Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.
We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.
To purchase hard copies of this book, please contact the representative in India:
CBS Publishers & Distributors Pvt. Ltd.
www.cbspd.com
|
customercare@cbspd.com
The Paris Agreement on climate change requires nations to keep the global temperature within the 2°C carbon budget. Achieving this temperature target means stranding more than 80% of all proven fossil energy reserves as well as resulting in investments in such resources becoming stranded assets. At the implementation level, governments are experiencing technical, economic, and legal challenges in transitioning their economies to meet the 2°C temperature commitment through the nationally determined contributions (NDCs), let alone striving for the 1.5°C carbon budget, which translates into greenhouse gas emissions (GHG) gap. This chapter focuses on tackling the risks of stranded electricity assets using machine learning and artificial intelligence technologies. Stranded assets are not new in the energy sector; the physical impacts of climate change and the transition to a low-carbon economy have generally rendered redundant or obsolete electricity generation and storage assets. Low-carbon electricity systems, which come in variable and controllable forms, are essential to mitigating climate change. These systems present distinct opportunities for machine learning and artificial intelligence-powered techniques. This chapter considers the background to these issues. It discusses the asset stranding discourse and its implications to the energy sector and related infrastructure. The chapter concludes by outlining an interdisciplinary research agenda for mitigating the risks of stranded assets in electricity investments.
Center for Energy and Environmental Policy (CEEP) at the University of Delaware, United States
Foundation for Renewable Energy and Environment (FREE), United States
*Address all correspondence to: jnyangon@udel.edu
1. Introduction
The power industry is in transition, and energy management systems are adapting to it. Recently, the rapid proliferation of distributed energy resources (DERs) (e.g., distributed generation such as residential solar photovoltaics (PV) and wind electricity, controllable loads, and energy storage), have transformed operational, planning, and regulatory dynamics. Low-cost natural gas in the US, Europe, and elsewhere continues to push gas-fired electricity generation to the top of the generation mix. To this end, governments continue to promote low-carbon technologies through ever-stringent energy policies, like renewable portfolio standards (RPS), net metering, feed-in tariffs, and carbon pricing initiatives and emission trading schemes like the European Union Emission Trading System (EU ETS), Switzerland Emissions Trading Scheme, emissions trading schemes in China and Australia, the Regional Greenhouse Gas Initiative (RGGI) in the nine U.S. states in the Northeast and Mid-Atlantic region, the Transportation and Climate Initiative (TCI) under consideration for transportation emissions in the Northeast and Mid-Atlantic, the California and Quebec’s Western Climate Initiative, among others. Furthermore, this growth in renewable electricity generation has been motivated by customers’ preference for distributed energy as a means to fostering grid reliability and system efficiency, cost reduction, and improved customer choice over their power supplies [1, 2, 3].
These efforts are in line with the 2015 Paris Agreement on climate change and its nationally determined contributions’ (NDCs) long-term goal of keeping the rise in global mean temperature to “well below [two degrees Celsius (2°C)] above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels” [4]. Moreover, limiting these temperature targets requires reaching net-zero global carbon dioxide (CO2) emissions “between 2060 and 2070,” with full decarbonization or “net negative CO2 emissions” realized before the end of the century [5]. From a policy perspective, this transition from carbon-intensive sources to low and non-carbon-emitting sources is continuing as high penetrations of distributed electricity, energy storage and management devices, and investment in new forms of flexible demand resources become connected to the power grid network. Such significant shifts threaten the fossil energy business model and could, in turn, result in the “stranding” of the carbon-intensive assets through retirement or devaluation [6, 7, 8]. In other words, meeting the Paris temperature targets necessitates turning existing fossil fuel investments into stranded assets and fossil fuel reserves into stranded resources. The concept of “stranding” or “stranded assets” has been explored broadly in extant literature, from environment-related risk exposure of coal assets [9] to “unburnable fossil fuel deposits” such as oil, gas, and coal and the risk of stranded assets [10, 11].
Bos and Gupta [12] define stranded assets as “assets that lose economic value well ahead of their anticipated useful life, whether that is a result of changes in legislation, market forces, disruptive innovation, societal norms, or environmental shocks” p. 1 and stranded resources as “resources which are considered uneconomic or cannot be developed or extracted as a result of technological, spatial, regulatory, political or market limitations, or changes in social and environmental norms” p. 2. On the other hand, Caldecott et al. [13] define stranded assets as those assets which “suffer from unanticipated or premature write-offs, downward revaluations or [conversion] to liabilities” p. 11. Policymakers and experts concur that this transition should be managed proactively and pragmatically because if done haphazardly, it could perpetuate the techno-institutional complex of “carbon lock-in” and path dependency, thus making future transitions difficult [14, 15, 16, 17, 18, 19, 20, 21]. On the other hand, if variable renewable energy resources like solar and wind electricity is introduced in significant quantities and not correlated exactly with peak load, it may create a unique challenge like the infamous California ISO’s “duck curve” shown in Figure 1. How should the energy sector respond?
Figure 1.
California ISO’s “duck curve.” Source: CAISO (2014).
This chapter is structured as follows. Section 2 discusses digitalization solutions and business model innovations and presents moral arguments for supply- and demand-side energy solutions, including sensors, meters, higher efficiency devices, and energy auditing, including measurement and verification strategies that can be utilized to improve energy management. Specifically, using ML (machine learning- and (AI) artificial intelligence solutions to support (a) tackling stranded assets in the transition to a low-carbon economy (b) real-time measurement of energy data, (c) manage data gathering and monitoring, (d) and proactively and accurately analyze the data gathered to detect changes in supply-demand imbalances and improve the situation promptly. Section 3 reviews the risks of stranding and identifies distinct opportunities for ML and AI applications in the energy sector. Section 4 emphasizes the impact of stranding risk factors on oil, gas, and coal resources and how this translates into the concept of “unburnable fossil fuel deposits.” Section 5 discusses how advances in ML and AI techniques might help tackle the risks of stranded carbon assets, and Section 6 concludes.
2. Leveraging digitalization and business model innovations for energy management
Today’s modern cities are sprouting with new industrial buildings and residential complexes. The consensus is emerging that dramatic growth in distributed renewable energy, and digitalization in economy and innovations, two megatrends of the twenty-first century, are critical strategies for climate change mitigation and changing the greenhouse gas (GHG) emission trajectories. Yet, while the electric power system is in transition, many of the vital power system challenges which confront governments and businesses, like access to a cleaner, more resilient, reliable, and affordable electricity, remain underfunded and unresolved. The increased deployments of energy management applications across the transportation, buildings, and industrial sectors, for example, reduce the cost of operation and consumption, lower energy losses, increase grid reliability, improve electric power production from carbon-free sources, and alleviate investment inefficiencies that could cause an energy-efficiency gap [22, 23, 24]. The continued growth of the fluctuated distributed generations (such as solar photovoltaics, wind turbines, electric vehicles, and energy storage systems) may perturb the network and create voltage drop/rise problems and in severe conditions, blackouts.
