Dedicated outdoor air system (DOAS) field guide for HVAC

What a DOAS is, and why ventilation gets its own unit

A dedicated outdoor air system (DOAS) is a separate piece of equipment that conditions only the ventilation outdoor air a building is required to bring in, and nothing else. It draws in the outside air, dries it down to a low dewpoint, tempers it, and ducts it to the spaces. The room's cooling and heating, the sensible load, is left to a second system running in parallel. One unit owns the fresh air and the humidity. Another owns the temperature.

That split is the whole idea. In a conventional setup, a single rooftop unit or air handler mixes a little outdoor air into a lot of return air and tries to do everything at once: cool the room, heat the room, and dehumidify the fresh air, all off the same coil and the same thermostat. A DOAS pulls the ventilation job out and gives it dedicated equipment that can be controlled on humidity instead of temperature.

You see DOAS on schools, offices, labs, healthcare, multifamily, and any building where the ventilation rate is high and the indoor humidity matters. It pairs naturally with energy recovery on the outdoor air and with parallel sensible systems like fan coils, chilled beams, VRF, or radiant. The Anvilfield ERV commissioning guide and the economizer and demand-control ventilation guide cover the pieces that bolt onto it.

DOAS

Dedicated outdoor air system, a unit that conditions only the ventilation outdoor air and handles the latent load

Latent load

The moisture load, the energy needed to remove water vapor from the air, separate from temperature

Sensible load

The temperature load, the heat that changes air temperature without changing its moisture content

What problem does a DOAS solve?

A DOAS solves the part-load humidity problem that a single mixed-air system cannot hold. A standard rooftop or VAV unit dehumidifies as a side effect of cooling. The coil gets cold, moisture condenses on it, and the air leaves drier. That works as long as the building needs cooling. The moment the sensible load drops, on a mild humid day, at night, or in a shoulder season, the thermostat is satisfied and the cooling cuts back, so the coil warms up and stops wringing out water. The fresh air keeps coming in wet, and the space drifts to 60 percent relative humidity and higher while the temperature reads fine.

That is the trap. The room is at 73 degrees and clammy, the occupants are uncomfortable, and the thermostat sees no problem because temperature is on setpoint. Mold complaints, sticky surfaces, and corroded equipment follow. The mixed-air unit was never sized to dehumidify when there is no sensible load to drive the cooling, and adding reheat to force it bolts a band-aid onto a system fighting itself.

A DOAS breaks that link. It dries the outdoor air separately, on a humidity control, regardless of whether the room needs cooling. The latent load enters the building already handled at the source, so the parallel sensible system can run only when there is a temperature load and never has to carry humidity it was not built for.

How is a DOAS different from a regular HVAC system?

A regular HVAC system mixes outdoor air with return air and conditions the blend off one coil, so ventilation, cooling, and dehumidification all ride the same control. A DOAS separates the two loads on purpose. The DOAS conditions the outdoor air and carries the building's latent load. A parallel system carries the room sensible load. This is the decoupling, and it works because in most climates the outdoor ventilation air is the primary source of moisture in the building.

Because the humidity rides in on the fresh air, a properly designed DOAS can take on essentially all of the space latent load plus a slice of the sensible, leaving the parallel sensible-only system to make up the rest of the temperature load. The two systems can then be sized for what each actually does, instead of one oversized box trying to cover the worst-case of both at once. That right-sizing is where the efficiency comes from.

The practical consequence on the job is that you commission a DOAS on dewpoint, not on supply temperature alone. A mixed-air unit is judged by whether the room hits setpoint. A DOAS is judged by whether the air it delivers is dry enough to absorb the building's moisture across the whole year, including the mild wet days when the sensible system is barely running.

The parallel sensible system

The parallel sensible system is whatever handles the room temperature load once the DOAS has taken the ventilation and the moisture off the table. Because the latent load is gone, this system runs dry, which means it can use a warmer chilled-water temperature and terminal devices that would sweat if they had to wring out moisture. That is the opening that makes chilled beams and high-temperature radiant practical.

