19 replies

Light Rub in Compressor at seal or not ?

JSDPAdvocate
Dear All expert,

I have the vibration problem of centrifugal compressor in MP stage driven by turbine. There were the fluctuated and sudden increased vibration many times in the past 10 months. Vibration data seem to show about the “light rub” problem in DE of MP compressor at seal but I'm not sure. I would like to summarize about some point of this problem from attached data.

1. All spiked vibration problems were shown by 1X rotating speed in MP casing. Both DE and NDE side vibrations of MP casing have contrast correlation. But there was no vibration problem in turbine and LP compressor unit. (See in vibration data 1)
2. For any spiked vibration, there was the fluctuated vibration before spiked vibration was occurred. (See in vibration data 1)
3. Lube oil and seal oil analysis report showed that oil conditions are rather normal in wear, oil properties and contamination.
4. Seal oil has high contamination in sludge and varnish. I had sent seal oil from seal oil trap both MP and LP casing to test gravimetric analysis. There are some coking deposits from seal oil both MP and LP casing in same quantity.
5. In 10-15 Dec 2013, Radial vibration in DE of MP casing slightly increased with no correlation factor. After that the spiked vibration was occurred and the 1X phase of both X,Y probe in DE of MP casing was changed 50 degree approximately. I’m not sure what problems were occurred in this time. (After this event, the vibration baseline on DE of MP casing rather higher than before and the problems occurred more frequently.)(See in vibration data 2)

I would like to inform you my idea about the possible causes as of this moment as follows.

1. There are the “light rub” on seal in DE of MP casing due to poor seal oil properties in contamination from oil oxidation or sludge and varnish. But I don’t know why I can’t see the difference of coking deposits from seal oil trap both MP and LP casing clearly.
2. There are the fluctuated of lube oil header pressure and DE MP casing bearing may be some abnormal. So the vibration problem was occurred in this bearing randomly.
3. The trend of vibration are increasing from fouling problem, aren’t it ?

Now the vibration probe(XI-392-04 A/B) and journal bearing temperature probe (TI-392-05 A/B) are fluctuating more after spiked vibration in 10 Apr 14. And I think the spiked vibration may be occurred again ?
So I should recheck another data to confirm this problem ? And please let me know if you have some advice for me or require some additional information.

Thank you in advance

Attachments

Hide
Files (1)
Take Action
  • Report Topic
  • Permalink
  • Printer Friendly Format
    • Original Post
    JSDPAdvocate
    quote:
    Originally posted by Ngo Dinh:
    Hi,
    As my view, I think that it is due to fluid whirl on 23rd Feb and light rub could be a case in 10 April.
    But could you please zoom more to see subsynchorous vibration. What type of bearing do you use for MP compressor?


    Hi Dinh,

    I have understand in your point.
    I have some question for your answer.

    Bearing, is tilting pad type, can not be occurred fluids-induced instability, can it ?
    You can see subsynchronous component in vibration data 2 at the first reply of this topic that the similar event was occurred in 26 Jan 2014.
    And Would you tell me about fluctuated vibration of this machine ?

    If the fluid whirl is occurred, why this problem show on this bearing (DE of MP compressor) only and how can I prove or avoid it?

    ,JSDP
    ()
    Take Action
    John from PA
    John from PALeader
    Contrary to Dinh, I do not necessarily see "fluid whirl on 23rd Feb" and I base that observation on the orbits and time base plots of that date/time. You do not see clear subsynchronous activity at the typical anticipated 40% to 47% of running speed. What is seen is a lot of what I term orbit "jitter". The full spectrum waterfall plots of 23 February, especially XI-392-04A/B, do show frequencies out to ±500 Hz and some activity down around 10 Hz (that's a best estimate). Can you generate a plot that defines that subsynchronous activity better, especially with respect to frequency?

    The scenario I see most likely from the data is a light rub at the DE bearing, likely from the coking that you feel exists.

    quote:
    And Would you tell me about fluctuated vibration of this machine?


    When coking occurs there is a build up of the coked material until such time as contact with the shaft occurs. At that time symptoms of a rub occur. Quite often, the coked material at the point of contact then breaks off and the vibration drops rapidly. Then the cycle repeats itself, often taking an extended period of time before it repeats. This can occur at the compressor seals as you suspect, but I have also seen it occur at the oil control rings of a tilt pad bearing. If I had to guess, I would say I have seen 25 to 30 confirmed occurences of the behavior over many years, mostly in machines where there is only a prime mover and a compressor. In those instances, there is a tendency that the problem occurs most often on the DE bearing and in my opinion that is due to the changing alignment situation while the machine is in operation. Several variables (ambient temperature, load, coupling type, etc) affect alignment at that connection point more than they will affect the NDE bearing. And when we affect alignment, we affect shaft position within the bearing, thus possibly causing the light rub scenario. By the way similar effects can sometimes be seen in steam turbines due to poor water chemistry causing the build up of deposits on steam seals.

    quote:
    Bearing, is tilting pad type, can not be occurred fluids-induced instability, can it?


