HVAC Air Balancing: How to Measure and Set Airflow

Ten percent. That is the entire margin the NEBB procedural standard allows between an outlet’s measured airflow and its design airflow — on total system flow and at every terminal. Yet across most commercial buildings in Greater Montreal, nobody knows the real airflow at a single diffuser, because it has never been measured. That gap is what makes HVAC air balancing at once the highest-return and the most poorly executed procedure in the trade.

HVAC Air Balancing: What Does It Actually Accomplish?

Balancing a system means delivering to each space the airflow it needs — no more, no less — by distributing the fan’s available flow across the branches. It is not comfort tweaking on the fly. It is a measurement, a comparison against a written target, an adjustment, then a re-measurement. With no target, there is no balance; there is just guesswork at the dampers.

The distinction matters because an unbalanced system charges you twice. It charges in energy: a fan pushing surplus air into two zones so a third gets enough is burning power for nothing. And it charges in service calls, because draft complaints, one overheated room, and odours migrating between floors keep coming back until the distribution is fixed. Closing a damper by feel to silence one complaint relocates the problem into three other spaces.

Tools you need

As a field benchmark, a common capture hood such as the testo 420 seats a 24 in × 24 in hood, reads up to roughly 2,350 CFM and has a built-in flow straightener — enough for nearly every ceiling diffuser in an office building, but not for a rooftop outdoor-air intake, which has to be measured in the duct.

Step 1 — Set Target Airflows Before Touching Anything

Pull the mechanical drawings, the air balance schedule and the unit data sheets. That is the easy case. The real case across Montreal is a building renovated in slices since the 1980s, whose drawings show a duct system that no longer exists.

You then rebuild defensible targets: use and occupancy of each space, the unit’s real capacity, and readings from any earlier report. Write those targets down — that document is what will make the result verifiable. Quebec’s regulation respecting occupational health and safety also imposes obligations on air quality at work stations and on capturing contaminants at the source: in a shop, a lab or a commercial kitchen, those requirements set part of your targets before any comfort calculation enters the picture.

Step 2 — Prepare the System

A balance performed on an unprepared system has to be redone. Before the first reading:

• clean or new filters — a loaded filter corrupts the entire starting point;
• belts, sheaves and bearings verified; a slipping belt drops airflow with no alarm;
• branch dampers open; NEBB requires at least one fully open path from the air-moving equipment to a terminal, otherwise you are throttling the whole system;
• building in occupied configuration: interior doors, ceiling tiles and transfer grilles in place, washroom and kitchen exhaust running.

Step 3 — Measure Total Flow, Then Every Outlet

Start with total airflow at the fan. A duct traverse in a straight section, following the Log-Tchebycheff rule, establishes the system’s true starting flow. It is the only number that tells you whether you are facing a distribution problem to correct or a fan that cannot deliver the required airflow — two diagnoses, two budgets.

Then log every outlet — supply, return, exhaust — adjusting nothing. That initial snapshot is what will later prove the gain. Compute measured-to-target ratio for each: below 90% the outlet is under-served, above 110% it is over-served.

Step 4 — Adjust From the Farthest Outlet Inward

The order is not a preference, it is fluid mechanics. Every throttling move redistributes flow across the whole system. So you start at the branch farthest from the fan, throttle the over-served branches in small increments, and remeasure after each move. A damper turned a quarter turn can shift several hundred CFM three rooms away.

Once the split holds, re-check the total and the critical outlets. Then mark each damper’s final position and issue a target / initial / final table per outlet. Without that report, the next technician starts from zero.

⚠ Safety

Changing exhaust changes building pressure. In a building whose mechanical room holds natural-draft gas appliances, added depressurization can reverse the draft and spill combustion products indoors. Combustion air and venting requirements under the CSA B149.1 code, in force in Quebec, are not traded against an airflow target: verify draft after any change to the supply-versus-exhaust balance. And never block a combustion air intake to “stabilize” a room.

Field Case: The Conference Room That Was Always Too Warm

Three-storey office building in Laval, constant-volume rooftop unit, recurring complaints about a top-floor conference room — too warm during meetings, freezing the rest of the day. Two earlier visits had blamed a faulty thermostat.

The measurements told a different story. Total airflow at the fan came in at 88% of unit capacity: acceptable. But the conference room was getting 61% of its target, while two private offices near the unit were running past 130%. The cause was mundane: during a partition retrofit, a branch damper had been closed to kill a draft in a corridor, and a heat recovery ventilator added later had shifted the split again with no readings taken afterward.

The fix: reopen the damper, progressively throttle the two over-served offices, remeasure the total. Every outlet landed inside the ±10% window. The thermostat had never been the problem.

Why Does a Balanced System Drift in Quebec?

Because our two extreme seasons load the ductwork differently. A balance done in July, with wet cooling coils and outdoor-air dampers in summer position, does not describe the same system in February, when the economizer closes and midwinter thaws cycle equipment several times a day. On variable-air-volume or heat-recovery systems, a second-season check at the critical outlets is well worth the hours.

The other driver is structural: Montreal’s commercial stock changes in small increments. Partitions moved, spaces converted, exhaust fans added on flat roofs already crowded with equipment — each change distorts the split, and almost none triggers a reading. A data logger left in place for a few days after a balancing campaign confirms the airflows hold under real conditions, including overnight and through setback sequences.

The Mistakes That Sink a Balance

• Adjusting with no written target. You get numbers, never a demonstrable result.
• Balancing on dirty filters. The starting point is wrong, so everything downstream is too.
• Ignoring return and exhaust. Supply alone says nothing about room pressure or the direction of air cascades.
• Throttling to fix insufficient total flow. If the fan cannot deliver, closing dampers does not create air — it raises static pressure and power draw.
• Changing filter class without re-verifying. A higher MERV rating adds pressure drop and distorts the split, with the farthest outlets losing the most.

Three Readings Before You Open a Damper

Pick the space that generates the most complaints in your worst-performing building. Measure three things before you open a damper: total airflow at the fan, airflow at that room’s outlet, and airflow at the outlet closest to the unit. Compared against targets, those three numbers immediately separate a distribution problem from a capacity problem — and you know whether the answer costs two hours of adjustment or a capital project.

For whoever signs the invoices, that is where the value sits: a balance report with real numbers converts an open-ended complaint budget into a dated, checkable list of corrections. That is what the Montréal Combustion team hands over on ventilation and air quality mandates — airflows measured, logged and reproducible, instead of a damper set by instinct that somebody will have to go find five years from now.