The short answer
Exercising a generator with no meaningful load proves only that it starts. A staged load bank test loads the set toward its rating, holds each step, and produces recorded numbers you can compare over time. On a diesel set it also does real mechanical good by raising cylinder and exhaust temperatures enough to burn off the deposits that light running leaves behind.
What wet stacking is and why light running causes it
A diesel engine is designed to run under load. Combustion temperature, cylinder pressure and exhaust gas temperature all depend on how hard the engine is working, and the fuel injection system is tuned assuming those conditions.
Run the same engine at a small fraction of its rating for an hour a week, and combustion never gets hot enough to burn the fuel completely. Unburned fuel and soot accumulate: on the cylinder walls, past the rings, in the turbo if fitted, and most visibly in the exhaust system, where it collects as a wet black residue. That last symptom is where the name comes from.
The consequences compound. Deposits glaze cylinder walls, which reduces ring seal, which reduces compression, which lowers combustion temperature further. Injector tips carbon up and spray poorly. Oil dilution from unburned fuel reduces lubrication. A set that has been lightly exercised for years can be in worse condition than one that was simply left alone.
The correction is the same as the prevention: load. Enough load, long enough, to bring exhaust temperature up and hold it there.
How a load bank test is actually run
A resistive load bank is a bank of heating elements with a controller that lets you switch load on in defined steps. It gives the generator a clean, unity power factor load with no motors, no electronics and nothing that can be damaged by whatever the set does.
The sequence:
- Pre-checks. Oil level and condition, coolant level, air filter restriction, fuel level and quality, belt condition, battery state, exhaust integrity, and clear airflow to the radiator.
- Connect and instrument. Load bank at the output, with metering on voltage, frequency and current per leg. Temperature monitoring on coolant or head, and a clear view of the exhaust.
- Start and stabilize. Run at no load until coolant reaches operating temperature.
- Step up. Apply load in stages, commonly a quarter, then half, then three quarters, then as near full rating as the set and the conditions allow. Hold each step long enough for temperatures to stabilize, typically a half hour or more at the higher steps.
- Record at each step. Voltage on each leg, frequency, current on each leg, temperature, oil pressure, and exhaust appearance.
- Step down and cool. Reduce load in stages, then run at light load to bring temperatures down before shutdown.
The whole test takes a good part of a day on a larger set. It is not a quick check, and it should not be treated as one. Full details of what the procedure covers are on our generator service page.
Reading the results: voltage, frequency, temperature
The numbers are the point. Here is what each one is telling you.
Voltage. Watch the momentary sag when a step is applied and the steady value once settled. A large sag that recovers slowly points to the automatic voltage regulator or to excitation. A steady value that drifts as temperature climbs points to the regulator as well. Legs that differ from each other by a meaningful margin under a balanced resistive load point to a winding or connection problem.
Frequency. Frequency is engine speed. Droop on load application is normal and recovery should be quick. Frequency that keeps sagging as load increases means the engine cannot make the power, which is a fueling, air or compression problem. Hunting frequency at a steady load points to the governor.
Temperature. Coolant temperature should climb with load and then stabilize. A temperature that keeps climbing at a steady load means the cooling system cannot reject the heat, and the cause is usually a fouled radiator core, a failing fan or belt, or coolant that has lost its heat transfer properties.
Exhaust. On a set that has been wet stacking, the early part of the test produces visible black or grey discharge as deposits burn off. That should clear as the test proceeds. Exhaust that stays dirty at full load after an extended hold is telling you the deposits are past the point where load alone will fix them, and mechanical work is due.
What the test does not tell you
Be clear on the limits, because a clean test report can create false confidence.
- It does not test the transfer logic. A resistive load bank connected at the output bypasses the automatic transfer switch entirely. Transfer timing, sensing and retransfer have to be tested separately.
- It does not prove fuel quality over a long run. A few hours does not surface water contamination, microbial growth in the tank or a filter that will clog at hour twenty.
- It does not substitute for oil analysis. Bearing wear, coolant intrusion and fuel dilution show up in a sample, not on a meter.
- It does not evaluate non-linear loads. A resistive bank has a unity power factor. Real loads include motors and switching supplies, and voltage regulation under those is a different question.
- It does not check the enclosure or mounting. Vibration isolation, exhaust routing and structural mounting are separate inspection items.
Treat the load bank test as one instrument in a program rather than the whole program.
Building a service interval around the test
The test is most valuable as a repeated measurement. One report is a snapshot. Three annual reports are a trend, and the trend is what tells you a set is degrading before it fails.
A workable program:
- Monthly: a short exercise run, ideally under whatever real load is available, with a log entry.
- Quarterly: visual inspection, fluid levels, battery condition, belt and hose check.
- Annually: oil and filter change, air filter, fuel filter, coolant test, and the load bank test with a written report.
- On any test anomaly: the specific follow-up the anomaly points to, whether that is regulator work, cooling system service or fuel system attention.
Keep the reports together and compare the same numbers year over year. Voltage regulation that has widened, frequency droop that has deepened, or a coolant temperature that stabilizes higher than last time are all early signals.
For fleets running several sets, running them through together makes the comparison easier and cuts the administrative overhead. Our fleet manager page covers how multi-unit scheduling is handled.
Scheduling the test around your season
Pick the trough in your operating calendar, not a convenient week.
A generator on a service body or a mobile operation is usually least critical during whatever period the business is slowest, and that is when the set should be out of service for a day. A generator that backs up a facility should be tested when a planned outage window is available, not on the assumption that nothing will need it.
Two scheduling notes. First, book ahead, because load bank equipment and bay time both need to be reserved. Second, allow for the follow-up. A test that finds a regulator problem or a cooling issue turns a one day booking into a longer one, and it is far better to have that happen on a planned week than during your busy season.
All generator work is performed in shop at the Yorba Linda facility. Eastvale operators reach it in roughly twenty-five to thirty-five minutes by SR-71 to SR-91, and a set that can be dropped and collected separately from the vehicle makes the scheduling easier still. Rate structure for this work is published on the labor rates page, and deposit terms are on the payment terms page.