In a highly electrified economy with high shares of variable solar and wind electricity systems, reducing systemic mismatches between the generation and energy demand assets in an efficient manner requires investments in smart energy management systems. Energy management systems consist of two main categories: (a) supply-side devices from the electric utility-side used to manage the fluctuation of the load demand such as substations, and (b) the demand-side management devices used to manage energy consumption and meet the available power from the generation side [25, 26, 27, 28]. Substations encompass transformers, switchgear, and protection, control and automation systems, and connect parts of the electric grid that operate at different voltage levels and managing these multidirectional power flows while ensuring reliability and security is critical as the share of decentralized and renewable energy increases. The rise of smart energy management systems, including ML, AI, big data, smart sensors, and the Internet of Things, is a boon not only to the electric power industry—especially in reducing operational costs and carbon emissions—but also the energy transition. For example, opportunities for leveraging digitalization for business model innovation in smart energy management and the corresponding implications for the power sector are substantial and untapped [29, 30].
Energy management is subject to barriers and limitations, which can delay its full market integration. These barriers include high cost of system implementation, inflexible fixed-price electricity tariff system and rate design, aging network’s infrastructure, and the need for bidirectional power flow, which is ideal for an intelligent grid network. As a result, energy management continues to have a prominent role in decarbonization. Using ML algorithms and AI optimization models, utilities and system operators can apply optimal dynamic pricing and energy storage resource to improve the management of the “duck curve” phenomenon. For example, Sheha et al. [31] applied game-theoretic models to show that leveraging a combined effect of dynamic pricing profiles and distributed electrical energy storage can help flatten the duck curve, thus solar energy can be increasingly added on the grid without resulting in grid failure. The duck curve problem arises when increasing solar penetration on the grid creates a dip in net load in the middle of the day as solar generation peaks and wind electricity is low, followed by a significant rise in peak in residential demand at sunset as, without any form of energy storage, solar electricity rapidly subsides, and customer consumption increases as citizens get home from work/school thus necessitating significant ramping of thermal generators [24]. Figure 1 shows California Independent System Operator’s (CAISO) widely known “duck curve” (Figure 1) [32]. Besides California, the “duck curve” phenomenon also occurs in energy markets with high solar electricity penetrations such as Italy, Germany, Hawaii, and others.
To eliminate the risk of over-generation and possibly, alleviate the “duck curve” problem, implementation of long-term solutions focusing on distinct opportunities for ML techniques, including distributed solar coupled with storage technologies and smart energy management, are emerging in various energy markets. At stake, according to Guidehouse Insights (formerly Navigant Research), is $278 billion in annual global market for the deployment of commercial and industrial (C&I) energy as a service (EaaS) solutions by 2028 [33].
The intergenerational issues associated with climate change identifies it as an externality associated with carbon dioxide and other GHG emissions because it involves costs that are borne by future generations who have not created the emissions [34, 35, 36, 37, 38]. Climate change economists have introduced the concept of “social costs of carbon,” which externalizes the externalities of these emissions by denoting the damages caused by them with a monetary value [35, 39, 40, 41]. It is for this reason that climate policy experts have advocated for a carbon price to achieve the “right price” as well as incentivize the investments in low-carbon technologies. Furthermore, from a policy, equity and regulatory point of view, scaling the deployment of low-carbon energy technologies inspires innovation in technological development, diffusion, transfer, and discourages the holding of dirty exhaustible assets (fossil fuel reserves) [42, 43], which are prone to becoming stranded due to perfect substitution, and disproportionately impact low- and moderate-income communities.
The risks of stranding of assets are likely to occur during the transition to a green economy. As van der Ploeg and Rezai [17] suggest two conditions are necessary for this transition to occur: (1), the unexpected future changes in the conditions likely to affect the economics of fossil fuel assets, such as customer demand, the social cost of carbon that values the climate externalities, and equity and efficiency considerations, must be present; (2) the cost of shifting around “the underlying capital stocks in the carbon-intensive industries to productive use elsewhere after the energy transition” must be too prohibitive or impossible to meet. Expectations about stranding carbon-intensive assets can occur due to sudden policy change, a breakthrough innovation in renewable energy technology such as energy storage batteries, which can lead to the stranding of fossil fuel-based financial assets since they directly pose a threat to the sustainability of the coal, oil, and gas-based business model.
With ML and AI techniques, energy operators can foster better short-term and long-term forecasting to improve electricity scheduling and integrated system planning, respectively. This would enable the utility operators and system managers to reduce their reliance on polluting, exhaustible fossil assets as well as proactively manage increasing amounts of distributed, low-carbon, variable energy sources like solar and wind energy. Additionally, the ML-AI-driven energy forecasts can provide accurate and optimal management of power grid fluxes to help operators proactively match demand-supply imbalances, manage uncertainties, as well as understand where, when and how many solar power systems [44, 45] and wind generation plants [46] should be built.
However, much of these forecasts employ domain-agnostic techniques, in which domain-specific scenarios are often less applied. For this reason, ML and AI algorithms of the future must incorporate weather-related innovations in climate science and weather modeling techniques in order to improve parametric and nonparametric estimates of both short- and long-term forecast uncertainty, for example, of variable generation and electricity demand [44, 46, 47, 48, 49]. For example, using a novel deep learning framework that combines wavelet transforms, stacked autoencoders, and long-short term memory, Bao et al. [50] produced stock price forecasts that outperform other similar models in both predictive accuracy and profitability performance. This notion can be extended to aid electricity demand forecasts that optimize intraday and day-ahead levelized cost and levelized avoided cost of electricity generation resources that minimize GHG emissions. More broadly, in the transition from the incumbent centralized electricity network to a distributed model that is underway, driven by the rapid growth of DERs, understanding the domain value of improved forecasts (e.g., to model electricity load in rural microgrids) across the quartiles of electricity market operation, matching of supply-demand imbalances, network control, and governance and administrative networks [2, 51] is an exciting challenge for ML and the debate on stranded assets.
4. Investments, stranding risk factors, and unburnable fossil fuel deposits
Going by Bos and Gupta [12] and Caldecott et al. [9]’s definition, stranded assets and stranded resources manifest in two main ways (1) devaluation through the unburnable fossil fuel resources, which must be kept in the ground to keep the long-term global temperature target to below 2°C above pre-industrial levels [4], and (2) premature retirement of exhaustible fossil capital assets due to climate policies, including the optimal social cost of carbon in the form of a carbon price [40]. Figure 2 shows the US annual electricity generating capacity additions and retirements from oil, gas, and coal power plants.
Figure 2.
Annual oil, gas, and coal electricity generating capacity additions and retirements. Source: EIA (2020).
