When evaluating a solar or wind project, one of the first questions investors and lenders ask is: how much energy will the project actually produce?
The answer is rarely a single number. Instead, renewable energy projects are typically assessed using probabilistic energy yield estimates such as P50, P75, P90 and sometimes P95 or P99.
For project finance, understanding the difference between P50 and P90 is particularly important because the production assumption flows directly into the financing structure:
A project that looks highly attractive using P50 assumptions can support significantly less debt when evaluated against a more conservative production case.
This article explains what P50 and P90 mean, why lenders care about them, and how they should be reflected in a renewable energy financial model.
P50 and P90 are probabilistic energy yield estimates that express the probability that actual annual production will exceed a particular level.
P50 is the production level that has a 50% probability of being exceeded. In simple terms, it represents the central or median production estimate.
If a solar project has a P50 production forecast of 150,000 MWh/year, this means there is approximately a 50% probability that actual annual production will exceed 150,000 MWh, and a 50% probability that it will be below that level.
P50 is therefore commonly used as an expected or base-case production assumption for project economics and investment analysis.
P90 is a more conservative production estimate. A P90 value has a 90% probability of being exceeded. For example, assume the same project has a P50 of 150,000 MWh and a P90 of 135,000 MWh.
The interpretation is approximately: there is a 90% probability that annual production will be at least 135,000 MWh.
Importantly, P90 does not mean that the project will produce 90% of P50. The "90" refers to the probability of exceedance, not to a percentage of P50 production. That distinction is one of the most common sources of confusion when discussing renewable energy yield assessments.
| P50 | P90 | |
|---|---|---|
| Probability of exceedance | 50% | 90% |
| Production estimate | Higher | Lower |
| Risk perspective | Expected / central case | Conservative case |
| Typical use | Investment returns, valuation | Debt and downside analysis |
| Revenue | Higher | Lower |
| CFADS | Higher | Lower |
| Debt capacity | Higher | Lower |
| Equity requirement | Lower | Higher |
The exact use of each case varies by market, technology, lender and financing structure. P90 is widely used in renewable project finance, but it should not be treated as a universal rule that every lender sizes every transaction on P90.
The reason is simple: energy production drives project cash flow. Consider a hypothetical solar project with 100 MW capacity, a €60/MWh PPA price, P50 generation of 150,000 MWh and P90 generation of 135,000 MWh.
At P50:
At P90:
That's a €900,000 difference in annual revenue before considering operating costs and other project-specific adjustments. The impact doesn't stop at revenue — it flows through the entire financial model.
This is where the topic becomes particularly important for financial modelling. A simplified project finance model looks like this:
Therefore, changing the production assumption from P50 to P90 can have a significant impact on the financing structure. For example, assume P50 CFADS of €7.0m and P90 CFADS of €6.2m, with a target DSCR of 1.30x.
The maximum annual debt service supported by the P50 case would be approximately:
Under P90:
The project therefore supports materially less debt under the conservative production case. This is why P50/P90 assumptions cannot simply sit in a technical appendix — they can directly influence the capital structure of the project.
One of the most important relationships in project finance is:
Rearranging:
Therefore, if CFADS decreases because the production assumption is more conservative, the debt service that the project can support also decreases — which ultimately reduces debt capacity. The underlying principle is well established in renewable project finance: lenders assess the cash flow available for debt service and apply a required coverage ratio when determining debt capacity.
| P50 | P90 | |
|---|---|---|
| CFADS | €7.0m | €6.2m |
| Target DSCR | 1.30x | 1.30x |
| Maximum debt service | €5.38m | €4.77m |
The difference may look relatively small at the annual cash-flow level. But when this is translated into a 15–20 year debt facility, the difference in total debt capacity can become substantial.
Imagine a lender sizes a loan entirely against the P50 production forecast. P50 represents the median expected production — which means, by definition, there is roughly a 50% probability that actual annual production will be below the P50 estimate.
From a lender's perspective, that isn't necessarily sufficient protection for a long-term loan. The lender wants to know: what happens to debt service coverage if the project performs below its central forecast?
P90 provides one way of incorporating production uncertainty into the financing analysis. Resource assessment literature, including guidance from NREL, notes that P50 and P90 assessments are used to quantify production risk relevant to a project's ability to service debt.
