How to calculate duct loss in unconditioned space
How leakage loss and conduction loss combine to reduce delivered capacity for ductwork routed through an attic or crawl space, and why a leakage test alone misses half the loss.
What this means
Ductwork routed through an attic, crawl space, or other unconditioned area loses delivered capacity two separate ways: leakage, where conditioned air physically escapes the duct before reaching a register, and conduction, where heat moves through the duct wall and insulation between the air inside and the unconditioned space outside even where the duct itself is perfectly sealed. A duct system can pass a leakage test cleanly and still deliver meaningfully less capacity than the equipment produced, because conduction loss operates independently of leakage and a leakage test alone does not measure it.
Equipment and model context
- Forced-air duct systems with any portion routed outside the conditioned envelope
- Worked figures illustrate the method and are not a rating for any product
This explains the two loss mechanisms and how each is estimated. It does not calculate loss for a specific duct system. That requires the actual duct surface area and insulation R-value exposed to unconditioned space, the temperature difference between the duct interior and that space, and measured or assumed leakage rates for the specific installation.
What this covers
- Why a duct system can pass a leakage test and still underdeliver capacity.
- How conduction loss is calculated independently of leakage loss.
- Why the temperature difference driving conduction loss is often larger than the difference indoors.
- What insulation R-value actually buys against a duct's conduction loss.
What changes the result
- Treating a passed leakage test as confirmation that duct loss has been addressed, when conduction loss through sealed but under-insulated duct continues regardless of leakage performance.
- Calculating conduction loss against the indoor-to-outdoor temperature difference rather than the actual temperature the duct's specific location reaches, which can exceed outdoor temperature by a wide margin in an unventilated attic.
- Underestimating how much duct surface area is actually exposed to unconditioned space, particularly for a system with long supply runs through an attic.
- Assuming a stated insulation R-value applies at installed condition, without accounting for compression or gaps that reduce its effective performance.
Why leakage and conduction are two separate calculations
Leakage loss is air that physically exits the duct system through a seam, a joint, or a connection before reaching its intended register, measured as a percentage of total airflow and checked with a duct blaster or similar pressurisation test. That test measures exactly what it measures, air escaping through physical openings, and nothing about it captures heat moving through an intact, sealed duct wall.
Conduction loss is heat transfer through the duct wall and its insulation, driven by the temperature difference between the air inside the duct and the unconditioned space surrounding it, occurring continuously along any duct run outside the conditioned envelope regardless of how well-sealed the duct is. A perfectly sealed duct with zero measured leakage still loses capacity through conduction if it runs through an attic or crawl space.
Why the driving temperature difference is often worse than it looks
Conduction loss calculation needs the actual temperature of the space the duct runs through, not the outdoor design temperature used for the building's load calculation. An unventilated or poorly ventilated attic on a sunny cooling design day can reach temperatures well above outdoor air temperature, sometimes by 40 degrees Fahrenheit or more, because the attic itself absorbs solar heat through the roof deck with limited means to shed it.
Using outdoor design temperature as a proxy for attic temperature understates the actual temperature difference the duct's cooling supply air experiences, which understates calculated conduction loss to match. The correct input is the space's own design temperature, which for an unventilated attic is a distinct, and often more severe, figure than outdoor air temperature alone.
What insulation R-value actually buys, and its limits
Insulation R-value is a measure of resistance to conductive heat flow, and higher R-value reduces conduction loss for a given temperature difference and duct surface area, following the same heat transfer relationship used for pipe insulation thickness calculations. Doubling insulation R-value roughly halves conduction loss for the same conditions, though the relationship is not perfectly linear once other resistances in the assembly are accounted for.
A stated R-value assumes insulation installed at its intended thickness and without compression, and duct insulation compressed by a strap, folded at a fitting, or gapped at a seam delivers less than its rated resistance at exactly those points, the same installation-quality caution that applies to pipe insulation. A calculated conduction loss based on the insulation's rated, undamaged R-value can understate real loss where installation quality has compromised it.
