When and where batteries arrive
: the small seat is filling, the room is still wide

← Back to Knowledge

Since this summer we have been hearing that batteries are already saturated. The price cap came down, and in Kyushu and Chubu clearing prices fell. Count it out and the answer splits in two. The small seat where batteries sit has indeed begun to fill. But even that seat is still half open, and in the building that is Japan's balancing market for ΔkW (EPRX) there are 7 rooms, of which the ones batteries sit in come to under 10% of the whole.

This page counts the supply side: when, where and how much battery capacity arrives. Procurement volume, the size of the room, is on 47-2, and where the price stops is on 47-4.

1. Batteries have 2 kinds of seat

← scroll horizontally →

The balancing market has 5 products (primary control reserve, secondary control reserve ① and ②, and tertiary control reserve ① and ②), but from a battery's point of view there are 2 kinds of seat.

One is the "primary offline frame." It can be offered without an online link to the transmission and distribution operator (TSO), and 100% of the awarded volume is batteries. It is a small frame of about 500 MW nationwide; in August batteries took 238 MW of it, and the remaining 259 MW (52%) sat as open seats that nobody won. Batteries of 10 MW and above have to connect online (COLUMN 15), so this is mainly a seat for batteries under 10 MW.

The other is the "online seats." Primary online, secondary ①, secondary ② and tertiary ① are procured inside the combined market, which puts primary through tertiary ① out together, and there thermal, pumped storage and batteries compete on the same bids. In this market the same power source can bid into several products at once (combined award), so the bars in Figure 1 add up to more than the physical capacity. The battery share has to be read product by product: 23% in primary online and 9% in the combined product. The rest is thermal and pumped storage, and there are almost no open seats online (6% in primary online). Awards in primary online are 65–72% thermal and 5–13% pumped storage. Thermal, which answers with governor-free (GF) response, is the master of this seat.

Extra-high-voltage batteries (10 MW and above) sit in the latter, the online seats.

2. The small frame has begun to fill — still half open. The online seats are taken, but they change hands

← scroll horizontally →

Figure 2 is the share of procurement volume that was not awarded, the shortfall rate. It is counted the same way as in COLUMN 16, area by area and slot by slot and then summed. The batteries-only frame (primary offline) went from 74% in April to 53% in August. That runs with the 67.9% for April–June in COLUMN 16, and awarded battery volume grew about 1.8-fold in 4 months. Even so, half of the frame is still open.

The online seats move differently. The shortfall rate for primary online went from 38% in April to 18% in June, and has been around 10% since July. The combined product has been almost zero since June. The seats are filled by thermal, pumped storage and batteries. But they are not seats hard-partitioned by procurement volume. In 30% of slots the awarded volume exceeds the procurement volume, and in August an average of 88 MW in primary online and 262 MW in the combined product was awarded above the procurement volume. Even when the seats are taken, a cheap bid gets in. The online seats are ones to look at not for "are they open" but for "can they be taken over." This is the seat that extra-high-voltage batteries will be moving into. If nuclear restarts stop thermal at night, the thermal share of this seat opens up (→ 47-2).

Looking at total bid volume alone, Tokyo, Chubu and Kyushu have had months above the procurement volume. Seats still count as open when measured by awards because most of the bidding does not reach an award (80–100% of the shortfall in those 3 areas in August). Because the same battery can stack bids across several products (combined award), the total awarded battery volume inside the same 30-minute slot reaches double the connected capacity of 84 × 10,000 kW in some months (a maximum of 1,685 MW in August). Total bid volume looks larger than the volume that can be seated. As in COLUMN 16, we line the counting up on awards.

By area, how full the seats are differs. Here "full" and "open" have to be looked at room by room.

← scroll horizontally →

Figure 3 lines up the shortfall rate by area for the 3 seats a battery can enter.

In the primary offline frame at the top, the fullest is Tokyo, with a shortfall rate of 20%. Chubu, Tohoku and Kyushu are 40–50% open, and Kansai, Chugoku and Hokkaido 60–80%. The reason is simple: the batteries that are connected and running are still concentrated in Tokyo, Chubu and Kyushu. The right way to read Hokkaido's openness is not "Hokkaido has seats to spare" but "batteries have barely arrived in Hokkaido yet." Hokkaido's primary offline seat is only 17 MW, and the award is 3 MW.

In primary online in the middle, thermal, pumped storage and batteries compete for the same seat. Hokkaido, Shikoku, Kyushu and Chugoku have shortfall rates of 1% or less, filled by pumped storage and thermal. The feeling that "Hokkaido is filled by pumped storage" is right as a statement about this seat. Tohoku and Hokuriku still have about a quarter open.

The combined product at the bottom is the largest seat, and it is nearly full in every area. Most of what sits there is thermal, and batteries are 9%.

