What do you pay, and what comes back?

The answer differs by project. So we have put here a calculator whose dials are made only from published figures. The reader is the one who moves them.

The eighteen questions covered why Japan needs batteries and why the composition of earnings is changing. What anyone who has read this far asks next is usually the same thing: so what do you pay, and what comes back? Project prices differ by project. So this article does not state a number. It lays out the dials that determine what comes back, using published figures, and puts them in a form you can move. Extra-high voltage or high voltage, the same calculator reads both.

Work it out yourself

What comes back is standby fees, seat fees and arbitrage, less costs. The calculator below uses only published levels as its dials. The initial values are 50 MW, 4 hours, ¥9 unit price and a 90% clearing rate. Move it to an aggregator's view (seat-fee-led, with a clearing rate around 20%) and the payback lengthens. The reader is the one who moves it.

Every dial comes from a published source. Unit prices from EPRX clearing results, seat fees from capacity-market clearing prices, the arbitrage spread from our own measurements on a 2 MW unit, construction costs from METI study-group materials (¥68,000/kWh average for subsidised projects; ¥49,000 system plus ¥14,000 construction above 50 MWh; ¥20,000–40,000 imported and unsubsidised). Whether two hours or four is decided by what you earn from. Primary control reserve (response within 10 seconds, sustained 5 minutes or more) and secondary I (30 minutes) can be served at two hours; if you want the three-hour composite product and a thicker seat fee, you need three to four hours, and construction cost rises accordingly.

We present three scenarios built on a project's actual figures — the area's tendered volume, how free the seats are, the connection year, the construction estimate, the measured clearing rate — alongside the project itself.

Where the clearing rate is decided

The vertical axis of the table (unit price) can be read from published figures. The horizontal axis (clearing rate) differs by project. It is set by how many MW you are relative to that area's tendered volume, and by whom you compete against.

High-voltage batteries under 10 MW have a bracket only batteries can sit in (primary offline) — about 500 MW per block nationally. As of August 2026, Tokyo is 80% full, Chubu and Kyushu 40–50%, and Kansai, Chugoku and Hokkaido are 60–80% empty (47-3). Extra-high-voltage units of 10 MW and above cannot enter that bracket. They sit in the online seats, where they bid against thermal and pumped storage. Online seats in Hokkaido, Shikoku, Kyushu and Chugoku have a shortfall rate under 1% — that is, they are filled by pumped storage and thermal — while Tohoku and Hokuriku have about a quarter free. Thermal clearing prices in FY2025 were ¥2.05–3.39 (EPRX). For a large battery to achieve a high clearing rate it has to bid below thermal, and that is what the lower rows of the table (¥3, ¥4.4) mean.

This is where our view and the aggregators' diverge. Aggregators work from the seat fee, treating balancing power as a top-up at a low clearing rate. We put balancing power at the centre and flex the clearing rate and unit price across three scenarios. Which is right can only be seen by setting the area's tendered volume (47-2) and how the seats are filling (47-3) against the project's own capacity. So for each project, we show which cell of the table it sits in.

Measured results for a 2 MW high-voltage unit

For 2 MW / 8 MWh high voltage, measured annual revenue can be published. Across nine areas over the most recent twelve months, roughly ¥40m–86m per 2 MW (¥20,000–43,000 per kW per year). Top and bottom differ by a factor of two, and the floor from the capacity market and arbitrage alone is ¥43.4m–58.41m, with balancing on top (see the column "The same 2 MW can earn ¥290m or ¥60m a year"). Construction cost is around ¥500–600m as a guide (Nikkei Energy Next); at ¥68,000/kWh it is ¥540m.

Buyers are both domestic corporates and foreign capital. For domestic corporates, the immediate expensing under the bold investment promotion tax measure, in force from 31 July 2026, is effective: where a company acquires the asset and uses it in its own business, the acquisition cost may be deductible in full in the first year (requirements and how to qualify are in columns 45 and 22). It does not change the timing of distributions or recovery. What it affects is the investor's own tax.

Published yield levels

For absolute yield levels, we cite what has been published. In a sensitivity analysis submitted by Mitsubishi Research Institute to a METI study group, where revenue is not fixed by contract, the base IRR is −1.5% even including capacity-market revenue (construction cost ¥60,000/kWh, four hours). At ¥50,000/kWh it is 0.4%, and at ¥30,000/kWh with wholesale spreads coming in above expectations, around 14%. Where revenue is fixed by contract, a general example of extra-high voltage at 47 MW / 188 MWh (four hours), layering the capacity market on a tolling contract with 75% debt, gives an equity return in the 18% range (column 30). Applying the per-kW table above to a two-hour build puts you outside those published levels. The difference lies in the view taken of the clearing rate and unit price, and we set that out per project in three scenarios. A yield quoted without its assumptions — whoever publishes it — is a number whose foundation has not yet been settled.