The common choices each have a fit. Fan coils are the workhorse, cheap and familiar, and they tolerate a little residual moisture. Four-pipe fan coils give you simultaneous heating and cooling across zones. VRF gives tight zoning and good part-load efficiency and pairs cleanly with a DOAS for the ventilation it cannot provide on its own. Chilled beams and radiant panels run with no moving parts in the space and stay quiet, but they live or die on the DOAS holding the dewpoint, because condensation on a chilled beam is a ceiling stain and a callback.

The choice is a design decision driven by the building, the zoning, and the budget, not a default. What matters in the field is that the parallel system is matched to the residual sensible load after the DOAS contribution, and that its surface temperatures stay above the space dewpoint the DOAS is maintaining.

Does a DOAS need energy recovery?

Most of the time, yes, and often the energy code requires it. A DOAS conditions 100 percent outdoor air, which is the most expensive air to treat, because there is no return air to dilute it. Recovering energy from the exhaust the building is throwing away to pre-condition that incoming outdoor air is how you make the ventilation affordable. A total energy wheel, an enthalpy wheel, transfers both heat and moisture from the exhaust to the entering air, so the DOAS coil starts from a milder, drier condition and does far less work.

ASHRAE 90.1 pushes this onto the drawings. The energy-recovery requirement is a table keyed to climate zone and the percentage of outdoor air, and a common trigger is supply airflow at or above 5,000 cfm with 70 percent or more outdoor air, with a minimum enthalpy recovery effectiveness commonly cited around 50 percent. A DOAS is 100 percent outdoor air by definition, so it lands in the required range on almost any sizable system. The thresholds have tightened across code cycles and stricter standards like ASHRAE 189.1 ask for more, so confirm the adopted edition.

On the latent-heavy job, the enthalpy wheel is doing real work on the moisture, not just the temperature. The Anvilfield ERV commissioning guide covers how to balance the two airstreams, control frost, hold cross-contamination down, and prove the measured effectiveness against the AHRI 1060 rating instead of assuming the wheel is recovering what the schedule claims.

Neutral supply versus cold supply

After the DOAS dries the air, you decide what temperature to deliver it at, and that choice splits into neutral or cold. Neutral supply reheats the dehumidified air back up to roughly room temperature so it enters the space without changing the thermal load. The fresh air shows up dry but does no cooling, and the parallel sensible system carries the entire room temperature load. Cold supply skips the reheat and delivers the air close to the temperature it leaves the cooling coil, so the DOAS air handles its own ventilation load plus a share of the space sensible.

Cold supply is usually the more efficient call, and the reason is direct. When you dehumidify air down to a low dewpoint and then reheat it to neutral, you throw away all the sensible cooling you just paid for. Cold supply keeps that cooling and puts it to work in the space, which shrinks the parallel sensible system and cuts energy. It is hard to justify a neutral-supply DOAS on energy alone once you account for the reheat penalty.

Neutral still has its place. It simplifies control where the parallel system must own all the zoning, and it avoids dumping cold air into spaces that cannot diffuse it well. The trade is comfort and zoning flexibility against the reheat energy. Make the call on the building, then commission to whichever strategy the design actually chose, because a cold-supply design balanced as if it were neutral will overcool or overheat.

Dewpoint control and reheat

A DOAS is controlled to a supply dewpoint, because dewpoint, not relative humidity, is what determines whether the air can absorb moisture in the space. The unit cools the outdoor air below its dewpoint at the coil, condensing water out, until the air leaves at a low enough dewpoint to carry the building's latent load. A common design target lands the leaving air near a 50 to 55 degree dewpoint, but the number belongs to the design engineer and the project, set from the space humidity target and the latent load, so verify it against the spec rather than carrying a habit number.

Reheat enters because the air coming off a coil cold enough to hit that dewpoint can be colder than you want to deliver, especially on a neutral-supply design or during low sensible load when cold air would overcool the space. The trap is paying twice: cooling the air down with the compressor, then heating it back up with a separate energy source. New gas or electric heat for reheat is the wasteful version and the energy code restricts it.

