Economizer and demand-control ventilation commissioning field guide for HVAC

Economizers and demand-control ventilation, and why both go quiet

An air-side economizer uses cool outside air to cool the building for free when the outdoor air is cooler or drier than the return air, so the compressors stay off or run less. Demand-control ventilation, DCV, modulates the outside air to the number of people actually in the space instead of ventilating for a full house all day. Two different ideas, two different paydays, and they share the same outside-air damper.

The economizer saves cooling energy. DCV saves the energy you spend conditioning ventilation air you do not need. On a packaged rooftop unit they often live in the same mixing box and run off the same controller, which is why a tech who understands one has to understand the other to set either correctly.

Here is the part that pays the bills and gets ignored. Both fail quietly. An economizer with a stuck damper or a dead sensor does not throw a code on most equipment. It just stops saving, and the building runs the compressors when it could be cooling for free, month after month, with nobody the wiser until someone puts a meter on it. The control is only as good as the commissioning nobody finished.

The air-side economizer: dampers, mixed-air sensor, and the sequence

An air-side economizer is a set of three linked dampers and a sensor that lets a unit cool with outside air. The outside-air damper opens to bring in cool air, the return-air damper closes to make room for it, and the relief or exhaust damper opens to dump the extra air the building cannot hold. The three move together, driven by one actuator or a linked set, so that as outside air comes up, return air goes down.

The mixed-air sensor sits downstream of where outside and return air blend, and it is the feedback the controller uses to modulate the dampers to a mixed-air temperature setpoint, commonly around 55°F. On a call for cooling, if the outdoor conditions are below the changeover, the controller opens the outside-air damper toward 100 percent and trims it to hold that mixed-air setpoint before it ever stages a compressor.

The sequence has an order. The unit calls for cooling, the high-limit check says free cooling is available, the dampers modulate to hold mixed air at setpoint, and only if the dampers cannot hold it does mechanical cooling stage on to finish the job. Get the order wrong, or let a bad sensor break any link in it, and the unit either ignores free cooling or chases it when it should not.

What is the economizer high-limit, and where do you set it?

The high-limit, or changeover, is the control that decides when outside air is too warm or too humid to be worth using, and it locks the economizer out above that point. Get this setpoint wrong and the economizer either gives up free cooling it could have used or drags in hot, humid air and makes the compressors work harder. It is the single setting that decides whether the economizer helps or hurts.

There are four common high-limit types. Fixed dry-bulb shuts the economizer off above a set outdoor temperature. Differential dry-bulb compares outdoor to return temperature and economizes whenever outdoor is cooler. Fixed enthalpy uses total heat, folding humidity into the decision so it does not pull in cool but soggy air. Differential enthalpy compares outdoor to return total heat. Enthalpy controls suit humid climates. Dry-bulb controls are simpler, and in dry climates they often beat enthalpy because they do not depend on a drifting humidity sensor.

ASHRAE Standard 90.1 ties the allowed high-limit type and setpoint to climate zone. For fixed dry-bulb the listed shutoff setpoints fall in roughly the 65°F to 75°F band depending on zone, with the drier zones allowed the higher numbers. Confirm the exact type and setpoint against the adopted edition of 90.1 and the local energy code, because the table has shifted across editions and the jurisdiction's version controls.

Why do economizers fail in the field?

Economizers fail because they are a moving mechanical control with sensors, sitting on a rooftop, that nobody looks at after startup. Field studies have found broken or non-functional economizers at high rates. A New Buildings Institute review of rooftop units found them non-functional or problematic on the order of 70 percent of the time, and other studies put the share of broken economizers even higher. The economizer is widely treated as the most common HVAC control fault, which is exactly why automated fault detection exists for it.

The failure list is short and repetitive. The outside-air damper sticks, seizes, or its actuator dies, so it never opens or never closes. The linkage between the actuator and the damper blade comes loose or slips, so the actuator strokes and the damper does not. The changeover sensor drifts or fails, locking the economizer out or holding it wide open. The controller is mis-set, the wiring is wrong, or someone disabled the economizer to chase a different complaint and never turned it back on.

The reason it stays broken is the reason it is worth a guide. A failed economizer rarely makes the building uncomfortable. The compressors just pick up the load the free cooling should have carried, the space stays at setpoint, and the only symptom is an energy bill nobody traces back to a stuck damper on the roof.

