GT Hydrogen Purity vs Gas Turbine Generator Output

S

Thread Starter

sriram

Dear all,

I have idea about how does the hydrogen purity affects gas turbine generator output. If i am not wrong it is this way by reading the threads from control.com and by Googling.

Hydrogen has molecular weight of 2, while Air is 29. it does mean Air weighs more than 14 times compared to Hydrogen. so even relatively low levels of air considerably increase the density of the gas mixture. Air is the most likely impurity to affect hydrogen within the generator casing. For each percentage point increase of Air impurity in Hydrogen will have a 14 times exponential increase in Gas density within the Rotor space in the Generator. Thus, the increased drag caused is also exponential in value. The best performance of a Turbine Generator (TG) can be derived when the Hydrogen purity within the TG is maintained near constant at 98 to 99.5%.

Even small reductions in hydrogen purity directly correlate to increased windage friction losses inside the turbine generators, which have a direct impact on a power plant's bottom line.

Impurities in a hydrogen supply increase the density of the gas, and the density of a gas affects its ability to remove heat.

This can be understand <pre>
Gdens = (Hpur x 1) + (Abal x 14.4)
Where: Gdens = Increase in gas density(%)
Hpur = Purity of hydrogen in generator casing (%)
Abal = Balance of impurity (air) in generator casing (%)</pre>
Roughly, for every 1 percent of air trapped in a hydrogen-cooled generator casing, there's an average 250 kW drop in power output.

Now the thing is that there was some experimentation going in our plant...how does hydrogen purity affects gas turbine generator output for every 1% drop in purity.....

Management decided to test the unit in combined cycle at 99%, check whats the performance. and subsequently for every 1% drop, check whats the unit performance again...like this ways from 99 to 95%.. we as a performance engineers can understand how difficult it is justify this 1% drop against unit output....

so they are asking us to make report for every 1% drop whats the unit performance at base load condition.

what is the wise way to convince them as many other factors should be in consideration while calculating the unit performance...simply ambient temp, humidity, pressure, power factor, inlet dp, exhaust dp, etc. the above said parameters should be constant to calculate the difference in unit performance for every 1% drop in hydrogen purity?

Please help some one.

CSA sir...how to respond to this type of challenges. give some wise answer.

Thank you in advance
 
sriram,

Testing can be done in all kinds of ways. Certain factors can be presumed to remain constant during the test, and other factors can be corrected for (ambient temperature, for example). It all depends on the length of the test and the number of variables to be "accounted" for during the test. If the ambient temperatures don't change very much during the testing period, then the use of correction curves (formulae) may not be necessary. If the turbine air inlet filters retain relatively the same differential pressure, and if the ambient humidity doesn't change very much, and if the axial compressor remains relatively clean (including the IGVs), and so on, and so on. Yes, there can be lots of variables for the turbine--but not so many for the generator (read on).

But, for me the question is: What is to be gained from this testing?

Is there a feeling that by reducing hydrogen purity the power output will increase? Because you've just pretty conclusively proven that increases windage losses will contribute to decreased power (watts; KW; MW)--and over time the losses can accrue to a significant amount. (I point to the fractional (less than one-half of one percent) increases being sold for F-class machines. The machines are so large (power output) and over time (10-, 20- or more years) the gains add up, making them "desirable."

Is there a feeling that operating costs can be reduced by reducing the amount of hydrogen being used to replace that lost during scavenging? Well, that's not going to change--because presuming the gassing rate of the seal oil remains the same (which it would for all intents and purposes), all other things being held constant (generator casing pressure; seal oil flow-rate; vacuum on the oil tank). So, decreasing hydrogen purity isn't going to appreciably decrease the cost of hydrogen used to maintain purity (whether it's at 99.7%, or 96.3%, or 95.1%). Since the gassing rate of the seal oil will remain constant regardless of the generator casing hydrogen purity the amount of hydrogen required to maintain casing purity--at any level--with remain the same once a stable purity is reached.

The thing you missed in your analysis (which was very good, by the way) was that hydrogen is seven to ten times better at transmitting heat than air. Transmitting heat means absorbing it from the generator internals, and "giving it up" to the water-to-hydrogen heat exchangers in the generator casing (so moving heat from a place where it's not desirable to a place where it's not damaging (cooling water)). So, hydrogen is much better for removing heat--which is the limiting factor for power production of any generator. If the heat developed as a result of current flowing in the windings (rotor and stator) can't be removed and increases the insulation of the windings will deteriorate faster. A generator can produce just about an unlimited amount of power--just keep increasing the torque from the prime mover and the electrical power will keep increasing, even past the generator nameplate rating (much past the generator nameplate rating!). Either the coupling between the generator and the prime mover will fail, or eventually the heat developed in the generator windings will cause the winding insulation to fail which will also lead to a catastrophic failure of the generator.