In the US Energy Information Administration (EIA) Annual Energy Outlook’s Reference case, natural gas-fired combined-cycle generation capacity will continue to be added steadily through 2050. Significant retirements of electric generation capacity, mostly from coal, occur by 2025, while approximately 117 GW of new wind and solar capacity additions could occur between 2020 and 2023 [52]. This means that without investing in heat rate improvement technologies by 2025 to increase their efficiency, coal-fired generation systems must retire to comply with the affordable clean energy (ACE) rule or become stranded assets. The AEO2020 Reference case also shows that the low cost of natural gas prices significantly contributes to the retirements of coal-fired and nuclear power plants by 2025. Diverse policy efforts, notably increasing state RPS targets, net metering policies, and declining capital cost profile of solar, are expected to incentivize and accelerated its growth through 2050 by making the investment case for widespread solar energy deployment attractive to investors, particularly when utility-scale and small-scale applications are considered. Table 1 summarizes the seven main drivers of stranding and the different aspects of stranded resources and assets.
Type of stranding
Nature of asset
Cause of stranded asset
Stranded resource
Liability
References
Economic
Viable projects receive investment (e.g., growing biofuels in deforested lands)
a
Increased market competition affects investment in the asset (e.g., falling oil prices leads to cuts in oil exploration investments)
When it becomes uneconomical to extract/convert the resource due to low demand
Premature stranding costs (e.g., decommissioning and phase-outs costs)
The different aspects of stranded resources and assets.
Increased efficiency could create higher overall demand referred to as the Jevons paradox. For example, a shift to electric vehicle model may lead to the rebound effect, resulting in increased vehicle miles travelled, and overall rise in GHG emissions [54, 56].
New technological breakthroughs like carbon capture, utilization and storage (CCUS) innovations can be used to extract CO2 from power plant exhaust and industrial processes.
According to the Potsdam Climate Institute, to meet the Paris Agreement of a temperature target below 2°C of global warming with the aim to limit it to 1.5°C, the global carbon budget of the total volume of CO2 emissions permitted by 2050 is 886 GtCO2 [74]. However, more than a third of this carbon budget has already been used up from burning fossil fuels, leaving a budget of around 565 GtCO2. In the case of a 1.5°C temperature limit or even lower, this budget would be drastically contracted. It is for this reason that national, state and local governments must prioritize low-carbon transformations; for instance, (i) ramping up renewable energy over the next two decades, (2) switching from oil to less carbon-intensive gas [5, 15, 25, 42, 48, 55, 75, 76, 77], and (3) keeping large global deposits of coal, oil, and gas reserves “in the ground” (Table 2) [11, 13, 79]. This call has led to the “keep fossil fuels in the ground” initiative, “fossil fuel divestment” campaign, and “unburnable carbon” resistance movement, as a way to compel companies which are active in hydrocarbons or with high coal, oil, and gas reserves in their portfolios to reinvest elsewhere [17, 63, 79, 80, 81, 82, 83, 84].
Total proved coal reserves at the end of 2019
Total proved oil reserves at the end of 2019
Total proved gas reserves at the end of 2019
Country
Reserves (million tons)
% World
Country
Reserves (billion barrels)
% World
Country
Reserves (trillion cubic meters)
% World
The United States
249,537
23.3%
Canada
170.8
9.8%
The Russian Federation
38.0
19.1%
The Russian Federation
162,166
15.2%
Venezuela
303.8
17.5%
Iran
32.0
16.1%
Australia
149,079
13.9%
Kazakhstan
30.0
1.7%
Qatar
24.7
12.4%
China
141,595
13.2%
The Russian Federation
107.2
6.2%
Turkmenistan
19.5
9.8%
India
105,931
9.9%
Iran
155.6
9.0%
The United States
12.9
6.5%
Indonesia
39,891
3.7%
Iraq
145.0
8.4%
China
8.4
4.2%
Germany
35,900
3.4%
Kuwait
101.5
5.8%
Venezuela
6.3
3.2%
Ukraine
34,375
3.2%
Saudi Arabia
297.7
17.1%
Saudi Arabia
6.0
3.0%
Poland
26,932
2.5%
The United Arab Emirates
97.8
5.6%
The United Arab Emirates
5.9
3.0%
Kazakhstan
25,605
2.4%
The United States
68.9
4.0%
Nigeria
5.4
2.7%
Turkey
11,525
1.1%
Libya
48.4
2.8%
Algeria
4.3
2.2%
South Africa
9,893
0.9%
Nigeria
37.0
2.1%
Iraq
3.5
1.8%
92.8%
90.1%
84.0%
Table 2.
Global reserves of coal, oil, and gas.
Source: BP Statistical Review of World Energy 2020 [78].
Notes: The total world proved coal, oil, and gas reserves at the end of 2019 were 1,069,636 million tons, 1735.9 billion barrels, and 198.8 trillion cubic meters, respectively. The total proved coal reserves include both anthracite and bituminous reserves and sub-bituminous and lignite reserves.
Table 2 shows the top 12 countries for each of the three fossil fuels. These coal, oil, and gas reserves represent 92.8%, 90.1%, and 84%, respectively, of the total global, proved reserves at the end of 2019 [78]. McGlade and Ekins [11] have computed a breakdown of the socially optimal distribution of stranded carbon assets that must be kept in the ground to meet the Paris Agreement temperature targets. They find that to have “a better-than-even chance of avoiding more than a 2°C temperature rise, the carbon budget between 2011 and 2050” must be kept at “around 870–1240 GtCO2” p. 187. This translates to approximately one-third of global oil reserves, half of the global gas reserves, and over 80% of global coal reserves of unburnable fossil fuels. Figure 3 summarizes the regional distribution of these unburnable reserves. These figures are in line with other estimates of the stranded coal, oil, and gas assets by other experts and organizations, that must be kept in the ground, to meet the 2°C Paris commitments [5, 10, 74, 85, 86]. However, while in the end, all carbon must be phased out, less-carbon intensive energy carriers like gas might continue to operate as a “bridging fuel” to the carbon-free economy, in tandem with renewable energy. When considering short- and long-term nature of technology rebound effects, path dependency in policymaking, and carbon lock-in in different markets [16, 20, 87, 88, 89, 90, 91, 92], adopting adaptive strategies, incorporating technology transfer, and incentivizing international collaboration in energy research, are vital stratagems for managing the distributional impact of this energy transition process as well as upstream value chain requirements (such as future nuclear baseload supply and renewables-based hydrogen generation).
Figure 3.
Percent of regional distribution of unburnable fossil reserves before 2050 for the 2°C scenario—data from McGlade and Ekins [11], Table 1.
5. Mitigating stranded assets risks using ML and AI techniques
Returning to physical and financial carbon assets at risk of being stranded, ML and AI techniques can provide appealingly pragmatic Pareto-optimal solutions for mitigating stranding risks among different policy aspects instead of using scalarization, thereby creating a balanced transition to lower-carbon technology [93, 94, 95]. What causes assets to strand? As discussed in Section 2, multiple factors, including economic, technological like disruptive innovation, political, regulation, spatial, and societal norms, or environmental shocks, can lead to asset stranding. Stranding is not just a loss in economic value but also an irreversibility of the investments. This means that if the investments wiped out is reversible and can be adjusted for other purposes such as retooling an obsolete coal power plant to be used as a hydro generation facility; then the assets have not stranded since they can be put to different profitable use [63, 80, 86].
