skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Value and Role of Pumped Storage Hydro under High Variable Renewables

Abstract

A project team, led by GE Research, was tasked by the U.S Department of Energy (DOE) to study the value and role of pumped storage hydro (PSH) under high variable renewables. The study was majority funded by DOE’s Office of Energy Efficiency and Renewable Energy (EERE) through a program managed by EERE’s Water Power Technologies Office. The project team for the study consists of three organizations – GE Research, GE Energy Consulting, and GE Renewables. PSH is an ideal complement to clean energy, as it can accommodate the intermittency and seasonality of variable energy resources such as solar and wind power. New PSH plants in areas with recently increased wind and solar capacity are expected to improve grid reliability while reducing the need for new fossil-fueled generation. This project aims to overcome a range of market barriers for PSH by helping stakeholders understand the benefits of PSH that are not well understood or quantified, by demonstrating the capability of new variable speed PSH (VSPSH) technologies, and by helping developers improve PSH revenues with the development of a new PSH scheduling tool. The study is particularly intended for utilities, Public Utility Commissions (PUCs), developers and regional planner organizations, as it exploresmore » the values and impacts of PSH, specifically in high-renewable penetration systems: Develop a PSH scheduling tool to co-optimize energy and ancillary services, taking into account price elasticity in the power market. Analyze and quantify the potential value of PSH under different system conditions. Develop a set of VSPSH models for transmission planners to study the impact of PSH on the grid. Investigate the dynamic capability of VSPSH and assess its impact on grid frequency response and transient stability. Investigate the PSH contribution to resource adequacy. STUDY FINDINGS Built a detailed model of the WECC area of the USA, projected to 2028 with wind & solar providing 50% of annual energy. The intention of modeling a year so near in the future was to examine such an aggressive case and especially test sensitivities by changing different parameters. Reserve Adequacy: The contribution of different storage durations was examined through Capacity value; namely the ability to serve additional load with the same level of reliability. The capacity value of storage was studied for two sites in Arizona and California, for many sensitivities and weather years. In all cases, 4+ hours of storage typically lead to capacity values of 100% of nameplate capacity, whereas 2 hours of storage are mainly valuable for cases when solar power penetration is high. Solar penetration and storage penetration have the largest effect on storage capacity value. Prior to performing the work, the expectation was that 100% of capacity value would only come with longer durations: the explanation is that the solar-heavy WECC system already has relatively limited time-frames where loss of load expectation is high. Grid Resiliency: the impact of variable speed technology (power regulation when pumping) was assessed for a major power loss event (-2760MW). A simulated 2GW VSPSH plant in Arizona in pumping mode was, alone, able to improve the frequency nadir to 59.55Hz from 59.5Hz and allowed Arizona to meet its frequency response obligation requirements. Such services are nevertheless not monetized today. A novel PSH Scheduling tool was developed, incorporating for the first time the impact of variable height differences between reservoirs (‘head’) and variable speed machine behavior. This tool was combined with iterative production cost modeling to account for price elasticity, i.e. the fact that PSH units have a significant size compared to the fleet and that bidding in or out such units can modify the market price. An example would be that when electricity market prices are low, a PSH unit committed to absorb power (pump) could increase the market price. The variable speed version of this tool was incorporated into the analysis of the WECC 2028 model, to examine the introduction of two different pumped storage facilities (one a 2000 MW, 20000 MWh facility in Arizona known as ‘Big Chino’ and the other a 500MW, 4000MWh facility in California known as ‘San Vicente’). In the base case of 50% wind & solar and ‘low’ storage levels (4.6GW / 8100GWh pumped storage hydro, 1.8GW / 3000GWh battery storage), just these two plants, for the year of 2028 alone, were seen to have a positive impact. Total WECC production costs were reduced by 182 M $ and 62 M $ respectively. CO2 emissions were reduced by 1.8 and 0.5 million tons respectively. The curtailment of other renewables and the number of starts requested of the thermal fleet were also significantly reduced. The PSH plants were configured to maximize annual operating profit (at 194 M $ & 69 M $ respectively). Configuring them to reduce WECC production costs is expected to give lower production costs. The biggest single drivers of PSH profit are the level of penetration of renewables and the volume of storage, profit is essentially coming from the Energy market. There is no ancillary service market in Arizona. Other sensitivities (gas prices, high/low year for hydro) are within +/-10%, except for exceptionally low hydro years. We can note a wide dispersion according to the scenario/plant applied: from 49 M $ to 232 M $. 2028 Electricity prices: 50% Wind & Solar gives prices <10 $/MWh between 8am & 2pm in both AZ & CA (often <5 $ /MWh), with evening peaks typically 7 to 9 times higher. 30% Wind & Solar: reduced impact overall compared to the 50% version, though solar curtailment is still high at 10% 30% Wind & Solar: prices still <10/MWh between 9am & 2pm in CA, <30 $/MWh for AZ between 8am & 2pm: relatively more Renewables being removed from AZ in this variant. High storage option explored: +14GW, of which 70% PSH & 30% Batteries, with PSH assumed to have 10 hours storage and Batteries 4 hours i.e. 9.8GW / 98GWh of PSH and 4.2GW / 16.8 GWh of Batteries. As a reference point, the current entire US storage fleet has a power capacity of 23GW today. The high storage option was judged to reduce the base-line value of storage by an average of 25% on a $ /kW basis across the WECC. The decrease in value varies widely by region. With more storage, WECC production costs decreased by 167 M $ & 48 M $. Evening peak prices decrease while mid-day prices do not change significantly except for CA, where prices increase close to 10 $ /MWh. Of particular interest is that simulations showed that PSH is not competing versus batteries but mainly against itself; equally, the biggest impact on battery revenue is other batteries and not pumped storage. Thermal generation: Building the 2028 theoretical mix involves both retirements and capacity additions (6GW of gas was notably added to ensure reserve margins). Even in the high storage scenario gas was not fully displaced. Variable speed pumped storage units do not currently exist in the U.S., though they are operated elsewhere in the world. To better simulate the operation and benefits of VSPSH in Grid dynamic simulations, two new VSPSH models have been created and incorporated into the PSLF library.« less

