Decision Framework: What Flips the Runtime Choice

This framework compares Caterpillar generator (industrial diesel) and KOHLER-SDMO generator (industrial diesel) based on three verifiable dimensions. The goal is not to declare a universal winner but to isolate the provenance-epistemic conditions — where the rating comes from and under what rule it was derived — that determine which machine keeps running when the load is real and the power is off.

1. The Rating Provenance Trap: Prime vs Standby

The number. Caterpillar diesel gensets are published with both prime and standby ratings. For example, a Cat C15 is rated 320–500 kW (standby) but its prime rating is lower — typically ~90% of standby. SDMO units like the D275 are listed as 250 kVA prime and 275 kVA standby. The mechanism. ISO 8528 defines prime power as the maximum power a generator can supply continuously (unlimited hours) at a variable load, with an average load not exceeding 70% of the prime rating. Standby rating is for the duration of an interruption at an average of 70% of the standby rating. If a spec sheet only shows standby kVA, the true continuous runtime at 80% load may be 30-40% shorter than you'd assume. Worked consequence. A facility that buys a "275 kVA" SDMO D275 and runs it at 220 kVA (80% of standby) for 48 hours is actually operating above the prime threshold — the engine oil temperature rises, fuel consumption increases, and the genset may derate or shut down on high coolant temperature after ~12 hours. A Caterpillar C15 sized at 400 kW standby but run at 320 kW (80% of standby) would face the same trap unless the buyer explicitly picked the prime rating. Reversal. For short-duration emergency runs (under 6 hours) that occur fewer than 50 hours per year, the standby rating is the right comparison — and both Caterpillar and SDMO deliver their full standby runtime without derating.

2. The Real-Load Runtime Killer: Transient Load Acceptance

The number. Caterpillar's EMCP 4.2 controller manages load steps and transient response, enabling the generator to accept a 30% block load in one step with a voltage dip under 15%. SDMO units with APM303 control can accept similar step loads, but the engine governor response (mechanical vs electronic) varies by model. The mechanism. Under ISO 8528-6, a generator must recover frequency and voltage within 5 seconds after a load step. If the governor is slow (e.g., mechanical on some SDMO models), a sudden motor start (like a 100 HP irrigation pump drawing ~90 kVA starting) can cause the frequency to sag below 47 Hz — triggering undervoltage protection and a shutdown. That stop is not a "runtime failure" in the fuel-exhaustion sense, but it kills runtime instantly. Worked consequence. Assume a 200 kVA load profile that includes two 50 HP motors starting simultaneously (starting kVA ~180 kVA each). A Caterpillar C15 with EMCP 4.2 and electronic governor will accept the combined transient and recover within 2 seconds. An SDMO D275 with mechanical governor may take 4-5 seconds — and if the voltage dip exceeds the protective relay setting, the unit locks out, ending the run. The runtime difference is not hours — it's zero vs 50+ hours. Reversal. In a load profile with only resistive or ramp-up loads (e.g., lighting, constant-speed pumps with soft starters), transient acceptance is irrelevant. Both generators will run uninterrupted until fuel runs out.

3. The Fuel-Consumption Table: Who's L/h Figure Is Actually Yours?

The number. Caterpillar publishes fuel consumption curves at 25%, 50%, 75%, and 100% of prime rating, for example ~62 L/h at 75% load on a C15. SDMO lists fuel consumption at 75% load for the D275 at about 55 L/h. The mechanism. The provenance of these figures is the ISO 8528 test cycle, which uses a fixed resistive load bank at constant ambient (25°C) and altitude (sea level). Real installations see higher backpressure (exhaust piping, silencers) and higher jacket-water temperature, which raise fuel consumption by 5–15% above the test-cycle value. Worked consequence. A facility at 1,500 m elevation (e.g., a mine in the Andes) that takes the SDMO D275's 55 L/h at face value will plan for a 24-hour run on a 1,500 L tank. Actual consumption at that altitude (derated by ~3.5% per 300 m above 900 m) may be 65–70 L/h — meaning the generator runs out of fuel at hour 20 instead of hour 27. Caterpillar's EMCP 4.2 displays real-time fuel rate, so the operator sees the drift and can adjust the tank schedule. Reversal. For installations at sea level with short exhaust runs and a generous tank (e.g., 2× calculated consumption), the published L/h figures are accurate enough. Both brands will meet the stated runtime within reasonable tolerance.

Non-Obvious Insight

Failure Mode / Counterexample

What breaks the framework: If the load is purely resistive (e.g., electric heating) and the altitude is sea level, both generators will run to the fuel exhaustion curve without derating. The only differentiator is the control panel's ability to report fuel consumption in real time — Caterpillar's EMCP 4.2 provides that; some SDMO models may not. But for a 2-hour run in a data-center standby scenario, that difference is immaterial.

Bottom-Line Threshold

If your continuous run exceeds 8 hours at an average load above 70% of the standby rating, specify the prime rating. If your load includes motor starts exceeding 30% of the generator's kVA rating per step, verify that the governor is electronic (not mechanical). If neither condition applies, the runtime difference between Caterpillar and SDMO is negligible — choose based on local service support and control-panel features.