Within ANSI/AIHA Z9.5, see Sections: 5.3.5 Exhaust stack discharge; 22.214.171.124 Combined Exhaust stacks; 5.1.1 Differential pressure and airflow between rooms; 3.3.1 Face velocities. [ANSI…Z9.5, 2002]
The recommended discharge velocity is 3000 feet per minute (fpm) for stacks without internal condensation.
This velocity is based most likely on the expected average wind speed. Wind modeling and staged exhaust fan operation can offer reductions in the energy consumed to produce this velocity. See Chapter 6.
"Two or more exhaust systems may be combined into a single manifold and stack…"
This translates into an energy-efficient design solution that will take advantage of right-sizing and manifolding exhaust systems. See Chapters 3 and 6.
A pressure differential between the supply and exhaust airflows to and from the laboratory creates the driving force that isolates the laboratory. The recommendation is that "...specifying quantitative pressure differential is a poor basis for design. What really is desired is an offset air volume."
While a 50 to 100 fpm airflow velocity is desirable from an energy-use point of view, this value is quite difficult to obtain and maintain in the "real world." The velocity pressure differential to create 50 fpm is only 0.00015 inches W.G. Typically, a laboratory maintains a differential of 0.01 to 0.02 inches W.G. Further discussion is presented in Chapter 3. [Roberts et al., 1992]
"The average face velocity of the hood shall produce sufficient capture and containment of hazardous chemicals generated under as-used conditions.
"An adequate face velocity is necessary but is not the only criterion to achieve acceptable performance and shall not be used as the only performance indicator.
"The mechanism that controls the exhaust fan speed or damper position to regulate the hood exhaust volume shall be designed to ensure a minimum exhaust volume of 50 cfm/foot of hood width, for a 24-inch deep hood (or 25 cfm per ft2 of hood work surface for different depth hoods) except where a written hazard characterization indicates otherwise." See Chapter 3 and 6 for additional discussion.