VAV box commissioning and airflow field guide for HVAC

What a VAV box does, and why commissioning it matters

A VAV box, a variable air volume terminal unit, throttles the supply air to a single zone to hold that zone's temperature. It is a damper in a sheet-metal box with a controller and an airflow sensor, sitting above the ceiling between the trunk duct and the diffusers it feeds. As the zone cools off or heats up, the controller modulates the damper between a minimum airflow and a maximum airflow to deliver exactly the air the space needs and no more.

Commissioning the box is setting and verifying those airflow setpoints and the control sequence, then proving the box actually does what the controller says it does. The cooling maximum is the design airflow for the zone. The minimum is the ventilation and reheat floor. Between them the box modulates, and on a box with reheat it adds heat once it has throttled down to minimum and the zone still calls for warmth.

Most VAV problems are not the box. They are a setpoint nobody verified, a flow sensor reading a number that does not match the air actually moving, or a reheat valve cycling while the box sits at full cooling airflow. The hardware is usually fine. The commissioning is what nobody finished, and the building gets occupied with boxes that have never been driven through their own sequence.

VAV box types

VAV boxes split into a handful of types, and which one you have changes what you commission and how the sequence behaves. The first split is cooling-only versus reheat. A cooling-only box is just a controlled damper on cool supply air. A reheat box adds a coil downstream of the damper, hot water or electric, so the box can warm the air when the zone needs heat at minimum airflow.

The second split is single-duct versus dual-duct. A single-duct box, the common case, takes one stream of cool supply air and throttles it. A dual-duct box has two inlets, a cold deck and a hot deck, and mixes them to hit the zone temperature, so it carries two dampers and often two flow sensors to commission.

Then there are fan-powered boxes, which add a small fan to draw warm ceiling-plenum air. A parallel fan-powered box runs its fan only on a call for heat, cycling plenum air in behind the primary air when the box is at minimum. A series fan-powered box runs its fan continuously through every occupied mode, so the zone sees constant air volume off the box while the primary air damper modulates upstream of the fan. Commissioning a fan-powered box adds the fan, its interlock, and on a parallel box the downstream backdraft damper to the list.

What is a pressure-independent VAV box?

A pressure-independent VAV box measures its own airflow and resets the damper to hold the airflow setpoint regardless of what the upstream duct pressure is doing. The controller runs two loops. A temperature loop reads the zone and sets an airflow target somewhere between minimum and maximum. A flow loop then drives the damper to make that exact airflow, reading the box's own flow sensor and correcting the damper as duct pressure swings. This is the standard for modern systems, and it is why the boxes do not fight each other.

A pressure-dependent box has no flow sensor. The temperature loop just commands a damper position, and the airflow at that position rides up and down with whatever static the trunk happens to be carrying. Close one box on the trunk and the others get more air. The pressure-dependent box is mostly a relic now, found on older or budget jobs, and it cannot be commissioned to airflow setpoints because it does not know its own airflow.

The whole pressure-independent scheme rests on the flow sensor being calibrated and reading true. If the sensor or its K-factor is wrong, the box holds an airflow setpoint that does not match the air actually moving, and it does it confidently, because the controller believes its own number. That is the single most common reason a commissioned box still delivers the wrong air. The control loop is doing its job against a measurement that lies to it.

Pressure-independent

A VAV box with a flow sensor that holds its airflow setpoint regardless of upstream duct pressure changes

Pressure-dependent

A box with no flow sensor that commands a damper position only, so airflow rides with duct static

Flow loop

The control loop that drives the damper to make the airflow setpoint the temperature loop calls for

The airflow sensor and the K-factor

The airflow sensor sits at the round inlet of the box and is usually a multi-point averaging ring or cross, with ports facing into the airstream to read total pressure and ports facing away to read static. The difference is velocity pressure, and the controller turns that velocity pressure into CFM. It does this with the box's K-factor, a calibration number the box manufacturer publishes for that size and sensor and prints on a label stuck to the box.

The relationship is simple and unforgiving: airflow equals the K-factor times the square root of the sensor's velocity pressure. The square root is what bites. Velocity pressure at the box inlet is tiny, often hundredths of an inch of water column, so a small error in the sensor reading or the zero swings the CFM noticeably, and the lowest airflows are where the sensor signal is weakest and least trustworthy.

