The Right Mix

The invention of the air-side economizer was a great boon to the HVAC industry as it made it possible to circulate large volumes of outdoor air when the conditions are favorable. The technology saved huge amounts of mechanical cooling energy, and money, in temperate climates and is a feature now required in most energy codes.

By Eugene DeJoannis, P.E., Technical Associate, VanZelm Heywood & Shadford, West Hartford, Conn. October 1, 2001

The invention of the air-side economizer was a great boon to the HVAC industry as it made it possible to circulate large volumes of outdoor air when the conditions are favorable. The technology saved huge amounts of mechanical cooling energy, and money, in temperate climates and is a feature now required in most energy codes. Its adoption, however, has not been without problems.

In cases where an air-handling unit (AHU) can only ingest the minimum required ventilation air, an arrangement can be implemented to mix that air fairly well with the return air, without the danger of freezing a water coil.

However, if the engineer wants to add a 100% outdoor-air-capable economizer, a new component is required: the so-called “mixing box.” The word mixing is somewhat misleading because the device does not actually mix the two air streams. Perhaps “connection box” would be more descriptive, because the equipment is simply a large section of air-handler housing that contains return- and outdoor-air dampers, duct connections and often the filter racks. Typically, the larger the AHU’s capacity, the poorer the mixing. The air streams must travel 6 to 8 feet toward each other to mix effectively, but the mixing boxes only allow a horizontal distance of 4 feet or less before the air streams reach a freeze-vulnerable water coil.

With a conventional mixing box, good mixing depends only on the kinetic energy of the two air streams, which are introduced at right angles to each other. Often AHU manufacturers try to enhance mixing by using parallel blade dampers for return and outdoor air (usually one large damper for outdoor air). Typically, the damper blades are arranged so that as they open, the air is directed toward the other damper. But parallel blade dampers are not linear throttling devices, so one must give up the more linear operation of opposed blade dampers to gain somewhat better mixing. As a result, use of parallel blade dampers does not assure good mixing.

If thorough mixing did occur, there would be no freeze danger at all because outdoor air is only required to mix the air stream to about 55in the early days of economizers until someone invented the freeze-stat—the extended temperature sensor that covers the surface of the cooling coil. The device shuts off fans and will close the outside air damper if freezing air reaches any part of its sensor tube. This innovation saved a lotof coils, but it left the building operator with no air supply. If the operator manually reduces the outdoor air volume to avoid freeze trips, the supply air temperature will not be low enough to cool a building’s interior. And if the operator does nothing, the AHU keeps shutting down and the spaceoverheats.

Economizer’s trip

In theory, when outdoor temperatures become cooler than a building’s return air, the economizer damper should fully open so that all the return air is indeed in the 55

No damage can occur until the cold air stream reaches 32

For example, with variable-air-volume (VAV) air handlers, the fan is at its minimum speed in very cold weather, so the streams have little kinetic energy. For large air handlers, the air streams have to travel the width or height of the mixing box in a relatively short horizontal distance to mix well. Moreover, the situation is dynamic; as mixing dampers modulate the geometry and air jet angles are constantly changing, invariably, some damper positions will cause that pesky freeze-stat to trip. To make matters worse, the fan speed variations on VAV fans overlay a whole new set of variables with different velocities for the same damper positions.

Before going further, it is important to dispel the myth that in sub-freezing outdoor conditions in “normal” buildings (occupant densities around 100 sq. ft. per person), the AHU will operate at its minimum outdoor airflow, and coil freezing should not be much of a problem. Indeed, if the 10% to 20% minimum outdoor air is not mixed well with the return stream, only a small amount has to reach the coil to do its damage or shut down the fan.

Moreover, if one considers the mixed air equation: OA% = (RAT-MAT)÷ (RAT-OAT) —where T is temperature and MA , RA and OA refer to mixed, return and outdoor air values—one can construct a table to chart the required percentages (See Table on page 69). If the needed supply temperature is 55

Introducing the Injecti-mixer

This begs the question, is there any reliable method of making the mixing box mix in the confines of the existing space? Many types of baffles have been tried, but results are unpredictable and problems recur as outdoor conditions vary. In struggling with this problem, engineers at VanZelm Heywood & Shadford developed a device that reliably mixes and, in many cases, can be added to existing mixing boxes. Instead of using the velocity of the two air streams to achieve mixing, this device, called the Injecti-mixer , depends on perforated mixing tubes that tunnel the cold outdoor air across the whole return-air stream.

