Why is it necessary to purge the filter in the HCl/Cl2 train and why do a leak check in two sections? Doesn’t this miss any possible leaks between the two halves (three-way valve) once reassembled? Can a single whole train check be done instead?

As you recall, SW-846, Method 0050 was designed to sample HCl/Cl2 from hazardous waste incinerators and municipal waste combustors, especially suited for those sources with wet scrubbers emitting acid particulate matter (e.g., HCl dissolved in water droplets).  As such, Method 0050 requires isokinetic sampling to insure a representative sample is extracted from the stack.  The water droplets containing the dissolved HCl  would be extracted in a representative manner from the passing stack gas around the nozzle.  However, after the droplets enter the gas sampling train, they may fall out in the optional cyclone or be retained on the heated filter.  To address this bias, the method calls for purging the sample train for 30 minutes (purge air through an Ascarite tube) to vaporize the liquid and purge any HCl in the cyclone or retained on the filter and pull it through the train and into the first three impingers.  Recent test by EPA have demonstrated that if visible moisture is still in the cyclone or on the filter, then increasing the probe/filter assembly to 177 C with additional purging of 15 minutes will insure complete removal of the HCl from the cyclone/filter to the impinger system.

The two tier leak check procedure identified in Method 0051 (remember, Method 0050 requires a standard FRM 5 full train leak check) requires a leak check of the probe and three-way stopcock, then a leak check from the first impinger through the rest of the sample train. The two tier leak check would allow the stack tester to leak check the probe and three-way valve only one time.  The probe/valve assembly would stay in the stack until the end of the testing day.  This would allow the tester to run several test, exchanging out the impingers several times and leak checking the impinger/meter box only without removing the probe/three-way valve from the port.  This saves some time when multiple runs/tests are being performed at the source.

Indeed, a single whole train leak check can be done each time, which is stricter than what Method 0051 requires.

For more information associated with FRM 26, please contact Mr. Terry Harrison, U.S. Environmental Protection Agency, MD-19, Research Triangle Park, North Carolina 27711 (919-541-5233).