Analyse electric fleet charging losses: Understand the measurement points

Charging losses describe energy drawn during charging that does not arrive as stored battery energy within a defined measurement boundary. A reliable analysis therefore needs suitable measurement points, the same time period and traceable energy values. The difference between a charging bill and the vehicle’s displayed driving consumption is not, by itself, a precise measurement of charging losses.

Two questions are useful for fleet management: how much purchased energy does driving require overall? And what portion of a measured energy difference can be attributed to charging? Both questions can be valuable, but they require different data.

Define the measurement boundary first

A meter at the site connection, the measurement from an individual charging point and vehicle data do not necessarily capture the same energy. Other consumers, conversion steps or shared facilities may lie between the measurement points.

The US Department of Energy also describes energy losses between electricity drawn from the grid and the battery in its energy-flow illustration. However, a general figure stated there cannot replace measurements of your vehicles under their operating conditions. Source: energy flow when charging electric vehicles.

State above every analysis which boundary it uses. “Energy at the charging point to measured battery input” is a different metric from “site electricity consumption per assigned kilometres”. The clearer the label, the lower the risk of a wrong conclusion.

Which data do you actually need?

A loss calculation requires energy measured at the defined input and a methodologically suitable measurement at the defined output. Both must belong to the same charging session. A vehicle’s percentage display is not a directly measured energy meter.

“Battery capacity multiplied by the change in state of charge” can at most provide a rough estimate, provided the underlying assumptions are known. The displayed percentages, usable capacity and meaning of the vehicle data supplied must be clarified. Do not publish such an estimate as an exact technical efficiency figure.

For cost management, another clearly named metric is often enough: purchased kilowatt-hours per 100 kilometres driven. This requires complete charging sessions and matching mileage data. Over short periods, a substantial change in battery charge level affects the result. A vehicle may finish the month with energy stored for the following month.

Worked example with a clear system boundary

Purely illustrative measurement example: For a defined charging session, 40 kWh is recorded at the specified input. A suitable, time-matched output measurement records 36 kWh within the defined battery boundary. The energy difference is 4 kWh.

Relative to the input energy, the calculation is:

• Difference as a percentage: (40 − 36) ÷ 40 × 100 = 10 per cent.

• Energy transferred within this measurement boundary: 36 ÷ 40 × 100 = 90 per cent.

The figures are deliberately chosen and make no claim about typical vehicles. The 4 kWh also requires examination of the measurement arrangement before it can be properly attributed. If the input measurement includes additional consumers, the difference includes their energy too.

If the output measurement is missing and only the 40 kWh from a charging receipt is available, this loss calculation is not possible. You can analyse costs and purchased energy, but you cannot determine a precise loss percentage.

Investigate anomalies without inventing causes

If purchased energy per kilometre rises, first check the data coverage: is mileage complete? Have charging sessions received later been assigned to the right period? Are there duplicate sessions or vehicles temporarily used differently?

Only then examine the operating conditions. Record the charging location, time window, temperature conditions, vehicle group and any unusual observations. Involve qualified personnel for technical anomalies. A single deviation does not establish a particular fault.

Keep the original values for clarification and mark corrections. Good data quality in fleet software prevents assignment errors from appearing as technical losses. If an anomaly recurs under documented conditions, you can arrange a targeted measurement.

Checklist for a reliable energy analysis

Before sharing a result, check:

1. Is the metric clearly named together with its measurement boundary?

2. Are input and output measured or merely estimated?

3. Do both values belong to the same vehicle and session?

4. Are additional consumers within the measurement boundary known?

5. Are the period, units and data version consistent, or are differences explained?

6. Are missing values and measurement uncertainties visibly disclosed?

7. Does the available data support the action being proposed?

For regular cost reports, you can also show purchased charging energy and its costs separately. The cost-per-kilometre calculation uses the charging costs actually recorded. This prevents a second blanket loss surcharge being added to costs that already include this energy consumption.

Use your existing charging and vehicle data to clarify which analysis is reliably possible: discuss your data requirements with StromNow.