This isn't a pitch. It's material for your own judgment.
Equipment makers, EPCs and aggregators all have a reason to sell to you. ScienceX holds no capital ties to any of them. On that footing, we set out the Japanese grid-scale storage market as it actually is — from primary sources (METI, the Agency for Natural Resources and Energy, OCCTO), as a company that buys and sells projects.
From revenue stacking across LDA, the capacity market and the balancing market, to grid connection, land, contracts, O&M, tax, subsidies and cybersecurity (JC-STAR) — the questions that matter, from entry to operation, with sources.
We publish because we are engineers.
Flood hazard maps are life-safety (evacuation) maps. Using MLIT's multi-stage inundation maps and the Flood Control Economic Survey Manual, a battery-storage site's location hazard can be translated into Expected Annual Loss (EAL) — a monetary figure. From a 2MW/8MWh model calculation to the ripple into insurance design, the veto / priceable / design-input three-tier sort, and a 14-item one-day workflow for a single candidate site.
As grid-scale batteries grow larger, cooling fan and PCS noise becomes a source of community complaints. Reviewing nighttime regulation values by zoning district and cost considerations for soundproofing measures.
Grid distance, available capacity, regulations, terrain, and landowner structure — assess your land's potential as a grid-scale battery site against 5 criteria.
On 1 September the ΔkW price cap fell from ¥15 to ¥10, with ¥7.21 shown as the next step, and since March procurement volume has moved from 3σ-equivalent to 1σ-equivalent. Counted from the published data of EPRX and JEPX: required volume is 0.55 and procurement volume 0.62, but the combined product is 0.97 — essentially flat. Batteries hold 9% of the combined product and 23% of primary online, so their seats come to under 10% of the market. Their own frame went from a 67.9% shortfall rate in April–June to 53% in August, and Kyushu cleared at ¥4.9 in August with the cap still at ¥15. The 5 concluding numbers here, with the evidence on 3 pages: procurement volume, supply and floor.
From delivery on September 1, 2026, the ΔkW price cap for primary reserve, secondary reserve 1 and the composite product falls from ¥15 to ¥10.00/ΔkW·30min. Aggregating all 158 days of EPRX awarded results, the central estimate of what disappears is −8.6% nationwide, ≈¥30 billion a year; primary offline — the product where batteries hold a 99% share — loses 22%, and Tohoku × offline 32%. The market moved first: August averaged ¥9.80, so the mechanical cut lands on Tohoku and Tokyo. With a 2MW/8MWh translation, a ¥7.21 stress test, and five things to do before September 1.
For a high-voltage 2 MW / 8 MWh battery, the all-market median is ¥270–290M/yr across Japan's nine areas, with Tohoku the overall leader at ~¥290M — roughly 80% of it from the balancing market. Hokkaido, the JEPX leader, could not capture it: its balancing market was too thin. The arbitrage map and the ΔkW map do not overlap. We fix all three markets from measured data, largest first, and trace the 3–5 year decay curve with real per-area numbers.
Balancing Market data across Japan's 9 power areas from EPRX — post-reform (Apr–Jun 2026) shortage rates, battery share, and area capacity-market prices to guide battery siting decisions.
An investment decision on a grid-scale storage asset comes down to the equity IRR. On the same extra-high-voltage 47 MW / 188 MWh, project IRR, equity IRR and DSCR move sharply across four contracts — balancing, capacity + wholesale, LDA and tolling. We show how fixed income sets the return, on stated assumptions and the published Japanese track record.
Same 5% rate on gross (total revenue) vs net (market profit) creates ¥1.25M+ annual difference. 3-market payment flows, 5 global contract models, and why operational quality matters more than fee rates.
EPRX operates Japan's balancing market. How it differs from JEPX and the capacity market, the trading-member structure, and the steps for a battery to participate.
Product design for primary through tertiary regulation, GF/LFC/EDC/GC control functions, online vs. offline connection differences, and how the FY2026 day-ahead trading transition affects battery operations.
LDA award rate is 20%, approximately 90% of market revenue is refunded. Examining the feasibility conditions and decision criteria for a "full merchant" strategy relying solely on 3-market stacking.
