Gas Turbine Atomizing Air Filter Differential Pressure


Thread Starter


The differential pressure across the atomizing air filter is constantly in alarm. The device summary states that it should alarm at 3.5 psid. The differential pressure gage reads 5 psid during operation before and after a filter replacement. Is 5 psid normal for a good filter, and is the device summary incorrect? I would like to know what other users have as an alarm set point.
Were the proper filters (same micron/flow-rating) used to replace the operating filters? There are a LOT of vendors saying their filters meet the OEM spec, but ask them what that spec is and they can't tell you. They just say, "It's a good as the OEM's--but it's less expensive." And the people buying the filters buy the line from the salesperson (or website), and buy the filters.

Unfortunately, the OEM doesn't usually provide spec's on the filters they provide with units, or even those that they sell as replacements. I've even seen filters sold by the OEM that didn't meet the spec (the OEM just publishes a spec on the internet and asks potential providers to give them a price quote for filters that meet the spec--there's no testing or proof required, other than the quote which implies the filters being quoted meet the spec).

Are you certain there isn't another root cause resulting in the 5 psi differential? Something amiss with valving, or instrument tubing? If the flow-rate through the filters is excessive, or the manifold for the differential sensing devices (gauge; transducer; differential pressure switch) isn't valved-in properly, or one of the sensing lines has been plugged with dirt and/or rust (along with moisture), then maybe the filters aren't the problem.

A lot of valves have to be moved during an off-line water wash, and sometimes they don't returned to the proper positions (usually because there is no site-specific procedure for checking off all the valves (either before or after a wash). Water can get into the filters and cause them to swell (depending on the type of filter material).

It's still possible the Device Summary setpoint is incorrect, but it's not typical. Hopefully someone can provide some information. Unfortunately, I don't have any F-class Device Summaries at this writing.

Hope this helps!