Wind, solar & battery costs have plummeted & energy storage installs are booming. Good timing for my new paper w/@dhariksm & @nsepulvedam on "Long-run system value of battery energy storage in future grids with increasing wind and solar generation"
👀➡️authors.elsevier.com/a/1bLLO15eiezz…
👀➡️authors.elsevier.com/a/1bLLO15eiezz…
Our new study out in @ElsevierEnergy's journal Applied Energy finds that the economic value of storage increases as variable renewable energy generation supplies an increasing share of electricity supply but that storage cost declines are needed to realize full potential.
We used a detailed electricity system planning model (energy.mit.edu/wp-content/upl…) to examine battery storage & determine key drivers that impact its economic value, how value changes w/increasing deployment over time, and implications for the long-term cost-effectiveness of storage.
What we found: Storage helps make better use of electricity system assets & that can defer or eliminate unnecessary investment. This ‘capacity deferral,’ or substitution of batteries for generation or transmission capacity, is primary source of storage value, explains @dhariksm
Other sources of storage value include: providing operating reserves, avoiding fuel cost + wear & tear incurred by cycling on and off gas plants, and shifting energy from low price periods to high value periods (energy arbitrage)—but the these sources are secondary in importance.
Importantly, we found that as more & more storage is deployed, the value of additional storage steadily falls. That creates a race between declining cost and declining value, and the cost of batteries must continue to fall if storage is to play a major role in electricity systems
We found that without further cost reductions, a relatively small magnitude (4% of peak demand) of short-duration (2-4 hr) battery storage is cost-effective in grids with 50-60% of electricity supply that comes from VRE generation...
“The picture is more favorable to storage adoption if future cost projections ($150 per kilowatt-hour for 4-hour storage) are realized” notes @dhariksm. At those costs, storage deployment may be economically justified at up to 8-16% of peak demand if wind/solar shares reach 60%.
We find, as you would expect, that increasing solar & wind deployment increases the value of storage, all else equal. In our modeling, we force incrementally more wind and solar with an energy share constraint similar to the way a renewable portfolio standard would...
However, if wind and solar penetration is accompanied by declining wind and solar cost (as we'd expect), the implications for storage value are more ambiguous and in some cases, cheaper renewables mean *less* value for batteries, as the quote below from @nsepulvedam explains
Since batteries derive much of their value from reducing wind/solar capacity needed to meet a given RE share, cheaper wind and solar could negatively impact storage value, which could create pressure to reduce storage costs in order to remain cost-effective, as @dhariksm notes ⤵️
This was not ALWAYS the case however. In some cases, cheaper renewables meant changes in WHERE & WHAT KIND of renewables were built, and that sometimes increased value opportunity for storage, eg by maximizing use of long-distance transmission to access cheap wind sites far away.
We also know battery storage projects are increasingly competing w/ `peaker' plants used to meet the highest demand periods and that competing w/more frequently used plants is harder, as @JulianSpector explains greentechmedia.com/articles/read/… We took a close look at this competition too.
We find that adding 1 MW of storage capacity displaces <1 MW of natural gas plant capacity. The reason: to shut down 1 MW of gas, storage must not only provide 1 MW of power, but also be capable of sustaining production for as many hours in a row as the gas capacity operates.
That means you need many hours of energy storage capacity (megawatt-hours) as well. Our study also finds that this capacity substitution ratio declines steadily as storage tries to displace more gas capacity, as my quote below explains.
Given the importance of energy storage duration to gas capacity substitution, we find that longer storage durations of eight hours generally have greater marginal gas displacement than storage with two hours of duration. Fig 8 again illustrates this ⤵️
However, we also find that the additional system value from longer durations (due to greater gas substitution) does not outweigh the additional cost of the needed energy storage capacity (added cost for more MWhs). See figure below for net value depiction (= value - cost)
“From the perspective of power system decarbonization, this suggests the need to develop cheaper energy storage technologies that can be cost-effectively deployed for much longer durations, in order to displace dispatchable fossil fuel generation,” says @dhariksm
To address this need, our same team (@nsepulvedam, @dhariksm, me + @TppMIT '19 Aurora Edington) have a follow-up paper under review now that provides the most extensive evaluation of the potential role and value of long-duration energy storage technologies to date. Stay tuned!
Final notes: Since bulk of storage value comes from capacity deferral -- reducing investment in underutilized wind, solar, gas or transmission assets -- that raises the question: how will storage owners monetize this value in practice?
Our findings thus highlight the importance of exploring reforms to electricity market structures or contracting practices that enable storage developers to monetize the value from substituting generation and transmission capacity—a central component of their economic viability.
That may require reforms to capacity markets & procurement of 'non-wires' alternatives to transmission (or distribution, which was out of scope for this study but part of another forthcoming work by the same study team).
The need to monetize the value of increased wind, solar or transmission utilization may also explain recent boom in integrated development of wind, solar, & energy storage projects, which may be a preferred method for capturing storage value in today's markets.
That's it. That's the thread. Thanks for reading!
You can check out the full paper (only modestly longer than this monster thread!) at this link for the next 50 days authors.elsevier.com/a/1bLLO15eiezz…
If you have trouble accessing or need a PDF, please message me for a personal use copy.
You can check out the full paper (only modestly longer than this monster thread!) at this link for the next 50 days authors.elsevier.com/a/1bLLO15eiezz…
If you have trouble accessing or need a PDF, please message me for a personal use copy.
