BFP PUMP SERGING AND IMPELLER CUTTING-OFF 3

[ryuchangmyong]

Let me say about problems of BFP Pump, its rated specification is follows.
1) rated flow : around 500m^3/hr
2) Min. flow : 200m^3/hr
3) Discharge Head : 240 bar
4) Centrifugal Type and 8 Stage
5) NHSPr : 8m
6) NPSHa : 16m
7) Test water temperature : 20 oC
8) Operation water temperature : 170 oC

1. Current Situation
When we get this BFP pump operation, we found so serious noise and vibration in a discharge ppiping line and it also appears on suction piping.( Currently, we call this phenomenon as Pump serging)
Currently, we guess that this caused by lower NPSHa at suction piping where deaerator was located at 16m height above BFP pump, In there are 3 sets of BFP pump, one of them is little bit better than others, so we believe that this was caused by NHPSa because each BFP head loss in suction line is different, thereby, short piping where located BFP is better relatively.
When we get pump test performence curve as attached file.

2. My opinion is follows at current pump operation
1) This pump should be getting right side inclined curve basically, so flow have to be decreased as much as head loss, otherwise, excessive head could cause suction head loss and thereby, serging occurs in suction line due to excessive pressure accumulation at min. flow.
and then this energy accumulation cause water hammering not only suction line but discharge cline together.

3. Query over this situation
Is it possible to get impeller cutting off in order to have inclined curve right side against parallel line between Min. flow(200m^3/h and rated flow(500m^3/h)
And how much we can expect decreased discharge head and flow rate at 5% Impeller cutting off.

 

[MJCronin]

IMHO, Most BFP problems are caused by poor piping design practices on the suction piping.

Multistage boiler feed pumps are unique and demand a bit more engineering knowledge and experience than simple single stage pumps.

As a minimum, BFP suction piping must satisfy two complex and related criteria:

1) The NPSHa must be greater than the NPSHr with a sufficient margin. ( I am not sure the margin is acceptable here)

2) Upon startup, the suction piping diameter and configuration must allow an adequate volume of liquid to flow into the pump very quickly. The liquid flow cannot undergo cavitation conditions within any stage of the pump during this startup event.

With multiple boiler feed pumps, it sounds like your suction piping is a mess.... Can you provide an isometric sketch of the suction piping with diameters ? What is the suction piping diameter compared with the suction nozzle of the pump ?

I have never seen a worse performance curve for pump operation......

Have you disassembled the pump and examined the first stage impeller ?



MJCronin
Sr. Process Engineer

 

[ryuchangmyong]

Dear MJCronin

Thank for your review, additionally this BFP pump inlet and outlet, piping nozzle size is follows.
1) Pump Inlet nozzle (10"), Velocity : 2.5 m/sec
2) Pump outlet nozzle (8")
3) Inlet nozzle piping line (14")
Currently, this pump was disassemblied, and have plan to get impeller trimming
in order for lower discharge head as attached performance curve.
So, I need to confirm if this impeller diameter to be cutting off (Except first impeller)
from 2nd vane to 8th vane, is it to get expected curve as shown on attached curve.

 

[LittleInch]

ryuchang,

I think you're missing the point here. From the information you've given us, I'm finding it difficult to work out how this pump works at all, but without more information on the suction/inlet system, you can trim the impellors all you like, but if the inlet system is flashing into steam as you describe, then all you will achieve by trimming the impellor is getting an unbalanced machine which has a worse performance than you had before. Only the pump vendor can tell you with any certainty what the curve would look like with smaller impellors, but I don't think the basic shape of the curve wil change, only a curve in parallel but further to th bottom left corner.

This sounds a bit like someone trying to drive a car with the handbrake on then changing the gearbox ratios, but still leaving the handbrake on and expecting the car to some how suddenly work better and not have a burning smell....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Artisi]

Check with the manufacturer, it may only need the final (No 8) impeller trimmed to get the head reduction needed.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[MJCronin]

Again, can you provide an isometric sketch of the suction piping to your pump ?

Do you have five to 10 diameters of straight piping upstream of the suction nozzle ?

In your IOM manual from the manufacturer, what was the recommendation for suction piping configurations ?

Was the pump used in another service ?

Some owners use a VFD for high energy, multi staged pumps. They are important for smooth startup.

Has this been considered ?

MJCronin
Sr. Process Engineer

 

[1gibson]

You need to make sure that the performance of pumps running in parallel matches very closely. If one was at max head tolerance and the other was at minimum, then the flow won't be shared evenly and one pump will run at a higher flow, there may be NPSH problems and cavitation. Check the motor amps and see if there is a large difference between the pumps.

 

[ryuchangmyong]

Thank you all for recommendation,
Let me tell you additional information alongwith Piping Sketch as attached.
1. Pump specification is follows.
1) Density : 0.888
2) Dgn Temp : 220oC
3) Operating Temp : 179oC
4) Suction flow rate : 528.3 m^3/h
5) DIscharge flow rate : 515m^3/h
6) Bleed off flow rate : 13.3m^3/h
7) Suction pressure : 11.29 kg/cm^2a
8) DIscharge pressure : 177.3 kg/cm^2g
9) Bleed off pressure : 61 kg/cm^2a
10) Total TDH, HP/IP : 1870m/ 560m
11) NPSHab : 16m
12) NPSHre : 8m
13) Surge margin : 5%
14) Shut-off pressure : around 1.2 time TDH

2. Current situation at commissioning operation in cold condition.
Surging(Noise and Vibration in suction line and discharge line) occurs at mininum flow rate around 200m^3/h and, when flow rate to be gradually increased to MCR, No surging occurs.

