National Institutes of Health 
Dr. Fredrickson 
9/21/78 
Page 3 
problems. There are a number of facilities which utilize HEPA 
filters in the exhaust system because there is some question 
as to whether the exhaust air is dispersed free of all build- 
ings and intakes. We know of at least six in service or in 
construction, some of which have been approved by NIH. Our 
proposed revision makes it clear the HEPA filter approach is 
acceptable whereas the PRG-NIH as presently written does not 
permit this approach. As may be seen from our comments below, 
an alternate to "dispersed free of buildings and intakes" is 
essential in many cases. 
D. The phrase "dispersed clear of occupied buildings and 
air intakes" in paragraph II-B-3-c-(7) needs better definition. 
Placing exhaust stacks on a roof and exhausting upward at velocities 
above 2,500 feet/minute as stated on page 148 of the Laboratory 
Safety Monograph doesn't always work. We recently participated 
in a test of a facility contained in a U-shaped building with 
a dormitory building closing off most of theopening of the U. 
The building roof is approximately 70 feet from the ground. 
The exhaust stack is 30 feet high. The wind was blowing NNE 
at about 10-15 knots. The exhaust velocity from the stack is 
about 2,300 feet/minute. When a smoke bomb was ignited in the 
exhaust stack a visible fraction of the smoke (perhaps 1%) 
was drawn down into the courtyard between the laboratory and 
the dormitory. Smoke could be seen 10 minutes after the test 
at a level equal to second story dormitory windows . This 
facility has HEPA filters in the exhaust system so we do not 
feel that the incomplete dispersion represents a health hazard. 
We do feel it illustrates the problem in determining whether 
adequate dispersion is obtained. 
We propose that a section be added to the guidelines defining 
a procedure for determining adequacy of exhaust dispersion. A 
suggested definition of adequate dispersal is "Not more than 1 
part of exhaust air in 10^ parts of air shall impinge upon any 
occupied building or air intake." This should be determined by 
wind tunnel tests simulating a range of wind speeds at each of 
a number of wind directions covering the entire compass at 
intervals of 22%°. Wind tunnel tests are required because wind 
speeds and directions are not so predictable and cooperative as 
to permit tests on the actual facility. 
The need for wind tunnel tests could be waived under a 
particular set of circumstances which we believe will provide 
adequate dispersion. All are required. 
1. The exhaust stack shall have a height of at least 60% 
[A-220] 
