The Twenty-Year ESA: Where a 200MW Battery Contract Actually Lives

Plus Power and the Tennessee Valley Authority signed a twenty-year energy storage agreement covering the 200MW/800MWh Crawfish Creek battery system in Alabama, and the figure that should arrest attention is not the megawatt rating, the megawatt-hour duration, or even the dollar value implicit in a multi-decade tolling-style commitment, but the tenor itself — twenty years on a lithium-ion asset whose underlying cell warranty horizon, even under the most favourable supplier negotiations, sits at roughly half that length. The ESA is therefore not a contract that runs in parallel with the asset's natural life; it is a contract that outlives the warranted technology by a factor of two, and every clause of consequence in such a document is, at root, a clause about who absorbs what happens between year ten and year twenty.

The configuration that produces this observation is specific and worth naming. TVA is a federal-corporate utility whose dispatch logic, balancing-area conventions, and rate construction sit outside the FERC-regulated investor-owned utility framework that dominates contract templates in PJM, ERCOT, or CAISO. Investment-grade counterparty credit is therefore present in an unusually clean form — there is no public utility commission rate proceeding that can re-trace the offtaker's underlying ability to pay — but the dispatch and curtailment economics of a TVA contract behave differently from what a pro forma drawn against a typical IOU template would assume. Lenders sizing debt service against the contract have to read the dispatch protocol, the ancillary services participation rules, and the curtailment compensation language as they actually apply within the TVA balancing authority, not as they would translate from an organized-market analogue.

The 4-hour duration ratio embedded in the 800MWh-over-200MW configuration is the second structural fact that controls how the contract behaves. A 4-hour battery is not an arbitrage asset in any meaningful sense; the round-trip economics of charging in low-price hours and discharging in high-price hours, even at the most generous spread assumptions, do not pay for the capital stack at the cell prices and balance-of-plant costs that prevailed when this project would have been procured. The 4-hour configuration is built around capacity value and ancillary services — frequency response, regulation, ramping reserves, and increasingly the suite of fast-acting reliability products that vertically integrated and federal utilities have begun to procure on long-tenor contracts. The revenue mechanic, in other words, sits closer to a tolling agreement than to a merchant exposure, which is why a twenty-year tenor was negotiable in the first place.

Once that frame is set, the augmentation clause becomes the centre of gravity for the entire document. A lithium-ion battery degrades on a curve that is well understood at year five and increasingly contested at year fifteen; the supplier's performance guarantee typically steps down to a contractual floor — often in the high seventies as a percentage of nameplate energy capacity — and once that floor is approached, the asset must either be augmented (additional cells installed to restore capacity), de-rated (the contract amends downward to reflect what the asset can deliver), or accept a liquidated damages exposure that compounds over the residual contract life. The ESA either specifies which of these paths is taken and who funds it, or it leaves the question open, in which case it is decided by litigation. There is no third option.

The negotiation around augmentation is rarely visible in press releases because it is not a single number. It is the joint specification of (a) the trigger — a measured energy capacity threshold below which augmentation must occur, (b) the funding mechanic — whether the developer carries the capex line item as a project cost, recovered through an indexed capacity payment, or whether the offtaker contributes through an augmentation fee adder, (c) the technology lock — whether augmentation must use cells of equivalent chemistry and supplier, which constrains the developer to a market that may not exist in year twelve, and (d) the performance reset — whether the post-augmentation guarantee runs from the original COD or from the augmentation event. Each of these four sub-clauses moves the lender's debt sizing materially. The developer that loses two of them at the term sheet stage typically loses the project finance economics altogether and has to fall back on a balance-sheet structure with a substantially worse return profile.

The performance step-down clause sits adjacent and is frequently confused with augmentation, but it operates on a different mechanic. Step-down governs what happens when the asset is below the contract guarantee but above the augmentation trigger — the band in which the offtaker is receiving less capacity than contracted but the developer is not yet obligated to install new cells. The contract has to specify whether the capacity payment scales linearly with delivered capacity, whether it scales with a deadband (a tolerance below which no payment adjustment occurs), or whether it triggers a discrete cliff at a defined threshold. Linear scaling is the most balanced from a financeability standpoint; cliffs concentrate cash-flow risk in narrow degradation bands that lenders price aggressively. The structural negotiation around the step-down curve is rarely visible to anyone outside the deal room, but it determines whether the project's debt service coverage ratio holds across the degradation envelope or punctures it in year fourteen.