The efficient version recovers the heat instead of buying it. Hot-gas reheat reclaims heat from the unit's own refrigerant circuit, and an energy wheel or a passive wrap-around heat pipe moves heat from the warm entering air to the cold leaving air for free. ASHRAE 90.1 reheat limitations also cap how warm DOAS supply air can be delivered when most zones are calling for cooling, commonly cited around a 60 degree limit even when recovered heat is used, so confirm the adopted edition. The rule of thumb on the job: if you are adding new heat to reheat air you just paid to cool, the design left recovery on the table.

Field example: a DOAS load split

Take a space that needs 2,000 cfm of ventilation outdoor air, with a design space target near 75 degrees and 50 percent relative humidity, which puts the space dewpoint around 55 degrees. To pull moisture out of the room, the DOAS has to deliver air drier than the space, so the design lands the supply dewpoint near 50 degrees off the cooling coil.

On a cold-supply scheme, that air leaves the coil around 52 to 53 degrees and gets ducted to the space without reheat. The DOAS now carries all of the ventilation latent load plus a measurable bite of the space sensible, because 53 degree air entering a 75 degree room is cooling as it mixes. The parallel sensible system, say fan coils or chilled beams, is sized only for the room sensible that the DOAS air does not cover. On a neutral scheme, that same air would be reheated to roughly 72 to 75 degrees, the sensible help disappears, and the fan coils carry the whole temperature load while reheat burns energy.

The numbers come from the psychrometrics of the actual design conditions, so treat these as the shape of the calculation, not values to copy. The lesson is that the supply dewpoint sets the latent capacity and the supply temperature sets how much sensible the DOAS donates, and the two are separate dials.

Controls: constant volume, demand ventilation, and CO2

The simplest DOAS runs constant outdoor air volume whenever the building is occupied: a fixed cfm of fresh air, dried to the supply dewpoint, every hour the schedule says people are in. It is easy to balance, easy to verify, and the right choice where occupancy is steady and predictable. The cost is that you condition full design ventilation even when the building is half empty.

Demand-controlled ventilation cuts that waste by modulating the outdoor air to the people actually present, usually inferred from CO2. As the space fills and CO2 rises, the DOAS brings in more fresh air. As it empties, the airflow backs off and so does the conditioning energy. On a DOAS the outdoor air rate is the whole job, so DCV acts directly on it, but it has to be set so the airflow never drops below the per-area ventilation floor the code still requires with the space empty.

Where CO2 sensors run the show, they have to be accurate and placed right. ASHRAE 62.1 sets accuracy and placement requirements for DCV sensors, commonly cited as accuracy within about 75 ppm and at least one sensor per ventilation zone, and it bars CO2-based DCV in spaces with CO2 sources other than people. The Anvilfield economizer and demand-control ventilation guide covers the sensor calibration, the differential setpoint, and the functional test in depth. Confirm the requirements against the adopted edition of 62.1.

How much outdoor air does a DOAS deliver?

A DOAS delivers the ventilation outdoor air rate the building is required to provide, which comes from ASHRAE Standard 62.1 for most commercial and institutional work, or the IMC where that is the adopted code. The rate is not a single number. The 62.1 ventilation-rate procedure adds a per-person component and a per-area component, so a crowded space like a classroom carries a high rate and a sparse one like a warehouse carries a low one, and the total is summed across the spaces the unit serves.

The reason the DOAS handles this so cleanly is that it delivers the ventilation air directly, at a known volume, instead of hoping a mixed-air unit pulls in enough outdoor air through a damper that drifts. You can measure the outdoor airflow at the unit and prove it against the design, which is exactly what 62.1 and most commissioning scopes ask you to document.

On the job, the outdoor air rate is the first thing to verify and the easiest to get wrong. A DOAS balanced low starves the building of fresh air and fails the ventilation requirement quietly. Balanced high, it wastes energy conditioning air nobody needs and can pressurize the building. Measure it, record it against the design rate per space, and confirm the standard and edition the project was designed to.

Why the energy code keeps pushing DOAS

The reason DOAS shows up on more drawings every cycle is the energy code, not fashion. ASHRAE 90.1 and the IECC have tightened the rules around ventilation energy, reheat, and energy recovery, and a decoupled system answers all three at once. It conditions only the air that has to be conditioned, it recovers the exhaust energy on that air, and it avoids the simultaneous-heating-and-cooling waste that a single mixed-air unit falls into at part load.

