme focuses on the environmental aspects of the clean energy transition. There are 4 sub-categories under the need for clean energy theme. “CO2 emissions” theme refers to the environmental viability of an energy source in terms of its emitted CO2. “Other emissions” theme refers to the environmental viability of an energy source in terms of Methane, NOx, and SOx emissions. “Fossil to Gas” theme refers to the transition out of other carbon-intensive fossil technologies such as coal and oil power plants to less polluting natural gas. Last, the other environmental repercussions of natural gas are themed up in “Other Environmental Effects”.
“Energy Costs” theme categorizes various costs in the clean energy transition literature. There are 3 sub-categories under the energy cost theme. “Transition Costs” theme refers to the total costs which should be expected to do the clean energy transition. “Energy Demand” refers to cost caused by global energy demand. When global energy demand increases, we also expect higher energy spending throughout the world. “Electricity Generation Costs” theme refers to the cost viability of a certain energy source when generating electricity.
Whether the natural gas is dominating over renewable energy technologies or enabling them is categorized under the “Crowd-out Effect” theme. If the natural gas siphons investments from renewables perpetually, this effect is called “crowd-out”. If the crowd-out effect is
16
prolonged, then we might be locked in on a fossil fuel technology path. Last theme, “Carbon lock-in” categorizes this dependency on fossil technologies. If the carbon lock-in happens, this would obstruct other zero-carbon technologies to emerge. At that point, we would be forced to direct our resources towards neutralizing carbon-emitting technologies’ effects and might miss the timing for a successful energy transition.
Table 1 shows how different authors have used relevant quotes under those themes.
Corresponding quotations which were used to build Table 1 could be found in Appendix 4.
a) Energy Reliability
i) Positive Direct Effects
In the literature, some authors (Ahmed, & Cameron, 2014; Baron, 2013; Colombo et al., 2016; van Foreest, 2010; Stephenson et al., 2012; Smil, 2015; Boersma, & Jordaan, 2017) only mentioned direct positive effects for Energy Reliability issue. Due to the current technology of energy grids and infrastructure, renewable energy sources need to be helped by other technologies to provide reliable energy. According to the literature review, natural gas, due to its flexible turning on-off technology, is considered as a nice candidate for providing secure and reliable energy. There are no negative or uncertain outcomes mentioned in the literature on natural gas' neither direct nor indirect effects.
b) Crowd-out and Carbon lock-in i) Uncertain Indirect Effects
Crowd-out effect and Carbon lock-in are by definition indirect effects. Natural gas might crowd-out renewable energy sources in different time frames but we cannot have a carbon lock-in immediately. Rather, these situations emerge when certain pathways are chosen by the decision-makers. The crowding out effect may show itself with a delay if natural gas invested over renewables persistently over time. Following this path may lead to Carbon lock-in. As a result, there are no direct effects mentioned on crowd-out or carbon lock-in Table 1.
There are a couple of authors (Colombo et al., 2016; Aguilera, & Aguilera, 2012) which have been categorized as mentioning positive crowd-out indirect effect in Table 2. These authors mention that, while the natural gas might help renewable energy with its positive direct effects, such as energy reliability or energy costs, the renewable energy sources will become
17
economically viable in the meantime through directed investments and learning effect. Figure 3 shows how learning effect decreases the cost of technology so that investing in that technology becomes more attractive.
Other authors (Ahmed, & Cameron, 2014; Baron, 2013; Verbong, 2014; Boersma, & Jordaan, 2017; van Foreest, 2011; Stephenson et al. 2012) mention that unless coupled with directed policy efforts, natural gas would divert investments from renewables causing a crowding-out effect. Although these different authors differ in their half-full or half-empty perspective, they mention the same key concept. The success of clean energy transition and whether the natural gas crowds out the renewable energy sources depend on how well directed policy efforts are handled (Arent et al., 2015; Baron, 2013; Stephenson et al., 2012; Colombo et al., 2016; Aguilera, & Aguilera, 2012). Figure 4 shows how carbon-tax and zero-carbon subsidies can enable renewables. If renewables are enabled by these directed policy measures, this would decrease the necessity of building natural gas or other fossil capacities.
