A return fan adds pressure to all the air coming back from the building, before it splits. An exhaust (relief) fan adds pressure only to the air that leaves. Move the sliders and follow the pressure, station by station, to see why that one difference decides which fan a system needs.
Same building in both drawings · 10,000 cfm supply · 600 cfm local exhaust + 400 cfm leakage · adjustable outdoor air and return resistance
Reading the drawings Air loses pressure through every duct, filter, coil and damper and gains it only across fans. Taps A–F read relative to outdoors = 0.00 in. w.g. + above outdoors − below outdoors Duct stripes move faster and thicker with more airflow; damper blades show how hard each damper is throttling; the plenum shading deepens with suction.
Selected operating point: 10% outdoor air · Airflow in cfm, pressures in in. w.g.
Illustrative steady-state pressures, not measured values or equipment ratings. The controls are assumed to hold +0.05 in. w.g. in the room and the selected airflows. Bars show static pressure rise, not energy use; see the worksheet for the air-power proxy and its caveats.
Each fan's static is the sum of the losses along the path it is responsible for, following the selection rules in the JCI/YORK application guide. The numbers follow the return-duct slider. This is the model's external static only: velocity pressure, system effect and internal unit losses are not included, and real selections must add them.
| Return-fan system | Exhaust-fan system | |
|---|---|---|
| Supply fan | ||
| Second fan | ||
| Air power at this operating point |
About the air-power row. Air horsepower = Σ (cfm × in. w.g.) ÷ 6356, using the static rises above. Divide by fan, drive and motor efficiency to estimate input power (a propeller relief fan may be under 40% efficient; a good centrifugal fan 60–70%), then weight by hours: the return fan runs every hour the supply fan runs, the relief fan only in economizer. The two robust comparisons are the endpoints: at minimum outdoor air the exhaust-fan system has one fan running, at 100% outdoor air the two systems do nearly the same work. In between, the split depends on how the mixing dampers are sequenced, so do not read a winner from the middle of the slider.
Whenever outdoor air comes in, the same amount has to leave. The three standard answers form a ladder, ordered by how much return-path resistance they can live with. The rules of thumb are from the JCI/YORK guide.
A weighted damper in the return plenum that swings open when the plenum rises above outdoor pressure. It only works if there is almost no return-duct loss between the space and the damper: a single-storey unit with an open return, exhausting a modest fraction of the supply air. The supply fan handles the whole system.
Return loss ≲ 0.10 in. w.g.Runs only when relief is needed, modulated on building static pressure: off at minimum outdoor air, full speed at 100%. The supply fan still pulls the return air through the return duct and return damper in every recirculating mode, so the return path has to stay modest. Cheapest to run; easy to control; favoured for VAV when the return path allows.
Return loss up to ≈ 0.8 in. w.g.Runs whenever the supply fan runs. It carries the return duct, holds the return plenum slightly positive so relief can leave through a damper in every mode, and gives the supply fan a mixing box that stays put. Used when the return path is long, and whenever return-side static has to be controlled regardless of duct length.
Return loss ≳ 1.0 in. w.g., or controlled return staticBecause the return path is long or restrictive. Trane's rule of thumb is to use a return fan when a ducted return adds more pressure drop than a reasonably sized supply fan can handle. What makes a bigger supply fan unreasonable is less the fan itself than what a deeply negative mixing box does to everything around it. The outdoor-air damper has to throttle a large pressure difference to pass a small minimum flow (see the damper section below), unit casings and access doors leak inward, rain and snow get pulled through the intake louver, and the whole recirculation path runs at higher differential. A return fan takes the return-duct loss onto itself and hands the supply fan a mixing box at a modest, steady pressure no matter what the economizer is doing.
Because relief has to work in every mode. With the return plenum held above outdoor pressure, relief air leaves through a modulating damper with no second fan loop to tune, and the building can be relieved even at minimum outdoor air if local exhaust shuts down.
