Residential Thermal Load, Taken Apart

What the Number on the Proposal Is Made Of

Three tons. Sixty thousand BTU per hour. The figure that lands on a proposal is a sum, and every term in that sum comes from a different physical mechanism with its own sensitivity to bad input. Conduction through opaque assemblies tracks area and assembly U-factor. Glazing tracks area, U-factor, orientation and shading. Infiltration tracks envelope leakage and the pressures that drive air through it. Occupants and plug loads add heat whatever the weather does. Solar gain through glass peaks at a different hour than the outdoor air. Two errors that happen to cancel will still hand you a total that is right for the wrong reasons, and a component breakdown that cannot tell you how much air each room needs.

A load calculation answers a narrower question than most homeowners assume. Design conditions are statistical: an outdoor temperature exceeded only for a small percentage of hours in a typical year, drawn from the climatic design data in ASHRAE Handbook - Fundamentals and carried into ACCA Manual J. Equipment sized to that condition will run flat out on the hottest afternoon of a hot summer and may fall slightly behind for a few hours. That is the intent, not a failure. Sizing instead to the peak-ever temperature - the record high, the once-a-decade cold snap - buys a handful of uncomfortable hours per decade and pays for them with poor part-load behaviour for the entire service life of the equipment. When a homeowner asks why the unit is not oversized just in case, that trade is the answer.

The Envelope: Conduction Through Everything Solid

Opaque assemblies are the easiest component to compute and the easiest to get quietly wrong. The error is rarely arithmetic. It is the takeoff: gross wall area entered where net area belongs, a bonus room over the garage counted as conditioned when its floor is not insulated, a cathedral ceiling averaged into a flat attic. It is also the assembly value. Nominal batt R-value is not assembly U-factor. Studs at sixteen inches on centre, headers, corners and partition intersections drive a continuous thermal bridge through the wall, and an R-19 label on the insulation does not survive contact with framing fraction. Manual J construction numbers already carry that adjustment; a spreadsheet built from insulation labels does not.

Below-grade and ground-coupled surfaces behave differently again. A basement wall does not see outdoor design temperature. It sees ground temperature that lags the season, varies with depth, and differs between the top course and the footing. Slab-on-grade loss concentrates at the perimeter, which is why slab edge insulation matters far more than anything under the middle of the pour. Crawlspaces sit wherever the vents and the insulation put them: a sealed, insulated crawlspace is inside the thermal envelope and a vented one is not, and the same house calculates differently depending on which it is. Walk the perimeter and look before assigning a construction number.

Installation quality is a real input, not a scruple. Batts compressed around wiring, gaps at rim joists, insulation stopped short at the top plate - each degrades a nominal assembly by an amount worth carrying explicitly. Grading the installation is a judgement call, and being honest about it on paper beats adding an unnamed fudge factor at the end that you cannot defend when your load comes back higher than the number the customer got from someone else.

Glazing: Two Loads Through One Opening

Windows carry two loads at once. Conduction through glass and frame runs on U-factor and behaves like a very poor wall. Solar gain runs on solar heat gain coefficient, area, orientation, latitude and whatever shades the glass: overhangs, adjacent buildings, mature trees, interior blinds. An NFRC label on the unit gives U-factor and SHGC directly. A replacement window with no label is a guess, and a guess on a house with a large west elevation is an expensive one.

Orientation is where the cooling peak actually comes from. East glass loads the morning, west glass loads the late afternoon and evening, and the whole-house peak seldom lands at the same hour as the peak for any single room. That mismatch is the entire argument for room-by-room work: a west-facing bedroom can need close to twice the airflow its floor area suggests, and a system sized correctly overall will still leave that room hot if the branch ducts were proportioned by square footage. Overhangs cut summer gain on south glass far more than on west glass, because the sun is low in the west at the hour that matters, and that detail disappears the moment glazing is entered as one lumped area.

Infiltration and Ventilation: Measured Air Versus Assumed Air

Air leakage carries the widest spread between assumption and measurement of any component. A blower door gives a measured result at fifty pascals - CFM50, or ACH50 once divided by volume - and that measurement is repeatable, defensible, and takes about an hour. Converting a fifty-pascal result to a natural infiltration rate at design conditions needs a correlation accounting for climate, building height, shielding and terrain. The correlation is approximate, but it starts from a measurement rather than from a category.

