Heat Load Calculation: The Work Behind a Capacity Figure

A heat load calculation produces a number with its assumptions attached: a rate of heat removal for one space, at the hour its load peaks. Detached from the assumptions, that number carries almost nothing. The working is the part worth having.

No method is a plain sum of what arrives. Heat entering a space does not all reach the air at once: some lands on floors, walls and furniture, then returns over the hours that follow. ASHRAE's own wording is that instantaneous heat gains at a given time do not necessarily equal the cooling load then.

Two published methods handle that delay, both ASHRAE's. The heat balance method resolves the energy balance at every room surface; the radiant time series method is a simplified form derived from it. ASHRAE calls the second suitable for peak design load work, not for annual energy simulations.

Standards also exist for what any acceptable method must cope with. ANSI/ASHRAE/ACCA Standard 183 covers peak load calculations in buildings other than low-rise housing; Manual J is the ANSI-recognised standard for equipment sizing loads in homes. Naming one of these tells you which discipline the figure came out of.

Accuracy has a ceiling, and the standard states it openly: even with reasonable procedures, the calculation can never be more than a good estimate of the actual load. The value lies in a disciplined estimate whose inputs are written down, replacing a guess with a position that can be checked.

Every calculation opens with a stated pair of conditions, indoor and outdoor. The indoor side is chosen: ASHRAE's procedure lists the room temperature to hold, the humidity to hold and the outdoor air rate. Those three belong to the designer, and each moves the answer.

The outdoor side is a percentile taken from a weather record. ASHRAE publishes climatic design conditions at annual frequencies of 0.4, 1.0 and 2.0 per cent, exceeded on average 35, 88 and 175 hours a year. Each derives from decades of station record, not a single hot afternoon somebody remembers.

That percentile decides when the system is allowed to fall behind. Size to the 1.0 per cent condition and the weather beats it for roughly 88 hours in an average year. Tightening the percentile buys those hours back and costs capacity in every other hour. Someone made that trade, and a competent proposal says which way.

Singapore publishes its own design conditions, and for commercial premises they sit inside the ACMV code. That document sets the outdoor condition and comfort range a mechanical design is worked to, with its current edition updated against the latest climate projections. What SS 553 governs is covered separately.

Homes fall outside that code, which leaves household sizing without a local reference. A contractor either adopts published values or assumes them without saying so; the second is common and rarely volunteered. Asking which outdoor condition the sizing assumed is fair, and the presence of an answer is itself informative.

Design conditions also explain honest disputes between quotes. A different indoor target, percentile or outdoor air rate moves the result, and nobody has to be careless. The gap only becomes a problem when nobody can trace it to a stated input. Comparing cooling capacity without comparing conditions says nothing.

The input list is long, and its length is the point. ASHRAE's procedure calls for building materials, component size, external surface colours and shape, read off the plans. Then location, orientation and external shading, followed by the indoor conditions, the outdoor air rate, and schedules for lighting, occupancy, equipment and appliances.

Those inputs fall into three classes, and the difference decides how much a figure can be trusted. Some are facts lifted from a drawing: areas, orientation, construction. Some are the designer's choices: indoor condition, outdoor percentile, ventilation rate. The rest are assumptions about how people use the space.

The third class is where two honest quotes separate. Occupancy, appliance use and operating hours are forecasts no drawing settles: a bedroom running as a home office carries a different equipment load from the same room used only for sleeping. A calculation that names its occupancy assumption can be argued with; one that hides it can only be accepted.

Outdoor air enters the calculation twice, under two names. Ventilation is air brought in deliberately, at a rate somebody sets; infiltration is air leaking through gaps, which nobody sets and every flat has. ASHRAE lists the two as separate components, and Manual J counts them separately as well.

Envelope construction is the input most often skipped on household work. It describes how readily the structure passes heat inward, a property of the building rather than the room. Two flats with identical layouts in blocks of different vintage do not share it, and nobody can supply it from a floor plan alone, which is why the survey exists.

A load calculation is a search across hours, and the answer is the largest result it finds. ASHRAE's methods are expressed as hourly summaries, reflecting 24-hour input schedules and profiles for each load variable: every component is analysed across those hours, then read for its maximum.

The hour that produces the maximum shifts with orientation. ASHRAE notes that a peak driven by solar gain can land when outdoor air temperature is nowhere near its highest, so west-facing and east-facing spaces reach their worst hour at different points. Summing only at the hottest outdoor hour will miss one of them.

Summing every room's peak overstates what a building needs, because those peaks land at different hours and never occur together. ASHRAE puts it plainly: a zoned system needs no greater total capacity than the largest hourly sum of simultaneous zone loads across a design day. A building peak is smaller than the total of its parts.

This bites in a flat served by one outdoor unit and several indoor ones. The rooms are rarely all at their worst in the same hour, and sizing the condenser by adding the indoor units buys for an hour that may never arrive. The surplus is paid for at purchase, then again in how the system behaves below its rating.

None of this hourly work happens by hand on a site visit. It runs in software, and the skill sits in what goes into the tool and how the output is read. Asking whether a contractor uses a load program is weak, because almost everybody has one; asking what was typed into it separates them.

Margin gets added at several points, and the multiplication is the problem. Someone rounds the occupancy up, someone rounds the appliance load up, the outdoor condition is nudged to the harsher percentile, and the equipment is selected one model above the calculated figure. Each step reads as caution on its own.

ASHRAE's instruction is unusually blunt: all load calculation inputs should be as accurate as reasonable, with no safety factors. Stacking safety factors at several levels, it says, produces an inflated and oversized load. ACCA handles the risk one step later, with size limits in Manual S.

Four modest cushions can leave a figure well clear of the real load, and the unit that follows is chosen for a space nobody lives in. Every downstream decision inherits the inflated number, from the outdoor unit to the pipe run to the electrical provision. What excess capacity does to a room is dealt with under oversizing.

A quote seldom shows its working, so read the questions that came before it. Inputs arrive by asking, and a calculation cannot exist without them. If nobody asked which way the openings face, which floor the flat is on, or how many people use the space, the figure came from a table, not the building.

Two questions separate an estimate from a calculation, and neither is confrontational: what outdoor condition did the sizing assume, and what occupancy and operating hours went into it? A contractor working from a method has both answers within reach; one working from floor area has neither. An estimate described honestly as an estimate is still useful.