The Tightened House That Grew Mold in the Closets

Why this matters

In one fall, a homeowner had new windows installed and the attic air sealed. In February she opened the north bedroom closet and found gray-green spots across the back wall, behind the winter coats, and the same in the hall closet on the same wall. She called the air sealing contractor, because the timing pointed straight at him. The timing was right and the instinct to deny it would have been wrong: the work had not caused a leak, it had removed most of the house's accidental ventilation, and nothing had replaced it. This case is reconstructed from the records the shop pulled together afterward, because the records are what turned an argument into a fix.

The record the shop started from

The contractor's office pulled four things before anyone went back to the house:

• The test-in and test-out blower door numbers from the energy program paperwork: 8.5 air changes per hour at 50 Pa (ACH50) in September, 4.6 ACH50 in November after the attic work and the new windows. See related: Blower Door Numbers Explained for Other Trades.
• The work order, which listed attic bypass sealing, top plates, the chimney chase, recessed light covers and the attic hatch. It listed no ventilation work, and no ventilation assessment.
• The homeowner's thermostat history. Her thermostat logs indoor humidity, and it had two winters of data.
• Her own description of the old windows: ice on the bottom of the glass most mornings in January, a towel on the sills.

The house: one story over a basement, 1,800 sq ft, three bedrooms, four people, electric range and electric water heater, a sealed-combustion furnace. No combustion appliance draws air from the house, which mattered later when the fix was chosen.

Reading the thermostat history

Two cautions before trusting it. A thermostat humidity sensor is not a calibrated instrument and can be off by a few points of RH. But it is the same sensor in the same spot both years, so a fixed offset cancels in the comparison: whatever the true values are, the house is carrying about 19 points more relative humidity at the same temperature. And the outdoor averages are within a degree of each other, so the weather did not do this.

At 68 F, 36% RH is a dew point near 40 F. 55% RH is a dew point near 51 F. About eleven degrees of dew point is a lot of water: at the same temperature, the moisture content of the air went up by the ratio of the two humidities, 55 divided by 36, about 1.5 times. See related: Relative Humidity Versus the Moisture Actually in the Air.

Where the water used to go

A family of four puts several gallons of water a day into the air from breathing, showers, cooking and drying. That water leaves a heated house three ways: carried out by air leaving the house, condensed on cold surfaces, or absorbed into materials. Before the fall, all three were working. The leaky house exchanged a lot of air with dry winter outdoor air, and the old windows were the coldest surfaces in the house, so they condensed and iced. The ice on the glass was the house dehumidifying itself, badly.

The work removed both. Cutting leakage nearly in half (4.6 against 8.5 ACH50 is 54% of the old figure) cut the winter air exchange by a similar proportion. And the new windows have warmer interior glass, so the glass stopped being the first place water landed. The water still had to go somewhere. It went to the next coldest surface. See related: The Cold Surfaces Where a House Condenses First.

Why the back of the closets

The shop set two temperature and humidity loggers for ten days in February: one in the living room, one on a shelf inside the north bedroom closet with the door shut, as the owner keeps it. A tech measured the closet's back wall with a surface thermometer at several points, checked against an infrared image of the whole wall.

The closet is a cold room. A closet on an exterior wall, packed with clothes and kept shut, gets no supply air and almost no mixing with the room. The coats insulate the wall from the room's heat. So the wall behind them runs close to what the wall assembly alone allows on a 20 F day, and on the north side there is no sun to help.

The arithmetic of the surface. Relative humidity at a surface is the room air's water vapor pressure divided by the saturation pressure at the surface's temperature. The house air at 68 F and 55% RH holds water vapor at about 1.29 kPa. Saturation at 61 F is about 1.83 kPa, so the open bedroom wall sees about 70%. Saturation at 52 F is about 1.32 kPa, so the closet back wall sees about 97%, within a degree of the house dew point of 51 F. The closet air logger agrees: the same 1.29 kPa at 60 F is the 73% it recorded.

A widely used design rule keeps surface relative humidity below about 80% averaged over a month, because mold on drywall paper does not need liquid water, only a surface that stays damp. The closet wall was at 97% for weeks. Last year's 36% at 68 F would have put that same 52 F wall at about 63% (vapor pressure about 0.84 kPa against 1.32), which is why there was no mold last winter on the identical wall.

What the record ruled out. No staining at the top of the wall or the ceiling, so not an attic leak or an ice dam. Moisture meter readings in the drywall were elevated only at the surface and only in the lower half of the cold zone, not behind it, so not a plumbing or exterior water leak. The attic sealing had not touched the walls at all. The pattern fit only surface condensation from indoor humidity.