Authors:
 [1];  [1];  [2];  [2];  [2];  [3];  [2];  [2];  [4]
  1. GE Research
  2. GE Gas Power
  3. GE Gas Power
  4. GE Renewable
Publication Date:
Research Org.:
GE Research
Sponsoring Org.:
USDOE Office of Energy Efficiency and Renewable Energy (EERE), Renewable Power Office. Water Power Technologies Office
OSTI Identifier:
1824300
Report Number(s):
DOE-GE-0008782
DOE Contract Number:  
EE0008782
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
13 HYDRO ENERGY; pumped storage hydro, Renewables, grid reliability, resource adequacy

Citation Formats

Shao, Miaolei, Guo, Xian, Bisceglia, Christina, de Mijolla, Genevieve, Rao, Shruti, Pajic, Slobodan, Ibanez, Eduardo, Bringolf, Mitchel, and Havard, David. Value and Role of Pumped Storage Hydro under High Variable Renewables. United States: N. p., 2021. Web. doi:10.2172/1824300.
Shao, Miaolei, Guo, Xian, Bisceglia, Christina, de Mijolla, Genevieve, Rao, Shruti, Pajic, Slobodan, Ibanez, Eduardo, Bringolf, Mitchel, & Havard, David. Value and Role of Pumped Storage Hydro under High Variable Renewables. United States. https://doi.org/10.2172/1824300
Shao, Miaolei, Guo, Xian, Bisceglia, Christina, de Mijolla, Genevieve, Rao, Shruti, Pajic, Slobodan, Ibanez, Eduardo, Bringolf, Mitchel, and Havard, David. 2021. "Value and Role of Pumped Storage Hydro under High Variable Renewables". United States. https://doi.org/10.2172/1824300. https://www.osti.gov/servlets/purl/1824300.
@article{osti_1824300,
title = {Value and Role of Pumped Storage Hydro under High Variable Renewables},
author = {Shao, Miaolei and Guo, Xian and Bisceglia, Christina and de Mijolla, Genevieve and Rao, Shruti and Pajic, Slobodan and Ibanez, Eduardo and Bringolf, Mitchel and Havard, David},
abstractNote = {A project team, led by GE Research, was tasked by the U.S Department of Energy (DOE) to study the value and role of pumped storage hydro (PSH) under high variable renewables. The study was majority funded by DOE’s Office of Energy Efficiency and Renewable Energy (EERE) through a program managed by EERE’s Water Power Technologies Office. The project team for the study consists of three organizations – GE Research, GE Energy Consulting, and GE Renewables. PSH is an ideal complement to clean energy, as it can accommodate the intermittency and seasonality of variable energy resources such as solar and wind power. New PSH plants in areas with recently increased wind and solar capacity are expected to improve grid reliability while reducing the need for new fossil-fueled generation. This project aims to overcome a range of market barriers for PSH by helping stakeholders understand the benefits of PSH that are not well understood or quantified, by demonstrating the capability of new variable speed PSH (VSPSH) technologies, and by helping developers improve PSH revenues with the development of a new PSH scheduling tool. The study is particularly intended for utilities, Public Utility Commissions (PUCs), developers and regional planner organizations, as it explores the values and impacts of PSH, specifically in high-renewable penetration systems: Develop a PSH scheduling tool to co-optimize energy and ancillary services, taking into account price elasticity in the power market. Analyze and quantify the potential value of PSH under different system conditions. Develop a set of VSPSH models for transmission planners to study the impact of PSH on the grid. Investigate the dynamic capability of VSPSH and assess its impact on grid frequency response and transient stability. Investigate the PSH contribution to resource adequacy. STUDY FINDINGS Built a detailed model of the WECC area of the USA, projected to 2028 with wind & solar providing 50% of annual energy. The intention of modeling a year so near in the future was to examine such an aggressive case and especially test sensitivities by changing different parameters. Reserve Adequacy: The contribution of different storage durations was examined through Capacity value; namely the ability to serve additional load with the same level of reliability. The capacity value of storage was studied for two sites in Arizona and California, for many sensitivities and weather years. In all cases, 4+ hours of storage typically lead to capacity values of 100% of nameplate capacity, whereas 2 hours of storage are mainly valuable for cases when solar power penetration is high. Solar penetration and storage penetration have the largest effect on storage capacity value. Prior to performing the work, the expectation was that 100% of capacity value