← Back to Knowledge

Let us put only published figures into the formula in the text. The subject is a high-voltage 2 MW / 8 MWh plant bidding into primary offline through an aggregator. The one question: using published figures alone, how many yen is the adder? The formula and what its words mean are in 50-2 Rules.

The answer first. Set in the Tokyo area, the adder is ¥1.411 in year 1 of operation, ¥0.959 in year 2, ¥0.809 in year 10 and ¥0.711 in year 17. When depreciation ends in year 18, the amount you may recover turns negative and the adder falls to the Type A ¥0.33. All figures are in ¥/ΔkW·30min (per promise to hold 1 kW open for 30 minutes).

Figure 1 The adder for 2 MW / 8 MWh in the Tokyo area over 20 years (estimate from published figures)

← Scroll sideways →

The adder for 2 MW / 8 MWh in Tokyo over 20 years ¥1.411 in year 1 of operation, ¥0.959 in year 2, ¥0.809 in year 10, ¥0.711 in year 17, and the Type A ¥0.33 from year 18. For reference, without deducting deemed capacity-market revenue it would be ¥1.637 in year 2 and ¥0.419 in year 18. 0 0.5 1 1.5 2 ¥/ΔkW·30min 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1.41 0.96 0.81 0.71 Estimate (base) Without deemed capacity revenue deducted (reference) Years from start of operation Type A ¥0.33
Yellow is the estimate. The grey dotted line is a reference value without deducting deemed capacity-market revenue. From year 18 the calculated value turns negative, so the Type A ¥0.33 is used. Our estimate (50-3).

1. The figures we put in

We included only items with a published unit price. Items without one are not zero but blank. They are left as spaces for your own plant's figures.

ItemValue usedSource and method
Construction cost¥68,000/kWh × 8,000 kWh = ¥544 millionMETI, “Current status and issues of stationary storage systems” (12 March 2025, Document 6), p. 5. Battery ¥41,000, PCS ¥6,000, other ¥7,000, construction work ¥14,000 (analysis by Mitsubishi Research Institute; COLUMN 40)
DepreciationStraight-line over 17 years: ¥32,096,000 a yearThe Ministerial Ordinance on Useful Lives, Appended Table 2, gives 17 years for “electricity business equipment — other equipment — mainly made of metal”. The only basis for putting grid-scale batteries on this line is press reports (second-hand accounts of the National Tax Agency's view)
Repairs¥5,000/kW/year × 2,000 kW = ¥10 million a yearAgency for Natural Resources and Energy, 21 June 2023, Document 7 (battery parameters for the Long-term Decarbonization Power Source Auction)
Generator-side charge (kW portion)Each TSO's unit price × 2,000 kW × 12 months. Tokyo: ¥100.12/kW·month, ¥2,402,880 a yearEach general transmission and distribution operator. Five of them raise their unit prices on 1 November 2026, so the new prices are used. Batteries are exempt from the kWh portion
Fixed asset taxTax rate 1.4%. Assessed value is acquisition cost × 0.936 (first year), then × 0.873 each year. Zero in year 1 of operation. Tax rounded to the nearest yenLocal Tax Act, Article 350; Fixed Asset Valuation Standards, Appended Table 15
Capacity market (deemed revenue)2,000 kW × adjustment coefficient 70.8% (from monthly averages before rounding; rounded to the nearest kW) = expected capacity 1,415 kW. × area price ¥15,111/kW = ¥21,382,065 a year. Deducted from year 2OCCTO: FY2029 adjustment coefficient table, Form 2, and main auction results
Wholesale-market revenue0See section 5 for why
Volume you can bid (expected bid volume)2,000 kW × 48 slots × 365 days × 90% = 31,536,000 ΔkW·30minOur placeholder (90% allows for charge/discharge constraints)
Personnel costs(your own figure)No published unit price
Outsourcing (O&M)(your own figure)No published unit price
Insurance, land rent(your own figure)No published unit price. Whether they can be included is also unsettled
Aggregator costs (identified fixed costs)(your own figure)Identify from the contract breakdown

Adding the blank items raises the adder by that much. With a divisor of 31,536,000, every ¥10 million a year of fixed costs adds about ¥0.32.

The items we did include also leave room in how they are set. Expected capacity in the capacity market is the 12-month average of “each month's adjustment coefficient × each month's dischargeable output”; taking dischargeable output as the 2,000 kW installed capacity is the upper end. We assumed the generator-side charge applies to the full 2,000 kW (with a contract for charging, the base becomes smaller). Fixed asset tax is not among the text's examples (unsettled), but it is a fixed cost by nature, and the Long-term Decarbonization Power Source Auction counts it in maintenance costs, so we included it. Taking it out lowers year 2 by ¥0.226.

