How Does an Oversized AC Unit Fail to Dehumidify: What You Control and What You Do Not

An oversized air conditioner fails to dehumidify when its sensible capacity drops indoor dry-bulb temperature to the thermostat setpoint before enough water has condensed on the evaporator coil to leave through the condensate drain, so net latent removal stays below the home’s latent load. The Air Conditioning Contractors of America Manual J procedure splits that load into calculated sensible cooling and calculated latent cooling from envelope, occupancy, infiltration, and climate data. ACCA Manual S requires the selected equipment’s latent capacity, read from Air-Conditioning, Heating, and Refrigeration Institute certified ratings or the manufacturer’s expanded tables at the actual indoor wet-bulb and outdoor dry-bulb, to cover that latent load without exceeding Manual S total-capacity limits. A clammy room at setpoint can also come from a system too small to catch up, from a blower left in continuous ON that re-evaporates coil moisture, or from a blocked condensate drain.

The diagnostic order is the one a prep hand uses on a dealer badge: lift the residue first. Fan mode, drain, and airflow are the residue.

Why does a house that hits the thermostat still feel clammy?

The thermostat reads dry-bulb air temperature. Comfort also depends on moisture. The U.S. Environmental Protection Agency’s mold and indoor-environment guidance tells occupants to keep indoor relative humidity below 60 percent, ideally between 30 and 50 percent.

Glenn C. Hourahan, P.E., then Senior Vice President of Research and Technology at ACCA, wrote that oversized cooling systems cycle off “long before moisture removal can be effected.” ASHRAE Handbook—Fundamentals, Chapter 14, defines the 1 percent cooling design temperature as the dry-bulb exceeded about 88 hours in a typical year. For Atlanta Hartsfield-Jackson, the 2021 ASHRAE tables list 91.6°F dry-bulb with a mean coincident wet-bulb of 73.6°F. Allison A. Bailes III, PhD, founder of Energy Vanguard, designs to 92°F there.

Hourahan’s example is blunt. If the calculated total load comes out to 31,500 Btuh, “nearly all contractors select a 3.0 ton (or greater) unit rather than a 2.5 ton unit.” A 2.5-ton nameplate is 30,000 Btuh, about 95 percent of that load. A 3-ton nameplate is 36,000 Btuh, about 114 percent. ACCA Manual S, in the single-speed rules still cited by ICC plan-review checklists, caps total cooling capacity at 115 percent of the Manual J total load. Hourahan wrote that “it is better for HVAC equipment to be 10% undersized than 10% oversized.”

Ed Janowiak, Manager of HVAC Design Education at ACCA, said he would pick closer to undersized than oversized. “The name of the game is dehumidification.” Oversize, undersize, and a fan or drain fault can all leave high humidity while the display reads the number you asked for.

How does an air conditioner actually remove moisture from indoor air?

Moisture leaves the house only when liquid water leaves the drain. Hugh I. Henderson, Jr., P.E., Principal at CDH Energy, with Don B. Shirey III and Richard A. Raustad of the Florida Solar Energy Center, stated that in a 2003 CIBSE/ASHRAE paper funded by the U.S. Department of Energy. The coil must run colder than the entering-air dew point. An amount they labeled Mo has to collect before the first pulse falls from the pan. Until that pulse, the coil is storing water, not exporting it.

They tested coils from 1.5 to 3 tons at the ASHRAE Test A point that matches AHRI Standard 210/240 cooling ratings: 80°F indoor dry-bulb, 67°F indoor wet-bulb, 95°F outdoor dry-bulb. Retained moisture clustered between 1.9 and 2.1 pounds. Condensate delay from a dry coil was 13.5 minutes on a 3-ton slanted slab, 16.3 minutes on a 2.4-ton A-coil at normal airflow, 23.5 minutes on a 1.8-ton vertical slab, and 32.5 minutes on the same A-coil with airflow cut to 1.5-ton behavior. On one cycle the psychrometric latent rate was 7.4 MBtuh; the condensate scale read 6.7 MBtuh.

That is why a five-to-ten-minute “coil warm-up” rule and a fifteen-to-twenty-minute “healthy cycle” rule fail as a diagnosis. Genry Garcia and Kaleb Saleeby of HVAC School described the coil’s approach to apparatus dew point as a process that “can take up to 10-15 minutes.” Henderson’s group measured a delay range more than twice as wide.

AHRI’s rating point is also not the house. ACCA Manual J’s default indoor cooling condition is 75°F at 50 percent relative humidity. Carrier’s 24ABC6 product data marks 63°F entering wet-bulb as the TVA indoor condition of 75°F dry-bulb / 63°F wet-bulb, and warns that the sensible columns assume 80°F entering dry-bulb: deduct 835 Btuh per 1,000 cfm for each degree below 80°F.

