Blog Wednesday 17th of June 2026

Four Things a Bakery Plant Manager Believes About 900 kW Diesels — Tested Against the Physics

Jane Smith I’m Jane Smith, a senior content writer with over 15 years of experience in the packaging and printing industry. I specialize in writing about the latest trends, technologies, and best practices in packaging design, sustainability, and printing techniques. My goal is to help businesses understand complex printing processes and design solutions that enhance both product packaging and brand visibility.

Table of Contents

Myth vs Reality · 900 kW Continuous-Process Diesel

Caterpillar generator C32 (830–1000 kW) against a comparably sized Perkins generator 4000-series set (600–1800 kW), at the ~900 kW tier that keeps a continuous-bake line moving when the grid drops. Four common beliefs, four corrections, each with the threshold that decides the buy.

An industrial bakery never really stops. Proofers hold temperature, the tunnel oven coasts on huge thermal mass, spiral coolers and mixers cycle around the clock. When the utility fails, a ~900 kW genset has to catch a running process mid-bake — and the folklore that surrounds these machines on the plant floor is mostly half-right, which is the most dangerous kind. Here are four beliefs, weighed against what the C32 and a Perkins 4000-series set actually do.

Myth 1

"The bigger the kW number, the more heat it dumps into my plant room."

The intuition feels obvious — more power must mean more heat — so the room gets sized off the kW figure. It is the wrong variable.

Reality. A genset's heat rejection is not its electrical output. It is the sum of three loss paths that must leave the building: jacket-water heat off the block, charge-air-cooler heat after the turbo, and radiator-and-fan airflow, plus the alternator's own copper and iron losses. Two ~900 kW sets can reject meaningfully different amounts of heat depending on engine efficiency and cooling-package design. You size the room to the published heat-rejection and combustion-air figures, never to the kW.

Worked consequence — drives the buy. Size the bakery's louvres and ventilation off "900 kW" and you may under- or over-build the room by a wide margin. Ask both Caterpillar (C32) and the Perkins 4000 packager for jacket-water heat, charge-air heat, radiated heat and combustion-air flow at your rating, then size the room to the larger set of those numbers. The machine that publishes lower total heat rejection for the same output needs less louvre area and less make-up air — a real construction-cost difference, independent of the kW both share.

Myth 2

"At 900 kW both sets will swallow my biggest motor start without blinking."

Because the running load fits comfortably, people assume transients are a non-issue. Transients live in a different budget than running load.

Reality. Block-load behaviour is graded by ISO 8528-5 and set by the alternator's excitation ceiling and the engine's governor/turbo response — not by spare running kW. A large dough mixer or a spiral-cooler fan bank started across-the-line pulls five-to-six-times its running current as reactive inrush. The Perkins 4000 is specifically tuned for high load acceptance on standby; the C32 is published at prime and standby ratings with documented dip behaviour. Both are capable — but only against a step they were sized for.

Worked consequence — drives the buy. Say the largest single step is a 180 kW mixer drive starting onto a bus already carrying ~500 kW of ovens and proofers. Across-the-line, its inrush kVA can momentarily demand far more excitation than its 180 kW running figure suggests. If either set is specified near its ceiling, voltage dips far enough to drop sensitive line-control PLCs — and a dropped PLC mid-bake is scrapped product, not a flicker. The buy decision: get the dip-and-recovery curve for that exact start from both brands, and size to the worst step, not the running total.

Myth 3

"A standby-rated set can just run the line for the whole outage."

Standby and prime get used interchangeably in conversation, so a standby-rated machine gets pressed into hours of continuous duty.

Reality. The ratings are a discipline, not a label. Caterpillar's standby rating is defined for the duration of a normal-source interruption at an average load near 70% of the standby rating — it is not an unlimited-hours prime rating. A bakery that loses grid for many hours, or runs the genset as planned prime power during peak tariffs, is in prime-rating territory, where the continuous-duty number is lower.

Worked consequence — drives the buy. If your outages are short and rare, buy on the standby rating and the C32's 830–1000 kW standby band or the 4000's standby figure covers you. But if the line ever runs on the genset for extended periods — long regional outages or deliberate prime operation — you must size on the prime rating, which means a larger frame for the same 900 kW continuous. Misreading this is the single most expensive sizing error at this tier, and it is identical between the two brands.

Threshold. If the genset will ever carry the line for more than the duration of a typical utility interruption — or run as planned prime power — size on the prime rating, not standby, for both machines.

Myth 4

"They're both diesels at 900 kW, so controls and integration are a wash."

Engines dominate the conversation; the control platform gets treated as an afterthought. On a process plant it is often the deciding difference.

Reality. Caterpillar fields the EMCP 4.2 control as a single defined platform — consolidated metering, diagnostics and management on one interface. A Perkins 4000 set is built by a gen-set packager around the engine, so its control platform varies by who assembled it. With the C32 you buy a documented control ecosystem; with the Perkins you buy an engine whose surrounding controls depend on the packager.

Worked consequence — drives the buy. A bakery integrating the genset into its plant SCADA, or planning to add a second set to parallel later, wants one consistent, documented controller. A standardised EMCP 4.2 fleet makes that a configuration task; a mix of differently packaged Perkins controllers can make it an integration project with new gateways and re-commissioning. If SCADA integration or future paralleling is on the roadmap, this is where the C32 earns its keep. For a single set that only starts, carries and stops, a competently packaged 4000 with a good third-party controller does the whole job — and the integration premium is wasted.

Belief	What actually governs it	Same or different across brands?
kW = heat in the room	Jacket-water + charge-air + radiated heat + air flow	Different — compare published heat-rejection figures
Big set = easy starts	Alternator ceiling + governor/turbo vs worst inrush	Same physics; verify each set's dip curve
Standby = run all outage	Prime vs standby rating discipline	Same rule; both must be sized on prime for long runs
Controls are a wash	Single platform (EMCP 4.2) vs packager-dependent	Different — decisive if integrating/paralleling

Step loads, inrush ratios and load figures above are illustrative, labelled as such; published power bands, ratings, the standby-load definition and EMCP 4.2 are manufacturer-stated.

The decision that survives the four myths.

Two beliefs are simply wrong (kW≠heat; big≠transient-proof) and you correct them the same way on both machines — by demanding the heat-rejection and dip-and-recovery numbers. The other two split the brands. Rule: if the genset will ever run the line longer than a normal utility interruption, size both on the prime rating. Then, if SCADA integration or a future paralleled second set is on your roadmap, choose the Caterpillar C32 on EMCP 4.2 for one documented control platform. If it is a single standalone set with no integration ambition, a well-packaged Perkins 4000 is the right machine and the control premium buys nothing. Below roughly 500 annual run hours, fuel economy decides none of it.

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. Caterpillar is a brand affiliated with this site; competitor names are used for identification only.

Share: LinkedIn Twitter WhatsApp

Four Things a Bakery Plant Manager Believes About 900 kW Diesels — Tested Against the Physics

"The bigger the kW number, the more heat it dumps into my plant room."

"At 900 kW both sets will swallow my biggest motor start without blinking."

"A standby-rated set can just run the line for the whole outage."

"They're both diesels at 900 kW, so controls and integration are a wash."

Leave a Reply

Recent Articles