Why both figures belong in the same design conversation
Leakage and conduction loss both reduce the capacity that actually reaches the conditioned space relative to what the equipment produced, and a load calculation or equipment selection performed without accounting for both can undersize the effective delivered capacity even where the equipment itself was sized correctly against the building's load. The practical response is minimising unconditioned-space duct routing where the design allows it, sealing rigorously to control leakage, and insulating adequately to control conduction, treating both as design decisions rather than only the leakage figure a code compliance test happens to measure.
Where duct routing through unconditioned space cannot be avoided entirely, calculating both loss mechanisms together gives a realistic figure for delivered capacity, which is the number equipment selection and comfort expectations should actually be built around rather than the equipment's rated output alone.
For one duct run with 15 percent leakage and R-6 insulation exposed to a 130 F attic while carrying 55 F supply air, this shows how leakage loss and conduction loss each reduce delivered capacity, and how the two combine.
- Leakage alone accounts for 15 percent of equipment output lost before any air reaches a register, the figure a duct leakage test would report.
- Conduction loss through the R-6 insulated duct wall against the 130 F attic condition adds a further 9 percentage points, bringing combined loss to roughly 24 percent of equipment output.
- A leakage test reporting 15 percent leakage, passing many code thresholds on its own, gives no information about the additional conduction loss stacked on top of it.
- The 130 F attic condition used here is not an extreme figure; unventilated attics can exceed outdoor air temperature by a wide margin on a sunny design day, which is why the conduction calculation cannot simply use the outdoor design temperature.
| Loss mechanism | Primary driver | What reduces it |
|---|---|---|
| Leakage loss | Unsealed seams, joints, and connections | Mastic sealing and verified duct blaster testing |
| Conduction loss | Temperature difference and duct surface area exposed | Higher insulation R-value and shorter unconditioned-space runs |
| Both combined | Total duct routing through unconditioned space | Routing ductwork inside the conditioned envelope where the design allows it |
Questions people ask about this
Does sealing a duct system address conduction loss as well as leakage?
No. Sealing addresses leakage specifically, closing the physical gaps air escapes through, and has no direct effect on conduction, which occurs through the intact duct wall and insulation regardless of how well the seams are sealed. A well-sealed but poorly insulated duct still carries meaningful conduction loss.
Is attic temperature the same everywhere in an attic?
Attic temperature varies with roof orientation, ventilation, insulation at the attic floor, and proximity to the roof deck, so a single representative figure is an approximation rather than a precise measurement at every point. Where the duct routing passes through a particularly hot section, such as directly under a south-facing roof slope, the conduction calculation for that section deserves the more severe local condition rather than a single attic-wide average.
How much does routing ducts inside the conditioned envelope actually save?
Routing ductwork inside the conditioned envelope eliminates both leakage loss to the outdoors and conduction loss to an extreme unconditioned temperature, since any leakage or conduction then exchanges with conditioned space air rather than attic or crawl space air, which is a materially smaller loss even where some inefficiency remains. This is why design guidance increasingly favours conditioned-space duct routing where the building's design allows it, rather than treating attic or crawl space routing as a cost-free default.
Does duct loss affect heating and cooling equally?
Both leakage and conduction loss apply to heating and cooling, though the magnitude differs because the temperature difference driving conduction loss, and the specific unconditioned space temperature, differs between a winter heating condition and a summer cooling condition. Each mode deserves its own calculation against its own design temperature difference rather than assuming one figure applies to both.
Evidence recordSource verification pending
government guidance, standards body publication · editorial review
This page is awaiting source verification against the documentation in its evidence record: United States Department of Energy and ASHRAE technical literature. Its documentation class and intended scope are shown here while that check is pending.
- Documentation class
- government guidance, standards body publication
- Scope of the definition
- Confirm against the exact model manual