Pump casting broken 2

[srisua]

Dear seniors

I have erected a positive displacement pump along with piping in one of our project. after completion of erection of piping we have observed a crack at pump discharge flange. Photographs of the same attached herewith. My customer is saying that this is happened due to mismatch of piping. I said that this much crack will not develop even pipe line alignment is not good, vibration can be developed. I am presuming that this crack developed due to defects in pump body casting. Request all of you to suggest what could be the reason?

http://files.engineering.com/getfil...1b9-aae13e157bd3&file=IMG-20150923-WA0008.jpg

http://files.engineering.com/getfil...213-487fd8b302ef&file=IMG-20150923-WA0006.jpg

Regards

 

[Artisi]

First up this is a horizontal split case pump,not a positive displacement pump.

As for the crack, difficult to speculate from a couple of photo's, could be misalignment, excessive load on the flange, vibration or a combination of all these features. When ever I see large diameter pipework supported on a timber chock I immediately think very poor installation techniques.
You need a well qualified engineer onsite to undertake a thorough and careful investigation of the complete installation.

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.)

 

[Artisi]

Having now looked at the photo's at higher resolution and a larger screen, I would say misalignment. It is obvious that the crack at the top is much wider than at the bottom, to me this indicates that the pump has been installed with the discharge flange loaded in an axial direction away from the pump with a higher load at the top of the flange compared to the bottom - appears the pipework may have been "forced" / "pulled" onto the pump flange to make the connection during installation -- or the high-tech pipe supports have settled into the ground following the initial installation resulting in loading the discharge flange to the point of failure.



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.)

 

[JJPellin]

I agree with Artisi. This casing is probably cast iron. The piping alignment is probably off drastically in the axial direction or there was a severe angular misalignment at the flange faces (open at the top). Break the flange loose and measure how far the pipe jumps and in which direction. This looks like a very stiff piping configuration. We would require that the two flange faces were parallel within 1/32" and the gap (unbolted) was the thickness of the gasket plus or minus 1/16".

Johnny Pellin

 

[MJCronin]

just my two cents.....

1) It looks like these pumps are on a Cooling Tower in Circ Water duty (supplying cooling water to a steam surface condenser) This is a notoriously bad application for this type of pump due to cavitation problems when either the cooling water becomes too hot or the water level in the tower drops too much. Many pumps have been ruined in this configuration. I am sure that these types of things will not occur at your plant.

2) Based on the size of the piping, the extreme reducers and the nozzle sizes of the pump, I assume that this pump has been "repurposed" from another installation and, possibly is being run at a higher speed than originally designed. What is the new speed for this pump ? Running a split-case at a higher speed will also increase the NPSHr, cavitation and induce instability problems in the future.

This small pump was forced into place between two massively stiff pieces of piping.

3) Regarding casing cracking, a good exercise for a new MBA would be to compare the costs associated with a (used?)cast iron pump replacement versus using a properly sized and selected DUCTILE IRON OR STEEL CASING pump.

You get what you pay for !!!!

Save Money !! Do it only once, the right way.... with competent people !

MJCronin
Sr. Process Engineer

 

[dbill74]

Looks to me like the alignment is off. Looks like the pump is new and the pipe is existing/old.

I would be curious to know what is supporting the blue air separator on the pump suction side. Like the pump it looks new. I wonder if it could be pulling the pump down on that side causing the discharge side to be misaligned.

 

[ScottyUK]

Just as the pipe leaves the tower basin there looks like there's an expansion bellows or something similar. I think I can see two of the four rods used to hold it in position for assembly at the 12 o'clock and 9 o'clock positions. Is the blue filter(?) supported on anything, or just cantilvered out from the pump?

 

[srisua]

Dear sirs,

After seeing your replies I had spoken to the fabricator who had erected pump and piping. As per him please note the following

1. Pump is not operated as of now. Hence no vibration generated.
2. The major mistake what he observed is at the discharge side of piping there was no support and piping including valve length is 4mtr and then it it going underground to connect header. Civil supports which you are seeing in photograph was done after erection of piping. Due to this piping weight flange might be broken.
3. Mr. Dbill, blue colour item is not air seperator, that is strainer to filter dust particles.
4. Yes Mr. MJcronin, this pump is for supplying of cooling water to condensor only. But plant is under construction, as of now pump was not started. Yes, I also always prefer steel casing pump instead cast iron body, but in this case pump was in customer scope and piping is in my scope. Hence we have erected the pump given by customer.
5. Mr. Scotty UK, blue filter was supported by foundation.

Now you are requested to give your views

Regards

 

[bimr]

With such poor workmanship, the crack may be due to a number of issues.

Pump vibration may be an issue. However, one would point to the improper installation of the pump. The pump foundation is too small and the pump does not appear to be anchored to the foundation.

The pipe supports are all inadequate and may have caused the problem.

The strainer appears to be hanging off the pump. This may have caused the crack.

The pipe coming out of the ground may have caused the crack due to forces of temperature expansion and thrust.

