We audit a lot of hospitality MEP briefs. Across the last 30 hotel projects we've reviewed in Sri Lanka and the Maldives, the requested standby genset capacity averaged 38% larger than what the actual demand profile justified. That's expensive — both as capex and as a lifetime of part-loaded generators running below their efficient operating range.
Below are the four assumptions that almost always inflate the number, and the diligence that reins them back in. None of this is exotic — it's just patient load-list arithmetic that gets skipped when the brief moves fast.
1. Connected load is treated as running load
The single most common error: tallying the nameplate kW of every motor, every chiller, every kitchen oven, every lift drive, and asking the genset to cover all of it simultaneously. In a 200-room hotel, the connected load might be 1,800 kVA, but the actual coincident running load — what the gear is drawing at the same moment under realistic occupancy — is usually 800 to 1,100 kVA. This is the diversity factor, and it's the lever that recovers the most capex.
Hotel diversity factors we see in practice: kitchens 0.45–0.55, guest-room HVAC 0.60–0.70, BOH lighting 0.55–0.65, lifts 0.30–0.40 (only one or two of four typically active in standby mode). Apply those before sizing.
2. Cold-start inrush is double-counted
A second-stage error: assuming all motor loads start simultaneously the moment the genset takes load. They don't. With a properly designed ATS sequence, you stagger major loads — chillers first, kitchen second, room AC third — so the inrush spike never coincides. The step-load the alternator actually has to swallow is much smaller than the sum of all locked-rotor currents.
For a Teksan or Force MTU set with full alternator step-load capability, sequenced starting typically allows a 25–30% reduction in the kVA that would otherwise be specified to handle a single worst-case simultaneous start.
3. "Spare" capacity gets stacked on top of an already-padded load
Once a load list has been padded for diversity errors, doubled for inrush, and uplifted by a 20% future-growth allowance, the resulting kVA recommendation is often 60% larger than what the hotel will ever draw. We always ask: where, specifically, does growth come from? If the answer is "we might add a banquet hall in five years," the right move is a riser with breaker capacity for a future genset paralleling — not a single oversized set running at 30% load forever.
4. The brief skips load shedding
Most hotel briefs don't differentiate essential from non-essential loads on the standby side. They should. With a basic load-shed scheme, banquet kitchens, swimming-pool heaters, decorative lighting and laundry can be dropped during a genset-only event. That alone usually cuts the standby kVA by 15–20% with no impact on guest experience.
What right-sizing looks like
For the 200-room four-star hotel we audited last quarter, the original brief called for 1,250 kVA prime + 500 kVA backup. Our recalculated demand, with sequenced starting, realistic diversity, load shedding for non-essentials and N+1 redundancy on a single set rather than two, came out to 800 kVA prime — a 36% reduction. The owner saved about US$95,000 in capex and another estimated US$8,000/year in fuel because the set runs closer to its efficient operating point during weekly tests and any extended outages.
The takeaway
Standby capacity is one of the few line items in a hotel MEP brief where rigour at the design stage compounds for the next 20 years. If your project is at the load-list stage, ask your MEP team to walk you through the diversity assumptions, the starting sequence, and what loads will be shed under a genset-only condition. If those three things aren't documented, the kVA number on the brief is almost certainly wrong.
NativeWay sizes generators across Sri Lanka, the Maldives and Bangladesh — for hotels, hospitals, factories and high-rises. If you'd like a second opinion on a load list, the team will audit it without charge. Get in touch.
