ARPA-E Energy Storage
Updated: Aug 15, 2019
ARPA-E held a two day workshop in Chicago from December 7th to the 8th inorder to discuss a potential funding opportunity being considered for long-duration energy storage (defined as 8-50 hours). Evolved Energy was invited to offer perspectives on the need and prospects for this type of storage.
Under consideration were chemical and thermal technologies with very low cycle cost and round trip efficiency greater than 50%. Excluded were mechanical technologies (Ex. pumped hydro) or power-to fuel processes (Ex. hydrogen).
The impetus for the meeting stems from the growing recgonition that wind and solar are the cheapest new sources of energy today and that long duration storage may have significant value facilitating their delivery in a system with very high penetrations of variable generation.
In addition to the delivered deck (attached below), the following were the main thoughts I walked away with after the workshop:
The Long duration storage (LDS) market is likely to develop post-2030 because that's when we reach 50%+ renewable penetrations in select locations. Most other applications prior are niche.
For LDS applications outside of, or prior to, a high renewables scenario, roundtrip efficiency is critical. For instance, arbitraging heat-rates between CTs and CCGTs you need high round trip efficiency for the economics to pencil.
For LDS applications at very high renewables, efficiency is much less important. Instead the question is what is the cost of your renewables and how cheap can you get the battery. You'd rather have $40/MWh solar and a $20/MWh dollar battery with 50% efficiency than $40/MWh solar and a $150/MWh battery with 90% efficiency. Efficiency being lower just means it is necessary to build more solar, which is cheap, perhaps cheaper than a different battery chemistry with higher efficiency.
Following from observation #1, near-term LDS requires a battery chemistry that is competitive to manufacture at low volumes. The market likely won't have 10s of GWh of demand in the 2020s for LDS, and thus, the business case should not be contingent on this.
Following from observation #1, near-term LDS the stack cost ($/kW) is still important because any company going after LDS in the 2020s will have to sell into a shorter duration market competitively to survive. Ex. a chemistry at $1000/kW and $10/kWh would be very valuable in a high wind 2040 scenario, but they would have a hard time competing with Li-ion in the meantime.
The biggest risk for LDS is that it turns out to be neither fish nor foul and that automotive batteries get used for short duration, high efficiency applications, and then power2X technologies are used to handle long-duration imbalance (curtailment). Handling curtailment in a high renewable scenario with non-firm loads that have high operating to fixed cost ratios and that produce valuable products (Ex. desalination, arc furnaces, electrolysis, etc.) may be more cost effective than round trip storage.