2. The breakdown for year 2

Figure 2 The breakdown for year 2 (Tokyo) — add, subtract, divide

← Scroll sideways →

The breakdown for year 2 (Tokyo) Fixed costs are depreciation ¥32.10 million, repairs ¥10.00 million, fixed asset tax ¥7.13 million and the generator-side charge ¥2.40 million, ¥51.63 million in total. Less ¥21.38 million of deemed capacity-market revenue leaves ¥30.25 million. Divided by the 31.54 million ΔkW·30min you can bid, that is ¥0.959. ¥ million / year Depreciation 32.10 Repairs 10.00 Fixed asset tax 7.13 Generator-side charge 2.40 Fixed costs ¥51.63M ① A year's fixed costs − ¥21.38M Capacity (deemed) ② Deduct (even if not bid) = ¥30.25M Recoverable limit ③ Remainder ÷ Volume you can bid 31.54M Adder ¥0.959 ④ Divide Left out (no published unit price): personnel, outsourcing, insurance, land rent, aggregator fixed costs. Adding them raises it.
Fixed costs are four items: depreciation, repairs, fixed asset tax and the generator-side charge. Personnel, outsourcing, insurance, land rent and aggregator fixed costs have no published unit price and are not included. Our estimate (50-3).
Tokyo, base (¥)Year 1Year 2Year 10Year 17Year 18
Depreciation32,096,00032,096,00032,096,00030,463,9990
Repairs10,000,00010,000,00010,000,00010,000,00010,000,000
Fixed asset tax07,128,5762,405,009929,429811,392
Generator-side charge (kW portion)2,402,8802,402,8802,402,8802,402,8802,402,880
Total fixed costs44,498,88051,627,45646,903,88943,796,30813,214,272
Less capacity market (deemed)0−21,382,065−21,382,065−21,382,065−21,382,065
Upper limit to recover44,498,88030,245,39125,521,82422,414,243Negative
Volume you can bid (ΔkW·30min)31,536,00031,536,00031,536,00031,536,00031,536,000
Adder (¥/ΔkW·30min)1.4110.9590.8090.7110.33 (Type A)

In year 1 the capacity market is not deducted, so fixed costs of ¥44,498,880 are divided as they are, giving ¥1.411. In year 2 fixed asset tax is added, but ¥21,382,065 of deemed capacity-market revenue is deducted, so the adder falls to ¥0.959. After that it falls slowly, only as fast as fixed asset tax shrinks each year.

When depreciation ends in year 18, fixed costs become ¥13,214,272, less than deemed capacity-market revenue. The text says nothing about years in which the upper limit turns negative (unsettled). From the sentence “after the fixed costs for the current fiscal year have been recovered, the certain amount shall be that of a Type A resource”, we use ¥0.33 here. In years where the calculated value falls below ¥0.33, we also use ¥0.33.

3. The biggest step comes in the year deduction starts

Figure 3 Where it drops, and by how much (Tokyo)

← Scroll sideways →

Where it drops, and by how much From year 1 to year 2 it drops ¥0.452 (deemed capacity-market revenue starts to be deducted). From year 17 to year 18 it drops ¥0.381 (depreciation ends). 0 0.1 0.2 0.3 0.4 0.5 0.6 ¥/ΔkW·30min −¥0.452 Year 1 → 2 −¥0.381 Year 17 → 18
Year 1 → 2 is the step where deemed capacity-market revenue starts to be deducted; year 17 → 18 is the step where depreciation ends. Our estimate.

The biggest drop in 20 years comes in the year deemed capacity-market revenue starts to be deducted. In this estimate that is year 2, a fall of ¥0.452 from ¥1.411 to ¥0.959. It is larger than the step in year 18 when depreciation ends (¥0.381, from ¥0.711 to ¥0.33), and that held in every one of the nine areas.

The position of the step moves, though, with the year deduction starts. What the text exempts is “resources that, being newly in operation, could not bid into the capacity market in time”, and it does not say up to which year that applies (unsettled). If deduction starts later, the step moves back to that year.

4. Comparing the nine areas

Figure 4 The adder in nine areas (dots for year 1, bars for year 2)

← Scroll sideways →

The adder in nine areas (years 1 and 2) The year-2 adder is Tohoku 0.798, Hokkaido 0.851, Kyushu 0.868, Chugoku 0.948, Hokuriku 0.957, Tokyo 0.959, Shikoku 0.965, Chubu 0.985 and Kansai 1.000 yen. Year 1 is ¥1.400–1.423. 0 0.4 0.8 1.2 1.6 ¥/ΔkW·30min Hokkaido 0.851 Tohoku 0.798 Tokyo 0.959 Chubu 0.985 Hokuriku 0.957 Kansai 1.000 Chugoku 0.948 Shikoku 0.965 Kyushu 0.868 Year 2 Year 1 (before capacity deduction)
The year-2 adder (bars) and year 1 (dots). The adjustment coefficient is the 12-month average of the monthly values for pure pumped storage, 4 hours, for FY2029 (our calculation from OCCTO's published table). The area price is the FY2029 main auction clearing price. Our estimate.
AreaYear 1Year 2Year 10Year 17Year 18
Hokkaido1.4190.8510.7010.6020.33
Tohoku1.4060.7980.6480.5500.33
Tokyo1.4110.9590.8090.7110.33
Chubu1.4080.9850.8350.7360.33
Hokuriku1.4060.9570.8070.7090.33
Kansai1.4231.0000.8500.7510.33
Chugoku1.4000.9480.7980.6990.33
Shikoku1.4190.9650.8150.7160.33
Kyushu1.4100.8680.7180.6190.33

