A finite pipeline, and a fixed deadline.
The projects that will be built under the 30% credit are already known. What is not known is which of them will find capital in time.
The July 4, 2026 beginning-of-construction deadline has already done its work. Between mid-2024 and that date, U.S. developers safe harboured an estimated 216 to 240 GWdc of utility-scale solar — enough to satisfy forecast national installations through the end of the decade — against an intended project spend widely estimated in the range of $200 billion.
Every one of those projects now shares one hard boundary. To keep the 30% investment tax credit, it must be placed in service by December 31, 2030.
The pipeline is closed-end. No meaningful new supply of eligible utility-scale solar enters the market after July 4, 2026. Every project that will be built with a 30% credit already exists today, with a known sponsor, a known site and a known interconnection position. This is a finite universe that can be screened, not an open-ended market that must be prospected.
The binding constraint has flipped. Before July 4, the scarce commodity was eligibility. After July 4, it is construction capital. A large share of the safe-harboured pipeline sits with sponsors who spent their equity on development and safe-harbor procurement and cannot now reach notice to proceed. Those projects are eligible, sited, permitted — and unbuildable.
How the window closed
01| Date | Event | Effect |
|---|---|---|
| Jul 4, 2025 | OBBBA enacted | Accelerated termination of the §48E and §45Y credits for wind and solar |
| Aug 2025 | IRS Notice 2025‑42 | 5% cost safe harbor eliminated for solar above 1.5 MW; the Physical Work Test becomes the durable path |
| Jan 1, 2026 | FEOC and material-assistance rules effective | Supply-chain qualification tightened; guidance still developing |
| Jun 6, 2026 | D.C. District Court vacates Notice 2025‑42 | Restores the 5% method on paper; appeal expected. Uncertainty, not relief |
| Jul 4, 2026 | Statutory BOC deadline passes | The eligible pipeline is now fixed and finite |
| Dec 31, 2027 | Placed-in-service deadline for projects that missed BOC | Not a realistic path for utility-scale development |
| Dec 31, 2030 | Four-year continuity wall for the safe-harboured cohort | The hard end of the opportunity |
Working backwards from 2030
02The deadline is not the constraint people think it is. Counting backwards from it produces a much shorter window.
| Milestone | Timing |
|---|---|
| Placed-in-service deadline, safe-harboured cohort | December 31, 2030 |
| Latest practical mechanical completion, allowing for energization and interconnection | Mid-2030 |
| Typical utility-scale construction duration | 12 to 18 months |
| Latest practical full notice to proceed | Q1 to Q2 2029 |
| Window during which financings must close | Q4 2026 to Q1 2029 |
Roughly ten quarters.
Every one of them consumes continuity runway against a date that cannot be renewed, extended or repurchased.
What follows from it
03The cohort is oversubscribed to capital, not to demand. Safe-harboured volume exceeds forecast installations through 2030, which is an industry acknowledgement that a material portion of it will never be built.
Meanwhile load growth, data centres above all, has left the offtake environment for the projects that do get built stronger than it has been in a decade. Capital scarcity, not demand weakness, is what is stranding assets.
Conventional construction lenders are underwriting the same finite set of projects with the same diligence apparatus, and are capacity-constrained in doing so.
The projects most likely to strand are not the worst projects. They are the ones whose sponsors are smallest.