Load Bank Testing: How to Verify a Generator Set's Capacity Before Handover
Load Bank Testing: How to Verify a Generator Set's Capacity Before Handover
A datasheet rating and an installed generator set are not the same thing. Load bank testing closes the gap before handover: step loading to rated output with temperatures.
What You Will Learn
Why The Test Exists
A generator set is bought on a rating and commissioned against a site. Those are two different things. The rating on the datasheet is established under reference conditions; the set is installed at a specific altitude, in a specific ambient temperature, behind a specific exhaust and air path. A load bank test is the procedure that closes the distance between the two, and it does so at the only moment when everyone involved is still in the same room and the equipment is still under warranty.
The case for doing it before handover rather than after is not technical, it is commercial. A set that cannot carry its rated load at the site will reveal that during the test at a point when the supplier is still responsible for the outcome. Discovered three months later, during an outage, the same finding is a dispute about who pays.
There is also a simpler reason that gets less attention. Many failures that are attributed to the machine are actually failures of the installation: an undersized cable, a restricted air path, a fuel line that cannot deliver full-load consumption, a control panel configured for a different voltage. A load bank test finds all of those, because it forces the installation to work rather than allowing it to idle.
What a Load Bank Test Proves, and What It Does Not
Being precise about the scope of the test is what keeps it useful. It is a strong instrument, and it is often asked to answer questions it was never designed for.
What it proves.
- That the set reaches and sustains its rated output at the actual site conditions, with the ambient
correction applied.
- That the cooling system removes the heat produced at full load - the radiator, the airflow path and the
enclosure ventilation all have to work together, and only a sustained run reveals whether they do.
- That the fuel system can deliver full-load consumption, from tank to injectors, without starving the
engine at the point of highest demand.
- That the governor and the voltage regulator hold frequency and voltage within tolerance as load is
applied and removed, and that transient response on load steps is acceptable.
- That the control system, the protection settings and - where fitted - the transfer switch behave as
designed, because the test provides a real load for them to act on.
- That the installation as a whole can carry the intended load, which is the question the owner actually
cares about.
What it does not prove.
- Long-term reliability. An hour at rated load demonstrates capability, not durability.
- That the set will still perform when it is dirty, when the air filter is part-loaded or when the radiator
core has accumulated dust. That is a maintenance question, not a commissioning one.
- Fuel consumption across the whole load range. Consumption can be measured at the test points, and only
those points.
- That the specified load profile will be met over time if the site's load grows beyond what was tested.
Stating those limits in the handover documentation is part of doing the job properly, because it prevents the test report from being read as a guarantee of everything.
Choosing the Load: Resistive, Reactive, or Both
A load bank presents an artificial load to the generator, and the type of load determines what is actually being tested.
Resistive load presents unity power factor. It loads the engine fully - the kW side - and is the appropriate instrument for verifying engine output, cooling and fuel delivery. It is also the standard method for clearing wet stacking on a lightly loaded set, because it raises combustion temperature to the range where deposits burn off.
Reactive load presents a lagging power factor and loads the alternator rather than the engine. It tests the alternator's ability to sustain voltage under an inductive load, which is what a site full of motors actually presents. Without it, a test can pass at unity power factor while the alternator behaves differently on the real installation.
Combined resistive-reactive load presents the set with a load closer to the site's actual power factor. It is the most representative test and the most expensive to arrange.
The practical recommendation is straightforward: verify the engine with resistive load at full kW, and where the site's load is significantly inductive - motors, transformers, welding equipment - verify the alternator with a combined test at the site's power factor. Testing only at unity power factor and describing the set as verified is an overstatement.
How the Test Is Run
The procedure matters as much as the equipment, because an incorrectly run test produces numbers that flatter the set.
Establish the reference conditions first. Record ambient temperature, altitude, and the set's nameplate rating. Those three values determine what full load means on the day, and they belong at the top of the report rather than in an appendix.
Step the load, and hold each step. The usual sequence is 25, 50, 75 and 100 percent of rating, holding each step long enough for temperatures to stabilise. The stabilisation time is part of the test, not dead time: coolant and exhaust temperatures respond slowly, and a step that is held for two minutes tests almost nothing.
Measure at every step, and record it. The measurements that matter are exhaust gas temperature, coolant temperature, oil pressure, frequency, voltage, current, power factor and, where it is being verified, fuel consumption. Exhaust gas temperature is the single most informative reading on a diesel set: it tells you whether the engine is working as hard as the load implies, and it is the number that reveals a fuelling or air restriction problem.
Test the transitions, not only the steady states. Applying and removing load is when the governor and the voltage regulator are actually challenged. Frequency dip on load application and recovery time should be recorded, because that is what the site experiences when a large motor starts.
Verify the protection settings. Overload, over- and under-voltage, over- and under-frequency, low oil pressure and high coolant temperature - all should be demonstrated or, where that is not practical during commissioning, documented as tested with the method and the result recorded.
Test the transfer switch under load where one is fitted. A transfer switch that is only tested unloaded has not been tested. The commissioning test should include at least one transfer and re-transfer with the set carrying load.
Site Conditions and Derating
Full load at the factory and full load at the site are different quantities, and the test has to be clear about which one it is measuring.
Altitude reduces available engine output. Air density falls with height, so less oxygen enters the cylinder per stroke and the engine produces less power for the same fuelling. A set rated at sea level carries a lower rating at two thousand metres, and the derating is specified by the engine manufacturer rather than negotiated on site.