In year 1 the adder is ¥1.400–1.423, and the only difference between areas is the unit price of the generator-side charge. From year 2 the gap widens, from ¥0.798 in Tohoku to ¥1.000 in Kansai. The larger the amount deducted as expected capacity × area price, the lower the adder — so Tohoku, Hokkaido and Kyushu come out lowest. Tokyo has a high area price, but its 4-hour adjustment coefficient, at 70.8%, is the lowest of the nine areas, so less is deducted.

5. What moves it, and how

Figure 5 How the year-2 adder moves when inputs change (Tokyo)

← Scroll sideways →

How the year-2 adder moves when inputs change (Tokyo) Base ¥0.959. Construction cost ±20%: ¥1.208 / ¥0.710; share you can bid 80% / 95%: ¥1.079 / ¥0.909; adjustment coefficient set at 100%: ¥0.679; no capacity-market deduction: ¥1.637. 0 0.5 1 1.5 2 ¥/ΔkW·30min Construction +20% 1.208 Construction −20% 0.710 Biddable share 80% 1.079 Biddable share 95% 0.909 Coefficient at 100% 0.679 No capacity deduction 1.637 Base 0.959 Base ¥0.959
The vertical dashed line is the base of ¥0.959. Our estimate.

We set wholesale-market revenue at zero. The text lists “wholesale-market revenue” as an example of revenue from other markets but does not say whether it means expected or actual revenue. Having assumed that 48 slots × 90% go to the balancing market, the plant cannot discharge on the wholesale market in those hours. Wholesale earnings enter the ceiling's first term as opportunity cost, so deducting them here too would count them twice. If you earn wholesale revenue in the remaining slots you do not bid, there is room to be asked to deduct that part (unsettled).

6. Falling to ¥0.33 partway through the year

The Type B adder applies only until the year's fixed costs are recovered. Once they are, it falls to the Type A ¥0.33, even partway through the year. When recovery counts as complete depends on how much of ΔkW revenue is counted towards it, and the text does not say (unsettled).

Figure 6 Falling to ¥0.33 partway through the year (Tokyo, year-2 example)

← Scroll sideways →

Falling to ¥0.33 partway through the year Once the year's fixed costs have been recovered, the adder for the rest of the period is ¥0.33. The timing depends on how much ΔkW revenue counts towards recovery: counting all of it, Tokyo's year 2 fills up in about 37 days; counting only the adder part, at the end of the fiscal year. April next March Reading A: count all ΔkW revenue Recovered in ~37 days → ¥0.33 for the rest (Tokyo, year 2, every slot won at ¥8.58) Reading B: count only the adder part Type B amount until the end of the year The text does not say which reading applies (unsettled). Ask the EGC secretariat.
Tokyo's year-2 upper limit is ¥30,245,391. If 2 MW wins every slot at September's Tokyo primary offline average of ¥8.58, each slot earns ¥17,160, so it fills up in about 1,763 slots — around 37 days. Our estimate.

If only the adder part counts, recovery finishes only at the end of the fiscal year. If all ΔkW revenue counts, the picture changes completely. Tokyo's year-2 upper limit is ¥30,245,391. If 2 MW wins every slot at September's Tokyo primary offline average of ¥8.58, each slot brings in ¥17,160, and the limit fills in about 1,763 slots — around 37 days. On that reading, you would bid under a ¥0.33 ceiling for most of the year.

7. For owners and buyers

Set from published figures, the adder is in the ¥1 range even in year 1 and below ¥1 from year 2 — an order of magnitude below the ¥10 cap. The figures on this page are a guide to the lower end of the ceiling; if you can identify blank items such as personnel and aggregator costs from breakdowns, it rises by that much. Whether you can identify the items is, directly, your room to price.

When buyers build their models, the safe choice is not to set the ΔkW price at the ¥10 cap but at the lower of the actual price in their area (50-4 Market) and the height the formula in the text can explain. How depreciation is set changes the adder, and the treatment when immediate depreciation is used is unsettled. The incentive is summarised in COLUMN 45.

Every calculation can be reproduced with our public script (scripts/column50/ceiling_calc.py in the scix-web repository; inputs in inputs.json in the same place).

Sources

Supervised by
Shinya Nakashima, Ph.D. (Eng.) — Representative Director, Science X Inc.

Works on the development, sale and technical due diligence of grid-scale battery storage plants in Japan. This column is written and supervised on the basis of that hands-on transaction and due-diligence practice. This English edition is a translation of the Japanese original; where the wording of the guideline matters, the Japanese primary sources govern.