500 MW INDUCED DRAFT FAN NOT RESPONDING TO BALANCING

MOTOR WITH JOURNAL BRG. THEN VOITH HYDRAULIC COUPLING WITH SCOOP CONTROL COUPLED WITH FAN  WHICH IS SIMPLY SUPPORTED ON JOURNAL BRG. FAN HAS IMPELLAR WITH 11 VANES. INLET GUIDE VANES ARE PROVIDED AT SUCTION.

1.Initially a 3kg weight (previous balancing wt.) was already in ID Fan-4c and vibration was 125 microns peak to peak (phase angle coming 345deg).

2.At initial trial 3kg weight was removed and vibration was taken. Vibration 230 microns phase angle 343 deg.

3. At same location of 3kg weight ,we added 7 kg weight as calculated by balancing diagram. But vibration didn't reduced. Vibration was 142 microns 260 deg.

4.In next trial as per balancing diagram we removed 7 kg weight and added 6.5 kg at 34 deg clockwise but vibration increased instead of reducing.Vibration was 210 microns at 250deg.( Vibration varying from 160 to 210 microns pk to pk).

5. Since balancing was not responding ,next weight came 4.7 kg in clockwise 42 deg but we placed it anticlockwise 42 deg.Vibration reduced to around 125 microns. Variation in vibration is there from 80 -120 microns pk to pk .(overall via. trend attached).Also at 85% scoop when fan speed was 508 rpm vibration was varying from 80 to 120.But with 100% scoop with rpm 560 rpm vibration was nearly constant at 100microns.

 

Attachments

Files (1)
Original Post

I haven't plotted it out but I'm in basic agreement with your "numbers" yielding a solution of about 6.5 kg.  However, the data you've attached uses units of acceleration, velocity and displacement.  Your discussion uses displacement.  Is the data from an accelerometer that has been integrated or are the displacement readings from proximity probes?  If proximity probes, are you compensating the data?

yes accelerometer .we are using emerson  CSI 2130 vibration analyser .we have both displacement and velocity values.since This Induced draft fan is slow rpm (560 rpm) so we have been using displacement values for balancing.Phase is taken using photo tachometer (pasting reflective tape).

I agree with John whether the data is compensated for run out or not? how much weight of rotor?

Before you balance rotor, have you tried to turn the rotor 180o and run to see how vibration and phase change? how about direction of rotor rotation?

As you said it look like rotor bow behavior but how about vibration at speed 100-150rpm?

No we haven't turned rotor by 180deg.We adopted the normal graphical balancing method.vibration was taken by emerson CSI 2130 (peak and phase) using photo tachometer.

Unfortunately we haven't observed vibration at 100-150rpm as already each time we had taken vib. ,fan was already reached 450+ speed. Speed is scoop controlled by voith hydraulic coupling.

 

I was asking about compensation in case you were using proximity probe readings.  

Using the same accelerometer and CSI 2130, have you ever had this (or similar) fan significantly lower than the initial 125 um?  Your 125 um at a 1X of 560 cpm is equivalent to 0.02 g peak.  There is obviously response to trial weights, but generally only to actions that take the amplitude higher.  It seems that you may have reached an end point that might be due to the transducer limitation.

I had a case many years ago in a power plant where we added temporary proximity probes for balancing.  The plant was absolutely amazed at how smooth we could get the units to run as compared to their efforts using a casing transducer.

Temp prox probes can be helpful sometimes for this.  Crusty used to be here, and he was a fan of plywood.  I had some prox's mounted to plywood (non-conductive when the case was ground); they were battery powered.  This is not difficult to do.  

One can look for bows better.  Runout is often a problem that needs to be dealt with for temporary mounts.

My balance calculation yields 6.56 kg at an angle of 2.4°, assuming that the original trial weight of 3 kg was placed at 0°. The CSI balance programs assume phase is positive opposite direction of rotation. I assume that since you are working with weight instead of weight X radius that the center of each weight is placed at the same radius or the same distance from the outer edge of the fan wheel. It would be good to measure the amplitude and phase at the other bearing during each balance run to be sure that the unbalance continues to be in a single plane. This can be useful information to determine whether a single weight should be placed on the backing plate or on the shroud or split between the two.

The cross channel phase does indicate single plane unbalance, but you did not give the amplitude ratio with that. Did you correct the cross channel phase angle for the axial direction for sensor orientation? All balance runs should be at the same shaft speed. Significant changes in vibration at other speeds would tend indicate a resonant structure or a balance resonance. Balancing ID fans can be problematic when balance changes as temperature changes. The fan wheel should be cleaned prior to balancing, otherwise deposits may fall off and create an unbalance that wasn't there previously.