Because something on the return side needs a controlled static. Return or exhaust VAV terminals, venturi valves, tracking pairs. This is the reason that has nothing to do with duct length; it gets its own section below.
The cost. It moves all the return air at all hours, and at minimum outdoor air most of its pressure rise is thrown away across the recirculation damper just to keep the mixing box negative. Trane lists exactly this as the return fan's part-load penalty, along with its remedy: reset the plenum-pressure setpoint so that one of the two dampers always stays nearly wide open. Two fans in series through a damper can also fight: if the return fan is allowed to push the mixing box above outdoor pressure, outdoor air stops entering and air blows out of the intake louver. Return-fan sequences limit fan speed or plenum pressure and interlock the return fan to the supply fan for exactly this reason, and building-pressure control is harder to tune than with a relief fan.
Because the supply fan can already handle the return path. Short ducts, an open ceiling plenum, a rooftop unit sitting on the space it serves. The relief fan is small, often a propeller fan, is off at minimum outdoor air, and modulates on building static with a loop that is easy to commission. The JCI/YORK guide favours it for VAV whenever the return path allows.
Select the supply fan for full recirculation, not for 100% OA. The supply fan must be selected for the greater of two cases: 100% outdoor air (intake path + unit + supply duct) and 100% return air (return duct + return damper + unit + supply duct). The second one almost always governs, because the exhaust fan only takes over the return duct once the return damper is closed. Drag the slider to minimum OA and watch the supply-fan rise.
Watch the relief side. A negative return plenum cannot relieve through a gravity damper, and a propeller fan cannot lift air out of a plenum sitting half an inch below outdoors: significant return loss calls for a centrifugal relief fan. With a variable-speed relief fan, check its minimum speed: at low supply airflow the fan's minimum flow can exceed the relief actually needed and pull the building negative, which a constant-speed fan with a modulating damper avoids. If the leaving path itself is restrictive (an exhaust-side energy-recovery wheel, a long exhaust duct, a silencer), a powered exhaust fan is needed whichever return arrangement you pick. And near 100% OA with a very negative return plenum, outdoor air can short-circuit backwards through the return damper straight to the relief fan, which is the "poor design" case drawn in the JCI/YORK guide.
Some systems need a return fan even with a short, low-loss return. Return or exhaust VAV terminals that track their supply terminals (laboratories, healthcare isolation suites, vivaria, pharmaceutical and cleanroom areas, or any zone with a controlled supply-minus-return offset), venturi valves, and any return-side device whose pressure drop is a large share of the available suction all depend on a known, steady inlet static. A pressure-independent return terminal needs a minimum negative static at its inlet to reach its maximum flow, and it needs that static not to wander as the economizer moves.
An exhaust-fan system has nothing controlling that variable. During recirculation the suction in the return duct is a by-product of the supply fan pulling through the return damper. In economizer mode it is a by-product of the relief fan, which is itself being modulated on building pressure. As return terminals close, the same suction is spread over less flow, the static at the remaining terminals shifts, the outdoor-air split drifts, building pressure moves, the relief fan responds, and the return static changes again. Every loop in that chain is controlling something else.
A return fan on a drive closes the loop. Its speed is modulated to hold a return-duct static setpoint at a representative sensor, the mirror image of supply-duct static control. The terminals see a stable inlet pressure and the return fan follows them as they modulate, just as the supply fan follows the supply terminals. Building pressure is then handled downstream of the return fan by the relief damper on the positive plenum, or by supply/return airflow tracking. Systems with dedicated general-exhaust fans apply the same principle on the exhaust side: the fan that serves the terminals is the one that holds their duct static.
Illustrative steady state: supply = 10,000 cfm. Separate local exhaust = 600 cfm; outward envelope leakage = 400 cfm. Both stay fixed here, so 9,000 cfm always comes back to the unit.
Yes. Different locations, different pressures.
