Mixing in a square and a rectangular duct regarding selection of locations for extractive sampling of gaseous contaminants

Tests were conducted to characterize the uniformity of velocity and tracer gas profiles in a square and a rectangular duct with respect to defining the suitability of locations for single point sampling of gaseous contaminants. Several configurations, such as a straight duct with unidirectional flow at the entrance section and straight ducts preceded by mixing elements (a 90 degrees mitered bend and double 90 degrees bends in S- and U-type configurations) were tested. Results are compared with those from circular ducts. For a straight duct of square cross section, which is not preceded by a mixing element, the coefficients of variation (COV) of tracer gas concentration at 19 duct diameters downstream of the gas release location is 143% (center release of tracer gas). COVs of velocity and tracer gas concentration in a straight square duct 9.5 duct diameters downstream of a 90 degrees mitered bend are 6% and 24.3% (top inside release), respectively, which does not meet the ANSI N13.1 limit of 20% for the tracer gas COV. In case of the rectangular duct with a 3:1 (width to height) aspect ratio, COVs of velocity and tracer gas concentration at 9 duct diameters downstream of a 90 degrees mitered bend are 29% and 62% (bottom inside release), respectively. A mixing element in a square duct comprised of two 90 degrees mitered bends in a U-configuration produces results similar to those obtained with a single 90 degrees bend. However, COVs of velocity and tracer gas concentration in a square duct 6 duct diameters downstream of an S-type double bend are 10.6% and 8.3% (top inside release), respectively, which comply with the ANSI tracer gas and velocity criteria for single point representative sampling. When mixing elements were employed in square ducts, the COV results were comparable with those of other researchers for circular ducts.