Self-consumption and self-sufficiency

Two ratios describe how a PV home and the grid share the work. They use the same energy flows but divide by different totals, and they often move in opposite directions. Both follow the definitions in the review of self-consumption policies by the International Energy Agency’s Photovoltaic Power Systems Programme (IEA PVPS).

Definitions

Self-consumption is the share of your PV production that is used in the home instead of being exported.

Self-consumption = PV used in the home / PV produced

In everyday terms it answers “how much of my sunshine do I use myself?” A high figure means little of it leaves the house. It says nothing about how much of your bill the sun covers.

Self-sufficiency, also called autarky, is the share of your electricity use that your own PV covers instead of the grid.

Self-sufficiency = PV used in the home / total electricity used in the home
                 = 1 - (electricity bought from the grid for your own use / total electricity used)

In everyday terms it answers “how much of my electricity does the sun supply?” A high figure means you buy little from the grid. It divides by what the home uses, not by what the panels make.

A small illustration with round numbers, not a result: PV makes 6,000 kWh and exports 2,000, so 4,000 kWh are used at home. The home uses 10,000 kWh and imports 6,000. Self-consumption is 4,000 / 6,000, about 67 %. Self-sufficiency is 4,000 / 10,000, which is 40 %.

The two pull apart as the array grows. A small array on a house that uses a lot has high self-consumption and low self-sufficiency: nearly all it makes is used, but it covers only a slice of the demand. A large array does the opposite.

EnergyFlow reports self-consumption, both as a percentage and as kWh of PV used on site, using the formula below. It does not report self-sufficiency as a figure. With no battery or car, you can work it out from the results as PV self-consumed divided by (PV self-consumed plus total grid import). With a battery or car, the results split grid imports by destination (house, battery, car) so you can leave out energy bought only to be sold again, and the answer is an estimate.

The formula EnergyFlow uses

Every hour the model compares PV with the house’s own demand: heating, cooling, hot water, ventilation and household electricity. Charging the battery or the car is not part of that demand. The calculation follows these steps:

PV surplus (hour) = max(0, PV - house demand)
Export from PV (hour) = min(grid export, PV surplus)
Self-consumption = (annual PV - sum of export from PV) / annual PV

The result is limited to between 0 and 100 %.

Two consequences follow from this definition.

First, PV that goes into the battery or the car counts as used on site. The model does not count it as exported even if the battery sells it later.

Second, only PV export reduces the ratio. When the grid export in an hour is larger than the PV surplus, the excess is attributed to battery discharge and then to the car. A battery that was charged from the grid and exports at a high price does not lower your PV self-consumption. This keeps the ratio meaningful when a controller trades grid energy.

Why optimising cost can lower self-consumption

The rule-based mode, a fixed order of priorities with no prices in it, always puts PV surplus into the battery first. That produces a high self-consumption by construction. The price-aware optimiser, which runs with a dynamic tariff (hourly prices) and a battery, is not trying to maximise a ratio. It minimises cost.

It may export PV in an hour when the export price is high, instead of storing it for later. It may fill the battery from cheap grid power overnight, leaving less room for the midday PV, which is then exported. Both choices lower self-consumption and can still lower the bill. A lower percentage under the optimiser is not a sign that something is wrong.

What changes it

There is no general figure, because the answer depends on your house. These are the directions of the main effects.

• PV size against demand. A larger array raises the kWh used on site but lowers the self-consumption percentage, because more of the output arrives when the house does not need it. Self-sufficiency usually rises with PV size.
• Roof directions. An east and west split moves production toward the morning and evening, when many homes use more power.
• Battery capacity and power. Both let more midday PV be used later, with less benefit once the battery can absorb a typical day.
• When loads run. Loads that coincide with sunshine, such as daytime EV charging at home, raise self-consumption. Heating and hot water do not shift to follow PV or prices.
• Season. Surplus is large in summer and small in winter, so annual figures hide a strong seasonal swing.
• Tariff and control mode (rules or optimiser), as described above.

To measure one effect, change only that input and compare two runs.

Limits of the model

• The model nets PV and demand over a whole hour. Netting over shorter periods would show less self-consumption, so hourly figures can be higher than minute-level metering would show.
• Household electricity is a modelled hourly profile scaled to the annual figure you enter. It is not your meter data.
• Energy lost inside the battery still counts as used. PV that is stored and partly lost on the way through the battery (its round-trip efficiency) is not exported, so it stays in self-consumption.