    Tilt pad bearings can exhibit fluid induced instability, but this is rare as opposed to the common two-axial groove plain sleeve bearing. The oil control rings that I previously mentioned can act as a very poor plain bearing if the shaft gets in close proximity to the surface. These oil control rings as you likely are aware are typically of non-ferrous material and one will be on each end of the bearing. Tilt pad bearings have also been reported to exhibit a frequency of 1/2X in the instance of pad flutter.
    ()
    Take Action
    Ngo Dinh
    Ngo DinhLeader
    hi John and JSDP,
    after seeing the data 2, i see the shaft position moving so much at two time and rubbing could be a case in this issue; however, i hope to have more information about the case for clear. there are:
    - have you checked the oil filter?
    - have you often taken the oil sample from tank for analysis and how about result?
    - does it have any heat source near bearing housing?
    - what kind of gas the compressor handle?
    - could you take the shaft centerline from time of 0 speed to the event occurred? i would like to see change much, eccentricity ratio? because in normal the oil seal clearance is bigger than bearing. and if vanish, slugde is problem, we could find something from bearing first.
    - hope to see more clear about subsynchronous frequency. i still concern that the orbit is like fluid instability and if fre. is about 0.1 or 0.15X, the problem could be stall/ surge as well.
    - have you ever operated with changing the oil pressure to see what different from vibration?

    it is just my thinking but i hope could help you. I also meet a case of coke deposit and i resolved it. but my case is turbine.
    for answer, pls see comment of john, he is experienced guy. i also learn from him.
    ()
    Take Action
    JSDPAdvocate
    Hi Dinh,John

    Thank you for your help. I had answer your questions but not completed. I think rub were occurred but the oil whirl weren't. As I have already learnt that rubbing is secondary effect of other problem. So I will try to check more data to specify root causes. If you have any advice for me, please tell me more.

    quote:
    - have you checked the oil filter?

    I didn’t check the oil filter. But from oil analysis, there are no difference properties between oil from tank and after filter. (See in oil analysis data)
    quote:
    - have you often taken the oil sample from tank for analysis and how about result?

    I have check oil sample from tank every month to check general test. Both lube oil and seal oil are rather normal in wear, contamination and oil properties. But the lube oil and seal oil have the sludge and varnish problem. There are some coke deposits from seal oil trap at MP and LP in same quantity, not clearly. (See in oil analysis data)
    quote:
    - does it have any heat source near bearing housing?

    The maximum operating fluids temperature is in the range of 95-105 oC. And the heat source may come from steam turbine, or not ? And what temperature does the oil can be heated to coking form ?
    quote:
    - what kind of gas the compressor handle?

    This compressor handle the quench water from naphtha LPG which is sent from quench system. The objectives of this cracked gas compression are compress cracked from 0.5 to 41 barg, knock out water and hydrocarbon, eliminate acid gas, remove moisture and mercury.
    quote:
    - could you take the shaft centerline from time of 0 speed to the event occurred? i would like to see change much, eccentricity ratio? because in normal the oil seal clearance is bigger than bearing. and if vanish, slugde is problem, we could find something from bearing first.

    I can’t collect vibration data from time of zero speed to the event because the machine cannot be stopped.
    quote:
    - hope to see more clear about subsynchronous frequency. i still concern that the orbit is like fluid instability and if fre. is about 0.1 or 0.15X, the problem could be stall/ surge as well.

    The sub-synchronous frequencies are random in the range of 0-0.1X. I can see the peak amplitude at 0.6-0.7X clearly. And I think the surge problem may not be occurred because the operation is in normal condition and the surge protection system is provided. I think the sub-synchronous component may come from external forces. (Impact, Adjacent component)
    quote:
    - have you ever operated with changing the oil pressure to see what different from vibration?

    I didn't try to confirm correlation between oil pressure and vibration but sometimes I can see the opposite relation between them. I think the oil pressure was affected from fluctuated vibration. So, oil pressure isn't the main cause of fluctuated vibration. is it ? Then I will try to change oil pressure to confirm vibration again.

