Resource Adequacy in the Desert Southwest
Nick Schlag, Adrian Au, et. al., February 2022 (Energy and Environmental Economics)
In the aftermath of recent blackouts in California and Texas, the subjects of reliability and resource adequacy have risen to national prominence. Regulators and policymakers – as well as the general public and media – have taken a keen interest in these topics, and many have questioned whether the industry is adequately prepared to confront the challenge of preserving reliability during a period of rapid transition. Yet despite its importance and the recent attention it has received, the topic of resource adequacy – and what will be needed to ensure it can be maintained during the transition to cleaner energy sources – remains an esoteric and poorly understood aspect of power system planning. This study sheds light on this important topic to support utilities, regulators, policymakers, and stakeholders in the Desert Southwest as they endeavor to plan, construct, and operate a reliable grid. The goals of this study are threefold:
1. Examine the current situation in the Desert Southwest in light of recent challenges in neighboring regions and identify any immediate risks to reliability in the region;
2. Identify and define best practices for resource adequacy planning that will provide a durable foundation for utilities’ efforts to preserve reliability within the region; and
3. Utilize these techniques to evaluate the region’s readiness to meet the resource adequacy challenges it will face over the next decade.
Study Highlights & Recommendations
► Load growth and resource retirements are creating a significant and urgent need for new resources in the Southwest region; maintaining regional reliability will hinge on whether utilities can add new resources quickly enough to meet this growing need and will require a pace of development largely unprecedented for the region
► An increasingly significant share of long-term resource needs is expected to be met with solar and storage resources, but a large quantity of “firm” generation capacity – including the region’s nuclear and natural gas resources – will also be needed to maintain reliability
► Substantial reliability risks remain as the region’s electricity resource portfolio transitions, most notably: weather- and climate-related uncertainties, performance of battery storage, and risks related to the timing of new additions
► To plan most effectively for resource adequacy, utilities should utilize the best practices identified in this study to the extent practicable, including the use of probabilistic methods to assess the need for capacity and the broad application of an ELCC methodology to assess the capacity value of all resources on an equivalent basis
New Challenges in Resource Adequacy in the Southwest
Due to a far-reaching combination of factors – technological, economic, policy, environmental and societal dynamics – the energy landscape of the Southwest region is in a period of rapid transformation. Many of these changes have direct implications on the utilities’ ability to maintain reliable electric service. Figure ES-2 summarizes six key trends that are fundamentally altering the Southwest’s energy system and will have large and immediate ramifications for resource adequacy planning in the region.
While each of these trends will impact utilities’ efforts to plan for reliability, the shift towards a portfolio more heavily reliant on renewables, storage, and distributed energy resources is notable because it will require advances beyond the simple techniques and common heuristics that have been used in planning for decades. The North American Electric Reliability Corporation (NERC) has described this transition “the greatest challenge to reliability”1 ; a growing body of research has shed light on the complex dynamics of how variable and energy-limited resources impact resource adequacy (illustrated in Figure ES-3):
1. As the penetration of variable resources grows and traditional generation retires, the periods in which the system is most vulnerable to reliability risks shift away from the traditional peak and toward periods of lower renewable production; this effect is exemplified by the shift in reliability risk to the evening net peak that occurs as solar penetration increases.
2. As the penetration of energy-limited resources grows, the risk of loss of load events will spread across an increasing number of hours; as the number of hours in which the system is at risk increases, the value of energy-limited resources with finite durations diminishes.
3. Variable and energy-limited resources exhibit complex “interactive effects,” meaning that the combined value of a portfolio of resources may differ from the sum of its individual parts.
Best Practices for Resource Adequacy Planning
The trends described above pose challenges to resource adequacy planners, but these challenges are not unique to the Southwest region. Utilities, regulators, and stakeholders throughout the country and around the world have already taken important steps to modernize their approaches to resource adequacy planning. “Best practices” continue to evolve as the understanding of these challenges advances and new information becomes available. However, the basic foundation of a robust framework for future resource adequacy planning is well-established and relies on the use of a loss of load probability (LOLP) model to (a) establish a planning reserve margin (PRM)requirement and (b) evaluate the effectiveness of resources using an effective load carrying capability (ELCC) methodology.
