A power plant can generate 1,000 megawatt-hours, yet homes and businesses will receive less than that amount. The difference is often called a grid loss—but the number is easy to misread unless you know where the measurement starts, where it ends and which denominator is being used.
Key takeaways
- Some electricity is inevitably converted to heat as it moves through conductors, transformers and other equipment.
- Published loss figures may also include metering errors, unbilled use or theft, depending on the dataset.
- Two percentages are comparable only when they cover the same network boundary, time period and denominator.
- A falling loss rate can improve efficiency, but it does not by itself prove that the grid is reliable, affordable or low-carbon.
Start with the electricity balance
Electricity accounting follows a simple idea: energy entering a defined system must be matched by energy delivered, used within the system, stored or lost. The difficulty is that reports do not always define the system in the same way.
A national balance may begin with total electricity output from power plants. A utility report may begin with electricity purchased or received at the edge of its network. A transmission operator may stop measuring where high-voltage power is handed to a distribution company. Each boundary produces a valid result, but the percentages answer different questions.
The World Bank indicator for electric-power transmission and distribution losses covers losses between supply sources and distribution points and from distribution to consumers, including pilferage. It expresses the result as a share of electricity output. That definition is broader than a purely engineering estimate of heat loss. See the World Bank indicator metadata.
Technical losses and non-technical losses are different
Technical losses
Current meeting resistance in wires produces heat. Transformers, cables and other grid equipment also consume a small amount of energy while operating. These losses cannot be reduced to zero, but utilities can limit them with suitable conductor sizes, efficient transformers, higher transmission voltages, balanced loads and careful network planning.
Distance matters, but it is only one factor. A heavily loaded line can lose more energy than a lightly loaded one because resistive losses rise with the square of current. Voltage level, conductor material, temperature, transformer efficiency and local demand patterns all affect the result.
Non-technical losses
The gap between electricity supplied and electricity recorded as sold can also come from inaccurate meters, data errors, unmetered connections, theft or electricity that is delivered but not billed. These are usually called non-technical or commercial losses. They are an accounting and governance problem as much as an engineering problem.
That distinction matters. Replacing a conductor may reduce technical losses, while better metering, billing and enforcement address non-technical losses. A single combined percentage cannot tell you which intervention is needed.
A 1,000 MWh example
Consider a hypothetical system during one day. The figures below illustrate the calculation; they are not data for a particular country or utility.
| Stage | Electricity | What changed |
|---|---|---|
| Gross generation | 1,000 MWh | Electricity measured at power plants |
| Power-station use | 20 MWh | Pumps, fans and other plant equipment |
| Electricity entering the network | 980 MWh | Net electricity available to transmit |
| Transmission loss | 30 MWh | Losses before the distribution boundary |
| Distribution and commercial loss | 50 MWh | Physical and non-technical gap |
| Delivered and recorded | 900 MWh | Electricity reaching recorded end use |
If the report divides the 80 MWh network gap by gross generation, the result is 8%. If it divides by the 980 MWh entering the network, the result is about 8.2%. Neither calculation is automatically wrong. The denominator must be stated.
The same principle explains why generation, net supply and final consumption should not be treated as synonyms. For a related distinction, see why renewable capacity is not the same as clean electricity generation.
Why country rankings can mislead
A higher published loss rate can reflect ageing equipment or overloaded networks. It can also reflect long rural feeders, dispersed islands, difficult terrain, incomplete metering or differences in data quality. A lower number may indicate an efficient grid, but it may also exclude parts of distribution or rely on estimates that are not directly comparable.
The U.S. Energy Information Administration estimates that U.S. transmission and distribution losses averaged about 5% of electricity transmitted and distributed from 2018 through 2022. Its method removes direct-use electricity because that power never enters the grid. The calculation note is as important as the percentage itself; read the EIA methodology and estimate.
Before comparing two places, check five items:
- Boundary: transmission only, distribution only or both?
- Denominator: gross output, net supply or electricity entering the network?
- Components: technical losses only or technical plus non-technical losses?
- Period: one month, one year or a multi-year average?
- Data quality: measured values, estimates or a mixture of both?
Why the issue matters in Southeast Asia
Electricity demand in Southeast Asia is growing rapidly, while renewable generation, data centres, cooling and electric mobility are changing where and when power flows. The IEA’s Southeast Asia Energy Outlook 2026 says transmission and distribution networks need to more than double in length by 2050 to keep pace with demand and more variable supply.
Expansion alone does not guarantee efficiency. New lines, substations, transformers, sensors and meters must be planned as one system. Regional interconnectors can help countries share electricity and balance renewable output, but they also create additional commercial boundaries where metering and loss allocation must be transparent. The IEA estimates that more than USD 300 billion of grid investment will be needed in Southeast Asia from 2025 to 2040 in its report on financing the ASEAN Power Grid.
Grid losses also interact with flexibility. Shifting demand away from stressed hours can reduce current on constrained lines and improve use of existing infrastructure. Our explainer on demand flexibility in Thailand shows why timing can matter alongside total electricity consumption.
What a useful grid-loss report should show
A credible public dashboard or utility report should publish both energy quantities and percentages. It should define the network boundary, separate transmission from distribution where possible, explain how own-use and distributed generation are treated, and disclose whether non-technical losses are included.
Trend data are more informative than a single-year ranking. A utility can then connect changes to specific work: transformer replacement, conductor upgrades, voltage optimisation, meter replacement, feeder reconfiguration or improvements in billing. Reporting uncertainty is also valuable, especially where meters are missing or data are estimated.
The practical conclusion
“Electricity loss” is not one universal measurement. It is the difference between two defined points in an electricity balance. To interpret it, ask where the power was measured, what the total includes and which denominator produced the percentage.
Once those definitions are clear, loss data become useful. They can reveal where grid investment, operational changes or better metering may deliver the greatest benefit—without confusing electricity generated with electricity actually delivered.
Image note: The featured image is an original AI-generated editorial illustration created for ClimateCrisis.net. It depicts generic electricity infrastructure in Southeast Asia and is not a photograph of a specific facility.