Notifications
No notifications yet.
How reported price-change events become daily and weekly comparisons. Each qualifying forecourt contributes one equally weighted price.
Core calculation
The source history records changes, not a fresh observation from every station every day. Fuel Price Pulse reconstructs a comparable daily position before calculating an average.
For each station and fuel, the final genuine price event recorded that day becomes the closing value.
The last recorded price remains in place until a later price-change event replaces it. No change event does not mean the station had no price.
One closing value per qualifying station is averaged for each fuel. Frequent reporters cannot contribute extra observations and dominate the result.
Weekly and 30-day movements use only stations with a reconstructable price at both endpoints. The matched station count is published with the change.
Two views of the market
The distinction matters when comparing a published weekly figure with the current Price Pulse dashboard.
Published weekly series
Historical prices are reconstructed from change events. A station and fuel enters the series from its first genuine event, and today's active status is not projected backwards.
Current public figures
The current figures in the press room, publisher page, UK Fuel Price Index and Price Pulse cards share one calculation. Each open, active forecourt contributes its latest accepted price for the selected fuel, dated within the previous 14 days. Missing, implausible or old prices, closed sites and invalid locations are excluded.
Freshness is checked separately for each fuel. A station's address or other details being edited does not refresh its price. The latest included price timestamp and the last completed feed import describe different events; neither means every station reported a new price then.
The UK Fuel Price Index looks back over the stations eligible for its current figures. It reconstructs one closing price per station and fuel, carrying the last valid price forward through days with no change. A station enters from its first valid record. An invalid later record removes that value until another valid record arrives. Historical carried values do not expire after 14 days; that rule selects the current cohort.
The graph includes daily station counts and a partial point for today. Changes over 7 and 30 days use only stations with an accepted current price and a reconstructable price at the historical endpoint. They can differ from subtracting two full-cohort averages.
This current-cohort lookback can change when today's coverage changes. Frozen Price Pulse editions instead preserve the historical coverage, payload and method recorded at publication. Use a dated edition when an article needs a fixed reference.
Coverage and validation
These checks keep obvious errors out and prevent very small groups from being presented as meaningful market comparisons.
A station and fuel enters the historical series from its first genuine price-history event.
Prices below 50p or above 300p per litre are excluded as implausible.
Retailer and postcode-area comparisons require at least five like-for-like stations.
A dated permanent closure removes a station from that date onward when the source provides one.
Reading the figures
Known limitations
Fuel Price Pulse keeps these limits visible so a reader can judge each comparison in context.
The current schema does not contain a complete dated history of station activation and deactivation.
Brand, postcode-area and motorway labels use each station's current classification because dated dimension history is not yet stored.
A change event proves that a price changed. It does not prove the source checked and resubmitted an unchanged price that day.
A withdrawn fuel is not currently represented by a null history event, so its last price may carry forward until another trustworthy status signal is available.
Industry bodies may publish sales-volume-weighted figures, which can differ from an equal-station average.
Retailer, area and motorway groupings reflect the classification available when the edition is generated, subject to the limitation above.
Reproducibility
Each edition stores a frozen data payload and methodology version. Later station updates do not rewrite an older report. CSV and JSON downloads expose the figures and station counts used in that edition.
Questions and corrections
Include the edition date, fuel and metric so the published payload can be checked directly.