It would probably take longer to mount proximity probes with or without plywood supports than it would take to balance the fan. I will let John and Bill do it that way! Who is "Crusty"? "The plant was absolutely amazed at how smooth we could get the units to run as compared to their efforts using a casing transducer." Wow I am impressed!

Walt

Walt Strong posted:

 

Who is "Crusty"? "The plant was absolutely amazed at how smooth we could get the units to run as compared to their efforts using a casing transducer." Wow I am impressed!

Walt

Crusty = Rusty C...from another forum.  

The power plant was so impressed that they added proximity probes to every fan (of similar type) in multiple plants.  

The full story is we were called into the plant because their internal vibration guy was headed on vacation and had not finished balancing (with casing transducers) multiple fans.  We were to complete the task.  Our guy had balanced some fans previously using proximity probes and had good effect data.  He mounted prox probes (likely glued with aluminum blocks) and called a one-shot solution.  He then just moved down the line, using the refined effect data to balance multiple fans.

I might add that this particular power gen facility took great pride in their people and had some very sharp people doing their vibration work.  Bill F., when with BN in Philadelphia, likely worked at the facility and even more likely knows the people I refer to.

 

we had balanced many similar ID fans earlier using same method. This particular fan had not followed logics.

e.g.. 1.it had 3 kg initial weight and vibration was 125microns  at 345 deg.

2.we removed that 3kg weight and vibration was 230 microns at 343 deg.

3.so we placed 7kg weight at same location and removed 3kg earlier weight.

vibration should have reduced below 125 microns logically but it was 142 microns at phase 260 deg.

 

 

The unbalance level could be the lowest reachable for this fan. Other possibilities: bearing attachment issues, improper clearancies (worn bearings, worn shaft at the bearing or rotor position...), shaft, rotor or coupling eccentricity, rotor changed rotation during procedure, uncleaned rotor, thermal welding shock, hydraulic coupling unbalance, misalignment, nearby resonance and many more.

TUSHAR000001 posted:

we followed all rules.

1. balancing at same speeds.

2. weight placed at same radius.

3. cleaned impeller before attempting balancing.

4. captured amplitude and phase at other planes too. (phase was same in all horizontal planes where vibration was more.)

 

On number 4., I would add that these other points can be used for the calculation as well.

On number 1 this is held strictly by most.  If equipment is operating near a critical or a data point is near a critical, this can cause trouble.  Machines have parameters that change, such as alignment; other types of machines have pressures that affect the location of the criticals.  So, balancing at the same exact speed may not mean the data is taken in the same spot relative to a critical.  I have no idea if this is your situation, but this can cause a big issue in balancing.  Fans are often near a critical -- just advice for what it is worth.

When one uses a data (or data points) further from a critical, the balance data has less sensitivity to the exact speed of the data (potentially less sensitivity to other parameters).  

========================= 

I did work in the Philly area on some turbines, but I didn't have my plywood then.  Times have changed, and one wouldn't need the plywood.  Virtual plywood would work today.

(1) Is this a center hung or overhung rotor?

Quote:          We adopted the normal graphical balancing method.vibration was taken by emerson CSI 2130 (peak and phase) using photo tachometer. Unquote

(2) Does the above quote mean the data was plotted on polar graph paper, or the use of the cosine or sine laws was used, or what? Does your 2130 meter not have the Fastbal Program?

(3) I noticed that some of the direction of weight movements were in the direction of rotation and some were against direction of rotation. At least that seems to be the case. Am I correct? If so, why did we do that?

(4) Where was your reflection tape located in reference to the where the "old" 3KG weight was?

(5) Was the tach light and transducer in the same plane and facing the same direction?

(6) Have you done a coastdown plot of the fan? Have you done a bump test on the fan (especially on the bearing where the transducer was placed during the balance job)?

Thanks and Have a Great Day,

Ralph

 

Quote from John of PA.......   Unquote

John,

Quote   

Ralph Stewart posted:

 

(5) Was the tach light and transducer in the same plane and facing the same direction?

Ralph

Ralph, why is this a consideration?  Something unique to balancing with the 2130?

Unquote

No. Nothing unique with the 2130.  Same basic balancing design as the old 2120, 2115, and 2110.

The main "consideration" is, determining where the 1st trial weight possibly would be located, without a mind set of "anywhere will do". After the 1st trial run, this positioning is irrelevant, unless  tach light and/or transducer is moved. IMO.

True, just placing a weight anywhere will work sometimes, but other times, "just any where" could prove to be dangerous. IMO.

