"Why does my 1000 kW airport set keep failing the witnessed block-load test — and would the other brand pass?"
An airport's new standby set — feeding the airfield lighting regulators and the terminal's life-safety loads — keeps dipping too far on the witnessed step test. The operator wants to know whether the constraint is the brand, or something upstream of it. The honest answer follows a chain.
This question feels like a brand question — Cummins generator QSK family versus Caterpillar generator C32 (830–1000 kW), both legitimate at ~1000 kW — but it almost never is at the start. A failed block-load test is the end of a chain of constraints, and replacing the engine before you've walked the chain often just buys an identical failure. Let me propagate the constraint from the witnessed dip backward to its real cause, then answer the brand part honestly.
Stage 1 — What the test actually measures
A witnessed block-load test under ISO 8528-5 applies a defined step and checks two things: how far voltage and frequency dip, and how fast they recover within class. The airfield lighting constant-current regulators are sensitive to the frequency notch; if it goes too deep, they fault. So the failing number is a transient depth, not a steady-state capacity problem.
Stage 2 — Propagate to the alternator and excitation
First link: sub-transient reactance and excitation
The depth of the voltage dip is set largely by the alternator's sub-transient reactance and how hard the excitation holds field during the step. A permanent-magnet-excited alternator sustains field current even as terminal voltage collapses; a self-excited scheme leans on residual voltage and recovers slower.
If the installed set uses a self-excited alternator and the airfield regulators trip on the dip, the constraint is excitation, not the engine brand. Cummins QSK sets are commonly specified with PowerCommand 3.3 driving the field; a C32 with EMCP 4.2 likewise depends on its alternator/excitation spec. Decision: before changing brands, check whether the failing set could be re-specified with PMG excitation and a faster AVR. That single change may pass the test on the same brand.
Stage 3 — Propagate to the fuel system
Second link: how fast torque arrives
The frequency notch depth is set by how fast the engine finds torque before speed sags. Modular Common Rail under PowerCommand 3.3 commands rail pressure on a fast digital loop; the C32's electronic control does likewise. A mechanically-governed engine, by contrast, would lag.
If the frequency dip — not the voltage dip — is what fails, the constraint is governor-plus-fuelling speed. Decision: confirm the engine is in its electronic-control configuration and the governor is tuned for the regulators' tolerance. If it already is, and the dip still fails, you have reached the genuine platform-difference question.
Stage 4 — Propagate to sizing, then answer the brand question
Third link: is the step simply too big for the frame?
If excitation and fuelling are both right and the test still fails, the step exceeds what a 1000 kW frame can hold within class — the constraint is sizing. The fix is a larger frame or a staggered start of the regulators, on either brand. Only once the step is within a 1000 kW set's capability does the brand comparison become decisive: at that margin, the platform whose integrated control gives the tightest, most repeatable transient — and whose AmpSentry protection holds through the regulators' inrush — passes with more headroom. Decision: that integrated-control margin is where the Cummins QSK tends to lead; a C32 with equivalent PMG excitation and tuned EMCP 4.2 can also pass, but with less integration headroom to spare.
Topology/standards per the cited standards; all product ratings are manufacturer-stated values from the cited datasheets, current to 2026-06; derived/illustrative figures are labelled as such. This is not an independent head-to-head test. Cummins is a brand affiliated with this site; competitor names are used for identification only.
