The New Install That Would Not Dehumidify

The call

Six weeks after a changeout, a homeowner on the Gulf Coast called back: the house held 74 F all day and felt like a basement after rain. Towels did not dry, the hardwood felt tacky, and a thermostat in the hallway showed 63 percent RH. The old system, an 18-year-old single-stage unit, had never felt like this.

The new equipment was a 3-ton two-stage condensing unit on a variable-speed (ECM) furnace with a matched cased coil and a TXV. The install sheet showed a weighed-in charge and a subcooling check. On paper, a good install. This case is about why a correct charge and a correct temperature were not the measure that mattered, and about the order of eliminations that got there without selling the customer a dehumidifier.

First reading: put moisture in one unit

Relative humidity moves with temperature, so it cannot compare a hallway to a supply register. The tech converted everything to dew point before comparing anything.

• Outdoors: 88 F, dew point 72 F.
• Return grille: 74 F, 63 percent RH, dew point about 61 F.
• First supply register: 58 F, 90 percent RH, dew point about 55 F.

So the coil was taking the air down about 6 F of dew point, 61 minus 55, while it ran. The temperature split, 74 minus 58, was 16 F. That split is the number that lied. Sixteen degrees looks like a healthy cooling system, and every tech who had looked at this system had seen it and moved on. A split tells you sensible capacity. It says nothing about how much water the coil is wringing out, and that was the complaint.

Eliminations, in the order evidence allowed

Charge. The obvious one, and the one the customer's neighbor suggested. On high stage after 15 minutes: subcooling 10 F against the maker's target of 10 F, superheat 9 F at the evaporator outlet with the TXV in control. Charge was right. Adding refrigerant to a TXV system on target raises subcooling and head pressure; it would not have made the coil colder. Eliminated, with numbers on the ticket.

A moisture source the system could not keep up with. The tech walked the house for the usual suspects: a dryer vent disconnected under the floor, bath fans that did not move air, an open crawlspace vent, a recent leak. The dryer vented outside, bath fans pulled a tissue, the slab house had no crawlspace. Indoor dew point at 61 F sat well below outdoor at 72 F, which says infiltration was present (it always is) but the house was not wide open. A source would show as an indoor dew point near or above outdoor, or one room much wetter than the rest. Neither. Not the driver.

Oversizing and short cycling. The classic cause of a cool, clammy house is a system that hits setpoint in eight minutes and shuts off before the coil gets cold and drains. The thermostat's run history ruled it out the opposite way: low-stage cycles of 40 to 55 minutes through the afternoon, with high stage coming on for part of the hottest hours. Long cycles are the condition dehumidification needs. Oversizing does not produce this history. Eliminated.

Continuous fan. A fan set to ON blows across a wet coil between cycles and puts much of the water back in the house. The thermostat fan mode read AUTO, and the furnace's off-delay was the factory default. Eliminated.

Airflow. Last, because it takes the most work to measure and because it was the one remaining that could explain long cycles that cooled well and dried poorly. Total external static on high stage read 0.50 in wc. Against the furnace's published blower table at the selected cooling setting, that static put delivered airflow at about 1,350 cfm, which on a 3-ton system is 1,350 divided by 3, or 450 cfm per ton.

The furnace board's cooling airflow switches were set to the maker's highest cooling profile. The installer had set it that way on purpose: more airflow means more sensible capacity and a better efficiency number on paper, and in a dry climate that is a reasonable choice. Here it was the cause.

Why 450 cfm per ton made a wet house

At high airflow the evaporator runs warmer. The tech read saturated suction temperature at the outdoor unit: about 48 F on high stage. Air moving fast across a coil that warm spends less time below its dew point, so more of the coil's work goes into dropping temperature and less into condensing water. The sensible heat ratio rises, the latent share falls. The system satisfied the thermostat's temperature call while the moisture load accumulated.

Two-stage equipment makes this worse rather than better when airflow is set high. Low stage is the stage that runs long and should be doing the drying, but if the blower's low-stage airflow scales off the same high profile, low stage gets the same warm-coil problem with less refrigerant capacity behind it. See related: Airflow Per Ton Targets and What Moves Them, which owns the 350 to 400 cfm per ton band and what pushes a system toward each end of it.

