Problem in crack gas compressor start up

[Mohammad Ka]

Hi everyone
During start up of an ethylene plant, discharge temperature of the crack gas compressor (fifth stage) was drastically increased, meanwhile very high amount noises like fast opening and closing of a valve was observed. Moreover the compressor faced with insufficient flow. In this case, opening of the antisurge valve was 100%.It should be noted that there is no FE on antisurge loop Of CGC stage 1 and no piping pockets in the suction. However, a pumping trap is installed in upstream of suction line CGC stage 1. Antisurge valve with a difuser is specified to be a low noise
Other trends could be listed as below:
1- Outlet temperature of crack gas from quench tower was increased about 4-5 deg.C.
2-Level of discharge and suction drums of the crack gas compressor (stage 4) were increased during starup.
3-Discharge temperature of 5th stage of the CG compressor was increased to 130 C.
4- Fluctuation of flow was observed in all five stages of the compressor.
5-Alongside the flow fluctuation some small fluctuinations of pressure drop was observed in caustic tower.
6- No vibration in CG compressor was detected by DCS.

The following activities to remedy have been done:
* Antisurge valve was opened and fully inspected . No failure was detected.
* All suction and discharge drums, check valves, orifices, cooling water exchangers and caustic tower in the line were inspected and no chocking or unusual thing was seen. 
* About 1000kg mixed yellow oily water was observed in cooling water exchanger of 5th stage of the crack gas compressor. Antisurge valve is branched from outlet of this exchanger. The water was completely drained. 
*Stage 5 of the compressor was inspected by a borescope camera and no failure was detected.

After these remediations and start the plant for the next three times, the noise, insufficient flow and increased temperature were again repeated.
I was wondering if anyone could propose a solution? 

 

[rotw]

I made a request file to dropbox:

https://www.dropbox.com/request/pKL7lKMiB2IR5h4EJBJ4

As for the construction aspect, I would also need the following additional information; if you could provide it, it would be appreciated:
- Number of impellers per each process section and their tip diameters ;
- Which case is the design case and/or the certified case;
- The design speed (rpm).

PS: please also blur / make invisible any reference to client / company / project location, etc.
So this remains anonymous as a sort of 'academic' exercise only.
Thanks!

 

[Mohammad Ka]

georgeverghese
here is the data for ASV 1 & 2:
ASV 1 : 187 t/h (min.) , 277 t/h (normal)
ASV 2 : 80 t/h (min. at 80% speed) , 277 t/h (normal)
Do you believe that these specifications may cause problems in start-up of the compressor?
And can you please describe more your previous note "You can only ramp up pressure on 5D when there is sufficent fresh feed from the furnaces." ?

 

[rotw]

Mohammad Ka
First off, thanks for sending out the docs. Well received. Also thanks for the work to consolidate all the info. It is presented clearly.

This goes far beyond my expertise, but just tentatively so we explore new causes: Could it be the case that controllers are fighting each other; the performance controller and anti-surge controller must be decoupled. If not, when performance controller provide new speed set point to turbine to ramp up, anti-surge controller and performance can have antagonist actions. So the gains and other parameters are tuned so both are dis coupled. I think otherwise this may lead to instabilities. Possibly due to process reasons the process was too much upset for the dis-coupling to be robust enough and be able to cope with that. Again, liked I said, one more speculation here.

By the way, would be able to provide on top of the documents you send already the following map:

Outlet pressure and temperature vs. flow for the design case?
Thanks again for your efforts in preparing and sending out the material.

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[hacksaw]

Your Item 4

"Fluctuation of flow was observed in all five stages of the compressor."

What is happening in the Gas Cracking Furnaces?

 

[Mohammad Ka]

hacksaw
No abnormal conditions were observed in furnaces during the runs. Outlet temperature, pressure and gas composition all were in the normal range.

 

[Mohammad Ka]

rotw
The turbine speed controller and antisurge controller has been decoupled and no interference has reported yet.

