Set the right indicators before calculating self-consumption
Self-consumption rate: definition, formula and field interpretation
The self-consumption rate measures the share of PV production consumed on site, excluding export. Formula: self-consumed PV energy / PV production. For a defensible self-consumption calculation, work at hourly resolution. Cross-check estimated or measured production (inverter, production meter) with grid import (Linky) to isolate the energy actually consumed at the right time.
Self-production rate (coverage): definition, formula and customer-facing interpretation
The self-production rate relates this self-consumed energy to the site’s total consumption. Formula: self-consumed PV energy / total consumption. This is the indicator that speaks to the client, because it translates into an average coverage of their consumption. Present it as an annual ratio, never as a promise of autonomy.
Why increasing one often reduces the other (with a simple example)
Without a battery or controlled water heater tank, oversizing mainly increases daytime surplus. You improve coverage, but you reduce the self-consumption rate.
| Case | PV production (kWh/year) | Self-consumed (kWh/year) | Self-consumption rate (%) | Self-production rate (%) |
|---|---|---|---|---|
| 3 kWp | 3 300 | 1 400 | 42 | 31 |
| 6 kWp | 6 600 | 1 800 | 27 | 40 |
Gather solid input data for a defensible self-consumption calculation
Consumption profile: ideally hourly (otherwise day/night) and how to obtain it
For a serious self-consumption calculation, start from an hourly load curve. The cleanest approach is a Linky export via the Enedis portal or API access (with the client’s mandate). Failing that, rebuild a day/night profile from peak/off-peak tariffs, then align it with the billed monthly kWh.
Production curve: orientation, tilt, shading and location to document
Document the azimuth, tilt, DC power and inverter AC limits. Record shading with time-stamped photos (morning, noon, evening) and note the horizon. For estimation, use a weather-radiation tool such as PVGIS and keep the hourly output, not just the annual figure. To go further on the impact of the installation, see orientation and tilt.
Special cases to include: electric heating, water heater, swimming pool, air conditioning, electric vehicle
Identify controllable loads and their actual schedules. Electric heating and air conditioning follow temperature. The water heater depends on the contactor and reheating cycles. Swimming pools and EVs create peaks. Without sub-metering, make your assumptions explicit and plot the dedicated load curve.
Reliable method: cross-check production and consumption hour by hour
Hourly calculation principle: self-consumed = minimum(production, consumption) for each hour
For a defensible self-consumption calculation, work on an hourly series. At each hour, self-consumed energy is the smaller of the two values. Self-consumed(h)=min(Production(h), Consumption(h)). Surplus(h)=max(0, Production(h)-Consumption(h)).
Why a simple annual ratio overstates self-consumption (the classic trap)
An annual ratio ignores the synchronisation between production and uses. A site that produces mostly in the middle of the day and consumes mostly in the evening may look “good” on an annual basis. Hour by hour, however, you will see a great deal of exported surplus. This shortcut often inflates the result, especially without hot water tank control or a battery.
Practical steps to perform the calculation and present a traceable result
Aim for a traceable result with sourced inputs and archived files.
- Retrieve a load curve from the smart meter or a logger. Fill gaps, justify atypical values.
- Generate a production curve using a recognised tool (production curve PVGIS, manufacturer software) and document orientation, tilt and shading masks.
- Align the timestamps, calculate the hourly minimum, then aggregate by month and year.
- Attach a table of assumptions and the method to the quote. You secure your argument against low-cost competitors.
Improve the self-consumption rate without making things up: control and storage
Controlled hot water tank: real impact on the self-consumption calculation depending on schedules
A controlled hot water tank does not “create” self-consumption, it shifts a load. For a defensible self-consumption calculation, cross-check PV production and power draw hour by hour. The heating element (often 1.8 to 3.0 kW) requires a heating window aligned with the solar peak. Also check night reheating (anti-legionella, boost) which worsens the result.
Battery: what it changes on both rates, and the limits (efficiency, power, season)
The battery increases the self-consumption rate by capturing midday surplus, and it can raise the coverage rate too, but only if the charge/discharge capacity and power keep up with the peaks. Factor the round-trip efficiency and depth of discharge into the hourly model. In winter, low production mechanically limits the gain, even with a large capacity. To quantify sizing and profitability, rely on a complete profitability calculation for a domestic storage battery.
Surplus: grid export, surplus sale and on-site uses (2026 watchpoints)
Treat surplus as a balance output, not as a “bonus.” Since 05/06/2026, the self-consumption bonus is no longer part of the case argument for submitted applications. Surplus sale remains governed by a 20-year purchase contract, at a very low level (order of cents). Without storage, prioritise controllable on-site uses, but document the scheduling and power assumptions to avoid an over-optimistic estimate.
Make your figure “solid” in a proposal: assumptions, evidence and margins
Clearly list your assumptions and data sources (readings, scenarios, variations)
For a defensible self-consumption calculation, state plainly what is measured and what is estimated. Attach the evidence, and date it. This protects you in the event of a dispute, and frames the scope.
- Reference consumption, period covered, hourly resolution, atypical days excluded or kept.
- Estimated production, orientation, tilt, shading, loss assumptions.
- Controlled loads, selected control rules, usage constraints, contractual limits.
Show a range of results rather than a single figure (summer/winter, weekday/weekend)
Present a realistic range with three cases: conservative, central, favourable. Distinguish at least summer and winter, weekday and weekend. Remind the client that pushing self-consumption can reduce coverage, and vice versa. Your proposal becomes easy to read, without making unrealistic promises.
Avoid building a profitability argument on an “assumed” profile: how to secure your message
Do not base your margin on a “typical” curve. Anchoring the message in measured data changes everything. Obtain a real history, trace your adjustments, and state what depends on behaviour. If you do not have hourly data, say so, widen the range and explain why.




