Improving the accuracy of a constant concentration tracer gas system.

Air infiltration flows into different zones of a building can be measured with the constant concentration technique by injecting a metered amount of tracer gas t o hold the concentration of the gas constant. The control and estimation algorithm used to calculate the injection rate is designed using classical transform and optimal estimation methods. The ability of the control algorithm to keep the concentration constant and to accurately measure time varying infiltration flows is demonstrated using digital computer simulations and laboratory experiments. Field demonstrations then complete the confirmation that all components of the total system are performing as designed, and that the desired accuracy targets have been achieved. The details of how constant concentration system accuracy targets were attained in the Princeton constant concentration tracer gas measurement system are outlined in this paper. Before the total system accuracy goals could be achieved it was necessary to focus attention on the commercial SF6 detection unit based upon the principles of gas chromatography and electron capture. Gas flow paths, sequencing and critical times in a given sample analysis all directly impinge on the total system function and ultimate accuracy. Some of the points discussed are: internal air leakage, valve switching, and calibration of various subsystems.