With respect to the unburnable carbon, stranding occurs when coal, oil, and gas companies, who have already committed heavy capital investment in related infrastructures such as exploitation, exploration, and pipelines, become hit by a sudden drop in commodity prices, leading to the stranding of their capital stocks. This could also happen when a government establishes an unanticipated Pigouvian fee, promoted by Pigou in a seminal article [96], on GHG emissions to correct for the unpriced environmental externality, either via a carbon price [97, 98] or a market-based emissions cap-and-trade mechanism [99]. This can have negative consequences for the market valuation of the upstream and downstream fossil fuel-based businesses and producers of electricity, leading to forced write-offs of their carbon assets [21] or their capital stocks getting stranded. For example, following the passage of the Powerplant and Industrial Fuel Use Act (FUA) in 1978 in response to the Arab oil embargo of 1973, a significant shortage of natural gas occurred, leading to a drop in natural gas-fired generation capacity additions. The unintended consequence of this policy-driven change in the national electricity generation mix was a shift to coal-fired generation capacity in the intervening years, leading to a rise in energy-related long-term carbon emissions.
In recent years, research shows that ML and AI are broadly powerful tools for technological progress that can be applied with a high impact in mitigating the transition to low-carbon technologies, especially in tackling the problem of stranded assets in the electricity sector. Power generation and demand forecasting is one area in which ML and AI techniques can improve policy vagaries and uncertainty about future demand, thereby mitigating the risk of stranding [17]. Below are the 10 distinct opportunities for ML and AI applications in the energy sector that include:
Electricity scheduling and dispatch: Improving electricity scheduling and dispatch mechanisms using ML and AI tools amidst increasing variable DER generation, storage, and flexible demand.
Energy data analytics and informatics: Using ML supervised models, e.g., that employ regression-based techniques on cellular network data, to generate information about low-data settings and determine where electricity power lines can be placed in regions unmapped, and help improve energy access [100].
Energy materials research: Applying ML, AI, optimization techniques, and physics to better understand the science of energy material’s crystal structure, to accelerate materials discovery for solar fuels that improves harnessing of energy from variable natural resources [101].
Natural gas methane detection and prevention: Employing ML and AI techniques to detect and prevent the leakage of methane from natural gas pipelines and compressor stations.
Nuclear fission and fusion: Application of ML and deep networks to speed up inspection of nuclear power plants and help design next-generation smart, modular nuclear reactors [102, 103].
Solar PV design and innovations: Using ML techniques to design controllable movable solar panels that maximize electricity production, for example, in bifacial solar modules and dual-orientation racking techniques [53, 104, 105, 106, 107].
Solar PV technical and economic potential estimation: Using ML to help estimate technical and economic potential of rooftop solar PV, e.g., by optimizing Light Detection and Ranging (LIDAR)-Geographic Information Systems (GIS) imagery-rendering of size and location data for rooftop solar panels [108, 109] .
Wind power management and monitoring: ML-driven condition monitoring (such as dimensionality reduction algorithm like Principal Component Analysis—PCA) of wind turbine blades, including optimization of blade fault detection, power curve monitoring, and temperature monitoring [110, 111].
Integrated transportation planning: Using AI and ML to improve vehicle engineering, shared mobility, and shift to lower-carbon options, like rail. In the long-term, ML and AI applications can support integrated intelligent infrastructure through planning, maintenance, and operations to make transportation more efficient though the GHG reduction, provide better demand forecasts, and support smart transit policy efforts such as autonomous vehicles, alternative fuels and electrification (e.g., electric vehicles, and vehicle-to-grid algorithms), and predicting battery state and degradation rate using supervised learning techniques [112, 113, 114, 115, 116].
Urban energy planning: With ML and AI applications, available building1 energy use data can be extrapolated to predict energy use at the city level. Furthermore, ML is uniquely capable of supporting improvements in “smart energy frameworks for smart cities” [25], including building codes, informing policymakers about utilizing urban rooftops for solar PV electricity generation [55, 108], retrofitting strategies using automated performance control [117], public-private partnerships to improve low-and moderate-income (LMI) stipulations and equitable electricity access [15, 64].
The above list is by no means exhaustive. The transformation to a low-carbon economy is occurring at an expanding rate. The technical innovations accompanying these carbon-free energy sources such as solar, wind, hydro, and geothermal energy is driving down the cost of these technologies as production increases and knowledge accumulation results from learning by doing. As a result, they are yielding substitutes for coal, oil, and progressively rendering coal, oil, and gas capital stock obsolete. It is expected that as this shift continues, new opportunities for ML and AI applications will become available, including in modeling consumer behavior and facilitating sustainable behavior change energy consumption action [3, 65, 118, 119], estimating and predicting the marginal emissions of residential energy utilization and thermal comfort in buildings in real time, on a scale of hours [57, 118], and game-theoretic modeling and design of socially beneficial energy policies like social norms, public opinions, stakeholder engagement, and education efforts [120, 121, 122]. Other breakthrough innovations might displace fossil fuels leading to stranding, and creating opportunities for ML-based electricity pricing techniques and rate design to set dynamic pricing of carbon, electricity, and consumer choice [1, 123, 124, 125, 126, 127], and multi-objective optimization to compute Pareto-optimal solutions for climate engineering, climate informatics, and solar geoengineering [58, 128, 129, 130]. There is a possibility that these technological innovations could create a sudden improvement in market evaluation of the renewable energy industries, while some assets of related carbon-intensive industries become stranded due to obsolescence, write-offs, or retirements.
Following the passing of the Paris Agreement on climate change, nations committed to keeping the global temperature below 2°C. Achieving this temperature target means coal, oil, and gas producers face stranding more than 80% of all these proven fossil fuel reserves and existing investments becoming stranded assets. These threats lead policymakers and market analysts to conclude that market evaluation and capital investments of some of these carbon-intensive firms risk being stranded, unless they fundamentally change their business models per the risk of asset stranding, to cushion themselves from unanticipated economic, technological, political, regulatory, spatial, social, and environmental changes, resulting in cheap renewable substitutes for coal, oil, and gas. A pragmatic and proactive response by governments is urgently required in the form of NDCs and climate policies to guide this transition, and that puts nations on a sustained path to the 1.5 or 2°C “carbon budget.” Such a process should avoid a disruptive and unorderly energy transition and macro shocks. Using ML and AI techniques to tackle the risks of stranded carbon assets and related infrastructure can enrich and inform this praxis. Stranded assets are not new in the energy sector; the physical impacts of climate change and the transition to a low-carbon economy have generally rendered redundant or obsolete electricity generation and storage assets. Low-carbon electricity systems, which come in variable and controllable forms, are essential to mitigating climate change. These systems present distinct opportunities for machine learning and artificial intelligence-powered techniques, making their applications prominent.
