What a Project-Finance Model Actually Needs

Financing a capital-heavy asset — a solar plant, a facility, an infrastructure project — is a different discipline from raising equity for a startup. The lender isn’t betting on upside; they’re lending against a ring-fenced set of cash flows and asking one question over and over: will this service its debt, even if things go a little wrong? The model has to answer that on every page. Here’s what a real project-finance model requires.

1. Revenue that’s contracted, not assumed

Equity investors accept revenue projections; project-finance lenders want revenue that’s contracted. The gold standard is a long-term offtake agreement — a power-purchase agreement (PPA), a lease, a take-or-pay contract — that fixes price and volume for years. The model builds revenue from those contracts, and it accounts for the buyer’s creditworthiness (a contract is only as good as the counterparty behind it).

For BRE’s 932 MW solar project, revenue came entirely from two long-term PPAs at a fixed price with annual escalation, covering 100% of net generation — insulating the model from merchant power-price risk. That’s what “bankable revenue” looks like.

2. Generation (or output) modeled net, not peak

The revenue line rests on how much the asset actually produces. A credible model uses net output — after losses, degradation, and downtime — not the nameplate best case. Modeling peak output is the fastest way to lose a lender’s trust, because they’ve seen it a hundred times and they’ll adjust it down themselves.

3. Debt sizing and amortization

The model determines how much debt the cash flows can support and lays out the repayment schedule. Key structural choices show up here: the interest rate, the tenor, whether interest is capitalized during construction, and when debt service begins. BRE’s model carried a ~$380M fully-amortizing senior facility at a fixed 7%, with a one-year construction period during which interest capitalized and repayment began only in the first full operating year — aligning payments with stabilized revenue.

4. The DSCR — across the whole tenor

The debt-service coverage ratio is the number the deal turns on: cash available to service debt, divided by debt service due. Lenders want it comfortably above 1.0x — typically a 1.20–1.30x minimum — and they want to see it every year, not on average. The figure that matters is the lowest it ever reaches. BRE’s model held a DSCR climbing from 1.46x to 1.89x, never dropping below 1.46x. (We wrote a whole piece on how lenders read a DSCR.)

5. A cash-flow waterfall

Project finance runs on a waterfall — the strict order in which cash is applied each period: operating costs first, then debt service, then reserves, then whatever’s left to equity. The model has to show that waterfall, because it’s the mechanism that protects the lender: debt gets paid before anyone takes a distribution.

6. Sensitivities and a downside case

No project-finance model is complete without stress cases. What happens to the DSCR if output is lower, costs are higher, or — for BRE — a tax incentive doesn’t apply? The model tested CAPEX with and without the local incentive and ran a curtailment sensitivity, so the coverage story held even in the downside. A lender doesn’t fund the base case; they fund the case that survives the downside.

The bottom line

A project-finance model is the deal. Contracted revenue, conservative output, sized debt, a DSCR that holds across the tenor, a cash-flow waterfall, and real sensitivities — miss any one and a serious lender walks. That’s the standard our financial modeling practice builds to, from mid-market equipment financing up to nine-figure project finance. If you’re financing a capital-heavy asset, book a 30-minute call and we’ll build the model that makes it bankable.