Stack the three together and the open seats for a battery are first in its own frame (60–100% in Kansai, Chugoku, Hokkaido, Hokuriku and Shikoku, and 40–50% in Tohoku, Chubu and Kyushu), and then online in Tohoku and Hokuriku. What is left online is a displacement contest against thermal and pumped storage. The battery share of the combined product rose from 6% in April to 9% in August. The seats have not filled up; they are widening as they fill. That is the shape seen from the supply side.

3. What arrives is a part of what is applied for

← scroll horizontally →

Seeing a figure like "3,500 × 10,000 kW of battery connection applications" makes it look as though all of it will arrive. It will not. The supply side is built in 4 layers.

At the top, contract applications are about 3,527 × 10,000 kW (end of March 2026). Below that, the cumulative volume to the end of FY2029 that has got as far as a construction cost contribution contract is about 1,500 × 10,000 kW. The batteries that developers have entered in the Supply Plan as their own plans are 167 × 10,000 kW in FY2026 and 631 × 10,000 kW in FY2030. At the bottom, what is actually connected and running is 84 × 10,000 kW (end of March 2026).

The gap between applications and connected capacity is 42-fold. That gap is what is called grid-capacity hoarding (speculative applications), and countermeasures began in January 2026: requiring land rights with a connection study application, raising deposits, capping the number of applications per developer, and answering early once it is clear the construction cost will not work. Applications are being narrowed. What actually arrives is somewhere between the bottom 3 lines of Figure 4.

4. Where the high-confidence supply arrives — the Long-term Decarbonization Power Source Auction

Some batteries are all but certain to arrive: those awarded in the Long-term Decarbonization Power Source Auction (LDA). Across 3 rounds, batteries account for 76 projects and 3,713 MW (round 1: 1,092; round 2: 1,370; round 3: 1,251 MW, excluding pumped storage). They are committed to begin supplying by the end of the fiscal year 4 years after contracting, and as of the end of March 2026 no battery has exited.

Location can only be read from the project names in the list of awarded power sources, and many projects in rounds 2 and 3 carry only a company name, so 1,626 MW of the 3,713 MW (44%) is "unknown." Of what can be read, the weight is in the 2 eastern areas — Tohoku 775 MW and Hokkaido 457 MW — with Tokyo at zero. What arrives in those 2 areas heads for Tokyo over the interconnectors in the next section. Read the breakdown by area on the basis that 40% of it is unknown.

5. The size of the vessel, and the volume arriving

Figure 5 lays the supply-side lines (the Supply Plan and construction cost contribution contracts) over the size of the vessel (procurement volume). The vessel has 3 tiers: the band at the bottom is the battery seat (primary offline, about 50 × 10,000 kW), the dashed line in the middle is about 680 × 10,000 kW for primary, secondary ① and combined together, and the dotted line above is about 1,140 × 10,000 kW for all products.

If it arrives as the developers plan, FY2030 is still 631 × 10,000 kW and does not reach the middle line. If everything that has got as far as a construction cost contribution contract arrives, the middle line is passed in FY2027. Whichever line is followed, the dotted line above — the vessel as a whole — is not reached even in FY2030.

The seats fill in order. The battery seat (primary offline) this year into next, the openings in primary online (Tohoku and Hokuriku) after that, the combined product beyond. And as written on 47-2, procurement volume is defined to be increased once competition is confirmed. Bidding thickening is itself the condition for the vessel to widen.

6. Seats are tied together across areas

The balancing market procures area by area, but where there is room on the interconnector a battery in the neighboring area can clear. A battery in Tohoku can sit in a Tokyo room. The years in which the interconnectors widen are already set.

In 2027/11 Tohoku–Tokyo goes from 573 × 10,000 kW to 1,028 × 10,000 kW (in stages, depending on the state of generation). By FY2028 the frequency converter linking Tokyo and Chubu goes to 300 × 10,000 kW. In June 2030, the Kanmon interconnector goes from 300 × 10,000 kW to 600 × 10,000 kW. In FY2030, a new line between Chubu and Kansai.

The shortfall rates in Figure 6 are for the batteries-only room (primary offline). The LDA batteries concentrated in Tohoku and Hokkaido can reach Tokyo's large room from November 2027. Batteries in Kyushu and Chugoku reach Kansai from June 2030. Which areas are full and which are open gets repainted along this calendar. For secondary ① alone, cross-area procurement starts in FY2027, and when the interconnectors are congested the market splits.

7. On low-voltage batteries

It has been pointed out that low-voltage (under 50 kW) grid-scale batteries are increasing with entry into the balancing market in mind. But the minimum bid unit in the market is 1 MW, and a low-voltage battery can bid only once an aggregator has bundled about 20 of them in the same area. A rise in the number of projects is not in itself a rise in supply capacity.