Fault detection and diagnostics for the economizer

Automated fault detection and diagnostics, FDD, watches the economizer and reports when it stops behaving, because the failures are invisible without it. Some energy codes now require it. California's Title 24, Part 6 requires economizer FDD on air-cooled unitary equipment above a cooling-capacity threshold, around 54,000 Btu/h, that has an economizer, and recent editions of ASHRAE 90.1 carry an FDD requirement as well. Confirm the trigger and the edition against the code the jurisdiction adopted.

What FDD watches is the list of ways an economizer breaks. It flags a failed or out-of-range air temperature sensor, the economizer not economizing when free cooling is available, the economizer running when it should be locked out, a damper that is not modulating, excess outside air, and a damper stuck open, closed, or at minimum. It checks the actuator for a disconnected or unplugged linkage. Then it reports the fault somewhere a person will see it, to a DDC front end or annunciated at the thermostat.

FDD does not fix anything. It makes the silent failure loud. The value is that a stuck damper now generates an alarm instead of an invisible run of wasted compressor hours, so a building operator can send someone to the roof before the savings have leaked away for a season.

What is demand-control ventilation, and what is the CO2 setpoint?

Demand-control ventilation modulates the outside air to the real occupancy of a space instead of ventilating for design occupancy whenever the unit runs. In a CO2-based system, a sensor in the space reads carbon dioxide as a proxy for how many people are present, and the controller opens the outside-air damper above the minimum as CO2 rises and closes it back toward minimum as the space empties. You stop conditioning ventilation air for a crowd that is not there.

The setpoint is where people get it wrong. The old rule of a fixed 1000 ppm is a myth that overventilates or underventilates depending on the outdoor level. What matters is the differential, the indoor CO2 above the outdoor CO2, because the steady-state difference is what tracks the per-person ventilation rate. Outdoor air is commonly around 400 to 450 ppm and rising, so a sensible target is a differential, not an absolute number. ASHRAE 62.1, in its 2022 edition, added differential CO2 limits aimed at DCV for this reason.

Where DCV is required comes from the energy side. ASHRAE 90.1 mandates DCV for densely occupied spaces, with the trigger in recent editions at a design occupancy of about 25 people per 1000 ft² in spaces over 500 ft² served by an economizer, a modulating outside-air damper, or a large design outside-air flow. The exact threshold has tightened across editions, so confirm the adopted one.

Minimum outside air and the floor DCV cannot cross

DCV modulates outside air, but it can never go below the minimum the space needs for ventilation, and that floor comes from ASHRAE 62.1. The Ventilation Rate Procedure sets the breathing-zone outdoor airflow as two parts added together: a people rate times the number of people, plus an area rate times the floor area. The people part is what DCV is allowed to modulate. The area part is fixed, because it covers the off-gassing of the building and its contents, which happens whether anyone is in the room or not.

That is the line a lot of DCV setups cross by accident. When the space empties, the controller can pull the outside air down to the area component, but not to zero. Drive it below the area-based minimum and you are no longer ventilating the building itself, only its missing occupants, and the indoor air quality the standard is protecting falls out from under you.

Setting and verifying the minimum is a measurement job, not a damper-position guess. You set the minimum outside-air damper position so the unit delivers the design minimum CFM at the intake, then confirm it with a measurement, the same way a balancer proves outside air during a test and balance. The air balancing guide covers measuring and proving outside air at the unit. The commissioning job is to confirm DCV rides above that floor, never under it.

How do you measure the actual outside air?

You measure outside air, you do not trust the damper position, because a damper at 20 percent open is not 20 percent of design flow. Measuring it is the hard part of the whole job, and that difficulty is exactly why minimum outside air gets faked on so many startups. The intake is built for low face velocity on purpose, to keep rain and snow out, and a low velocity is the worst case for accuracy on a hood or an anemometer. The intake is rarely a clean straight duct you can traverse, and on a rooftop the wind is swirling the air at the louver while you read it.

Three approaches get used. A direct traverse of the outside-air duct, where one exists with straight run to traverse, is the cleanest. A velocity grid across the louver face with the free area applied works when there is no duct, with all the error a turbulent, low-velocity louver brings. The temperature-mixing method backs the outside-air fraction out of the mixed-air, return-air, and outside-air temperatures, and it only works when the spread between return and outside is wide enough to be meaningful, so it is a winter tool, not a mild-day one.

