UNABLE TO LOAD TURBO EXPANDER OF ASU

[ARUNKUMAR T]

Dear all,

We are having a turbo expander in cryogenic air separation plant. Few days ago, the lube oil pump tripped due to some electrical issues and the turbine tripped. Mean while the bearing temperatures of the turbine and booster bearings gone upto 180 deg C. After rectifying the electrical issues, we re started the turbine and the vibration levels are found to be high ( 56 microns in place of 4 microns under normal conditions). Hence we replaced the turbine cartridge with new cartridge.

On running the new cartridge, the rotor reaches its full speed of 58000 rpm for inlet pressure of 35 bar G and IGV opening of 34%. This is highly unusual as the normal operating condition is that the roto will reach max. 52000 rpm for inlet pressure of 42 bar and IGV opening of 50%.

Since the equipment has reached its full speed for such lower inlet conditions , we are unable to load the equipment further.This affects the plant's production drastically.

What may be the reason for such abnormality????

Note : Old and new cartridges are of same specification and of same make

 

[rotw]

one question, excuse my ignorance but is the indicated inlet pressure meant at the suction of the compressor or the turbine?
is your IGV system hydraulically actuated and fed by a common lube oil circuit (control oil and lube oil) or is it pneumatic ? just speculating here: if the IGV does not adjust to 50% (suppose this is the set point), then the capacity controller would increase the speed to reach the desired outlet pressure. Again just brainstorming..

 

[georgeverghese]

Someone has made some changes to the capacity controls on the other side of this turbo expander - this controls the speed of the turbine also. The other side of this turbine, which is commonly called the speed brake side, may be an electric power generator, a gas booster compressor etc. The more lightly loaded the speed brake side, the faster the turbine will spin for a given power extraction on the turbine side.

 

[ARUNKUMAR T]

@rotw & georgeverghese:

The indicated pressure is the inlet pressure of the turbine.
The other side of the turbine is a gas booster.
We do not have any auto capacity controller, IGV (pneumatical) of the turbine is controlled manually.
Let me explain our process for better understanding.
# We will start the turbine by opening the IGV manually (upto 50%) with inlet gas at 42 bar where it expands to 5 bar causing a Temp. drop of 50 deg.C
# we will watch the other side of the turbine (gas booster) to do its work - Compresses the gas from 42 bar to 60 bar. We do not give any set point or any other auto control input to the booster. The above mentioned compression will be pertaining to the speed of the turbine (normally 52000 rpm).
# And now, we have opened the IGV of the turbine only upto 35 % with inlet gas at 35 bar, but the rotor has reached its full speed of 58000 rpm.
# And also for such higher speed the gas booster is compressing its fluid from 41 bar to 47 bar only. This outlet pressure of the gas booster is important for us to achieve full production from the plant.

Thus we are not getting the rated discharge pressure of 60 bar from the booster, but the rotor has reached its full speed.

 

[MortenA]

It does not make much sense that the pressure on the compressor side of the TX is lower? The compressor dont create pressure but head. If there is no "resistance" it will generate a low pressure/high flow. How is the pressure on the compressor side controlled. I can see that you state that you dont control it - but why is it lower? Extreme example - if you let the gas out to atmosphere from the compressor - right at the nozzle, the compressor couldn't generate any pressure but it could generate flow. The resulting duty would be low though.

Best regards, Morten

 

[rotw]

On compressor side at least, seems strange to me that - all conditions being equal* - the compression ratio became that low while speed increased to the extend you mentioned. (*I assumed you are not in deep choke, means the system resistance your booster is facing has not shifted).
Thus my question: how confident are you on the accuracy/reliability of the measured 58000 rpm speed?
How is you speed measurement system set up / what type of speed probes are you using; my point is, is there any chance that things may have been affected/upset during bearing replacement procedure ? Also, has the inspection revealed any noticeable damages on the other rotor components apart from damage on bearings?

 

[georgeverghese]

It does look like the reduced brake side booster compressor Pd is the reason for the higher than normal speed of this machine as suspected.
Some one has changed the setpoint on this brake side compressor discharge PIC from 60barg to 47barg - why?

 

[rotw]

quoted
# we will watch the other side of the turbine (gas booster) to do its work - Compresses the gas from 42 bar to 60 bar. We do not give any set point or any other auto control input to the booster. The above mentioned compression will be pertaining to the speed of the turbine (normally 52000 rpm).
unquoted

 

[georgeverghese]

Well, something has caused the backpressure on the brake side booster compressor discharge to drop from 60barg to 47barg - some process change further down the line ? It may even be some change upstream of the booster.
This change is a significant process modification, and there is nothing stated so far on what this may be. Anyway, it is now clear that this is the cause for not being able to load up the expander to design flow at 52000 rpm.
The author of this process change in this plant is probably well aware of this bottleneck.

 

[georgeverghese]

A lube oil failure on any high speed rotating machine is a serious mechanical integrity loss event, and it may be that there are still remaining issues on the mechanical stability / balancing of this machine even after the cartridge replacements. High thrust and vibration loads approaching shutdown limits may be still be present as the machine is loaded up and this may have forced the plant operators to reduce the loading on the brake side compressor.
A complete dissembly, repair and dynamic rebalancing of this turbine-compressor assembly at 52000rpm or so at the manufacturer's shop may be a likely scenario for this machine.

 

[georgeverghese]

Due to high thrust loads remaining on this machine after the bearing cartridges were replaced, the automatic thrust load balancing mechanism may have prevented further loading of the brake side gas compressor by internal recycling in order to keep net thrust loads within limits. Check the expander maintenance manual for these details. There should be DCS readings on the thrust loads on this machine.

 

[ARUNKUMAR T]

Dear All,
on thorough analysis of the entire system, we found that the speed transducer is malfunctioning - it is giving higher feedback of the speed than the actual (Reading 58000 rpm in place of the actual 39000 rpm).
Hence all through this period we were running the equipment at 39000 rpm only. Not as what the transducer read 58000 rpm.
Operating parameters of the turbine and the gas booster are normal for this actual speed of 39000 rpm.

We replaced the speed transducer with new one.
Now we loaded the equipment to 52000 rpm and we are getting 60 bar at outlet of gas booster.

Thus our problem in loading the equipment is solved now.
Still, the reason for the malfunctioning of the speed transducer is unknown.

 

[rotw]

Good !

 

[georgeverghese]

So thrust loads upon restart would have returned to normal after the cartridges were replaced - that should have showed up on the net thrust load readout. That in turn, would have pointed to some problem other than machine imbalance.