The Door Gap You Never Measured: What a Shop Door Perimeter Actually Costs

Yo, let me tell you a story. I'm hauling a slab of walnut up the stairs to my second bedroom — the one the landlord thinks is for guests — and I pass my own shop door. And I stop, because I can feel air moving. Not a draft you'd write a letter about. Just air. Moving out of my shop, past the jamb, and settling into the rest of my apartment. The strip of light under that door is maybe a quarter inch. I've walked past it a thousand times. That night I actually looked at it.

Here's what that strip is. Build-Construct's door-soundproofing guide puts a number on it: a quarter-inch gap under a standard 36-inch door is roughly 9 square inches of open space. Nine square inches. The guide calls it the equivalent of leaving a small window open. Multiply gap height by door width and you've got your own leak area. Mine is a small window I've been leaving open for years.

You Found It, You Dry It

Before you go buy anything, look at the perimeter as a path, not a decoration. Four sides: two jambs, the head, the threshold. Each one leaks for a different reason. The threshold leaks because the door bottom clears the floor by design so it doesn't drag.

Now here's the part that costs you on the house side. If your shop shares a wall with the house — attached garage, converted porch, that second bedroom — the U.S. Department of Energy treats that boundary as an air barrier. The 2021 IECC Table R402.4.1.1, which the DOE's October 2024 insulation and air-sealing page cites, names "garage separation" as a component, sealed with foam backer rod plus caulk or low-expansion spray foam. So the gap isn't just your problem. It's a named line in code.

Want to know if you've got it bad? PNNL's Building America Solution Center publishes the test. Hold the house at 50 Pa with a blower door, keep all operable garage openings closed, and measure house pressure relative to the garage with a second manometer. The house-to-garage differential should exceed 45 Pa to pass the EPA Indoor AirPlus criterion. Then the confirmation: record CFM50 with the overhead garage door closed, open it, record again. If the difference is more than 6%, the house-to-garage air barrier fails even if the first test passed. BASC warns the pressure test can give a false positive if the garage-to-outside path is leakier than the house-to-garage path. That comparison exists to catch it. You don't own a blower door. But now you know how the boundary is supposed to hold, and you can feel it with a wet back of your hand on a cold morning.

The Heat Bill Doesn't Care About Your Intentions

How much does this actually cost? The honest answer is: nobody in my budget has measured my door, and you shouldn't trust a stranger who claims to have. But the published ranges are worth having.

ENERGY STAR states that sealing air leaks and adding insulation can provide up to a 10% savings on annual energy bills — and names weatherstripping doors as one of the simple fixes. This Old House attributes to the U.S. Department of Energy the figure that air sealing doors and windows can reduce heating and cooling costs by more than 20%, while its own contributor Sal Vaglica says proper weatherstripping saves 10 to 15%. Those are two different numbers in the same article. Present them as a range, not one clean figure. The DOE's own page says caulking and weatherstripping often pay for themselves in one year or less, which is the standard I'd hold a $40 fix to.

The context that makes it mean something is the leakage share. Build-Construct's blower-door explainer states that 25 to 40 percent of total heat loss or gain in a conditioned building is attributable to air leakage, and that a typical existing home measures 5 to 15 ACH50 with an effective leakage area of 50 to 150 square inches. My door is one small window in that total. It's not the whole bill. But it's a cheap place to start, and Build-Construct notes sealing work typically cuts leakage 20 to 40 percent in one pass. Those are trade figures, not code — treat them directionally.

What the Perimeter Actually Costs

Here's where the numbers get fun, because the fix is embarrassingly cheap relative to the problem. Fixr's January 2025 pricing gives materials per linear foot: foam tape $0.16–$0.25, felt $0.25–$0.50, tubular $0.25–$0.50, V-seal $0.25–$0.50, gasket $0.60–$1.75, interlocking metal channels $6–$10, and door sweeps $13–$20 per linear foot. Bulb thresholds run $8–$30 each; magnetic runs $22–$25 per linear foot. A single door at the low end comes in around $55, and the national average for all the doors and windows in a 2,000 sq ft home is $350, range $200–$600. Handymen charge $50–$100 an hour, and labor is the bulk of it, per Fixr.

Want a concrete bottom seal? The M-D Building Products 67967 kerf-style 36-inch vinyl replacement door bottom, two-pack, is listed at $26.99 on Amazon and seals gaps up to 3/8 inch. That price was what the listing showed when it was fetched — marketplace, third-party seller, not MSRP. For a garage-scale threshold, the GaraDry 3/4-inch PVC threshold kit, 8 ft 3 in, lists at $96.69, which works out to $11.72 per foot, with a 5-year warranty and a 6,000 lb vehicle limit — that's a driveway door, not your shop door, but it shows where the ceiling sits. The Papillon EPDM bottom seal, 20 ft, lists at $38.97 ($1.95 per foot), forms a ½-inch seal and is rated −40°F to 284°F.

