Latest from IAQ & Ventilation

S&P USA Ventilation Systems
HPAC Engineering
David Schurk
Ventilation Rate Sensors for IAQ-primary
Johnson Controls
270° Six-Way Valve and Actuator

The Impact of System Type on the ADPI and Air-Outlet Devices

April 1, 2005
In 1967, the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) introduced the Air Diffusion Performance Index (ADPI),

In 1967, the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) introduced the Air Diffusion Performance Index (ADPI), a single-number means of relating temperatures and velocities in an occupied zone to occupants' thermal comfort and optimizing ceiling-diffuser and air-outlet selection. The ADPI is calculated by mapping a cube from the floor of a space to an elevation of 6 ft and a distance from perimeter walls of at least 1 ft. Temperatures and velocities are recorded on the four elevations as prescribed in ANSI/ASHRAE Standard 113-1990, Method of Testing for Room Air Diffusion, at evenly spaced intervals in the module. The effective draft temperature then is calculated for each location. The ADPI value is the percentage of the points that meet the condition of -3 to +2 F, with a room velocity of 70 fpm or less.

Although much has been written about it over the years, the ADPI has been easier to discuss than to apply. With the help of computers, however, application is getting easier. In practice, ADPI calculations typically are not used in the selection of diffusers, as zone-by-zone analyses of air patterns are difficult in terms of design time. Difficulties in applying the ADPI, therefore, result in both interior and perimeter zones being designed in a general manner. Problems may arise in zones with widely shifting loads or in situations in which loads are not as planned.

Increased focus on indoor-air quality in the design of air-distribution systems, as well as the introduction of extremely high-mixing diffusers with tested air-diffusion performance in the 90-plus-percent range, is renewing interest in the ADPI as part of diffuser selection. Not often discussed is the relationship between diffuser selection and the ADPI, given the type of system (variable temperature or variable volume) employed.

This article will provide an overview of ADPI selection and recommended ranges for various kinds of air-outlet devices and examine the codependency of system type and air-outlet-device type.


According to 2001 ASHRAE Handbook-Fundamentals,1 the ADPI is intended for sedentary (i.e., office) settings and utilized when an HVAC system is in the cooling mode. It should be applied only for spaces with an 8- to 10-ft-high ceiling.

Different diffusers can incorporate distinctly different airflow patterns, or isovels. ADPI values vary by diffuser type, as well as the amount of cooling a space requires, an often-overlooked fact that has great implications for the type of system chosen to serve a load.

Referring again to 2001 ASHRAE Handbook-Fundamentals,1 air quantity and pattern and diffuser size determine jet throw, T. Jet throw is the distance from the center line of a diffuser to the point at which the maximum velocity of the air stream is reduced to a selected velocity in the stream cross-section, called the terminal velocity (Vt). For ceiling diffusers and sidewall grilles, the reference terminal velocity is 50 fpm; therefore, manufacturers catalog isothermal throw data to 50 fpm (T50 isothermal). This applies to isothermal conditions without boundary walls affecting air jets. The characteristic room, or jet, length (L) is the distance from the center line of a ceiling diffuser to the boundary or wall in the direction of airflow.

When air streams from neighboring diffusers collide, the characteristic length is half the distance from one diffuser to the other, plus the distance from the ceiling to the occupied zone. Typically, the occupied zone is up to 6 ft above the floor.

Figures 1, 2, and 3 depict ADPI ranges and peaks for circular ceiling diffusers with circular airflow patterns and sidewall grilles. The data are from Table 4 on Page 32.14 of 2001 ASHRAE Handbook-Fundamentals.1

In Figure 1, note that, at a space-cooling requirement of 40 Btuh per square foot, the peak ADPI rating is 88, with an isothermal-throw-to-characteristic-length ratio (T50 isothermal/L) of 0.8. For an ADPI rating of 80, the ratio of isothermal throw to characteristic length ranges from 0.5 to 1.5 at a cooling load of 40 Btuh per square foot. This cooling load reflects an approximate delta-T of 20 F.

Figure 1 also shows typical turndown ratios for single-duct variable-air-volume (VAV) systems with or without reheat. For these systems, the highest ADPI for the ratio of isothermal throw to characteristic length is to the right of the curve (Point A). As the load decreases, the ADPI shifts to the left until the lowest acceptable value is found (Point B). Note that as the VAV unit throttles down, the curve shifts to a lower load range — in this case, from 40 to 20 Btuh per square foot. Diffusers for a VAV system, therefore, should be selected for the ratio of isothermal throw to characteristic length at maximum flow. Given the load curves of round ceiling diffusers, an acceptable ADPI can be maintained through a wide range of turndown.

For optimum room comfort with a fan-powered, constant-volume, variable-temperature system, the best isothermal-throw-to-characteristic-length ratio is 0.8 (Figure 2). This represents peak performance at about 80 percent of the length of the room, from the center of the diffuser to the wall or medial line. Note that as delta-T decreases, maximum ADPI increases (Point C to Point D).

Figures 1 and 2 show that a constant-volume, variable-temperature reheat system, such as a series fan box, has distinct comfort advantages when utilized with ceiling diffusers. As the temperature differential decreases, the ADPI for a given flow rate or ratio of isothermal throw to characteristic length increases.

The same sort of analysis is suitable for other types of diffusers. Figure 3 shows that sidewall grilles can be used in constant-temperature, VAV cooling systems. The data shows that for the highest comfort level to be obtained, the ratio of isothermal throw to characteristic length needs to be 2.0 at maximum airflow. As airflow decreases, the ratio drops to 1.3. For optimum constant-volume performance, the desired isothermal-throw-to-characteristic-length ratio is about 1.6.

Remember that ADPI curves are relevant for cooling and related to the isothermal data cataloged per ANSI/ASHRAE Standard 70-1991, Method of Testing for Rating the Performance of Air Outlets and Inlets.

Diffusers for constant-volume systems should be selected for the ratio of isothermal throw to characteristic length at maximum load. For these types of systems, given the load curves of round ceiling diffusers, the ADPI will increase as load decreases. This is why one can maintain acceptable comfort levels in a space over a wide range of loads when using a constant-volume system and a network of lesser-performing diffusers (i.e., perforated face diffusers). Diffuser-performance characteristics are more critical with variable-volume systems.


What does all of this tell us? We can see clearly that diffuser selection is more critical to comfort with VAV systems than it is with constant-volume systems. The lesson to be learned is that diffuser selection is not a variable independent of system selection. Diffusers need to be selected in concert with the type of system with which they are being applied.


  1. ASHRAE. (2001). 2001 ASHRAE handbook-fundamentals (ch. 32). Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers.

Leon Kloostra is chief engineer for Titus, for which he has worked for 48 years. During the early 1960s, he helped develop the single-number index now known as the Air Diffusion Performance Index (ADPI). He is a member of the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) and ASHRAE Technical Committee 5.3 Room Air Distribution. He can be contacted at [email protected].

For HPAC Engineering feature articles dating back to January 1992, visit