Chapter 1　Where did 20 GW come from?

Key points

• "About 20 GW by 2040" first appeared in a handout at the eighth meeting of the public–private council on 26 November 2024 and was fixed in the strategy two days later. The outline of 10 September had no field for the deployment volume; the draft of 8 October showed "●GW" with the number blanked outA
• The 20 GW figure comes from a demand estimate by a consulting firm commissioned by the government. The estimate multiplies the roof, wall and window area of each building type by an "installation coefficient", then by an adoption rate from a survey of about 500 respondents. The coefficients are based on expert interviews; the report shows no statistical or measured basis for themA
• The 20 GW is the demand expected if generation cost falls to 14 yen/kWh or below. In the same strategy, however, the secretariat estimates the 2040 generation cost at "the mid-15 yen range". The target's premise and the government's own estimate do not matchA
• The demand study assumes a 10-year equipment lifetime; the generation-cost review the following month assumes 20 years. Two government documents differ by a factor of two on the lifetimeA
• The report itself notes that the survey results "include respondent bias"A

1. The day the number appeared

The Public–Private Council on Next-Generation Solar Cells held its first meeting on 29 May 2024 and adopted the Next-Generation Solar Cell Strategy on 28 November 2024. Of the ten meetings, the eighth (26 November) was the first whose handouts carried a numerical domestic targetA · EV-045.

The handouts let us follow how the number arrived. The operating rules at the first meeting listed "setting deployment and price targets" as an agenda item, and the first meeting's paper put "whether the deployment target is at a level that can realistically be achieved" among the points at issueA · EV-037, EV-038. The outline at the sixth meeting (10 September) showed only the axes and legend of the demand-estimate chart, marked the values "under review and re-organisation (commissioned study)", and had no column for the deployment outlook in its roadmap tableA · EV-041. The draft at the seventh meeting (8 October) added the fields "domestic: about ●GW / overseas: about ●GW" and "●yen/kWh to ●yen/kWh", with the numbers blankedA · EV-042. Forty-nine days later, at the eighth meeting, those fields read "about 20 GW", "500 GW or more" and "10–14 yen/kWh or below"A · EV-045. The summary minutes of the eighth meeting contain no record of the secretariat explaining why that level was chosen, nor of alternative levels being compared. The only numerical remark recorded is that "setting concrete figures such as the deployment target of about 40 GW in 2040 … is important above all for giving predictability to companies and investors"A · EV-046.

The strategy sets the targets as follows. Domestic deployment of about 20 GW in 2040, and about 40 GW or more "if costs fall substantially". Overseas demand of 500 GW or more. Alongside these it defines a "self-sustaining level", at which perovskite solar stands without subsidy, of 10–14 yen/kWhA · EV-002.

2. How the estimate was made

How was 20 GW calculated? A note on p.22 of the strategy outlines the method: area by location × an installation coefficient × the adoption rate obtained from a survey. Roofs and walls are converted to output at 150 W per square metre, windows at 75 W, and walls and windows are multiplied by a factor of 0.1875A · EV-003.

The work was done under a 2024 contract from the Ministry of Economy, Trade and Industry (METI), the "Survey on Demand for Next-Generation Solar Cells". The contractor was Boston Consulting Group; the report is dated January 2025 and was posted in METI's library of commissioned reports in August 2025A · PS-120. The council's handout page does not link to itA · EV-029.

The report's estimate has three steps.

1. Find the total roof area in Japan for each building type: about 2.29 billion square metres for detached houses, about 360 million for manufacturing sites, and so on
2. Multiply by an "installation coefficient", the share on which perovskite cells could additionally be installed: 10% for detached houses, 37.5% for manufacturing sites, 34% for ordinary warehouses, 40% for gymnasiums, about 15% for primary schools
3. For each generation-cost level, multiply by the share of about 500 survey respondents who said they would install at that price

The report states that the coefficients are based on "expert interviews". It does not say which experts were asked, what they were asked, or why detached houses get 10% and manufacturing sites 37.5%A · EV-008. On the survey, the report itself notes that the results "include respondent bias". Surveys of purchase intent are known to over-represent interested respondents and to produce higher rates than actual purchasesC · CL-005.

The resulting demand curve shows about 21 GW at a generation cost of 16 yen/kWh, about 26 GW at 14 yen, about 37 GW at 10 yen and about 44 GW at 7 yenA · EV-002. The 20 GW target corresponds roughly to the value at about 16 yen.

3. An estimate that misses the self-sustaining level

Now look at another part of the strategy. It sets the self-sustaining level, at which perovskite is installed without subsidy, at 10–14 yen/kWh and explains the 20 GW as demand at that level. On p.22 of the same document, the secretariat estimates the generation cost in 2040 at "the mid-15 yen range"A · EV-004.

So the premise of the target (14 yen or below) and the forecast of the body that set the target (mid-15 yen range) do not agree within the same document. Reading the demand curve at the mid-15 yen range gives a value somewhat below 20 GW.

The strategy addresses the gap with a learning curve: when cumulative deployment reaches 10 GW, equipment cost falls to 140,000 yen/kW; at just over 80 GW it falls to 60,000 yen/kW, at which point generation cost reaches 10 yenA · EV-005. The domestic target of 20 GW alone does not reach 80 GW, so arriving at 10 yen depends on Japanese companies capturing part of the 500 GW overseas marketA · CL-013.

