Whole-House Dehumidifier Installation and Setup

Purpose

A ducted dehumidifier earns its keep only if it draws from a space that is actually wet, discharges where the dry air gets distributed, and runs on a control that is not fighting the thermostat. Miss one of those and the unit runs all day, adds sensible heat, and the customer calls back saying it made things worse. The quieter failure is a unit hung over a finished ceiling with untested overflow protection: it does not fail on install day, it fails in year three onto drywall.

What this produces is a unit sized from measured entering conditions rather than a headline pint number, ducted to a measured airflow, drained with a tested float switch, and handed over with a dated before-and-after reading.

Scope

Covers residential and small light commercial ducted dehumidifiers: sizing confirmation, placement, process and discharge ducting, condensate, control wiring and blower interlock, restoration, performance verification and setpoint.

Does not cover the diagnosis that decided a dehumidifier was the answer; that belongs to the humidity complaint investigation SOP, and nothing gets installed off a phone call. Does not cover portable units, crawlspace encapsulation or duct sealing. Where an atmospherically vented appliance shares the space this unit will draw from, the combustion air verification SOP runs first, because pulling air out of that room is the condition that makes a natural draft appliance spill.

Roles and responsibilities

Procedure

1. Confirm sizing off the manufacturer's performance table at the measured entering conditions. Read pints per day at the recorded space temperature and RH, not off the box. Acceptance: a written table value with its entering condition beside it. The residential nameplate test condition moved to 65 F from the older 80 F standard, so figures from either side of that change are not comparable. What wrong looks like: a pint number with no condition attached. Stop rule: no table value, no install. Hazard: none at this step, a decision made at the tailgate.

2. Kill power and prove it dead at the air handler and at the circuit the unit will land on. Prove the meter on a known live source, prove the conductors dead, prove the meter live again, per NFPA 70E-2021, 120.5, with work practices at 29 CFR 1910.333(b)(2). Acceptance: 0 volts phase to phase and phase to ground on a meter that read correctly before and after. Hazard: without that check a blown-fuse meter reads a believable zero, and the blower coasts, so the air handler is also isolated under 29 CFR 1910.147 before a hand goes in. Stop rule: no qualified person on site under 29 CFR 1910.332 sends the electrical work to an electrician.

3. Set the unit to the manual's service clearances with a drain path that falls the whole way. Acceptance: clearances meet the manual on every side that specifies one, unit level within tolerance, continuous fall from drain outlet to termination with no upward leg. What wrong looks like: a unit shoehorned where the filter cannot come out, which becomes a unit whose filter never gets changed. Stop rule: no clearance, no set, relocate. Hazard: this is the lifting step, so an attic install stands on a rated platform and walks joists rather than ceiling, and a two-person lift is two people.

4. Cut in the process (intake) duct, then measure what it actually pulls. Size to rated process airflow, keep flex short and pulled tight, draw from the wet space on the drawing. Acceptance: measured process airflow within 10 percent of rated cfm, read at the unit in normal running state. What wrong looks like: compressed flex and a reading well under rated, which surfaces later as a unit that never satisfies. Stop rule: outside 10 percent, correct the duct and re-measure. Hazard on the cutting: opening duct with internal fiberglass liner liberates fibers, so cut with the blower off, and where liner is shedding the control is a respirator under a written program per 29 CFR 1910.134, not a nuisance mask.

5. Duct the discharge so dry air leaves and does not come straight back. Land it into the supply plenum downstream of the coil or a dedicated dry-air run, away from the intake. Acceptance: with the unit running, intake air reads within 2 F and 3 percentage points RH of the room it is meant to sample. What wrong looks like: intake warmer and drier than the room, so the space stays wet while the unit reports success. Stop rule: on a short circuit, separate the terminations or add a damper. Hazard: plenum edges cut the back of a reaching hand, so gloves stay on and you deburr before reaching in twice.

6. Pipe the condensate, then test the overflow device by making it work. Trap and pitch per the manual and the mechanical code your jurisdiction has adopted, and fit the auxiliary pan and switch that code requires above finished space. Acceptance: a poured pint drains with no backup inside a minute, and lifting the float shuts the unit off within the manual's stated delay. What wrong looks like: a switch fitted and wired to nothing, which passes a visual inspection. Stop rule: a switch that does not stop the unit is rewired and retested before the job closes. Hazard: water and line voltage share this cabinet, so the pour test runs with control power off.

7. Wire the control and set the interlock so the unit and the thermostat are not fighting. Land the humidistat or the thermostat's humidity terminal, decide whether a humidity call also energizes the blower, and mount the sensing control in conditioned space at breathing height away from a register. Acceptance: a humidity call starts the unit, and the blower if configured, inside the manual's delay, and a cooling call does not cancel it. What wrong looks like: a humidistat in the mechanical room, reading the machine instead of the house. Stop rule: if the controls fight, get the diagram for that pairing rather than guessing at a jumper. Hazard: the cabinet still holds line voltage, so the disconnect stays open while hands are inside.