Assumed leakage, by contrast, is a tick box: tight, average, loose. Two houses of the same vintage answering to the same description can differ by a factor of three in measured leakage, and infiltration is often a large share of the heating load. On a retrofit where the air sealing is already done, the assumed value will usually be far too leaky and the resulting equipment far too big. Where a jurisdiction requires envelope leakage testing for code compliance anyway, as many now do under adopted editions of the International Energy Conservation Code, the number already exists. Use it instead of a category.

Mechanical ventilation is a separate line, not a subset of infiltration. Outdoor air brought in deliberately to satisfy ASHRAE Standard 62.2 arrives whether the envelope leaks or not, carries its own sensible and latent load, and in a humid climate the latent share of it can dominate. Heat or energy recovery reduces that load without deleting it, and the recovery effectiveness belongs in the calculation rather than in the estimator's head.

Internal Gains: Heat the Weather Does Not Control

People and equipment add heat on a schedule that has nothing to do with the weather. Manual J assigns occupancy from bedroom count rather than from what the homeowner reports, which ties the calculation to the house instead of to the current family - a sensible convention when the equipment will outlast the occupants. Appliances, lighting and plug loads contribute a fixed sensible base. Occupants contribute both sensible and latent, and in an otherwise tight house they are the first place latent load enters at all.

Internal gains matter more as envelopes improve. A well-insulated, tight house has a small envelope load, so a kitchen full of people and a south-facing great room can swing the cooling requirement by a much larger fraction than the same gains would in a house from the 1960s. On the heating side those same gains are a credit that a design-day calculation deliberately declines to take, because the load has to be met at four in the morning with the house asleep and every appliance off.

The Sensible and Latent Split

Every cooling load splits into sensible heat, the part that moves a thermometer, and latent heat, the part that condenses moisture out of the air. The ratio of sensible load to total load - the sensible heat ratio - decides what equipment can actually satisfy the house. Latent load arrives with infiltration and ventilation air, from occupants, and from indoor sources such as cooking and bathing. In a dry climate the latent term is small and almost any coil meeting the sensible number will do. In a hot, humid climate the latent term is a large fraction of the total, and a coil selected on total capacity alone can hold the thermostat setpoint while leaving the house at sixty percent relative humidity.

Equipment does not carry one sensible heat ratio either. Manufacturers publish expanded performance data giving total and sensible capacity across entering wet-bulb, entering dry-bulb, outdoor temperature and airflow - a matrix, not a single rating. Selecting against a humid-climate load means reading that matrix at the design entering conditions and at the airflow you actually intend to deliver, then confirming that sensible and latent capacities both land where the load needs them. Reducing airflow per ton shifts capacity toward latent; raising it shifts capacity toward sensible. That lever is part of the selection, and ACCA Manual S is where the method for using it lives.

Ducts Outside the Thermal Envelope

Ductwork outside the thermal envelope is a load component in its own right, and often a large one. A supply trunk in a vented attic sees air temperatures well above outdoor design on a summer afternoon, so conduction through the duct wall runs on a temperature difference much larger than the one across the ceiling underneath it. Leakage compounds the loss. Supply leaks dump conditioned air where it does nothing, return leaks pull hot, humid attic air straight into the coil, and the latent penalty from return-side leakage in a humid climate is brutal.

Two responses exist and they are not equivalent. One is to account for the loss: add the duct load, size the equipment larger, and keep paying for it every hour of every season. The other is to bring the ducts inside the envelope, or bring the attic inside it, or at minimum seal and insulate to a level the calculation can then credit. Adding capacity to cover leaky attic ducts sizes the equipment for a defect and locks that defect into the utility bill. Measuring duct leakage, required for code compliance in a growing number of jurisdictions, turns the argument into a number.

Why the Heating Load and the Cooling Load Size Different Things

Heating and cooling loads are not two views of one quantity. The heating design load is conduction plus infiltration plus ventilation at winter design temperature, with no credit taken for sun or internal gain, and it sets burner input or heat pump capacity at a stated outdoor temperature. The cooling design load adds solar gain, internal gain and the latent term, and it sets not only compressor capacity but the airflow the system must move and the coil that air has to cross.