3. I believe that this pump have some problems in performance curve at mininum flow due to parallel line of H.Q curve (that should be inclined line)
Moreover, I find out that surge margin is little bit small of 5% considering suction pressure loss.

4. If we change suction line to 20" from current 14" as long as 5m just front of BFP pump,
is this better than trimming impeller diameter, In my opinion, trimming impeller can change performance curve, this is my key concern in this matter.

 

[ryuchangmyong]

Dear All
Let me show you BFP pump ISO line, as attached file.
1) Pump Inlet nozzle (10"), Velocity : 2.5 m/sec
2) Pump outlet nozzle (8")
3) Inlet nozzle piping line (14")
I got a realized that suction line is smaller considering NPSHav even if Pump vendor they followed requirement of API code, Current flow velocity is around 2.5m/s,
Can we getting down of velocity to 1.0 m/s with change of inlet size.

 

[Artisi]

OP: I haven't read you post all that closely as it's not an application I would ever want to get involved with, but will offer the suggestion that if the pumps are over capacity - reduce the output by impeller/s diameter reduction. Plus you need to pay close attention to 1gibson advise of ensuring all pumps are prefoming very closely in terms of flowrate.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

ryuchang,

Can you post the pump curve please and any details for the impellors you have.

There is something strange happening here at your lower flow rates which isn't clear from the information supplied. It tends to indicate that your issue is a pump related one if the issue is at low flow rates.

The other option is to ask the vendor for his CAVITATION curve, which can be quite different from the NPSH curve. I suspect something is happening in the pump at your lower flows which is either causing the liquid to heat up further such that the water boils or that the pump at that flow requires higher pressure to avoid cavitation.

At that temperature (180C), what do you calculate your vapour pressure to be? My estimate is somewhere between 9 and 10 kg/cm2.

what is your suction pressure graph doing during these "surging" events?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[LittleInch]

This article might help explain better than I can.

http://www.chemicalprocessing.com/articles/2006/093/

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[ryuchangmyong]

Dear LittleInch

Let me show you performance and NPSHr as shown on Vendor data.
Actually, there is different performance during commissioning test. Currently I'm sure that NPSHa is less than NPSHr at mininum flow as shown on attached performance curve at my first post.
NPSHr : 8m
NPSHa : 16m
However, at mininum flow condition, there might be additional pressure loss or NPSHr is more greater than Vendor data.

 

[LittleInch]

ryuchang,

That curve is more or less what I would expect.

If you read the article carefully, it shows that in some instances - may be this one maybe not - the Cavitation curve can detach from the NPSH curve (normally 1 to 2 m above it) and increase dramatically at lower flow rates.

Therefore ask the vendor specifically for the Cavitation curve. You might need to get hold of the technical office and not the sales person to get what you need, but it looks clear to me that your system struggles below 200m3/hr and this is a possible cause.

To fix it you either need to modify the first impellor or just run it higher than 200 and put in a minimum flow bypass??


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Artisi]

Once again without analysing all your data, I think you need to look at the actual operation in terms of pump performance suitability at the elevated temperature - as I would suggest that the performance testing has all be carried out at ambient temperature and actual site conditions are completely different.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[1gibson]

That curve is very flat at low flows. These are made up numbers and an exaggeration, but a difference of 2m in head might cause one pump to produce 150 m3/hr and the other to produce 250 m3/hr. No mystery there about surging at low flows, one glance at the curve tells you these are not well suited for parallel operation below say 300 m3/hr. Increase your min flow to 300 m3/hr if possible.

 

[MJCronin]

Again, multi-staged boiler feed pump problems are more complex than the average NPSH problems for smaller pumps....

With all of the money and time you have invested, I suggest that you consider hiring an experienced pump consultant to get a quick opinion. Find one that has solved high energy fluid flow pumping problems in the past.

You have not stated if you have contacted the pump vendor..... What has the pump vendor advised you about the installation ?

At the end of this problem , please respect us..... and share your final solution.



MJCronin
Sr. Process Engineer

 

[Artisi]

I agree with 1gibson, up the minimum flowrate or increase the bypass, running pumps of this power input at throttled flow is asking for trouble.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[ryuchangmyong]

Dear ALL

Thanks for your recommendation until now, I will let you know what happen after impeller trimming in a week. Pump vendor have a plan to trim impeller 2nd to 8th to avoid surging at minimum flow.

 

[ryuchangmyong]

Dear ALL

I wanna post the BFP performance curve after impeller trimming, this curve nothing improved with impeller trimming, on the contrary, Actually, this Curve get more worsed against before.
Once again, this BFP pump had problem with surging in a minimum flow range, however, after impeller trimming, surging range is increased and flowrate to be hunting in normal flow range.
We got this BFP pump impeller trimming from 2nd impeller through 8th impeller leaving 1st impeller
in order to improve surging probelems at minimum flow.
I believe that NPSHre was increased and, Q, H, L was also decresed as much as principle of similarity as below.
Q2=Q1(N2/N1) * (D2/Q1)^3
H2=H1(N2/N1) * (D2/D1)^2
L2=L1(N2/N1) * (D2/D1)^5