Dispatch optionality is the third sub-mechanic, and it is the one that most often surfaces as an enforcement dispute rather than a contract negotiation. The ESA grants TVA the right to dispatch the battery within parameters — daily cycle limits, depth-of-discharge limits, calendar constraints around grid events — and the developer carries the obligation to maintain availability against those parameters. The friction surfaces when the offtaker's dispatch behaviour, while technically within the contract envelope, accelerates the degradation curve in ways that the original performance model did not anticipate. A battery cycled daily at 95% depth-of-discharge degrades materially faster than the same battery cycled three times weekly at 80%, and if the contract caps cycles but not depth, or vice versa, the developer absorbs a degradation acceleration that re-prices augmentation into year eight rather than year twelve. The dispatch envelope, defined precisely, is a financeability clause masquerading as an operational protocol.

End-of-life liability is the fourth sub-mechanic and the one that has only recently begun to receive serious attention in long-tenor ESAs. A 200MW/800MWh facility at end of life carries a decommissioning cost that is non-trivial — the cells themselves have residual value if the second-life and recycling markets in the late 2040s develop as currently projected, but the balance-of-plant, the inverters, the thermal management infrastructure, and the site restoration obligations cost real money to retire. The contract has to specify whether the developer escrows decommissioning reserves over the contract life, whether the offtaker contributes through a decommissioning adder, and what happens to the residual cell value — does it accrue to the developer as a captured value, to the offtaker as a contract setoff, or is it shared on a defined formula. Lenders increasingly require the decommissioning reserve mechanic to be defined at financial close, which means the negotiation that should happen in year twenty actually happens in year zero.

The structural fragility this configuration produces is that all four mechanics — augmentation, step-down, dispatch optionality, end-of-life — interact with each other and with the underlying degradation model in ways that are not separable. A contract that is balanced on augmentation but loose on dispatch optionality produces a developer that is augmenting more frequently than the financial model assumed, and the augmentation funding mechanic, even if favourably negotiated, does not catch up. A contract that is balanced on step-down but rigid on technology lock for augmentation produces a developer that is contractually obligated to source cells from a supplier who has discontinued the relevant SKU, and the resulting force majeure dispute occupies the next eighteen months of the project's operating life. The fragility is not in any single clause; it is in the joint distribution of the four clauses against a degradation curve that is itself uncertain at the long end.

BEIREK's work on the developer side of long-tenor storage agreements is, in practice, the work of structuring these four sub-mechanics so that lender-grade certainty survives cell degradation, warranty expiration, and the technological turnover that inevitably occurs across a twenty-year contract life. The headline tariff is the easiest part of the negotiation; the augmentation funding mechanic, the step-down curve, the dispatch envelope, and the decommissioning reserve construction are where the value is captured or surrendered, and they are where contract lifecycle management has to start at the term sheet stage rather than at COD. We structure the augmentation clause so that the funding mechanic is indexed to a defined cell-price benchmark rather than left as a fixed cash obligation that becomes regressive as battery costs decline; we negotiate the dispatch envelope so that cycle and depth constraints are jointly bounded against a contractual degradation curve that the offtaker has accepted; we build the decommissioning reserve construction so that lenders see a defined cash mechanic at financial close rather than a year-twenty obligation hanging in the balance sheet.

The work continues across the contract life because the mechanics that were negotiated at term sheet do not enforce themselves. Augmentation triggers have to be measured against capacity test protocols that the offtaker has the right to challenge; step-down calculations have to be reconciled against availability data that the offtaker's metering may not capture identically to the developer's; dispatch envelope breaches have to be flagged in a manner that does not blow the relationship but does preserve the contractual claim. Contract lifecycle management on a twenty-year ESA is not a documentation function; it is an enforcement function disguised as a documentation function, and the developer that treats it as the former typically discovers in year nine that the latter is what they actually needed.

The closing proposition is straightforward enough to put as a question. If the augmentation clause in a twenty-year storage agreement is the single largest determinant of whether the asset is financeable, refinanceable, and exit-able at fair value, why does it remain — in most term sheets that cross our desk — the clause that receives the least drafting attention until lender comments arrive and the negotiation has to be reopened from a weaker position?