The National Renewable Energy Laboratory (NREL) has released a report titled "Impacts of Extensive Energy Storage Deployments on Grid Operations" from the Energy Storage Futures Study (SFS). The Energy Storage Future Study (SFS) is a multi-year research project at the lab that explores the role and impact of energy storage systems in the development and operation of the U.S. electric power industry. The Energy Storage Future Study (SFS) aims to study the potential impact of advances in energy storage technology on the deployment of utility-scale and distributed energy storage systems, as well as on future power system infrastructure investments and operations.
This report uses cost-driven scenarios from the National Renewable Energy Laboratory's (NREL) District Energy Deployment System (ReEDS) model as a starting point to examine the operational impact of grid-scale energy storage systems and their relationship to renewable energy generation. . To this end, the commercial production cost modeling software PLEXOS was used to evaluate the hourly operation of five scenarios with cumulative deployment of 210GW to 930GW of energy storage systems in the United States by 2050. The study found that between now and 2050, energy storage will play an important role in the power system – by storing electricity at the lowest marginal cost (usually excess generation from solar or wind facilities) and at the highest daily Power generation during net load. Deploying and operating energy storage systems helps to integrate variable renewable energy and provide continuous and reliable power by providing a vital resource.
The Energy Storage Futures Study (SFS) series provides data and analysis in support of the U.S. Department of Energy's "Energy Storage Grand Challenge," an effort to accelerate the development, commercialization, and utilization of next-generation energy storage technologies and maintain U.S. A comprehensive program for global leadership in energy storage. The Energy Storage Grand Challenge employs a use-case framework to ensure that energy storage technologies can cost-effectively meet specific needs, incorporating a broad range of technologies across multiple categories: electrochemical, electromechanical, thermal energy storage, power generation, construction, and power electronics.
The Energy Storage Futures Study (SFS) series of reports was written by the National Renewable Energy Laboratory of the U.S. Department of Energy (DOE). The research was sponsored by the U.S. Department of Energy's Office of Energy Efficiency and Renewable Energy and was conducted using computational resources from the National Renewable Energy Laboratory. The views expressed in this article do not necessarily reflect the views of the U.S. Department of Energy.
Summary
Due to the rapidly falling costs and the enormous potential value of energy storage systems, one could see energy storage systems of up to hundreds of gigawatts of installed capacity being deployed on the grid in the future. The Energy Storage Futures Study (SFS) report aims to explore the potential role and impact of energy storage systems in the growing power sector in the United States.
The assessment builds on a previously published Energy Storage Futures Study (SFS) report, in which the National Renewable Energy Laboratory (NREL) added new capabilities to its publicly available District Energy Deployment System (ReEDS) model. Research shows that by the end of 2050, the installed capacity of energy storage systems deployed in the United States is likely to exceed 125GW, which would be more than five times the cumulative installed capacity of energy storage systems currently deployed in the United States, even according to the most conservative estimates.
This analysis moves into scenarios with high rates of energy storage deployment in District Energy Deployment Systems (ReEDS) through detailed production cost modeling to observe hourly, daily, and yearly operations and the value of associated energy storage systems. Overall, the power system scenario for this scenario works successfully, indicating that there is no need to worry about load balancing of the grid by the end of 2050. The successful load balancing demonstrates that various improvements to district energy deployment systems (ReEDS) in previous work are effective in envisioning these future scenarios.
Studies have shown that energy storage systems are highly consistent with the availability of solar power generation facilities, which have predictable power generation cycles, which are well aligned with the need for charging and discharging of energy storage systems. On the other hand, the generation cycle of wind power is not stable, and there are often prolonged overproduction, which may last for hours or days, which is much longer than the duration of the energy storage systems studied here. While energy storage systems can play a key role in harnessing solar power and wind power, synergies with solar power facilities are more consistent. The energy storage system in the grid is deployed and operated in conjunction with renewable energy power generation facilities, which can effectively provide energy time shift and reduce peak load services. Despite the low annual capacity factor of the energy storage system, which is inherently limited by its charging demand, it has a very high utilization rate, which indicates the great contribution of the energy storage system to the resource adequacy of the power system.