Sen and von Schickfus [62] calculate that €1.61 billion of security reserve or €13.38/MWh subsidy, is required to compensate coal energy assets in Germany at the risk of becoming stranded. Given the threats of sudden changes in the stringency of carbon policies and related abrupt repricing or retirement of fossil fuel assets, they also find that investors generally do care about stranded asset risk, but that they also expect to be financially compensated for stranded assets. This analysis highlights the threat of stranded asset risk in the coal industry and the need for understanding the interaction between policymaking and investors’ expectations. For example, the International Renewable Energy Agency (IRENA) [131] estimates that to meet the Paris Agreement’s 2°C temperature target, $1.9 trillion in electricity generation assets would be stranded after 2030. The report concludes that stranding will disproportionately affect $7 trillion in upstream energy infrastructure, of which three-quarters are in oil production. Institutional investors must tap ML and AI techniques ML to improve energy planning and system efficiency (e.g., detect and prevent the leakage of methane from natural gas pipelines, speed up inspection of nuclear power plants, and improve electricity scheduling and dispatch mechanisms). Given the vital role of the energy sector and its interrelation with the rest of the economy, using ML and AI to tackle stranded electricity assets is emerging as a cost-effective derisking strategy. Stranding and the risk of stranded carbon assets is a growing challenge requiring an interdisciplinary approach that brings together ideas from engineering, economics, and policy fields, as well as quantitative opportunities of ML, AI, optimization, and dynamical systems, to address interpretability, uncertainty quantification, and integration questions.
References
1.Nyangon J, Byrne J. Diversifying electricity customer choice: REVing up the New York energy vision for polycentric innovation. In: Tsvetkov PV, editor. Energy Systems and Environment. London, UK: IntechOpen; 2018. pp. 3-23
2.Pérez-Arriaga I, Knittel C. Utility of the Future: An MIT Energy Initiative Response to An Industry in Transition. MIT Energy Initiative: Boston, MA; 2016
3.Taminiau J, Banks JP, Bleviss D, Byrne J. Advancing transformative sustainability: A comparative analysis of electricity service and supply innovators in the United States. 2019;8(4):e337
4.IPCC. Global warming of 1.5°C, an IPCC special report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. In: Masson-Delmotte V, Zhai P, Pörtner HO, Roberts D, Skea J, Shukla PR, Pirani A, Moufouma-Okia W, Péan C, Pidcock R, Connors S, Matthews JBR, Chen Y, Zhou X, Gomis MI, Lonnoy E, Maycock T, Tignor M, Waterfield T, editors. Geneva, Switzerland: World Meteorological Organization; 2018
5.Rogelj J, Schaeffer M, Meinshausen M, Knutti R, Alcamo J, Riahi K, et al. Zero emission targets as long-term global goals for climate protection. Environmental Research Letters. 2015;10(10):105007
6.Riahi K, Kriegler E, Johnson N, Bertram C, den Elzen M, Eom J, et al. Locked into Copenhagen pledges—Implications of short-term emission targets for the cost and feasibility of long-term climate goals. Technological Forecasting and Social Change. 2015;90:8-23
7.Verbong GPJ, Beemsterboer S, Sengers F. Smart grids or smart users? Involving users in developing a low carbon electricity economy. Energy Policy. 2013;52:117-125
8.International Energy Agency. World Energy Outlook 2016. Paris, France: International Energy Agency (IEA); 2016
9.Caldecott B, Dericks G, Tulloch DJ, Kruitwagen L, Kok I. Stranded Assets and Thermal Coal in Japan: An Analysis of Environment-Related Risk Exposure. Oxford, UK: University of Oxford’s Smith School of Enterprise and the Environment; 2016
10.Carbon Tracker Initiative. Unburnable carbon: Are the world’s financial markets carrying a carbon bubble? 2011. Available from: https://carbontracker.org/reports/carbon-bubble/ [Accessed: April 21, 2020]
11.McGlade C, Ekins P. The geographical distribution of fossil fuels unused when limiting global warming to 2°C. Nature. 2015;517(7533):187-190
12.Bos K, Gupta J. Stranded assets and stranded resources: Implications for climate change mitigation and global sustainable development. Energy Research and Social Science. 2019;56:101215
13.Caldecott B, Howarth N, McSharry P. Stranded Assets in Agriculture: Protecting Value from Environment-Related Risks. Oxford, England: University of Oxford’s Smith School of Enterprise and the Environment; 2013
14.Loorbach D. Transition management for sustainable development: A prescriptive, complexity-based governance framework. Governance. 2010;23(1):161-183
15.Bridge G, Bouzarovski S, Bradshaw M, Eyre N. Geographies of energy transition: Space, place and the low-carbon economy. Energy Policy. 2013;53:331-340
16.Klitkou A, Bolwig S, Hansen T, Wessberg N. The role of lock-in mechanisms in transition processes: The case of energy for road transport. Environmental Innovation and Societal Transitions. 2015;16:22-37
17.van der Ploeg F, Rezai A. Stranded assets in the transition to a carbon-free economy. Annual Review of Resource Economics. 2020
18.Sovacool BK. The history and politics of energy transitions: Comparing contested views and finding common ground. In: The Political Economy of Clean Energy Transitions. GB: Oxford University Press; 2017. pp. 16-35
19.Byrne J, Lund PD. Clean energy transition—Our urgent challenge: An editorial assay. Wiley Interdisciplinary Reviews: Energy and Environment. 2017;6(1)
20.Aklin M, Urpelainen J. Political competition, path dependence, and the strategy of sustainable energy transitions. American Journal of Political Science. 2013;57(3):643-658
21.Bertram C, Johnson N, Luderer G, Riahi K, Isaac M, Eom J. Carbon lock-in through capital stock inertia associated with weak near-term climate policies. Technological Forecasting and Social Change. 2015;90:62-72
22.MacDonald S, Winner B, Smith L, Juillerat J, Belknap S. Bridging the rural efficiency gap: Expanding access to energy efficiency upgrades in remote and high energy cost communities. Energy Efficiency. 2019
23.Liu Y, Yao X, Wei T. Energy efficiency gap and target setting: A study of information asymmetry between governments and industries in China. China Economic Review. 2019;57:101341
24.Nyangon J. Distributed energy generation systems based on renewable energy and natural gas blending: New business models for economic incentives, electricity market design and regulatory innovation [PhD dissertation]. College of Engineering, University of Delaware; 2017
