How to Think About Capacity Versus Output

Why this matters

A nameplate number is a measurement somebody made once, in a laboratory, at conditions written into a rating standard. Your job site is not that laboratory. When a customer says the equipment cannot keep up, there are two completely different findings hiding behind one complaint: equipment that is failing to produce what it should at today's conditions, and equipment that is producing exactly what it should while today's conditions have moved the number down.

The first is a repair. The second is a design conversation about load, supplemental heat, or distribution. Selling the wrong one wastes a compressor or a heat exchanger, and selling the other one leaves the customer cold with a clean bill of health in their file.

The number that separates them is not the nameplate. It is the manufacturer's expanded performance table read at the conditions you actually measured.

What "rated" is a measurement of

Every rated capacity carries a hidden clause: at the conditions defined by the applicable standard. Unitary air conditioners and heat pumps are rated under AHRI Standard 210/240, which sets cooling capacity at 95 F outdoor air with 80 F dry bulb and 67 F wet bulb entering the indoor coil, and heating capacity at two outdoor points, 47 F and 17 F. Gas-fired central furnaces are rated to the applicable gas appliance standard, ANSI Z21.47 / CSA 2.3, with the seasonal efficiency figure produced under a separate Department of Energy test procedure in 10 CFR Part 430.

The rating point is not a promise about your building. It is a fixed reference so two pieces of equipment can be compared to each other. What you need for diagnosis is the expanded performance data behind the rating: the table or curves that give capacity across a range of outdoor temperatures, entering conditions, and airflows. That table exists for essentially every piece of equipment sold and most techs have never opened one.

Four things that move output away from the rating

The outdoor condition. An air-source heat pump produces less heat as outdoor temperature falls, because the outdoor coil has less temperature difference to work with, and a condenser rejects less as ambient rises. Both are normal behaviour, not faults.

The entering condition on the load side. Cooling capacity is quoted at a specific entering wet bulb and moves substantially with humidity, because part of that capacity is latent. A coil measured on a dry 90 F day and a humid 90 F day gives two different numbers, both correct.

Flow. Airflow or water flow away from the rating basis changes both capacity and the split between sensible and latent. Low airflow on a cooling coil reduces total capacity and shifts what remains toward latent; high airflow does the reverse.

Supply and site conditions. Altitude, fuel supply pressure, and voltage outside the equipment's range each cap what the machine can do, and none of them appear on any performance table. They have to be measured.

The gate, stated in full

Unit of analysis: one system, at one measured operating point, in a steady window, with flows measured at the equipment rather than assumed.

Gate: the equipment is the limit when measured delivered output is more than 10 percent below the manufacturer's expanded performance value at the same measured conditions AND the measured flow is within 10 percent of the table's basis. Both halves must hold.

If output is low but flow is outside 10 percent, the flow is the finding and the output comparison is not yet interpretable, so correct flow and re-measure before applying the gate again.

When the gate fires, verify the energy input before condemning any component: fuel firing rate against the rating plate for a combustion appliance, refrigerant charge and electrical supply for a vapor-compression system. An input shortfall produces an output shortfall that looks identical to a failed component and costs a fraction as much to fix.

Basis check, every time: confirm whether the published table includes indoor fan heat before comparing it against an air-side measurement, because an air-side reading always includes it. Comparing a table that excludes fan heat against a measurement that includes it flatters the equipment by exactly the fan's input.

Safety before either of these measurements

• If you smell gas, everyone leaves the building immediately. No switches touched, no lights, no phone used inside, call the gas utility's emergency number from outside. Do not return to shut anything off.
• Opening a pressure tap on a gas train releases gas at the fitting. Close the appliance manual shutoff first, install the fitting, then open the valve, and leak-check the connection with a listed leak-detection solution afterward. Never check with a flame.
• Carbon monoxide is odorless and the route is inhalation. Wear a personal CO monitor whenever you are working on a firing appliance, and do not alter firing rate or combustion air without a calibrated analyzer. A building CO alarm gets the same response as a gas odor.
• Reaching into an outdoor unit or an air handler. Lock and tag the disconnect under 29 CFR 1910.147 for the mechanical and stored-energy hazard, and wait for the fan to stop before your hand crosses the guard.
• Metering at an energized board or checking supply voltage is electrical work under 29 CFR 1910.333(b)(2). Prove the meter live, prove the circuit, prove the meter again, per NFPA 70E-2021, 120.5.
• Placing temperature probes in a supply plenum puts your hand near a hot heat exchanger; let the appliance be off while you install them. Keep the probe out of line of sight of the heat exchanger or it reads the radiant surface rather than the air, an error the measurement card covers in detail.

Case one: the gate does not fire

Air-source heat pump, nameplate three ton. Complaint: cannot hold temperature on cold nights, and the customer believes the unit is failing.

Measured on a 17 F night, indoor entering air 68 F, airflow measured by traverse at 1,150 cfm against a table basis of 1,200 cfm, supply air 84.8 F.

Delivered, air side: 1.08 x 1,150 x (84.8 - 68) = 1,242 x 16.8 = 20,866 Btu/h, call it 20,800.

The manufacturer's expanded table for this model at 17 F outdoor and 68 F entering, on a basis that includes indoor fan heat, gives 21,500 Btu/h.

