Grid-Scale BESS Q&A

Japan's balancing market (EPRX) — the market through which transmission and distribution operators procure the reserve capacity used to keep grid frequency stable — imposes an upper limit, or price cap, on bids. The cap has been lowered in stages: from JPY 19 through last year to JPY 15 today, falling to JPY 10 from autumn 2026, with the government signaling a further reduction to around JPY 7 (these are ΔkW prices — the price per kW of capacity held on standby).

The essential point is that the cap is a ceiling, not the same thing as actual cleared prices. A lower cap does not mechanically reduce revenue by the same amount; what matters is where cleared prices actually sit, which can be verified against the auction results that are publicly disclosed.

When evaluating a revenue simulation, start with two checks: is the model using the price cap itself as an assumption, and is it anchored to published cleared prices? A credible projection starts from a discount to actual market levels and builds in year-on-year price declines. Beyond that, testing a conservative scenario in which prices fall all the way to the announced final cap level — and confirming the project still holds up — allows you to demonstrate resilience to regulatory change in concrete numbers.

This is one of the questions we hear most often from investors. The short answer: price assumptions should be anchored to published market results, not to any particular operator's business plan. EPRX (the Electric Power Reserve eXchange), which operates the balancing market, publishes clearing results; since the March 2026 market reform (the shift to day-ahead trading and 30-minute blocks), the nationwide weighted average for battery storage has been in the low JPY 9 range. A conservative model starts at or below that actual level.

Next, the clearing rate. Not every bid is awarded, so a projection built on winning every block is optimistic. The basic discipline is to discount revenue by a clearing rate (the share of bids that are actually awarded). In addition, for products such as Tertiary Reserve 2, where some blocks are simply not tendered at all, you also need to multiply by the share of blocks actually procured. Whether this two-stage discount is present makes a substantial difference to the credibility of any projection.

A further test is whether the revenue case depends on the skill of the aggregator (the operator that trades the battery in the market on the owner's behalf). A model whose IRR rests on a particular operator's trading prowess cannot be verified once that assumption is questioned. If the assumptions are built on published market results instead, then even a future tightening of rules on bidding blocks would show up simply as a lower clearing rate — and its impact can be read straight off a standard stress table.

The capacity market pays not for energy generated (kWh) but for keeping supply capability (kW) available on demand. The main auction is held four years ahead of the delivery year, and once a bid clears, the price for that year is fixed. This means revenue for the next several delivery years can be treated not as a forecast but as contracted, locked-in income. Prices differ by area — in the Kyushu area, for example, the FY2028 delivery year cleared at JPY 13,177/kW and FY2029 at JPY 15,112/kW.

Because prices are fixed four years in advance, capacity market revenue serves to underpin a storage project's annual fixed costs. Even if market-linked balancing revenue fluctuates, a project whose fixed costs are largely covered by fixed capacity income has considerably greater downside resilience.

That said, we are candid about the risks: future auction prices may come in lower, the scheme itself may be redesigned, and there are non-delivery penalties if the contracted capacity cannot actually be provided. In practice, non-delivery risk is addressed by including an availability guarantee in the O&M contract (a clause under which the O&M provider guarantees equipment uptime) combined with manufacturer warranties. In our view, the sound approach is to treat capacity market revenue as a fixed-cost underpinning rather than as the main engine of returns.

Differences in headline returns usually reflect differences in revenue structure, not differences in optimism. There are two main revenue models for battery storage. One is the tolling model — leasing the operating rights of the facility to a utility or large energy user under a long-term, fixed-price contract. Revenue is locked in, so returns are lower, and the typical structure uses debt leverage to lift the equity return. The other is the merchant model — earning directly in the balancing market and wholesale power market. Because the investor takes market risk, the expected return is correspondingly higher.

Comparing these two on the same yardstick makes merchant returns look "too high." But the real question is not the level of the return; it is how the assumptions were built. Are prices anchored to published market results? Is a clearing-rate discount applied? Are annual price declines built in? If the assumptions are conservatively constructed, a high expected return is simply the compensation for the risk being taken.

One more structural factor matters: the shape of the payback. A project whose capital recovery is concentrated in the early years of operation depends far less on price forecasts ten or twenty years out — tightening the distant assumptions barely moves the return. Conversely, a project that recovers capital thinly over the long term is highly sensitive to those far-out price assumptions. Of course, the merchant model is directly exposed to falling market prices, so we recommend looking for projects that disclose the break-even point — the price level at which capital recovery becomes at risk under stress scenarios.

