Wednesday, February 13, 2013

History of Air Motion and GSA “Peach Book” Specification

In the early 70’s, The GSA (Government Services Administration) developed a set of performance based specifications for buildings, referred to as the “Peach Book”. One of these specifications was that the measured air speed throughout the space shall lie between 20 and 50 fpm at all points. Originally, they specified an air velocity meter that was incredibly inaccurate below 100 fpm. We located an anemometer that was accurate down to 5 fpm but was directionally sensitive. Nonetheless, carefully oriented (using smoke to assure the predominant direction), we developed a repeatable test procedure that eventually became the basis for ASHRAE Standard 113. Using this procedure, we discovered that the specification was not able to be met with any air distribution system. Eventually, a modified specification was agreed upon that allowed 20% of the measured points to exceed 50fpm, and 40% to be below 20, as long as the average was between 20 and 50fpm. 

The measurement devices have greatly improved, and if anything, they are more sensitive than what we used in the GSA tests in '78. The ASHRAE Fundamentals Handbook ADPI data was taken with heated sphere anemometers at Kansas State in the 60's. All those had been calibrated so that they were accurate at low airspeeds, using careful measurements. In the GSA tests, we had a number of distributed 100 watt loads and the designs called for either 0.6 or 0.9 cfm/sf, and resulted in meeting the new GSA specification. The testing was conducted by Dr. Paul Miller (the ADC Engineering Consultant at the time) and myself, with GSA personnel witnessing all the tests.

The diffuser was a continuous ½”, 2 slot linear diffuser that resulted in a two dimensional air pattern throughout the space. The highest air speeds were recorded under the diffuser, in an upward direction, as expected. In attempting to meet the original specification, we also tested an array of ceiling diffusers and found that at some location, air speeds were always less than 20, and at others (under the diffusers) exceeded 50 fpm.

In the course of running over 1000 air distribution tests (reported in two ASHRAE Technical papers), we found that average room airspeed in the comfort zone is essentially proportional to the room load with any ceiling diffuser type, as long as the primary jet doesn't enter the occupied zone. Therein lays the problem. Some air outlets at low flows are unable to maintain sufficient "Coanda" to prevent the negative buoyancy of the air jet from falling into the space. This is often referred to as "dumping" and we find that perforated face diffusers are the most likely to exhibit this behavior. A number of other types of air outlets are seldom prone to this behavior.

At design flow on the other hand, if the diffusers are too close together, throws collide and drop into the occupied zone. By using ADPI analysis, it is possible for the designer to identify in advance, using published throw data, how different diffusers respond to changes in airflow rate, inlet size, and diffuser separation. If the predicted ADPI is greater than 80% at a given set of conditions, the primary jet(s) are not entering the occupied zone. We have also seen that when the ADPI is greater than 80%, average room air speeds are never greater than 40 fpm at flow rates less than 1.2 cfm/sf. This analysis, of course, is for interior zones in cooling mode.

Perimeter zones are another matter, and while average air speeds in heating are typically 20 fpm or lower, there are always high air speeds at the floor near cold windows. These will always exceed 30 fpm somewhere. (Unadjusted linear slots at the window, which is almost always the case, are likely to deliver jets exceeding 100 fpm into the occupied space, in either heating or cooling mode).

I would expect that anyone taking data today, using modern omnidirectional anemometers and meeting the requirements of either ASHRAE 113 (or the ISO 7726) specification, would get similar results.

Authored by: Dan Int-Hout, Chief Engineer Krueger