 

[rotw]

yet, the quick opening and closing of the valve is a matter to question.
Reference to your OP: "fast opening and closing of a valve was observed".

Few possibilities I can think of:

- Process upset could be coped with by control system;
- Process upset exceeded control system capabilities (instruments, calibration, characterization, not designed for the signals fed to system);
- Process upset triggered control system to behave unexpectedly which in return made process situation much worse;

I just have the impression that the control system was 'fighting' to keep the train in operation, that is attempting to fulfill the set points value from master control (DCS or operator). Thus we see quick oscillations of the parameters, until it finally trips.

As precision, I meant performance controller not speed controller. The performance controller is the one that provides speed set point value to speed controller that further actuates the turbine governor. The performance controller should be 'the PIC on stage 1S'; its set point should be train suction pressure. Anyway your response is noted.

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[Mohammad Ka]

rotw
Thank you for your response
In case of "fast opening and closing of a valve ...", a valve here refers to the check valve after the caustic tower and not a control valve.
Appreciating for clarify the subject, the performance controller does not interfer the antisurge controller.

 

[rotw]

Any update about the following map - if possible:
Outlet pressure and temperature vs. flow for the design case?


Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[georgeverghese]

Will take a closer look at these in a few days time. Am travelling overseas at the moment on a personal matter.

 

[georgeverghese]

Back from overseas trip. The dropbox link is not accessible to me, but maybe we can still manage without them? My email address is apachen01@hotmail.com if you would want me to see them anyway.
Your earlier response says each furnace has a capacity of 30t/hr each for 3 furnaces, but your company website says 1267 e3 t/annum, which converts to 50t/hr/furnace based on 350days run time per year. So total compression capacity for CGC would be approx 150t/hr. ASV 1 and 2 appear to have sufficient recycle capacity even when there is zero fresh feed from the 3 furnaces.
However, you also said earlier that ASV 2 is not activated by low pressure on stage 5. A PIC on stage 5S should feed its output signal to the antisurge algorithm to open the ASV 2 through an auto signal selector. This will avoid having the ASV 2 going too near the surge line and waiting for the antisurge algorithm to act.

 

[rotw]

Mohammad Ka

I did some work on the material, but due to time limitation I could not do more at this time.
Have modeled your 1 stage only to do some tests.

- You appear to have a double flow machine as first stage (3+3 impellers) while your map is for the total flow. Is this correct?
- I have some doubts on the gas composition for normal operation. I doubt the heavy HC's would lead to a MW=18.6 g/mol. Please check?

Anyhow, I have assumed a gas composition so to arrive to the same MW=18.6 g/mol as for Normal Case (this is not accurate but just as first attempt). I do not get close to the discharge temperature you've provided, either because of assumed compounds or this provided temperature data may need to be reverified.

After I model Normal case, I run a different case, for which I have made a sensitivity and used gas composition such as to obtain MW=20 g/mol (same as your failed start up gas composition).

Preliminary Findings:

If I keep the operating speed identical as Normal and inlet conditions identical as Normal, except gas set to MW=20g/mol; then:
The discharge pressure goes slightly up but the point moves near surge (I got it on a control line set at 12%).
Because stage #1 is not equipped with individual anti-surge valve, it is needed to do the exercise for each stage 1 to 4; but I believe the overall map is more stringent, as this will have shorter margin to surge than the case where each stage would have its own anti-surge valve.

So, I suspect that a gas of MW = 20 g/mol instead of 18.6 g/mol will cause surge.
To get a confirmation, it is needed to model accurately all gas compounds and then run simulation for stages 1 thru4 and also stage 5 separately.




Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[rotw]

As continuation of my previous post....

Some notes:

- Something looks strange in the trends. The temperature indication downstream of stage #5 does seem to be erratic and departing from the value we would expect based on the compression ratio. I suspect a possible erroneous indication from the temperature transmitters. For example, on 2nd unsuccessful start-up, the temperature cannot be 85 deg C. that is too low; it could be that the temperature at inlet of the stage#5 is not stable which may have cause such behavior but this hypothesis was previously discarded if I am correct. Note that I assumed for all unsuccessful start ups, a gas of relatively constant gas molecular weight. So please check this: temperature transmitters and acquisition downstream stage 5.