Probabilistic methods for resource adequacy analysis (or LOLP models) were first popularized in the middle of the twentieth century when planners recognized the usefulness of measuring risks to reliability statistically based on probabilities of extreme weather events and power plant outages. Today and in the future, reliability outcomes will continue to depend on weather variability (and its impacts on load, renewables, and other resources) and generator availability; the idea of a probabilistic approach to measuring reliability risk remains fundamentally sound, and the methods established in this early era serve as a foundation for the future of resource adequacy planning. However, the complexity of the probabilistic simulations needed will increase significantly as an unavoidable consequence of the transition to a portfolio that is less reliant upon conventional firm resources. The future of resource adequacy depends upon continued enhancements to probabilistic methods and data that capture this complexity, including simulation of chronological operations and resource interactions and considering weather variability, energy use limitations, and evolving load patterns.
While rigorous probabilistic modeling is essential to planning for resource adequacy for a power system, it is also important to understand how individual resources contribute to system reliability. To this end, a complementary capacity-based accounting construct akin to the familiar “planning reserve margin” will also remain useful. The key to robust capacity accounting is that all megawatts of capacity – both the requirement and the contribution of resources – be denominated in terms of “perfect capacity,” a unit of capacity that is available in all hours of the year at full capacity. The use of this fictional benchmark to both establish the requirement and count resources towards it provides for a balanced, technologyagnostic framework that values each resource based on its relative contribution to system needs.
Within this framework, the capacity value assigned to each resource (or portfolio of resources) should be determined using an ELCC methodology, which relies on the same LOLP modeling techniques to determine the amount of perfect capacity that provides an equivalent value to system reliability. Properly applied, an ELCC-based framework for capacity accreditation naturally accounts for the oft-cited complications that will arise in this transition, including the “shift to the net peak,” the need to account for energy sufficiency as well as capacity, and the saturation effects and diversity benefits that accrue to portfolios of variable and energy-limited resources. ELCC is therefore broadly viewed as the cornerstone of a robust approach to capacity accreditation and has quickly gained widespread usage within the industry.
This study relies on E3’s Renewable Energy Capacity Planning (RECAP) model, a chronological loss-ofload probability (LOLP) model to analyze the evolution of resource adequacy needs of the Southwest over the decade. The analysis addresses three questions over this time horizon:
How much new capacity is needed to ensure resource adequacy in the region?
How effective are different types of resources in meeting this need, considering their specific constraints and limitations?
Do the utilities’ current resource plans, as reflected by the portfolios produced in their IRPs, position the region to meet resource adequacy needs in the future?
Regional Demand Forecast
A region’s demand for electricity – in particular, its highest “peak” demand – is the main driver of its capacity needs for resource adequacy. For this study, an hourly future load forecast representative of the Southwest region as a whole was developed through aggregation of individual utilities’ annual load forecasts and historical hourly load shapes. In aggregate, peak demand in the region is forecasted to grow significantly in the coming years due to net migration to major population centers in the region, increased adoption of electric vehicles, and the growing trend of new large commercial and industrial customers. Based on utilities’ projected impacts of these trends, regional coincident peak under "typical” weather conditions is expected to grow by roughly 700 MW per year across the study horizon, reaching 26,700 MW by 2025 and 31,700 MW by 2033. Of course, more extreme weather conditions that occur during some years could result in even higher peak demands; this possibility is captured in the analysis by simulating hourly load shapes under 70 distinct weather years to capture potential year-to-year variability in extreme temperatures and peak demand.
