Popular claim — and why it’s misleading
“A Caterpillar generator runs longer on a full tank than a Cummins generator of similar rating.” This shows up in contractor circles and occasionally in tender notes — the idea that Cat iron has a lower specific fuel consumption at the same load, so you stretch runtime. It sounds plausible because Caterpillar’s 3500-series was originally developed for continuous mining and marine, where thermal efficiency mattered. But the statement collapses when you fix the single variable that actually governs runtime: real load as a fraction of the nameplate standby rating. No other twist matters as much.
In this piece we walk three dimensions — fuel consumption at a given load point, tank capacity vs. rating, and control of parasitic load — and show how the “longer Cat runtime” myth survives only when you compare mismatched ratings or ignore the duty cycle effect. The conclusion is a single decision rule.
Dimension 1: Fuel consumption at 70% load — the “real” load point
NFPA 110 and ISO 8528 both accept that a standby generator runs at an average load of about 70% of its standby rating during a utility outage. At that point, many gensets operate near their brake-specific fuel consumption (BSFC) sweet spot. Let’s look at published numbers for two units that often compete in the 2 MW class.
| Parameter | Cummins QSK60 (2000 kW standby) | Caterpillar 3516 (1750 kW standby) |
|---|---|---|
| Standby rating (60 Hz) | 2000 kW | 1750 kW |
| Prime rating (60 Hz) | 1800 kW | 1600 kW (approx., based on 90% of standby) |
| Fuel consumption at 70% load (illustrative) | ~385–400 L/h (roughly 102–106 gal/h) at 70% of 2000 kW = 1400 kW load | ~345–365 L/h (approx. 91–96 gal/h) at 70% of 1750 kW = 1225 kW load |
| Specific fuel consumption at 70% load | ~0.275 L/kWh (illustrative, based on 385 L/h ÷ 1400 kW) | ~0.285 L/kWh (illustrative, based on 350 L/h ÷ 1225 kW) |
Fuel consumption values are manufacturer-stated or derived from datasheet curves; labelled illustrative where rounded.
Mechanism: The Cat 3516 at 70% of its lower standby rating (1225 kW vs. 1400 kW) burns fewer litres per hour — that’s simple arithmetic. But the specific consumption (L/kWh) of the QSK60 is slightly lower (roughly 0.275 vs. 0.285), meaning the Cummins engine converts fuel into electrical output with marginally better thermal efficiency at that load point. The difference stems from the QSK60’s Modular Common Rail (MCRS) fuel injection, which permits higher injection pressure and more precise timing across the load range. The Cat 3516 uses unit injectors with mechanical actuation (EUI), which is robust but not as flexible at part load.
Worked consequence: If you have a 1400 kW real load (say a data center with two UPS modules plus chillers), the Cummins QSK60 runs at 70% of its standby rating and delivers about 385 L/h. The Cat 3516 cannot sustain 1400 kW continuously at its standby rating (1750 kW × 0.7 = 1225 kW, but the engine is rated for a maximum standby of 1750 kW — it can handle 1400 kW, but that’s 80% of its standby rating, which exceeds the typical 70% average for sustained use). At 80% load, the Cat’s specific consumption rises to about 0.295 L/kWh (illustrative), yielding about 413 L/h. So the Cummins actually burns less fuel per hour at 1400 kW than the Cat. That inverts the myth.
When this dimension reverses: If the real load is below 1000 kW (e.g., a 900 kW load), both gensets run well below their sweet spot. At 45% of rating, the Cat 3516’s specific consumption may degrade to ~0.32 L/kWh, while the QSK60’s MCRS maintains ~0.30 L/kWh. The gap narrows. But the Cat still burns fewer litres per hour because it’s delivering less kW (900 vs. 900 — same load). At identical load, the Cummins had a slight edge in efficiency. The myth holder would need to claim “Cat always runs longer,” but that only holds if the Cat’s tank is larger or the load is matched to its lower rating. That’s not a myth — it’s a selection bias.
Dimension 2: Tank capacity vs. rating — the arithmetic trap
Popular belief: “A Cat 3516 has a larger standard fuel tank, so it runs longer before refuelling.” Let’s check the standard skid tank options.
| Standard tank option | Cummins QSK60 | Caterpillar 3516 |
|---|---|---|
| Integral day tank (skid-mounted) | ~1500 L (approx. 400 gal) | ~2000 L (approx. 530 gal) |
| Double-wall sub-base tank | Up to 4000 L (1060 gal) optional | Up to 5000 L (1320 gal) optional |
| Runtime at 70% load (integral tank) | ~3.9 h (1500 L ÷ 385 L/h) [derived] | ~5.8 h (2000 L ÷ 345 L/h) [derived] |
| Runtime at 70% load (max optional tank) | ~10.4 h (4000 L ÷ 385 L/h) [derived] | ~14.5 h (5000 L ÷ 345 L/h) [derived] |
Runtime derived from tank capacity ÷ illustrative fuel consumption at 70% load; see note above.
Mechanism: The Cat 3516 standard skid tank is about 33% bigger than the Cummins standard tank (2000 L vs. 1500 L). Combine that with the Cat’s lower throughput at its own 70% load (345 L/h vs. 385 L/h), and you get a 50% longer runtime at the Cat’s natural operating point. This is the root of the myth: the Cat package is configured with a larger tank because its engine family (C15/3500) was originally designed for prime power in remote mining, where 8–12 hour shifts between refuels were standard. The QSK60, in contrast, is often spec’d for mission-critical standby in refineries and data centres where an external day tank or fuel farm is common, so the integral tank is smaller.