In February 2025 the Agency for Natural Resources and Energy's Working Group on Generation Cost Verification published a reference value for perovskite solar in 2040 of 15.3 yen/kWh (16.4 yen including policy expenses). In the same paper, utility-scale crystalline silicon in 2040 is 7.9 yen (8.4 yen) and residential silicon 7.6–10.4 yenA · EV-011, EV-012. On the government's own outlook, perovskite in 2040 costs about twice as much as silicon.

4. A 10-year lifetime and a 20-year lifetime

Among the assumptions behind a generation cost, how many years the equipment lasts is one of the most powerful.

The demand-study report assumes a 10-year lifetime. Its other inputs are an initial cost of about 251,000 yen/kW, operating cost of 5,000 yen/kW a year, a 3% discount rate, 15% conversion efficiency, and capacity factors of 14.5% on roofs, 8.4% on walls and 7.5% on windows. On these assumptions the present generation cost is 28.0 yen on roofs, 42.9 yen on walls and 57.5 yen/kWh on windowsA · EV-007. The report describes these inputs as "estimates based on expert interviews".

The reference value from the generation-cost working group a month later assumes a 20-year lifetime, the median of hearings with six companiesA · EV-011.

Going from 10 to 20 years roughly halves the cost per kWh for the same initial outlay. The 15.3 yen reference value for 2040 rests on 20 years. Sekisui Chemical, the first company to bring a product to market, states a durability of 10 years for the product it launched in March 2026B · EV-025. Today's product lifetime is close to the demand study's assumption and half the cost review's.

Which of the two assumptions was used for the policy decision, and how the difference was reconciled, is not explained in the public recordA · CL-004.

5. What was said where the target was set

The council's summary minutes are public but do not name speakers. Calls for a target appear early. At the second meeting (19 June) someone said that "predictability of long-term demand is essential; the government should set a long-term target", and at the seventh that "government and industry need to set an ambitious target so that as much deployment as possible goes ahead"A · EV-039, EV-043. There were also cautious remarks on cost: at the sixth meeting, "for 2040 we should consider a cost-reduction target more demanding than the current estimate"; at the seventh, "the Green Innovation Fund's 14 yen/kWh holds only under certain conditions"A · EV-028, EV-043. Who said what on what grounds, and how the secretariat responded, cannot be read from the minutes.

The contractor that produced the demand estimate presented at the third meeting (26 July 2024) and the fifth (26 August 2024). Neither presentation is posted on the handout pageA · EV-029. Within the public record as it stands, we found no record of the numerical target being deliberated in open sessionB · CL-001.

6. What we know so far, and what we do not

What we know

• The 20 GW figure is demand estimated from expert-interview coefficients and an intent survey, and assumes a cost reduction to 14 yen or below
• The strategy's own 2040 cost estimate does not reach that premise
• The September outline had no deployment field; the October draft had a blanked field; the 26 November draft carried the number. Neither the handouts nor the minutes give the reason for the level chosen
• The expected replacement of existing capacity with tandem cells changed from 49.1 GW (2035–2040) in the outline and the October draft to 66.8 GW (2032–2040, including 17.6 GW from efficiency gains) in the November draft. The difference is explained by the change in period and baseA · CL-020
• The demand study and the generation-cost review assume lifetimes of 10 and 20 years respectively
• The calculation sheets, the basis of the coefficients, and the survey questions and tabulations have not been published

What we do not know (being checked)

• How alternative target levels were compared between the seventh and eighth meetings, and who chose about 20 GW
• Where the installation coefficients come from
• What the 2040 generation cost becomes if recalculated with a 10-year lifetime

We are pursuing these through administrative document disclosure requests and through the open questions sent on 29 October 2026 (reply deadline 28 November). This chapter will be updated as answers arrive.

Sources

Unless stated otherwise, costs are in yen per kWh. URLs and retrieval dates for every document are in the evidence register and primary-source list (Japanese). A meeting-by-meeting comparison of the handouts and minutes of meetings 1–8 is in the source file 20gw-trace-kanmin-01-08.md (Japanese).

Pages in this series

• Series home
• Chapter 1 Where did 20 GW come from?
• Chapter 2 What the council discussed (Japanese)
• Chapter 3 Who decided, who received (Japanese)
• Chapter 4 Comparing generation costs
• Chapter 5 Demand created by subsidies (Japanese)
• Chapter 6 How the press reported it (Japanese)
• Chapter 7 What is happening abroad
• Chapter 8 What the same money could have bought
• Chapter 9 Open questions to the government (Japanese)
• Policy timeline (Japanese)
• Public support register (Japanese)
• Evidence register and primary sources (Japanese; CSV)

Related pages

• Rethinking perovskite solar from zero (Japanese) — Checking the technical, economic and institutional premises one by one.
• Reviewing MRI's perovskite industrial-policy column (Japanese) — The 'Team Japan' argument against primary sources.
• JC-STAR and grid cyber-resilience: an effectiveness check — An earlier I-S3 investigation using the same evidence-register method.

Citation and reuse

The CSV and Markdown source files (index) are released under CC BY 4.0. You may copy, adapt and use them commercially if you credit I-S3 Co., Ltd. and link to this page. Quotations from the text with attribution are welcome.

Anyone named in this series who wishes to rebut or correct a statement may write to our contact form or s_masuda@i-s3.com; we publish rebuttals at equal length.

Suggested citation

Masuda, S. (2026). "Chapter 1　Where did 20 GW come from?". I-S3 Co., Ltd., published 2026-10-29, https://www.i-s3.com/perovskite-policy-investigation-ch1-20gw-origin-en.html