Steam Turbine -Loss of Performance 1

[smckennz]

Hi there
I have been asked to look at the possibility of repairing a steam turbine which has has severe damage to one row of blades.
It is a 1960's English Electric turbine driving a generator.
MCR is 3225kW(e). MER is 2085kW(e). Turbine speed is 6000rpm
Inlet steam is 415psig at 390 Deg C, 734 Deg F. Exhaust vacuum is 28.05"Hg. There is some passout, but I havent got details yet.
It is 9 stage with first stage being (Curtis, I think)impulse double blade row.
Something has partly wiped out one blade row; the fourth stage, I think.
My questions are:
Is there a method that I can use to estimate the performance/efficiency reduction caused by loss of this stage?
If the blades are severely damaged, am I better to pull out the whole row together with the diaphragm and diaphragm nozzles or just dress off the damaged blades?

The turbine instrumentation is fairly poor and the steam supply is less than the turbine rating, so an open valve performance test is unlikely to help much.

Thanks

Steve McKenzie

 

[poetix99]

For reasons mostly related to mechanical reliability, I recommend that you consult the OEM (or their successor), or a reputable consultant who could inspect the damage and evaulate the consequences of some choices that you have.

How did a single stage in the middle of the steam path sustain damage without additional damage upstream or downstream of that stage? (There are some possible physical explanations, I am suggesting however that you ensure that there is no other damage to your steam path.)

With respect to performance and mechanical reliability, your best choice is to replace the blades with new ones; OEM blades (or "knock-offs" as a second choice).

(I'll refer to the original stage numbering in what I have to say below.)

With respect to performance, you're better off "dressing off" the damaged blades. That is not making any judgements whatsoever about the mechanical reliability of reusing the damaged blades. There would of course be a reduction of the stage 4 efficiency, again impossible to quantify in an e/m chat room, and possible to only roughly estimate when actually looking at the damage. Restoration of the leading edge and the convex side are the most important areas of the rotating blades in an impulse turbine such as yours.

If the blades are reasonably well cleaned up, I think a stage efficiency reduction of 5% - 20% for an impulse turbine is a good place to start. Because your first stage (Curtis stage) and last stages are taking a good bit more than an equal portion of the energy, and because of the so-called "reheat effect", the overall impact on performance will be less than a straight average of the stage efficiencies, somewhere around 0.25% - 1.5%

[A useful text for steam turbine performance is "Evaluating and Improving Steam Turbine Performance" - K.C. Cotton, ISBN 0-9639955-1-0 (tel. 1.518.384.7885 or 1.518.371.6259). Although it is written specifically for much larger machines than your 3MW turbine, many of the general principles apply. There is consideration of the efficiency reduction due to deposits and erosion of a steam path which, with some imagination may be extrapolated to Foreign Object Damage.]

If the judgement is made that the rotating blades should be removed, you must decide whether to also remove the stationary nozzle diaphragm. Performance impact is less if you remove it, but it is necessary to consider the increased loading on stage 3. There will no longer be as much of a downstream restriction to flow at stg.3 (stage 5 nozzle area is greater than stage 4 nozzle area); the pressure drop and available energy on stage 3 will increase. It is possible that the stg.3 blades would be overloaded, or even possible (though unlikely) that the stg.3 diaphragm would be overloaded, and have to much downstream deflection.

The blade loads on the downstream stages (5-9) will also increase, but by much lesser amounts.

 

[smckennz]

Hi Poetix

Thanks for that. Have ordered Cottons book after reading a few good reviews.
Your thoughts mirror my own uninformed sentiments.
In passing, only the first stage is Curtis. The remainder are about 50% reaction as appears normal in smaller machines.
Assessing such a small efficiency loss by measurement is impracticable in a working cogen (process driven + passout) environment. Similarly, a consultant will hedge his opinion with "ifs and buts" even with a visual inspection. 1% of 3000kW is only 30kW so I will leave it alone.
Nobody will tell me what happened to the fourth stage, so I guess a small bolt dropped out of someones pocket during the last overhaul.

Cheers

Steve

 

[poetix99]

I'll say that if indeed stages 2 - 9 are ~50% reaction, then the resultant drop-off in turbine output might be greater than I'd originally guesstimated.

The effect of damage to a rotating blade row increases as the pressure drop across the blade row (i.e. reaction) increases. This could be as much as 2 or 3% of the total output in your case, with 50% reaction, and if the blades are not-so-well cleaned up. However, please note that I am in murkier water with these conjectures. [Cotton's book discusses these effects at length.]

If the "pass-out" steam is downstream of the damage, but still superheated, and if you have any record of the temperature of the pass-out steam before and after damage, you might be better able to assess the effect of the damage.

I'll also mention (perhaps too late) that it is important to have verified that the trailing edges of the stage 4 stationary vanes are in good condition. Foreign objects can easily "clang around" between stator and rotor, with the effect of closing off the stator vanes.

I had assumed that the entire turbine was of a nominal "impulse" design. This means only that the vast majority of the stage pressure drop occurs across the stationary blade row. Impulse stage designs might be single rotating row, or double rotating row - velocity compounded or Curtis stages.

Neither impulse not reaction turbines are necessarily "normal". There is a general tendency for European OEMs to emphasize reaction designs, and for US OEMs to emphasize impulse designs, but there are many notable exceptions to this "rule".

 

[smckennz]

As power prices have been low for a few months, things have lost their urgency. But another "dry" is forecast.
My preliminary sums indicate we have lost as little as 200kW or as much as 400kW, depending on the load/efficiency curve profile for the turbine (which I do not have).
Will be visiting the site in 2 weeks to take some measurements.

Cheers

Steve McKenzie

 

[walkerny]

Steve,
The OEM is long gone on English Electric turbines. We have manufactured replacement blading (rotors, stators, control stages)for similar obsolete turbines.

Joe Walker / Stork H&E Turbo Blading. Steam turbine blades, buckets, nozzles, gas turbine compressor blades, vanes, IGV, rotors, stators.

 

[Cookie]

If your unit has 9 rows of reaction blades and you remove one you can roughly estimate 1/9 loss of power. You can dress blade to some degree. If the blades are badly damaged they can be reverse engineered, manufactured and installed within a 2-3 week period. This work can generally be completed during a major inspection of the overall unit.
If you remove the blades from this row machine them off leaving the root of the blade in place. This will protect the root so that new blades can be installed at a later date. Leave the diaphram in place.