Hello,
I am looking for someone to help me out in making operating cases for multiple pumps. I know this is a very lame question to post in this forum. I am new to pipe stress calculation and trying to learn things while working on a project. If we have 3 pumps A,B,C in a system and only 2 would be operating (say A&B) at any given time, then how should one simulate the temperature profile of the lines to/from stand-by pump C? I would greatly appreciate the help. Thanks.

Generally speaking, you're best off taking the segments from the tee on the suction, through the pump, and out to the tee on the discharge side, as "ambient". You will need at least four cases to run

1. All Pumps Running
2. Pump A Idle
3. Pump B Idle
4. Pump C Idle

I generally run case 1 for the purpose of sizing spring hangers. Typically, these systems suffer becuase of the difference in growth between the running pump and idle pump piping.

Now, if you have a set of hot pumps (say 300°F+), you'll often find that the spared pump(s) is kept at temperature via a warm-up bypass line that allows a small flow backward through the pump. The reason for this is to avoid the potential for cracking due to thermal shock, but it has the added benefit for us of getting rid of the big temperature differential between the running and idle pumps.

Refer to WRC-449 "Guidelines for the Design and Installation of Pump Piping Systems" which has some excellent piping stress guidelines. For example, what temperatures to use for the idle pump piping, or "warm-up bypass" piping.

Isn't warm up piping arrangement across standby pump in hot service a costly proposition from operating cost view point?

I recollect loading boiler feed pump offers requiring such warm up piping to compare with the alternatives not requiring the same from the pump manufacturers.

If by providing some flexible piping layout we can avoid warming up piping, just kept in P&ID for avoiding cold standby pump in piping analyses, we can claim a value addition of the present worth of the power cost saved for keeping the standby pump hot for the plant lifetime. Moreover, for non-lethal service pumps, one can approach pump vendor to increase allowable say, 50% before finalising the pumps order.

regards
sam

personally I prefer warm bypass piping... even with a flexible, well supported piping arrangement the pump will have to cycle up to the op temp, and usually this is not done in a slow controlled manner... I have seen both approaches the bypass seems to work out the best for the pumps.... one mans opinion

idle pump piping?
loops?

Yer all mad.

is it not the case that non running pumps are usually for run and standby service.
therefor if one pump requires maintanace, then
the other one is brought on line?

In any case, the process is temporarally interupted or de-cycled to allow for switch over.

Then again, if its a nuclear reactor............

keine vestein?

verstehen kein?

unt fuhrer?

typo sorry....


ASnyhow I do not get the point of your post...

"
Yer all mad.

is it not the case that non running pumps are usually for run and standby service.
therefor if one pump requires maintanace, then
the other one is brought on line?

In any case, the process is temporarally interupted or de-cycled to allow for switch over."

My point was when the spare is kept toasty and then put into service in an unplanned push the button manner (which is typical) because the bearings and all the internal parts are warm and ready to go, there is a distinct advantage to the bypass keep the stand by pump warm approach.

No, not usually. It is uncommon to interrupt the process. Switchover requires a 'dance' of valving & start/stop on the spared pumps. (Chemical Process Industry norm.....)

Surely the case of the cold pump has to be evaluated. There may be an instance of a failed seal and this has to be replaced. Then your bypass warm up piping will be isolated, the pump will cool to let the chaps work on it so temperature differentials will exist.

Yes of course, but I still prefer to keep the standy pump toasty.

Analysis should always reflect real world scenarios and equipment loads should be met for all cases.