would only come with longer durations: the explanation is that the solar-heavy WECC system already has relatively limited time-frames where loss of load expectation is high. Grid Resiliency: the impact of variable speed technology (power regulation when pumping) was assessed for a major power loss event (-2760MW). A simulated 2GW VSPSH plant in Arizona in pumping mode was, alone, able to improve the frequency nadir to 59.55Hz from 59.5Hz and allowed Arizona to meet its frequency response obligation requirements. Such services are nevertheless not monetized today. A novel PSH Scheduling tool was developed, incorporating for the first time the impact of variable height differences between reservoirs (‘head’) and variable speed machine behavior. This tool was combined with iterative production cost modeling to account for price elasticity, i.e. the fact that PSH units have a significant size compared to the fleet and that bidding in or out such units can modify the market price. An example would be that when electricity market prices are low, a PSH unit committed to absorb power (pump) could increase the market price. The variable speed version of this tool was incorporated into the analysis of the WECC 2028 model, to examine the introduction of two different pumped storage facilities (one a 2000 MW, 20000 MWh facility in Arizona known as ‘Big Chino’ and the other a 500MW, 4000MWh facility in California known as ‘San Vicente’). In the base case of 50% wind & solar and ‘low’ storage levels (4.6GW / 8100GWh pumped storage hydro, 1.8GW / 3000GWh battery storage), just these two plants, for the year of 2028 alone, were seen to have a positive impact. Total WECC production costs were reduced by 182 M $ and 62 M $ respectively. CO2 emissions were reduced by 1.8 and 0.5 million tons respectively. The curtailment of other renewables and the number of starts requested of the thermal fleet were also significantly reduced. The PSH plants were configured to maximize annual operating profit (at 194 M $ & 69 M $ respectively). Configuring them to reduce WECC production costs is expected to give lower production costs. The biggest single drivers of PSH profit are the level of penetration of renewables and the volume of storage, profit is essentially coming from the Energy market. There is no ancillary service market in Arizona. Other sensitivities (gas prices, high/low year for hydro) are within +/-10%, except for exceptionally low hydro years. We can note a wide dispersion according to the scenario/plant applied: from 49 M $ to 232 M $. 2028 Electricity prices: 50% Wind & Solar gives prices <10 $/MWh between 8am & 2pm in both AZ & CA (often <5 $ /MWh), with evening peaks typically 7 to 9 times higher. 30% Wind & Solar: reduced impact overall compared to the 50% version, though solar curtailment is still high at 10% 30% Wind & Solar: prices still <10/MWh between 9am & 2pm in CA, <30 $/MWh for AZ between 8am & 2pm: relatively more Renewables being removed from AZ in this variant. High storage option explored: +14GW, of which 70% PSH & 30% Batteries, with PSH assumed to have 10 hours storage and Batteries 4 hours i.e. 9.8GW / 98GWh of PSH and 4.2GW / 16.8 GWh of Batteries. As a reference point, the current entire US storage fleet has a power capacity of 23GW today. The high storage option was judged to reduce the base-line value of storage by an average of 25% on a $ /kW basis across the WECC. The decrease in value varies widely by region. With more storage, WECC production costs decreased by 167 M $ & 48 M $. Evening peak prices decrease while mid-day prices do not change significantly except for CA, where prices increase close to 10 $ /MWh. Of particular interest is that simulations showed that PSH is not competing versus batteries but mainly against itself; equally, the biggest impact on battery revenue is other batteries and not pumped storage. Thermal generation: Building the 2028 theoretical mix involves both retirements and capacity additions (6GW of gas was notably added to ensure reserve margins). Even in the high storage scenario gas was not fully displaced. Variable speed pumped storage units do not currently exist in the U.S., though they are operated elsewhere in the world. To better simulate the operation and benefits of VSPSH in Grid dynamic simulations, two new VSPSH models have been created and incorporated into the PSLF library.},
doi = {10.2172/1824300},
url = {https://www.osti.gov/biblio/1824300}, journal = {},
number = ,
volume = ,
place = {United States},
year = {2021},
month = {3}
}