After completing a HGP and Magic inspection on our GE 7FA/7FH2 we are dealing with low CE H2 purity. The CE H2 seal was removed for the Magic and inspected and deemed OK by the OEM to be returned to service - faces were cleaned, etc, clearances checked and within tolerance.

After introducing H2 back to the generator the CE H2 purity did not return to "normal" - typically we see ~97% on both CE and TE. Our new "normal" is ~92% on the CE when on turning gear. We have run the unit once and we did see the purity on the CE drop to the mid to high 80's and also noticed the casing purity dropped to 90% while the TE never dropped below 96%.

So far we have:

1) Confirmed the analyzer - calibration and confirmation check.

2) Completed a "bucket" test as per GEK 107241 section G. Found the gas side seal oil flow to be ~2 Gpm. We have not heard back from GE if this is acceptable. Total flow while on turning gear is ~14 Gpm.

3) Confirmed our valve line up and also confirmed the CO2 bulk tanks have been properly isolated.

4) Confirmed seal oil DP measurement is within spec - 6 psid.

We don't expect any issues with H2 purity. We use a bulk storage tank which feeds H2 to both gas turbines and our steam turbine. All other units are reading "normally" with regards to H2 purity.

I am concerned we have a seal installation issue on the CE.

I have looked through a few posts on this topic but I haven't found what other users ultimate resolution was.

If anyone could share their similar experiences it would be greatly appreciated. Also if I haven't provided enough information please let me know and I can provide more data.

 

Scavenging rates do not have to be equal on both ends of the generator. In fact, normal differences in seal oil flow-rate to each end of the turbine will require that in order to be able to make the purities of both ends equal that scavenging flow-rates will need to be different.

It does sound as if the purging/charging wasn't done too carefully; the results you reported after the first run are pretty typical of an incomplete purge/charge cycle.

The whole purpose of having adjustable scavenging rates is to be able to compensate for differences in seal oil flow rates, and for seal oil wear over time. More seal oil flow means more liberated air; it's that simple--at least in the Seal Drain Enlargement tanks at each end of the generator. The drop in casing purity was probably caused either by unpurged air/CO2 which was mixed with the hydrogen when the rotor was rotated, or using the CE end analyzer to monitor casing purity, or some combination of the two.

Unfortunately, when purging/charging it's all too common for oil to make it's way into the generator casing and even into the hydrogen monitoring lines. This can cause all manner of havoc and erroneous reading. Also, getting moisture in the sensing lines when the generator is open is another problem which occurs in some environments when proper precautions aren't taken during outage work.

Lastly, there is a good possibility that there is indeed a seal ring installation problem after the outage that is contributing to or causing the problem. Only a visual inspection, likely with measurements required, will tell for sure.

But, a good starting point is always to try increasing scavenging on the enlargement tank with the lowest purity for at least 24-48 hours. Increase the scavenging rate, recording the change, and wait for a few hours under steady state conditions. Steady state conditions means don't change the scavenging rate while the unit is on gear, then start the unit and load it up to 70% load; make the scavenging rate change while the unit is on gear and expected to remain on gear for the duration of the test, or while on load and expected to remain on load for the duration of the test--in other words, when seal oil flow rate should remain relatively constant).

Increase the scavenging rate, recording the changes, until the purity increases. If it doesn't increase, then that likely means there's a problem with the analyzer or the piping/tubing, or some other condition.

Once the purity increases to an acceptable level, start incrementally reducing the scavenging flow-rate and monitor the purity. Hopefully, you can get back to an acceptable level of scavenging while maintaining an acceptable scavenging flow-rate (hydrogen consumption). If not, if the purity starts decreasing shortly after you start reducing the scavenging flow-rate then it's a pretty safe bet that there's something wrong with the seal ring/seal ring springs causing a higher than normal seal oil flow-rate.

The way to increase purity is always to increase scavenging flow-rate--temporarily, until the acceptable purity flow-rate is achieved--and then decrease the scavenging flow-rate to an acceptable level (hydrogen consumption) once the purity is acceptable. If the purity can't be maintained with a low, acceptable scavenging rate then something is wrong with the system, usually the seal rings/seal ring springs, etc.

Or, it might be hydrogen purity, but you say that's not a problem.

There's not many ways for air or other impurities to get into a generator casing filled with hydrogen at approximately 2 barg. Air doesn't "leak" into a generator casing at 2 barg pressure. Hydrogen leaks out. So, about the only way for purity to be low is for all the impurities not to have been removed during the purging/charging process, or for impure hydrogen to be introduced into the generator casing, or for "excessive" seal oil flow-rates to cause excessive air to be liberated into the seal drain enlargement tanks.

And--providing the hydrogen is pure--increasing scavenging flow-rate(s) should be all that's required to bring purity up. If scavenging flow-rate(s) can't be returned to acceptable levels without a decrease in purity, then something is amiss. Usually with seals.