High-performance and low-energy building targets sharpen the case. Once a project is chasing a stretch code, a green-building standard, or a net-zero energy goal, the part-load humidity and reheat losses of a conventional system become the loads you can no longer afford. Decoupling lets the designer size each system tight, run the sensible system at efficient water temperatures, and recover energy on the ventilation, which is where the modeled savings come from.

What this means for the field is that a DOAS is usually carrying a compliance path, not just comfort. If the unit is not delivering its design outdoor air, not recovering at its rated effectiveness, or burning new heat for reheat, the building may be out of step with the energy model it was permitted on. Commission it like the energy case depends on it, because it does.

DOAS in high-density and high-load buildings

The buildings driving the most DOAS work right now are the high-density and high-internal-load ones, and the comfort spaces attached to data centers and AI compute facilities are a clear case. The white-space servers reject huge sensible heat that is handled by close-coupled cooling, not comfort air, but the people spaces around them, the offices, network operations rooms, and support areas, still need ventilation and humidity control. A DOAS gives those spaces dry, conditioned fresh air without dragging the whole comfort system into the sensible fight the IT load is already winning.

The general pattern holds anywhere the sensible load is large and concentrated while the ventilation is comparatively small: separate the two so each system is built for its real job. A high-density office or a lab with heavy equipment gets a tight, dry ventilation supply from the DOAS and a sensible system sized for the equipment heat, instead of one oversized unit lurching between extremes.

One caution for these spaces. CO2-based demand-control ventilation assumes people are the CO2 source, and rooms with combustion, gas suppression, or other non-occupant CO2 sources break that assumption, which is exactly why 62.1 restricts DCV there. On a technical space, verify the ventilation control logic against what is actually in the room before trusting a CO2 reading to set the airflow.

Commissioning a DOAS

Commissioning a DOAS is proving the unit delivers the design outdoor air, dry enough, with the recovery working, the reheat behaving, and the parallel system balanced against it. The most common failure is not broken hardware. It is a unit that was installed, started, and never measured, so nobody knows whether it holds the dewpoint on a mild wet day, which is the exact day it matters.

Start with the outdoor airflow, measured at the unit against the design rate per space. Then verify the supply dewpoint under load, not just the supply temperature, because temperature alone tells you nothing about whether the air is dry. Confirm the energy recovery is performing by reading temperatures and, where you can, humidity across the wheel or core and comparing the effectiveness to the rating, the same way the ERV commissioning guide lays out. Drive the reheat and check it is recovering heat, not buying it, and that it does not run when there is no call for it.

Then balance the parallel sensible system against the DOAS contribution, because the two interact. If the DOAS is on a cold-supply strategy donating sensible cooling, a parallel system balanced as if the DOAS were neutral will overcool. Test the building across conditions if you can, including a part-load case, because the part-load humidity is the whole reason the DOAS exists and the only time you will catch it slipping.

Maintaining it once it is in service

A DOAS hands the owner a short list of maintenance items that, skipped, quietly erase its performance. The energy wheel is first. A wheel that is dirty, slipping its belt, or stuck slows the recovery and the unit's coil suddenly has to do far more work, which shows up as higher energy and, in winter, a wheel that frosts and a supply that drifts. The wheel seals wear and let exhaust leak into the supply, so they get inspected, not assumed.

Filters are next, on both the outdoor air and the exhaust. The DOAS pulls 100 percent outdoor air, so it loads filters with everything outside, and a loaded filter cuts the airflow below the ventilation rate the building is required to deliver. The owner who lets filters go is starving the building of fresh air without knowing it.

The condensate drain is the one that bites. A DOAS condenses a lot of water, because removing the latent load is its job, so it produces more condensate than a comparable mixed-air unit. A plugged drain backs water into the unit, onto the floor, or into the ductwork, and a P-trap that dries out or was never primed lets the unit pull air through the drain. Check the drain, the trap, and the pan as routine, not after the ceiling stains.

What to document

A DOAS that nobody documented is a system the next technician has to reverse-engineer at the worst possible time, on a humidity complaint in August. The record is what proves the unit was set to the design and lets the next person tell a drifting setpoint from a real fault.

