MCS PV performance estimates

How the MCS annual-generation estimate is built, what assumptions it standardises and what it cannot guarantee.

An MCS PV performance estimate is a standardised forecast of annual AC electricity generation. It lets a customer compare proposed systems on a consistent basis. It is not a guarantee of what one future year will produce.

As at 22 July 2026, the current MCS solar PV installation standard is MIS 3002 Issue 6.0, dated 18 March 2026.

The standard calculation

MIS 3002 expresses the estimate as:

annual AC output in kWh = array kWp × Kk × shade factor

The inputs are:

array kWp, the sum of module data-plate power at standard test conditions
Kk, a location-specific annual yield value selected for postcode zone, orientation and pitch
shade factor, a multiplier representing expected shading loss

The result is AC energy for a year. It is not the inverter peak power, a daily target or the amount the home will use itself.

Measuring orientation, pitch and shade

MIS 3002 defines orientation as azimuth relative to due south and inclination as degrees from horizontal. It requires those site characteristics to be measured or determined from suitable plans.

Where the horizon is obviously clear and there is no near or far shading, the standard permits a shade factor of 1.00. Otherwise the designer must use a suitable shading method. An alternative to the MCS procedure must be demonstrably equivalent or better.

Arrays on different roof planes should not be hidden inside one average orientation. Their kWp and yield conditions need to be handled so the total reflects each plane. The same applies where one group has materially different shade.

Site visit and remote assessment

MIS 3002 recognises that drawings, maps or photographs can establish some inputs, while an accurate shade assessment will usually need the site to be examined. If a proposal is based on remote information, the customer must be warned that later site findings can change the performance estimate.

The pre-installation estimate should be revised when the final module count, orientation, inverter, shade or layout differs. An estimate left unchanged after a design alteration is no longer evidence for the installed system.

What the simple method does well

The MCS method gives a comparable calculation from auditable inputs. It avoids sales forecasts based only on a national “kWh per kWp” rule and forces orientation, pitch, location and shade into the estimate.

It is deliberately simpler than a full interval simulation. A detailed model can also represent:

module and cell temperature
inverter efficiency across power and voltage
string mismatch and DC cable loss
clipping and export curtailment
snow, soiling and availability
module-level electronics
hourly weather and horizon geometry

If a detailed model is presented, keep the MCS result as the standard comparison and list the extra inputs and losses. A more precise-looking number is not necessarily more accurate if its assumptions are hidden.

Estimate, expected value and guarantee

Solar resource varies from year to year. Actual output also depends on equipment availability, grid outages, dirt, vegetation, faults and any curtailment. A good forecast represents a central expectation under stated assumptions, not a minimum that every year must exceed.

Avoid wording such as “will generate” unless the contract genuinely guarantees it and defines measurement, exclusions and remedy. “Estimated annual generation” is the accurate description.

Generation-estimate inputs

Check that competing estimates use:

the same installed kWp definition
the actual postcode zone
measured roof-plane orientation and pitch
a site-specific shade assessment
the final inverter and array layout
clearly separated MCS and proprietary-model results
the same treatment of clipping, degradation and export limits

A higher generation figure may reflect a larger array, a different method or simply a more optimistic assumption. Ask for the calculation sheet, not just the final kWh.

After installation

The handover pack should show the as-built estimate and inputs. Compare actual energy over meaningful periods, allowing for weather and downtime. One poor month cannot establish underperformance, and a sunny first year does not prove the estimate was conservative.

For diagnosis, retain monthly generation, inverter availability and event logs. A weather-normalised performance assessment is more informative than comparing the monitoring app directly with one annual number.

Keep self-consumption separate

Annual generation answers how much PV energy the system is expected to make. Self-consumption answers how much of that energy is used on site. Bill savings also depend on import and export prices.

These are separate calculations. A generation estimate must not silently assume that all predicted kWh avoid the retail import price.

Applies to

Solar

Last reviewed

22 Jul 2026