Finally, the report also found that energy storage systems improved the efficiency of a variety of power assets. For example, in the future grid scenarios studied, energy storage systems reduce the carbon emissions of the power system by replacing coal-fired and natural gas-fired power generation facilities with excess power generation from renewable energy sources such as wind and solar power. In addition, energy storage systems can reduce pollutant emissions from fossil fuel power generation facilities, which may adversely affect the health of local residents, especially those living near fossil fuel power generation facilities. Energy storage systems will also improve and stabilize the operation of the grid, increasing the utilization of certain transmission lines while reducing congestion on power lines. How energy storage affects nearby transmission by increasing or decreasing usage depends on local conditions, and the study also found that energy storage generally increases the utilization of transmission assets. These findings suggest that further analysis should take into account the unique interactions of energy storage systems and transmission lines when deploying and operating energy assets simultaneously.
Overall, the analytical results of the Energy Storage Futures study show an increasing opportunity for long-duration energy storage systems to play an important role in future power systems. The analysis shows that deploying more long-duration energy storage systems can improve operational efficiency by reducing excess generation, reducing fossil fuel generation facility operations and emissions, and increasing transmission system utilization. Additionally, energy storage systems play an important role in providing capacity at peak net load times. Energy storage future research also investigates the role of long-duration energy storage systems, especially under highly decarbonized grid conditions, such as those close to 100% clean energy.
1. Introduction
The deployment of renewable energy sources such as solar power and wind power is increasing globally, driven by falling costs and the achievement of renewable energy targets. At the same time, falling battery costs have led to a growing interest among energy developers in deploying energy storage systems to provide grid services, including energy transfer and peak shaving. The National Renewable Energy Laboratory's (NREL) Future of Energy Storage study assesses energy storage deployment pathways through 2050 and finds that energy storage systems will be a significant contributor to large-capacity power systems, with cumulative U.S. energy storage deployments by 2050 The installed capacity of the system will reach 132GW. This energy storage deployment is primarily driven by a combination of energy value (energy transfer) and capacity value (providing electricity when the power system needs it).
When deployed on electric power systems, energy storage systems can provide many benefits, including energy transfer, reducing the operation of fossil fuel generation facilities, providing ancillary services, promoting system resource adequacy, and potentially delaying transmission line or other power system infrastructure upgrades . In providing energy transfer, energy storage systems, as net energy consumers, are able to store available energy at a lower price so that it can be released when electricity prices are higher. As the share of zero marginal cost renewables in the power system continues to grow, one major use of energy storage systems is to store electricity during periods of excess generation from renewable generation facilities, another benefit is to use stored electricity to avoid or reduce startups Expensive fossil fuel power generation facilities. Energy storage systems can also provide ancillary services or serve as backup power sources, providing a reliable source of power for the power system. Additionally, the energy storage system can meet peak demand by discharging it when the power system needs it.
Finally, energy storage can provide other benefits that are difficult to quantify, such as avoiding or delaying upgrades to transmission or distribution systems, especially in areas where traditional generation resources may be difficult to site.
Most power system models can simultaneously represent only a small fraction of these potential benefits, which are widely distributed in scale and time. When considering future investments, power system planning or capacity expansion models consider these different value streams as much as possible to offset up-front costs. However, planning models often fail to account for detailed operating parameters or fully capture all the services an energy storage system (or other asset) might provide. Running an analysis often requires using multiple models with different objective functions and times. Production cost modeling, for example, is one such tool that can provide detailed insights into grid operations, including future power systems that may look very different from today's power systems. For example, in some future research scenarios for energy storage systems, energy storage deployments reach hundreds of gigawatts, while the United States has cumulatively deployed 23 GW of energy storage systems by 2021, but the vast majority come from pumped hydro power generation facilities. So what will the operation of an energy storage-intensive power system look like, and how is it different from today? How will the different types of energy storage systems interact with each other? How do these operations vary by season, scenario, and storage configuration?
