What Is Block Loading Capacity in a Generator — and Why It Matters

Ask most buyers what size genset they need and you will get an answer in kVA. Ask what happens in the first two seconds after the mains fail and the room usually goes quiet.

That gap is where a surprising number of expensive installations go wrong. A set can be correctly sized for everything the site draws, pass its factory test, run beautifully on a load bank — and still trip the moment a fire pump starts on a dead bus. The rating was never the problem. The problem was how much load the set could swallow at once.

What block loading actually means

Block loading is the application of a large load in a single step, rather than gradually. The genset is running unloaded or lightly loaded; a contactor closes; and a substantial chunk of demand appears instantaneously.

Block loading capacity is how much of that the set can take. It is normally quoted as a percentage of the prime rating for the first step — so a 500 kVA set with 60% first-step acceptance can absorb roughly 300 kVA in one block, then more once it has recovered.

The distinction that matters:

• kVA rating answers how much can this set carry?
• Block loading capacity answers how much can it be handed at once?

A set will happily carry a load it cannot accept in one step. That is not a contradiction, and it is the single most common misunderstanding in genset specification.

Why the engine cannot simply take it

When a block of load lands, the engine immediately slows. The governor detects the falling speed and calls for more fuel. On a naturally aspirated engine that mostly works — the air is already there, the extra fuel burns, torque rises, and speed recovers.

On a turbocharged engine it does not, at least not straight away. The turbo is driven by exhaust gas, and there is not much exhaust gas until the engine is already working harder. For a second or two the engine has been asked for fuel it cannot burn, because the air has not arrived yet. Torque lags, speed keeps falling, and frequency dips further and for longer than it would on an NA engine.

That single mechanism explains most of what follows. Highly rated engines — more kW squeezed from each litre of displacement — lean harder on the turbo and generally accept smaller steps. Modern emission-compliant engines run tighter, leaner air-fuel control, and step-load response is frequently what gets traded away to meet the limits. It is entirely normal for a current CPCB IV+ set to accept a smaller first block than the older, dirtier engine it replaced at the same kVA.

What "acceptable" means: ISO 8528-5

There is a standard for this. ISO 8528-5 defines how a generating set must behave during transients, and sorts sets into performance classes — you will see G1, G2 and G3 quoted most often.

Each class sets limits on four things:

1. How far frequency may deviate when load is applied or removed
2. How quickly frequency must recover to within a steady-state band
3. How far voltage may deviate
4. How quickly voltage must recover

G2 is the ordinary commercial default and is adequate for most building services. G3 is the one to specify where connected equipment is genuinely sensitive to excursions — telecom, IT, medical and process loads. The classes tighten as the numbers go up, and a set that meets G3 will generally cost more, because meeting it means a better governor, a better alternator, or a larger engine than the kVA alone would suggest.

The practical instruction is simple: ask which class the quoted set meets, and get the actual dip and recovery figures rather than the class label alone. A supplier who can produce a load acceptance table for that exact model is telling you something useful. One who answers in kVA is not.

Where this bites, in real Indian installations

Block loading is not an exotic edge case. It is the normal condition on a mains-failure start, because everything the changeover switch reconnects arrives at the same instant.

Fire pumps are the hardest case in most buildings. They are large, they are direct-on-line by design in many installations, and they must start on a dead bus — that is the entire point of them. A fire pump is frequently the single load that decides the genset size for a hotel or hospital.

Chillers and large AC compressors hit hard on restart, particularly if several are configured to restart together after an outage.

Lifts in a residential tower restart simultaneously, and each one is a motor start.

Cold storage is deceptive. The compressors that were running when the power failed all want to restart at once, into a system where head pressure has not equalised, which makes their starting demand worse than the nameplate suggests.

UPS systems are their own category of trouble. A UPS with a large rectifier step can present a substantial block, and — more painfully — many UPS units watch the incoming frequency and simply refuse to accept genset supply if it wanders outside their window. The result is the complaint every engineer has heard: the genset is running fine but the UPS won't take it. That is almost always a transient response problem, not a fault in either machine.

Fuel stations are a small-scale version of the same thing. Dispensing pumps, an air compressor and forecourt lighting are individually modest, but they are not coordinated with each other — the compressor starts whenever a customer pulls in for air, on top of whatever else is running.

What going wrong looks like

The failure is rarely dramatic. It usually reads as an intermittent, unexplained nuisance:

• Underfrequency trip. The set starts, takes the load, and shuts down on a low-frequency alarm. Often blamed on the panel.
• Contactors dropping out. Voltage dips far enough that coils release, so loads disconnect themselves mid-step — which sometimes rescues the set and makes the fault look random.
• Motor starters tripping on the largest motor while everything else stays up.
• UPS refusing to transfer, or transferring and immediately reverting to battery.
• Repeated failed starts where the set holds on a light night load and fails during the day.

The tell is that all of these are load-dependent and step-dependent, not duration-dependent. A set with a genuine capacity problem struggles after hours of running. A set with a block loading problem fails in the first two seconds and then behaves perfectly.

How to size for it

Start by finding the worst single step, not the total.

1. List the loads that arrive together on a mains-failure start. That is your first block, and it is decided by your changeover and control scheme, not by your connected load.
2. Identify the largest single motor in that block and use its starting demand, not its running demand. Direct-on-line starting draws several times full-load current.
3. Take the design peak as busy-hour running load plus that largest start, and compare it against the set's stated first-step acceptance — not against its kVA rating.
4. Check the derating for your site's ambient temperature and altitude. Acceptance figures are quoted at reference conditions, and a set in a hot, poorly ventilated plant room does not meet them.

Then reduce the step if you can:

• Stage the starts. A step-start sequence bringing essentials on first and large motors in later, on timers or a PLC, is the cheapest fix available and frequently removes the problem entirely.
• Soft starters, star-delta or autotransformer starting cut motor inrush substantially.
• VFDs cut it dramatically, at the cost of harmonics that need their own attention.
• Oversize the alternator, not the engine, if the symptom is voltage dip rather than frequency dip. Voltage behaviour during a step depends heavily on the alternator's subtransient reactance.

Where a single unavoidable load dominates and cannot be staged, the honest answers are a bigger set or two sets in parallel. Running two sets so that one is available to absorb a step is a legitimate reason to have both online — which is a different question from whether both need to run all the time, covered separately here.

Five questions worth putting in writing

Before signing off on a set, get these in the quotation rather than in conversation:

1. The load acceptance table for that specific model — percentage of prime rating per step, not a generic brochure claim.
2. The ISO 8528-5 performance class the set meets.
3. Frequency dip and recovery time at your stated worst-case step.
4. Voltage dip and recovery time at the same step.
5. Whether those figures apply at your site's ambient temperature and altitude, or at reference conditions.

Two sets quoted at the same kVA, from the same brand, in the same enclosure, can differ substantially on all five. That difference is invisible on the rating plate and decisive in the plant room.

The short version

The kVA rating is a steady-state answer to a steady-state question. Almost nothing about a mains failure is steady state. If your site has a fire pump, a chiller bank, lifts, cold storage or a UPS, the number that determines whether the changeover succeeds is not the rating on the nameplate — it is how much load the engine can absorb in the first instant, and how fast it recovers.

Size for the step, not just the sum. Then ask for it in writing.

Load acceptance figures, performance classes and derating vary by engine, alternator, governor and site conditions. The guidance here describes the method — always obtain the specific figures for the exact model quoted, from the manufacturer or an authorised dealer, before making a purchase decision.

If you are still working out what your total demand looks like, our genset load calculator sizes from an equipment list and accounts for the starting surge of the largest motor.