The thermostat was the second half. The communicating-capable thermostat had a dehumidify function and a dehumidify output wired to the furnace's dehumidify terminal. The function was switched off in the installer settings. With it on, the thermostat would have told the furnace to slow the blower further whenever indoor RH sat above its setpoint during a cooling call, which is exactly the condition this house lived in.

The fix

1. Power off and prove dead before touching the board. The service switch at the furnace off, the board's line voltage terminals proven dead with a tester checked on a known live source before and after, per NFPA 70E Article 120 in the current edition. Dip switches sit inches from line voltage terminals and the door switch is easy to defeat by leaning on it.

2. Cooling airflow set to the maker's humid-climate setting. In this furnace's manual that is the profile delivering nominal 350 cfm per ton with a slow ramp at the start of each cycle, so the coil gets cold before full airflow arrives. Panels on, power restored, standing to the side of the cabinet as the board powered up.

3. Dehumidify enabled at the thermostat, setpoint 50 percent RH. The tech then checked the furnace's dehumidify logic in the manual, because makers differ on whether the terminal is energized to call for dehumidification or energized to mean no call is needed. Wired with the wrong sense, the furnace runs in reduced airflow all the time. A forced call with RH above setpoint dropped the blower, and raising the setpoint above room RH restored it, which proves the sense is right.

4. Re-measured after the change. Total external static on high stage fell to 0.32 in wc. The blower table at that static and the new setting read 1,050 cfm, 1,050 divided by 3 is 350 cfm per ton. Saturated suction dropped to about 43 F. Supply read 54 F, a split of 74 minus 54 or 20 F.

5. Charge re-verified at the new airflow. The maker's subcooling target is stated at rated airflow, and moving airflow moves the readings. Subcooling read 11 F against the 10 F target, inside the maker's tolerance, superheat 8 F. No refrigerant added. If it had read outside tolerance the tech would have charged to the target at the new airflow, logged per the shop's refrigerant record SOP.

6. Coil freeze margin checked. Dropping airflow cools the coil, and in dehumidify mode it drops further. At 43 F saturated suction on high stage there was margin above freezing, and the maker's minimum airflow in dehumidify mode was respected by using its own setting rather than an invented one. A coil that ices in dehumidify mode on a mild night is the failure to watch for on a follow-up.

Confirmation

The tech left a temperature and RH logger in the hallway and came back seven days later. Indoor held 74 F at 52 percent RH through the week, a dew point of about 55 F, down from about 61 F. The logger showed RH climbing on the two days of afternoon rain and falling back within a few hours of each, which is the system responding to load rather than falling behind it. The customer reported towels drying and the floor no longer tacky.

The return-to-supply dew point depression on the follow-up visit, with the system on low stage in dehumidify mode, measured about 11 F, against 6 F before the change.

What the install process changed

Nothing was broken. Every component was correct and the charge was right. The setup choice that suits a dry climate was applied in a humid one, and the commissioning sheet had no line that would have caught it: it recorded a split, a static and a subcooling, all of which looked normal at 450 cfm per ton.

The shop added two lines to its cooling commissioning sheet for humid-climate installs: delivered cfm per ton with the board setting that produced it, and return and supply dew point with the depression between them. It also added the dehumidify function to the thermostat setup checklist, with the sense check from step 3. A sheet that measures only temperature will approve a system that cannot dry a house.

Where the answer flips. In a dry climate, or in a house with a dedicated whole-house dehumidifier, the high airflow profile can be the right choice and this fix would cost capacity for nothing. And if the run history had shown eight-minute cycles instead of 40 to 55 minute ones, airflow would have been a secondary issue behind an oversized system, and slowing the blower would have helped only a little.

References

• See related: Airflow Per Ton Targets and What Moves Them
• See related: How to Estimate Blower Airflow From Static and the Fan Table
• See related: HVAC Superheat and Subcooling Method Reference
• See related: The Refrigerant Added and Leak Record SOP
• NFPA 70E Article 120, current edition