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

One Spec Pulls the Rest: Tracing a 700 kW Genset Decision Through Its Own Constraints — Caterpillar C32 vs SDMO D830

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.
Comparison Teardown · ~700 kW Industrial Diesel · How one constraint forces the next

Like-for-like at roughly 660–750 kW: a lower-band Caterpillar C32 (published 830–1000 kW) against the KOHLER-SDMO D830 (750 kVA prime / 825 kVA standby — call it ~600 kW prime, ~660 kW standby). The point of this teardown is not four independent scores. It is to show how fixing one number forces your hand on the next, until the whole spec is decided by a single root choice.

Most genset comparisons read like a checklist: kW here, fuel there, cooling somewhere else, as if each line could be optimised on its own. On a real ~700 kW install for a paper-mill auxiliary bus, that is not how it behaves. The decisions are chained. Pick a duty class and you have fixed the usable kW; fix the usable kW and you have fixed the cooling load; fix the cooling load and you have fixed the room and the fuel curve. So instead of scoring the C32 and the D830 dimension by dimension, this teardown follows the chain — and the brand that wins is the one that lets the chain resolve cleanly.

link_1_→_sets_everything_downstreamduty_class:_prime_or_standby_decides_the_usable_kw_before_you_compare_badges">

Link 1 → sets everything downstreamDuty class: prime or standby decides the usable kW before you compare badges

Mechanism. A genset publishes more than one number for the same hardware. Caterpillar generator's standby rating is defined for the duration of a utility interruption at an average load near 70% of that standby rating — it is not a continuous figure. The prime rating, available for unlimited annual hours at variable load, sits lower. SDMO generator splits the D830 the same way: 750 kVA prime, 825 kVA standby. So the very first question — how many hours a year, at what duty — silently sets which column of the datasheet you are even allowed to read.

Worked consequence — drives the buy. A paper-mill auxiliary bus that must hold during every grid dip and ride out multi-hour interruptions, several hundred hours a year, is a prime duty, not standby. Read that way, the D830's usable number is its ~600 kW prime figure, and a lower-band C32 must likewise be specified on its prime rating, not the headline 830–1000 kW standby band. If you compared the C32's standby figure against the D830's prime figure you would "prove" a gap that does not exist. The buy decision this forces: declare the duty class first, in writing, and demand both vendors quote the same class. Everything below this line inherits the kW you just fixed.
When this reverses. If the bus is genuinely emergency-only — a few exercise hours plus rare short outages, average load under 70% — then standby is the correct class, the usable kW jumps, and a smaller, cheaper frame covers the same connected load. The chain restarts from a higher kW and a lower price. Misjudging this one link is the most expensive error in the whole exercise, because every downstream number is sized off it.
link_2_→_inherits_the_kw_from_link_1transient_headroom:_the_fixed_kw_now_caps_how_big_a_step_you_can_take">

Link 2 → inherits the kW from Link 1Transient headroom: the fixed kW now caps how big a step you can take

Mechanism. Once duty class fixes the usable kW, the transient budget is no longer free — it is a fraction of that fixed number. How large a sudden load step the set absorbs without voltage or frequency sagging past ISO 8528-5 limits depends on the alternator's excitation ceiling and the engine's torque rise, both scaled to the rating you locked in Link 1. Choose a tighter prime rating and you have also, without a separate decision, tightened the step you can survive.

Worked consequence — drives the buy. The mill's worst step is a large refiner motor or a recovered section line restarting across-the-line — a reactive inrush of roughly five-to-six times running current for a second or two. On a ~600 kW prime machine, that surge can momentarily demand a large slice of alternator capacity. Because Link 1 already spent your headroom on continuous duty, the same step that a standby-rated machine would shrug off can now drop voltage far enough to drop the section contactors. The decision: hand both the C32 and the D830 your exact worst start and require the dip-and-recovery curve at the prime rating you fixed above — not at the standby figure, which would flatter both. On the C32 you can step up within the 830–1000 kW family to buy back ceiling without changing control platforms; on the D830 you move to the next SDMO frame.
When this reverses. Put soft starters or VFDs on the refiner drives — common on modern mill retrofits — and inrush collapses toward 150–200% of full-load current. The transient link goes slack, the kW you fixed in Link 1 is no longer threatened by steps, and the chain's pressure moves downstream to heat and fuel.
link_3_→_inherits_the_kw_and_the_prime_dutyheat_rejection:_continuous_prime_load_fixes_how_much_heat_the_room_must_swallow">

Link 3 → inherits the kW and the prime dutyHeat rejection: continuous prime load fixes how much heat the room must swallow

Mechanism. A diesel set sheds waste heat by three paths that all must leave the room — jacket water off the block, the charge-air cooler after the turbo, and radiator-and-fan airflow — plus the alternator's own copper and iron losses. The size of that heat load is set by the load you run continuously, which Link 1 fixed at the prime rating. A prime duty means the cooling package works near its ceiling for hours, not the brief excursions a standby set sees. The rated kW only holds if the package can dump that heat at the room's actual inlet temperature against the static pressure the ducting imposes.

