TODAY’S STUDY: An Energy Storage Case Study In Nevada
The Economic Potential for Energy Storage in Nevada
Ryan Hledik Judy Chang Roger Lueken Johannes Pfeifenberger John Imon Pedtke Jeremy Vollen, October 1, 2018 (Brattle Group)
This study identifies the amount of energy storage that can be incorporated cost-effectively into Nevada’s future electricity resource mix.
In 2020, up to 175 MW of utility-scale battery storage (with 4-hour storage capacity) could be deployed cost-effectively statewide.
By 2030, the economic potential for utility-scale storage increases to a range from 700 MW to more than 1,000 MW, depending most significantly on the extent to which storage costs decline over time.
Behind-the-meter (BTM) storage could add up to 30 MW of storage capacity by 2030 under favorable conditions, and this could further increase by up to 40 MW through the provision of cost-effective utility-administered BTM storage incentives
Nevada Senate Bill (SB) 204 (2017) requires the Public Utilities Commission of Nevada (PUCN) to “determine whether it is in the public interest to establish by regulation biennial targets for the procurement of energy storage systems by an electric utility.”1 The Nevada Governor’s Office of Energy (GOE) commissioned this study to provide information for the PUCN when evaluating at what levels energy storage deployment would be economically beneficial for the state of Nevada, and whether procurement targets for energy storage systems should be set and, if so, at what levels. To assess the value of energy storage in Nevada, our study considers the range of benefits summarized in Table 1.
In this study, we account for a number of critical considerations when assessing the value of energy storage:
• Various value streams. Capturing one value stream for storage can mean foregoing opportunities to fully capture some of the other value streams. Co-optimizing the operation of energy storage relative to the available multiple value streams is therefore important to accurately estimate total storage benefits. We have utilized Brattle’s bSTORE modeling suite to account for these tradeoffs. The resulting “stacked” values estimated in this report are additive because we have considered areas where overlapping usages may not occur consistently.
• Uncertainty in costs and benefits. Energy storage technology is rapidly developing, and the value streams that it can capture are similarly in a state of evolution. It is important to account for uncertainty in the costs and benefits of storage when establishing future storage procurement targets. We use a range of costs to consider the possibility of relatively rapid versus slow cost reduction for storage. We use a scenario-based approach to consider a range of future developments influencing the benefits storage can provide.
• The relationship between storage quantity and benefits. The incremental cost-effectiveness of energy storage decreases as its market penetration grows. This is because the opportunities to provide services such as frequency regulation and local distribution capacity deferral saturate as more storage is added to the power system. In prior energy storage research, we have found that capturing the decreasing marginal value of adding storage is a critical consideration when quantifying overall value and cost-effective storage potential.2 Our approach accounts for this relationship.
• Degree of foresight in battery utilization. Modeling approaches often rely on optimal operation of the storage technology, assuming perfect foresight of system conditions. Our approach accounts for real-world limitations on foresight of future system conditions, and considers how imperfect foresight affects storage operations.
Our methodology is applicable to a broad range of energy storage technologies including, for example, various battery technologies, flywheels, compressed air storage, hydroelectric pumped storage, or thermal storage. To focus the analysis on a representative range of storage costs and performance characteristics, we simulate storage deployment of lithium-ion batteries, which are the predominant energy storage technology currently being deployed and contracted. More specifically we analyze lithium-ion batteries with 4-hour storage capacity.
Consistent with the applicable current law and NV Energy’s 2018 Integrated Resource Plan (IRP), our study assumes NV Energy remains the utility responsible for serving most retail customers in Nevada. We assume that: (1) generating resources currently dedicated to serving Nevada loads at their cost of service would continue to be used to serve loads even if they will be subject to competitive pressures in the future, (2) new generation additions and retirements are consistent with NV Energy’s IRP, and (3) the transmission available without wheel-out charges between balancing areas remains limited to that available in today’s Energy Imbalance Market (EIM) footprint.
If Nevada retail customers were able to choose their power suppliers in the future, the total amount of generating resources needed to serve Nevada’s electricity demand would not change. Thus, we do not need to assume that all of the current retail customers must be served by NV Energy or that all generating and new storage resources must be owned by NV Energy. Rather, we focus primarily on how Nevada, as a state, will supply its electricity customers and how the state as a whole may use energy storage as a resource to help meet state-wide system needs and policy objectives. When analyzing the benefits of storage, we evaluate the cost of producing electricity to serve Nevada lectricity users, regardless of who are the retail suppliers. Any changes to the cost of producing electricity account for the costs of operating power plants located inside Nevada (regardless of ownership) and the net costs of purchased power from other entities to serve electricity users in Nevada.
Energy storage can be incorporated cost-effectively into Nevada’s future power supply mix. Under the assumptions used in this study, a statewide deployment of up to 175 MW of utility-scale storage could be cost-effective in 2020 if storage costs are at the lower end of the expected cost range. By 2030, declining battery costs and evolving system conditions increase this estimate of cost-effective potential to at least 700 MW and possibly exceeding 1,000 MW at the high end. The development of these estimates accounts for constraints that limit the operation of the storage devices relative to that of a peaking unit, in particular limits on battery storage discharge duration.
