How many significant figures should a preliminary wind-energy estimate report?

The short answer

A preliminary estimate should not display more confidence than its inputs support. Reporting many significant figures can make uncertain resource, loss and operating assumptions look precise. Choose a presentation that communicates the decision range while retaining sufficient internal precision for correct calculations.

Separate calculation precision from evidence quality

A spreadsheet can calculate a long decimal result from approximate inputs. Those digits describe arithmetic, not measurement accuracy or model validity. Keep internal calculations unrounded where appropriate, then round the reported result in a way consistent with the assessment purpose.

Record the main uncertainties alongside the central estimate. A reader should not have to infer confidence from the number of digits in the annual kWh total.

Show ranges where they affect the decision

Use supported sensitivity cases or a suitable uncertainty method to show how resource, curve and loss assumptions influence the outcome. Do not assign a statistical confidence label to an arbitrary high-and-low scenario range.

For an illustrative comparison, a tiny difference between two proposal totals may be immaterial when the same plausible input variation changes both totals much more. The report should make that limitation visible rather than ranking alternatives solely by the last displayed digit.

Preserve reproducibility behind the summary

Keep the underlying inputs, method and unrounded outputs in the assessment archive. The concise report can then use sensible rounded values without losing the ability to investigate a later revision or compare a changed configuration.

Before procurement, identify which evidence gaps must be resolved and which remain acceptable for the decision. An honest preliminary range can support that choice more effectively than a precise-looking single value. Neither presentation establishes a product guarantee, a verified payback or future annual production without the relevant supporting evidence.