A wrong K-factor reads the wrong CFM at every setpoint, in proportion, so the box holds 480 CFM thinking it holds 600, or the reverse. The K-factor is specific to the box model and inlet size, so it cannot come from memory or from the box next to it. Pull it from the label or the manufacturer's catalog for the exact unit. The amplification factor, how much the sensor multiplies the true velocity pressure, is baked into that published K, which is why you do not invent one.

K-factor

The manufacturer's calibration constant that converts the inlet sensor's velocity pressure to airflow for that box size

Velocity pressure (VP)

The pressure from air motion the inlet sensor reads, converted to CFM through the K-factor

Flow ring / cross

The multi-point averaging sensor at the box inlet that reads total and static pressure to derive velocity pressure

Minimum and maximum airflow setpoints

A VAV box runs between three airflow numbers, and commissioning sets all of them in the controller. The cooling maximum is the design airflow for the zone, the most air the box delivers when the space is calling hard for cooling. The minimum airflow is the floor the box throttles down to. The heating airflow, on a reheat box, is the air the box delivers when it is heating, which on a modern sequence is a separate, higher number than the cooling minimum.

The minimum is the setpoint that gets argued over, because it serves two masters. It has to deliver the zone's ventilation minimum, the outside-air share ASHRAE Standard 62.1 says the occupants are owed, and it has to be low enough that the box is not overcooling a space that does not need the air. Set the minimum too low and you starve ventilation. Set it too high and you reheat a flood of cold air all winter, paying to cool it and heat it at once.

ASHRAE Standard 90.1 limits how much air a reheat box may reheat, to stop that simultaneous heating and cooling. The common figures you will see are a minimum on the order of 0.4 CFM per square foot, or 30 percent of the zone peak, or the ventilation requirement, whichever governs. The exact thresholds and the edition shift, so confirm the project's basis of design and the adopted energy code before you enter a number, because the minimum is where the energy code and the ventilation code both have a say.

Cooling maximum

The design airflow for the zone, the most air the box delivers on a full cooling call

Minimum airflow

The throttled-down floor, set to meet ventilation without overcooling, bounded by 62.1 and 90.1

Heating airflow

The airflow a reheat box delivers in heating, a separate setpoint from the cooling minimum on a dual-max sequence

Reheat, the dead band, and simultaneous heating and cooling

Reheat energizes when the box has throttled down to its minimum airflow and the zone still calls for heat. The sequence runs cooling first: as the zone warms, the box opens toward maximum. As it cools, the box closes toward minimum. Between cooling and heating sits the dead band, a temperature range where the box holds minimum airflow and the reheat valve stays shut, so the box is neither heating nor cooling and the controller is not chasing its tail across the setpoint.

Below the heating setpoint, the box reheats. On the older single-maximum sequence, it does this by opening the reheat valve while the box sits at one fixed minimum airflow. That works, but if the minimum was set high, the box is heating a large volume of cold primary air, which is exactly the simultaneous heating and cooling the energy code is trying to kill.

The trap field crews find is a reheat coil running while the box is up at full cooling airflow. That means the dead band or the heating logic is broken, or the flow sensor is lying and the box thinks it is at minimum when it is wide open. Either way you are pouring hot water into a stream of cold air, and the zone never settles. The reheat should only ever fire from the minimum or the heating airflow, never from the cooling maximum.

What is the dual-maximum reheat sequence?

The dual-maximum sequence gives a reheat VAV box two different maximum airflows, one for cooling and a separate, lower one for heating, instead of pinning the minimum at the heating airflow. It is the modern reheat strategy in ASHRAE Guideline 36, and it exists to let the cooling minimum drop way down without starving heat.

Here is the mechanism. On the old single-max sequence, the minimum airflow had to be high enough to carry the zone's heating load through the reheat coil, so the minimum was stuck up around 30 to 50 percent of design even when cooling. The dual-max breaks that link. In cooling, the box throttles to a true low minimum, often just the ventilation floor. In the dead band it holds that low minimum. On a call for heat, it first opens the reheat valve at the low minimum, and only if that is not enough does it raise the airflow up to a separate heating maximum while modulating the coil. So the box rarely sits at a high airflow with the coil on.