Within these ranges, the optimal storage procurement target will depend on the state’s evolving actual need for new generating capacity. Thus, the incorporation of similar storage scenarios into NV Energy’s resource planning process would be valuable to further confirm these conclusions.
The findings of our analysis are summarized in the figures below. Figure 1 illustrates the total state-wide ratepayer benefits and costs at various levels of storage deployment as well as the composition of the major storage-related value streams that affect utility ratepayers:3 (1) avoided generating capacity investments; (2) production cost savings (related to supplying energy and ancillary services as well as avoided curtailments of renewable generation); (3) the benefit of deferred T&D investments; and (4) avoided distribution-system customer outages. Not included in Figure 1 are (5) societal emissions-related impacts (since they do not affect utility rates currently), which result in societal-emissions-cost decreases of $0.7 to $10.6 million in 2020 and decreases of $1.6 to $27.0 million in 2030; and (6) other benefits, such as voltage support and T&D energy losses, which are too small to affect the conclusions about cost-effective levels of storage deployment in the state.
As shown, total 2020 benefits exceed total costs only at the low end of deployments analyzed, and only if the low end range of installed storage costs can be realized.4 In 2030, total benefits exceed total costs across the full range of cost projections and deployment scenarios, although the net benefit of incremental additions in 2030 drops to zero at 700 MW for the high battery cost scenario, as shown below.
Figure 2 below shows the incremental net benefits of storage at various deployment levels. This perspective is useful for identifying the point at which the benefits of incremental storage additions equal the costs of those additions. Storage additions beyond that deployment level are uneconomic, as incremental costs will exceed incremental benefits. As shown, up to 175 MW of storage deployment are cost effective in 2020 at the low end of the storage cost range. By 2030, the cost effective deployment level exceeds 1,000 MW at the low end of projected cost, with 700 MW being cost effective at the high end of projected costs.
The estimates of cost-effective storage potential are based on quantitative analyses that capture the primary drivers of storage value at the grid level: avoided generation capacity costs, reduced energy costs, reduced ancillary services costs, avoided T&D capacity costs, and reliability improvements (i.e., customer outage avoidance). We do not include the value of estimated avoided emissions when evaluating the amount of storage that would be cost-effective from a system perspective.
Implementation of and Nevada’s participation in a regional power market may reduce the value of storage due to lower production cost savings associated with increased resource diversification that would be achieved through having a market that spans a larger region. The resource adequacy needs associated with serving Nevada loads may not be reduced and other value streams are unlikely to be affected. If the implementation of a regional market were to reduce the production cost savings by half and not affect other value streams, the cost-effective level of storage deployment in 2030 would fall from a range of 700 MW to greater than 1,000 MW (without a regional market) to a range of 400 MW to greater than 1,000 MW (with a regional market).
In addition to the utility-scale and distribution-system-level applications discussed above, storage could add value as a customer-side, behind-the-meter (BTM) application. The avoidance of demand charges and peak-energy charges in the retail electricity bill of large (commercial and industrial) customers likely will be the primary driver of BTM storage adoption within the study’s 2030 time horizon. Some “baseline” level of BTM storage adoption would happen irrespective of any utility storage procurement initiatives based on specific targets. To remain consistent with the scope of this study, we quantified the cost-effective incremental increase from this baseline BTM storage adoption level that could result from a utility-administered BTM storage incentive program offered to retail customers. In return for an incentive payment, customers would allow the utility to control their storage device for a limited number of hours of the year to address resource adequacy (i.e., generation capacity) requirements.
We considered a range of assumptions that would influence BTM storage adoption, such as battery cost, adoption rate, magnitude of utility incentive payments, and the composition of the commercial and industrial (C&I) customers in the state. At estimated 2020 BTM storage costs, BTM storage adoption in the absence of a utility incentive program could be up to 7 MW. The introduction of cost-effective utility incentive programs could incrementally increase these estimates by up to 24 MW. At 2030 BTM storage costs, baseline BTM storage adoption is estimated to be up to 31 MW without the incentive program, which would incrementally increase between 6 MW and 39 MW with availability of cost-effective utility incentive programs. Results of these BTM storage potential cases are summarized in Figure 3. These values are incremental to the adoption potential estimates for utility-scale storage (including front-of meter distribution-level storage) described above.
In addition to the assumed utility incentive payments for resource adequacy, it is possible that BTM storage could provide additional sources of value, such as ancillary services or avoided T&D costs. Third party aggregators, utilities, or customers could monetize greater value under these conditions, thereby leading to increased BTM storage investments.
As these results show, energy storage can be a cost-effective addition to Nevada’s future mix of electricity resources, reducing system costs and benefitting consumers as a result. It can provide value across a range of applications and use cases, whether for resource adequacy, renewables integration, T&D investment deferral, or some combination of these and other benefits streams. This conclusion is robust across a range of modeled scenarios. The economically optimal levels of future deployment depend most significantly on the trajectory at which energy storage costs decline and new generating resources are needed to meet Nevada’s electricity demand.