SOLAR · PHOTOVOLTAIC
Solar & photovoltaic business plan: the complete guide
Practical guide · ~11 min read
The business plan of a solar project (or photovoltaic plant) turns its economics into numbers: how much it generates, how much it sells, what it costs to build and operate, how it is financed, and what it returns to the shareholder. Done well, it is the document that convinces a bank or an investment committee. Here is the method, brick by brick.
1. Yield: P50, P90 and degradation
It all starts with the energy produced. You estimate an annual yield (in MWh) from the site irradiation, the installed capacity (MWp) and the performance ratio. Two scenarios stand out:
- P50: median generation (50% chance of being exceeded) — the base case of the business plan.
- P90: a prudent generation level (90% chance of being exceeded) — the one the bank uses to size the debt.
You also apply an annual degradation of the modules (often 0.4 to 0.7%/yr) over the whole life (30 to 40 years).
2. Revenue: feed-in, PPA or market (merchant)
Revenue depends on how the electricity is valued:
- Feed-in / contract for difference: a tariff awarded through a tender, indexed, over the long term — highly visible revenue.
- PPA (Power Purchase Agreement): a bilateral contract with an offtaker, at a fixed or indexed price.
- Merchant: sale at market price (spot / forward price curve) — more potential upside, more risk.
Many projects combine a contracted period (feed-in/PPA) then a merchant period at end of life. The business plan must model that switch — and that is often where spreadsheets go wrong.
3. CAPEX and OPEX
The CAPEX (upfront investment) breaks down into several items. Modelling it line by line, and staging it across construction, changes the debt drawdown schedule and therefore the equity required.
| CAPEX item | Share of budget |
|---|---|
| Photovoltaic modules | High |
| Inverters, transformers, substations | Medium |
| Mounting structures / trackers | Medium |
| Civil works & cabling | Medium |
| Grid connection | Variable (very site-dependent) |
| Development, engineering, fees | Low to medium |
CAPEX is often expressed in €/kWp to compare projects, and has fallen sharply for ground-mount PV. The OPEX (annual operating costs) covers operation & maintenance (O&M), insurance, land lease, administration and inverter replacement. It is expressed in €/kWp/yr and grows with inflation: an underestimated OPEX inflates the margin over 30 years.
4. Financing: gearing and sculpted debt
A solar project is financed with senior debt + equity. Two sizing logics coexist:
- By gearing: a fixed percentage of CAPEX (e.g. 80%).
- By DSCR (sculpted debt): debt is calibrated to meet a target coverage ratio every year.
The DSCR (Debt Service Coverage Ratio) is the metric the bank watches: it measures the ability of cash flow to cover debt service.
5. Taxes
A credible solar BP includes the applicable corporate and production taxes for the jurisdiction (in France, corporate income tax, plus local and grid taxes such as CVAE and IFER). Forgetting a production tax, for example, overstates the returns.
6. Return metrics
From there you derive EBITDA, cash flows, then the key metrics: IRR (project and equity), NPV, DSCR and LCOE. They are detailed in our guide on IRR, NPV, DSCR, LCOE.
7. Common mistakes
- Using P50 to size the debt (the bank thinks in P90).
- Forgetting module degradation or the end of contracts (merchant switch).
- Neglecting taxes and working capital.
- A "non-auditable" spreadsheet where one wrong cell breaks the whole result.
8. Sensitivity analysis and scenarios
A business plan is never a single number, but a range. The variables that move a solar project's returns the most are usually: the power price over the merchant period, the yield (P50 vs P90), the debt rate and the CAPEX. Testing each one in isolation (sensitivity), then in combination (optimistic / base / stress scenarios), reveals the project's real robustness — exactly what a credit committee examines.
9. Estimate returns quickly
For a first order of magnitude before building the full model, use our solar ROI calculator: it estimates IRR, NPV, LCOE and DSCR live from capacity, price, CAPEX, OPEX and debt structure. For a bankable, consolidated business plan, see the solar business plan software.
FAQ
How do you build the business plan of a solar project?
Start from the yield (P50/P90), value the revenue based on the contract (feed-in, PPA or market), deduct OPEX and taxes to get EBITDA, then size the debt (by gearing or DSCR) and compute the flows to the shareholder. The key metrics are IRR, NPV, DSCR and LCOE.
What IRR should you target for a photovoltaic project?
It depends on risk and contract, but a ground-mount PV project secured by a long contract often targets a project (unlevered) IRR of roughly 5 to 8%, with an equity IRR lifted by debt leverage. The right benchmark remains the investor's cost of capital.
How long is a solar business plan?
You typically model the asset's operating life (often 30 to 40 years for PV), with a contracted phase (feed-in/PPA) followed by a market (merchant) period, and debt amortised over 15 to 20 years.
Why does the bank work with P90 rather than P50?
The bank sizes debt on a prudent scenario to make sure it gets repaid even in a poor year. The P90 (generation reached 9 years out of 10) gives that safety margin, while the P50 (median generation) is the base case for the returns shown to the shareholder.
How do you reduce the CAPEX per kWp of a solar project?
CAPEX per kWp falls with project size (economies of scale), component choice, grid-connection optimisation (often the most site-dependent item) and careful engineering. Comparing projects in €/kWp is the right reflex, but a very low CAPEX can hide higher OPEX or risk.
PUT IT INTO PRACTICE
Build this business plan in Wattvalio.
P50/P90 generation, offtake contracts (feed-in, PPA, merchant), debt, taxes, NAV valuation — from a single asset to the consolidated portfolio, with bankable figures.
Or estimate returns with the free calculator →Read also: IRR, NPV, DSCR, LCOE · ROI calculator · Solar BP software