That is where the energy comes from. Less air reheated, far less of the year spent cooling and heating the same air. Commissioning a dual-max box means verifying both maxima, the low cooling minimum, the order the box raises airflow versus opening the valve, and that it never jumps the airflow up before the valve has done its work. Drive it through the whole range and watch the sequence, because the dual-max is easy to misconfigure into acting like a single-max.

Dual maximum

A reheat sequence with separate cooling and heating maximum airflows, allowing a low cooling minimum

Dead band

The temperature range between heating and cooling where the box holds minimum airflow and the reheat valve is shut

Single maximum

The older sequence with one fixed minimum that doubles as the heating airflow, so the minimum runs high

ASHRAE Guideline 36 high-performance sequences

ASHRAE Guideline 36 publishes standardized, vetted control sequences for VAV systems, so the VAV box logic, the air handler logic, and the resets are written the same way on every job that adopts it instead of being reinvented by each controls programmer. For a VAV box that means the dual-max reheat, the zone temperature loop driving an airflow setpoint, and the box generating requests up to the air handler. For the system it means the static pressure reset and the supply-air-temperature reset that the boxes drive.

The point of Guideline 36 is energy and repeatability. The sequences were built to cut the simultaneous heating and cooling, the over-pressurized ducts, and the fixed setpoints that older controls baked in, and to do it with logic that a commissioning agent can actually test against a written script. When the spec calls for Guideline 36, you are not guessing at intent. The sequence is documented, the requests and resets are defined, and the functional test follows the sequence step for step.

Commissioning to Guideline 36 means verifying the box runs the published sequence, not a controls house approximation of it. Confirm the dual-max behavior, confirm the box issues cooling and reheat requests when it cannot make setpoint, and confirm those requests actually move the air handler's static and temperature resets. The exact addenda and edition matter, since Guideline 36 has been revised and added to, so test against the version the project specifies.

The controller and the building management system

Every pressure-independent box carries a DDC controller, a small direct-digital controller mounted on the box that runs the temperature and flow loops and the reheat. It reads the zone temperature sensor, the inlet flow sensor, and the damper position, and it drives the damper actuator and the reheat valve or electric stages. That controller is where the setpoints, the K-factor, and the sequence live, and it is what you connect to when you commission the box.

The controller talks to the building management system over a network, commonly BACnet, so the box's points roll up to the head end: zone temperature, airflow setpoint and actual, damper position, reheat command, and the occupancy mode. The BMS pushes the schedule and the occupied and unoccupied setpoints down, and it carries the resets that tie the box to the air handler. A point that reads wrong at the box reads wrong at the head end, so verifying the points map correctly is part of the work.

Commissioning leans on this network. You can command the box from the head end, force airflow setpoints, override the damper, and watch the flow sensor respond, which is how you drive the box through its range without standing on a ladder for every step. But the override has to be confirmed back at the box, because a point that maps to the wrong object reports a number that has nothing to do with the air actually moving.

Static pressure reset and trim and respond

The air handler does not run at one fixed duct static. On a Guideline 36 system the static pressure setpoint resets up and down based on how open the VAV dampers are, so the fan delivers just enough pressure for the boxes to make their airflow and no more. The method is trim and respond. The setpoint trims down at a steady rate, and when boxes start running out of damper and cannot hold their airflow, they generate pressure requests that respond the setpoint back up.

The reason this matters at the box is that a VAV box can only make its airflow if the trunk has enough static to feed it. A box that pegs its damper wide open and still cannot reach setpoint is, often, a box on a system whose static has been reset too low or whose duct was undersized. The box is doing everything it can. The air is not there to deliver.

The critical-zone or worst-case reset is the heart of it: the system holds just enough static to satisfy the one box that needs the most, so the rest sit with their dampers off the stops and the fan is not over-pressurizing the whole duct to feed the average zone. Commissioning the boxes and commissioning the reset are the same job from two ends. Verify the box generates a pressure request when it cannot make airflow, and verify that request actually moves the air handler's static setpoint. See the duct static guide for measuring and reading the static the boxes are working against.