To understand this concept, it is helpful to think in terms of a two-stage economizer. The first stage is similar to an oversized minimum outdoor air damper as used on field erected and custom AHUs. A plenum, installed around this damper, is connected to the Injecti-mixer tubes used to duct the cold air into the warm return stream (see Figure 1, p. 70). This damper is used as the minimum damper and as the first part of the economizer, providing up to 40% to 50% outdoor air. The second-stage economizer has no mixing tubes (to reduce pressure drop) and is only opened when outdoor conditions are safely above freezing. For a 55

Since air handlers differ greatly in internal dimensions, the engineering team wanted a simple geometry to make the Injecti-mixer design as quick as possible. Even with a simple square tube geometry, a speadsheet design tool was necessary to quickly size the components for multiple AHUs (see “Spreadsheet Sizing Tool Facilitates Injecti-mixer Design”). To insure the cold air is distributed evenly across the face of the return air damper, four or five tubes will usually be necessary. If the space between tubes for the return air is kept about equal to the tube’s width, the tubes will generally be 8 to 12 inches on a side.

However, when one examines the pressure losses from a large plenum into a tube of this size, it is evident that the tube entry loss is much higher than the loss through the 50% open perforated skin of the tubes. Somehow the tube had to be enlarged without blocking any more of the return air path. The answer was to stretch the tube in the direction of the return-air flow, from a square cross section into a hexagon, by adding two longer panels between the front and rear V-shaped sides. This enlargement is limited because there must be enough room left in the mixing box to remove and replace the filters, unless a side access filter rack is used. The tubes are made with a flange at one end to screw to the first stage outdoor air plenum. An access door for damper inspection is always provided in the plenum, with an inside light, and tubes can be removed if needed for access.

Packaged air handlers may limit the outdoor-air damper choice to a single large damper. For the two-stage economizer, it would be ideal to utilize two equal-sized full-width outdoor dampers, but if that’s not possible, there is a simple solution. With a single outdoor damper, the first stage plenum can be built to enclose the entire factory furnished damper. The flat bottom of the plenum is wide enough to mount the entry-ends of the Injecti-mixer tubes on it, in front of the return damper. The vertical side of the plenum slopes up to the forward edge of the AHU outdoor damper. A new second-stage outdoor damper is installed across the width of the sloped part of the plenum.

More design details

The sizing process involves applying the formulas explained in “Spreadsheet Sizing Tool Facilitates Injecti-mixer Design,”( p. 71), so that the pressure drop through the tube at full flow (40% to 50% OA) is in an acceptable 0.1 to 0.2 inches of water range (entry loss + loss through perforations). Tubes can sometimes be made longer than the return air damper height if the pressure loss through the perforations is too high.

When the tube geometry is determined, a control sequence is required to properly stage the dampers. Because modern electric damper actuators usually have an adjustable span and zero point, it is rather easy for the controls contractor to deliver the required staging of operators, even though the typical digital AHU controller only generates a single 4- to 2-milliamp or zero to 10-volt DC signal for the economizer damper. Using the zero to 10-volt output as an example, the first stage OA damper is adjusted to have its zero (closed) at zero volts DC and a span of 4 volts. The second stage damper is adjusted to start its stroke at 4 volts DC and have a 6-volt span, so that at 10 volts DC, it is fully open. The RA and OA dampers will be adjusted to the same zero and span settings as each other, but in order to pressurize the building, the start is delayed from the operation of the first stage OA damper. The dampers will start to stroke at whatever voltage brings the first stage to the required minimum position to make up exhaust, pressurize the building and provide ventilation for occupants. This setting is typically 1.5 to 2 volts, outdoor air 15% to 20% open.

Although the engineers attempted to minimize the pressure drop across the Injecti-mixer tubes in the sizing process, because some pressure drop is inevitable, this can be used as an advantage. A differential pressure sensor can be installed across the tubes to make a surprisingly linear airflow meter, which is useable for direct OA measurements as long as the second stage is still closed. But when the second stage opens, the process is well beyond the minimum OA situation and can be assured of providing adequate ventilation without measuring it.