Summary

• Batteries have 2 kinds of seat. Their own frame (about 500 MW) is 100% theirs; in the online seats they are 10–20%, and the rest is thermal and pumped storage.
• The batteries-only frame went from a shortfall rate of 67.9% in April–June to 53% in August. Tokyo is 80% full, while Kansai, Chugoku and Hokkaido are still largely open. The online seats are filled by thermal and pumped storage, but they are not partitioned by procurement volume and change hands on a cheap bid.
• What arrives is part of the applications (3,527 × 10,000 kW). On the developers' plans it is 631 × 10,000 kW in FY2030.
• The vessel as a whole (about 1,140 × 10,000 kW) is still wide even in FY2030. The seats move in the order primary offline → primary online → combined, and the interconnectors tie the areas together.

Where the price stops is on 47-4.

Reference table

Reference table 1 The primary offline seat and awards (August 2026, MW per 30-minute slot), with supply-side figures by area

Source: our calculation from EPRX trading results (2026/8 preliminary), Agency for Natural Resources and Energy (ANRE) Next-Generation Power Grid Working Group, 12th meeting, Document 2, and Organization for Cross-regional Coordination of Transmission Operators (OCCTO) Long-term Decarbonization Power Source Auction clearing results, attachments (rounds 1 to 3). LDA areas are aggregated only where they can be read from the project name; the 1,626 MW (44%) that cannot be read is "unknown"

Notes: the figures in Figures 1 to 3 are preliminary through August 2026. The shortfall rate is (procurement volume − awarded volume) ÷ procurement volume, counted area by area and slot by slot and then summed, with slots where the award exceeds the procurement volume treated as zero shortfall (rounding may differ from the nationwide aggregate in COLUMN 16). In Figure 3, "nearly full" is a shortfall rate of 20% or less, "half" is 20–50% and "open" is above 50%. The "volume under a construction cost contribution contract" in Figure 4 is read approximately off the bar chart in Attachment Figure 7-1 of the FY2026 Supply Plan compilation (as of the end of November 2025). The Supply Plan in Figure 5 joins the FY2025, FY2026 and FY2030 values with straight lines.

Sources

• EPRX, trading results (clearing data by area) (取引実績（エリア別約定データ）) https://www.eprx.or.jp/information/results.php
• ANRE, Next-Generation Power Grid Working Group (次世代電力系統WG), 12th meeting, Document 2, "Rapid grid connection of grid-scale batteries" (系統用蓄電池の迅速な系統連系) (6 August 2026), and 7th meeting, Document 1-1 (9 February 2026) https://www.meti.go.jp/shingikai/enecho/denryoku_gas/saisei_kano/smart_power_grid_wg/
• ANRE, Subcommittee on Mass Introduction of Renewable Energy and Next-Generation Power Networks (再生可能エネルギー大量導入・次世代電力ネットワーク小委員会), Document 1 (26 December 2025) https://www.meti.go.jp/shingikai/enecho/denryoku_gas/saisei_kano/pdf/078_01_00.pdf
• OCCTO, "Compilation of the FY2026 Supply Plan" (2026年度供給計画の取りまとめ) (30 March 2026), Table 3-1 and Attachment Figure 7-1 https://www.occto.or.jp/assets/various/kyoukei/torimatome/260330_kyokyukeikaku_torimatome/260330_kyokei_torimatome.pdf
• OCCTO, Long-term Decarbonization Power Source Auction clearing results (round 1, 26 April 2024; round 2, 28 April 2025; round 3, 13 May 2026) and the attachment "List of awarded power sources" (落札電源一覧). Round 3 attachment https://www.occto.or.jp/assets/various/capacity-market/jitsujukyukanren/2025_boshuyoukou_long/260513_longauction_youryouyakujokekka_kouhyou_besshi_ousatsu2025.pdf
• OCCTO, Long-term Decarbonization Power Source Auction capacity provision contract terms (容量確保契約約款) (as amended 3 September 2025), Articles 13 and 15, and "Status of capacity provision contracts" (容量確保契約の状況) (corrected version of 5 June 2026 and version of 26 August 2026)
• OCCTO, "Interconnector operational capacity, FY2026–2035," Attachment 1 (2026〜2035年度の連系線の運用容量) (1 March 2026) https://www.occto.or.jp/assets/renkeisenriyou/oshirase/2025/files/oshirase_1_2026-2035_unyouyouryou.pdf
• OCCTO, 117th Committee on Balancing Capacity and Supply-Demand Balance Evaluation, Document 4 (16 March 2026, cross-area procurement of secondary ①) https://www.occto.or.jp/assets/chousei_117_04.pdf
• EPRX, "Balancing market explanatory material, 2nd edition" (需給調整市場かいせつ資料 第2版) (13 March 2026, product requirements and minimum bid volume) https://www.eprx.or.jp/outline/docs/kaisetsu.pdf