Five gates to clear

1. FEFTA.The electricity business is a designated sector for inward direct investment, and acquisition by foreign capital is likely to require prior notification. Core sectors, however, are limited to generators of 50 MW or more in maximum output, so neither an extra-high-voltage project of a few tens of MW nor a 2 MW high-voltage unit is expected to fall into the core category on size. The procedure is heavy, but it is a gate that can be cleared (column 24).
2. Revenue levy.The Local Tax Act treats anyone selling electricity as an electricity supply business and taxes revenue even where there is no profit (standard rate 0.75%). Whether balancing market revenue falls within scope is settled by a prior query to the prefectural tax office (column 31).
3. Fire code and siting.Above 20 kWh, installation must be notified to the fire service, and outdoor units must in principle sit at least 3 m from buildings. Urbanisation control zones can host a unit if development permission and fire clearance are obtained, but it depends on the municipality. Hold pre-consultations before signing for the site.
4. Connection queue.Against 172 GW in connection studies, 640 MW is connected (48-13). Some projects face a three-to-five-year wait on grid-side works. What foreign capital should buy is a project with a confirmed connection date, or one already running — and that is what our technical due diligence examines.
5. Delegating operations.Bidding 47 blocks a day and handling capacity-market rules is not work for an owner to do personally. It is delegated to an aggregator, for 5–15% of market revenue (the only published rate is RE100 Denryoku's 5%). Choosing an aggregator that discloses its track record in detail also helps when attaching a bank.

Is there an exit?

There is. The buyers are the utilities and the capital around them. Kansai Electric Power has set a target of developing roughly 1 GW of storage in the early 2030s and formed a fund of about 250 MW and ¥6.5bn with Kinden and MUFG (June 2026). Tokyo Gas has stacked up three 20-year tolling contracts totalling 110 MW, with six projects and 300 MW planned for operation. Tanagawa 99 MW (40% held by Kansai Electric Power) attracted Japan's first fully merchant project finance, and 49 MW in Niigata attracted a ¥10bn bond. There is Itochu's storage fund (over ¥8bn), and an SPC covering six sites and 174 MW across eight companies including Fuyo General Lease. There are seats both for capital entering mid-development and for capital buying close to completion, and the price difference is the payment for time and construction risk.

Where we stand

We hold information on more than 90 projects and work on the technical due diligence and sale of storage plants that are connected or have a confirmed connection date. For foreign capital, extra-high voltage primarily, and high voltage too; for domestic corporates, structures in which immediate expensing works. Project-specific figures are presented alongside the project, in three scenarios that state explicitly which cell of the table above it sits in.

Sources

• EPRX clearing results (March–August 2026: primary control reserve for batteries, national weighted average ¥9.3; all-resource weighted average ¥4.4; median lowest award price in the ¥3 range); EPRX, "FY2025 Trading Results," 18 June 2026 (thermal primary control reserve, average clearing price ¥2.05–3.39); Agency for Natural Resources and Energy, 4th Stable Electricity Supply WG, Material 6, 14 July 2026 (¥10 ceiling; the next step of ¥7.21 is conditional)
• Our own aggregation from EPRX trading results (August 2026, preliminary): seats and awards in the primary offline bracket (47-3); primary online tendered volume of approx. 1,500 MW per block (47-2)
• OCCTO, Capacity Market Main Auction clearing results (for FY2029: Hokkaido, Tohoku and Tokyo ¥14,812/kW; Kyushu ¥15,112; Chubu ¥10,280; Hokuriku, Kansai, Chugoku and Shikoku ¥8,785); FY2029 adjustment coefficient table (31 July 2025)
• METI, 5th FY2024 Study Group on Expanding Stationary Storage Systems, Material 3 (30 January 2025: ¥68,000/kWh; system ¥49,000/kWh above 50 MWh); 3rd meeting, Material 3 (29 August 2024: imported and unsubsidised, ¥20,000–40,000/kWh); Nikkei Energy Next (initial investment for high-voltage 2 MW / 8 MWh class, ¥500–600m)
• Our column "The same 2 MW can earn ¥290m or ¥60m a year" (annual revenue for 2 MW across nine areas, approx. ¥40m–86m; floor ¥43.4m–58.41m; JEPX component ¥11,000–17,000/kW per year); column 30 (four hours, general example, 75% debt, equity IRR in the 18% range); column 40 (construction cost); column 07 (O&M 1–2%, aggregator fees); column 24 (gates for foreign capital); column 31 (revenue levy and siting); columns 45 and 22 (immediate expensing)
• Sensitivity analysis submitted by Mitsubishi Research Institute to a METI study group (base IRR −1.5% without fixing; 0.4% at ¥50,000/kWh; around 14% at ¥30,000/kWh with spreads above expectations)
• METI, "Bold Investment Promotion Tax Measure," in force 31 July 2026; Ministry of Finance notices on designated and core sectors for inward direct investment; Local Tax Act (revenue levy on electricity supply business, standard rate 0.75%); Fire Service Act and fire prevention ordinances
• Kansai Electric Power and MUFG press release, 16 June 2026; Nikkei BP, June 2025 (Kansai Electric Power, approx. 1 GW in the early 2030s); Tokyo Gas press release, 30 June 2025; MUFG Bank, 7 May 2025 (Tanagawa 99 MW); Mizuho Securities, 17 April 2026; Nihon Chikudenchi and Ricoh Leasing, 27 April 2026; Itochu, 30 September 2024; Fuyo General Lease, 26 March 2026
• RE100 Denryoku (aggregator fee 5%)