25.Nyangon J. Smart energy frameworks for smart cities: The need for polycentrism. In: Augusto JC, editor. Handbook of Smart Cities. Switzerland: Springer; 2021. pp. 1-32
26.Lopez-Rodriguez I, Hernandez-Tejera M, Lopez AL. Methods for the management of distributed electricity networks using software agents and market mechanisms: A survey. Electric Power Systems Research. 2016;136:362-369
27.Kuiken D, Más HF. Integrating demand side management into EU electricity distribution system operation: A Dutch example. Energy Policy. 2019;129(June 2019):153-160
28.Stavrakas V, Flamos A. A modular high-resolution demand-side management model to quantify benefits of demand-flexibility in the residential sector. Energy Conversion and Management. 2020;205:112339
29.Shomali A, Pinkse J. The consequences of smart grids for the business model of electricity firms. Journal of Cleaner Production. 2016;112(Part 5):3830-3841
30.Wang H, Wang S, Tang R. Development of grid-responsive buildings: Opportunities, challenges, capabilities and applications of HVAC systems in non-residential buildings in providing ancillary services by fast demand responses to smart grids. Applied Energy. 2019;250(1):697-712
31.Sheha M, Mohammadi K, Powell K. Solving the duck curve in a smart grid environment using a non-cooperative game theory and dynamic pricing profiles. Energy Conversion and Management. 2020;220:113102
32.CAISO. What the Duck Curve Tells Us about Managing a Green Grid. California, United States: California Independent System Operator (CAISO); 2014
33.Guidehouse Insights. Energy as a Service Overview. United States: Guidehouse Insights; 2019
34.Stern N. The Economics of Climate Change: The Stern Review. Cambridge, UK: Cambridge University Press; 2006
35.Zhen Z, Tian L, Ye Q. A simple estimate for the social cost of carbon. Energy Procedia. 2018;152:768-773
36.Iribarren D, Martín-Gamboa M, Navas-Anguita Z, García-Gusano D, Dufour J. Influence of climate change externalities on the sustainability-oriented prioritisation of prospective energy scenarios. Energy. 2020;196:117179
37.Holahan R, Kashwan P. Disentangling the rhetoric of public goods from their externalities: The case of climate engineering. Global Transitions. 2019;1:132-140
38.Byrne J, Lund PD. Sustaining our common future: Transformative, timely, commons-based change is needed. Wiley Interdisciplinary Reviews: Energy and Environment. 2019;8(1)
39.Tol RSJ. A social cost of carbon for (almost) every country. Energy Economics. 2019;83:555-566
40.Rozenberg J, Vogt-Schilb A, Hallegatte S. Instrument choice and stranded assets in the transition to clean capital. Journal of Environmental Economics and Management. 2020;100:102183
41.Byrne J, Wang Y, Lee H, Kim J. An equity- and sustainability-based policy response to global climate change. Energy Policy. 1998;26(4):335-343
42.Roberts C, Geels FW, Lockwood M, Newell P, Schmitz H, Turnheim B, et al. The politics of accelerating low-carbon transitions: Towards a new research agenda. Energy Research and Social Science. 2018;44:304-311
43.Rogge KS, Reichardt K. Policy mixes for sustainability transitions: An extended concept and framework for analysis. Research Policy. 2016;45(8):1620-1635
44.Voyant C, Notton G, Kalogirou S, Nivet M, Paoli C, Motte F, et al. Machine learning methods for solar radiation forecasting: A review. Renewable Energy. 2017;105:569-582
45.Das UK, Tey KS, Seyedmahmoudian M, Mekhilef S, Idris MYI, Van Deventer W, et al. Forecasting of photovoltaic power generation and model optimization: A review. Renewable and Sustainable Energy Reviews. 2018;81:912-928
46.Wan C, Song Y, Xu Z, Yang G, Nielsen AH. Probabilistic wind power forecasting with hybrid artificial neural networks. Electric Power Components & Systems. 2016;44(15):1656-1668
47.Nyangon J, Alabbas N, Agbemabiese L. Entangled systems at the energy-water-food nexus: Challenges and opportunities. In: Rao P, Patil Y, editors. Reconsidering the Impact of Climate Change on Global Water Supply, Use, and Management United States. IGI Global; 2017. pp. 145-165
48.Nyangon J, Byrne J, Taminiau J. Assessment of price convergence between natural gas and solar photovoltaic in the U.S. electricity market. Wiley Interdisciplinary Reviews: Energy and Environment. 2017;6(3):1-20
49.Lightner EM, Widergren SE. An orderly transition to a transformed electricity system. IEEE Transactions on Smart Grid. 2010;1(1):3-10
50.Bao W, Yue J, Rao Y. A deep learning framework for financial time series using stacked autoencoders and long-short term memory. PLoS One. 2017;12(7):1-5
51.Gui EM, MacGill I. Typology of future clean energy communities: An exploratory structure, opportunities, and challenges. Energy Research & Social Science. 2018;35:94-107
52.EIA. Annual Energy Outlook 2020 (AEO2020) Reference Case: Electricity. Washington, DC: U.S. Energy Information Administration (EIA); 2020
53.Byrne J, Nyangon J, Hegedus S, Chajes M, Taminiau J, Ahmed N, et al. Feasibility Study of City-Scale Solar Power Plants Using Public Buildings: Case Studies of Newark and Wilmington Delaware with Early Investigations of Bifacial Solar Modules and Dual Orientation Racking as Tools for City-Scale Solar Development. Newark, DE: CEEP, University of Delaware; 2019
54.Munyon VV, Bowen WM, Holcombe J. Vehicle fuel economy and vehicle miles traveled: An empirical investigation of Jevon’s Paradox. Energy Research and Social Science. 2018;38:19-27
55.Byrne J, Taminiau J. Utilizing the urban fabric as the solar power plant of the future. In: Droege P, editor. Urban Energy Transition. 2nd ed. New York City: Elsevier; 2018. pp. 31-49
56.Sorrell S. Jevons’ Paradox revisited: The evidence for backfire from improved energy efficiency. Energy Policy. 2009;37(4):1456-1469
57.Ngarambe J, Yun GY, Santamouris M. The use of artificial intelligence (AI) methods in the prediction of thermal comfort in buildings: Energy implications of AI-based thermal comfort controls. Energy and Buildings. 2020;211:109807
58.Tahmasebi P, Kamrava S, Bai T, Sahimi M. Machine learning in geo- and environmental sciences: From small to large scale. Advances in Water Resources. 2020;142:103619
59.Rubin ES, Azevedo IML, Jaramillo P, Yeh S. A review of learning rates for electricity supply technologies. Energy Policy. 2015;86:198-218
60.Nyangon J. The U.S. shale gas revolution and its implications for international energy policy. Green Monitor. 2015;3(1):184-190
61.McClellan J, Keith DW, Apt J. Cost analysis of stratospheric albedo modification delivery systems. Environmental Research Letters. 2012;7(3):034019
62.Sen S, von Schickfus M. Climate policy, stranded assets, and investors’ expectations. Journal of Environmental Economics and Management. 2020;100:102277