In this way, proper outdoor airflow is ensured, even on VAV fans which turn down to as low as 30% flow in winter. In spaces with highly variable occupancy, this measurement can be used to set an absolute minimum outdoor air flow—to make up exhaust and pressurize the space—and the indoor-outdoor CO 2 difference can be used to increase the OA setpoint to track the actual occupancy.

Optimizing outside air

It’s easy to forget, but in buildings with a large interior space that always needs cooling, the AHU will not be operating at 15% OA, but at some higher fraction, even when it is 0

The Injecti-mixer provides an economical solution that can easily be installed.

Mixed Air Matrix

Outside Air % = (Return Air Temp. – Mixed Air Temp.)/(Return Air Temp. – Outside Air Temp.)

Desired Mixed Air Temp. = 55°

Expected Return Air Temp. = 73°

Outside Air (Deg. °F)
Required Outside Air %










45.0% — Freeze Safe







Spreadsheet Sizing Tool Facilitates Injecti-Mixer Design

A spreadsheet design tool is extremely helpful to repeatedly perform the geometric calculations that determine the pressure drop through the tubes. The number and size of the tubes should be adjusted within the following guidelines.

Limit the return air velocity at the minimum outdoor airflow to 2,500 feet per minute (fpm), and preferably less than 2,000 fpm. High velocity is useful for turbulent mixing, but can impose a high return side DP.

The delta-P through the holes in the tube walls depends on the outside air velocity through the holes that are not capped, and the free area ratio of the plate, as follows:

%%TRANGL%%P holes = C x Vel Press. = 4 x (Velocity/4,005)

This is for a 50% free area, where the hole area = half of surface area. Loss coefficient is halved for each 10% increase in free area ratio—e.g. C=2 for 60% open. (See SMACNA Table 14-17, HVAC Systems Duct Design, 1990)

The %%TRANGL%%P of the entry into the tube depends on the ratio of the tube entry area to plenum area and the entry velocity as follows:

%%TRANGL%%P entry = 0.5 x K x Velocity Pressure, for a sharp edged tube entry

Mixing boxes operate at about 0.10 inches of water column negative. If possible, adjust tube dimensions to keep the total %%TRANGL%%P (entry + holes) & 0.2 in. water column negative to avoid excessive fan energy use. (SMACNA 14-12B)

A spread sheet example is shown below. Note that even with the trailing edge of the tubes capped, the tube entry loss is usually the larger of the two DPs.

Economizer Protection Tips

If a designer is expecting a large percentage of outside air in freezing weather, there are a number of options available to prevent freezing of steam or water coils. For example:

Vertical tube (tube-in-tube) steam coils with dual traps and 18-in. drip legs. Be sure to properly size the coil and control valves (two valves in parallel for flows more than 400 to 500 lb./hr.), and use a vacuum breaker after the control valve.

Add a low-head pump across water coils and beyond the control valve to circulate water at 3 ft. per second in the tubes. Central pumps can also be made to do this if valves are opened when freeze danger is imminent. With poor mixing, freezing might occur at an average mixed air temperature of 30

Drain vulnerable cooling coils and add some glycol solution to be sure no raw water remains in the bottom rows.

Use a glycol mixture of 25% or more in the first coil in the system to see the outdoor air stream. Lower concentrations are prone to bacterial growth.

Use integral, external or internal face-and-bypass type preheat coils which maintain full design water flow in freezing conditions and bypass air flow around the coil for temperature control.

Select a static mixing device when ordering the air-handling unit (AHU). It increases the unit’s length by an amount about equal to its height, to achieve mixing, but it avoids the frustrating problems of freeze trips and overheated spaces in cold weather.

Use a direct expansion of refrigerant into the cooling coil which will not freeze, but does not provide a modulating discharge temperature as chilled water coils do.

Use fossil fuel-fired heating instead of water or steam coils.

When considering these options, remember that for buildings with large interior spaces which have little if any winter heat losses, the AHU is almost always cooling, and the preheat coil will be cold! Do not count on heat from the AHU heating coil to protect itself or the cooling coil.