First half: 20,800 against 21,500 is 3.3 percent below the table. That is well inside 10 percent, so the first half of the gate fails.

Second half: 1,150 cfm against the 1,200 cfm basis is 4.2 percent low, inside 10 percent.

The gate does not fire, and the finding is that the equipment is producing what it is supposed to produce. Note that the nameplate three ton figure, 36,000 Btu/h, never entered the comparison. At the 17 F condition this unit was never going to make that number and it was not designed to.

The customer's complaint is still real. What it describes is a balance point: the outdoor temperature below which the heat pump's declining output crosses under the building's rising load. The correct next step is a heat loss calculation plotted against this table, not a teardown. Reporting "no fault found" without that plot leaves the customer where they started.

Case two: the same gate, and it fires

Gas furnace, nameplate 100,000 Btu/h input, 80,000 Btu/h output, nameplate temperature rise range 40 to 70 F. Complaint: long run times and the space falling behind on cold mornings.

Measured at full fire, steady, with airflow by traverse at 1,300 cfm and a temperature rise of 45 F, which sits inside the nameplate rise range and therefore tells you airflow is consistent with the rating basis rather than starved or excessive.

Delivered, air side: 1.08 x 1,300 x 45 = 1,404 x 45 = 63,180 Btu/h.

The comparison value has to be on the same basis. Air-side measurement includes the blower's own heat, and this blower draws 0.5 kW, which is 0.5 x 3,412 = 1,706 Btu/h. So the expected air-side figure at rated output is 80,000 + 1,706 = 81,700 Btu/h.

First half: 63,180 against 81,700 is 22.7 percent below expectation, past the 10 percent threshold.

Second half: rise inside the nameplate range at a measured airflow consistent with the rating basis, so flow is not the finding.

Both halves hold. The gate fires, and the equipment is the limit.

Now verify the input before condemning anything. Clocking the gas meter: 2 ft3 in 96 seconds gives 2 x 3,600 / 96 = 75 ft3/h, and at the utility's posted 1,030 Btu per ft3 that is 77,250 Btu/h of input, which is 77 percent of the nameplate 100,000.

Run that through the appliance's own conversion. At the 80 percent thermal efficiency implied by its nameplate pair, 77,250 of input should give about 61,800 of output, and adding the 1,706 of blower heat gives an expected air-side figure of 63,500 Btu/h. Measured was 63,180, agreeing within about half a percent.

That agreement is the whole diagnosis. The furnace is converting the fuel it receives at exactly the efficiency it should. It is simply not receiving the fuel. Nothing inside the appliance is failing, and a heat exchanger or a control board replaced here would have changed nothing.

The cause sat at the gas train: manifold pressure measured below the rating plate value, traced to supply pressure sagging under load through an undersized branch shared with a newer appliance. Corrected at the piping and the regulator, then re-verified.

Verification after correction, same 1,300 cfm: clocked 2 ft3 in 74 seconds, so 97.3 ft3/h at 1,030 Btu per ft3, about 100,200 Btu/h of input, essentially nameplate. Temperature rise 58 F, inside the nameplate range. Air side: 1,404 x 58 = 81,432 Btu/h, which is 99.7 percent of the 81,700 expected on the same fan-heat-inclusive basis used for the original comparison.

Both the input and the output were re-measured. Confirming only one of them would have left open whether the appliance was still converting properly at the higher fire.

What the gate does not handle

Three classes of equipment where the single-point comparison above is the wrong tool, and using it anyway produces a confident wrong answer.

Variable-capacity equipment. On inverter-driven or modulating equipment, output at any instant is a commanded value, not a maximum. Measuring 40 percent of rated capacity says nothing about the equipment until you know what it was told to produce. Force a known operating state through the service interface, or compare against the control's own reported command, before comparing anything to a table.

Staged equipment measured on the wrong stage. A two-stage appliance measured on low fire against a full-fire table shows a shortfall that is entirely bookkeeping. Confirm which stage is running and read the matching column.

Conditions outside the published table. Manufacturers publish over the range they tested. At an extreme outdoor temperature or an airflow well outside the tabulated band there is no comparison value, and extrapolating past the last row invents one. Say so on the ticket and fall back to the equipment's own commissioning record at similar conditions.

One further limit catches good techs: the gate tells you whether the equipment meets its own specification, not whether that specification was ever right for the building. A system that passes at every condition and still leaves the space uncomfortable was sized against the wrong load, and no measurement at the equipment will reveal that.

References

• AHRI Standard 210/240 for the rating conditions applied to unitary air conditioners and heat pumps
• ANSI Z21.47 / CSA 2.3 for gas-fired central furnace ratings, and the Department of Energy test procedure at 10 CFR Part 430 for seasonal efficiency
• Manufacturer expanded performance data for the specific model, including the basis for whether indoor fan heat is included
• 29 CFR 1910.147 (mechanical isolation and stored energy) and 29 CFR 1910.333(b)(2) (electrical work); NFPA 70E-2021, 120.5 for live-dead-live proving
• See related: What a BTU Actually Is; How Heat Gets Into a Space You Are Trying to Condition; The Temperature You Measure and the Temperature That Matters