I am also wondering if they had balanced this fan prior to this time and the 3KG weight is the weight and position which worked the previous time, and that is the reason for placing the 1st trial weight of 7 KG in the same position as was the 3KG, instead of calculating a new trial weight position. To me, this information would be very important to know, as would other information. But, that is just my thinking.

Thanks and Have a Great Day,

Ralph

Ralph Stewart posted:

Quote from John of PA.......   Unquote

John,

Quote   

Ralph Stewart posted:

 

(5) Was the tach light and transducer in the same plane and facing the same direction?

Ralph

Ralph, why is this a consideration?  Something unique to balancing with the 2130?

Unquote

No. Nothing unique with the 2130.  Same basic balancing design as the old 2120, 2115, and 2110.

The main "consideration" is, determining where the 1st trial weight possibly would be located, without a mind set of "anywhere will do". After the 1st trial run, this positioning is irrelevant, unless  tach light and/or transducer is moved. IMO.

True, just placing a weight anywhere will work sometimes, but other times, "just any where" could prove to be dangerous. IMO.

I am also wondering if they had balanced this fan prior to this time and the 3KG weight is the weight and position which worked the previous time, and that is the reason for placing the 1st trial weight of 7 KG in the same position as was the 3KG, instead of calculating a new trial weight position. To me, this information would be very important to know, as would other information. But, that is just my thinking.

Thanks and Have a Great Day,

Ralph

Yes this fan was balanced previously in 2012 and 3 kg weight is previous balancing wt. .It was running with vibration in range of 80 microns .(alarm limit kept 100microns for this fan). Recently around 3 months back ESP pass Renovation work was going on and one ESP pass out of 4 ESP pass was isolated. At that time in routine vibration monitering  we found this fan vibration varying from 120 to 150 microns at 95% scoop of hydraulic coupling (fan speed around 528 rpm). After all 4 ESP pass available vibration was on higher side and varying. At 100%  scoop (speed 560 rpm)vibration is stable but in alarm range. 

So to start with we removed the previous 3kg weight as trial reading. And after removing that weight vibration at increased from 125 to 230 microns and at same phase angle so we added 7kg weight. But vibration came as 142 microns.

we had pasted reflective tape as reference so in all reading ref is same. Also traducer and photac were placed at same location in each reading. 

From previous experience these induced draft fan have shagging problem.

Quote  from John of PA    "Having the "tach light and transducer in the same plane and facing the same direction" is not a requirement for placing a trial weight that would reduce vibration, as opposed to a position that would increase vibration." Unquote

I do not think I said it was a "requirement". Or did I?. 

Sorry if I mis-lead anyone. To me it is a simple way to make an educated effort as opposed to just throwing a weight on and hoping it is not on the "heavy spot".  If one ever places a weight on the heavy spot and the system is running at or near its critical frequency....... one might can imagine what might happen. But that is another story there,.... pre-determining the resonant frequency in relation to its running speed(s).

After 29 years and hundreds of balance jobs with the use of the CSI balance program, I find it works quite well for me. 

What works for one, may not work for all.

But, like has been said, it is not a "requirement".

Just my opinion.

Thanks and Have a Great Day,

Ralph

 

 

Quote from BECAR    "Ralph, I know this trial weight position method exists, but I have used it only one time and never again. I was using CSI 2120-2 at that time. However the vibration level increased. Has anyone some similar experience?" Unquote

BECAR,

I am sorry it did not work for you.  I am not sure how you "used it" (possibly filling in the blanks in the programs setup). The first and second tried I tried "it", the amplitude increased also, so I developed my own, "fill in the blanks setup" on a "hard-copy" sheet.

As we all know, once the job passes the 1st trial eight positioning run, it's smooth sailing then, the analyzer does the rest for us.  All we have to do in make sure we are measuring accurately with the new weight size and positioning.

Sorry if I lead one to believe the "alignment" is a requirement. It just "helps" me to get close to the "light spot" on the 1st attempt. If I can not "align", I will do the angular "math" difference and mentally adjust from there.

I believe, but that is just my thoughts, that placing a 1st trial weight just "anywhere" is not the way to do it. I know if it is running well below its critical, the 1st weight position is simple.

What works for one, may not work for all.

But like has been said, it is not a "requirement".

Just my opinion.

Thanks and Have a Great Day,

Ralph

5. Since balancing was not responding ,next weight came 4.7 kg in clockwise 42 deg but we placed it anticlockwise 42 deg.Vibration reduced to around 125 microns.

Does this mean the 4.7Kg is 8 Degs (42 minus 34) anticlockwise from original or 42 Degs from original?