With the recent emergence of hydraulic fracturing technologies, natural gas prices have been getting competitive compared to other fuels. These competitive prices have been persuading more investors to divert their focus from renewable technologies (Ahmed, & Cameron, 2014; Arent et al., 2015; Baron, 2013). In addition, urgent demand for energy supply complicates things and reduces our sensitivity to the fuel choice (Ahmed, & Cameron, 2014; Arent et al., 2015; Baron, 2013; van Foreest, 2011).
The investment competition is not only in between natural gas and other fossils but rather it is in between natural gas and all sources of energy including renewables, nuclear or fossils (Ahmed, & Cameron, 2014; Boersma, & Jordaan, 2017). Any diverted investments from renewable energy can be
Figure 4 Directed Policy Efforts CLD Figure 3 Learning Effect CLD
18
seen as lost opportunities which might have increased the learning rate’s positive effect on renewable energy costs. This might force renewables to be economically unviable in the long term. Thus, some of the authors question whichtechnologies the investments are diverted from to enable natural gas (Ahmed, & Cameron, 2014; Arent et al., 2015; Baron, 2013). Hence,
without policy integration and directed efforts, it would be hard to assume that every natural
gas investment is done “at the expense of other fossil fuels” (Arent et al., 2015). Current direct positive effects of natural gas and the speed of development might lead to a default position of relying mainly on fossil fuels in developing countries (Ahmed, & Cameron, 2014). Sheer energy demand does not discriminate the type of fuels because that demand needs to be satisfied for the sake of economic growth (Ahmed, & Cameron, 2014). Likewise, Baron (2013) mentions that the share of renewables has remained unchanged since 1990 because additional fossil capacities have also been added to the mix. Cheap natural gas prices increase the relative cost of investing in renewable technologies (Baron, 2013; Arent et al, 2015; Ahmed, & Cameron, 2014). A similar incident has already been seen in the US when wind technology producers have been reporting substantial losses due to the competitive natural gas prices caused by hydraulic fracturing production rates (Ahmed, & Cameron, 2014; Newell, & Raimi, 2014). If one choice of technology dominates over the others economically, this choice might keep dominating the alternative technologies because of the learning rate effect. The investments directed towards this favorable choice might ensure its leadership position in its economic viability due to decreasing costs. Hence, using the path of least resistance strategy and investing in the most economically viable choice at every turn might endanger the sustainability goals and clean energy transition. This behavior is shown in Figure 5. If natural gas is chosen for its economic viability instead of renewable energy sources, this would make
19
natural gas technologies more affordable and economically viable in the future as well. At the same time, when a technology is not chosen, their relative position against the other technologies will deteriorate.
c) Energy Costs
i) Positive Direct Effects
One of the challenging complexities of direct-indirect effects of natural gas resides in the Energy Costs theme. According to some authors (Colombo et al., 2016; Flavin, 2008; van Foreest, 2010; Smil, 2015; Boersma, & Jordaan, 2017; Arent et al., 2015), there are many positive direct effects of natural gas under this theme. Firstly, natural gas technologies are more economically viable than current renewable energy technologies (van Foreest, 2010; Smil; 2015; Boersma, & Jordaan, 2017). Due to the high costs of renewable energy, investing in them would threaten an increase of the electricity costs on the demand side (Flavin, 2008; van Foreest, 2010; Smil, 2015; Boersma, & Jordaan, 2017; Arent et al., 2015; Colombo et al., 2016). Thus, natural gas presents us with an opportunity of providing cheap energy for the demand side. According to IEA (2019), in 2017, %13.2 of the world population still did not have access to electricity. One of the main reasons for not having access to electricity is affordability. If the clean energy transition intends to be just, decision-makers also have to consider the affordability of the electricity. Natural gas, due to its relatively low costs in both fuel prices and initial investments, is considered as a candidate as a next step in the clean energy transition in the literature.
