Renewable Energy · Intermediate level · 3 min read ·
What are P50 and P90 scenarios?
P50 and P90 are two readings of the same production projection. P50 is the value expected to be exceeded in half of the years: the central scenario. P90 is the conservative one, expected to be exceeded 9 years out of 10. The gap between them shows how uncertain the projection is.
In 30 seconds
- P50 is exceeded in half the scenarios; P90 in nine out of ten.
- P90 is always lower than P50: it's the prudent reading.
- More uncertainty in the sun or the data pushes the two values further apart.
Why should you care?
A production projection isn't a fixed number: the sun changes from year to year. P50 and P90 show you the center and a reasonable floor.
- You know whether a production figure is the central or the prudent one.
- You see how far a bad year's energy could move from a normal one.
- You compare projects knowing which scenario each one uses.
Think of it as your commute to work
Half of the days you get there in 30 minutes or less: that's your P50. But for an important meeting you leave 40 minutes ahead, because that gets you there on time 9 days out of 10: that's your P90. Same trip, two ways to plan.
How does P50 and P90 scenarios work?
- 1
Production is simulated
Software such as PVsyst uses many years of sun data, the design and the losses to estimate annual energy.
- 2
The P50 comes out
It's the central value: half the years are expected to produce more and half less.
- 3
Uncertainty is measured
How much the sun varies between years is combined with the margin of error in the data and the model.
- 4
The P90 is calculated
That uncertainty is taken off the P50 to reach a value exceeded 9 years out of 10.
The formula
P90 ≈ P50 × (1 − 1.282 × σ)
- P50
- = annual production in the central scenario
- σ
- = combined uncertainty, as a percentage of P50
- 1.282
- = statistical factor for a 90% probability
An example with numbers
A 1 MWp system with a P50 of 1,314 MWh a year, as in the GW, MW and kWp lesson, and a combined uncertainty of 7%. Illustrative figures, unrelated to any project.
| P50 | 1,314 MWh a year |
|---|---|
| Discount (1.282 × 7%) | 9% |
| P90 (1,314 × 0.91) | 1,196 MWh a year |
| Difference | 118 MWh a year |
In 9 years out of 10, at least 1,196 MWh is expected. A very bad year can still fall below it: P90 isn't an absolute floor.
Where do you see it at Circular Urban?
Each solar project's sheet shows the estimated annual production from the design simulation. When a technical study reports scenarios, it helps to know which one a figure belongs to. It's a reference projection: real production depends on each year's weather and the project's performance.
- Each solar project's sheet, in Energy Production: Annual Production.
- Each solar project's sheet, in Risks & Mitigants.
What changes in your finances once you get it
You ask which scenario it is
Faced with a production figure, you check whether it's P50 or P90 before drawing conclusions.
You plan with a margin
You also run your numbers with the prudent scenario, so a weak year doesn't catch you off guard.
You read the gap
If P50 and P90 are far apart, the projection carries more uncertainty.
Common mistakes
- Myth
P90 is the least the project can produce.
- In reality
It's a value expected to be exceeded 9 years out of 10. In one year out of ten, production can fall below it.
- Myth
P50 means the project runs at 50% of its capacity.
- In reality
The 50 refers to probability, not capacity: it's the value exceeded in half the years.
- Myth
If a year comes in below P50, the projection was wrong.
- In reality
By definition, half the years are expected to fall below P50. What matters is the average over several years.
Test yourself
A project has a P50 of 2,000 MWh and a combined uncertainty of 5%. What is its approximate P90?
Keep learning
This content is educational. Every participation carries risks inherent to the project, including the possibility of partial or total loss, and results are subject to each project's actual performance.