    Please help me to check the oil analysis again for confirmation that I see the oil analysis result aren't clear to indicate the coke problem at MP compressor and the lube oil and seal oil are rather normal in wear, contamination and oil properties. (See in oil analysis data)

    JSDP

    Attachments

    Hide
    Files (1)
    ()
    Take Action
    Ngo Dinh
    Ngo DinhLeader
    Hi JSDP,

    The my question from beginning until now is why oil coke can form.
    as limit of my knowledge, the oil coke deposit is phenomenon of oil oxidation and the good condition for oil coke deposit forming is high heating (normally more than 200oC) and/or the quality of oil is bad. The vanish/sludge is one of element which can be lead to coke forming. That is the reason I suggest you check the oil filter and quality of your oil.
    - About changing pressure to see the vibration correlation, it is that if the coke deposit or slightly rubbing occur when you change oil pressure, the vibration should be changed.
    - about shaft centerline, I mean that you should take from the starting time to operation time to see the eccentricity of the shaft with bearing wall which if it is too high, the slightly rubbing is sensitive for occurring. Besides, the hot alignment could see more clearly in the issue.
    - The heat source could be from the turbine also but in normally the drive end of turbine is rear front with negative pressure therefore the heating is not much. I do not know how about your arrangement.
    ()
    Take Action
    JSDPAdvocate
    Hi Dinh

    quote:
    as limit of my knowledge, the oil coke deposit is phenomenon of oil oxidation and the good condition for oil coke deposit forming is high heating (normally more than 200oC) and/or the quality of oil is bad. The vanish/sludge is one of element which can be lead to coke forming.

    From your information, I think the cause isn't coke forming because there are no heat source over 200 oC. But seal oil have problem about sludge contamination which may be the cause of this vibration problem ?
    quote:
    - About changing pressure to see the vibration correlation, it is that if the coke deposit or slightly rubbing occur when you change oil pressure, the vibration should be changed.

    You mean that if the coke deposit occur, the vibration is fluctuated while lube oil pressure is changing. This phenomenon is always occurred with this problem ?
    quote:
    - about shaft centerline, I mean that you should take from the starting time to operation time to see the eccentricity of the shaft with bearing wall which if it is too high, the slightly rubbing is sensitive for occurring. Besides, the hot alignment could see more clearly in the issue.

    OK, I will check the eccentricity ratio again to confirm the chance of rubbing. Because I don't have information about bearing dimension.
    quote:
    - The heat source could be from the turbine also but in normally the drive end of turbine is rear front with negative pressure therefore the heating is not much. I do not know how about your arrangement.

    In my opinion, the coke deposit cannot be formed because the heat source isn't too much for occurrence. But I think the cause of problem is the sludge contamination in seal oil too much. But I don't sure about the effect of sludge in seal oil with vibration level in MP compressor. And now the vibration is more fluctuated than before after spiked vibration in 10 Apr 2014.(20-25 um.)

    Thank you very much with your helpful discussion for me. I don't have any additional information now. Please tell me if you have another advice.
    And I will ask and inform you again if there are any progress.

    JSDP
    ()
    Take Action
    JSDPAdvocate
    Hi John

    I'm not clear in your answer.
    quote:
    And Would you tell me about fluctuated vibration of this machine?
    "If I had to guess, I would say I have seen 25 to 30 confirmed occurences of the behavior over many years, mostly in machines where there is only a prime mover and a compressor."

    25 to 30 , You mean that vibration level are fluctuating in the range of 25-30 um for coke forming problem ?

    JSDP
    ()
    Take Action
    John from PA
    John from PALeader
    quote:
    Originally posted by JSDP:
    Hi John

    I'm not clear in your answer.
    quote:
    And Would you tell me about fluctuated vibration of this machine?
    "If I had to guess, I would say I have seen 25 to 30 confirmed occurences of the behavior over many years, mostly in machines where there is only a prime mover and a compressor."

    25 to 30 , You mean that vibration level are fluctuating in the range of 25-30 um for coke forming problem ?

    JSDP


    I was estimating that i have seen and been able to confirm the behavior in 25 to 30 separate and distinct machines.
    ()
    Take Action
    Ngo Dinh
    Ngo DinhLeader
    hi JSDP,

    I think that the alignment and bearing clearance should be checked. Though this, you should have more clear picture. Alignment you can check through SCL as mentioned.
    besides, I think if the sludge of oil effect to the seal, it also effects to the bearing. So, if alignment and bearing clearance is not perfect, the sludge and varnish of oil would form a layer on bearing wall which could make temp. be higher than normal. So, the oil use for seal and bearing is the same source? how about alignment, bearing clearance is important question.
    ()
    Take Action
    JSDPAdvocate
    quote:
    Originally posted by Ngo Dinh:
    hi JSDP,

    I think that the alignment and bearing clearance should be checked. Though this, you should have more clear picture. Alignment you can check through SCL as mentioned.
    besides, I think if the sludge of oil effect to the seal, it also effects to the bearing. So, if alignment and bearing clearance is not perfect, the sludge and varnish of oil would form a layer on bearing wall which could make temp. be higher than normal. So, the oil use for seal and bearing is the same source? how about alignment, bearing clearance is important question.