A ramp-rate restriction is a limit that the transmission and distribution operator places on how quickly a storage facility may change its output — for example, a maximum of a few percent of rated output per minute — in consideration of grid stability. It is stated in the connection study reply and becomes a binding constraint under the interconnection agreement. Because it directly determines eligibility for fast-response products such as Primary Reserve, it is a frequent topic of investor questions.

The first thing to understand is that this is not something the operator can clear unilaterally through EMS (energy management system) settings. Because the restriction is written into the contract, relaxing it requires the grid operator's agreement. The practical routes are negotiation based on a power-factor change (adjusting the ratio of active to reactive power) or consultation supported by submission of the control specifications. Battery hardware itself is capable of extremely fast response, so the real issue lies in the contract and grid operations, not in equipment capability.

On the regulatory side there is a tailwind: Japan's grid code review is moving toward exempting Primary Reserve response from ramp-rate restrictions, with FY2027 as the target — but this is still under deliberation, not a decision. If an investment case relies on revenue from fast-response products, the outcome of the relaxation negotiation will drive returns, so the practical focus becomes how to secure it in the transfer agreement — for example, by making written confirmation of the relaxation a condition precedent to closing.

JC-STAR is a national labeling scheme that certifies the security conformance of IoT products. In storage investment reviews we are frequently asked whether the batteries need JC-STAR, and the answer is straightforward: the battery units themselves are out of scope.

The reason lies in the nature of the scheme. JC-STAR is a network security framework, and equipment with no network functionality simply cannot be certified under it. What falls in scope is the network-exposed equipment — the PCS (power conditioning system, which converts between DC and AC), the EMS (energy management system), gateways, and other communicating devices. The BMS (battery management system) inside the battery and local control functions are generally out of scope, and in practice the accepted interpretation is that if the upstream aggregator's system is JC-STAR-conformant, individual certification of the devices beneath it is not required.

Looking ahead, JC-STAR certification is expected to become a requirement in public-sector tenders, so whether equipment is compliant is becoming a checkpoint at the equipment-selection stage. The scheme is evolving, so please confirm the latest scope and requirements when finalizing a specific equipment configuration.

FEFTA (the Foreign Exchange and Foreign Trade Act) regulates investment by foreign investors in Japanese companies, and the electricity business is among the designated sectors subject to notification. As a rule, FEFTA procedures apply regardless of project size — even for a relatively small storage business. The trigger is the transfer of SPV shares to a foreign investor (or its Japanese entity), and the filing obligation rests with the buyer.

On timing, the key point is this: the purchase agreement may be signed before the notification is filed, but closing (payment of the balance and transfer of the interest) cannot take place until the notification has been accepted. In practice, the purchase agreement therefore expressly makes FEFTA clearance (acceptance of the notification) a condition precedent (CP) to closing. Note that merely incorporating a Japanese company (an SPC) is outside FEFTA's scope, but where a foreign investor establishes a company in Japan for the purpose of the electricity business and then invests, notifications may be required both at incorporation and at share acquisition — two filings in total. The safest settlement structure is a simultaneous exchange through an escrow account (a third-party-administered account) after the notification has been accepted.

On a future resale: if the project is sold on to foreign capital, that buyer must file a fresh notification; a sale to a Japanese entity generally requires no additional filing. You may hear views such as "small projects are exempt" or "a single project doesn't trigger FEFTA," but the applicability of exemption categories and post-transaction reporting is highly fact-specific. Where there is any doubt, filing a prior notification is the safe course — and legal counsel should always be consulted on a deal-by-deal basis.

The starting point is that the main noise source at a storage facility is not the battery units but the PCS (power conditioning system — the equipment that converts between DC and AC). Battery enclosures are typically more than 10 dB quieter than the PCS, so noise mitigation is, in practice, a question of how the PCS is handled.

PCS noise levels vary widely by manufacturer and model. Some common large units run at around 85 dB, quieter models at around 60 dB, and newer designs advertising roughly 45 dB have recently come to market. A 10 dB difference is perceived as roughly twice as loud, so equipment selection itself can be the single most effective noise measure.

Applicable regulations differ by municipality and zoning: as a guide, residential zones may require around 40–45 dB at night, and even outside residential zones a limit of around 55 dB at the site boundary can apply. Where a large PCS is to be installed in a tightly regulated area, the response combines layout planning that places noise-emitting equipment as far as possible from dwellings, sound barriers, low-noise equipment, and adjustments to the operating schedule. One further point: community relations are not solely a matter of decibel readings. Whether a project has followed a careful process — such as resident briefings before construction — is itself worth verifying during investment due diligence.