Permanent airflow stations, thermal-dispersion or pitot-array, are the fix when the design needs continuous outside-air measurement for DCV or code compliance. Whatever you use, name the method in the record, because each one carries its own error and the next person needs to know how the number was gotten. The air balancing guide walks the outside-air measurement in detail.

The damper, the actuator, and the leakage class

The damper is where the economizer physically wins or loses, and a leaky damper loses around the clock. ASHRAE 90.1 requires low-leakage dampers on the outside-air and relief openings, with a maximum leakage commonly cited at 20 CFM per square foot at 1 in. wg for the larger dampers, tested to AMCA 500D, with a looser allowance for small dampers. A high-leakage damper that is supposed to be closed at minimum still bleeds outside air, which means you heat or cool air you never meant to bring in, every hour the unit runs.

The actuator drives the blades, and the linkage connects the two. Both are common failure points. A spring-return actuator gives you a defined failure position when it loses power or signal, and that position matters: outside-air dampers are usually set to fail closed so a dead actuator does not dump unconditioned air or freeze a coil in winter. End switches and position feedback let the controller and any FDD know the damper actually reached the commanded position instead of binding halfway.

Check the linkage by hand and by eye, not by the command. The actuator can stroke its full travel while a slipped set screw leaves the blade barely moving. That mismatch is one of the most common economizer faults there is, and it reads as a perfectly happy actuator on the controller while no air moves at the louver.

Relief, exhaust, and building pressure on economizer

When the economizer opens to bring in a flood of outside air, that air has to leave somewhere, or the building pressurizes hard and the doors blow open. The relief or exhaust path is the half of the economizer people forget. On full economizer the unit might be pulling 100 percent outside air, and all of it has to be relieved, so the relief damper or the relief and exhaust fans have to track the outside-air damper as it opens.

Get the relief wrong and the symptoms show up at the building, not the unit. Too little relief and the building goes strongly positive on economizer, exterior doors are hard to pull open or stand cracked, and the unit cannot actually bring in the outside air it is calling for because there is nowhere to put it. Too much exhaust running when the economizer is at minimum and the building goes negative, pulling infiltration through the envelope.

Verify the relief as part of the economizer test, not separately. Drive the unit to full economizer, watch the relief path open in step, and read building pressure relative to outside with a manometer. The building pressurization section in the air balancing guide covers the targets and how exhaust and makeup air interact. On the economizer side the job is confirming the relief opens enough to let the free cooling actually happen.

How do you commission an air-side economizer?

You commission an economizer by driving it through every mode and proving each one with a measurement, not by watching the controller report it is fine. The functional test walks the unit through mechanical cooling, integrated economizer, and full economizer, and confirms the changeover and the dampers at each step.

Run it in order. Force the high-limit to think outside air is warm and confirm the economizer locks out and the unit cools mechanically with the outside-air damper at minimum. Then force a free-cooling condition below the changeover on a cooling call and confirm the dampers drive open and the unit holds mixed-air setpoint on outside air alone, with the compressors off. Then add load past what the dampers can carry and confirm mechanical cooling stages on while the economizer stays open, which is the integrated mode. At each step, verify the changeover acts at the right point, watch the dampers actually stroke at the louver, confirm the sensors are calibrated, and measure the outside-air CFM at minimum and at full economizer.

The fault you are hunting is the gap between commanded and actual. The controller will tell you the damper is at minimum. Your job is to confirm the damper is at minimum and that minimum is the design CFM. Force the modes, measure the air, and write down what each mode actually did, because a functional test nobody documented is a startup, not a commissioning.

Commissioning demand-control ventilation

Commissioning DCV starts at the CO2 sensor, because sensor drift is what kills these systems. A CO2 sensor that has drifted high holds the damper open and throws away the savings. One that has drifted low starves the space of ventilation. Check the sensor against a known reference or fresh outdoor air, not against itself, and confirm it reads outdoor air near the real outdoor level before you trust anything it says indoors.

Then prove the modulation does what it should. Inject a known CO2 source or use real occupancy to raise the level, and confirm the outside-air damper opens above minimum as CO2 climbs and returns toward minimum as it falls. Watch the two ends. As the space empties, the damper has to settle at the minimum-outside-air floor and not below it, which is the link back to 62.1 that most DCV setups get wrong. As the space fills to design occupancy, the damper has to open enough to deliver the full design ventilation rate.

The high-occupancy case is the one to test hard, because that is when underventilation actually harms people. Confirm the system can reach design outside air, not just modulate around the middle. A DCV setup that looks fine at half occupancy and cannot open far enough when the room is packed has failed at exactly the moment it mattered, and the only way you find that is by driving it there during commissioning.

