Duct Blaster vs Smoke vs Pressure Pan Leak Location Method Decision Matrix

Why this matters

Duct leakage robs a system of a large fraction of the air it moves, drives comfort complaints, and fails code commissioning tests. But "the ducts leak" and "the leak is here" are two different jobs. A duct blaster quantifies total leakage. A pressure pan localizes which registers feed the leaky branches. Smoke (or fog) makes the leak visible at the joint. Reaching for the wrong tool means you certify a number you cannot fix, or chase a leak you cannot measure. This matrix sorts the three by whether you need a total CFM, a branch-by-branch ranking, or a finger-on-the-joint location.

The options

A duct blaster (duct leakage tester) is a calibrated fan that pressurizes the entire duct system to a reference pressure, typically 25 Pa, and reads the airflow required to hold that pressure: CFM25. That CFM25 is total duct leakage, the code-reportable number. Run it total (to outside plus inside) or "leakage to outside" by also running a blower door. It gives a magnitude, not a map.

A pressure pan is a shallow box with a manometer port that seals over one register at a time while the duct blaster (or blower door) pressurizes the house or ducts. The pan reading in Pascals at each register ranks how leaky each branch is. A near-zero pan reading means a tight branch; a high reading flags the run to investigate. It localizes to the branch, not the joint.

Smoke or theatrical fog is pushed into the pressurized duct; it streams out at every leak. It is the visual, point-of-leak method, the one that puts the leak in front of the homeowner. It needs the ducts pressurized and accessible, and it confirms a repair instantly when no smoke escapes.

When the duct blaster wins

Use the duct blaster whenever you need a number: code commissioning, a rebate or program that specifies a CFM25 threshold, before/after proof of a sealing job, or a baseline on a comfort complaint. It is the only one of the three that produces a defensible, reportable total leakage figure. Run it on a system the design and code reference, seal off the air handler and registers as the test protocol requires, and reach the reference pressure before reading. It tells you the system fails; it does not tell you where. Pair it with a pressure pan or smoke to find the leaks it just proved exist.

When the pressure pan wins

The pressure pan wins when total leakage is high and you need to know which branches to attack first. Pan every supply and return register while the system is pressurized; the highest Pascal readings point to the leakiest runs. This turns a whole-house "the ducts leak" into a prioritized punch list: seal the three worst branches and re-test. It is fast and needs no attic crawl to rank the runs. It localizes to a register's branch, so you still have to find the actual joint in that branch, which is where smoke earns its place. A pan reading near the building pressure means that branch communicates strongly with unconditioned space; a near-zero reading means it is effectively sealed.

When smoke wins

Smoke is the finder. Once the duct blaster says the system leaks and the pressure pan says which branch, push fog into the pressurized ducts and watch where it streams: a slipped collar, an unsealed boot, a torn flex inner liner, a panned joist return drawing from the attic. It is the strongest method for showing the customer the leak and for confirming the seal held (no smoke after mastic and tape). It needs the system pressurized and reasonable access to the suspect run, and fog in unconditioned space can be hard to read in a windy or dusty attic.

Reading the numbers together

The three methods reinforce each other when you keep their units straight. The duct blaster gives a magnitude in CFM25; treat that as the score, not the diagnosis. The pressure pan gives a Pascal reading per register; a pan reading approaching the reference building or duct pressure means that branch communicates almost fully with unconditioned space, while a reading near zero means it is effectively sealed. The ranking from the pan tells you the order to attack, but the pan cannot tell a slipped collar from a torn flex liner; only smoke at the pressurized joint does that. So a clean workflow reads the blaster to decide whether sealing is even warranted, reads the pan to prioritize the worst branches, and uses smoke to find and prove each joint. A common field mistake is to smoke-test first on a whole house with no blaster number; you find some leaks, seal them, and have no defensible proof of how much you reduced total leakage. Quantify before you chase, so the customer (and any rebate program) sees the before-and-after CFM25 delta the work actually earned.

Field decision flow

1. Need a reportable total or a before/after proof: duct blaster to CFM25. Document the number against the applicable threshold.
2. Total leakage is high and you want a repair priority: pressure-pan every register while pressurized; rank by Pascals.
3. Branch identified and you need the exact joint, or you want the customer to see it: smoke the pressurized run and trace the stream.
4. After sealing, re-smoke the joint (no stream) and re-run the duct blaster (lower CFM25) to confirm and document the fix.
5. For "leakage to outside," combine the duct blaster with a blower door so interior-to-interior leaks do not inflate the number you report.

Quantify, prioritize, locate, confirm. Blaster, pan, smoke, blaster again.

References

• ANSI/RESNET/ACCA Standard 380, Testing Airtightness of Building, Dwelling Unit, and Sleeping Unit Enclosures, and Airtightness of Duct Systems.
• ACCA Manual D, Residential Duct Systems, for duct design and leakage context.
• ASHRAE Standard 152, Method of Test for Determining the Design and Seasonal Efficiencies of Residential Thermal Distribution Systems.
• 2021 International Energy Conservation Code (IECC), Section R403.3, duct sealing and leakage testing requirements.