- It does not seem possible to ramp-up this compressor train to nominal speed with a gas MW=20 g/mol. This would put stages into surge, anti-surge valves would open and even then - the maximum casing allowable working pressure would be exceeded on stage #5 and possibly excess of power requirement.

So, the plausible start up scenario to me is; start up with heavier gas (MW=22) should be okay until train has reached at minimum speed, then stabilize, then start to build new gas in the loop, and this new gas would need to have lower MW so that it permits to ramp up to nominal speed (MW of that gas should be around 18~19).

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[Mohammad Ka]

georgeverghese,
Just three furnaces are used for start up. During the normal operation the number of furnaces in service shall be increased to meet the design capacity.
As far as I remembered, I noted that " The ASV controller uses suction temperature and pressure, discharge flow, temperature and pressure to adjust the valve opening." at 13 Sep 19 05:53. However, It could be misunderstood or misspelled that ASV 2 is not activated by low pressure on stage 5. Already there is a PIC as you mentioned on 5S with the same function.

 

[Mohammad Ka]

rotw,
Thank you for your effort to simulate the phenomenon.
1- Yes, the first stage is a double flow machine.
2- I checked the MW and it is valid based on the vendor, design and laboratory data.
3- I have the same problem with discharge temperature in simulation of the CGC and it could be solved by adjusting the temperature data somehow.
4- Generally speaking, the heavier gas, the lower force is required for compression. It does mean that the CGC would be operated in far away from the surge line with higher MW. If we supposed that the higher MW gas moves the operating point to the surge line, 100% opening of ASV 1 & 2 would solve this problem. Please consider that MW from 18.6 g/mol (in normal operation) to 22 g/mol (in start up) is recommended by vendor.
5- The TTs, PTs and FTs have been checked and all OK.
6- Normally from the start up to the normal operation the gas MW is decreased but I do not understand that how the lower MW could permit to ramp up the compressor while the higher MW NOT ?

 

[rotw]

2- I checked the MW and it is valid based on the vendor, design and laboratory data.
The total MW is certainly valid, but as example on stage#1, normal gas: 49.439% C2H4 + 32.51% C2H6 cannot lead to a MW=18.6 g/mol.
Therefore, I think the gas composition must be re-checked.

3- I have the same problem with discharge temperature in simulation of the CGC and it could be solved by adjusting the temperature data somehow.
Please clarify what you mean by adjusting temperature data?

5- The TTs, PTs and FTs have been checked and all OK.
Noted, yet I cannot resolve the conflict between compression ratio and resulting discharge temperature in the trends.

4- Generally speaking, the heavier gas, the lower force is required for compression. It does mean that the CGC would be operated in far away from the surge line with higher MW. If we supposed that the higher MW gas moves the operating point to the surge line, 100% opening of ASV 1 & 2 would solve this problem. Please consider that MW from 18.6 g/mol (in normal operation) to 22 g/mol (in start up) is recommended by vendor.
6- Normally from the start up to the normal operation the gas MW is decreased but I do not understand that how the lower MW could permit to ramp up the compressor while the higher MW NOT ?

Yes, what you say is correct, if you start up away from surge it is not problem but when you say 100% opening of surge line will solve the problem, I think generally speaking, this depends on the sizing of the anti-surge valves (CV's). In your case, I suspect CV is sufficient if MW=22g/mol was already taken into account for start-up sequence by OEM. If you start along surge control line, then you cannot ramp up.

Yet, when you reach the nominal speed line with heavier gas, you cannot build up pressure by closing the anti-surge valves with a high gas MW because at some point you will exceed design limits, so you would need to have enter into the process loop a lighter gas and build a lower MW (18.6 g/mol) to further reach the nominal operating point. Do you agree?