Worked consequence — drives the buy. Two nominally 700 kW-class sets behave differently in a hot mill hall at, say, a 45 °C inlet with a partly fouled louvre. The package specified for that ambient holds its prime rating; the one specified to a 25 °C test cell quietly derates — and because Link 1 committed you to running near prime for hours, that derate is not a transient, it is a thermal trip building over minutes on a long outage. The action: spec the cooling package on both machines to the measured worst-case hall inlet and external static pressure, and get the derate-versus-ambient curve in writing at the prime rating. SDMO offers soundproofed enclosures across the range; an enclosure that quiets the set but chokes airflow tightens this link further, so verify the rating with the enclosure fitted. Note how this link could not even be evaluated until Links 1 and 2 had fixed the duty and the kW.
When this reverses. Mount the set outdoors in a weatherproof enclosure breathing cool ambient, or in a louvred plant room sized for the package, and airflow headroom is abundant. The heat link goes slack despite the prime duty, and the chain's last pressure point is fuel.
link_4_→_the_end_of_the_chainfuel:_the_prime_hours_and_the_fixed_load_set_the_only_number_that_compounds">

Link 4 → the end of the chainFuel: the prime hours and the fixed load set the only number that compounds

Mechanism. Diesel burn is roughly load multiplied by brake-specific fuel consumption (bsfc). Link 1 fixed the load near prime; the duty fixed the hours in the hundreds per year. bsfc is not flat — most diesels are best per kWh near 70–85% load and thirstier at low part-load. So once everything upstream is locked, fuel is the one cost that keeps growing every hour the set runs.

Worked consequence — drives the buy. Suppose the mill's continuous draw averages ~520 kW on a ~600 kW prime machine — roughly 87% load, near the efficient zone. Across several hundred prime hours a year, even a couple of percent difference in bsfc at that point, times 520 kW times the hours, is a meaningful annual diesel line — illustrative, but it compounds for the life of the set. The decision the chain finally lands on: ask both vendors for the fuel table at your fixed load point, not at 100%. Caterpillar publishes C32 builds optimised either for low fuel consumption or low emissions; the low-fuel build pays back precisely because Link 1 committed you to high prime hours. Had the duty been standby, this link would carry almost no weight — proof that fuel's importance was decided four links upstream.
When this reverses. If the duty were emergency standby with a few hours a year, fuel economy would settle nothing and acquisition cost plus local service reach would close the deal — quite possibly for a well-supported D830. The same hardware, a different Link 1, and the whole chain inverts.
Chain linkCaterpillar C32 (lower band, ~600 kW prime)KOHLER-SDMO D830 (750/825 kVA)
Duty class (root)Quote prime; 830–1000 kW band is standby — don't read across classes750 kVA prime / 825 kVA standby — match the class to the C32
Transient (inherits kW)Dip curve at the prime rating; step up in-family for ceilingSame physics; move to next SDMO frame for ceiling
Heat (inherits duty)Spec package to measured hall inlet + static pressureSame; verify rating with chosen soundproof enclosure fitted
Fuel (chain end)Low-fuel build pays back under high prime hoursConfirm bsfc table at your fixed prime load point
ControlsEMCP 4.2 — consolidated metering & diagnosticsAPM403 standard on larger units (auto/manual, V & fuel metering)

The 45 °C inlet, 520 kW average, 87% load, inrush ratios and step figures above are illustrative, labelled as such for like-for-like reasoning; published power bands, prime/standby ratings, the 70%-standby-load definition and control platforms are manufacturer-stated.

Decision rule

Resolve the chain in order, never out of it. First declare the duty class: if average load exceeds 70% of rating or annual run-hours clear roughly 250, you are buying prime — read the prime column on both the C32 and the D830 and never compare a standby figure to a prime one. Then require the dip-and-recovery curve at that prime rating for your worst single start; if its inrush exceeds 40% of prime kVA, fit soft starters or step up a frame. Then spec cooling to the measured hall inlet, and only then weigh fuel — choosing the Caterpillar C32 low-fuel build with EMCP 4.2 when prime hours are high and you want in-family headroom, or a well-supported SDMO D830 when the duty is genuinely standby and acquisition cost plus local support decide it. Get Link 1 wrong and every number below it is wrong too.

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 *

Recent Articles