Integrated vs non-integrated economizer

An integrated economizer runs free cooling and mechanical cooling at the same time. A non-integrated one locks the economizer out the moment a compressor starts. Integrated is the efficient way and the one ASHRAE 90.1 requires for most equipment, because partial free cooling still counts. When outside air can carry part of the load but not all of it, an integrated economizer holds the dampers open and lets the compressor make up only the difference, so the unit pulls the smallest possible amount of mechanical cooling.

A non-integrated economizer throws that away. The first stage of mechanical cooling slams the economizer to minimum, so on a mild day just past the point where free cooling alone can hold setpoint, the unit runs full mechanical cooling with the free-cooling dampers shut, when it could have been running the compressor against a pre-cooled mixed-air stream. The penalty lands exactly in the mild weather where the economizer should be earning the most.

Confirm integration during the functional test. Drive the unit to the point where free cooling cannot hold setpoint alone, then add load and watch whether the compressor staging closes the economizer or leaves it open. If the dampers slam shut when the first compressor starts, you have a non-integrated sequence or a control set up wrong, and on equipment 90.1 requires to integrate, that is a finding.

Sensor calibration and drift

Every sensor the economizer and DCV depend on drifts, and an uncalibrated sensor wastes more energy than the control was ever going to save. The dry-bulb temperature sensors, the enthalpy sensors with their humidity element, and the CO2 sensors all wander from true over time, and because the failures are silent the drift goes unnoticed until someone checks.

The enthalpy and CO2 sensors are the worst offenders. A humidity element on an enthalpy changeover ages and reads wet or dry, which moves the changeover point and either locks out good free cooling or pulls in air that should have been rejected. CO2 sensors using older technology can drift a meaningful fraction in a year, which directly moves how far DCV opens the damper. This is why dry-bulb changeover often beats enthalpy in a dry climate: fewer sensors to drift, and the one that remains is easy to check.

Recalibrate on a cadence, commonly annual, and confirm the manufacturer's interval and method, because some CO2 sensors self-calibrate against background and others need a gas reference. Check the temperature sensors against a known thermometer in the same airstream. The math is blunt. A changeover sensor reading a few degrees off, or a CO2 sensor reading a couple hundred ppm off, gives back more in wasted compressor hours or wasted ventilation than the whole control strategy was saving. The calibration is not maintenance you skip to save money. It is the maintenance that protects the savings.

Free cooling and the data center envelope

Airside economization is a large lever on data center efficiency, where it shows up directly in PUE. A facility that can run its cooling on outside air for a big share of the year runs its compressors and chillers far less, and in cool or cold climates the economizer hours dominate. The same mixing-box physics as a rooftop unit, scaled up and run almost continuously, which is why free-cooling hours are a headline number in data center cooling design.

What opened up more of those hours was the thermal envelope. ASHRAE TC 9.9 widened the recommended and allowable temperature and humidity ranges for IT equipment across its editions, with the A1 through A4 classes letting operators accept higher inlet temperatures. Every degree of allowable inlet temperature adds economizer hours, because outside air that used to be too warm to use is now inside the envelope. Many locations can economize for a large fraction of the year inside the recommended range, and more with the wider allowable range.

The commissioning discipline carries straight over. The changeover logic, the damper and actuator integrity, the relief path, and the controls that decide when to bring in outside air are the same problems at data center scale, with bigger stakes on the energy and bigger risk if humidity or contamination control on the outside air is not handled. The duct static pressure guide covers the airflow side of data center cooling and the underfloor plenum that delivers it.

Energy savings, payback, and the retro-commissioning find

The savings from an economizer and DCV are real, and they are also the first thing to disappear when nobody checks. A working air-side economizer can cut cooling energy substantially in a climate with enough mild hours, and DCV cuts the conditioning load on ventilation air in spaces whose occupancy swings, like classrooms, gyms, theaters, and conference rooms. The combination is one of the better paybacks in commercial HVAC when both actually run.

The catch is the word actually. The savings exist only while the damper strokes, the changeover is set right, and the sensors read true. A stuck or disabled economizer saves nothing, and because the building stays comfortable, the lost savings never generate a complaint. That gap between assumed and delivered is why economizers are a favorite target of retro-commissioning.