In the trends you have provided, I cannot read clearly the pressure, temperatures. The start up process is dynamic in itself but I suppose there was enough time for variable to stabilize so the readings should be static-like when I look at the trend and the duration. Yet, there is conflicts between measured temperature and expected values based on compression ratio which I could not comprehend.

Let me know your thoughts.

Added:

- You have indicated the check valves have been checked, this includes downstream Stage#5. This points remains very unclear (check valve chattering).
Maybe downstream stage#5 is a sort of 'feedback loop', so that the flow must feed forward immediately with downstream system quickly relieved from hot gas to ensure temperature is stabilized. From trend, it looks temperature increases exponentially. Anti-surge valve might not be sufficiently fast to relief downstream system (for instance like a hot gas by pass valve would do) versus fast train speed ramp-up. Not sure then how gas temperature would effect check valve proper functionning, if at all.

 

[georgeverghese]

Will assume the ASV on stage 5 IS activated by low pressure on 5S without waiting for the antisurge algorithm to act.

In most of these trends you have sent, compressor speed is approx 4600rpm = 85% of max normal speed, and stage 5S pressure is approx 17barg in most cases. Pls check with your instrumentation engineer if the ASV2 has sufficient recycle capacity at this operating condition. We see from the compressor design case startup performance trends that the operating point at 13000m3/hr for stage 5 is quite close to the surge point for this case at 11000m3/hr.

 

[Mohammad Ka]

rotw,
Thanks for your notes.

1- The mentioned gas composition is based on weight percent so in mole percent is 32.6% C2H4 + 20% C2H6
MWavg= 0.326*28.05 + 0.2 * 30.07 = 15.16
It sounds the data is correct.

2- For simulation of the compressor I used the ploytropic head and efficiency from the performance curve by reading the graph. I believe that reading the data from a graph may produce some error. So I adjusted these data in a narrow range to meet the discharge condition of the CGC. However, there is still deviation between simulation discharge temperature stage 5 and design value!

3- Can you please explain more about this conflict?

4- Certainly as long as the compressor is in surge control line, it could not be ramped up and this problem was seen in this case. However, I checked the CVs of both ASVs for different MWs. Increasing MW from 18 to 22 g/mol would grow the passing flow rate in ASV 1 and 2 by about 10 %. It means that the CVs are sufficient for this range of MW. Its should be noted a gas with lower MW has higher amount of light components and it tends to occupy more space in the valve rather than heavier component which causes to reduce the passing flow rate. So the higher MW is better than the lower MW for start up.
As long as you do not exceed design pressure limit, where this is valid in our case, a higher MW gas could build up higher pressure in comparison to a lower MW.
5- I mentioned that the check valve in upstream of 5S (Stage 5 suction) was chattering (between tower and the CGC stage 5) and NOT in downstream.

 

[rotw]

Mohammad Ka,

1- Noted. Thanks for your explanation.
2- Noted. Which type of simulation are you using? do you have a vendor simulator for your cracked gas train or third party one or else?
3- If I look at 2nd unsuccessful start-up in your trends, I use same data at inlet flange of stage#5
Pin = 10.22 bara (I assumed 1 bar pressure drop between stage 4 and stage 5)
Tin = 323.15 K
Massflow = 89805 kg/h

I consider Speed = 4650 rpm and MW= 19.9
Pout (predicted)= 18.7 bar-a (from trends = 18.9 bar-g)

Outlet temperature (predicted) = 373 degK (from trends = 358 degK) -> that is 15 degK !!
This is what I meant by conflict. By the way, the point is very much positionned on control line.

Anyway, I modeled all 5 stages now.
If you like me to check a particular scenario, I can do (if this is useful for you). You could also use the dropbox and I could give you access to dropbox if needed.


Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[rotw]

One question:

Is the maps you have provided 'as predicted' or 'as tested' ?

That might solve my 'conflict'...


Life is not about waiting for the storm to pass. It's about learning dance in the rain.