ii) Negative Indirect Effects 1. Transition Costs
Although, natural gas can be a good candidate for directly reducing energy costs, a couple of concerns for the indirect effects arise in the literature. Natural gas is considered as a transition fuel. Meaning, after an amount of time, we need to transition out of natural gas to zero-carbon technologies to reach our sustainability goals and avoid a climate crisis. Every dollar that has to be spent on natural gas will be a short term solution for the world's energy demand (Ahmed, & Cameron, 2014; Parkinson, 2013; Channell, 2012; Colombo et al., 2016; Boersma, & Jordaan, 2017). When we absolutely have to transition into zero-carbon technologies to avoid a
20
climate crisis, new investments have to be allocated again for the next step of the transition. Some authors (Ahmed, & Cameron, 2014; Baron, 2013; Colombo et al., 2016; van Foreest, 2011; Boersma, & Jordaan, 2017) mention that this would make the whole clean energy transition more expensive as an indirect effect. Baron (2013) argues that carbon taxes are not at the levels that they should be. Even if carbon taxes were at sufficient levels, they would still need to be directed at renewable energy sources to improve the learning effect. Although this investment diversion through subsidies would be beneficial for renewable energy sources, it also increases the total energy costs on the fossil and natural gas energy production side (Baron, 2013), and thus would also do so on the demand side. The increase in electricity prices on the demand side would endanger the affordability of electricity as well.
2. Energy Demand
Arent et al. (2015) mention a feedback relation between energy production and the world GDP in his research. If more energy is produced for the same price, this results in more wealth which in turn creates a demand for more energy as a result of the increased production. This increased energy demand will require more capacities being installed, including fossils (Arent et al., 2015). This effect is called energy rebound (Madlener, & Alcott, 2009; Greening, Greene, & Difiglio, 2000). The energy rebound issue is mostly associated with energy efficiency in the academic literature. Madlener and Alcott (2009: 371) define rebound as “the additional energy consumption enabled by energy efficiency increases”. Greening, Greene, and Difiglio (2000: 389) describes rebound in this way: “Gains in the efficiency of energy consumption will result in an elective reduction in the per unit price of energy services. As a result, consumption of energy services should increase partially offsetting the impact of the efficiency gain in fuel use”. Madlener and Alcott (2009: 1) mention that energy efficiency strategies might raise economic growth, but as an environmental or an energy policy strategy could “backfire” by leading into more resource use. Although energy rebound seems to relate to the energy efficiency in the literature, cheap energy prices by natural gas might lead to the same outcome. This behavior is represented in Figure 6. If we are able to produce more energy by using
21
cheap natural gas, this would have a positive effect on the world economy. In turn, we would increase our production and consumption rates. This behavior would lead to increased energy demand. Increased demand would keep our energy costs at the same level if not increase.
3. Electricity Generation Costs
Ahmed, & Cameron (2014) mentions that in order for renewable energy sources to balance their variability issue, we also have to invest in natural gas which in turn would increase the total costs for renewable energy. If every renewable energy source needs to be coupled with a natural gas plant, this would increase the market entry barriers for renewable energy investors. In addition, if renewables have to be coupled with natural gas, this would also increase the cost of electricity generation. When the electricity generation costs increase, this would reflect on our electricity bills on the demand side (Flavin, 2008; van Foreest, 2010; Smil, 2015; Boersma, & Jordaan, 2017; Arent et al., 2015; Colombo et al., 2016). Not to mention, these natural gas production facilities would not be activated for the sole reason of helping renewable energy sources with variability. If there is a power plant and an energy need, we would also try to utilize these natural gas capacities to compensate that demand.