The room setpoint is maintained by balancing outdoor air against relief, local exhaust and leakage. It is not the supply-fan discharge pressure, and raising fan pressure does not by itself pressurize the room.
This +0.05 example is not a universal building-pressure target. Weather, stack effect, use and envelope construction all matter, and the JCI/YORK guide suggests staying at or below about +0.10 so doors still close.
More suction in the mixing box means more throttling to hold the same outdoor airflow. A damper sized to pass full-economizer airflow may run nearly shut at minimum ventilation. With strong suction, small blade movements or blade leakage admit too much air; actuator resolution, linkage play and loop tuning all become consequential. Stable minimum-OA control depends on damper sizing, leakage class and the pressure range the damper has to work across, and that range is set by the return arrangement.
At a fixed damper position and roughly constant air density, airflow follows approximately Q ∝ √ΔP across the intake path. Quadrupling the pressure difference passes roughly twice the airflow through the same opening. A minimum position balanced at one mixing-box pressure is wrong at every other one, and in an exhaust-fan system the mixing-box pressure changes with return-duct loss, damper positions and, in VAV, with supply airflow.
That is a comparison between operating points, not an uncontrolled-flow prediction from this animation. Here the controls are assumed to adjust the dampers and hold the requested OA flow; the blade angles drawn are an illustration of how far each damper has to close to do it.
Damper authority is the ratio of the fully open damper pressure drop at design flow to the total pressure drop of its subsystem, the duct section between two relatively constant-pressure points. For the OA damper that is the outdoor-to-mixing-plenum path, not the whole supply system. Plenum pressure alone does not establish authority; damper sizing, its characteristic (parallel or opposed blade) and the other intake losses matter. A large pressure drop across a nearly closed damper is not its design authority.
That definition follows the ASHRAE Handbook, Fundamentals of Control and Belimo's damper application guide.
Practical responses: select the damper and actuator for the actual pressure range and the minimum-to-maximum airflow; consider a separate minimum-OA damper alongside the economizer damper; measure and control OA flow directly where it matters. A return fan makes minimum-OA control easier in a restrictive-return system because the mixing-box pressure stops depending on the return duct, but it does not guarantee stable ventilation by itself. Trane's Engineers Newsletter 42-2 draws the same pressure profile for a single-zone VAV unit and shows why the OA damper must open as mixing-box suction falls, and how airflow measurement closes the loop.
"Exhaust fan" here means the air handler's economizer relief fan (Trane and ASHRAE Guideline 36 call it a relief fan), not a separate toilet or process exhaust fan. The return fan runs whenever the supply fan runs. Whether the relief fan can shut off at minimum outdoor air depends on the building's other outflows; in this example local exhaust and leakage exactly cover minimum OA, so it can.
Airflow. Supply is held at 10,000 cfm in every mode. Local exhaust (600) and outward leakage (400) are fixed, so 9,000 cfm returns to the unit. Recirculation = supply − outdoor air. Central relief = outdoor air − local exhaust − leakage. Real leakage varies with building pressure and the controls have to accommodate that. Constant supply isolates the fan-arrangement question; a VAV system also changes supply and return quantities, which changes every pressure here.
What the sliders mean. The outdoor-air slider sets an airflow fraction, not a damper position or a fan speed. The return-duct slider changes duct resistance at constant flow. Damper blades in the drawings show an illustrative percent-open computed from Q ∝ A·√ΔP, normalised at each damper's wide-open reference point (intake path 0.30 in. w.g. at 10,000 cfm; recirculation path 0.25 at 9,000; relief damper 0.05 at 9,000). Duct stripe speed and thickness scale with airflow; fan rotation speed scales with each fan's airflow.
Pressure budget (in. w.g. static, relative to still outdoor air; 1 in. w.g. ≈ 249 Pa). Room A = +0.05. B = A − return-duct loss, identical for both arrangements because the return flow is identical. Filter/coil loss D → E = 0.60; supply path F → A = 1.00, so F = +1.05.