The Incentive Screen:
What Current Law Actually Provides

The least reliable number in most power proposals is not the equipment price or the output estimate. It is the incentive line, because it rests on law that has been amended twice in four years, on qualifications nobody has documented, and on program budgets that close without asking permission.

Section 01Where incentive lines go wrong

Every proposal for on-site energy arrives with an incentive line, and that line does more work than any other number on the page. It converts a capital cost the buyer could verify with three phone calls into a net cost the buyer cannot verify at all without reading tax law. That asymmetry is why the line deserves its own discipline, separate from the engineering and separate from the financial model.

In study work the failures repeat in three recognizable forms. The first is the stale statute: a model built under the law as it stood when the template was written, presented under the law as it stands today. The federal clean-energy credit regime was restructured in 2022 and amended again in July 2025; a spreadsheet that predates either event is citing a code section that no longer says what the cell comment thinks it says. The second is the assumed qualification: the proposal books the headline credit plus every statutory bonus at maximum value, with no evidence that the specific project satisfies any of the tests those bonuses require. The third is program optimism: state and utility incentives carried as standing entitlements when they are, in fact, appropriations with budgets, waitlists, administrators, and closure dates.

The corrective is not sophistication. It is a single page we call the incentive screen: every claimed incentive on its own line, and for each line three entries. The statutory or program basis, by citation. The as-of date on which that basis was last verified. And the qualification status of this specific project against that specific line. The screen does not compute value; the financial model does that. The screen establishes which lines have earned the right to be in the model at all.

Section 02The federal baseline: 30 percent, for qualifying property

Start with what current law actually provides, as of this paper's date. The federal investment tax credit for clean-electricity property stands at 30 percent of eligible basis for qualifying property. Two code sections govern: section 48, the legacy energy credit that continues to apply to projects that began construction under its rules, and section 48E, the technology-neutral clean electricity investment credit that governs the current era.^1,2 The July 2025 amendments, enacted as Public Law 119-21, rewrote significant parts of the regime, which means any citation that predates them is provisional until rechecked.

Every word in the phrase "for qualifying property" is a test. Qualification turns on three things. Property class: the statute defines which facility types and which storage property are eligible at all, and conventional combustion generation is generally not among them. Construction timing: which section applies, and whether the project fits inside its eligibility window, depends on when construction begins under rules the Internal Revenue Service defines with precision. And statutory conditions: the 30 percent figure is the statute's alternative rate, available to facilities below a one-megawatt output threshold, to projects grandfathered by early construction dates, and to projects that document compliance with prevailing-wage and apprenticeship requirements; the base rate without any of those conditions is 6 percent.¹ A proposal that writes "30 percent" without stating which condition the project satisfies has not stated a fact. It has stated an intention.

The July 2025 amendments sharpened the timing stakes. Wind and solar facilities now face a placed-in-service deadline at the end of 2027 unless construction began within twelve months of enactment. Fuel cell property beginning construction after 2025 receives a flat 30 percent that sits outside the emissions-rate test and that the statute itself bars from being increased or adjusted by any other provision of the section. New restrictions tied to foreign-entity involvement in supply chains and ownership attach their own documentation burden.¹ None of this fits in a footnote, which is the point. The statutory-basis column exists because the statute moves.

Section 03Adders exist in statute; a screen assumes none

The bonus adders are where proposals inflate. Current law provides three principal increases above the headline rate, and each is real money for projects that genuinely qualify. The energy community increase adds 10 percentage points for projects placed in service within statutorily defined categories: certain brownfield sites, statistical areas with fossil-employment history and elevated unemployment, and communities tied to retired coal infrastructure, all mapped in agency guidance that shifts as designations update.¹ The domestic content increase adds another 10 points for projects that certify statutorily specified shares of domestically produced steel, iron, and manufactured products; the required manufactured-product share steps up by construction-start date, from 40 percent for the earliest cohort to 55 percent for construction beginning after 2026, and the certification rests on supplier data most vendors have never been asked to produce.¹

The third increase is different in kind. The low-income communities bonus adds 10 or 20 percentage points for facilities below a statutory size limit, but it is not claimed; it is awarded. The program allocates from an annual capacity limitation of 1.8 gigawatts, by application, in defined windows.³ For the 2026 program year, the application window opened February 2 and closed August 7, two days before this paper's publication date.³ A proposal that books that bonus for a project that never applied, in a year whose window has closed, is not describing an incentive. It is describing a wish.

The screen's rule for adders is uniform: booked at zero until this project's qualification is documented, in writing, against the current version of the test. Run the economics with and without every adder, and insist that the project clear the bar unadorned or carry an explicit, dated qualification plan for each point of uplift it claims. This is not cynicism about the adders, which are substantial and were enacted to be used. It is sequencing. Qualify first, then book. The marketing genre runs in the other order.

An incentive that has not been qualified is not a discount. It is a hope with a percentage sign attached.

Section 04Timing is a qualification, not a footnote

The same equipment at the same site can carry materially different credit value depending on when construction begins under federal rules, which makes the calendar a qualification of its own. In August 2025 the Internal Revenue Service issued Notice 2025-42, tightening how wind and solar projects establish the beginning of construction: demonstrating physical work of a significant nature is now the primary route, and the five-percent expenditure safe harbor survives only for small solar facilities.⁴ Rules of this kind are exactly why the screen carries an as-of date on every line. A begin-construction strategy that was standard practice in June can be a diligence finding by September.

