Pump Start-Up Pipe Stress 1

[JackofallMech]

In pipe stress analysis is it typical to take into consideration pump start-up thrusts or is it rare that start-up thrusts cause more stress than already operating systems?

 

[BigInch]

In some cases you might want to consider the starting or stopping torque and thrusts too. Most of those forces would be taken out by the anchor bolts if rigidly bolted to a foundation, but if the pump is spring mounted a good portion of that torque will be resisted by any rigidly connected items. The same could be said for thrust, but to answer your question, I haven't seen either of those done for plant or pipeline design, but those pumps always have very rigid foundations. For submersible pumps suspended by the discharge pipe at great distances below ground, it is possible that the only torque resistance is provided by piping suspending the pump. There I have seen starting and stopping torque considered as a load to the pipe.

http://virtualpipeline.spaces.msn.com

 

[JackofallMech]

Is there usually an initial pulse force generated by the fluid when the pump kicks on and acceleratates the fluid from rest, getting it up to speed?

When I was talking about a thrust I was actually thinking in the axel direction rather than torsional.

thanks for you input though

 

[JackofallMech]

Sorry Biginch

I read your comments a second time and realized that you did address the axial thrust.

regards

 

[BigInch]

I also thought you were not interested in the pressure pulse transmitted into the pipe, because usually people refer to that as waterhammer, or surge analysis etc., rather than "pipe stress". There can also be a "radial pressure" spike, in pipe stress terms.

http://virtualpipeline.spaces.msn.com

 

[JackofallMech]

Thank Biginch

Now that I understand the lingo a little better. Is surge analysis common practice for determining pipe stresses? Do most pumps create a pressure pulse at start-up that can be a cause for concern or is it usually insignificant when compared to stresses on an up and operating system?

Regards

 

[BigInch]

You've hit on probably the biggest discrepency in pipe stress/hydraulic analysis, being that most pipe stress calculations ignore surge stresses completely, using only the design pressure and temperature change and most hydraulic analysis ignores the axial and bending stresses due to imposed loads and temperature changes.

The codes don't help much with the many and often confusing cases to analyze, but do attempt to deal with this apparent discrepency by setting surge pressure limits slightly above design pressures and temperature stress limits well below yield stresses. Yup, good old safety factor. Surge pressures (usually) add directly to the design pressure as the maximum operating pressure case is also the case normally taken as the base event when initiating a surge event too, so the end result of the hydraulic analysis is design pressure + surge. But, that's the end of it. Hydraulic analysis does not consider much more than pressure allowable, ignoring even the axial pressure pipe stresses (closed end stresses) created by the surge and the base case axial stresses and bending moments caused by imposed loads and temperatures.

The pipeline codes for example allow a 10% overpressure (above design pressure, or more appropriately the Maximum Allowed Operating Pressure) before other measures must be taken, such as adding a pressure relief system to limit overpressures to 10%, or increasing the wall thickness and consequently upping the the MAOP until design + surge pressures are less than the new MAOP + 10%.

Most centrifugal pumps with electric motors start fast enough to create a pressure surge of some kind. Electric motors generally have a starting torque 50% higher than the running torque. That fact shouldn't be too terribly significant for the pipe stress engineer, since the hydraulics engineer should have limited the pressure to 10% of the design pressure, or recycled the design back to the pipe stress engineer if he could not hold it within that range.

Note that there is at least one "if" in there, so its probably a good thing for the pipe stress guys to keep in the back of their minds that a 10% allowance is given for surge pressures (in the case of ASME B31.4 & .8 pipeline codes, check your code if different from those). Its also common that surge pressures are still not considered at all in some cases, even though surge analysis is required by most all design codes.

http://virtualpipeline.spaces.msn.com

 

[LSThill]

BigInch (Petroleum)

Additional piping Stress Analysis may be required using other software application like BOSS FLUID.





L S THILL

 

[JackofallMech]

Biginch

Do you have any refences that you could recommend for calculating surge forces that I can apply to stress analysis? I would think the pump manufacturer should have some of the info I need.

 

[BigInch]

If you can find my e-mail, "contacts" section on my MSN webspace page and let me know your e-mail address (don't post it here), I can send one along to you.

http://virtualpipeline.spaces.msn.com

 

[stanier]

Another Code matter that you may take into account is that the pump station may be inside the boundary fence and another code such as ASME B31.3 applies rather than the pipeline code.

ASME B31.3 allows for an increase in longitudinal stress up to 1.33 times design under a certain number of conditions. One condition is that the allowance is subject to the Owner's approval. All other stresses have to be taken into account. The conditions are too long to include here.

For pipeline with rubber ring joints the surge pressure is taken into account in design of concrete thrust blocks.

The question of fluid of fluid structure interaction is complex. T Some sophisticated software can provide answers but generally it needs experience to apply the boundary conditions.

Books by Casti and Becht cover the use of this code much better than I can explain in one posting.

As for surge analysis I would recommend Fluid Transients in Pipeline Systems by Thorley.

I strongly suggest you engage some one who understands the analysis and the design for surge . It is not an area for the enthusiastic amatuer.

Geoffrey D Stone FIMechE C.Eng;FIEAust CP Eng

http://www.waterhammer.bigblog.com.au