63.Ansari D, Holz F. Between stranded assets and green transformation: Fossil-fuel-producing developing countries towards 2055. World Development. 2020;130:104947
64.Heiskanen E, Johnson M, Robinson S, Vadovics E, Saastamoinen M. Low-carbon communities as a context for individual behavioural change. Energy Policy. 2010;38(12):7586-7595
65.Nyangon J, Byrne J. Spatial energy efficiency patterns in New York and implications for energy demand and the rebound effect. Energy Sources, Part B: Economics, Planning, and Policy. 2020;15(3):1-26
66.Shaker RR, Altman Y, Deng C, Vaz E, Forsythe KW. Investigating urban heat island through spatial analysis of New York City streetscapes. Journal of Cleaner Production. 2019;233:972-992
67.Zhou W, Wang J, Cadenasso ML. Effects of the spatial configuration of trees on urban heat mitigation: A comparative study. Remote Sensing of Environment. 2017;195:1-12
68.Schallenberg-Rodriguez J. Renewable electricity support systems: Are feed-in systems taking the lead? Renewable and Sustainable Energy Reviews. 2017;76(September 2017):1422-1439
69.Hultman N, Rebois D, Scholten M, Ramig C. The greenhouse impact of unconventional gas for electricity generation. Environmental Research Letters. 2011;6(4)
70.Byrne J, Nyangon J, Deblauwe H, Oster C, Shin S, Xu J, et al. Measuring urban sustainability through common indicators and peer city benchmarking: Assessing sustainability assets for performance improvement and economic and environmental progress in Delaware. SSRN Online Journal. 2017. Newark, DE
71.Lele SM, Brondízio ES, Byrne J, Mace GM, Martínez Alier J, Lupp JR. Rethinking Environmentalism: Linking Justice, Sustainability, and Diversity. Cambridge, MA: MIT Press; 2019
72.Taminiau J, Nyangon J, Lewis AS, Byrne J. Sustainable business model innovation: Using polycentric and creative climate change governance. In: Fields Z, editor. Collective Creativity for Responsible and Sustainable Business Practice. Hershey, PA: IGI Global; 2017. pp. 140-159
73.Kennedy CA, Stewart I, Facchini A, Cersosimo I, Mele R, Chen B, et al. Energy and material flows of megacities. PNAS. 2015;112(19):5985-5990
74.Meinshausen M, Meinshausen N, Hare W, Raper SCB, Frieler K, Knutti R, et al. Greenhouse-gas emission targets for limiting global warming to 2°C. Nature. 2009;458(7242):1158-1162
75.Goldthau A. Rethinking the governance of energy infrastructure: Scale, decentralization and polycentrism. Energy Research and Social Science. 2014;1(March 2014):134-140
76.Byrne J, Taminiau J, Kim KN, Lee J, Seo J. Multivariate analysis of solar city economics: Impact of energy prices, policy, finance, and cost on urban photovoltaic power plant implementation. Wiley Interdisciplinary Reviews: Energy and Environment. 2017;6(4):1-16
77.Verdolini E, Vona F, Popp D. Bridging the gap: Do fast reacting fossil technologies facilitate renewable energy diffusion? Energy Policy. 2018;116:242-256
78.BP. The Statistical Review of World Energy 2020. 69th ed. London, UK: BP; 2020
79.Larrea C, Murmis MR. Unburnable Carbon and Biodiversity: A Global Fund for Keeping Fossil Fuels in the Ground in Biodiversity Hotspots of Developing Countries. In: Fossil Fuel Supply and Climate Policy Conference. Quito, Ecuador; 2016
80.Caldecott B. Introduction to special issue: Stranded assets and the environment. Journal of Sustainable Finance & Investment. 2017;7(1):1-13
81.Bullard N. Fossil Fuel Divestment: A $5 trillion Challenge. New York, NY: Bloomberg New Energy Finance (BNEF); 2014
82.Jakob M, Hilaire J. Unburnable fossil-fuel reserves. Nature. 2015;517(7533):150-151
83.Piggot G. The influence of social movements on policies that constrain fossil fuel supply. Climate Policy. 2018;18(7):942-954
84.Bratman E, Brunette K, Shelly DC, Nicholson S. Justice is the goal: Divestment as climate change resistance. Journal of Environmental Studies and Sciences. 2016;6(4):677-690
85.Millar RJ, Fuglestvedt JS, Friedlingstein P, Rogelj J, Grubb MJ, Matthews HD, et al. Emission budgets and pathways consistent with limiting warming to 1.5 °C. Nature Geoscience. 2017;10(10):741-747
86.van der Ploeg F, Rezai A. The risk of policy tipping and stranded carbon assets. Journal of Environmental Economics and Management. 2020;100:102258
87.Seebauer S. The psychology of rebound effects: Explaining energy efficiency rebound behaviours with electric vehicles and building insulation in Austria. Energy Research and Social Science. 2018;46:311-320
88.Fischedick M, Byrne J, Hermwille L, Taminiau J, Luhmann H, Stelzer F, et al. Reflections on the state of climate change policy: From COP21 to cities. In: Lele S, Brondizio ES, Byrne J, Mace GM, Martinez-Alier J, editors. Rethinking Environmentalism: Linking Justice, Sustainability, and Diversity. Cambridge, MA: MIT Press; 2018. ISBN: 9780262038966
89.Font Vivanco D, Kemp R, van der Voet E. How to deal with the rebound effect? A policy-oriented approach. Energy Policy. 2016;94:114-125
90.Ivanova D, Vita G, Wood R, Lausselet C, Dumitru A, Krause K, et al. Carbon mitigation in domains of high consumer lock-in. Global Environmental Change. 2018;52:117-130
91.Wei T, Zhou J, Zhang H. Rebound effect of energy intensity reduction on energy consumption. Resources, Conservation and Recycling. 2019;144:233-239
92.Fan R, Luo M, Zhang P. A study on evolution of energy intensity in China with heterogeneity and rebound effect. Energy. 2016;99:159-169
93.Choi D, Ahn Y, Choi DG. Multi-criteria decision analysis of electricity sector transition policy in Korea. Energy Strategy Reviews. 2020;29:100485
94.Antunes CH, Henriques CO. Multi-objective optimization and multi-criteria analysis models and methods for problems in the energy sector. In: Greco S, Ehrgott M, Figueira JR, editors. Multiple Criteria Decision Analysis: State of the Art Surveys. New York, NY: Springer New York; 2016. pp. 1067-1165
95.Wang N, Heijnen PW, Imhof PJ. A multi-actor perspective on multi-objective regional energy system planning. Energy Policy. 2020;143:111578
96.Pigou AC. A special levy to discharge war debt. The Economic Journal. 1918;28(110):135-156
97.Creti A, Jouvet P, Mignon V. Carbon price drivers: Phase I versus Phase II equilibrium? Energy Economics. 2012;34(1):327-334
98.Chesney M, Gheyssens J, Pana AC, Taschini L. Environmental Finance and Investments. 2nd ed. Switzerland: Springer; 2016
99.Coase RH. The problem of social cost. In: Gopalakrishnan C, editor. Classic Papers in Natural Resource Economics. London: Palgrave Macmillan UK; 1960. pp. 87-137
