Airflow CFM Reads High but Rooms Starved: Leak Location Decision Tree

Why this matters

A blower that moves rated or above-rated CFM while the rooms stay starved means the air is going somewhere other than the registers, and the entire diagnosis is about locating where the supply air escapes before it reaches the space. High measured CFM at the air handler with low register delivery is the signature of supply-side duct leakage: air leaving the blower into a leaky trunk, plenum, or boot connection and spilling into an attic, crawlspace, wall cavity, or chase instead of the room. It can also be a distribution problem where the air goes to the wrong rooms (a few registers hog the flow while others starve), or a measurement that reads fan flow but not delivered flow. The trap is reading a healthy fan curve and declaring airflow fine; fan CFM is not register CFM. The fix depends on where the leak is, so the work is methodical leak location: compare total fan flow to the sum of register flows, then walk the supply path from plenum to boot looking for the gap.

Symptom presentation

A flow grid, hot-wire reading, or blower-table estimate at the air handler shows CFM at or above the rated value, yet a flow hood at the registers totals well below that, and the complaint rooms feel weak. Delta-T at the coil may be normal or slightly wide. The attic or crawlspace housing the ducts may feel conditioned (a sign air is leaking there). Some registers may read near design while others read far below, pointing to where the upstream leak or imbalance sits. Energy use is high because the system conditions unintended spaces.

Quick checks

Measure total fan CFM, then measure each register with a flow hood and sum them. A large gap between fan flow and the register total is leaked air. Walk the supply trunk and branches in the attic or crawlspace by hand and with a smoke pencil at seams, the plenum-to-coil joint, and each boot-to-drywall connection. Feel for conditioned air spilling into unconditioned space. Check for disconnected flex at boots and crushed runs that force air out of upstream leaks. Note which registers starve to localize the upstream break, because a starved cluster of registers usually shares a common upstream trunk section where the break sits. As a rule of thumb a fan-to-register gap of more than 10 to 15 percent of rated flow is enough leakage to starve rooms and warrants a sealing pass before any other diagnosis.

Isolation tree

Branch 1, supply plenum or trunk leak. The plenum, the coil-to-plenum joint, and the main trunk seams are the highest-pressure points and leak the most. A large fan-to-register gap with conditioned unconditioned space points here. Smoke-test and feel the seams; seal with mastic and re-measure the gap.

Branch 2, boot and branch connection leaks. Flex disconnected at the boot, an unsealed boot-to-drywall joint, or a torn flex inner liner leaks air into wall cavities and attics. The register downstream of the break starves. Reconnect, seal, and clamp every boot.

Branch 3, distribution imbalance (no leak). If the register total roughly equals the fan flow but the distribution is lopsided, the air is reaching registers, just the wrong ones. Rebalance dampers and verify branch sizing; the starved rooms gain flow as the hogging branches are damped.

Branch 4, measurement artifact. A fan-flow reading taken at a turbulent location, or a blower-table estimate against the wrong static, can overstate CFM. Re-measure with a traverse or a calibrated flow grid at a clean location, and trust the summed register flows as the delivered number.

Branch 5, leak into a return-side cavity (cross-talk). On platform returns or panned-joist returns, supply leakage into the return cavity recirculates conditioned air and starves rooms while inflating measured fan flow. The fan sees high flow because the leaked air loops straight back through the coil without ever reaching a register, so the blower curve looks healthy while delivery collapses. Inspect panned returns and seal supply leaks into them, and treat a building-cavity return as a leak source until proven sealed.

Confirming diagnosis

The confirming measurement is the fan-flow-versus-register-sum comparison, ideally backed by a duct-leakage test (a pressurization test that quantifies total leakage and, with the registers sealed and unsealed, separates leakage to outside from leakage to inside). A large leakage-to-outside number confirms the supply-side escape. Smoke-testing each suspect seam and boot under blower operation locates the specific gaps. After sealing, re-run the fan-to-register comparison; a closed gap with register flows now summing near fan flow confirms the leaks were the cause and the rooms recover.

Remediation

Seal the plenum, trunk seams, and coil-to-plenum joint with mastic or mastic tape; reconnect and clamp disconnected flex at boots; seal every boot-to-drywall joint. Insulate ducts that remain in unconditioned space. For imbalance, rebalance dampers and resize undersized branches per Manual D. For panned-return cross-talk, seal supply leakage into the return cavity or convert to a ducted return. Re-verify with a follow-up duct-leakage test and a fresh fan-to-register flow comparison; the delivered register total should rise to meet the fan flow and the starved rooms should reach design.

References

• ACCA Manual D (duct design, leakage, and distribution balancing)
• ANSI/RESNET/ACCA Standard 380 (duct-leakage testing methods)
• ACCA Manual J (room load and required register CFM)
• SMACNA HVAC Air Duct Leakage Test Manual (duct leakage classification and testing)
• ASHRAE Handbook, HVAC Systems and Equipment (duct leakage fundamentals)