"The project cleared every regulation, yet failed the buyer's internal screening" — a familiar story in battery storage transactions. Beyond legal requirements, each buyer maintains its own internal siting criteria, and these vary widely from company to company. The same project can pass one buyer's review without issue while breaching another buyer's internal thresholds.

Examples we have encountered in our dialogue with buyers include: setback distance from dwellings (ranging widely, from companies that accept roughly 30 m subject to soundproofing measures, to those requiring 100 m or more between dwellings and the PCS); hazard maps (companies that draw a hard line at a given assumed flood depth); restrictions on eligible investment regions; and companies whose first check is whether the site falls within a strictly noise-regulated zone.

For sellers and developers, the implication is clear: regulatory compliance is only the starting point. Preparing measured setback distances, hazard information, zoning classification, and the applicable noise limits early in the process widens the pool of potential buyers and accelerates the sale. Conversely, investors who share their own criteria with the intermediary at the outset avoid spending time reviewing projects that will never fit.

That an EPC-unselected project carries completion risk is a fact that should be acknowledged plainly. What we can offer is a framework for distinguishing the quality of that risk. First, the nature of the remaining work. If the grid rights (interconnection agreement and payment of the grid connection cost contribution), land title, and long-lead equipment such as the substation gear are already secured, what remains is standard construction — a fundamentally different risk profile from an early-stage development.

Second, contractual risk transfer. If the EPC contract requires a completion guarantee and liquidated damages (LD — pre-agreed compensation payable by the EPC for schedule delays), and the contractor is selected from firms with the creditworthiness to stand behind those terms, completion risk is contractually shifted to the EPC. The contractor's own credit standing is critical here: a guarantee is only as good as the financial strength behind it. Choosing an EPC on price alone can render the guarantee meaningless. The selection criteria — completion guarantee, LD provisions, and counterparty credit — are points investors should verify during diligence.

Third, the role of performance testing. The performance tests required before participating in the balancing market are not a hurdle unique to any one project; they are a standard procedure common to all market participants. If the schedule properly reserves a testing and inspection period between grid interconnection and commercial operation, most of the "unexpected test rework" risk can be assessed simply as a question of schedule buffer.

The largest difference is consumption tax. A transfer of shares (SPC equity) is exempt from Japanese consumption tax, whereas an asset transfer (a sale of equipment and project rights) is a taxable transaction. Because extra-high-voltage storage projects involve large transaction values, the consumption tax impact is substantial — one practical reason why the equity transfer (M&A) route, acquiring the project-owning company outright, is so widely used. There are also projects where, for the seller's own tax reasons, an equity transfer is the only format on offer.

At the same time, an equity transfer means acquiring the company whole. Beyond the visible assets, the buyer may inherit existing contractual relationships, future payment obligations, and in some cases off-balance-sheet liabilities. Due diligence (the pre-acquisition investigation) therefore matters even more than in an asset deal. The right conclusion is not "equity transfer wins because there is no consumption tax," but a decision made after scrutinizing exactly what is being inherited.

Please note that tax treatment can vary with the terms of the transaction and the circumstances of the parties; always confirm the specifics with your tax and legal advisors. Acting as a buy-side facilitator, we help organize the information needed to choose the right structure.

Yes, it is. In the market we do observe transactions in which a facility, after grid interconnection, participates in the balancing market for several months to build an operating track record and is then sold. Rather than "build and sell immediately," the strategy is to demonstrate with measured results that the asset earns before selling it.

Why does a track record command value? For buyers — financial investors especially — the greatest uncertainty is whether the facility will actually earn in the market as projected. However refined the model, a projection is still a projection. Even a few months of actual clearing, response, and revenue records converts part of that uncertainty into measured data, opening the project to a more conservative class of buyers. The same asset can attract a different buyer universe, and a different price level, depending on whether that record exists.

We would be candid about the caveats, however. The premium attributable to a track record varies greatly by project, timing, and market conditions — there is no generalizable rule of thumb. A track record also cuts both ways: if the first few months of operation underperform, the price becomes harder to defend, not easier. And during the holding period before resale, the owner bears market price risk. This is best understood not as "season it and it will always sell higher," but as one strategy for widening the range of exit options.

This is one of the questions we receive most often from investors. In grid-scale storage transactions, alongside the "asset transfer" route (selling equipment and land individually), it is common practice to acquire the entire equity interest in the SPC — a special purpose company established solely to hold the project — in what is effectively an M&A transaction.