Nearly all current battery projects use non-firm connections. Analyzing the impact of output curtailment on revenue and latest developments regarding capacity and balancing market participation eligibility.
Despite both being "capacity market" mechanisms, the Main Auction and LDA have fundamentally different design philosophies. Organizing the criteria for deciding which to participate in.
Grid-scale battery revenue cannot be completed in a single market. How you combine three markets determines business success or failure.
For those considering entry into the battery business, we explain the basic structure of LDA through the 3rd round regulatory changes, based on primary sources.
The Bold Investment Promotion Tax Incentive came into force on 31 July 2026 and the nine regional METI bureaux are taking applications. A four-question 30-second test, the 31 March 2029 confirmation deadline and the five-year clock that starts at confirmation, and the four points that decide it for a storage station — the 17-year useful life that becomes the denominator of the 15% ROI test, where the lease line falls, and why the in-service date rather than the payment date fixes the year. Plus the six questions still under enquiry.
Whatever its output, a storage station is a business-use electrical facility that needs a safety regulations filing and a chief electrical engineer. The applicant is the retail electricity provider, the grid connection cost contribution is fixed only after you apply, and the “4-year rule” appears in no document. 49.5 kW is one twentieth of the 1 MW minimum bid. Five assumptions checked against the statutes, the government Q&A and the ten utilities’ rules — and where the line falls against high voltage.
Overseas capital is starting to view Japan's high-voltage grid-scale battery storage (2MW/8MWh class) as an acquisition target. Yet once a buyer moves, the first obstacle is not returns or siting but whether the capital can reach Japan at all. We map the exit and entry capital gates (FEFTA), scarce grid capacity and land, rights sales versus completed assets, the equipment exit (JC-STAR and the U.S. lists), and land and tax, all from primary sources.
Japan's plan to require JC-STAR ★1-certified equipment for grid connection is fixed in direction, but the governing text is not yet promulgated. For EHV/HV the trigger is the interconnection contract application and applies from April 2027; LV under 50 kW from October 2027. A practitioner's read of device scope (PCS/EMS), the status of overseas-made PCS, and what is still open.
Japan's FY2026 Bold Investment Promotion Tax Incentive. The five gates a grid-scale battery storage owner must clear to actually use immediate depreciation — plus how tax deferral works, how contract structure (tolling vs. merchant) decides eligibility, and capital-gains tax at exit. Explained plainly, from primary sources.
Even at a 1/3 headline rate, the 17-year disposal restriction, 5-year retention, and 3-year outcome reporting erode the effective subsidy rate. A 12-item judgment checklist, cost decomposition matrix, and three strategy archetypes (Kinokawa / Maibara Koto / Hirohara). Co-produced with Yuta Katou (Enememo).
Interconnection study applications surged 6x year-over-year. With increased deposits and mandatory land title submissions, capacity hoarding will be eliminated. What this means for already-developed projects.
What is for sale changes across four stages — land only, rights with a connection study response, interconnection approval secured, and an operating asset — and with it the order of magnitude of the price, the documents to check and the work left after closing. Who sells and why they let go, who buys and what they base the price on, and what moves in what order from teaser to NDA, due-diligence pack, letter of intent and name change — down to the difference between brokering a deal and buying to resell, and the ways a deal stalls. Written from the side of a company that buys projects and resells them.
Why do two identical 2 MW / 8 MWh plants carry different price tags? This column breaks down the four acquisition structures used in Japan — completed-asset sale, rights transfer, SPC equity, and operating secondary — and the price ladder behind them (rights ¥150–200M, completed ¥700–800M, with track record ¥800M–1.0B), separating hard cost from the price of risk transfer. It also explains why buying an operating plant does not automatically bring its market qualifications with it (only the capacity market provides for express succession), how tax, construction-licensing and contract succession differ between asset and equity deals, and what the balancing-market price-cap cut to ¥10 from 1 September 2026 means for buyers.