100.Gershenson D, Rohrer B, Lerner A. A new predictive model for more accurate electrical grid mapping. 2019; Available from: https://engineering.fb.com/connectivity/electrical-grid-mapping. [Accessed: May 23, 2020]
101.Butler KT, Davies DW, Cartwright H, Isayev O, Walsh A. Machine learning for molecular and materials science. Nature. 2018;559(7715):547-555
102.Chen F, Jahanshahi M. NB-CNN: Deep learning-based crack detection using convolutional neural network and naïve Bayes data fusion. IEEE Transactions on Industrial Electronics. 2018;65(5):4392-4400
103.Kates-Harbeck J, Svyatkovskiy A, Tang W. Predicting disruptive instabilities in controlled fusion plasmas through deep learning. Nature. 2019;568(7753):526-531
104.Tillmann P, Jäger K, Becker C. Minimising the levelised cost of electricity for bifacial solar panel arrays using Bayesian optimisation. Sustainable Energy & Fuels. 2020;4(1):254-264
105.Ripalda JM, Buencuerpo J, García I. Solar cell designs by maximizing energy production based on machine learning clustering of spectral variations. Nature Communications. 2018;9(1):5126
106.Bayesian optimization for maximum power point tracking in photovoltaic power plants. In: 2016 European Control Conference (ECC) Aalborg, Denmark: IEEE. 2016
107.Kopecek R, Shoukry I, Libal J. Cost/kWh thinking and bifaciality: Two allies for low-cost PV of the future. Photovoltaics International. 2015;(4th Quarter):88-97
108.Taminiau J, Byrne J. City-scale urban sustainability: Spatiotemporal mapping of distributed solar power for New York City. WIREs Energy and Environment. 2020:1-24. DOI: 10.1002/wene.374
109.Walch A, Castello R, Mohajeri N, Scartezzini J. Big data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty. Applied Energy. 2020;262:114404
110.Jiménez AA, García Márquez FP, Moraleda VB, Gómez Muñoz CQ. Linear and nonlinear features and machine learning for wind turbine blade ice detection and diagnosis. Renewable Energy. 2019;132:1034-1048
111.Stetco A, Dinmohammadi F, Zhao X, Robu V, Flynn D, Barnes M, et al. Machine learning methods for wind turbine condition monitoring: A review. Renewable Energy. 2019;133:620-635
112.Davis SJ, Lewis NS, Shaner M, Aggarwal S, Arent D, Azevedo IL, et al. Net-zero emissions energy systems. Science. 2018;360(6396):eaas9793
113.Zeng W, Miwa T, Morikawa T. Application of the support vector machine and heuristic k-shortest path algorithm to determine the most eco-friendly path with a travel time constraint. Transportation Research Part D: Transport and Environment. 2017;57:458-473
114.Kaack LH, Vaishnav P, Morgan MG, Azevedo IL, Rai S. Decarbonizing intraregional freight systems with a focus on modal shift. Environmental Research Letters. 2018;13(8)
115.Sovacool BK, Kester J, Noel L, Zarazua de Rubens G. Are electric vehicles masculinized? Gender, identity, and environmental values in Nordic transport practices and vehicle-to-grid (V2G) preferences. Transportation Research Part D: Transport and Environment. 2019;72:187-202
116.Kester J, Noel L, Lin X, Zarazua de Rubens G, Sovacool BK. The coproduction of electric mobility: Selectivity, conformity and fragmentation in the sociotechnical acceptance of vehicle-to-grid (V2G) standards. Journal of Cleaner Production. 2019;207:400-410
117.Taminiau J, Sanchez CD, John B, Shin S, Xu J. Risk mitigation in energy efficiency retrofit projects using automated performance control. In: Nyangon J, Byrne J, editors. Sustainable Energy Investment: Technical, Market and Policy Innovations to Address Risk. London, UK: IntechOpen; 2020
118.Kara EC, Roberts CM, Tabone M, Alvarez L, Callaway DS, Stewart EM. Disaggregating solar generation from feeder-level measurements. Sustainable Energy, Grids and Networks. 2018;13:112-121
119.Papadopoulos S, Kontokosta CE. Grading buildings on energy performance using city benchmarking data. Applied Energy. 2019;233-234:244-253
120.Safarzadeh S, Rasti-Barzoki M. A game theoretic approach for assessing residential energy-efficiency program considering rebound, consumer behavior, and government policies. Applied Energy. 2019;233-234:44-61
121.Giest S, Mukherjee I. Behavioral instruments in renewable energy and the role of big data: A policy perspective. Energy Policy. 2018;123:360-366
122.Zhang J, Zhang W, Song Q, Li X, Ye X, Liu Y, et al. Can energy saving policies drive firm innovation behaviors? Evidence from China. Technological Forecasting and Social Change. 2020;154:119953
123.Faruqui A, Arritt K, Sergici S. The impact of AMI-enabled conservation voltage reduction on energy consumption and peak demand. The Electricity Journal. 2017;30(2):60-65
124.Mowlaei ME, Saniee Abadeh M, Keshavarz H. Aspect-based sentiment analysis using adaptive aspect-based lexicons. Expert Systems with Applications. 2020;148:113234
125.Faruqui A, Leyshon K. Fixed charges in electric rate design: A survey. The Electricity Journal. 2017;30(10):32-43
126.Wei S, Chongchong Z, Cuiping S. Carbon pricing prediction based on wavelet transform and K-ELM optimized by bat optimization algorithm in China ETS: The case of Shanghai and Hubei carbon markets. Carbon Management. 2018;9(6):605-617
127.Faruqui A, Hajos A, Hledik RM, Newell SA. Fostering economic demand response in the Midwest ISO. Energy. 2010;35(4):1544-1552
128.Han J, Ahn Y, Lee I. A multi-objective optimization model for sustainable electricity generation and CO2 mitigation (EGCM) infrastructure design considering economic profit and financial risk. Applied Energy. 2012;95:186-195
129.Ma J, Cheng JCP, Jiang F, Gan VJL, Wang M, Zhai C. Real-time detection of wildfire risk caused by powerline vegetation faults using advanced machine learning techniques. Advanced Engineering Informatics. 2020;44:101070
130.Jakariya M, Alam MS, Rahman MA, Ahmed S, Elahi MML, Khan AMS, et al. Assessing climate-induced agricultural vulnerable coastal communities of Bangladesh using machine learning techniques. Science of the Total Environment. 2020:140255
131.IRENA. Perspectives for the Energy Transition: Investment Needs for a Low-Carbon Energy System. Abu Dhabi: International Renewable Energy Agency (IRENA); 2017
Notes
The IPCC classifies mitigation actions in buildings into four categories: carbon efficiency (switching to low-carbon fuels or to natural refrigerants); energy efficiency (reducing energy waste through insulation, efficient appliances, better heating and ventilation, or other similar measures); system and infrastructure efficiency (e.g. passive house standards, urban planning, and district cooling and heating); and service demand reduction (behavioral and lifestyle changes).
Written By
Joseph Nyangon
Submitted: 30 July 2019Reviewed: 27 July 2020Published: 10 March 2021
Open access peer-reviewed chapter