Again, other potential problems include oil and/or moisture in the analyzer piping/tubing.

Please write back to let us know how you fare.

 

CSA - Thank you for the reply.

Based on our low casing purity numbers during our initial run up - we haven't run the unit since - we completed a "bleed and feed" on our generator and exchanged approximately 33% of the volume. We also used a portable H2 analyzer and took a sample from the bottom of the generator and measured 100% H2. In addition we have increased our CE scavenge rate by 2500 cc/min - total flow on low scavenge is now 5000 cc/min. The CE purity has risen to 95.5% and leveled out.

Our total seal oil flow is moving around from 7 Gpm to 18 Gpm while on turning gear which is concerning to me as this leads me to believe, as mentioned previously, we are dealing with an improper seal installation.

The OEM is recommending we "shock" the seals by increasing the seal oil DP to 20 - 30 psid, momentarily, to put the seal back in place. I am somewhat skeptical that this will work and I have some concerns with the risk associated which performing such a procedure. Have you had any experience, or for that fact heard of, with this procedure? In fact I think the seal is "floating" and I can't comment on if increasing the dp in this manner will move the seal or not but even if it does what is to say it will not move again?

My suspicion at this moment was the seal was not properly torqued and we are dealing with increased flow on the CE gas side seal leading to low purity as well as evidence of H2 in the bearing drain enlargement.

I would like to hear of your experience, or lack thereof, with the proposed "shock" procedure.

Thanks again.

 

Seeing this is a 7FA, it's pretty unlikely the site does not have a services agreement with GE of some kind. If so, you may be bound to try what GE recommends.

What do you have to lose? You're either going to try what they suggest, or you're going to have to purge the H2, then purge the CO2, then pull the CE bearing apart, perform any necessary repairs, reassemble the bearing, purge the air with CO2, then charge with hydrogen. That's a minimum of two days' work. With a good crew.

If GE feels like they have a solution, just be sure you get a detailed plan for implementing the suggestion--and a TA who's done it before or is confident and competent.

I've never heard of this, but GE has LOTS of experience with 7FH2 generators, and the MAGIC inspection is an example of some of the new processes and procedures.

Let us know how you fare.

 

We have resolved our low purity/high gas side seal oil flow issue.

Prior to going into the H2 seal we increased the DP, as suggested by the OEM, with no appreciable results. After this test we decided we had exhausted our options and we had to go into the H2 seal.

We pulled the seal for inspection - all clearance checks were within spec prior to removal. Upon close inspection there were some indications, burrs, etc. The seal was dressed and deemed fit for service. Upon seal installation - prior to the upper half bearing being installed - the generator was pressurized with air and the seal oil was returned to service using one of the main AC lube oil pumps. To accomplish this we blanked the lube oil line to the CE. We performed a bucket test with the unit off turning gear and the results from the test, as per the OEM, were satisfactory @ 0.5 GPM.

We have operated the unit since the "repair" and all parameters are normal.

I wish I had more to report but this is it - it appears that the minor indications may have caused the seal to hang-up resulting in higher than normal gas side seal oil flow and ultimately lower CE H2 purity.

 

GJRS,

Thanks very much for the feedback. There really is only one way for purity to be lowered and that is through an elevated seal oil flow-rate. With the generator casing at 2 barg, air cannot leak into the generator casing. And if the purity is only low at one end of the generator (in your case, the CE, Collector end--opposite the TE, Turbine End) it's most likely the result of excess seal oil flow causing more entrained air to be liberated in the seal oil drain enlargement tank.

This is why the two H2 purity monitors are located on the two shaft seal drain enlargement tanks. This is where the first indications of excess seal oil flow will manifest themselves as decreasing purity: at the shaft seal oil drain enlargement tanks. If left unattended increased seal oil flow-rate would eventually contaminate the casing purity as well (presuming the source of H2 being used to charge and maintain the casing is pure).

Thanks very much for writing back! "Feedback is the most important contribution!"(c) here at control.com. It's what makes threads valuable for weeks and months and years after the original problem has been solved.

Wouldn't it have been great to have had individual shaft seal oil flow-rate indications, one for the Collector End and one for the Turbine End? That way, you would not have had to do a "bucket test" (which I presume is putting a large bucket of some sort under the float trap after removing a bit of piping to collect the oil as it flows out of the enlargement tank). Because there is only one Seal Oil Flow-rate meter for both Turbine- and Collector-end seals it's impossible to know which end is flowing more oil or how much more oil either end is flowing than the other (without individual flow-rate meters). That's why it's important for operators to monitor and record seal oil flow-rates; it should be easy to spot a decrease in purity when the seal oil flow-rate increases. Trending things like this can provide lots of useful information, and even point to possible problems before they become significant issues (though not in this particular case).