Why are battery projects worth hundreds of millions of yen bought with cash rather than debt? We test the “batteries can’t be collateralised” myth point by point, then map the five domestic debt archetypes that emerged between 2023 and 2026 — Himeji, Hirowara, Nishigo (¥6.2bn / 20 years), Tanagawa, Niigata Yamaya (¥10bn bond) and a 14-site high-voltage portfolio. From there: what lenders actually doubt in a revenue model, and the three verifiable numbers — market outturn, the announced policy path and the contracted revenue ratio — that answer them.
Against 172 GW under connection study, only 640 MW is actually grid-connected — most deals on the market still exist only on paper. For a first-time buyer of grid-scale storage, this guide sorts fact from claim across nine domains and 43 checklist items: grid rights, land, permits, equipment cybersecurity, the site, power contracts, and the contractual vessel — what to check before buying, with which document, and where the line lies.
One and the same battery is labeled differently by four separate laws — the Electricity Business Act, the Local Tax Act, the depreciation rules, and the City Planning / Agricultural Land Acts. Size, siting and tax are legally coupled and can't be chosen independently. A legal-and-tax map for investors and owners, read through an AC ~2 MW / DC ~8 MWh single-site model.
Why three terms — Grid Connection Study, Wheeling Service Contract, and Generation Output Adjustment Supply Contract — are structurally prone to confusion. The 66 characters of §8(2)(ho) (April 2024), the April 2026 boundary without transitional measures, seven failure modes observed in the field, and a 13-item progress-depth checklist for pre-transfer DD — all reconstructed from primary sources.
5 layered contracts — wheeling, GSA, retail, BG, market participation. Pumped-storage measure cuts wheeling fees by ¥10M/year; 5-layer 21-contract architecture yields ¥200M differential over 20 years. Includes 11-item pre-transfer checklist.
The components that make up battery development rights pricing and the framework for evaluating their value. Entirely different logic from solar FIT "ID resale" — understand the fair value of these rights.
LDA-type and full merchant-type projects have fundamentally different IRR and DSCR profiles. A framework for the cash flow structure that determines project finance feasibility.
Grid-scale batteries are neither "bigger is better" nor "smaller is easier to start." Revenue structure, financing, and risk characteristics fundamentally differ by scale.
Compare leaving it idle, selling the rights, or changing partners — why waiting costs the most, and the practical steps to sell development rights.
How to separate land from rights, file a name change with the grid operator, and 5 contract considerations versus solar FIT rights sales.
Answer “what does a storage plant cost to build” with a single number and you will usually be wrong. METI puts the subsidised 2024 average at 68,000 yen/kWh; the same set of documents also records 20,000–40,000 yen/kWh for projects using overseas equipment without a subsidy. The gap is not error but condition. We split 68,000 yen/kWh into battery 41,000, PCS 6,000, other 7,000 and construction works 14,000, then show why the unit cost falls above 50 MWh, how the total for a 2 MW / 8 MWh plant moves between 272 and 544 million yen, what a one-third subsidy leaves, and where the public sensitivity analysis puts the profitability line.
A decision framework for grid-scale storage O&M. The law names an execution party in only two places — substation safety supervision and interconnection cybersecurity — leaving DC-side correction and communications a contractual "gap." Mapped with a responsibility matrix, plus the 5x unplanned-outage count, equipment-level JC-STAR requirements, product-by-product communication requirements, and international LTSA availability guarantees.
Solar's "movables + business interruption" package doesn't transfer to battery storage. Six insurance layers — property, machinery, BI, liability, cyber, theft — with insured-value design, BI indemnity criteria, and premium benchmarks.
Site screening, community engagement, regulatory confirmation, boundary survey, interconnection study, grading, and EPC. A practitioner's complete guide to battery-specific development processes distinct from solar.
If you want to change the battery or PCS manufacturer after acquiring development rights, does the interconnection study need to be redone? Explaining the OCCTO "interconnection study necessity confirmation" system and practical aspects of equipment changes.
Even for the same 50MW battery, interconnection costs range from hundreds of millions to billions of yen. Understanding this "10x gap" structure is what separates success from failure in battery business.
Articles on next-generation energy (SMR, e-fuel), and more will be published in sequence.
For project-specific information that cannot be assessed from articles alone,
we will disclose details after signing an NDA following your inquiry.