To answer these questions, this work evaluates the detailed operation of scenarios identified by the District Energy Deployment System (ReEDS) model to explicitly realize scenarios with higher energy storage deployment rates on an hourly basis using the commercial production cost modelling software PLEXOS. operate. The study uses a range of scenarios to assess the role that energy storage systems may play on a daily, seasonal and annual basis, and how that role changes with the configuration of energy storage systems and the mix of system resources. (Note: PLEXOS is an economical optimization software that uses optimization techniques based on mathematical principles for prediction. PLEXOS provides the latest visualization capabilities, data processing and distributed computing methods to provide powerful high-performance simulation systems for electricity, water and natural gas. )
2. Methods and data
The primary goal of future research on energy storage systems is to examine future electricity paths that could see large and sustained deployments of energy storage systems in the United States. Understanding how large-scale transmission and generation systems evolve with energy storage deployment is at the heart of this research. The purpose of this modeling is to identify and evaluate the costs and benefits of various energy storage deployment paths. Considering the complexity of the power system, this cannot be achieved with a single-model large system model. To this end, the study took two main system models, "identification of the least-cost investment path" and "detailed simulation of projected future system operation," and combined them.
In the first step of the analysis, the District Energy Deployment System (ReEDS) was used to determine the minimum cost investment set for transmission and generation assets under various evolutions in technology cost and performance. The District Energy Deployment System (ReEDS) is used for long-term power system planning efforts because it integrates many different constraints and drivers of power sector change and investment, including technology and fuel prices, policies and regulations, technology performance constraints, fuel availability Limits, as well as changes in load shape and aggregate demand, to determine investment paths. The second step is to use PLEXOS to simulate the projected hourly timescale of the energy storage system between now and 2050 (under a given scenario). The results from PLEXOS can be used to assess future projected supply and demand of energy storage systems without any major challenges.
2.1 Five key scenarios in the analysis
The analysis is based on scenarios generated by the National Renewable Energy Laboratory's (NREL) District Energy Deployment System (ReEDS) Capacity Expansion Model, which represents the U.S. power system in 134 regions connected by aggregated transmission corridors for power system decommissioning and Generation, transmission and generation investment for the lowest cost system optimization. The model optimizes the investment in the power system, including the operation of the power system at each time scale. Other reports in this study contain a full discussion of the District Energy Deployment System (ReEDS) model, its inputs, and various improvements related to this work.
The research report's analysis focuses on the following five key scenarios implemented in District Energy Deployment Systems (ReEDS):
Reference Scenario (Ref): This scenario follows all cost and technology evolution to 2050 as a reference assumption.
• Low-Cost Batt: This scenario sets the lowest cost of the battery.
• Low-Cost PV Scenario (Low-Cost PV): This scenario sets the lowest cost of solar power facilities.
• High natural gas cost/low cost battery scenario: This scenario assumes high cost natural gas power generation and lowest cost battery.
• Zero-Carbon Scenario: This scenario reflects a zero-carbon energy scenario in future research that achieves more energy storage deployment (in MW) than the four scenarios above.
Figure 1 shows the installed capacity and annual power generation by type of power generation facility in five scenarios from 2020 to 2050. All scenarios show an increase in installed capacity of solar, wind, and energy storage systems, largely due to falling technology costs for these three resources. These sources of energy have replaced nuclear, coal and, in some cases, natural gas generation. Natural gas power generation is most evident in the high natural gas cost/low cost battery scenario, which uses a high natural gas cost and a zero-carbon scenario that calls for replacing all coal-fired and natural gas-fired power generation facilities by 2050. The other three scenarios (reference scenario, low-cost battery, and low-cost solar) follow the middle of the natural gas cost trajectory, resulting in significant gas-fired power generation still in place through 2050.
Installed installed capacity of energy storage systems by technology type under the same scenario and year. It should be noted that the deployment of energy storage systems with a duration of more than 12 hours was not considered in the energy storage future study. In 2020, almost all long-duration energy storage systems will be existing pumped hydro power generation facilities. 2030 will see a dramatic increase in the deployment of shorter-duration battery storage technologies (2- and 4-hour durations), as these technologies are considered to have the lowest cost given their smaller installed capacity. 4-hour battery storage systems continue to dominate through 2040, but are starting to see deployments of 6-hour battery storage systems. By 2050, battery storage systems with a duration of 4 hours remain the predominant energy storage technology in all scenarios, but some scenarios (especially zero-carbon scenarios) show the deployment of longer duration batteries, such as 8 hours duration and 10-hour battery energy storage system. Due to falling costs, longer-duration battery storage systems will become more cost-competitive in the coming years. Longer-duration battery storage systems are also more valuable in the case of a zero-carbon scenario, as they enable energy transfer over longer periods of time and provide higher capacity credits. Finally, even the least deployed energy storage system in the modeled time frame (the reference scenario) will see a roughly 10-fold increase in the installed capacity of energy storage systems deployed by 2050.