Ambient temperature does the same. A set rated at 25 °C ambient carries a lower rating at 40 °C, both because the air is less dense and because the cooling system has less temperature difference to work with.
The enclosure affects both. A canopy that is correct for a temperate site may restrict airflow enough to matter at a hot one. This is why air path verification belongs in the test rather than in an assumption.
So the report should state the correction. "The set reached rated output at site conditions" is a meaningful statement only if it says which conditions, and whether the rating tested was the nameplate figure or the derated figure for that site. Where derating applies and the site's load still fits, that is a satisfactory outcome; where it does not fit, that is exactly the finding the test exists to produce.
Commissioning Versus Recurring Testing
The two are not the same test, and conflating them wastes money in one direction and misses failures in the other.
The handover test is a full acceptance test. Stepped to 100 percent, held, measured, recorded, witnessed and signed. It answers the question the purchase depended on: can this set do what it was bought to do, here.
Recurring testing is a condition check. A period in service, at a load that experience shows is appropriate for the installation, with the measurements compared against the commissioning baseline. Deterioration shows up in the comparison, which is why the baseline has to exist. A recurring test without a commissioning record to compare against produces numbers that nobody can interpret.
The interval follows the duty. A standby set that is exercised without load needs load bank testing more often than a prime power set that carries a real load every day. A set that is tested monthly with no load is accumulating unburnt fuel and deposited carbon, which is the mechanism behind wet stacking at low load - and a periodic load test is the maintenance item that removes it.
Whoever runs it should produce the record. For recurring testing, comparing measurements against the commissioning baseline matters more than the absolute values. That comparison is only possible if the baseline was written down properly in the first place.
What the Handover Report Must Contain
A test that is not written down has limited value, because the numbers are the asset. Six sections make a report that can be used years later by people who were not present.
- The equipment and its identity. Model, serial numbers of the set, engine and alternator, hour meter
reading at the start of the test, and the nameplate ratings.
- The site conditions on the day. Ambient temperature, altitude and any relevant note about the
enclosure or the installation, including whether the rating tested was corrected for those conditions.
- The test schedule. Every load step, its percentage, its duration, and the load bank type - resistive,
reactive or combined, and at what power factor.
- The measurements. Recorded at each step, not summarised as a pass. Temperature, pressure,
frequency, voltage, current, power factor, and fuel consumption where measured.
- The transitions and the protection tests. Frequency and voltage dip on load steps, recovery times,
and the protection settings demonstrated with their results.
- Deviations and sign-off. Anything that did not meet the specification, whether it was corrected
during the test or remains open, who witnessed the test and when.
The report travels with the machine. When the set is sold, relocated or serviced by a different team, that document is the reference that makes every subsequent test interpretable.
What To Have Ready Before the Test
A load bank test that arrives unprepared will be shortened, and a shortened test is a missed opportunity.
- A load bank of adequate capacity, and enough cable to reach the set safely.
- Confirmed ratings: nameplate, and the corrected rating for the site's conditions.
- The engine manual's cold and hot checks, so the test can include whatever the manufacturer specifies.
- Fuel in the tank for the expected consumption plus margin, calculated from full-load consumption, not
from the tank's nominal capacity.
- A written test schedule agreed with whoever is witnessing, so nobody discovers on the day that a step or
a measurement was expected.
- A defined position on who does what if a defect is found, because the value of the test is precisely
that it finds defects while the responsibility is still clear.
For projects we supply, the test schedule, the site condition correction and the report format are agreed before the set ships, so commissioning day is about measurement rather than negotiation.
FAQ
What is a load bank test for a generator?
A load bank test presents the generator with an artificial electrical load so that its output can be verified. It confirms that the set reaches and sustains its rated output at the site's conditions, that the cooling and fuel systems cope at full load, and that the governor, voltage regulator, control panel and - where fitted - transfer switch all function under a real load.
Does a generator have to be load bank tested before handover?
It is the only reliable way to prove an installed set can carry its intended load at its own site conditions. Without it, the set is accepted on the strength of a datasheet rating established under reference conditions, which is a different quantity from performance at the actual altitude, ambient temperature and installation.
Resistive or reactive load bank - which do I need?
Resistive load at unity power factor tests the engine: output, cooling and fuel delivery. Reactive or combined resistive-reactive load tests the alternator under a lagging power factor, which is what sites with motors, transformers or welding equipment actually present. For a site with significant inductive load, a unity power factor test is not a complete verification.
Why did the set not reach rated output during the test?
The usual causes are altitude and ambient temperature derating, an air path restricted by the enclosure or the installation, a fuel system unable to deliver full-load consumption, or an instrument reading taken before temperatures stabilised. The report should state the conditions and the corrected rating so that the result can be judged against what was actually expected at that site.
How often should load bank testing be repeated?
The interval follows the duty. A standby set that is exercised without load needs it more often than a prime power set carrying a real load daily. For a standby installation, periodic load bank testing belongs in the maintenance schedule as a recurring item with a defined interval, alongside the hour-based service items.
Talk to Our Engineers About Your Project
For projects we supply, the test schedule, the site condition correction and the report format are agreed before the set ships, so commissioning day is about measurement rather than negotiation.
This article is for informational purposes only. Specifications and availability may vary by region. Please contact NovArk Power for a confirmed quotation tailored to your project requirements.
