In all cases tests must be done with reference to the agreed specification of the generating set. Prior to operational tests, environmental data must be recorded and a preliminary inspection is specified. This includes safety checks, earth connections and guarding, insulation tests, fluid levels checks etc. On initial start-up emergency stop system must first be checked, followed by frequency, voltage and phase rotation checks, and an inspection for leaks and vibration.

Only after these preliminary checks are load tests started. These include load duration tests or a 'heat run', with recording of steady-state voltage and frequency followed by load acceptance tests, when transient responses to load changes are recorded.

The standard defines three performance classes - G1, G2 and G3. A further class, G4, is reserved for performance criteria which are agreed between the supplier and the buyer.

Each performance class has different criteria for a range of characteristics of the generating set. G1 is the least stringent and generally applies to small, simple generating sets intended to supply unsophisticated loads. G2 is broadly equivalent to commercially available power, while G3 is intended for sets which are powering strategically critical loads, or those which particularly require a stable and accurate power supply such as hospitals and data centres.

Engine governing is measured by testing frequency, and alternator voltage regulation is measured directly. Specified characteristics relating to frequency include steady-state variation, dip when maximum power increase is applied, rise when 100 % power is removed and the time before the frequency returns within limits in both cases. Voltage characteristics again include the permissible dip when maximum power increase is applied, the rise when 100 % load is removed and the recovery time.

Some of these criteria are as follows:

The maximum power increase for these tests, expressed as a percentage of the rated load of the set is determined by the characteristics of the engine, and the match between the engine and the alternator. Traditionally, naturally-aspirated engines were tested with 100 % load acceptance, whereas turbo-charged engines were tested with a 60 % power increase. However the standard defines a more complex formula based on engine parameters, and in practice this value is now usually determined by the manufacturer.

Other tests can also be specified, extending the scope of those set in the standards. These include cold-start load acceptance, simulated motor starting loads and fibre parallel running, for example.

Non-unity Power Factor Testing

In practice, almost all generating sets see a non-unity power factor load when in normal use. Virtually all typical loads of any but the tiniest generating sets include inductive and motor loads. Even loads such as fluorescent lighting, which have capacitive components, have ballast chokes to ensure that their power factor is near unity or slightly lagging.

Almost all but the smallest generating sets are designed and rated at a lagging power factor, usually 0.8, and virtually all set builders operate quality systems to ISO9001:2000. While this standard allows for organisations to set their own systems and procedures, it is difficult to argue against the concept that a product can be fitted with a rating plate, stating a load capacity at a power factor of 0.8 if it has not been tested at the nameplate rating.

When professional engineers and consultants are involved in specifying a power supply for a project, they are becoming increasingly aware of this, and require that a set is tested to the standards, and at the nameplate rating. This means non-unity, or resistive/inductive load testing. ISO 8528 specifies that test reports should note if tests have been done at a power factor which is different than the rated one. Usually this means that tests done with a purely resistive load can be considered incomplete.

Both quality standards and the ISO 8528 for engine-driven gen-set testing demand that complete testing is carried out. Professional engineers and consultants have responded by specifying such tests. The equipment, with control, instrumentation, data capture and analysis systems are available from leading load bank specialists such as ASCO.

There is no doubt that in the future more tests will be done to ensure that generating sets comply with specification, will accept load in service under a managed maintenance regime, and will operate in an environmentally acceptable way with optimum fuel efficiency and minimum pollution.