d) Need for Clean Energy i) Positive Direct Effects
According to the literature, the most challenging subject of using natural gas as a transition fuel resides in understanding how natural gas will impact the environment. Natural gas' foreseeable direct positive effects on environment qualify it as a transition fuel. Natural gas approximately emits half the CO2 emissions (Baron, 2013; Colombo et al, 2016; Flavin, 2008; van Foreest, 2010; Stephenson et al., 2012; Smil, 2015; Boersma, & Jordaan, 2017; Aguilera, & Aguilera, 2012) and also less NOx and SOx emissions (Colombo et al, 2016; Smil, 2015; Boersma, & Jordaan, 2017) compared to other fossil fuels. Natural gas can be considered as a better next step in world's energy production compared to other fossils. As long as natural gas is replacing the other fossils, we can say this would have a positive effect on the environment in terms of emissions. There are success stories of reducing emissions by using natural gas in the US (Arent et al., 2015; Stephenson et al., 2012; Boersma, & Jordaan, 2017) and the Netherlands (Smil, 2015). In addition, some authors mention that a reduction of emissions might be expected if
22
natural gas should ever become dominant in Asia well (Stephenson et al., 2012; Boersma, & Jordaan, 2017). Natural gas’ possible direct emission reductions are shown in Figure 7. If we use natural gas instead of other fossils, this would decrease the emissions due to natural gas’ less polluting effects. The transition from other fossils to natural gas to achieve this direct emission reduction is shown in Figure 8. Cheap natural gas prices are forcing investors to transition out of other more expensive and more polluting fossil fuels.
ii)
Negative or Uncertain Indirect Effects
1. CO2 Emissions
Although the US reduced its emissions with the boom of shale gas, it also made coal exportation prices cheaper because of the competition. This cheap coal has been exported to Europe (Baron, 2013; Verbong, 2014) and made it more viable to keep the European coal plants still running (Verbong, 2014). This raises the question of shifting the burden. Figure 9 shows the system archetype for “shifting the burden”. Cheap natural gas is the symptomatic solution for reducing other fossils locally. However, this enables cheap coal prices for exportation as a side effect and increases the capacity for the other fossils globally. In turn, problem symptom is not affected at all. Already extracted coal, if not used in the location where it was
Figure 9 Shifting the burden system archetype (Senge, 1990)
23
extracted, will be used in someplace else in the world. Arent et al. (2015: 17) argues on this point by saying "To the extent that abundant gas occurs locally, changes in GHG emissions in one region can have consequences for emissions in another region". Hence, Arent et al. (2015) discuss that without policies directed specifically on enabling renewable energy sources, it is uncertain how net CO2 emissions will be in the future. Moreover, Smil (2015) argues that natural gas has been a key element in decarbonizing electricity production, but achieving to stay under the 450 ppm goals might not be feasible by just depending on natural gas. Furthermore, Stephenson et al. (2012) discuss on British Columbia (B.C.) case and points out that if B.C. relies solely on natural gas for energy production, in order to reach its sustainability goals, they have to do a great deal of CO2 emissions reduction on other sectors such as industries, buildings, vehicles or population.
To sum up, according to many scientists, natural gas, although being a relatively cleaner fossil fuel, cannot be used as a silver bullet strategy for mitigating all of our CO2 emissions risks. Another uncertain aspect shows itself in our calculations of the CO2 and other emissions. In scenario and modeling studies that were taken into account in this research use only the emissions caused by fuel-burning (Xiao, Niu, & Guo, 2016; Brouwer, 2015; Eggelston, Buendia, Miwa, Ngara, & Tanabe, 2006). This is understandable since there is more data on this phenomenon when investigating industry reports (IPCC, 2007). However, there are also indirect emissions which are emitted throughout a fuel's life cycle (Arent et al, 2015; Stephenson et al., 2012; Boersma, & Jordaan, 2017). Namely, when extracting, producing and transporting a fossil/hybrid fuel, there are still indirect GHG emissions. Although the CO2 shares of these emissions are recognized as inconsiderable when compared to emissions caused by burning other fossils (Wood, Gilbert, Sharmina, Anderson, Fottitt, Glynn, Nicholls, 2011), they still contribute to the total emissions emitted throughout the life-cycle of a fuel. In order to expand our assumptions in our scenario and modeling practices, scientists need more data on the life-cycle emissions so that they can include these
24
numbers in their studies. Figure 10 shows how overlooked life-cycle emissions of natural gas are affecting the system.