Trim and respond

A reset method that lowers a setpoint at a fixed rate, then raises it in response to requests from zones that cannot keep up

Static pressure reset

Resetting the air handler's duct static setpoint to the lowest pressure that still lets the boxes make airflow

Critical zone

The worst-case box driving the reset, the one needing the most static or coldest air at any moment

How do you verify VAV box airflow?

You verify VAV box airflow by measuring the air actually delivered to the zone and comparing it to what the controller says the box is making, at both the minimum and the maximum setpoint. The controller's number comes off the inlet flow sensor and the K-factor. The real number comes off a flow hood at the diffusers or a duct traverse downstream of the box. The two should agree within a reasonable band, and when they do not, the controller is the one that is wrong until proven otherwise.

Command the box to its cooling maximum, let it settle, and read every diffuser on the box with the flow hood. Sum them and compare to the controller's reported CFM. Then command the minimum and do it again, because a box can read close at max and badly off at min where the sensor signal is weakest. The minimum is the harder, more important check, since that is where ventilation lives and where the square-root error in the sensor is largest.

Two real-world catches. Duct leakage between the box and the diffusers means the hood at the registers reads less than the box truly delivered, and on a small box that leakage can run over 100 CFM, more than 20 percent of the minimum. A duct traverse just downstream of the box, before the takeoffs, takes the leakage out of the comparison. And the box has to be in steady state, not mid-stroke, or you are comparing a moving target. This is the same flow-hood and traverse work the air balance runs, so the box airflow verification and the balance reconcile against each other.

The functional test

The functional test drives the box through its whole sequence and confirms it does what the written sequence of operation says, point by point. You force the box to minimum, to maximum, and into reheat, and at each step you watch the damper, the flow sensor, the reheat valve, and the zone response. The test is not a glance at the head-end graphic. It is making the box move and confirming the hardware follows.

Run the steps in order. Drive a cooling call and confirm the damper strokes open and the airflow climbs to the cooling maximum, with the flow sensor tracking. Drive into the dead band and confirm the box holds minimum with the reheat valve shut. Drive a heating call and confirm the reheat valve opens at minimum airflow first, and on a dual-max box, that the airflow only rises to the heating maximum after the valve has opened. Then check the failure mode: drop power or air to the actuator and confirm the damper goes to the position the sequence demands, fail-open or fail-closed as specified.

Point-to-point comes first, before the sequence. Confirm the zone sensor reads true against a handheld, the damper command matches the actual blade position, the flow sensor reads zero with the fan off and a real number with it on, and the reheat command actually opens the valve. A sequence test on top of a bad point just chases a ghost. Verify the points, then verify the sequence the points feed.

Calibrating the controller reading to the measured airflow

When the controller's airflow disagrees with the hood or the traverse, you calibrate the box so its internal CFM matches the measured air. The controller carries an adjustment for this, an offset and a gain on the flow reading, or a field correction to the K-factor. You enter the measured airflow at known setpoints and let the controller re-fit its reading to reality. After that, the box holds setpoints that mean the same thing the balancer measured.

Calibrate at more than one point. A single-point correction at maximum can leave the minimum off, because the sensor's error is not flat across the range, and the minimum is where it matters most. The common practice is to read the box at a high airflow and a low airflow, give the controller both measured values, and let it set the gain and offset so it tracks across the span. Then re-read to confirm the correction took.

This is also where the controls tech and the balancer have to actually talk. The balancer takes the hood readings, the controls tech enters the correction, and if they work in separate weeks off separate numbers the box ends up calibrated to a measurement nobody can reproduce. Do it together, record the measured values and the correction applied, and the box is calibrated to a number that has a name and a date behind it.

Occupancy and demand-control ventilation

A VAV box does not hold the same minimum around the clock. The schedule and occupancy mode change it. In occupied mode the box holds its ventilation minimum and modulates for comfort. In unoccupied mode it can drop the minimum or close down to setback, since there is nobody to ventilate, and reopen on a night or morning warmup or a tenant override.