Why does an oversized AC unit fail to dehumidify if cycle-time rules are not the test?

Oversizing still changes the opportunity for that first condensate pulse. Extra sensible capacity drives room dry-bulb down faster. If the on-period is shorter than the time that coil needs to start draining, the water on the fins is still in the house.

Henderson’s field plot of a single-speed heat pump with the supply fan left on is the mechanism in numbers. At a runtime fraction of 1.0, steady-state sensible heat ratio was 0.76. Below a runtime fraction of about 0.4, the coil provided only sensible capacity. The National Electrical Manufacturers Association DC 3 thermostat standard they used allows a maximum of 3.6 cycles per hour. That is a modeling input, not a pass/fail test for a residence.

Variable-capacity and two-stage compressors can linger at a lower capacity. Henderson’s group found a two-speed residential system in a Florida house showed “very little latent degradation.” ACCA’s Daytona Beach heat-pump example listed a high-stage cooling capacity of 22,700 Btuh (102 percent of load) with a latent cooling capacity of 5,406 Btuh (244 percent of the latent load).

Lowering the thermostat does not repair the ratio. A single-stage oversized unit can satisfy 68°F as quickly as it satisfied 72°F and still leave a poor sensible-to-latent split.

How do you measure calculated sensible load, latent load, and equipment latent capacity?

Floor area per ton is sales shorthand. Hourahan listed 500 to 600 square feet per ton, 350 to 450 cfm per ton, and “latent load = 30 percent of total capacity” among the shortcuts that “generally result in greatly miss-sized HVAC equipment.” Manual J wants orientation, glass, insulation, tightness, ducts, occupants, and the outdoor design pair.

ACCA Manual J counts occupants as bedrooms plus one and assigns 230 Btuh sensible and 200 Btuh latent per person, a split Bailes walked through from the standard. Four people in a three-bedroom house are 920 Btuh of occupant sensible load and 800 Btuh of occupant latent load before infiltration. Infiltration latent load uses Manual J’s 0.68 × cfm × grains-difference relationship.

Bailes published a worked house. The Manual J cooling load was 21,341 Btuh. He inferred an actual sensible load near 18,862 Btuh from how indoor temperature held through an Atlanta design-day peak, and he recorded a latent cooling load of 3,803 Btuh. The manufacturer’s expanded data for his Mitsubishi system listed only 718 Btuh of latent capacity. Indoor relative humidity sat in the mid-to-upper 50s without a whole-home dehumidifier. Bailes’s reading is that the published latent number is wrong.

Equipment latent capacity is total minus sensible at the same row of the expanded table. Carrier’s 24ABC630 with a CNPV*3117A indoor coil, 1,000 cfm, 95°F outdoor:

| Indoor entering wet-bulb | Total capacity | Sensible capacity | Latent capacity (total − sensible) | |---|---|---|---| | 67°F (AHRI indoor wet-bulb) | 28,600 Btuh | 20,800 Btuh | 7,800 Btuh | | 63°F (Carrier TVA indoor, 75°F dry-bulb) | 26,550 Btuh | 19,900 Btuh | 6,650 Btuh |

Drop indoor wet-bulb and latent capacity collapses even though the nameplate is still “2.5 tons.” An ICC/ACCA Manual S verification example uses a different machine at 85°F outdoor, 1,000 cfm, and 63°F entering wet-bulb: total cooling capacity 28,400 Btuh, sensible heat gain 23,321 Btuh, sensible cooling capacity 21,600 Btuh. Manual S matching happens on that design row.

Blower airflow is a commissioning measurement against the manufacturer’s installation instructions for that coil, not a generic 400 cfm per ton claim. The same Carrier table publishes 875, 1,000, and 1,125 cfm for that 2.5-ton match. At 95°F outdoor and 67°F wet-bulb, latent capacity is 8,750 Btuh at 875 cfm and 6,790 Btuh at 1,125 cfm.

How do fan, airflow, and drainage faults produce the same clammy room?

Henderson, Shirey, and Raustad’s central finding is off-cycle evaporation when the supply fan keeps running after the compressor stops. Moisture that condensed during the on-period returns to the airstream. Citing Kosar, Witte, Shirey, and Hedrick in the March 1998 ASHRAE Journal, they reported that hourly simulations which ignore that degradation under-predict space humidity by 5 to 10 percent relative humidity. AUTO fan still showed some latent degradation. Constant ON was worse.

An undersized system fails differently. It may run nearly continuously and still miss setpoint on the design day. If expanded-data latent capacity is below the Manual J latent load, humidity stays high. Bailes’s Atlanta heat pump held 74°F on a design afternoon at 88 percent of the calculated total load.