    Hi Dinh

    The oil for seal and bearing is not the same source. So I think the rub may be occurred on seal due to high sludge contamination. Because the rub is the secondary effect, I should recheck alignment and bearing clearance again.

    JSDP,
    ()
    Take Action
    MYASMessenger
    Hi JSDP & everybody

    Interesting case... Cool

    Let me pls start my observations from the Comp DE SCL. it fluctuates by 40 microns left and right while comp NDE doesn't.
    second is that when ever the high vibration starts it is always preceeded by very low subsynch vibration. this isn't evident in the spectrum but it is in the orbits. You can see that in the first document page No(18) orbit No (3) and in the second document page 1 orbit No (3).

    based on that i believe Comp DE bearing is loose or damaged and any upset in the process due to loading or unloading will cause excitation of a low frequency component due to low turbulence in the compressor flow. this is followed by magnified 1X force due to unrestrained shaft due to bearing loose or babit damage.

    this doesn't rule out the cocke buildup theory. I saw somehow a simillar case in robert eisenmann book in chapter 5 case No 13.

    Thanks
    yousef
    ()
    Take Action
    JSDPAdvocate
    quote:
    Originally posted by MYAS:
    Hi JSDP & everybody

    Interesting case... Cool

    Let me pls start my observations from the Comp DE SCL. it fluctuates by 40 microns left and right while comp NDE doesn't.
    second is that when ever the high vibration starts it is always preceeded by very low subsynch vibration. this isn't evident in the spectrum but it is in the orbits. You can see that in the first document page No(18) orbit No (3) and in the second document page 1 orbit No (3).

    based on that i believe Comp DE bearing is loose or damaged and any upset in the process due to loading or unloading will cause excitation of a low frequency component due to low turbulence in the compressor flow. this is followed by magnified 1X force due to unrestrained shaft due to bearing loose or babit damage.

    this doesn't rule out the cocke buildup theory. I saw somehow a simillar case in robert eisenmann book in chapter 5 case No 13.

    Thanks
    yousef


    Hi Yousef

    Thank you for your advice and useful information.

    I try to search about the robert eisenmann book that you recommended. The name of this book is 'Machinery Malfunction Diagnosis and Correlation, Chepter 5: Analytical Rotor Modeling, Case 13: Gas Turbine Response Correlation, or not ?

    If my information about the robert eisenmann book is not true, please advise me about this book.

    JSDP
    ()
    Take Action
    JSDPAdvocate
    quote:
    Originally posted by MYAS:
    Sorry it is case history 14 : "Charge Gas Compressor with Internal Fouling"

    Hi Yousef,

    Thank you for your useful book. There are a lot of interesting information.

    You mean that my cracked gas compressor in MP stage may have internal fouling problem because the vibration was high only DE of MP stage that is the discharge side like this case history ?

    But there was no details about fluctuated and spiked vibration like my case in this case history. Now my problem is about the spiked vibration in a short period that cause the tripped machine if the vibration is over the vibration trip level.

    From this case history, there are about the calculated model which show the result of calculated orbits. Do you have any information about how to calculate the model ? I think this calculated model is rather hard.

    JSDP
    ()
    Take Action
    JSDPAdvocate
    quote:
    Originally posted by MYAS:
    Sorry it is case history 14 : "Charge Gas Compressor with Internal Fouling"


    Hi MYAS,

    Now, I have additional information about this problem.
    Cracked gas compressor was occurred high spiked vibration 2 times about 50-55 um again in 17 Oct 2014 and 11 Jan 2015. But there are some data which differ from previous data.
    --->> Vibration was high at DE side of C-300MP with large forward precession orbit. But NDE side had lower vibration level with 'reverse precession' orbit.
    Note: More information are in attached file.

    Now I think the problem is rotor rub which like a partial rub which something at NDE side. (Not metal-to-metal contact)
    What do you think about this problem with additional information ?

    Thank in advance
    JSDP,

    Attachments

    Hide
    Files (1)
    ()
    Take Action
    John from PA
    John from PALeader
    quote:
    Originally posted by JSDP:

    What do you think about this problem with additional information ?

    Thank in advance
    JSDP,


    You need to go back and review the replies you've already had. I for one don't discount the possibility of a rub, but still feel you need to investigate coking on a seal or a tilt pad bearing if used.
    ()
    Take Action

    Add Reply

    Views: 323
    Likes (0)

    Membership Required

    We're sorry. You must be signed in to continue. Sign In or Register
    ×
    Are you sure you want to remove from your Block List?
    ×
    Loading...
    ×
    When you block a person, they can no longer invite you to a private message or post to your profile wall. Replies and comments they make will be collapsed/hidden by default. Finally, you'll never receive email notifications about content they create or likes they designate for your content.
    Note: if you proceed, you will no longer be following .
    ×
    Advanced Search
    Advanced Search