Energy storage's role in decarbonization will depend on duration, cost cuts
Further cost reductions will be necessary to justify widespread storage deployment for decarbonization purposes, Argonne and MIT researchers found.
Energy storage has been hailed as the missing link and even an essential ingredient to higher levels of wind and solar power, but a new paper from Argonne National Laboratory and the Massachusetts Institute of Technology questions that premise.
The authors of the report, “The value of energy storage in decarbonizing the electricity sector,” conclude that the value of shorter-duration storage technologies, up to about two hours, is only justified by generation cost savings under the most stringent carbon emissions limits, and even then, only at low storage penetration levels. Hence, continued innovation and cost declines for lithium-ion batteries and other electrochemical energy storage technologies will be necessary to economically justify large-scale deployment in future low-carbon power systems.
In short, “costs have to improve before we see ubiquitous adoption of storage,” says Jesse Jenkins, a PhD candidate at the Institute for Data, Systems and Society at the MIT Energy Initiative, and one of the authors of the report along with Audun Botterud and Fernando J. de Sisternes at Argonne.
The report aims to fill a gap in previous studies that examine the role of storage in the electric grid from narrower perspectives. The authors of the Argonne-MIT study say their work adds to the existing literature “by providing a more complete assessment of the economic value of energy storage through jointly capturing the short- and long-run interaction between storage, renewable energy, and other zero-carbon electricity sources and their relative contributions to meet demands for energy and operating reserves along with emission reduction objectives.”
Estimating storage's value to decarbonization
The authors came up with a methodology they could use to run “experiments” for different levels of carbon restrictions.
They modeled a power system with the demand and wind and solar resource data of the Electric Reliability Council of Texas, but with a “Greenfield” configuration. That means they selected the entire generation mix from scratch, using pulverized coal, combined cycle and open gas turbines, nuclear power, wind turbines and solar PV. They increased demand consistent with 2035 projections.
In running the model they assumed no transmission network constraints and increasing levels of energy storage capacity and increasingly stringent limits on CO2 emissions.
They modeled scenarios ranging from 0 GW to 30 GW of storage – 0% to 30% of peak demand – with emissions limits ranging from 200 metric tons of CO2/GWh to 50 tCO2/GWh, which are about 60% 90% below prevailing 2013 emissions rates in the United States (489 tCO2/GWh) and the European Union (337 tCO2/GWh), respectively.
Running their models, the team found that energy storage delivers value by increasing the cost effective penetration of renewable energy, reducing total investments in nuclear power and gas-fired peaking units, and improving the utilization of all installed capacity. But they found that for energy storage systems with a two hour capacity, that value exceeds current costs only under strict emissions limits. The implication is that substantial cost reductions are needed to justify large-scale deployments.
Jenkins says the threshold cost for wider storage penetration is about $500/kWh for an installed two-hour storage system that could last 20 years. That compares with current costs in the $700/kWh to $1,000/kWh range for Lithium-ion battery systems, which typically have a useful life of between six to 10 years using current technology.
In contrast, a storage system with a 10-hour storage capacity delivers value consistent with the lower range of current costs of pumped hydroelectric storage, a mature technology. For emerging long duration energy storage technologies, such as liquid metal or flow batteries, the cost target for more ubiquitous adoption is $150/kWh to $200/kWh, Jenkins estimates. Since long duration batteries have larger energy reservoirs, the cost target per unit of energy storage capacity is lower than the two-hour storage resource.
The authors used pumped storage costs as a cost comparison for long-duration batteries because existing long-duration batteries, such as flow batteries, are still “pre-commercial,” Jenkins says.
The authors still see and important role for storage in decarbonizing the electricity sector, but storage is not necessary if a more diverse mix of flexible, low carbon power sources are deployed, particularly the flexible dispatch of nuclear power plants, which is used in Germany and France.
Without a flexible source of low-carbon energy, natural gas-fired plants would be necessary to meet operating reserves in scenarios with high penetrations of wind and solar power. But using gas-fired generation as a bridge has its limits because of the diminishing returns imposed by gas’ CO2 emissions.
“The bridge runs out without a zero-carbon source of flexible generation,” Jenkins said.
But there are also limits on the value of energy storage. The authors found the marginal value of energy storage declines as penetration increases.
“There is no single threshold,” Jenkins says, but “rather a gradual decline in the marginal value of each additional increment of storage. As you add more and more storage to a system, you start eliminating the most costly generation assets and reducing the price spread between high and low price hours.”
Storage part of a 'diverse mix'
In concluding, the authors say, “there is no silver bullet to decarbonize the electricity sector: the least-cost generation mix includes a diverse mix of resources and wind, solar, and flexible nuclear technologies co-exist in the optimal low-carbon generation portfolio, regardless of the level of energy storage."
However, they say, “energy storage is only strictly necessary to meet tight emissions limits in the absence of flexible dispatchable zero-carbon generation technologies."
For deploying storage in the current market, Jenkins says the best strategy for most companies is to target niche applications, such as locations with transmission constraints.
There is also a value in being a first mover. There's a race then between the declining cost of storage and the declining value as storage penetration rises, Jenkins says.
Otherwise, the authors say optimal economic penetration of storage will be limited, unless costs continue to decline well beyond current targets or in specific locations where storage systems deliver significant additional value to electricity systems not considered in the current paper, such as avoidance of transmission or distribution costs.
While the current paper focuses on the generation sector, Jenkins is working on a similar paper that will evaluate energy storage in the context of the transmission and distribution sector. He says that paper should be finished in about a year.
If energy storage can scale up and drive costs down, it could scale as quickly as solar power has over the last decade or so, Jenkins said.
"The storage industry is about where solar power was in 2005."
Follow Peter Maloney on Twitter