 

[srisua]

Dear Mr. Bimr,

Pump has not started as of now and hence vibration is not an issue. Pump is properly anchored to foundation.

Yes, as you said inadequate supports may caused

Strainer is not hanging, it is properly anchored

There is no chance to for thermal expansion and medium is cooilng water.

Regards

 

[Artisi]

The views have already been given, strain being applied to the casing from misaligned pipe work, unsupported load into the case etc.

You are saying that the pump has never been operated (run)- is that correct?????

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.)

 

[srisua]

Mr. Artisi

Yes sir, pump has never been operated. This project is under construction. pump was supplied by customer, we have erected the same. Further piping was designed and erected by us. After few days of erection, we have noticed crack.

Regards

 

[Artisi]

My view doesn't change re loading the pump casing by poor pipework installation - although it might be possible, but very unlikely that the casing had a casting fault resulting in an easier failure. I would suggest that the pipework is immediately removed from both sides while carefully observing what happens as the flanges are relieved of any pipe loads, although now that the case has been reshaped (cracked and moved from its original as cast form) you may not notice any undue difference between the pipe flange and the pump flange. The pump supplier should be contacted to witness the dismantling and possibly a metallurgist be invited to look at the cracked casing.

The wider gap of the crack at the top of the discharge branch is very telling which to me says, misalignment of the pump flange / discharge pipe flange by a considerable amount.

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.)

 

[srisua]

Mr. Artisi,

your suggestions are noted and thank you very much.

Regards

 

[MJCronin]

Me Srisua,

Many installations of this type incorporate large robust rubber expansion joints at the suction and discharge of the pump.

These expansion joints serve two purposes; during operation, they isolate sensitive equipment from vibration and they facilitate installation and fitup. (Yes, I know that the manufacturer states that they should not be used solely to mitigate pipe strain)

Since there has already been damage to the pump during installation and the system must be reworked, please consider these as a possibility.

http://www.garlock.com/en/products/rubber-expansion-joints

I still firmly believe that there will be cavitation problems.

While I understand that it is not your business, I would like to see the pump curve and the expected operating point.

MJCronin
Sr. Process Engineer

 

[bimr]

srisua,

Don't understand the statement "There is no chance to for thermal expansion and medium is cooling water."

You say that the system has never been operated.

On one end of the pump piping, you have a concrete tank which obviously will not move. On the other end, you have pipe coming out of the ground.

If this pipe is empty and sitting out in the sun as pictured, there has to be some change in temperature during the day. There may be as much as 10,000 lbs force daily. This could snap the pump fitting.

http://www.pumpsandsystems.com/topics/piping/basics-pipe-thermal-expansion

Is this installation in Russia.

 

[MJCronin]

What is the speed of the motor driving the pump ?

What are the sizes of the pump inlet and outlet and the size of the piping connecting the butterfly valves ?

MJCronin
Sr. Process Engineer

 

[srisua]

Kind attn: Mr. BIMR

Dear Sir,

This installation is in India. Project is under construction, we have erected pump and piping just 15 days back. Pump has never been operated as of today. As it is winter in India, temperature is 30 deg C only. I should have not used the word "medium is cooling water" as pump is never operated. However I dont expect any thermal expansion in 30 deg C. As all of you suggested there may be problem in installation i.e, mismatch and lack of pipe supports. But as our fabricator is so experienced and done so many installations, i am also suspecting defects in pump casting?

Kind attn: Mr. MJcronin

Dear Sir,

Herewith I am attaching pump General arrangement drawing. Motor speed is 1488 RPM. Please note pump is never operated as of now.

Suction and discharge sizes of pump is 400NB and 300NB

http://files.engineering.com/getfil...-afd5-4051-9a16-c7807acc6a7e&file=cw_pump.pdf

Regards

 

[desertfox]

Hi

That's an interesting failure but I would doubt any flaws in the pump casing and agree with others that undue stress has been exerted on the casing either through misalignment or differential expansion of the components (expansion and contraction could easily happen between day and night temperatures).
Another aspect of that failure might lie in what order and level of bolt preload each bolt in the flange has been tightened to.
The pump installation manual should state the maximum bolt torque for the flange bolts and the order that the bolts should be tightened in, if this hasn't been adhered to then the cracking in the flange you're seeing is possible.
If I was in your position I would ask for the procedure the installer used to erect the pump and any documentation that was used, I suspect the answer you seek lies in that process.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[srisua]

Kind attn: desertfox

Dear Sir,

Yes, as you said there might be undue stress due to inadequate pipe supports. I am also suspecting casting defects in pump. I dont feel thermal expansion between day and night temperatures as in India, at the moment maximum 30 degC and minimum is 20 deg C.


Regarding bolt tightening, i just had spoken to fabricator, he had not used any torque wrench, only hand tightened. As we installed so many pumps like this as of now, we have not asked installing procedure and erected as we have done earlier.

My customer asked to replace the pump as we have erected though he supplied the pump. At present i have taken a stand and trying to convince my customer that this is happened due to defects in casting.

Regards