2.2 Concept of PLEXOS
Production cost modeling starts with the District Energy Deployment System (ReEDS), which includes transmission systems, types of renewable and conventional generation facilities, installed capacity and location, as well as hourly load and variable generation data. This data is passed to the production cost model along with the hourly operation of the energy storage system. The study used PLEXOS, a commercial production cost model, to simulate the hourly operation of a future electricity system as determined by the District Energy Deployment System (ReEDS). The goal of a production cost model is to optimize the dispatch of generation resources to meet load more cost-effectively, subject to all constraints such as renewable resources, transmission availability, and operational practices.
The District Energy Deployment System (ReEDS) simplifies scheduling in its algorithms to inform investment decisions for decades to come. This allows the model to take into account key operational constraints and challenges, but still be simple enough to keep computation tractable. However, because the District Energy Deployment System (ReEDS) does not explicitly model the 8,760-hour dispatch on an annual time horizon, it cannot address many of the operational constraints that are critical to ensuring energy balance and optimal operation of energy storage. Therefore, PLEXOS simulates hourly operations of grid expansions identified by the District Energy Deployment System (ReEDS) to measure operating costs and verify supply and demand balance. Evaluate basic transmission adequacy using a simplified regional approximation of the transmission network, as well as evaluate the dispatch of energy storage systems. An energy storage system is defined in PLEXOS as installed capacity (MW) and duration (h) and location determined by the District Energy Deployment System (ReEDS). Consistent with all service assumptions of the District Energy Deployment System (ReEDS), the energy storage system is co-optimized to provide energy and ancillary services.
The analysis covers U.S. states and includes three separate simultaneous interconnect regions: Eastern Interconnect, Western Interconnect, and Texas Interconnect. Because this geographic footprint is broad, capacity expansion modeling and production cost modeling simplify the representation of the transmission system. In this modeling, the U.S. power system consists of 134 regions connected by transmission lines. Therefore, both the District Energy Deployment System (ReEDS) and PLEXOS will restrict the flow between, but not within those areas.
The goal of the production cost modeling step is to test the operational impact of capacity expansion scenarios, provide a detailed picture of hourly dispatch across scenarios, and gain insight into the optimization of energy storage capacity dispatch.
For the purposes of this research report, the main outputs of the production cost model include:
• Identify any unserved loads or unserved reserves, which may indicate resource adequacy issues.
• The composition of electricity generation on different time scales, from hourly to yearly.
• Total variable operating costs of electricity generation (including fuel costs, start-up and shutdown costs, and variable O&M costs)
• Scheduling and utilization of energy-constrained resources such as pumped hydro and battery storage systems.
• Resource portfolio provides reserves annually and at any given time.
• Use of simplified representations of transmission systems and congestion.
• Scheduling during specific periods (eg, high renewable energy/low load periods and low renewable energy/high load periods).
3. Operational results of future scenarios with high energy storage deployment rates
This section will discuss the energy storage systems deployed in the five scenarios of the Energy Storage Future Study, understanding the important role that energy storage systems play on a daily, seasonal, and annual basis. The changing role of energy storage systems over time is also investigated, starting with the representation of the system in 2020 and modelling in 2050.
3.1 The role of energy storage systems in annual operations
Energy storage is fundamentally different from most generation facilities on bulk power systems because it is not technically a generation resource. In order to provide power to the power system, the energy storage system must first store energy, and there is an efficiency loss due to power conversion. Due to the need for recharging and the associated efficiency loss, the capacity factor map uses points to represent the average annual capacity factor stored in each scenario. The annual capacity factor remains below 25% (and in some cases below 10%) in all years and scenarios. To understand the capacity factor trends in Figure 3, first consider the reasons for deploying energy storage systems in a capacity expansion model, including:
• Energy transfer resources: Diurnal energy storage can enable energy arbitrage, especially as more and more low-cost resources such as solar power generation facilities become available for charging. As battery costs fall, more energy conversion opportunities become economical.