Demand-control ventilation, DCV, takes this further by resetting the minimum on real occupancy instead of a fixed schedule. A CO2 sensor in the zone, or a people count, tells the system how many occupants are actually present, and the box's outside-air minimum scales with them. An empty conference room drops to a low ventilation floor. A packed one drives the minimum up to cover the people in it. ASHRAE Standard 62.1 recognizes this dynamic reset, and it saves the energy of ventilating empty space.

Commissioning DCV means proving the reset actually moves the box, not just that the CO2 sensor reads a number. Force a high CO2 value and confirm the minimum airflow rises. Drop it and confirm the box settles back to its base ventilation minimum, not to zero. The failure that hides here is a CO2 sensor that drifted or a reset that is mapped but does nothing, so the box reads a fancy graphic at the head end while holding one fixed minimum all day. Verify the linkage end to end, the same way the economizer and outside-air work get verified during the balance.

Critical-space and data center VAV

Some zones cannot tolerate the loose tolerances a general office accepts, and the VAV commissioning tightens up accordingly. A lab, an operating room, an isolation room, or a cleanroom runs a directional pressure requirement on top of the airflow numbers, so the box has to hold a tight supply-to-exhaust offset to keep air flowing the right way across a door. There the flow sensor calibration is not a nicety. It is the thing keeping contaminated air out of a clean space or in a dirty one.

On these zones the boxes often run in tracking pairs, a supply box and an exhaust or return box, where the controller holds a fixed CFM difference between them so the room stays positive or negative no matter how the airflow modulates. Commissioning a tracking pair means verifying the offset holds across the whole range, not just at one airflow, and that a fast change in one box does not let the room flip pressure before the other catches up.

Data center and equipment-cooling zones lean on tight, stable airflow for a different reason, keeping the inlet temperature inside the band the equipment wants, with redundancy so a single box or fan failure does not starve the load. The commissioning emphasis shifts toward stability and failure response. Verify the box holds its airflow without hunting, and verify what happens on a loss of a unit, because in a critical space the failure mode is the test that matters most.

Why can't my VAV box make airflow?

A box that cannot reach its airflow setpoint with the damper wide open is almost always short on static, not broken. The trunk feeding it does not have the pressure to push the design air through an open damper, because the static reset has driven the setpoint too low, the duct to the box is undersized or crushed, or the air handler itself is not making pressure. Check the static at the box inlet before you touch the controller. No pressure, no air, and no setpoint change fixes that. The duct static guide walks the static diagnosis.

The other failures have their own signatures. A hunting box, one whose airflow swings and never settles, is usually a flow loop tuned too aggressively or a flow sensor seeing turbulent air because the box was installed right off an elbow with no straight inlet duct. A stuck damper reads a command that does not match the blade position, so the point-to-point catches it. Reheat running at full airflow means the dead band or heating logic is wrong, or the flow sensor is lying and the box thinks it is at minimum.

And the quiet one: a wrong K-factor. The box holds its setpoint perfectly, the controller is happy, and the air delivered is simply wrong because the conversion from velocity pressure to CFM is off. Nothing alarms, nothing hunts. You only catch it by putting a hood or a traverse on the box and finding the measured air does not match the reported air. That is why the airflow verification is not optional. It is the only thing that finds the box that is confidently wrong.

The TAB and commissioning interface

Test and balance and commissioning are two scopes that meet at the VAV box, and keeping them straight saves a lot of finger-pointing. The TAB technician sets and measures the airflow: confirms the box makes its minimum and maximum, takes the hood readings, and feeds the measured numbers into the controller calibration. The commissioning agent verifies the sequence: that the box runs the dual-max, generates its requests, and responds to occupancy and resets the way the sequence of operation says.

The handoff runs one direction. You cannot commission a sequence on a box whose airflows were never set, so the balance is usually a prerequisite for the functional test. The CxA reads the TAB report, sees which boxes made design and which did not, and targets the functional testing at the ones with findings. A box the balancer flagged as unable to make airflow is a box the CxA tests against the static reset, not against the box controller.

The deliverable on both sides is a record. The balance report shows design versus measured airflow at min and max per box. The commissioning record shows the functional test results, the sequence verified, and the issues logged and closed. The two reconcile, and a box is not done until both say so. See the air balancing guide for the TAB side of this work and how the report gets certified.