8. Restore power, close it up, and re-prove the protections you disturbed. Covers and doors go on first, then close the disconnect standing to the hinge side rather than square in front. Acceptance: unit starts on a call with no nuisance trip, and both protective functions re-tested live, meaning the float switch stops the unit when lifted and the blower door interlock stops the air handler when opened. Hazard, at the moment the customer is nearest: a fault you did not find announces itself at re-energization, so occupants and pets are out of the room and covers are on before the handle moves. Stop rule: any trip, buzz or smell means open the disconnect and find it first.

9. Verify performance across the unit and set the setpoint against the coldest surface in the house. At steady state, record entering and leaving dry bulb and RH at the unit and convert both to dew point. Acceptance: leaving dew point at least 10 F below entering dew point at rated airflow, and a space setpoint of 45 to 55 percent RH, taking the lower half where single glazing or a cold slab will condense. What wrong looks like: a leaving dew point barely below entering, meaning low airflow, low charge, or a coil read before it was cold. Stop rule: under 10 F, do not hand over, re-check step 4 airflow first. Hazard: a fan is turning behind the grille, so probes go through the test port, never past the guard.

When the site does not match

No drain path in reach: a condensate pump is acceptable only with its auxiliary switch wired and tested at step 6, and a pump without one is not an option. No spare circuit: the unit does not land on the air handler circuit as a convenience, the job pauses and an electrician is booked. Vented appliance in the draw space: stop and run the combustion air verification SOP before this unit is ducted at all.

The record this produces

The commissioning sheet is filled in live, not rebuilt at the truck. Fields: space temperature and RH before install with the location taken; the performance table value and its entering condition; model and serial; configuration as built with any deviation and who authorized it; measured process airflow against rated in cfm; the intake-versus-room check from step 5; drain and float switch results as pass or fail, with the retest where one failed; entering and leaving dry bulb, RH and dew point at the unit; setpoint left on the control; date and time.

Three people read it later. The tech returning for the follow-up, who needs the same points and heights for the comparison to mean anything. The service manager, deciding whether a callback is a warranty defect or a load the customer added. And the shop, if the customer later disputes the house was ever wet.

Worked pass: 2,400 square foot two-story over a conditioned basement

July complaint of a sticky basement and a musty smell. The investigation recorded the basement at 78 F and 62 percent RH, a dew point of about 64 F, against a main floor at 75 F and 55 percent RH.

Step 1. Performance table at 78 F and 62 percent RH entering reads 82 pints per day for the selected unit, whose nameplate says 70 pints at the 65 F test condition. Value and condition recorded. Pass.

Steps 2 and 3. Disconnect open, conductors 0.0 volts phase to phase and phase to ground, meter proved live before and after. Unit set on the basement platform, 12 in clear at the filter side against a 12 in minimum, continuous fall to the laundry standpipe. Both pass.

Step 4. Process airflow reads 205 cfm against a rated 250 cfm. 205 divided by 250 is 0.82, so 18 percent low, outside the 10 percent gate. Step 4 fails and takes its stop rule. Cause: 25 ft of 8 in flex, compressed at two joists, hard turn at the collar. Replaced with a 20 ft pulled-tight run and a hard 90. Re-measured 242 cfm; 242 divided by 250 is 0.968, so 3.2 percent low, inside 10 percent. Pass on retest, both readings recorded.

Step 5. Intake reads 77.4 F and 61 percent RH against the room at 78 F and 62 percent RH, so 0.6 F and 1 point apart, inside the gate. Pass.

Step 6. A poured pint clears in about 20 seconds. Float lifted and the unit does not stop; the switch is fitted with its leads capped behind the cabinet. Rewired per the diagram and retested, unit stops in 5 seconds against a stated delay of up to 10. Pass on retest, both results recorded.

Steps 7 and 8. Humidistat at 4 ft in the basement living area, off the register; humidity call starts unit and blower, cooling call runs alongside without cancelling it. Covers on, disconnect closed from the hinge side, no trip, and both protections re-proved live: float lifted stops the unit, blower door opened stops the air handler. Both pass.

Step 9. Entering 78 F and 61 percent RH, dew point about 64 F. Leaving 96 F and 24 percent RH, dew point about 53 F. 64 minus 53 leaves 11 F of depression against a 10 F gate, which passes on the gate rather than clear of it, so the reading is repeated at the next visit before the setting is called settled. Setpoint 50 percent. Pass.

Follow-up at 24 hours: basement 76 F and 51 percent RH, dew point about 57 F, down from 64 F at the same location and height.

References

• Manufacturer's installation manual and performance table for the specific unit, read at your measured entering conditions rather than the nameplate rating condition
• The mechanical code your jurisdiction has adopted, for condensate disposal and the auxiliary drain pan requirement over finished space
• 29 CFR 1910.333(b)(2) with NFPA 70E-2021, 120.5 for live-dead-live; 29 CFR 1910.147 for isolating the blower; 29 CFR 1910.134 for respiratory protection under a written program
• See related: the humidity complaint investigation SOP, which owns the diagnosis that selects this equipment, and the combustion air verification SOP