Run the screen identically across candidate technologies and the honest differences surface on their own. Solar arrives with the deepest incentive experience and, where timely, the fullest adder menu; it now also carries calendar risk that belongs in its underwriting, because eligibility windows are statutory and construction schedules are not. Storage continues under the technology-neutral credit with a longer statutory runway, subject to its own conditions and to the same foreign-entity restrictions. Fuel cells, under the 2025 amendments, carry a flat 30 percent with no emissions test and no adders: simpler to underwrite, capped by design, and no credit hedges the price of the gas beneath them. Combustion engines and turbines generally sit outside the clean-electricity credit entirely, which is worth saying plainly: their case must close without a credit line, and on capital cost and delivery speed it sometimes does. A credit is a reason to sharpen the comparison. It is never a reason to pick the machine before the site has spoken.

Section 05Program incentives are perishable: the California example

State and utility programs fail differently from statutes: they run out. California's Self-Generation Incentive Program was for many years the state's principal incentive for on-site generation and storage. On December 15, 2025, the program administrator announced that the ratepayer incentive budgets, the generation budget among them, would close to new applications on December 31, 2025, and they did.^5,6 A proposal dated 2026 that carries a generation line from that program is stale on its face, whatever the rest of its analysis says.

The lesson generalizes. A program incentive is an appropriation with an administrator, a budget, a waitlist, and a closure date, and every one of those is a fact with a timestamp. The screen therefore treats program lines more strictly than statutory ones: administrator-verified status, on a stated date, within days of the decision the number is feeding rather than the quarter the template was built. Closure is also information in its own right. As program budgets close, the arithmetic shifts toward the statutory instruments and, more importantly, toward the version of the project that works with no incentives at all.

Section 06The screen: three columns and six rules

The screen itself is deliberately plain. One page. Every claimed incentive on its own line, three columns per line. Column one is the statutory or program basis: the code section or program handbook the line rests on, by citation. A line that cannot name its section is not a line; it is a rumor. Column two is the as-of date: the date the basis was last verified against current law or a live administrator source, which is never the date the template was built. Column three is the qualification status of this project against this line, in one of three honest states: qualified, with the evidence named; not qualified, booked at zero, with the next verification step and its owner; or closed, removed from the model entirely.

Laid out for a hypothetical California project in August 2026, a screen looks like this. The rows are illustrative, not advice.

Six rules govern how the screen is used:

1. Never net an unqualified incentive into a headline number. Present the gross cost, then the qualified value, separately. A single blended "net cost" hides exactly the assumption that needs daylight.
2. Run the economics with and without each line. If the recommendation flips on a line whose status is "not qualified," the decision is not ready, and the screen has just told you what the next unit of work is.
3. Give every line an owner and a dated next action. "Confirm energy community status" is not a status. "Location test against current designations, by counsel, by September 15" is.
4. Put tax positions in front of qualified tax professionals before capital moves. The screen organizes the questions for counsel. It does not replace counsel, and nothing in this paper is tax advice.
5. For tax-exempt and public owners, stop earlier. How any federal credit becomes value in a particular owner's hands is a matter of ownership structure and tax law that belongs with qualified tax counsel, and public-agency procurement and finance routes are agency-specific and counsel-specific besides. A screen prepared for such an owner says so on its face and books nothing until counsel has spoken.
6. Date the screen and re-run it at every capital gate. Two of its three columns are perishable by construction. A screen that has not been refreshed since the last statutory amendment is a historical document.

Section 07The six-question audit

Any owner, board member, or finance officer can audit the incentive content of any proposal in fifteen minutes with six questions. No tax background is required to ask them; a defensible proposal answers all six without flinching.

1. Does every incentive line cite its statutory or program basis?A code section or a program handbook, by name. No section, no line.
2. Does every line carry an as-of date after the most recent amendment?For the federal credit regime, that means after July 2025. An earlier date means the line was written under a different law.
3. Is qualification status stated per line, with evidence for anything booked above zero?Qualified, not qualified, or closed. "Eligible" without documentation is a synonym for not qualified.
4. Are all bonus adders booked at zero unless this project's qualification is documented?Stacked maximums are a marketing convention, not an underwriting one. Each adder is its own test with its own evidence file.
5. Do the economics run with and without incentives, and does the recommendation survive the without case?If not, the proposal should say so plainly and show the dated qualification plan that closes the gap.
6. Has a qualified tax professional reviewed the tax positions?And if the owner is tax-exempt or public: is qualified tax counsel engaged before any credit value is booked at all?

Section 08What the screen protects

The screen's output is modest: one page. What it protects is not modest. It protects the board from approving a net cost built on an unqualified stack. It protects the project from a diligence process that reprices it at the worst possible moment, because lender and investor diligence will run exactly this screen, on exactly these columns, with less charity. And it protects the decision itself, because a project that pencils only with every adder booked at maximum is not a project with upside. It is a risk with good formatting.

Current law provides 30 percent for qualifying property, adders that must be earned one test at a time, and program budgets that close. That is not a discouraging picture; it is simply the picture, and every number in it carries a date. Write the screen, date it, and make every line prove itself. The alternative is to discover, late and in front of the people funding the project, which lines could not.