Loss Factor Calculation

Answer: Calculate a combined retained fraction by multiplying the retained fractions for declared sequential losses. To calculate total loss, subtract that product from one. First verify each percentage’s denominator, time period and included effects. A report’s irradiance, DC-energy and AC-energy stages are different quantities; their percentages cannot all be added or multiplied indiscriminately.

Start with the denominator

For a deduction from a stated input:

Loss fraction = (input − output) / input

Retained fraction = output / input = 1 − loss fraction

Keep input and output in the same units and over the same period. A kW rating and an annual kWh total cannot be used as a dimensionless loss ratio. Record whether the calculation concerns radiation at the array, module DC energy, inverter AC output or a delivery meter.

The panel-derate-factor definition explains retained-factor terminology and software conventions. This page focuses on reconstructing the calculation rather than choosing an industry-wide factor.

Two valid calculations with different percentages

Assume a hypothetical 20,000 kWh starting energy, a first sequential deduction of 5%, and a second deduction of 4% of what remains.

The retained factor is 0.95 × 0.96 = 0.912. Total loss is 8.8%, or 1,760 kWh. Adding the two conditional percentages would give 9%, which overstates this example’s loss.

However, the two energy deductions can also be expressed against the same original 20,000 kWh denominator: 5% and 3.8%. These original-basis contributions add to 8.8%. The rule is not “percentages are never additive”; it is “identify whether the percentages share a denominator.”

These numbers are a worksheet, not measured plant losses or recommended model inputs.

Does the order matter?

For fixed scalar retained fractions, multiplication is commutative: 0.95 × 0.96 equals 0.96 × 0.95. The previous blanket claim that multiplication order changes the result was incorrect.

The attribution does change. Reversing the worksheet puts a 4% deduction first: 800 kWh lost, 19,200 kWh retained. The following 5% deduction is then 960 kWh. The final energy remains 18,240 kWh, but the named contributions differ.

In a physical simulation, the factors may not remain fixed when inputs change. Temperature depends on operating conditions; inverter clipping depends on the resulting power. Do not assume that physically rearranging effects preserves their independently recalculated contributions.

Reconstruct a loss diagram using its own definitions

PVsyst’s current loss-diagram documentation states that its displayed loss percentages use the previous energy stage as the reference. Its explanation also warns that individual attribution can be difficult when electrical effects interact. The numerical reconstruction should follow the report’s energy stages, rather than assume every category is independently removable.

Practical checks:

1. Preserve the report version, energy values and displayed percentages.
2. Note the preceding quantity for each percentage.
3. Recalculate one stage at a time and retain unrounded intermediate values.
4. Compare the reconstructed final quantity to the reported final quantity.
5. Explain remaining differences using rounding, variable definitions or modeled interactions; do not silently force the last row to match.

A sensitivity run with one changed assumption can be more useful than treating a loss-diagram line as guaranteed recoverable energy.

Reference energy needs a reference irradiance

A dimensionally explicit reference expression is:

E_reference = P_reference × (H_POA / G_reference)

H_POA is plane-of-array irradiation in kWh/m² over the period; G_reference is reference irradiance in kW/m². Their ratio has units of hours. For a hypothetical 100 kW array, H_POA = 1,500 kWh/m² and G_reference = 1 kW/m² give 150,000 kWh reference energy.

That is a reference quantity, not proof of an achievable theoretical maximum or a delivered-energy forecast. Use the chosen method’s incident or effective radiation definition consistently. PVsyst’s grid-system variable definitions distinguish reference energy used for PR from the nominal-energy starting point of its array loss diagram.

If the corresponding final energy is hypothetically 120,000 kWh, the ratio to that reference is 0.80. Calling it performance ratio requires the appropriate energy boundary, irradiation input and method. It is not automatically the same as every aggregate software loss factor, and 0.80 is not a universal acceptance target.

Keep a calculation register

Do not treat an arbitrary factor table as evidence for a current project. Cleaning returns, voltage-drop requirements, module clearances and equipment performance each require their own inputs. A first-week comparison also needs actual weather and availability; an unexplained difference is not automatically a universal commissioning failure.

When reviewing a solar generation and financial workflow, ask for the register and exported results for the demonstrated project. The PVsyst report guide gives a related evidence-handoff checklist. No software automatically establishes project approval or financing acceptance simply by displaying a waterfall chart.

Questions about loss calculations

Why do sequential loss percentages multiply?

Each conditional deduction acts on the quantity remaining from the previous stage. Multiplying the retained fractions reproduces that stated sequence.

When can loss percentages be added?

When they represent non-overlapping deductions expressed against the same original denominator. Convert them to that common basis first and check that no mechanism is counted twice.

Is the combined loss factor always performance ratio?

No. Performance ratio uses a defined final yield and irradiation-based reference yield. A combined factor may cover only selected deductions at another boundary.

Source review: September 30, 2026. Examples are hypothetical arithmetic; actual project methods and acceptance conditions require separate review.