THE EFFECT OF CONCENTRATION GRADIENTS ON DEDUCING A CONTAMINANT GENERATION RATE-FUNCTION

If a room containing a contaminant emission source is well-mixed, and if a series of concentration measurements are made with a real-time sa mpling device, it is possible to deduce the contaminant generation rat e function G(t), mg/sec. However, if the space is not well-mixed, and if the real-time sampler inlet is located within a concentration gradi ent near the emission source, the deduced G(t) function can differ mar kedly in shape and magnitude from the true G(t) function. This article describes the theoretical difficulty in deducing G(t) from concentrat ion measurements by examining hemispherical diffusion and two generati on rate functions-a constant generation rate and an exponentially decr easing generation rate. A scenario is posed in which the room volume i s 25 m(3), the supply/exhaust air rate is .042 m(3)/sec (6 nominal air changes per hour [ACH]), the eddy diffusivity coefficient is .0033 m( 2)/sec, and measurements are made at 1.0 m from the emission source. F or a constant generation rate G, the deduced G(t) function increases f rom zero to a maximum that is 3.8-fold G, then gradually decreases to a steady-state value that is 2-fold G. For the exponentially decreasin g emission rate function with an initial rate G(0), the deduced G(t) f unction increases from zero to a maximum that is 3.5-fold G(0), and th en gradually declines to zero. It is shown that both the eddy diffusiv ity coefficient and the measurement distance from the source affect th e shape and magnitude of the deduced G(t) function. Therefore, to vali date a G(t) function deduced from serial contaminant concentrations in the workplace, one must either establish that the workplace is well-m ixed, or demonstrate that a scaled version of the emission process in a well-mixed test chamber produces a similar G(t) function.