2. Other Emissions
Howarth, Santoro, and Ingraffea (2011) especially point out the importance of direct and indirect methane emissions since methane also contributes significantly to the global warming potential. According to IPCC AR4 (2007), the 20-year global warming potential of methane is 72 times effective than CO2. One source of indirect methane emissions is methane flaring. Smil (2015) explains that methane flaring takes place in refineries, gas plants, and during well tests. According to Smil (2015), natural gas still is a cleaner fuel compared to oil and bitumen in terms of methane flaring emissions. However, in order to decrease the uncertainty on the total effect of natural gas on the global warming potential, Smil (2015) mentions that we have to make comparative studies on natural gas’ life-cycle methane emissions and total anthropogenic methane emissions. Furthermore, Howarth et al. (2011) argue that shale gas could perform %20 worse than coal over a 20-year time frame and %100 percent worse than coal over a 100-year time frame in methane emissions. Boersma & Jordaan (2017) also point out to the uncertainties of the magnitude of methane emissions from shale gas life-cycle. Stephenson et al. (2012) refer to two different reports with two different outcomes that state shale gas produces higher life-cycle emissions than conventional natural gas over a 100-year timeframe. IEA (2011) reports this increase to be in between 3.5% and 12% whereas Shell Global Solutions reports this increase to be in between 1.8 and 2.4% over ‘‘wells-to-wires’’ (Stephenson et al., 2012). As shown, there are conflicting numbers in different reports.
3. Fossil to Gas
Although, a switch from other fossils to natural gas presents us with an opportunity to reduce the climate impact of the electricity production, handling this switch is not that simple due to the uncertain dynamics of the system. First uncertainty in the literature presents itself in the fuel prices.
Colombo et al. (2016) mention that cheap natural gas prices and ample supplies might enable the first step in the –other fossils to gas– clean energy transition and also might be an incentive to enable other zero-carbon technologies further down the road if complemented
25
with necessary measures. Likewise, Aguilera and Aguilera (2012) mention that if a carbon tax is implemented on coal, coal's relative position against the natural gas would deteriorate and further enable natural gas investments. Both Colombo et al. (2016) and Aguilera and Aguilera (2012) argue that competitive natural gas prices against coal would become an incentive for a switch from other fossil fuels.
Contrarily, Ahmed and Cameron (2014) refer to Tracking Clean Energy Progress (2013), stating out that US shale gas boom had the opposite effect on Europe. Competitive natural gas prices in the US pushed the American coal industrialist to export coal to Europe for fairly cheap prices. Verbong (2014) confirms this by saying that these cheap coal prices temporarily halted the operations in European natural gas plants to enable cheaper coal plants. Verbong (2014) even mentions that these cheap coal prices have postponed some of the promised transition decisions out of coal plants in Europe. This means that cheap natural gas prices in a local region had a controversial effect on the global system by making coal prices competitive elsewhere (Arent et al., 2015).
Fuel prices' interesting dynamic effects are not only confined to this issue. Arent et al. (2015) mention that when natural prices were not competitive enough in 2013-2014, we have seen a shift to cheaper coal again. Arent et al. (2015) also points out to an opposing example by referring to several studies (Newell, & Raimi, 2014; Energy Modeling Forum, 2013; McJeon, Edmonds, Bauer, Clarke, Fisher, Flannery, Hilaire, Krey, Marangoni, Mi, Riahi, Rogner, & Tavoni, 2014). These studies mention the energy-economy feed-back. When the energy is produced at a relatively cheap price, this has a positive effect on the growth. In turn, a growing economy demands more energy in all forms including fossils. This issue caused by cheap energy
26
prices resembles the energy rebound’s effect mentioned in Figure 6.
This paradox has certainly an interesting dynamic. When natural gas prices decrease, there is a shift back to coal. When natural gas prices increase, there is a shift back to coal. This might refer that a shift from other fossils to natural gas might be insensitive to the natural gas
price. The natural gas’ fuel prices dynamics have been shown in Figure 11. Even if natural gas
can help the transition locally, it can hinder the transition at the same time globally.
4. Other Environmental Effects
Besides emissions, natural gas has also other negative environmental effects. Some of these negative effects can be traced back to the hydraulic fracturing process. Hydraulic fracturing process needs excessive use of water as well as the necessity of using chemicals in the water (Smil, 2015; Stephenson et al., 2012). Smil (2015) mentions an instance of water contamination in Ohio, Arkansas, Texas, and Oklahoma due to the hydraulic fracturing process.