This is an important point. You should not think of P50 as the optimistic case and P90 as the bad case. Instead, the two cases answer different questions.
Represents the central production expectation. It is particularly useful for project valuation, investment returns, base-case economics, equity IRR, NPV and business planning.
Represents a more conservative production outcome. It is particularly useful for debt sizing, DSCR analysis, downside analysis, financing discussions and assessing debt-service resilience.
P50 asks: what do we reasonably expect the project to produce? P90 asks: what happens to the project's financial structure if production is materially lower than the central estimate?
There is an interesting consequence for equity investors. Suppose debt is sized conservatively using a downside production case while equity returns are evaluated using the expected P50 case.
The difference between those cases can create significant equity upside: P90 determines conservative debt capacity, while P50 determines expected project cash flow.
If the project performs around its P50 expectation, the project may generate more cash than required to service the conservatively sized debt. That additional cash ultimately benefits equity, subject to the project's distribution restrictions and financing documents.
This is one reason why understanding the relationship between production assumptions, debt sizing and equity returns is so important when building a project finance model.
The concept applies to both technologies, but the underlying uncertainty can differ.
Solar production is influenced by factors such as solar irradiation, weather variability, system losses, soiling, availability, degradation, curtailment and modelling uncertainty.
Wind production is influenced by wind resource, long-term wind variability, measurement uncertainty, wake effects, turbine availability, losses, curtailment and modelling assumptions.
The resulting P50/P90 relationship therefore needs to come from a project-specific energy yield assessment, rather than simply applying a generic percentage reduction.
This deserves emphasis. A common shortcut is "P90 = 90% × P50." That is incorrect.
P90 is a probability-based estimate, not a fixed percentage reduction from P50. The difference between P50 and P90 depends on the project's underlying uncertainty distribution, which can be influenced by:
Therefore, the P90 value should come from the appropriate energy yield assessment rather than a generic financial modelling assumption.
A professional renewable energy financial model should make it possible to understand the financial consequences of different production cases. For example, the model might include production scenarios (P50, P75, P90, and P95/P99 where relevant) and financial outputs (revenue, EBITDA, CFADS, DSCR, debt capacity, equity contribution, project IRR, equity IRR).
This allows an investor or lender to move from:
That's much more useful than simply displaying a P50 and P90 number on a technical report.
Consider a hypothetical 100 MW solar project with a €60/MWh PPA price, P50 generation of 150,000 MWh, P90 generation of 135,000 MWh, OPEX of €1.5m and a target DSCR of 1.30x.
P50:
P90:
The P90 case therefore supports approximately 12% less annual debt service than the P50 case in this simplified example. And because debt service is sustained over many years, that difference can have a significant effect on total debt capacity.
This is an illustrative example only; actual debt sizing depends on the financing structure, timing of cash flows, interest rate, tenor, sculpting methodology, reserve requirements and applicable lender criteria.
The gap between P50 and P90 can itself be informative. A large gap indicates greater uncertainty around expected production, which can lead to:
Conversely, a relatively narrow distribution may support greater confidence in the project's production forecast. However, there is no universal "acceptable" P50-P90 spread — it needs to be assessed in the context of the specific technology, resource assessment, geography and financing structure.
P50 and P90 are not simply two alternative production forecasts. They represent different ways of looking at project risk. The most important relationship to remember is:
And in a project finance model:
This is why a renewable energy financial model needs to connect energy yield assumptions directly to the financing structure. A model that calculates an equity IRR from P50 production but never tests debt service under a conservative production case can give a misleading picture of the project's financing risk.
A professional project finance model should ideally allow the user to switch between production cases (P50, P75, P90) without manually changing formulas. The selected case should flow automatically through generation, revenue, CFADS, DSCR, debt sizing and equity returns.
This makes it possible to quickly answer questions such as:
This is exactly the type of functionality that turns a basic renewable energy spreadsheet into a proper project finance model.
At Project Finance Templates, we build professional financial models for solar and wind projects with the core project finance mechanics required to analyse project economics and financing.
Our models include functionality for areas such as energy production, revenue, CAPEX and OPEX, CFADS, debt sizing, DSCR, debt repayment, equity returns, IRR and sensitivity analysis.
Explore our renewable energy financial models — with P50/P75/P90 production cases flowing straight through to debt sizing, DSCR and equity returns.