Why Your Generator Burns More Fuel Than Expected

Diesel generators are often said to use more fuel than the datasheet suggests. Sometimes the engine really is burning too much. More often the expectation was wrong, because the datasheet figure applies at a load the set never reaches, or the measurement is unreliable. This guide takes the usual causes in order, load, machine, measurement, and ends with a fuel test you can trust.

Litres per hour versus litres per kWh

Two numbers describe fuel use, and confusing them causes many disputes. The fuel rate in litres per hour (L/h) tells you how fast the tank empties. The specific fuel consumption (SFC) in litres per kWh (L/kWh), or grams per kWh, tells you how efficiently the engine turns fuel into electricity. A set at light load burns fewer litres per hour than at full load, but more litres for every kWh it delivers.

Modern diesel sets typically achieve a full-load SFC in the region of 0.21 to 0.25 L/kWh. The site's calculators use 0.24 L/kWh as a default. The exact figure for a given engine is in its manufacturer's fuel consumption table, usually quoted at 25, 50, 75 and 100 % load.

Cause 1: low load factor and the part-load curve

A diesel engine has fixed losses that do not shrink with load: friction, pumping, cooling fan and auxiliary drives. At light load those losses are a larger share of the fuel burned, so SFC rises. The default part-load curve in our fuel consumption calculator expresses that as a multiplier on the full-load SFC:

Compare 25 % and 75 % load on the same 100 kW set. At 75 % it burns 18.5 L/h, more than twice as much per hour as the 8.4 L/h at 25 %. Per kWh, though, the picture reverses: 0.336 L/kWh against 0.247 L/kWh, so every unit of energy costs about 36 % more fuel at 25 % load. This curve is generic and deliberately conservative; it is not any manufacturer's data, and real engines differ.

So an expectation based on the full-load SFC is always too low for a lightly loaded set. At a site averaging 25 % load, the set will use about 40 % more than the load multiplied by 0.24 L/kWh, with nothing wrong with the engine.

Cause 2: an oversized set

Light load is often a sizing decision. Suppose a site has a steady 30 kW load. On a 100 kW set it runs at 30 % load; on a 40 kW set it runs at 75 %:

On the default curve the smaller set uses roughly 23 % less fuel for the same work. The real difference depends on the two engines' own full-load SFC, which may not be identical, so compare manufacturers' tables. Oversizing usually comes from adding nameplate ratings, adding margin twice, or sizing for a motor start that a soft starter or variable-speed drive could have reduced. The generator sizing calculator shows whether running load or starting governs.

Cause 3: wet stacking at light load

When a diesel engine runs for long periods at light load, combustion temperatures stay low. Some fuel does not burn completely, and it combines with soot and lubricating oil to form a dark, oily residue in the exhaust system, around the turbocharger and on the valves and injectors. This is called wet stacking. Signs include black, oily drips at exhaust joints, persistent smoke when load is applied, and rising oil consumption. Over time it fouls injectors and glazes cylinder liners, which itself makes combustion worse and raises fuel consumption further.

Manufacturers commonly advise against prolonged operation below roughly 30 % of rated load, but the limit and the recommended remedy differ between engines, so the specific maker's operation and maintenance manual is the reference. Remedies include periodic running at higher load, using a load bank, reducing the set size, or moving to cyclic operation with a battery, where the set runs at a healthy load and then stops.

Cause 4: maintenance and fuel condition

• Air filters. A blocked filter starves the engine of air, so combustion becomes rich, smoke increases and fuel per kWh rises. Dusty sites need more frequent changes than the standard schedule.
• Injectors and fuel system. Worn or fouled injectors atomise fuel poorly. The engine needs more fuel for the same output and often smokes or knocks.
• Fuel quality and water. Water enters tanks through condensation, poor seals and contaminated deliveries. It promotes microbial growth, blocks filters and damages injection equipment. Drain water separators regularly and keep tanks full and sealed where practical.
• Cooling. A radiator clogged with dust, a slipping fan belt or a failed thermostat makes the engine run too hot or too cold. Either moves it away from its designed combustion conditions.
• Intake and exhaust restriction. Hot intake air from poor ventilation or recirculation, and excessive exhaust back pressure, both reduce engine efficiency.

Cause 5: confusing kVA with kW

The engine supplies real power, so fuel follows kW, not kVA. A 100 kVA set rated at PF 0.8 is an 80 kW machine. If 40 kW is treated as 40 % of "100", the true load factor of 50 % is understated, and the wrong point on the curve is used. The reverse mistake also happens: a controller showing 50 kVA at PF 0.6 means only 30 kW, so an estimate based on 50 kW expects too much fuel. A low power factor does add some current-related losses in the alternator and cables, which the engine must supply, but its main effect is to use up kVA capacity rather than fuel. The kVA to kW converter helps keep the two apart.

Cause 6: measurement errors, leaks and theft

• Dipstick versus flow meter. A dipstick reading converted with a rough tank chart is easily wrong by several percent, especially near the top and bottom of the tank. A fuel flow meter is more accurate, but the return line matters: many engines send surplus fuel back to the tank, so a meter on the supply line alone reads high unless the return flow is subtracted.
• Tank shape. In a horizontal cylindrical tank, a centimetre of level near the middle holds much more fuel than a centimetre near the bottom. A linear conversion from level to litres gives wrong results. Use the manufacturer's calibration table or a calculated one.
• Temperature expansion. Diesel expands roughly 0.08 % for every degree Celsius. Fuel delivered cool and measured after a hot day in an outdoor tank will show more volume than was delivered: 1000 L warming by 25 °C gains about 20 L. Over a short test, temperature swings can hide or exaggerate a real difference.
• Leaks and theft. Weeping joints, a leaking tank or fuel removed from an unattended site all appear as "consumption". Remote fuel-level monitoring that logs sudden drops while the set is stopped is a practical way to detect them.

How to run a proper fuel test

The aim is a measured L/kWh figure over a real operating period, which you can compare with the manufacturer's data at the same average load.

1. Choose a period long enough to make reading errors small, ideally a week or more of normal operation.
2. At the start, record the tank level at a known temperature, the controller's cumulative kWh counter and the running-hours counter. Note any fuel added during the period.
3. At the end, record the same readings at a similar time of day, so that fuel temperature is comparable.
4. Fuel used = start volume + fuel added − end volume. Energy = end kWh − start kWh. Average load = energy ÷ running hours.
5. Compute L/kWh and compare it with the expected SFC at that average load factor.

Here is an illustrative calculation for a 100 kW set:

A result like this points to the machine or the measurement rather than the load: check the filters, injectors and fuel system, verify the tank chart and look for leaks. Had the measured figure been close to 0.336 L/kWh, the engine would have been behaving normally and the real issue would be the low load factor. Enter your own measured load and the manufacturer's fuel table in the fuel consumption calculator for a like-for-like comparison, and use the electricity cost calculator to put the cost per kWh in context.

Summary

1. Judge fuel use in L/kWh at the actual load factor, not in L/h against the full-load figure.
2. Expect noticeably higher SFC at light load, and avoid oversizing.
3. Treat prolonged light-load running as a wet stacking risk and follow the maker's guidance.
4. Keep filters, injectors, fuel and cooling in good order.
5. Use kW, not kVA, for fuel estimates.
6. Measure carefully over a long enough period, using the controller's kWh counter and a calibrated tank chart or a correctly installed flow meter.

This guide is general education. Engine-specific limits and maintenance intervals come from the manufacturer's documentation.