Two product classes I want to keep straight. Build-Construct puts spring bronze at $10–$20 per door, which sits far below Fixr's $6–$10 per linear foot for interlocking channels. Different products, different jobs. Don't pick one to argue with the other.

The install detail that separates a real seal from a peel-and-stick patch: This Old House describes a grooving tool with a V-shaped base cutting a slot 1/8 inch wide and 3/16 inch deep in the corner formed by jamb and stop, and a spline roller driving the strip in. That's a jamb seal. Felt typically lasts only a year or two, per the same guide. Tubular rubber, vinyl or silicone is what it recommends for high-use door bottoms. V-strip is described as highly effective and long-lasting for tops and sides.

Dust Is a Fine-Dust Problem, Not a Chips Problem

Now the part that actually bothers me, because it's not the heat bill. It's the dust I'm breathing and the dust my family is breathing in the next room. My dust collector runs at the machine. A door gap is the path the system did not capture.

OSHA's Woodworking eTool gives the exposure limits that gap is fighting against: the PEL for nuisance dust is 15 mg/m³ total dust and 5 mg/m³ respirable fraction, 8-hour TWA. NIOSH's recommended exposure is 1 mg/m³ total dust. The NIOSH Pocket Guide lists wood dust as a potential occupational carcinogen, and OSHA notes that oak, mahogany, beech, walnut, birch, elm and ash have been reported to cause nasal cancer in woodworkers, particularly at high exposures. That's the register this debate belongs in.

To see why it's a fine-particle story: Build-Construct's air-filtration guide attributes table-saw sawdust to 100–500 microns, sanding dust to 10–50 microns, and respirable dust to 0.5–10 microns, with fine dust staying airborne 20 to 60 minutes after work stops. That's the window in which your perimeter gap is moving dust into the house. The same source's sizing rule for a shop air cleaner is a minimum 4 to 6 air exchanges per hour, 8 to 10 preferred for heavy sanding — a 20 × 30 ft shop with 10 ft ceilings needs 600 CFM at 6 ACH. Note the caveat: those particle bands come from a trade blog, not an extension service.

Noise: You Can't Seal Half a Door

The db-by-seal-type table from Build-Construct is the one number set I'd hang on the wall. Universal sweep: 3–6 dB, 5–10 years. Automatic drop seal: 8–12 dB, 10–15 years. Brush sweep: 2–5 dB, 3–7 years. Interlocking threshold: 10–15 dB, 15–20 years. That's one publisher's numbers, not standardized lab data, so hold them loosely.

Build-Construct also gives the door-panel baseline — a hollow-core interior door blocks about 20 to 25 dB, a solid-core about 30 to 35 — and states that a 1/8-inch gap around a standard door reduces its sound-blocking performance by half. That's why sealing the perimeter matters more than swapping the slab. Half of a solid door's performance walks out the gap.

The scale context matters here. The decibel scale is logarithmic; 10 dB is a tenfold change in sound power. So a 3–6 dB sweep and a 10–15 dB interlocking threshold aren't "twice as good" as each other. And OSHA 1910.95 sets the action level at 85 dBA for an 8-hour TWA, with permissible exposure at 90 dBA for 8 hours and 95 dBA for 4. A better door seal can cut what reaches the house. It doesn't replace hearing protection in the shop.

The Perimeter Fix I'd Actually Do

Here's the plan, and it's cheap. Run a hand along the jamb reveal on a cold morning. Where you feel air, you've got a leak path. Buy one roll of tubular or V-seal for the jambs and head, and one door sweep or a kerf-style bottom like the M-D at $26.99 for the threshold. If you've got a proper jamb stop and a grooving tool, that's the seal that lasts. If you don't, foam tape is $0.16–$0.25 per linear foot.

Then — and this is the part I got wrong for years — run a bead of caulk at the base of the jamb where it meets the floor and at any penetrations in the shop-to-house wall. The DOE names "garage separation" in 2021 IECC Table R402.4.1.1 and specifies foam backer rod plus caulk or low-expansion spray foam for construction gaps. In a shop where I can't fit a dust collector, that's the fix that fits my situation.

One last move if you share a wall with the house: run a theatrical smoke machine in the shop with the house depressurized and watch where the smoke pulls. PNNL's BASC recommends exactly that for locating leaks at the garage-to-house interface. Standing in a second bedroom watching smoke move toward the wall will tell you more than any spec sheet. Then go buy the cheapest roll of weatherstripping on Fixr's list, install it, and measure the air on your hand again. If it's still moving, the perimeter wasn't the only hole.

The door gap isn't nothing. It's the cheapest thing you'll measure this week.