Finally, air is probably going to have humidity--moisture--in it. (Unless they are going to inject dry, instrument air into the machine to reduce the hydrogen purity.) Humidity (moisture) in hydrogen-cooled generator casings is not a good recipe for long generator internal life. If the machine has a hydrogen dryer (other than the little absorption thingies in the hydrogen control cabinet which only serve to "protect" the hydrogen purity sensors--the discharge of which is usually sent to atmosphere, not back to the generator casing), then it would have to be put in service more often--requiring CO2 for purging and hydrogen for make-up. If the machine doesn't have a hydrogen dryer, then one should probably be installed and used frequently if the purity is maintained at, say, approx. 95%. There's a cost for piping and installation and wiring and controls--and it's not cheap. Operating one requires a good, understandable, site-specific procedure, and training--both of which also come at a cost.

Without really understanding what the intent of the testing would be, it's really hard to provide any rebuttal or justification other than yours added to what's above.

Generators are really dumb devices--they convert torque to amperes, and in the process generate heat (because of the current flowing in the windings (rotor and stator)), and the heat has to be removed for the generator to last any appreciable length of time. It's the ability to cool the generator that really determines its power output rating--the ability to maintain a particular power output indefinitely without excessive damage, that is.

Hydrogen is used as one medium for cooling because it is so light (low windage losses), has a high heat transfer ability (much higher than just about any other medium), and allows for a smaller generator to be utilized for the same power output (saving real estate, foundation and construction costs as well as reducing the cost of the generator, though it does increase the cost of the auxiliaries (seal rings; end bells; hydrogen scavenging system; seal oil system; hydrogen and CO2 piping; etc.).

Generators don't really care what's being used to cool them--air, or hydrogen, and sometimes water circulated through passages in the stator/windings--it just converts torque to amperes and it's the generator "operator's" responsibility to make sure the generator temperatures remain within limits so the generator insulation doesn't get damaged and require a shutdown to repair (or worse). Generators don't even really care if the heat isn't removed--they will just keep producing amperes until the winding insulation deteriorates and allows internal shorts and grounds (and in some cases the air gap gets reduced to the point of rubs/interference of the rotor and the stator).

I've often wondered why small hydrogen-cooled generators were installed at many paper mills, and a colleague just explained it to me. The air (environment) at many paper mills is corrosive (they have to "seal" a lot of the electronics to prevent corrosion and damage) and to use an air-cooled generator would be to invite a LOT of generator maintenance (which paper mills aren't known for spending money on--maintenance, that is). An air-cooled machine of the same output wouldn't be that much larger, but it would entail much more frequency--and costly--maintenance.

Look--if your management want to do the test, and if they want to interpret the data with their own slant, as an employee there's little you can do. Complaining or trying to tell the emperor he's hot wearing any clothes (a popular European fairy-tale--with a fairy-tale moral/ending) will not bode well for you during your next performance appraisal. You can't even say, "I told you so!" if they decide to run the machine with low hydrogen purity and there are "premature" problems--because they will not interpret the "prematurity" as having been caused by their decision (as a result of their risk assessment/management).

That is, unless they pay a lot of money for a consultant who has the fortitude to write a proper report--which doesn't get discredited by Management....

Best of luck! Let us know how you fare in your endeavour to encourage sane, logical thinking and risk assessment.
 
S
what is to be gained from this testing??
CSA sir, the main intention behind doing this test is come to a breakeven point to maintain hydrogen purity which will give more power output...

till date our usual practice is to maintain hydrogen purity above 95% (when it falls below 95 we are bleeding and feeding something like venting and filling hydrogen( we are using bottled hydrogen from 3rd party and purity quality is upto the standards)...

Management telling if we maintain 99% hydrogen purity output will be much more instead of what practice we are doing now keeping at 95%?? so all this testing is for this only...because hydrogen bottle cost also almost 500 usd..
 
suryapower,

Welcome to the discussion; it's a friendly one.

I'm presuming the Hydrogen Control Panel still uses manually-adjusted needle valves for the scavenging function--and that Seal Oil is supplied from the Lube Oil reservoir. Finally, it's presumed the hydrogen purity sensors are valved to the hydrogen side seal areas (where the purity should be lowest); to measure generator casing purity at least one of the purity sensors has to have it's sensing valves switched to draw off the generator casing.

Air gets into the generator casing from Seal Oil gassing--that is, the release of air which is entrained with the oil, which comes from being in contact with air in the Lube Oil- and Hydraulic system return lines and the Lube Oil reservoir. Seal Oil flows from the seals into the hydrogen-side of the seal and releases the entrained air (gasses) into the hydrogen side of the seal. So, presuming the purging/charging was done well and resulted in a high hydrogen purity, and because the generator casing is at a higher pressure (usually around 2 barg) than the atmosphere the only way for air to get into the generator casing and contaminate the hydrogen (reduce the hydrogen purity) is from the air released from the Seal Oil. SO, the seal oil flow-rate, and the amount of air entrained in the Seal Oil will affect hydrogen purity.