Changing Characteristics of Customer Demand
Changing customer preferences and increasing customer engagement also has implications for how utilities plan for resource adequacy. Distributed energy resources (DERs) – including solar, energy storage, and demand response capable devices such as programmable thermostats – are growing in popularity, and their adoption changes how customers consume – and produce and store – electricity. NER ’s 202 LTRA succinctly summarizes the opportunities and complexities resulting from increased deployment of DERs:
“Distributed energy resources DER growth promises both opportunity and risks for reliability. Increased DER penetrations can improve local resilience and offset peak electric demand on the [bulk power system]. However DER can also increase variability and uncertainty in demand and therefore requires careful attention in planning for resource adequacy and energy sufficiency. DERs also increase the complexity of operating the BPS as operators often lack visibility into the effect of the DER on loads. Consequently, there is an immediate concern to ensure that data transfer, models, and information protocols are in place to support BPS planners and operators.”15
In many ways, the effects of DERs on resource adequacy will parallel the impacts of utility-scale non-firm resources: they are generally variable and/or energy-limited and will be subject to saturation effects and interactive effects. A durable framework for resource adequacy must therefore account for the impacts of customer-sited resources in a manner that accounts for their contributions to resource adequacy consistent with methods applied to utility-scale resources.
Electrification of new end uses will also have implications for future resource adequacy planning. Transportation electrification is already occurring today, but electrification of buildings and industry may follow as the imperative to electrify in pursuit of economy-wide decarbonization intensifies. Growing shares of these new end uses will further add complexity to resource adequacy planning, as the shape of electricity demand will evolve in the future. Transportation load impacts are both uncertain and complex, since they depend on customer driving behavior, charging infrastructure availability (home vs. workplace vs. public), charging speed (high-power rapid charging vs. slower overnight charging), charging costs, and electricity rate design.
Electrification of building loads will further increase resource adequacy needs in deeply decarbonized electric grids due to its outsized impact on winter loads. The addition of load during winter heating seasons further compounds the challenge that planners will have to ensure adequacy during the winter as well as the summer. A recent analysis of wind and solar droughts – defined as week-long anomalies of low wind and solar resource availability – in the Western Interconnection notes that these periods tend to occur during the coldest periods of the year, during which demand for space heating would be highest:
“Compound wind and solar droughts occurred seasonally when [heating degree days] were largest and the synoptic circulation associated with the compound drought events exacerbates this to a small degree. This means that the electrification of heating could potentially make these compound wind and solar droughts high stress events on a hypothetical underlying energy system (though this may be simultaneously mitigated by global warming).” 16
But while meeting newly electrified loads will likely require additional resources, the addition of these new loads also offers opportunities for new demand-side flexibility. Electric water heating has already proven itself as a flexible load resource and space heating may provide similar demand response opportunities as space cooling has (one of the primary demand response resources today). Industrial customers often have savvy energy managers dedicated to minimizing energy costs, who are likely to unlock relevant load flexibility opportunities…
Regional Resource Portfolios
This study focuses on the ability of two future resource portfolios to meet the region’s reliability needs in two snapshot years (2025 and 2033); the installed capacity of different resources in each of these scenarios is illustrated in Figure ES-7. All four portfolios incorporate retirements of coal and natural gas resources as currently planned by utilities within the region, totaling 1,400 MW of capacity by 2025 and 5,400 MW by 2033. New additions vary according to scenario:
The “Existing Committed Resources” scenarios include only new resources that have executed contracts with utilities and/or requisite regulatory approvals, which include roughly 3,000 MW of new solar and 1,200 MW of new energy storage.
The “IRP Portfolios” scenarios include all future resource additions captured in utilities’ current IRPs (or comparable planning processes) in addition to the existing & committed resources; these additions total roughly 10,000 MW of new installed capacity by 2025 and 35,000 MW by 2033, comprising large amounts of solar and storage and smaller amounts of wind, geothermal, demand response, and natural gas.
