Daily minimum-maximum (min-max) price spread contracts (“DMMPS” contracts) may be necessary to bring about the energy storage revolution that many believe is critical for ultra-low carbon power grids.
The US energy market does not feature wide enough spreads for merchant investors in energy arbitrage-focused storage projects to make sufficient returns. Too much risk – not enough reward. But DMMPS contracts could be part of an important basket of solutions to enable some long-term revenue certainty and its corresponding drop in expected investor returns.
Trading on DMMPS elicits something new from the power contract provider community and maybe even new entrants. That is, providers need to hold a power forecast view on daily min-max pricing spreads which is dependent on advanced power modeling, and not just a view on the dynamics of a conventional power market trading book. DMMPS may be an intriguing product for volatility-hunting investors with power modeling capabilities, such as hedge funds, to make profitable yet risk-managed power market bets.
Today’s Trading Products Are Insufficient to Enable Arbitrage-Based Financings
Energy storage is increasingly being viewed as the likely mid- to long-term solution to balance out low marginal cost intermittent generators (solar and wind). In wholesale power markets, this means the storage systems are performing energy arbitrage. They are charging when their nodal (price at their exact substation) energy price is low and discharging when high.
But how is the new era ushered in? Do we expect investors to take merchant risk on the future pricing spread of their storage investments at substantial market scale (tens of GWs)? If so, we’re going to wait a very long time for this arbitrage revolution.
One could argue that the revolution is upon us with the very large storage (often solar+storage) deals in California and elsewhere (e.g. NV, HI, AZ, and OK) that have been announced in recent years. But these projects shift power on a set schedule, which is a blend of energy arbitrage and energy capacity, but much more the latter. And none of these projects hold up to transparent market pricing on energy-only or energy-plus-energy capacity.
For energy storage to truly be at grid-balancing scale, market pricing and associated financial instruments are going to have to be the main enablers. Currently though, commodity trading desks do not yet have the right tools in the toolbox to help. They generally contract against a structured power “book”, which basically can be conceptualized as 288 month-hour averages (12 months x 24 hours) in any trading year, going out about 12 years. Trading month-hour prices against each other yields far from sufficient revenue.
To make real money on energy arbitrage, an energy storage project has to game daily minimum-to-maximum pricing. The hours of charging and discharging need to constantly change for anywhere close to optimal revenue, and not just by month, but often from one day to another.
Recent Historical US Power Market DMMPS
In Table 1 below the min-max (minimum hourly price vs. maximum hourly price) spread of some of the larger hubs and zones in US power markets for 2017 thru 2019. Both month-hour spread and the daily minimum vs. maximum priced hourly spread (DMMPS) are provided.
Figure 1 depicts the month-hour spreads (orange bars) versus the ratio of DMMPS to month-hour spreads. One can plainly see that there is no clear relationship between DMMPS and month-hour spreads. For instance, ERCOT North had the highest month-hour spread but the 2nd lowest ratio.
Comparing Energy Price Min-Max Spreads (2017-2019 data)
Therefore, performing energy arbitrage against month-hour spreads (that which is contractable today from commodity trading banks) is not only insufficiently economic, but also irrelevant to the revenue stream that such an energy storage project needs, namely the energy arbitrage from DMMPS.
The Money-Maker (Node) vs. the Trading Basis (Hub/Zone)
For energy arbitrage-focused storage projects to make sufficient returns, the energy arbitrage would have to be performed on a nodal basis. Only wisely-selected nodes have high enough arbitrage pay-out.
Though, any hedge fund or other financial entity offering a DMMPS contract would likely only do so with settlement based on a hub or zone, for two reasons:
(1) Modeling Risk: It’s highly unlikely the energy storage project investors and the DMMPS contract provider would take the same aggressive view on arbitrage at the target nodes.
(2) Liquidity: The DMMPS contract provider will want a highly liquid trading point (i.e. hub or zone) to be able to backstop the design of their contractual product.
Energy arbitrage-focused storage project developers will carefully select nodes that they expect to yield much larger spreads (2x to 3x over hub or zone) in the reasonably modellable first two or three years of operations. Investors, however, will probably take the prudent expectation that price spread will decay toward the mean (i.e. hub/zone price) thereafter.
So, once one layers in the intellectual honesty of what’s viable, it’s prudent to limit market pricing expectations below those of the more accurately modellable first few years. We believe a node forecasted to hold a DMMPS of 150% (2.5x) above hub/zone node in year, might be reasonably anticipated to average only about 60% (1.6x) over a 10-year project lifespan (1.6x average note: starting at 2.5x for first three years, reducing linearly to 1x by year six, and then remaining at 1x years six to ten). We apply this 1.6x average node-over-hub/zone DMMPS factor and the ramp down from 2.5x to 1.0x in our economic analysis in the section below.
