The well-established Bureau of Mines chilled mirror method to determine the temperature at which hydrocarbons condense out of a gas stream is both labor intensive and subject to operator variability. In recent years, equipment has been introduced to automate this process. However, the equipment has not been rigorously tested independently under controlled conditions to compare its performance vs. the chilled mirror method. Controlled laboratory testing confirmed the repeatability, uncertainty, recovery time (between samples), and impact of sample filtration on the dew point calculations of streams both free of contaminants (Phase 1) and containing a methanol/water mixture and water vapor (Phase 1).
In gas streams free of water and methanol, the analyzers performed in good agreement with the Bureau of Mines (BOM) chilled mirror method, particularly in the range of dew points the units were designed to detect. Contaminated streams were more challenging to characterize, both for the chilled mirror method and the analyzers, however the main deviation between the BOM manual and the automated approaches was with methanol/water mixtures. Additional work has been recommended to better characterize this performance, but the results for the clean gas streams clearly showed the analyzers to be functionally equivalent to the BOM method.