Another negative effect of extracting natural gas is that it leaves the ground susceptible to tremors and earthquakes (Smil, 2015; Boersma, & Jordaan, 2017). Smil (2015) mentions an incident where the public attributed the earthquakes in Oklahoma region to hydraulic fracturing. Also, Boersma and Jordan (2017) mention that after the tremors in the Netherlands, the public has pulled back its support from natural gas extraction.
5. Combined CLD of natural gas’ direct and indirect effects.
In Figure 12, the previous CLDs are combined to clearly depict the dynamics at play. The combined CLD structures the debate and thoughts on natural gas and completes our answer to research question 1 by explaining the debate with the help of Table 1 and Figure 2.
As can be seen from the Figure 12, there are different loops counteracting with each other and without further research, it would be hard to figure out which behaviors or loops will be dominant over the other. This dynamicity makes it harder for scientists to validate the consistency of their arguments. If the positive direct effects would be dominant in the clean energy transition, natural gas might help the transition. On the contrary, if the indirect loops would be dominant, natural gas can obstruct the clean energy transition.
There are 4 specific conflicts between the direct and indirect effects of natural gas when the CLD is considered in detail. The first issue presents itself between the relation of direct
27
emissions and indirect emissions. Although we are certain that natural gas’ direct emissions are
lower (Baron, 2013; Colombo et al, 2016; Flavin, 2008; van Foreest, 2010; Stephenson et al., 2012; Smil, 2015; Boersma, & Jordaan, 2017; Aguilera, & Aguilera, 2012), we cannot say the same for the indirect CO2 and methane emissions (Ahmed, & Cameron, 2014; Stephenson et al. 2012; Smil, 2015; Arent et al., 2015; Baron, 2013; Boersma, & Jordaan, 2017). The uncertainty on the issue especially shows itself in the methane emissions with conflicting numbers in different reports (Howarth et al., 2011; Stephenson et al. 2012; Boersma, & Jordaan, 2017). CO2 and its corresponding fuel burning emissions seem to be our main concern in scientific studies in regards to climate change, but life-cycle emissions of all GHGs require an investigation to determine whether our assumptions on natural gas’ benefits are well-founded.
Second issue presents itself in the relation in between Transition from Other Fossils to
NG (Arent et al., 2015; Colombo et al, 2016; Flavin, 2008; Stephenson et al., 2012; Smil, 2015;
Boersma, & Jordaan, 2017; Aguilera, & Aguilera, 2012) and Spillover Effect (Ahmed, & Cameron, 2014; Arent et al.,2015; Verbong, 2014; Baron, 2013; Boersma, & Jordaan, 2017). In here, uncertainty is caused by the natural gas’ the fuel price. There are conflicting evidence for both sides of the issue (Arent et al., 2015; Newell, & Raimi, 2014; Energy Modeling Forum, 2013; McJeon et al., 2014; Ahmed, & Cameron, 2014; Verbong, 2014). Regardless of the natural gas price, the world has seen a shift back to coal in some instances. In order to be confident to say cheap natural gas would help the clean energy transition, this dynamic has to be investigated further.
The third issue presents itself in the Energy Rebound Loop. We are not certain whether cheap electricity production would increase the energy demand and push the total energy costs up with the demand (Arent et al. 2015, Madlener, & Alcott, 2009; Greening, Greene, & Difiglio, 2000). Energy rebound effect calculates the percentage of the saving which will return back as resource use reduction (Madlener, & Alcott, 2009; Greening, Greene, & Difiglio, 2000). We need to investigate the ratio of the energy rebound caused by fuel prices so that we can be certain to say whether the cheap energy prices would help or hinder the transition.