• Capacity resource: Energy storage is a reliable resource that helps meet peak net demand periods.
In some cases (reference scenarios and low-cost solar scenarios), the capacity factor increases slightly over time due to: (1) a relative increase in the deployment of longer duration battery energy storage systems, providing higher Average capacity factor; (2) More deployment of zero marginal cost resources (solar and wind) for arbitrage. In other cases (low-cost batteries and high gas cost/low-cost batteries), the average coefficient of energy storage capacity declines after 2030, despite increased deployment of longer-life battery storage systems in some cases; This occurs in scenarios with low-cost battery assumptions, as it may be economical to employ low-cost batteries even if there are fewer arbitrage opportunities. In these four scenarios, the trend in capacity factor is mainly driven by the first role (energy transfer) played by the energy storage system. However, the fifth scenario (the zero-carbon scenario) shows the lowest annual capacity factor to 2050. In this case, the time to run the energy storage system in the high gas cost/low cost battery scenario is only about 55 days. Given the significant reduction in all-weather in a zero-carbon scenario, the energy storage system operates for less time, thereby reducing the average capacity factor. In this case, the second role played by the energy storage system (as a capacity resource) is similar to that of a seldom-operated natural gas peaker.
The energy storage system still plays the first role (energy transfer). Arbitrage opportunities in zero carbon, while occurring less frequently, are lucrative because energy storage systems are generating electricity at zero marginal cost, replacing electricity from expensive fossil fuel generation facilities with renewable electricity (about $200 at marginal cost) /MWh).
Although Figure 3 shows a lower annual capacity factor, the triangles in the graph (showing the average energy storage capacity factor for only the first 10 net load hours in each region) tell a different story. The calculations revealed high capacity factors during these times, often well above 75%. These hours represent periods when the power system is under supply stress, and energy storage systems contribute significantly during these periods. In most cases, the energy storage capacity factor in the first 10 net load hours is the lowest in 2050 due to higher storage deployment capacity. As the installed capacity of energy storage systems on the power system increases, newly deployed energy storage systems have diminishing returns, indicating a trend toward saturation. Similar to the annual capacity factor, the average factor for the first 10 payload hours is the lowest in the 2050 zero-carbon scenario for the same reasons that the annual capacity factor is the lowest in this scenario.
Finally, the report examines the relationship between the percentage of energy storage capacity emitted daily and the variable generation contribution in a high gas cost/low cost battery scenario in 2050 (shown in Figure 5). Here, the collection of Regional Energy Deployment System (ReEDS) regions is plotted as a point; doing so divides the footprint into 18 U.S. regions, somewhat approximating the RTO/ISO footprint or market area. Each zone is a point where the percentage of daily discharge capacity is compared to the percentage of total solar power generation (left graph) or wind power (right graph). Although the graph shows a large variation, there is a clear relationship between an increase in solar power generation and an increase in the percentage of daily discharge capacity (or daily energy cycle), and between an increase in wind power generation and a decrease in the percentage of daily discharge capacity obvious relationship. The positive correlation of solar power generation shows that as solar power generation increases, there is an increase in demand for energy storage systems that can be recharged with excess electricity from solar power facilities during the day and discharged during the morning and evening peak periods. On the other hand, an inverse relationship for wind power generation is also seen. Previous analyses have shown that excess power generation from wind facilities typically lasts for hours or even days, far exceeding the duration of the energy storage systems deployed in this study. During these wind excess periods, the energy storage system is fully charged and does not need to be discharged until it stops during the excess generation period, reducing the average daily capacity.
Describe how energy storage systems can effectively provide peak reduction services (and energy transfer) across all configurations and grid combinations. Although an energy storage system has a low annual capacity factor and is inherently limited by its charging needs, it has a very high utilization rate (in many cases more than 75%).
