Reading an energy statistic without being fooled

Energy debates are conducted with numbers that mean different things to different speakers. A handful of distinctions resolve most of the confusion.

Capacity versus generation

The most common and most consequential confusion.

Capacity is the maximum rate at which a plant can produce, measured in watts. A gigawatt of solar and a gigawatt of nuclear are the same capacity.

Generation is the energy actually produced over a period, measured in watt-hours.

The link between them is the capacity factor — actual output divided by theoretical maximum output over the period.

Typical capacity factors differ enormously: solar in the region of ten to twenty-five percent depending on latitude, onshore wind twenty-five to forty, offshore wind higher, nuclear frequently above ninety.

Which means a gigawatt of solar produces a small fraction of the annual energy a gigawatt of nuclear does.

Headlines about capacity installed are therefore not headlines about energy delivered, and both are cited selectively.

Energy versus electricity

The second most common error, and it flatters progress substantially.

Electricity is a portion of total energy use. Transport, industrial heat and building heat consume large quantities of energy directly as fuel.

A country reporting that renewables supply a large share of its electricity may be supplying a much smaller share of its total energy.

Both figures are legitimate; conflating them is not, and it happens constantly.

Primary energy and the substitution problem

A subtler issue that cuts the other way.

Primary energy accounting counts the thermal energy content of fuels. A thermal power station converts a minority of that to electricity and rejects the rest as heat.

Wind and solar produce electricity directly with no thermal losses.

Counting them on the same basis therefore understates their contribution, because a unit of renewable electricity displaces several units of primary fossil energy.

Different statistical agencies handle this differently — some apply a substitution factor, some do not — which is why the same country's renewable share differs between sources.

When comparing figures, check which convention is used.

Levelised cost and what it omits

The standard cost metric, and it is incomplete in a specific way.

Levelised cost divides lifetime cost by lifetime output, producing a cost per unit of energy.

It does not account for when the energy is produced.

Electricity delivered during peak demand is worth more than electricity delivered at midday alongside everything else. A source with a low levelised cost that produces when the system does not need it is worth less than the number suggests.

This is why value-adjusted metrics and full system cost modelling have become more common, and why levelised cost comparisons between dispatchable and variable sources are misleading on their own.

Emissions intensity and boundaries

Where the boundary is drawn determines the answer.

Operational emissions only, or full lifecycle including manufacturing and decommissioning?

Territorial emissions, or consumption-based including imported goods? Countries that have offshored manufacturing look substantially better on the first than the second.

Are upstream methane emissions from gas extraction counted, and at what global warming potential over what time horizon? Methane's relative impact differs by a large factor between twenty-year and hundred-year framings.

None of these choices is wrong. All of them change the number, and the chosen framing usually favours the speaker's argument.

Percentage of what

Always ask.

Percentage of new capacity added, of total capacity, of electricity generated, of total energy, in a given hour, on a given day, or over a year — these differ by very large margins.

Records for instantaneous renewable share on a windy Sunday afternoon are genuine and are not the annual figure.

Averages across time and space

Two further traps worth naming.

Annual averages conceal the periods that determine system design. A grid can meet total annual demand with renewables and still require substantial firm capacity for a still, cold week in winter.

Citing the annual figure to argue that the winter week is solved is a common move and does not follow.

Geographic averages behave the same way. A national renewable share tells you nothing about a region with weak transmission links, where local generation and local demand must balance more closely.

The useful figures are therefore distributional rather than central: what happens in the worst week, in the least-served region, rather than what the year-round national average was.