It really doesn't take much for the hydrogen seals to be "disrupted". If I recall they're made of brass (a soft metal) and if the mechanics drop them or mishandle them during assembly they can be dinged or have small burrs. I believe there have been some improvements to the springs holding the halves together, but that was sometimes a problem, also. Probably the reason the OEM was suggesting over-pressurizing the seals momentarily to try to get them to re-seat and be aligned better under the springs.

Again, thanks for the feedback! Hope we were of a little help in your resolution process.

 

Hi CSA,

We have some number of 9FA's with recent AGP outages. I have been working in power generation industry for 5 years and H2 Purity has always been a mystery to me.

At one of the units for example, the purity for both CE and TE have dropped only 2% in 3 months, [Unit-A for further reference]

but at another unit the purity for both CE and TE drop (still keeps dropping) 3% in only 1 month, every month. [Unit-B for further reference]

We operate between 96% and 93% H2 purity. When it drops below 93% (alarm state) we manually 'purge' H2 into the generator, by dropping the generator casing pressure for about 5 PSI, by opening the vent valve, and purge H2 into the recently vented volume. This process takes about 2-3 days until we get the purity back up to 96%. We use pressurized bottle skids (each skid is 2200 PSI / 79 cubic meters) and we use nearly 3 of them during this process. Come to think about the consumption and the cost of H2. (PS. Do you call this process 'feed & bleed', if not what is it? I'm foreign to the lexicon).

You said it can't be the case that the air is entering through the seal as there is positive pressure on the generator side, but on some another platform someone mentioned this:

// As hydrogen leaks out of the seals, a flow path is created for water vapor to enter the generator. This can be explained by water vapor seeking equilibrium. In other words, the partial pressure of water vapor in the turbine oil is greater than in the generator. As water vapor seeks equilibrium, water vapor enters the generator, even though the hydrogen gas is at a higher pressure. This can be further understood by studying Fick's Law of thermodynamics that addresses diffusion of gas molecules. //

That makes sense to me. Also I'm wondering can it be because of the lube oil mist eliminator fan's vacuum, taking air into the generator through some flange or other connection?

For about seal oil flow, here is comparison between Unit A and B.

(first values are for unit A, second values are for unit B)

Seal oil flow (GPM) 13 / 11
Seal oil pressure at seal oil inlet skid (Kg/cm2) 5,75 / 5,80
Seal oil pressure at generator flange C.E (Kg/cm2) 3,80 / 3,60
Seal oil pressure at generator flange T.E (Kg/cm2) 3,75 / 3,90
Lube oil pump discharge pressure (Kg/cm2) 6,90 / 6,95
H2 Dew Point temperature (Fahrenheit) 96 / 92
Generator rotor fan differential pressure (inH2O) 11 / between 9-12*
Generator common cold gas temperature (Fahrenheit) 101 / 103
Generator casing pressure (PSI) 44 / 42
H2 scavenging header pressure (PSI) 134 / 136

I have seen no damaged seal rings in ANY outages so far. Clearances are according to GE practice, I'm not skeptical about this one either.

So what can it be? Thanks in advance for sharing your knowledge.

Best regards.

 

Jose,

I don't have any experience with the phenomenon mentioned on another forum/website. Water is made up of two molecules of hydrogen, and one of oxygen.

Are you saying that there is LOTS of water vapour in the area of the hydrogen seals? Or that water vapour is being released from the seal oil?

Have a look at the generator seal oil piping diagram to see if there's a possible way for the L.O. Mist Eliminator to draw air into a generator or the seal drain enlargement tanks?

Are you measuring the casing hydrogen purity, or the seal drain enlargement tanks?

I'm not sure what process you're using, or what it's called, but it certainly sounds like it's consuming a LOT of hydrogen for very little purity increase.

Have you tried varying scavenging flow rate(s)? If so, what were the results?

How long have the units been operating without problems?

When did the problems start--after the AGP outages?

 

CSA,

I'm saying that it is possible. It's not just a phenomenon to me. It's due to molecular energy and velocity of the H and O molecules, seal oil in between the air and the H2 could serve as a membrane "in addition to" vapour in the area of seals so they could mix up easily.

I don't think water vapor is being released from the seal oil. Water content of turbine oil sample is zero.

I mean ALL purity values including CE, TE and casing. They increase/decrease altogether. My readings are based on H2 analyzer readings, I haven't conducted another reading with a portable analyzer. But the C&I team keeps the analyzer calibration up to date, so I'm not skeptical about the readings.

For some weird reason automatic scavenging doesn't work well for us even though we adjust the flow rate. We feel like we need to get the purity back up to 96% quicker than auto scavenging provides us, so we do it manually.

Units have been in operation for more than 110 thousand fired hours. The problem occurs sometimes at one unit, sometimes at another. But always after outages.

Regards.