Capture it per unit. Record the design and measured outdoor airflow, the supply dewpoint setpoint and the measured value under load, the energy recovery type and its measured effectiveness against the rating, the supply strategy and whether reheat is recovered or new heat, and what the parallel sensible system is and how it was balanced against the DOAS. Note the ventilation standard and edition the design was based on, and who verified each number.

Common mistakes

• Setting the DOAS on supply temperature instead of dewpoint, so the air is the right temperature but not dry enough and the space stays humid.
• Mismatching the parallel sensible system to the DOAS contribution, so a cold-supply unit overcools or a neutral design leaves the room warm.
• Skipping energy recovery on 100 percent outdoor air, which drives the conditioning energy up and can miss the ASHRAE 90.1 requirement.
• Never verifying the outdoor air rate, so the building is over or under ventilated and the ventilation requirement fails quietly.
• Buying new gas or electric heat for reheat when recovered heat was available, wasting energy and tripping the reheat limits.
• Ignoring the condensate drain and trap on a unit that produces more water than a mixed-air system, so it backs up and floods.
• Running chilled beams or radiant while letting the DOAS dewpoint slip, so the cold surfaces sweat and stain the ceiling.

Field checklist

Standards and references

The ventilation rate a DOAS delivers comes from ASHRAE Standard 62.1 in most commercial and institutional work, through its ventilation-rate procedure of a per-person plus per-area calculation, or from the adopted mechanical code where that governs. The demand-control ventilation sensor accuracy and placement rules also live in 62.1, along with the restriction against CO2-based DCV where there are CO2 sources other than people.

The energy side comes from ASHRAE Standard 90.1 and the IECC. Energy recovery on outdoor air is required by a 90.1 table keyed to climate zone and percentage of outdoor air, commonly cited at Section 6.5.6.1, with a minimum enthalpy recovery effectiveness often around 50 percent. ASHRAE 90.1 also limits reheat and caps how warm DOAS supply air may be delivered when most zones call for cooling. Stricter standards, including ASHRAE 189.1, ask for more recovery. The thresholds shift between editions, so confirm the adopted edition and any local amendments before citing a number on a submittal.

Thermal comfort targets, the temperature and humidity the design holds, trace to ASHRAE Standard 55. Energy recovery devices are rated and certified under AHRI Standard 1060, which is what you check measured effectiveness against. Above all of these, the equipment manufacturer's data and the project's design engineer and specifications control the actual setpoints and capacities. Cite the standard that governs the point, and let the project documents and the manufacturer's listing override any rule of thumb.

Units, terms, and conversions

DOAS work crosses a few unit systems, because the same air state shows up on a psychrometric chart, a controls graphic, and a manufacturer's selection differently.

Airflow is in cubic feet per minute (cfm) on most drawings, or liters per second or cubic meters per hour in metric sources. Dewpoint and dry-bulb temperature are in degrees Fahrenheit here, degrees Celsius elsewhere. Humidity shows up as relative humidity in percent, as dewpoint, or as a moisture content in grains of water per pound of dry air, which is the unit latent capacity is often figured in. Energy recovery is given as effectiveness, a percentage of the available enthalpy difference recovered, and the latent load itself is energy, in Btu per hour or kilowatts.

Dewpoint

The temperature at which air becomes saturated and water condenses, the true measure of how dry the air is

Enthalpy recovery effectiveness

The share of the total energy difference between exhaust and outdoor air that the recovery device transfers

Total energy wheel

A rotating enthalpy wheel that transfers both heat and moisture between the exhaust and the entering outdoor air

Reheat

Adding heat to dehumidified air to raise its supply temperature, ideally from recovered heat rather than new energy

Sensible heat ratio

The fraction of a cooling load that is temperature versus moisture, which sets how the DOAS and parallel system split the work

Grains per pound

Moisture content of air by weight, used to size the latent load the DOAS has to remove

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is an energy recovery ventilator (ERV)?
• What is an air-side economizer?
• What is the ventilation rate procedure?
• How much exhaust does a restroom need?
• What is a chilled beam?