Key facts
- Energy return on invested (EROI) compares electricity produced over a plant's life with the energy needed to build, fuel and decommission it.
- Gas-centrifuge enrichment needs roughly 40-50 kWh per SWU, about 2% of what gaseous diffusion needed (World Nuclear Association).
- A harmonised review of light-water reactor studies reported a median of 12 g CO2-eq/kWh; another review gave 10-130 g with an average of 65 g.
- Weißbach et al. reported an EROI of about 75 for a reference plant, and critics of that approach report far lower values; results depend heavily on assumptions.
What an energy balance counts
An energy balance asks how much energy it takes to build a power plant, supply it with fuel for its life, run and decommission it, and how that compares with the electricity it delivers. Two metrics recur: the energy payback time, the period of operation after which the plant has produced the energy invested in it, and the energy return on invested, or EROI, the ratio of lifetime energy output to lifetime energy input.
A related life-cycle result is the greenhouse-gas intensity, which tracks emissions rather than energy but is often used as a proxy because most fuel-cycle emissions stem from energy use. The metrics are related but not identical, and studies differ in whether they count heat or electricity and whether inputs are weighted by their primary-energy equivalent.
Where fuel-cycle energy is used
For a light-water reactor the main inputs are plant construction, mining and milling, conversion, enrichment, fabrication, operation and maintenance, and decommissioning and waste management. Which dominates depends on ore grade, enrichment technology and reactor lifetime. Lenzen's review, which compared published studies, noted that most emissions in the nuclear fuel cycle arise in processing stages upstream and downstream of the plant itself. Weißbach et al. found that overall energy demand was dominated by uranium extraction and enrichment, and argued that with the shift of enrichment from diffusion to centrifuge, extraction would become the dominant item, particularly as ore grades decline.
The step that changed most over time is enrichment. The World Nuclear Association gives roughly 40-50 kWh per SWU for gas centrifuges and 2,400-2,500 kWh per SWU for gaseous diffusion. Applying those figures to the 6,871 SWU needed for one tonne of 4.5% product (see the SWU calculator), enrichment takes about 275-345 MWh by centrifuge and about 16.5-17.2 GWh by diffusion.
The WNA gives about 360,000 kWh of electricity per kilogram of fuel at a burn-up of 45,000 MWd/t, or 360 GWh per tonne. On those figures, enrichment is about 0.1% of the electricity the fuel produces with centrifuges and about 4.6-4.8% with diffusion. These ratios are illustrative arithmetic from the cited factors, not results of a full life-cycle study.
Range of published results
| Study | Metric | Reported result |
|---|---|---|
| Lenzen (2008), review of published studies | Life-cycle greenhouse-gas intensity, LWR and HWR | 10-130 g CO2-eq/kWh, average 65 g |
| Warner and Heath (2012), systematic review and harmonisation, LWR | Life-cycle greenhouse-gas intensity | Median 12 g CO2-eq/kWh after harmonisation (13 before), based on 27 references |
| UNECE (2021), life-cycle assessment of electricity options | Life-cycle greenhouse-gas intensity of conventional nuclear | About 5.1-6.4 g CO2-eq/kWh |
| Weißbach et al. (2013), reference plant | EROI | About 75 (about 105 with 100% centrifuge enrichment); payback about two months |
These figures cannot simply be averaged because the methods, boundaries and base years differ. The spread between the review ranges reflects, in part, studies that use older diffusion-based enrichment or low-grade ore assumptions, and in part the choice of whether, for instance, background processes are modelled with process data or economic input-output tables. Warner and Heath also estimated that, if uranium ore grades were to decline, median results by 2050 could range between about 9 and 110 g CO2-eq/kWh, according to the abstract of their paper.
Why the estimates disagree
Several methodological choices drive the divergence. First, the enrichment technology assumed: diffusion-era data produce higher energy intensity than centrifuge data. Second, ore grade and mining method: low-grade ore requires more energy per kilogram of uranium, and assumptions about future grades range widely. Third, plant lifetime and capacity factor: Weißbach et al. assumed 8,000 full-load hours per year and a 60-year life, and note that assuming shorter lives lowers EROI. Fourth, system boundaries: whether back-end waste management, decommissioning, and the energy used to produce the plant's materials are included, and whether a heat-to-electricity weighting is applied.
The same paper disputes earlier analyses based on work by Storm van Leeuwen, arguing that the mining energy in those analyses was overstated and diffusion enrichment assumed; the authors report that such analyses yield EROI values lower by a factor of about 20. Those analyses continue to be cited by critics of nuclear power, and the authors of the review literature do not agree on a single value. Readers wanting to compare studies should look at each study's boundary, enrichment mix, ore grade and lifetime assumptions before comparing headline numbers.
Reading the numbers with care
An energy balance describes inputs and outputs; it does not measure cost, risk or waste. It also applies to a specific plant type and fuel route. Studies that cover reactor types other than light-water designs, advanced fuels or reprocessing routes are outside the figures above. The UNECE report itself focuses on conventional nuclear generation rather than advanced designs. For cost aspects of the same chain see fuel cycle costs, and for the process stages themselves see the fuel production chain.
Frequently asked questions
What is EROI?
Energy return on invested is the ratio of the energy a plant delivers over its lifetime to the energy needed to build, fuel, operate and decommission it.
How much energy does enrichment use?
The World Nuclear Association gives roughly 40-50 kWh per SWU for gas centrifuges and 2,400-2,500 kWh per SWU for gaseous diffusion.
Do life-cycle studies agree on nuclear's energy balance?
No. Reviews report ranges, for example 10-130 g CO2-eq/kWh in one review and a harmonised median of 12 g in another, and EROI estimates differ by large factors depending on assumptions.
Why do results differ between studies?
Differences stem from enrichment technology, ore grade, plant lifetime, capacity factor and what is included inside the system boundary.
Is the greenhouse-gas intensity the same as the energy balance?
No. They are related because many emissions result from energy use, but the metrics are calculated differently and should not be interchanged.
Sources and further reading
- Energy Conversion and Management (Lenzen) — Life cycle energy and greenhouse gas emissions of nuclear energy: A review (2008)
- Journal of Industrial Ecology (Warner and Heath) — Life Cycle Greenhouse Gas Emissions of Nuclear Electricity Generation: Systematic Review and Harmonization (2012)
- UNECE — Life Cycle Assessment of Electricity Generation Options (2021)
- Energy (Weißbach et al.) — Energy intensities, EROIs, and energy payback times of electricity generating power plants (2013)
- World Nuclear Association — Uranium Enrichment (2024)
- World Nuclear Association — Economics of Nuclear Power (2024)
Related topics