Has there been any change in the lube system - flow, viscosity, type? Just seems like it might be "damping" since everything else should be the same.

I have done enough ID fan balancing jobs to know that many things can go wrong. The very first thing to do when the balance result is not as expected, besides checking your calculations (including input data), is to remove the balance weight and repeat the baseline run. If the baseline run is not repeatable, then something has changed with the vibration characteristics of the machine. That "something" can be a number of things, but you first have to realize that it happened and try the balance procedure again!
Walt

Tushar, in your initial post you state:

1.Initially a 3kg weight (previous balancing wt.) was already in ID Fan-4c and vibration was 125 microns peak to peak (phase angle coming 345deg).

2.At initial trial 3kg weight was removed and vibration was taken. Vibration 230 microns phase angle 343 deg.

3. At same location of 3kg weight ,we added 7 kg weight as calculated by balancing diagram. But vibration didn't reduced. Vibration was 142 microns 260 deg.

You should have STOPPED right there!  Nothing you did after that run is going to work. When phase shifted from 345 / 343 to 260 degs, your "system" had changed.  As Walt stated, the ONLY thing you could do was go back to #2 and see if the data repeated, that is, removed all weights and see if you still had 230 microns @ 343 degrees.  If not, something in the "system" had changed.  You have to determine what changed. If your reference data is not repeatable, you simply cannot balance the fan (or whatever).

--------------------------------------

The trend is your friend.

MachineryAnalysis.org

Ralph Stewart posted:

TUSHAR000001,

Thanks for your reply.

Were you using the "Fastbal" program during your balance job? Or just using the amplitudes and phase readings to calculate the solutions (which will work as just as "good".).

Just wondering.

We are not using any program ,only polar plot we are using with amplitude  and phase.

Rusty Cas posted:

Tushar, in your initial post you state:

1.Initially a 3kg weight (previous balancing wt.) was already in ID Fan-4c and vibration was 125 microns peak to peak (phase angle coming 345deg).

2.At initial trial 3kg weight was removed and vibration was taken. Vibration 230 microns phase angle 343 deg.

3. At same location of 3kg weight ,we added 7 kg weight as calculated by balancing diagram. But vibration didn't reduced. Vibration was 142 microns 260 deg.

You should have STOPPED right there!  Nothing you did after that run is going to work. When phase shifted from 345 / 343 to 260 degs, your "system" had changed.  As Walt stated, the ONLY thing you could do was go back to #2 and see if the data repeated, that is, removed all weights and see if you still had 230 microns @ 343 degrees.  If not, something in the "system" had changed.  You have to determine what changed. If your reference data is not repeatable, you simply cannot balance the fan (or whatever).

--------------------------------------

The trend is your friend.

MachineryAnalysis.org

I too feel system has changed , and ID fan rotor has also got bow. Phase difference between Fan DE brg. pedestral and NDE pedestral is 178 deg . So please suggest what action should we take. 

Attempt a fresh balancing by removing all weights by repeating O1??

Also how to confirm thermal bow ,and balance it?

Mr Tushar I have following questions.

1.Type of fan axial with static blades ,axial fan with hydraulic operated blade pitch or radial fan.

2.if fan is axial with static blade profile weather you have inspect the impeller and measure the blade thickness .

minimum blade thickness reduce to 50℅ of original replace the impeller.

3.if fan is radial check the ash deposition inside the cone area.

 

 

S TIWARI posted:

Mr Tushar I have following questions.

1.Type of fan axial with static blades ,axial fan with hydraulic operated blade pitch or radial fan.

2.if fan is axial with static blade profile weather you have inspect the impeller and measure the blade thickness .

minimum blade thickness reduce to 50℅ of original replace the impeller.

3.if fan is radial check the ash deposition inside the cone area.

 

 

1.its radial double susction fan , suction side has inlet guide vane (motor operated).

speed of fan is controlled by voith hydraulic coupling (scoop control)

3. First step is we ask maintenance dept to clean ash diposition and inspect impellar.

 

Quote       "Phase difference between Fan DE brg. pedestral and NDE pedestral is 178 deg ."    Unquote

When you say, "Phase difference between Fan DE brg. pedestral and NDE pedestral is 178 deg" are the readings on the actual pedestals frame in the horizontal direction and same plane, or are the transducers  on the bearings themselves in the horizontal direction and same plane, where the balancing data is being taken?

Do you have the horizontal amplitudes and phase from both the DE and NDE of the fan?

Thanks and Have a Great Week End,

Ralph

Add Reply

×
×
×
×