Which load governs depends on climate, and the consequence lands on the equipment that does not govern. In a cold climate the heating load is larger, the furnace is selected for it, and the blower attached to that furnace may deliver considerably more air than the cooling coil wants. In a hot, humid climate cooling governs, and the smallest furnace available in the matched product line can still be substantially oversized for the heating load. Heat pumps sharpen the conflict, because one machine has to satisfy both, which is why the balance point - the outdoor temperature at which capacity meets load - is a design output rather than a rating printed on a box. Variable-capacity equipment widens the acceptable window without removing the need to know both numbers.

Screening Estimate Versus Full Room-by-Room

A screening calculation and a room-by-room calculation answer different questions, and both answer honestly within their limits. Screening takes floor area, climate, a broad envelope description and a glazing fraction, and returns a whole-house load good enough to tell you whether the existing three-ton unit is plausible, whether a competing proposal sits in a defensible range, or roughly what budget to discuss on the phone. It cannot produce room airflows, cannot size a branch duct, and cannot support a permit.

A full Manual J takes the real takeoff: every assembly with its area and construction number, every window with orientation, U-factor, SHGC and shading, a leakage figure preferably measured, ventilation rate and recovery effectiveness, and the orientation of the house on its lot. It produces the room-by-room breakdown that feeds duct design under ACCA Manual D and equipment selection under ACCA Manual S. Where a jurisdiction, a utility programme or an energy code requires a load calculation, that is what it requires. A screening number is not a substitute, and submitting one as though it were will cost you the plan review and the schedule that depended on it.

1. Conditioned floor area and volume, room by room, with ceiling heights noted wherever they vary
2. Each opaque assembly by area and construction, carrying framing fraction and installation grade
3. Every window and door by orientation, area, U-factor, SHGC and external shading
4. Envelope leakage - measured CFM50 where it exists, with the conversion method named
5. Mechanical ventilation rate, equipment type and recovery effectiveness
6. Duct location, insulation level, and measured or assumed leakage
7. Indoor design conditions and the outdoor design temperatures used, with their source stated

When the Load Meets the Equipment Catalogue

Capacity comes in steps. Loads do not. The calculated load lands somewhere between two catalogue sizes, and Manual S sets out how far above the load a selection may sit and how to read manufacturer data to confirm it. Nominal tonnage is a label rather than a capacity: rated output depends on outdoor temperature, entering air conditions and airflow, and the figure that matters is capacity at your design conditions, not the one embedded in the model number.

Selection is also a matching exercise. Outdoor unit, indoor coil and blower form a rated combination, and substituting any one of them changes the published performance. Latent capability, airflow at the external static pressure the duct system actually imposes, and, for heat pumps, capacity at the cold end of the range all belong to the selection rather than to a later troubleshooting visit.

Oversizing, Short Cycling, and the Clammy House

An oversized cooling system satisfies the thermostat quickly and then stops. Short run times are the mechanism behind nearly every complaint that follows. Dehumidification happens only while the coil is wet and air is moving across it; a coil that has just started cooling is still chilling down, and much of the moisture condensed during a short cycle re-evaporates off the fins as soon as the blower stops. A correctly sized unit runs long, steady cycles at design conditions and removes moisture continuously. The oversized one holds dry-bulb setpoint and leaves the house clammy, a complaint diagnosed as a refrigerant charge problem far more often than as what it actually is.

Cycling costs more than comfort. Compressor starts draw heavily and stress the machine, the mismatch between capacity and load at part load - which is most of the season - drags seasonal efficiency below the rated figure, and room temperature swings widen as the system slams on and off. On the heating side an oversized furnace produces the same pattern with the addition of noticeable drafts at start-up. None of this shows up on the day of commissioning, which is precisely why oversizing survives from one job to the next.

The Square-Footage Rule: Why It Survives, What It Costs

Square-footage sizing persists because it is fast, free, and usually produces equipment that cools the house. A rule of one ton per some fixed area, calibrated by local habit, encodes decades of experience with one climate and one construction era. Against the housing stock it was calibrated on, it lands close enough often enough that the contractor using it rarely finds out which jobs it failed.

The failures are systematic rather than random. The rule carries no term for orientation, so a house with a glass wall facing west and its mirror-image neighbour facing east receive identical equipment. It carries no term for envelope quality, so a deep-energy retrofit gets sized as though the work never happened, frequently by a margin approaching a factor of two. It carries no latent term, so it fails hardest in exactly the humid climates where latent performance decides whether the customer is satisfied. Producing a single whole-house number, it says nothing about which room gets which airflow. And where a permit, a rebate or a warranty claim depends on documented sizing, it produces nothing usable at all.