A detailed study into the cost of energy storage has been published in Joule online magazine and summarised by Vox. The article focuses on what it refers to as the “energy storage capacity cost”. This metric represents the cost levels that storage must reach to economically support renewable generation as back-up. The analysis models examined extended periods of low renewable energy output across four different locations. This was studied along with the associated costs of renewable generation. Using this data, they identified target cost thresholds for energy storage that would make it feasible to retire fossil-fuel-based backup infrastructure.
In Europe, there’s a weather pattern called the kalte dunkelflaute, a cold, dark period that can last about two weeks and happens every few years. These periods vastly reduce renewable energy output. However, shorter, smaller-scale versions of this event happen much more often. They can take place several times a year, or even every night when there’s little wind and no solar power available.
Defining energy storage capacity cost
One area that lacks clarity in the Joule article is the exact definition of “energy storage capacity cost.” It is unclear whether this refers to:
- Capital cost per megawatt-hour (MWh) generated annually
- Levelised Cost of Storage (LCOS)
- Levelised Cost of Electricity (LCOE), inclusive of input electricity costs
The article lists figures in dollars per kilowatt-hour ($/kWh), which can be converted to $/MWh by multiplying by 1,000. For a grid aiming for 100% availability, the target energy storage capacity cost is stated as $10–12/kWh ($10,000–$12,000/MWh). For 95% availability, the threshold rises to $150/kWh.
Applying this to a hypothetical 40MW, 200MWh plant operating 4.5 hours per day for 350 days a year would yield 63,000MWh annually. With a capital expenditure of $60 million, this equates to a capex of approximately $0.95/MWh generated. Adding a 5% annual cost of capital raises this to $1/MWh.
Costing up Green CAES
Storage technologies such as our Green CAES model benefit from increased duration. This is because longer-duration plants are more cost-efficient per MWh delivered. In contrast, metrics like LCOS and LCOE are more dependent on electricity throughput and do not benefit as directly from increased duration.
Current estimates suggest an LCOS of $68/MWh and an LCOE of $110/MWh for this example plant. This places the plant well below the $20/kWh benchmark referenced in the Joule analysis, even under extremely low utilisation scenarios (e.g. 13.5 minutes per day).
These figures indicate that the cost of energy storage using technologies like Green CAES is already competitive with long-term battery storage. Large-scale, long-duration storage are more practical than battery systems. This is due to raw material limitations, particularly with lithium, cobalt, and rare earth metals and battery degradation over time.
February 2020
