Hydrogen Cooled Generator Purging

B

Thread Starter

bfishe01

I was just looking for some general values that others purge hydrogen from their generator. IE flow rate that CO2 is introduced to the generator? If the unit still on turning gear? What is the casing pressure during purge? during regular operation? Time required to purge generator and casing volume?

The reason I ask is we recently did a purge going into an outage and when the generator casing was opened hydrogen was found. The point of the generator that was open was on the bottom as well where I would assume the lowest concentration would be found, however we had >25% hydrogen. We suspect this is due high CO2 flowrate? Any insight?
 
bfishe01,

What was the CO2 pressure regulator set for?

Purging should normally be done when at zero speed, so as not to allow the generator rotor fans to mix the hydrogen and CO2.

After the unit was purged of CO2, was the CO2 then purged using air?

What instrument was used to measure for hydrogen in the generator casing when it was opened? Hand-held instrument, or the generator hydrogen purity monitor (stationary) through some valves? If the casing pressure was near zero, then the stationary hydrogen purity monitors (which, I believe, are normally only capable of measuring hydrogen in air) wouldn't work properly, because they rely on casing pressure to have flow through the sensor elements.

Was the instrument which measured the hydrogen in the bottom of the generator casing after purging set for hydrogen in air, or how was it set?

If a hand-held device was used, is the calibration certain?

Scavenging is entirely dependent on seal oil flow-rates, since air is liberated from seal oil (that's the only way air can get into a generator at 2 barg (30 psig)--from the seal oil which is at a higher pressure than casing pressure and contains entrained air, which is approx. 20% oxygen).
 
The pressure regulator setting was varied throughout the purge and experienced lots of issues with freezing due to differential across regulator. The most common setting for the regulator would have been 100PSI to try and accommodate CO2 flow to the generator and to reduce differential across regulator.

The purge was done on turning gear (5RPM) and this was also suspect to cause mixing at gas interface.

The unit was purged of CO2 using dry instrument air.

The online analyser was used to measure hydrogen content while hydrogen purging was occurring (H2 in CO2 was setting on analyser). The sample piping setup to this analyzer was changed for the purge with the sample venting to atmosphere vs being recirc’d to generator. This value was also confirmed by two handheld units that had its sampling tubing in online analyzer vent. The handhelds and online analyzer all have up to date calibrations.

The sample for the hydrogen analyzer comes off the top of the machine. (As valved in for purging)

The CO2 that was used for this purge was 4x-5x what manufacturer has specified for degas.

Generator casing pressure for purge was dropped to 25 kPa (~450kPa operating pressure) wondering if this pressure was increased it would help to assist in reducing short circuit of CO2 and pressure differential for better sampling?
 
bfishe01,

I don't know your experience or knowledge, and I'm providing a lot of this information because a lot of people read these posts and try to learn something from them. So, I'm not trying to insult you or your site personnel or practices, but I do frequently find there's too little time allotted and the people performing the procedure are not very familiar with the process and the why's.

Mixing is most likely the reason for the problem, though I would also suggest that part of the problem is trying to use low casing pressure when analyzing through the "on-line" system. Some generator manufacturers used to provide portable gas analyzers which were much better at low pressures, and could be hooked up to points closer to the generator casing. It can be hard to purge 1/4-inch tubing with low pressure, leading to mixed results. When using hand-held sensors, a LOT of people don't really know how to use them (they don't use them very often, and haven't read the fine manual, and don't take the measurements properly--and are usually too fast to measure and don't wait for a proper response, and don't take multiple readings when they get a suspicious reading).

I wonder also if the proper valving was used for purging. Gas density differentials are used to help with purging. CO2 is MUCH heavier than hydrogen, so is to be admitted to the bottom of the generator casing and "pushes" the hydrogen up and out of a vent at the top of the generator casing. Then when air is used to purge CO2, it's admitted to vent at the top of the generator casing and used to "push" the CO2 down out of the bottom of the generator casing.

The BIGGEST problem with purging and charging of hydrogen casings is impatience. For GE-design heavy duty gas turbines with hydrogen-cooled generators, when an automatic hydrogen purge is initiated, the unit is allowed to go to zero speed BEFORE purging begins (the vent valve is opened to release hydrogen, and the CO2 valves is opened to admit CO2). I've always found when manually purging hydrogen from a generator casing that it's best to manually open the hydrogen vent first and allow the casing pressure to decrease significantly, then start admitting CO2. And, this only after going to zero speed to prevent mixing caused by the generator rotor fans. It's my personal belief--and experience--that by allowing the hydrogen pressure to fall to near zero in the generator casing before admitting CO2 that it's easier and faster to get to a low H2 in CO2 measurement. Then switch the valves to admit air into the casing to purge CO2 out through the bottom.

When charging, again--patience is key. At least 8 hours, if not 12, should be allotted for the entire process. And the people doing the charging (and the purging) should know and understand why they are performing each step of the procedure (understanding gas densities and how they're used to assist with the process, as well as valve changes, and regulator settings). Just handing someone a generic procedure and expecting them to accomplish the task in 4 hours when they don't really understand why they're doing this or that is not going to have a good outcome.

Purging/charging times vary greatly between sites, and the amount of gases used can also vary greatly. This is most often caused by un-familiarization and too little time being allotted for the procedures by planners and managers who don't understand the process very well, either.

As for the freezing of pipes and regulators, I have actually found galvanized pipes to have occlusions caused by poor galvanization which cause "orifices" in the pipe which cause pressure drops/temperature drops which lead to freezing on the outside of the pipe and on the inside, as well. Most pipe-fighters don't look down the inside of a 3/4-inch Sch. 40 or 80 pipe to make sure it's clear before they cut it up and weld it in place. But, with the international supply of materials and components these days, the "quality control" (and I use that term VERY loosely) tends to be very poor.

Also, many regulators are not chosen for the desired flow-rates--which aren't usually published for many generator systems, so we rely on the packager to provide the proper regulator for the application. And, it may not be the proper regulator. Regulators which aren't used very frequently can also fail more often. And, as with all things, international supplier don't always have the best quality (some are VERY good; some are VERY bad; and there's a lot of in between, too).

Hope this helps! It's really hard to be specific when there are so many intangibles and variables.
 
Top