The reason is that the core value of a storage project lies in its bundle of contracts: the position under the interconnection agreement with the utility (the grid connection rights), the land rights (superficies or leasehold), and the various permits. In an asset transfer, each of these must be re-assigned individually with counterparty consent; acquiring the company itself transfers them comprehensively. Market practice is one project, one SPC, with the project ring-fenced from the credit risk of the seller's other businesses and investors (so-called bankruptcy remoteness). There are also projects that, for the seller's own tax reasons, are offered on an equity-transfer-only basis. For foreign investors, the usual structure is two-tiered — incorporating a Japanese stock company which then acquires the SPC equity — combined with the notification procedures under FEFTA (the Foreign Exchange and Foreign Trade Act).

On tax, a transfer of shares or equity interests is exempt from Japanese consumption tax. In an asset transfer, consumption tax applies to the equipment component, whereas in an equity transfer it does not — a difference that becomes material at the transaction sizes typical of extra-high-voltage projects.

To be candid about the flip side: an equity transfer means buying the vehicle whole, so any debts or contractual defects the SPC has incurred in the past come with it. Due diligence on off-balance-sheet and contingent liabilities, and the design of representations and warranties in the purchase agreement, matter even more than in an asset deal.

Exits fall into three broad categories: (1) hold the asset and recover the investment in full; (2) build an operating track record in the market after commissioning, then sell; or (3) sell to the next buyer once capital recovery is substantially secured.

Under option (1), holding, merchant-model projects can be designed to target full capital recovery in roughly four years under the base case (the so-called self-amortizing structure, covered in a separate section) — meaning the investment can stand on its own without depending on an exit. Option (2), season then sell, involves participating in the balancing market (the market for the reserve capacity that keeps electricity supply and demand in balance) after grid interconnection, building a period of operating results, and then selling. A completed facility with no operating history and an operating asset that can demonstrate real market revenue are valued differently by buyers — and the market does show transactions consistent with acquiring operating projects, participating in the market, and subsequently selling. Option (3) suits investors seeking an exit on a multi-year horizon, since it aligns the timing of capital recovery with the timing of sale.

We would add the honest caveats. Resale prices depend on future market and regulatory conditions, so there is no guarantee that a track record will always command a higher price. And the size of any "track-record premium" is something to be substantiated with project-specific numbers — it cannot be quoted as a general figure. If you are considering an acquisition with an exit in mind, we would encourage a discussion based on the terms of the specific project.

"Self-amortizing" describes a design that targets full recovery of the invested capital early in the operating life — around year four under the base case. Rather than betting on whether a 20-year revenue forecast proves correct, the investment decision is made to stand or fall on the reliability of revenue during the few years it takes to recover capital.

The revenue structure is what makes this possible. One of the pillars, the capacity market (the market for securing generation and supply capability, where auctions fix prices four years ahead of delivery), becomes locked-in revenue once cleared. In one example among the projects we handle, this fixed income alone covers roughly 90% of annual fixed costs, underpinning the cost side of the business. For the variable revenue from the balancing market, our modeling policy is to anchor price assumptions to published market results — not any operator's plan — starting below actual market levels and building in annual price declines.

The consequence of this structure is that the bulk of the return is determined in the near-term years, where prices are either already contracted or verifiable against published market results. Even materially lowering the price assumptions for the distant future has a comparatively small effect on the outcome. That is the substance behind the phrase "not dependent on long-term forecasts."

This does not mean the risk is zero. If market prices in the recovery-critical early years fall well below expectations, capital recovery is pushed back. That is precisely why we disclose a stress table that extends to the break-even point — the price level at which the investment merely returns its principal — so investors can verify the resilience with their own eyes.

This is one of the most substantive challenges we receive from institutional investors. Let us say first: where the criticism that "the scenario range is too narrow" is warranted, we accept it plainly and respond by expanding and disclosing a wider stress table. Our policy is not to defend assumptions by hiding them, but to show the break-even point and let investors judge.

That said, the rule of thumb that "the base case is really the upside case" applies to some assets and not to others. The heuristic was honed in the underwriting of assets whose capital recovery is spread thinly over 20 years — projects premised on long-term fixed-price contracts or long-range forecasts. In such structures, aggressive assumptions about the distant future are enough to quietly turn the base case into an upside case, and in that context the criticism is entirely fair.

Early-recovery (self-amortizing) projects, however, have a different structure. The bulk of the return is determined within the next few years, where prices are either already contracted or can be calibrated against published market results, so even sharply reducing far-future price assumptions does not move the outcome much. Put differently, the arena in which "is the base case too bullish?" should be debated has been compressed from 20 years out to the next few years.