What to document

A commissioned box that nobody can reconstruct later is a box that gets re-commissioned. The record is what answers the question two years out when a zone runs cold and the question is whether the box was ever right. Capture enough that the next tech can pick up the baseline without standing on a ladder.

For each box, record the box tag and the size, the design minimum and maximum CFM against the measured minimum and maximum, the K-factor used, the reheat type and whether it fired correctly at minimum, the calibration correction applied, and which sequence was verified. The table below is the core per-box record. Tie it to the balance report and the functional test results so the three documents agree on the same numbers.

Common mistakes

• Trusting the controller's airflow without ever putting a hood or traverse on the box to confirm it.
• Entering a wrong or generic K-factor instead of the published value for the exact box size.
• Setting the minimum airflow too low to meet the zone's ventilation requirement under 62.1.
• Leaving reheat able to run at full cooling airflow because the dead band or heating logic is wrong.
• Commissioning the boxes without verifying the air handler static reset that feeds them.
• Calibrating only at maximum, so the minimum, where ventilation lives, is left reading off.
• Mapping a CO2 or occupancy reset to the head end that looks live but never moves the box minimum.
• Calling a hunting box a controller problem when it was installed right off an elbow with no straight inlet.

Field checklist

Standards and references

ASHRAE Guideline 36 owns the high-performance control sequences for VAV systems, the dual-max reheat and the trim-and-respond resets a commissioned box is tested against. It has been revised and added to, so verify the edition and the addenda the project specifies before you test to it. ASHRAE Guideline 0 and the commissioning process guides describe how the commissioning itself is run, the functional testing and the issues log that close a system out.

ASHRAE Standard 62.1 sets the minimum ventilation rates the box's minimum airflow has to deliver. ASHRAE Standard 90.1 sets the energy limits that cap how much air a reheat box may reheat, which is what drives the low cooling minimum and the dual-max in the first place. The specific minimum airflow thresholds, commonly cited around 0.4 CFM per square foot or 30 percent of zone peak, move between editions, so confirm them against the adopted energy code and the basis of design rather than treating the round number as fixed.

On the airflow side, NEBB and AABC publish the test and balance procedural standards and certify the firms that set and measure the box airflows, and ASHRAE Standard 111 covers the measurement methods behind the readings. Above all of it sit the box and controller manufacturer, whose published K-factor and setpoint ranges govern the box, and the project specification and engineer of record, who set the sequence and accept the work. Cite the body that owns the point, and confirm the edition, because these documents revise on their own cycles.

Units, terms, and conversions

VAV commissioning carries its own vocabulary, and the same idea reads differently across a controls submittal, a balance report, and a sequence of operation.

Airflow is CFM, cubic feet per minute, in the field, and liters per second or cubic meters per hour in metric sources, where 1 CFM is about 0.472 liters per second. The flow sensor's velocity pressure is inches of water column, in. wg or in. w.c., where 1 in. wg is about 249 pascals, and at a box inlet it runs in hundredths of an inch. The box goes by VAV box, VAV terminal, or terminal unit. The control sequence goes by the dual-max, the Guideline 36 sequence, or simply the sequence of operation, and the air handler reset goes by trim and respond.

VAV box

Variable air volume terminal unit; a controlled damper, flow sensor, and optional reheat that conditions one zone

Pressure-independent

A box that holds its airflow setpoint regardless of upstream duct pressure, using its own flow sensor

K-factor

The manufacturer's calibration constant converting inlet velocity pressure to CFM for that box size

Min / max airflow

The throttled floor and the design ceiling the box modulates between; min carries the ventilation requirement

Reheat dead band

The temperature range where the box holds minimum airflow with the reheat valve shut

Dual maximum

A reheat sequence with separate cooling and heating maximum airflows, allowing a low cooling minimum

Guideline 36

ASHRAE's standardized high-performance control sequences for VAV systems and air handlers

Trim and respond

The reset that lowers the air handler static or temperature setpoint until zone requests push it back up

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is air balancing?
• How do I measure total external static pressure?
• What is an air-side economizer?
• What is an energy recovery ventilator (ERV)?
• What is a fan coil unit?