The fourth issue presents itself in the Crowd-out mechanism (Ahmed, & Cameron, 2014; Baron, 2013; Verbong, 2014; Boersma, & Jordaan, 2017; Arent et al., 2015; Baron, 2013;
28
29
Colombo et al., 2016; Aguilera, & Aguilera, 2012). Crowd-out might occur when all these dynamic issues represented in the CLD interacts with each other. As the literature suggests, due to low prices, natural gas might crowd-out renewable in the short-to-medium term (Aguilera, & Aguilera, 2012; Arent et al, 2015; Boersma, & Jordaan, 2017). In addition, Spillover effect and
Energy Rebound might push the decision-makers to give precedence to natural gas and even
other fossils. This suggests that there will be less learning effect rate for the renewable energy sources making them less commercially viable in the future.
e) Computational Model
i) The boundary of the Computational Model
The dynamics that revealed in the literature review presented a handful of interesting issues to investigate in the computational model. As the boundary of this thesis work, there needed to be a selection of the dynamics mentioned in the CLDs to investigate in the theoretical model effectively.
One of the significant topics on natural gas is whether these investments will crowd-out renewable technologies. The crowd-out effect can trigger a carbon lock-in and obstruct a successful transition. Thus, the crowd-out structure has been selected as the boundary of this research. Crowd-out happens for a reason. Namely, the natural gas’ preferable costs and learning rate’s effect on costs might enable natural gas even further. Thus, cost and learning rate structures are also in the boundary of this computational model. In order to check whether the clean energy transition is successful in terms of the sustainability goals, emissions structure is also within the boundary of this computational model.
The computational model consists of 8 sectors and the structure of the model, the related equations and data assumptions can be found in Appendix 6 & 8. The overview of the boundary of the model is presented in the bull’s eye diagram (Ford, 1999) in Table 2. Bull’s eye diagram shows endogenous, exogenous, and omitted variables of the model, and thus creates a visual depiction of the boundary of the model and possible future extensions for the model. Bull’s eye diagram also includes the data sources where the corresponding literature and data source for the parameterization and equations of the model.
30
ii) The structure and equations of the Computational Model
The selected structures for the computational model are learning rate, crowd-out, costs of technologies, investments for technologies, need for clean energy, and emissions. One-factor learning rate formula which focuses on installed capacities has been used in the computational model (McDonald, & Schrattenholzer, 2001; Viguier, Barreto, Haurie, Kypreos, & Rafaj, 2006; Batinge, 2015). The crowd-out effect presents itself as dynamicity in between relative costs and benefits of natural gas compared to other technologies (Ahmed, & Cameron, 2014; Baron, 2013; Verbong, 2014; Boersma, & Jordaan, 2017; Arent et al., 2015; Baron, 2013; Unruh, 2000). For the cost of technologies, levelized cost of electricity (LCOE) have been used (IPCC, 2014; IEA, 2010). LCOE aggregates the cost of electricity production by dividing the time costs by life-time electricity production (Ueckerdt, Hirth, Luderer, & Edenhofer, 2013). LCOE includes any costs from capital investments to fuel prices aggregated within. For the investment decision function, cost-benefit analysis has been used (Levin, & McEwan, 2000; Garber, & Phelps, 1997). The cost-benefit analysis compares the costs for a specified amount of benefit. Meaning, the benefit of a specific technology is divided by the cost of the technology. The result gives us the cost-benefit advantage for a specific technology. In this model, investment weights are calculated by comparing the costs of technologies for one unit of benefit (1 exajoule of electricity). This benefit of 1 exajoule is divided by the cost ratio of a specific technology and this result calculates the investment decision weight for that specific technology. There are three types of technology in this computational model: renewables, natural gas, and other fossils (coal). The need for clean energy structures is founded upon the “directed policy measures” for the clean energy transition in the literature (Arent et al., 2015; Baron, 2013; Stephenson et al., 2012; Colombo et al., 2016; Aguilera, & Aguilera, 2012). The corresponding model structure for the need for clean energy changes how we perceive the costs of renewable technologies as the global emissions increase. The emissions of fossils and natural gas are calculated by the emission factor for direct fuel burning (Xiao, Niu, & Guo, 2016; Brouwer, 2015; Eggelston et al. 2006). The emissions from electricity are then calculated into global emissions by using a varying range for electricity production’s emissions share (IEA, 2018; IPCC, 2014).
31 Table 2 Bull's eye diagram
