Skip to content
allanninal.dev/build
Collaborate

Part 5 of 7 · Fuel log auditor series ~5 min read

What the fuel data says about the vehicles

The queries are the visible output and the vehicle findings are the valuable one. Fuel consumption is a continuous measurement of how hard a vehicle is working to do the same job, which is as close to a health check as a fleet gets for free.

Key takeaways

  • A sustained economy decline in one vehicle is a maintenance signal, not a driver signal.
  • Utilisation findings are often larger than efficiency ones.
  • Cost per mile needs unattributed fuel in it, or it is understated.
  • Report per vehicle over time, never as a ranked table.
  • The best outcome is a booked service, not a conversation with anybody.

Decline as an early warning

One van's seasonally adjusted fuel economy declining over four quartersA bar chart with four bars showing van five's seasonally adjusted miles per gallon. January to March: thirty-four point one. April to June: thirty-three point four. July to September: thirty-two. October to December: thirty point six. A note says that is ten per cent over a year, steadily, and nobody noticed from the driver's seat.020406080~34.1Jan-Mar~33.4Apr-Jun~32Jul-Sep~30.6Oct-DecVan 5, miles per gallon, seasonally adjustedTen per cent over a year, steadily. Nobody noticed from the driver's seat.
Fig 1. One vehicle over a year with the seasonal effect removed. A gradual decline of this shape is invisible day to day and is one of the clearer signals a fleet produces.

Ten per cent of a van’s fuel bill is real money, and the causes are usually cheap: underinflated tyres, a clogged filter, a dragging brake, a sensor that has drifted. All of them are found at a service and none of them announce themselves.

The output is therefore a maintenance booking rather than a conversation, which is worth emphasising because the instinct on seeing declining economy is to look at who has been driving it. In a fleet with stable driver assignments a slow decline in one vehicle is almost always the vehicle.

Utilisation

Three vans with very different annual mileages and the same fixed costsThree boxes stacked on the left. Van two: twenty-four thousand miles, labelled heavily used. Van six: eleven thousand miles, labelled average. Van nine: two thousand four hundred miles, labelled barely used. All three converge on Same fixed costs covering insurance, tax and depreciation, and that leads down to Van nine's cost per mile is ten times van two's. A note says the efficiency question is usually smaller than the do-we-need-it question.Van 224,000 milesheavily usedVan 611,000 milesaverageVan 92,400 milesbarely usedSame fixed costsinsurance, tax,depreciationVan 9's cost per mileis ten times van 2'sThe efficiency question is usually smaller than the do-we-need-it question.
Fig 2. Three vehicles by annual mileage. The largest saving available in most small fleets is not efficiency; it is a vehicle that is not needed.

Fuel data reveals this almost as a side effect, because a vehicle that is barely being fuelled is barely being used. It is often a genuinely uncomfortable finding — the spare van that somebody insists is needed for peak weeks — and it is a question worth asking with numbers attached.

Cost per mile, honestly

Two things get left out and both understate it. Unattributed fuel is real fuel and it belongs in the fleet total even when it cannot be assigned to a vehicle. And fixed costs — insurance, tax, depreciation, the maintenance from Day 118 — are usually larger than fuel and are frequently reported separately.

A cost per mile that includes only fuel is a number that makes every vehicle look cheap and makes the utilisation finding invisible, which is exactly the finding that matters most.

What the report says

The quarterly page

  • Fuel purchased: £18,400 across 214 transactions.
  • Unattributed: £900, 4.9%, down from 11% last quarter.
  • Queries raised: 19. Answered: 17. Data errors corrected: 14 of those.
  • Vehicles with a sustained decline: 1 — van 5, down 10% over the year, service booked.
  • Utilisation: van 9 covered 2,400 miles at a fully-loaded cost of £1.94 per mile.
  • No driver rankings appear in this report, deliberately, and this line stays here so nobody adds one.

The third line is the honest summary of what this system mostly does: fourteen of nineteen queries were data errors. That is not a disappointing result, it is the expected one, and every one of those corrections improves the figures that produce the fourth and fifth lines.

The second line is the process metric. Unattributed fuel falling from eleven per cent to five per cent means the matching and the queries are working, and it is the number that determines whether anything else on the page can be trusted.

The line that stays in the template

The last item is a small piece of institutional memory. Somebody will ask for a driver league table roughly once a year, and a line in the standing report explaining why there is not one answers the question before it is asked, with the reasoning already written down.

Next: what all of this costs to run.

All posts