What the work should have included

The IRC requires whole-house mechanical ventilation where a dwelling tests below 5 ACH50 (R303.4, as adopted by your jurisdiction). This house tested at 4.6. Whether that section reaches an alteration to an existing house, rather than new construction, is your jurisdiction's call, and many do not apply it to a window and attic retrofit. Many weatherization and utility programs require a ventilation assessment to ASHRAE 62.2 after air sealing regardless, and this one's paperwork had a box for it that nobody checked.

The ventilation target. ASHRAE 62.2 (2016 and later editions) sets a total rate of 0.03 cfm per sq ft of floor area plus 7.5 cfm per person, counting occupants as bedrooms plus one: for this house, 0.03 times 1,800 is 54, plus 7.5 times 4 is 30, so 84 cfm total. The standard allows a credit for the house's own measured leakage, which reduces the fan rate; at 4.6 ACH50 the credit is modest, and the contractor sized the fan to the full figure rather than leaning on it. See related: Why Air Sealing Changes What a House Needs for Ventilation.

The fix, in the order it was done

1. Cleanup came first, by people protected for it. Disturbing mold releases spores, and wiping or scraping is exactly that. The patches totaled under 10 sq ft across both closets, which is within the size EPA's homeowner guidance treats as manageable without a remediation firm. The contractor's crew wore N95 or better respirators under the company's written respiratory protection program (29 CFR 1910.134, which 1926.103 applies to construction work such as the drywall replacement), gloves and goggles, bagged the clothes for laundering, scrubbed the paper face with detergent and water and dried it. Drywall with paper that came away soft was cut out and replaced. A larger area, or any staining that recurred after cleaning, would have gone to a remediation contractor.

2. Continuous mechanical ventilation. The house already had a bath fan, noisy and rarely used. The contractor replaced it with a quiet fan rated for continuous duty, ran it continuously, and measured its delivered flow at the grille with a flow hood: 86 cfm against the 84 cfm target. Exhaust-only ventilation slightly depressurizes the house; that was acceptable here because no appliance draws combustion air from the house. In a house with a draft hood appliance it would not have been chosen without a spill test. See related: Whole-House Ventilation Strategies for Non-HVAC Trades.

3. The closets were opened to the room. Solid doors on the two north closets were swapped for louvered doors, and the owner moved storage so nothing sits flat against the exterior wall. That warms the back wall toward the room's temperature.

The confirmation. The loggers went back in for ten days in late February, outdoor average 22 F. Living room 68 F and 40% RH; the closet back wall at 57 F. At 40% the house air's vapor pressure is about 0.94 kPa, and saturation at 57 F is about 1.59 kPa, so the wall now sees about 59%. Both levers moved the same number: the drier air lowered the top of the fraction, the warmer wall raised the bottom. No regrowth at the spring check. The outdoor average ran 2 degrees warmer than the first logging, which by itself helps the wall a little, so the contractor treats the next January as the real test and has a reminder set to check.

What the air sealing contractor says at the sale now

The shop's change was not technical. Its techs already knew all of this. What changed is when the customer hears it: before the contract, not after the mold.

SALES: When we're done, the house will leak a lot less.
       That's the point, and it also means the house
       stops airing itself out by accident.
OWNER: Is that bad?
SALES: Only if nothing replaces it. Your old windows
       were icing up every winter, and that ice was water
       that had nowhere else to go. Tighten the house,
       and that water goes to the next cold spot, like
       the back of a closet on the north wall.
OWNER: So what do we do?
SALES: We test before and after. If the after number
       is tight enough, the quote includes a quiet fan
       that runs all the time. It's in this line here.

The fan line is now on every air sealing estimate, priced as an option that turns mandatory if the test-out number lands below the shop's trigger of 5 ACH50, matching the IRC figure. The office also schedules a humidity check call for the first February after any job that tightens a house by more than a quarter.

References

• International Residential Code, 2021 edition, R303.4 (mechanical ventilation below 5 ACH50), as adopted by your jurisdiction
• ASHRAE Standard 62.2, Ventilation and Acceptable Indoor Air Quality in Residential Buildings, 2016 or later edition, as required by your program or jurisdiction
• EPA, A Brief Guide to Mold, Moisture, and Your Home; 29 CFR 1910.134 for respirator use by employees, applied to construction work by 29 CFR 1926.103
• See related: The Cold Surfaces Where a House Condenses First
• See related: How to Find the Cold Spot Behind a Closet Mold Complaint