Square Peg and Round Hole: Market Pricing and Merchant Return Expectations
Many energy arbitrage-focused energy storage projects will do their best to optimize revenue against ancillary services, so there is actually multi-level arbitrage going on (arbitraging ancillary services vs. energy arbitrage). But if we really expect widespread energy storage deployment to balance intermittent generation, the revenue for energy arbitrage is really going to have to “drive the bus.” Thus, to take a clear-eyed look at this issue, let’s run the math on energy arbitrage-only scenarios.
Below in Table 2, we are looking simply at required DMMPS revenue for an arbitrage-performing 1-hr capacity duration (i.e. power capacity = energy capacity) energy storage project. Below are some key assumptions behind Table 2:
- RETURN RATE (IRR): We use an average private equity expected return rate of 25% for a totally merchant energy arbitrage project with no project or back-leveraged debt.
- PROJECT LIFE: Revenue is over 10 years, so that means 1 cycle per day (1 full charge; 1 full discharge), which about matches common cycle-based warranties often pegged around 3,500 cycles.
- DISPATCH ACCURACY: We assume very intelligent, but not perfect arbitrage dispatch. So, we only monetize 85% of the revenue of what perfect dispatch against the market DMMPS would yield.
- PROJECT COSTS: Future projects’ capital expense (“CapEx”), roundtrip efficiency, and operating costs were derived directly or via correlations from NREL and the DOE(1)(2) reports released in the past two years.
- We are using a 3.4x multiplier, the DOE’s historical (2017 COD projects) cost ratio between medium-duration (0.5 to 2 hrs) and long-duration (≥2 hrs) battery projects, for converting NREL’s 4-hr low-case $/kWh pricing forecast to a 1-hr $/kWh forecast.
- Of NREL’s pricing cases, their low-case prices most closely matches 7X Energy’s project cost analyses based on recent confidential battery OEM pricing offers.
Table 2: 10-year Monetized DMMPS (Revenue) and Associated Market DMMPS Required for Expected Investor Return on a Merchant Arbitrage-Focused Storage Project
Table 2 makes it clear that the future DMMPS required for this kind of energy storage investment are way higher versus than what we’ve observed in recent years. The largest 2017-2019 historical DMMPS (listed previously in Table 1), $107/MWh in ERCOT North, miles below the hub/zone DMMPS required in Table 2 (average is $411/MWh).
Financeable Scenarios with a DMMPS Contract
But what if we can push down the financier’s expected internal rate of return (IRR) low enough to make the hub/zone DMMPS a bit more viable? That’s how a DMMPS contracting product could be so useful.
Table 3 below shows the same 2022 to 2028 COD energy storage projects at a 10% IRR, roughly indicative of an energy storage project with significant (75% of capital stack) project debt (@ 5% interest rate), with the contract DMMPS strike price equal to the debt service payment.
While we don’t have a strong sense of what a DMMPS contract premium would be, we estimate it at effectively about 10% of the strike price on an annualized basis. The contract premium for all contract years will likely need to be paid all upfront, so this 10%-of-strike-price value is actually an “effective” annual cost. We expect the DMMPS contract provider to require a minimum dispatch performance, but with a reasonable margin of error.
Table 3: 10-year Monetized DMMPS (Revenue) and Associated Market DMMPS Required for Expected Investor Return on Arbitrage-Focused Storage Project with DMMPS Contract
The hub/zone DMMPS average ($247/MWh) listed above is still much higher than what we’ve historically seen ($70/MWh on average). Particularly considering the DMMPS contract provider has to believe the hub/zone spread in the Market DMMPs Averages column will materialize. But it’s certainly way closer to historical hub/zone DMMPS than in the merchant case ($177/MWh above instead of $341/MWh above). And one must keep a few other points in mind that could make the economics more realistic.
- Nodal DMMPS may be increasing because of growing solar, and in particular, wind project clusters. These areas have nodes which experience negative pricing more often.
- Changes in ISO/RTO policy that are fairer to energy storage may be implemented (e.g. why are settlement price floors in the neg-$100s but price caps in the many $1000s).
- Almost all energy-arbitrage performing energy storage projects deployed in this decade will have their revenue augmented by ancillary services. This would push up the average hub/zone DMMPS that is relevant to their energy arbitrage actions above the annual average.
The energy storage future will stretch project developers, OEMs, and power contract providers toward a host of new innovations. Focusing contract solutions on how the money can actually be made will be necessary to bridge the present with the future. DMMPS contracts, or other conceptual brethren, may not be for those financial groups looking to walk only the beaten path, but for those looking to get a foothold in a major global growth space. It may be a little scary, but DMMPS ain’t for wimps.
1 U.S. Department of Energy (July 2019). “Energy Storage Technology and Cost Characterization Report” https://www.energy.gov/sites/prod/files/2019/07/f65/Storage%20Cost%20and%20Performance%20Characterization%20Report_Final.pdf
2 U.S. Energy Information Administration (May 2018). “U.S. Battery Storage Market Trends” https://www.eia.gov/analysis/studies/electricity/batterystorage/pdf/battery_storage.pdf