Drainage faults mimic low dehumidification because water that never leaves the pan never leaves the house. Henderson’s definition of useful moisture removal is condensate in the drain. A dry drain while the coil is wet is a plumbing problem wearing an oversized-AC costume. A timed catch during a known compressor on-period is the field measurement.

| Failure | Thermostat | Indoor RH | Condensate on a timed run | First check | |---|---|---|---|---| | Oversized single-stage | Hits setpoint; short on-periods | High, swinging with cycles | Little until on-time exceeds coil delay | Manual J vs nameplate; log cycles | | Undersized | Misses setpoint on hot days; long run | High if latent capacity is short | Drain may run; house still wet | Expanded latent vs Manual J latent | | Fan ON | Hits setpoint | High after compressor stops | Water on coil, then re-evaporated | Set fan to AUTO; repeat RH log | | Drain or pan fault | May hit setpoint | High; musty pan | Coil wet; little water at outlet | Clear trap, slope, pump; measure volume |

What can you control, and which remedy actually fits?

The occupant controls fan AUTO versus ON, filter condition, thermostat setback, and whether anyone has logged humidity with a calibrated instrument through a full afternoon. Hourahan listed high unoccupied setpoints among the mistakes that let moisture get ahead of the coil. A one-point hygrometer reading cannot establish a cycling pattern. A technician controls measured cfm, charge, coil cleanliness, and whether a humidistat is wired. Two-stage and variable-capacity compressors need the thermostat that can hold them in low stage. Last summer’s outdoor dew point and the nameplate on the pad are Manual J and Manual S inputs, not thermostat settings.

Replacement is the right remedy when a current Manual J shows the installed total capacity well above the Manual S cap and the logged on-periods at part load never reach condensate. The new selection is a Manual S match at the design indoor wet-bulb, with latent capacity at or above the latent load.

A whole-home dehumidifier is the right remedy when the air conditioner can handle the sensible load and still cannot cover the latent load at peak dew-point weather. Hourahan’s method is a second Manual J run at a muggy part-load outdoor condition, “say, 83°F and 95% relative humidity.” He also described slow weather at 70°F to 75°F and 80 percent-plus relative humidity, when there is “no sensible load, and no call for cooling.” Dedicated dehumidification does not excuse a clogged drain or a fan left ON.

If the expanded ratings at the measured wet-bulb still show latent capacity below the Manual J latent load, Hourahan’s next step is dedicated dehumidification, not another bump in nameplate tons.

FAQ

Can an oversized AC cause humidity problems?

Yes. Glenn Hourahan, P.E., of ACCA wrote that oversized cooling equipment satisfies the thermostat and cycles off before moisture is removed. Hugh Henderson, P.E., of CDH Energy measured net latent capacity collapsing as runtime fraction falls, especially with the fan left ON. The usual complaint is a clammy room at setpoint.

Is it bad to have an oversized AC unit?

In humid climates, extra sensible capacity shortens compressor run time and cuts condensate. ACCA Manual S has limited single-speed cooling to about 115 percent of the Manual J total load for that reason. Hourahan found equipment routinely 50 to 200 percent larger than needed. Extra starts remain even in dry climates.

Is it better to oversize or undersize an AC unit?

When moisture control is the issue, Hourahan wrote it is better to be 10 percent undersized than 10 percent oversized. Ed Janowiak, ACCA’s Manager of HVAC Design Education, said he would pick closer to undersized: “The name of the game is dehumidification.” Manual S still requires latent capacity to meet the latent load.

Why would an AC not be dehumidifying?

Oversizing is one cause. Henderson, Shirey, and Raustad showed continuous fan operation re-evaporates coil moisture after the compressor stops. A clogged condensate drain holds water in the pan. Low blower airflow, a dirty coil, or a refrigerant undercharge can keep the coil above dew point. One hygrometer snapshot cannot separate them.

Can blower fan speed affect AC dehumidification?

Yes. Carrier’s 24ABC630 expanded ratings at 95°F outdoor and 67°F indoor wet-bulb show latent capacity of 8,750 Btuh at 875 cfm and 6,790 Btuh at 1,125 cfm. Lower airflow raises the moisture share. Continuous ON after the compressor stops does the opposite, Henderson found.

When does a whole-home dehumidifier solve a latent-load problem?

When Manual J latent load at peak dew-point weather exceeds the air conditioner’s latent capacity at the indoor wet-bulb, or when evenings sit at 70°F to 75°F and 80 percent-plus RH with almost no sensible load, as Hourahan described. ACCA treats dedicated dehumidification as independent humidity control, not a substitute for an unclogged drain.