Distribution of a site's photovoltaic power in consumption

How is the photovoltaic electricity distributed among consumers?

1. Electricity consumption is divided into two parts: that coming from the grid and that coming from the photovoltaic (PV) installation.
2. To determine these two shares, Climkit defaults to the input meter, which has two flows: draw from/feed back to the grid, as well as consumption meters.
3. The share of photovoltaics in consumption is then calculated as follows: PV share = sum of consumption - draw from grid. The site's autonomy rate is obtained by the formula: PV share / sum of consumption.
4. Climkit distributes the PV share among consumers every 15 minutes by applying the site's autonomy rate to their individual consumption. The share not covered by PV is then supplied by the grid.
5. By basing the calculation on the input meter flows, it is ensured that the draw from and feed back to the grid are accounted for in accordance with the DSO's billing and remuneration, and that the draw from the grid is distributed fairly among consumers.
6. The photovoltaic production is therefore deduced using the following formula: Production = sum of consumption - draw from grid + feed back to grid. The site's self-consumption rate is obtained by the formula: PV share / production.

Why is there feed back to the grid and draw from the grid within the same 15-minute period when production equals consumption?

1. If consumption is greater than production at the beginning of the period, electricity is drawn from the grid. But if consumption decreases at the end of the period, the surplus production is fed back to the grid.
2. If we did not rely on the input meter, we would have 100% site autonomy during those 15 minutes. However, this would not reflect the reality accounted for by the DSO, which bills for draw from the grid at the beginning of the period and remunerates feed back to the grid at the end of the period.

Why is there some production at night?

1. Since there are losses and metering discrepancies (for example, the input meter often measures a lower draw from the grid than all consumption meters, even without production), these discrepancies are absorbed in the calculation of the deduced production.
2. If the draw from the grid is higher than the consumption, production is observed at night. This means that the input meter is less precise than the set of consumption meters.
3. Conversely, if the draw from the grid is lower than the consumption, a negative production is obtained, indicating that part of the consumption is not measured by a meter.
4. If these values remain minimal, they represent normal losses in the installation that can be ignored, as they slightly reduce production without affecting the draw from the grid and thus the grid's share in consumption.

What to do in case of large discrepancies between the draw from the grid and consumption?

1. If the difference between the draw from the grid without production and the consumption is significant, it indicates that at least one consumption point is not measured, meaning at least one meter is missing.
2. Pending the installation of an additional consumption meter, a rule meter is created to deduct this "unmeasured" consumption. This rule meter can then be added to the site's common meter or directly assigned to a billing point.
3. By deducting this unmeasured flow, the input meter, consumption meters, and production meter are accounted for.

Why not create a rule meter and deduct the unmeasured flow in all cases?

1. This rule meter would absorb all small differences and thus sometimes record positive, sometimes negative values, which would influence the grid's share in consumer consumption, and it would no longer exactly match the quantity billed by the DSO.
2. Furthermore, if the rule meter is assigned to the common billing point, it would increase or decrease the common consumption, which would no longer match what is actually measured by the common meter.
3. In conclusion, even if this would make the charts more consistent (without night-time production), the unmeasured flow should only be deducted if it is genuinely an unmeasured consumer. In all other cases, the production flow is deducted, which absorbs discrepancies and losses, while remaining aligned with the input meter as accounted for by the DSO.

How does a battery work and what is its impact on self-consumption?

1. The installation of a battery allows for the storage of surplus photovoltaic (PV) electricity produced on a site. When PV production exceeds instantaneous consumption, the surplus is stored in the battery.
2. Once the battery is fully charged, any further surplus is fed into the electricity grid.
3. When consumption exceeds solar production, the battery discharges to supply the building's consumers. This mechanism significantly increases the self-consumption rate, as solar electricity produced during the day is also available at night.
4. When the battery is empty, the electricity supplement is automatically drawn from the grid.

Why are there differences between the data from the photovoltaic inverter, the DSO, and the Climkit platform?

It is entirely normal to observe discrepancies between the data displayed on the Climkit platform, those measured by the inverter, the battery, or the Distribution System Operator (DSO) meter.

Several reasons explain these differences:

1. Meter Tolerance: Certified meters (e.g., MID) have an accuracy between 0.5 and 1%. Other meters, such as certain "smart meters" integrated into the inverter, may be slightly less accurate.
2. Metering Type: Direct metering (meter connected directly to the circuit) is more accurate than indirect metering with current transformers (CTs). For best results, CTs adapted to the actual measured current should be used. In practice, DSOs often install oversized CTs, which can cause underestimation at low currents.
3. Calculation Methods: Electricity self-consumption on a PV site is generally calculated from different measurements and is not directly recorded. Systems may deduce certain values from others, which leads to differences: for example, an inverter may estimate building consumption from production and the input measurement (draw from and feed back to the grid), while another system may calculate production based on measured consumption.
4. Measurement Location and Losses: The production indicated by the inverter corresponds to electricity generated in direct current (DC), whereas Climkit measures that actually fed in as alternating current (AC) into the building's grid, after conversion. DC/AC conversion and cables result in a 3 to 5% loss.
   When using an MV/LV transformer, losses are approximately 5%.
5. Measurement Frequency: Systems measure and transmit data at different intervals (every minute, every 5 or 15 minutes, at fixed or random times, etc.), which can cause small differences, especially if consumption varies rapidly. Furthermore, rounding figures can lead to slight deviations over a period's total.
6. Battery Presence: If the site includes a battery, the method of metering stored or returned energy varies depending on the systems, particularly at night when the battery discharges. Small amounts of energy may be fed into or drawn from the grid without always being taken into account by the battery monitoring system.

In summary, discrepancies of a few percent (or a few kWh) between two metering systems are normal and do not indicate an error or malfunction.

How to verify the accuracy of the measurements?

Climkit regularly checks the consistency of its measurements. The simplest test is to examine the data at night: without solar production, the sum of individual consumptions must correspond to the energy imported from the grid (main meter). This "night test" is a good indicator of system operation.

For any other system, you should contact the installer to check the configuration and proper functioning of the hardware.

Finally, to precisely compare two systems, it is recommended to export and compare the load profiles (in 15-minute steps) over several days. This data, available on the Climkit platform (Excel file), allows for a detailed analysis of any differences.