+49 69 8700 4500 | [email protected]
Blog Wednesday 17th of June 2026

Six Sites, One Question: When a Caterpillar C32 Beats an SDMO D830 at ~660 kW — and When It Doesn't

author avatar
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.
Decision Framework · ~660 kW Industrial Diesel

A proof-by-cases for the awkward 600–700 kW band, where the KOHLER-SDMO D830 (750 kVA prime / 825 kVA standby ≈ 660 kW) meets the bottom of the Caterpillar C32 (830–1000 kW) and the top of the C15 (320–500 kW). No single winner — six concrete sites, six verdicts.

There is a frustrating gap in the diesel range right around 600–700 kW. The KOHLER-SDMO D830 lands squarely in it at about 660 kW standby. Caterpillar generator, by contrast, brackets it: the C15 tops out near 500 kW and the C32 starts near 830 kW. So the real decision is not "which 660 kW set" — it is whether to take the SDMO generator unit that fits the number exactly, or buy a Caterpillar that either runs a little hard at the top of a C15 or a little easy at the bottom of a C32. That choice flips by site. Here are six, each with the rule that decides it.

Case 1 — The growing colocation hall

A small edge data hall draws ~480 kW today on a near-constant IT-and-cooling load, but the operator has signed two tenants for next year that will push it past 650 kW. A C15 at the top of its band fits today and runs out of room next year.

Worked consequence. Buying to today's 480 kW means a forced second purchase and a controls re-integration in eighteen months. A C32 at the bottom of its 830 kW band carries today's load at roughly 58% and absorbs the tenant growth without a new machine. The C32's light loading also keeps it inside its best transient margin for the data hall's step loads.
Pick: Caterpillar C32. When a known load ramp will cross 650 kW within the depreciation window, buy the headroom now on one EMCP 4.2 platform rather than a perfectly-sized D830 you outgrow.

Case 2 — The fixed-load water-pumping station

A municipal pumping station runs a flat ~600 kW whenever it runs, for years, with no growth plan. The D830 sits at about 73% standby load here — comfortably in its efficient zone.

Worked consequence. A C32 carrying 600 kW sits near 72% of its 830 kW band too, so neither is loafing. But the C32 is a physically larger, costlier frame bought for headroom this site will never use. The D830 matches the duty and the budget, and SDMO's APM403-class control covers a standalone station's needs.
Pick: SDMO D830. Flat load, no growth, single set — the exactly-sized machine wins on capital and footprint. Paying C32 prices for unused headroom is wasted money.

Case 3 — The motor-heavy aggregate plant

A quarry's crushing line throws large across-the-line motor starts onto the bus — a single 200 kW crusher motor with five-to-six-times locked-rotor inrush, on top of conveyors already running.

Mechanism: voltage dip per ISO 8528-5 is the inrush reactive surge against the alternator's excitation ceiling.

Worked consequence. That 200 kW start at ~5.5× inrush is a large reactive bite. On a D830 sized to ~660 kW the surge is a serious fraction of the alternator ceiling; on a lightly loaded C32 (830 kW) the same surge is proportionally smaller, so the dip is shallower and the line's contactors hold. The headroom that looked wasteful in Case 2 is exactly what keeps this plant running.
Pick: Caterpillar C32. When the worst inrush kVA exceeds ~40% of a 660 kW alternator's rating, the larger frame's ceiling is the deciding spec — unless you fit soft starters, which sends you back to Case 2.

Case 4 — The remote site where parts arrive by truck, not next-day

A processing site sits a day's drive from the nearest major service depot. Uptime depends less on the machine than on who can reach it and how fast a part ships.

Worked consequence. Heat rejection and transient margin are identical on paper, but a coolant pump that fails on a Sunday is decided by the local service network, not the datasheet. Whichever brand has stocked parts and a trained technician within reach wins outright — and at this site that is the SDMO dealer with regional inventory.
Pick: SDMO D830 — here, because local support is stronger. The verdict would flip the instant the Caterpillar dealer network is the closer one; this case is decided by geography, not engineering.

Case 5 — The hot, air-restricted retrofit room

A genset must drop into an existing plant room with undersized louvres; measured worst-case inlet air is ~45 °C against meaningful static pressure.

Mechanism: rated output holds only if the cooling package (jacket water + charge-air + radiator/fan airflow, plus alternator losses) can reject heat at the room's real inlet temperature and pressure.

Worked consequence. A D830 sized close to its rating has little thermal margin to give up to a hot room, so it derates first. A C32 carrying the same ~660 kW at the bottom of its band already runs cooler relative to its package and tolerates the hostile room better — the same headroom logic as Case 3, applied to heat instead of inrush. But note SDMO's soundproof enclosure must be verified fitted; a quiet enclosure that chokes airflow worsens either machine.
Pick: Caterpillar C32 — the under-loaded larger frame keeps full power in the hot room. Fix the louvres and ambient drops to spec — then this collapses back into Case 2 and the D830 returns.

Case 6 — The tight-budget standby set that runs 20 hours a year

A warehouse needs a code-required standby set for a ~620 kW life-safety-plus-essential load, exercised monthly, almost never called on for a real outage.

Worked consequence. Fuel burn (≈ load × bsfc × hours) is trivial at 20 hours a year, so the low-fuel-build argument earns nothing. Transient and thermal margins matter only for the brief test. The decision is pure acquisition cost and code compliance — and the exactly-sized D830 is the cheaper, smaller machine that passes the inspection.
Pick: SDMO D830. When run hours are negligible and the load won't grow, headroom you never use is cost you didn't need.
SiteDeciding factorVerdict
Growing colo hallLoad crosses 650 kW in-windowCaterpillar C32
Flat pumping stationFixed load, no growthSDMO D830
Aggregate plantLarge inrush vs alternator ceilingCaterpillar C32
Remote siteLocal service & parts reachSDMO D830 (geography)
Hot retrofit roomThermal margin in 45 °C inletCaterpillar C32
20-hour standbyAcquisition cost, negligible hoursSDMO D830

Load percentages, inrush ratios and ambient figures are illustrative, labelled as such; published power bands, ratings and control platforms (EMCP 4.2, APM403) are manufacturer-stated.

The rule the six cases share

At ~660 kW the D830 fits the number and the C32 buys margin. Choose the C32 whenever any one of three things is true: a known load ramp will pass 650 kW inside the depreciation window; your worst inrush event exceeds 40% of a 660 kW alternator's kVA; or measured room inlet exceeds 40 °C with real static pressure. If none of those holds — flat load, mild room, soft-started motors, low hours — the exactly-sized SDMO D830 is the right machine, and local service reach can override everything above. Headroom is only worth its price when one of those three thresholds is crossed.

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.

Leave a Reply

Your email address will not be published. Required fields are marked *