For that reason, we do not engage in an unwinnable argument about where the base case ought to sit. Instead, we deliver the model with prices, clearing rates, and other assumptions exposed as input cells, switchable between bear, base, and bull cases, so investors can test it under their own assumptions. The validity of those assumptions can be checked directly against the market's published results.

Our answer, in short, is that the comparison is usually being made across different playing fields. Most of the projects cited as "past track record" are based on tolling agreements — structures in which the operating rights to the facility are leased to a counterparty such as a utility under a long-term, fixed-fee contract, with the counterparty absorbing market price risk. One frequently cited example is a project with a 20-year fixed-price tolling agreement, capacity market revenue accounting for roughly 60% of income, and 80% debt financing, producing an IRR (internal rate of return) of 18%. Thick fixed revenue supports leverage, and that is the return it produces — a perfectly rational design.

A merchant (market-linked) project, by contrast, leaves market price risk with the investor. It is a basic principle of finance that the party bearing the risk earns the higher expected return, so "the IRR is higher than tolling-based track records, therefore it is optimistic" does not follow. If anything, a merchant project offering only tolling-level returns could be viewed as inadequate compensation for the risk being assumed.

The real question is not the absolute level of the return but how the assumptions were set — specifically, how the price assumptions relate to published market results: above or below actual cleared levels, and how future declines are built in. Our policy is to anchor to the published actuals and start from a level further discounted below them, a cross-check anyone can perform.

In fairness, we would add that the tolling model has clear virtues: stable revenue and easier access to bank financing. For investors seeking steady income, tolling is genuinely the better fit in many situations. What matters is knowing which risk you are taking, and comparing projects on the same playing field.

Grid-scale battery storage transactions involve large sums, and the object of sale is not merely equipment but a bundle of contracts — so the mechanics of settlement often determine whether a deal succeeds. Here are the points investors most often ask about.

First, the seller's initial screen is the buyer's funding. In practice, a striking number of storage sale processes collapse because the buyer's financing fails to materialize, so buyers who have secured funds — or can evidence them — are strongly preferred. Buyers relying on debt should expect to be asked early for proof of financing certainty and payment capacity. The size of the deposit is a matter for negotiation in each transaction, but the common approach is to put down a deposit of a certain percentage to solidify one's negotiating position — and exclusivity or preferred-negotiation rights are, as a rule, not granted free of charge. In equity-transfer deals, where there is no differentiation on equipment specification, price and certainty of funds become the axis of competition.

Settlement, in our standard practice, is a simultaneous exchange through an escrow account (a settlement account administered by a third party): payment of the balance and delivery of the equity and documents occur at the same moment, so neither party is ever exposed alone.

Where the buyer is a foreign investor, prior notification under FEFTA (the Foreign Exchange and Foreign Trade Act) may be required. In that case the purchase agreement itself may be signed before the filing, but closing (payment of the balance and handover) cannot occur before the notification is accepted. The safe practice is to state acceptance of the notification expressly as a condition precedent (CP) to closing in the purchase agreement, and to build the review period into the transaction timeline.

Completion risk — the risk that the facility is not finished as planned — cannot be reduced to zero, but the greater part of it can be transferred by contract. The standard protection is to require, in the contract with the EPC (the contractor responsible for engineering, procurement, and construction as a single package), a completion guarantee and liquidated damages (LD) provisions, placing the risk of delay or non-completion on the EPC. A guarantee is only meaningful if its issuer has the credit to honor it, so EPC selection requires a contractor of sufficient scale and creditworthiness to stand behind a completion guarantee. Equally, the effective size of the completion risk depends on how far the project has progressed — whether the grid rights, the land, and the long-lead equipment (such as extra-high-voltage substation gear) are already secured, leaving only standard construction work outstanding.

Before commercial operation there is performance testing. This is a standard procedure common to every market participant, not a special hurdle for any one project. A sound plan can be recognized by whether a period for testing and inspection is explicitly reserved in the schedule between grid interconnection and the start of commercial operation.

Once operating, the two pillars are the O&M contract (the operations and maintenance agreement) and manufacturer warranties, working in combination. The capacity market, for example, imposes penalties for failing to deliver the contracted capacity — and the standard protection is the availability guarantee in the O&M contract (under which the O&M provider guarantees the share of time the equipment is capable of operating) combined with the manufacturer warranty. Equipment failures themselves are addressed through manufacturer warranties providing equivalent-product replacement, and some projects carry extended warranties as well.

To be candid: warranty durations, scope, and exclusions differ substantially by manufacturer and by project. Do not take comfort from the phrase "it comes with a warranty" — make verification of the original warranty terms a mandatory item in due diligence. We ourselves confirm the warranty terms on each project before presenting it.