Sizing hot water storage for Indian hotels: start from the draw-off, not the bed count

Every hotel hot-water enquiry we receive arrives with the same number attached: the bed count. It is the first thing the developer knows, the first thing the architect puts on the drawing, and the first thing a supplier multiplies by a litres-per-room figure to produce a quotation. It is also, on its own, close to useless for sizing a heat-pump system.

The reason is straightforward. A heat pump is not a geyser. A geyser has a large element and a small tank, so it answers demand by brute force — it can raise the temperature of its contents quickly and it does not much care when the demand arrives. A heat pump has a modest, steady heat output and relies on storage to bridge the gap between what it can make per hour and what the building wants per minute. Size the plant against a daily total and you will be right about the energy and wrong about the shape. The shape is what breaks.

What the bed count actually hides

Consider two 120-room properties in the same city with identical occupancy. The first is a business hotel next to an airport. Its guests check out between 05:30 and 07:00 to make morning flights, and very nearly all of them shower in a ninety-minute window. The second is a leisure resort where guests drift down to breakfast between 08:00 and 11:00 and shower whenever they feel like it.

Same beds. Same litres per day. The business hotel's peak draw-off is roughly three times the resort's, and if you size both from the same rule of thumb one of them runs out of hot water on its second morning of operation. We have been called to that building. It is not a pleasant commissioning visit, and the fix — retrofitting storage into a plant room that was drawn without space for it — costs considerably more than getting the number right on paper.

Laundry compounds this. An in-house laundry running two shifts is a large, predictable, schedulable load, which makes it the single most useful thing in the whole model: it can usually be moved. A kitchen is the opposite — smaller, but stubbornly tied to service times that nobody will shift for the sake of a plant room.

The method

We build a draw-off profile in hourly buckets across a design day, then size two things against it independently.

Storage covers the worst contiguous stretch where demand exceeds what the plant can deliver. Not the daily total — the deficit. You integrate the shortfall across the peak and that integral, plus a margin for stratification losses and for the fact that a cylinder is never usefully empty, is your minimum useful volume.

Plant is sized to refill that storage inside the recovery window before the next peak. For a business hotel that window is generous: nothing much happens between 09:00 and 17:00. For a hospital, where there is no quiet hour, it is not, and the plant grows accordingly.

Doing it this way routinely produces a smaller heat pump and a larger cylinder than the litres-per-room approach. That trade is almost always the right one in India. Storage is cheap, passive, has no moving parts and no failure mode more exciting than a leaking flange. Compressor capacity is expensive, draws demand charges, and spends most of its life cycling at part load if you have bought too much of it.

A worked profile for the airport hotel above comes out roughly like this:

{
  "peakWindow": "05:30–07:00",
  "peakDrawLitres": 8400,
  "plantOutputLitresPerHour": 2600,
  "deficitLitres": 4500,
  "storageSelected": 6000,
  "recoveryWindowHours": 4.5
}

The 6,000-litre selection is not the deficit rounded up for comfort. It is the deficit plus the unusable bottom of the cylinder plus an allowance for the return leg of the circulation loop, which brings water back at a temperature that is neither hot enough to serve nor cold enough to be good for the heat pump.

The circulation loop is part of the load

That last point deserves its own paragraph, because it is the most commonly omitted item in a hotel hot-water model. A dead-leg policy that guarantees hot water at the tap within a few seconds means a recirculation pump running continuously, and every metre of that pipework loses heat. In a large property the standing loss on the loop can be a meaningful fraction of the total daily energy — we have measured properties where it exceeded the guest draw-off overnight, because at 03:00 there is no guest draw-off at all and the loop is still turning.

For a heat pump this matters twice over. It is load, and it is return temperature. A loop returning at 48 °C into the bottom of a cylinder destroys the stratification the system depends on and pushes the heat pump to operate against a warmer source than it needs to, which costs you COP precisely when you can least afford it. Getting the return injected at the right height in the cylinder, or into a dedicated tempering volume, is worth more than a percentage point or two of nameplate efficiency.

What we ask for

If you send us a hotel to size, the useful inputs are these, roughly in order of how much they change the answer:

• The check-out pattern — even approximate. "Mostly corporate, early flights" moves the number more than an exact room count.
• Laundry: in-house or outsourced, and if in-house, the shift timings and the machine list.
• Kitchen covers and service windows.
• Pool and spa, if any, with surface area and whether the pool is covered overnight.
• The incoming cold-water temperature by season. In Jaipur this swings from around 12 °C in January to near 30 °C in June, and the lift the heat pump has to provide swings with it. A system sized on an annual average is undersized for four months of the year.
• The plant-room envelope and the roof area available for outdoor units, because the answer has to physically fit.

Notice what is not on that list: the bed count in isolation. It appears, but as one input among several rather than as the whole model.

Where the rules of thumb came from

None of this is a criticism of the engineers who wrote the litres-per-room figures. Those numbers came out of a world of gas boilers and electric calorifiers, where plant was cheap to oversize, storage was small, and recovery was fast enough that the shape of the demand genuinely did not matter much. The rule was a reasonable abstraction for the technology it described.

Heat pumps break the abstraction because their economics invert it. The thing you are trying to minimise is no longer installed capacity but operating hours at bad conditions, and the lever for that is storage and scheduling, neither of which a per-room figure can express.

So the rule of thumb has not become wrong so much as it has become a rule about a different machine. Keep it for the first conversation, by all means — it is a useful sanity check and it tells you whether a project is a 10 kW problem or a 300 kW one. Just do not let it reach the purchase order.

Hero photo via Unsplash.