Worked consequence: If you buy the QSK60 with its standard skid tank and load it to 1400 kW, you get 3.9 hours — less than four hours. For many standby codes (NFPA 110, Level 1, Type 10), you only need 2 hours of fuel on site; 4 hours is more than adequate. But if you’re prime-power in a grid-tie reversal scenario, 4 hours may not cut it. The Cat with standard tank at 1225 kW gives ~5.8 hours — enough for an overnight shift. This is a real advantage for remote prime power, not for standby.
When this dimension reverses: The Cat’s larger tank is an advantage only if you accept its lower rating. If the real load is 1400 kW, the Cat’s consumption jumps to ~413 L/h (see Dimension 1), and runtime on a 2000 L tank drops to ~4.8 hours — still more than the QSK60’s 3.9 hours, but the gap shrinks. If you max the optional tank on both, the Cat’s advantage is 14.5 vs. 10.4 hours — but in practice, a 5000 L tank is a separate line item, not a “standard” feature. The myth collapses when you compare the same load class with comparable tank volumes. The real question: are you comparing the package as sold, or the engine as a component? The myth exploits the package difference.
Dimension 3: Parasitic loads and control logic — the hidden throttle
Another myth layer: “Caterpillar’s EMCP 4.2 controller runs fewer parasitic loads, so more fuel goes to the generator.” Let’s check the actual power draw of the control systems.
Mechanism: The Caterpillar EMCP 4.2 control board consolidates management, diagnostics, and metering with a typical draw of about 50–75 W. The Cummins PowerCommand 3.3 digital control includes automatic start/stop, AmpSentry protective relay, and paralleling interfaces — typical draw is about 60–80 W. Both are negligible relative to a 1500–2000 kW load (0.004% of rating). The real parasitic load comes from radiator fans, fuel pumps, and battery chargers. On the Cat 3516, the fan drive is belt-driven from the engine, consuming about 30–40 kW at full speed (roughly 2% of rating). The QSK60 uses a similar belt-driven fan, but offers an optional variable-speed electric fan kit that reduces fan power to ~15 kW at part load.
Worked consequence: At 70% load, the Cat’s fixed fan consumes ~35 kW continuously; the QSK60 with variable-speed fan draws ~15 kW. That 20 kW difference is about 1.4% of the 1400 kW load. Over a 10-hour runtime, that’s 200 kWh of energy that must come from fuel. At the QSK60’s specific consumption (0.275 L/kWh), that translates to about 55 L of extra fuel burned by the Cat every 10 hours — roughly 0.5% of total fuel. Again, tiny. Not a runtime game-changer.
When this dimension reverses: In extreme ambient temperature (50°C), both fan systems run at full speed, eliminating the variable-speed advantage. Then the parasitic difference disappears. But the myth that “Cat’s controller saves fuel” is not supported by any datasheet. The EMCP 4.2 does not significantly reduce fuel draw.
Decision rule (single variable funnel): which generator runs longer?
Take the real continuous load (kW) you must supply.
If real load ≤ 1200 kW: A Cat 3516 (1750 kW standby) will run longer than a QSK60 (2000 kW standby) because it operates closer to its sweet spot and its standard tank is larger. The myth holds — but only because you selected a smaller-rated machine for the load.
If real load > 1200 kW: The QSK60 matches or beats the Cat on fuel efficiency (lower L/kWh), and with optional tanks the runtime gap narrows to ~10–15%. The myth becomes a selection artifact.
Threshold: 1200 kW is the crossover point for these two specific models. Below it, Cat wins on runtime; above it, Cummins does — or at least ties.
This rule applies to standby or prime power with a fixed fuel tank. With external fuel farms, runtime is essentially unlimited for either — then the myth is meaningless.
Failure mode and non-obvious insight
Non-obvious insight: The myth persists because it compares the Cat 3516 (a 1750 kW machine) against the QSK60 (a 2000 kW machine) with the standard tank. If you do a like-for-like comparison at 1750 kW (using the QSK60’s prime rating of 1800 kW), the Cummins’ higher efficiency (0.275 vs. 0.285 L/kWh) means it burns about 3.5% less fuel per kWh. Over a 10-hour run at 1225 kW, that’s about 42 L less fuel for the QSK60 — but the Cat’s larger standard tank (2000 L vs. 1500 L) still gives it longer absolute runtime because the tank volume difference (~33%) dominates the efficiency difference (~3.5%). The real variable is tank size, not engine efficiency. The myth inverts only when you equalize tank volume. Most buyers never specify equal tanks — they take the standard package, so the myth self-reinforces.
Failure mode: If the generator is used for grid support at low loads (e.g., 10–20% of rating for hours), both engines suffer poor specific consumption (>0.35 L/kWh). The larger engine (QSK60) throttles worse because it has a higher idle fuel rate. At 200 kW load, the QSK60 burns roughly 70 L/h, the Cat 3516 burns about 60 L/h — the Cat wins on runtime again, but both are horribly inefficient. The myth fails to consider that no operator runs a 2 MW genset at 200 kW for extended periods; they would choose a smaller unit. The myth holds only in the wrong application.
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.