Scavenging is done through purity sensors and the manually-adjusted scavenging flow-rate needle valves, and the scavenging flow meters, from the area where the hydrogen purity should be lowest: the hydrogen side seal drains. And because a small amount of "gas" (which is hopefully mostly air) is continually vented through the scavenging system which would reduce the generator casing pressure, pure hydrogen has to be introduced to the casing to both maintain casing pressure and to help with maintaining purity (because circulating hydrogen will inevitably carry some air with it into the casing and around the areas where only hydrogen is supposed to be present).

So, the only way for air to get into the generator--under normal operating conditions--is from Seal Oil. Therefore, the Seal Oil flow-rate, and the amount of air entrained in the Seal Oil flow, determines the amount of scavenging required--which in this context refers to the amount of hydrogen required to maintain casing pressure, because, again, hopefully the majority of the air is being vented (scavenged) to atmosphere. (The amount of entrained air in Seal Oil can be considered to be relatively constant under normal operating conditions.)

And, Seal Oil flow-rate--for all intents and purposes--doesn't change much unless the Seal Oil Differential Pressure is increased, or the seals wear or the springs loosen or the seals weren't installed properly after generator inspection. Seal Oil flow-rate does change with time and wear--but presuming the wear is nominal, the seal springs don't loosen or crack, and the Seal Oil Differential Pressure is maintained fairly constant, the Seal Oil flow-rate will remain relatively constant. And this is regardless of the desired purity--be it 99.5% or 95.0%--the purity will only change with changes in Seal Oil flow-rate.

Regardless of the purity the Seal Oil flow-rate doesn't change very much under normal operating conditions. That's very important to understand--especially in the context of this thread. The intent of scavenging is to set the scavenging flow-rate (total--both ends of the generator) so that the purity remains relatively constant--meaning the amount of gas flowing through the scavenging system to atmosphere (through the manually-adjusted scavenging needle valves and purity sensors) matches the amount of gassing from the Seal Oil (the amount of air being released from the Seal Oil flowing in the hydrogen side of the seals).

And once a desired purity is achieved for the current Seal Oil flow-rate/gassing the manually-adjusted scavenging needle valves do not need further adjustment. Unless Seal Oil flow-rate changes--which is why it's <b>SO</b> important to periodically record and analyze Seal Oil flow-rate, so that if purity changes it can be corresponded to a change in Seal Oil flow-rate. (Many GE-design hydrogen-cooled generators only have one (1) Seal Oil flow meter, so if the flow changes at one end more than the other it can't be determined which end is experiencing the higher flow. Significant changes in Seal Oil flow-rate should be logged to the Operations Log, and the Maintenance Department should be notified. AND, <b>any</b> changes to the manual scavenging needle valve position(s) should also be logged.)

And, again, pure hydrogen is added to the generator to maintain casing pressure as hydrogen--and air--are continually released to atmosphere through the scavenging system.

If it's necessary to manually "bleed and feed" hydrogen to maintain 95% purity then there's probably a LOT of hydrogen being wasted--more than would normally be lost through the scavenging system if the scavenging needle valves were adjusted to maintain a particular hydrogen purity, be it 99% or 96% or 95%.

Having said all of the above--manual adjustment of the scavenging needle valves takes a LOT of patience, and sometimes many days to achieve a relatively constant purity (at any percentage).

Now, it's already been proven--the least windage (torque) loss occurs at the highest purity. So, if the intent is produce as much power as possible--then the purity needs to be as high as possible. It's impossible to have 100% purity because Seal Oil is always flowing--and, again, the hydrogen purity sensors are usually set to monitor the purity at the lowest possible places in the generator (the hydrogen side seal drains).

I contacted a colleague in the USA (where hydrogen prices are probably among the highest in the world) and the retail price of a standard bottle of hydrogen (approximately 300 cubic feet, or approx. 8.5 cubic meters) is almost USD200.00. (Of course, power plants get a quantity discount, and I was told that can reduce the cost to approx. USD100.00 per bottle. And we're only talking about the cost of the hydrogen--not including the steel cylinder (bottle).) So, if I could sell a USD100.00 bottle of hydrogen for USD500.00, I would get into the hydrogen import/export/selling business--because that's a 400% profit!

Finally, I would venture to say that a three or four or even five percent difference in hydrogen purity can't be measured on most turbine-generator outputs. It's pretty insignificant--and this is perhaps the most important fact. (Of course, over a span of 20 or 30 years, it might add up to a significant amount of money due to windage losses--but that can only really be estimated.) But, I wouldn't say the generator output would be "much more" at 99% than at 95%. But, if the Seal Oil flow-rate is the same for 99% purity and 95% purity, the hydrogen "consumption" would be the same. Manual "bleeding and feeding" must require more hydrogen than if the scavenging needle valves were properly adjusted to maintain a particular purity--regardless of that purity.

And that's a wrap. Hope this helps!!!

(WHEW! That's a lot of typing to say what was said in my previous response to this post.)
 
suryapower/sriram,

By the way--both of my responses refer to GE-design hydrogen-cooled generators provided with heavy duty gas turbines and combined cycle steam turbines. (Actually, the principles apply to just about any hydrogen-cooled generator which uses oil as the sealing medium on the generator shaft.)
 
Top