Summary Reliability Statistics
Outputs from the LOLP simulation of each of these four scenarios are summarized in Table ES-1. These results inform several notable observations:
As of 2021, the region’s loads and resources were roughly in balance; the frequency of expected unserved energy events was slightly higher than a traditional “one-day-in-ten-years” reliability standard. This reliability benchmark (an LOLE of 0.1 days per year) is used throughout this study as a reference point for resource adequacy.
Existing and committed resources alone will be insufficient to meet the region’s reliability needs. Without additional resources, the region’s resources would be insufficient to meet demand 2 days each year – far more than envisioned in a “one-day-in-ten-year” standard. Filling this gap will require close to 4,000 MW of new effective capacity by 2025 and over 13,000 MW by 2025.
The utilities’ plans for new resources, as reflected in their IRPs, appear sufficient to meet regional reliability needs as defined by a “one-day-in-ten-years” standard under Base Case assumptions of load and resource performance.
1. Load growth and resource retirements are creating a significant need for new resources in the Southwest region
2. Utilities’ current resource plans have identified enough resources to maintain regional reliability over the next decade
3. A significant share of long-term resource needs is expected to be met with solar and storage, which together are well-suited to meet a large portion of the region’s loads on summer peak days
4. The Southwest will continue to rely on a large quantity of “firm” generation resources to maintain resource adequacy; the region’s remaining nuclear and natural gas resources will be crucial to meeting the need for firm resources through the study horizon and beyond
5. Substantial reliability risks will accompany the transition of the region’s electricity resource portfolio; managing and responding to these risks will require continuous efforts to refresh resource adequacy planning as more information becomes available and utilities gain more experience operating new resource portfolios
This analysis finds that utilities’ IRPs in aggregate will position the region to meet regional resource adequacy needs. In the absence of any systemic deficiency that can be traced to current planning conventions, this study concludes that no immediate changes to utility planning practices are needed to maintain reliable electric service.
This finding notwithstanding, utilities should continue to advance their resource adequacy planning practices to take advantage of new information and modeling techniques. These improvements will enable utilities to mitigate the risks identified herein and improve their efforts to balance planning for reliability portfolio alongside affordability and sustainability objectives. Most importantly, utilities should implement the resource adequacy planning “best practices” as identified in this study to the extent practicable, including:
Assess the need for capacity using a probabilistic analysis framework that captures the range of potential energy demands under an increasingly volatile climate and should update this analysis periodically as new information becomes available or as load shapes change.
Apply an ELCC methodology to assess the capacity value of all resources in their portfolios on an equitable basis, capturing all of the risks and limitations to resource availability that are well understood and quantifiable.
Additionally, in recognition of the uncertainties and associated risks identified in this report, utilities should regularly update inputs and assumptions in their resource adequacy planning.
Ensure load forecast captures plausible weather conditions that reflect the best available climate science. The upward climate trend and associated changes to the distribution of extreme weather conditions will have major implications on the abilities of the utilities’ portfolios to supply their needs to an acceptable level of reliability.
Align planning assumptions used to characterize each resource with expectations for performance under extreme heat. The extreme heat conditions that drive resource adequacy challenges in the Southwest region may also impact the availability of generation, both through increased risk of plant outages and degradation of plant output. Utilities should ensure their planning reflects an understanding of these impacts for all types of resources; to the extent these effects are material, they could represent a correlated risk to resource adequacy.
Gather and incorporate real-world information on performance of emerging technologies. In the absence of historical data, performance assumptions for nascent technologies like battery storage are often idealized in resource adequacy modeling. Replacing idealized assumptions with real-world performance data will improve utilities’ abilities to value the capacity contribution of these resources accurately. A centralized database with records of battery storage outages such as NERC’s Generation Availability Data Set for other technologies would provide significant value to utilities’ planning efforts throughout the country.
Finally, in recognition of the increasing systemic threats posed by catastrophic extreme weather events and common mode failures – both of which are difficult to incorporate into a probabilistic analysis framework – utilities should supplement probabilistic resource adequacy studies with resilience planning studies that examine the potential consequences of extreme weather and/or system contingencies…