The classic retro-commissioning find is exactly this: a building full of economizers that were commissioned at startup, drifted or failed within a few years, and have been running on mechanical cooling ever since while the energy model assumed free cooling. The fix is often cheap, a new actuator, a fresh sensor, a corrected setpoint, a reconnected linkage, against savings that show up immediately on the bill. The economizer is the most common HVAC fault and one of the cheapest to fix, which is the whole argument for putting a meter and a functional test on it instead of trusting that it still works.

What to document

A commissioned economizer that nobody documented is a startup, and it gives the next person nothing to check against. Record enough that the operator, the next agent, or you in two years can tell whether it still does what it did the day it was set.

Capture the unit, the high-limit type and setpoint, the minimum outside-air CFM and how it was measured, the economizer modes verified and what each did, the CO2 setpoint or differential and the DCV floor, the sensors and their calibration dates, the damper failure position, and any deficiency. The table below is the core record.

Common mistakes

• Leaving the outside-air damper stuck or leaking, so the economizer never opens or never closes.
• Skipping sensor calibration, so a drifted changeover or CO2 sensor wastes more than the control saves.
• Running a non-integrated sequence that locks the economizer out the moment a compressor starts.
• Driving DCV below the area-based minimum outside air, so the building itself goes unventilated.
• Setting DCV to a fixed 1000 ppm instead of a differential above the outdoor CO2 level.
• Forgetting the relief path, so the building over-pressurizes and the economizer cannot bring in the air it calls for.
• Trusting the commanded damper position instead of measuring the actual outside-air CFM.
• Setting the high-limit by habit instead of the climate-zone value the energy code allows.

Field checklist

Standards and references

ASHRAE Standard 90.1, the energy standard, is where most of the economizer rules live: when an economizer is required, the high-limit changeover types and climate-zone setpoints, the integration requirement, the low-leakage damper limits tested to AMCA 500D, and in recent editions the economizer FDD requirement. ASHRAE 90.1 also mandates demand-control ventilation for densely occupied spaces above a design-occupancy threshold that has tightened across editions, commonly cited around 25 people per 1000 ft².

ASHRAE Standard 62.1, the ventilation standard, sets the minimum outside air the economizer and DCV have to respect. The Ventilation Rate Procedure gives the breathing-zone outdoor airflow as a people component plus an area component, and the 2022 edition added differential CO2 limits aimed at CO2-based DCV. The area component is the floor DCV cannot cross. California's Title 24, Part 6 requires economizer FDD on unitary equipment above a cooling-capacity threshold with an economizer, and other jurisdictions are adding similar requirements.

On the measurement and balance side, NEBB and AABC publish the procedures and certify the firms that verify outside air and damper performance, and AMCA owns the damper-leakage and fan-performance test methods. For data center cooling, ASHRAE TC 9.9 sets the thermal envelope that governs how many economizer hours are available. Cite the body that owns the point, and confirm the edition and section, because all of these revise on their own cycles and the adopted version with local amendments controls.

Units, terms, and definitions

Economizer and ventilation work carries its own vocabulary, and the same idea reads differently across a controls submittal, an energy code, and a balance report.

Outside air is OA or OSA, return air is RA, and mixed air is MA. Airflow is CFM in the field and liters per second or cubic meters per hour in metric work. The high-limit goes by changeover or lockout. Enthalpy, total heat, is in Btu per pound of dry air. CO2 is in parts per million, ppm, and DCV runs on the differential above outdoor, not an absolute number. Minimum OA is the ventilation floor from 62.1.

Economizer

A control that uses cool or dry outside air for free cooling when conditions allow, cutting compressor run time

Changeover / high-limit

The setpoint that locks the economizer out when outside air is too warm or too humid to use

Enthalpy

Total heat of the air, sensible plus latent, in Btu per pound, used by enthalpy changeover to account for humidity

DCV

Demand-control ventilation, modulating outside air to actual occupancy, commonly using CO2 as the occupancy proxy

CO2 differential

Indoor CO2 above outdoor CO2, the value that tracks per-person ventilation, not a fixed 1000 ppm

Minimum outside air

The ventilation floor from ASHRAE 62.1, a people component plus an area component, that DCV cannot go below

FDD

Fault detection and diagnostics, the automated monitoring that reports economizer faults a person cannot see

Integrated economizer

An economizer that runs at the same time as mechanical cooling so partial free cooling is not wasted

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 economizer fault detection?
• How thick should pipe insulation be?
• How many footcandles do you need for an office?