Surge Pressure- When is it too much for a system?

[SF24]

This is my fist post so I hope this is in the right forum section.
I have also searched this site and the web for an answer with no luck.

I am currently working on a pump station/ force main design and have called for a 10” surge relief valve to be installed on the header. The client has asked if the surge relief valve is needed or if it can be removed from the design. I submitted the design information to GA Industries who has performed an analysis and confirmed the need for one.

I have a basic understanding of surge (transient) pressures and how to calculate them on a simple system (Joukowsky’s Equation). And that surges can cause failure and fatigue to the system.

My questions are then:
1. How much is too much surge pressure? When it exceeds the pressure rating of the lowest rated component of the system? Or some percentage thereof?
2. Also same for fatigue? When is it too much pressure that it will cause to much fatigue over the life span of the system?

I am sure there is a code somewhere that states this.
Basically what I am getting at is…. At what point can I say there is not enough surge pressure potential or enough fatigue potential to justify a surge relief valve.

Thanks for any help!!!
Walt

 

[bimr]

Your question is too complicated to answer in this forum. One would need to know the extent of your pumping system to answer.

You should be using good design techniques such as soft start equipment on pumps, silent check valves, reasonable pipe velocity, air release valves, etc. to minimize the potential for surge.

Note that the highest surge will probably occur as a result of a power failure.

If you have a large pump station, it may be cost effective to model the pump station application.

For smaller pump stations, the reference book listed below will provide you an outline of the surge analysis required.

Pumping Station Design by Robert L. Sanks

Here is a paper on pipe surge that I noticed on the web. I believe that the writer posts on this forum.

http://www.engineersaustralia.org.a...7-0528-B401-36CA-F0E6D07C1841&siteName=ieaust

 

[ch81pc]

I have used pumps with flywheels instead of surge protection on a few installations. Much less maintainance, smoother starts and stops even in event of power failure. Just a thought.

 

[SF24]

Thanks for the responses guys.

The system will have VFD's, cushioned swing checks, several ARV's and a flow velocity that is reasonable even with both pumps pumping at the same time.

I recieved the following from GA Industries on their method for determining the need for surge protection

"In determining the need for a surge relief valve, we typically consider the following:

1. If the potential surge pressure (in a 'worst case' surge event, i.e. power failure shutdown when pumping at maximum flow) exceeds the rating of the piping, a surge relief valve is definitely called for.

2. If the potential surge pressure approaches the rating of the piping and/or is much higher than the normal pumping pressure, a surge relief valve is typically recommended to protect the piping over long term usage. The sudden development of a surge can be considered a 'shock load'. The long-term effects of repeated 'shock loads' can lead to pipe fatigue and eventually failure, even if the surge pressure is within the pipe rating. This is the situation in "your project" (edited to remove Project name).

Pipe of any material generally does not gain strength over time. Fatigue in PVC piping isn't well defined in the literature. I have attached two different papers that attempt to address the subject from different viewpoints.

3. If the potential surge pressure is well below the rating of the pipe and/or only nominally greater than the pumping pressure, then a surge relief valve may not be required. Many other factors must be considered if foregoing surge protection (e.g. - age of piping, presence of pressure sensitive instruments attached to piping, pipe support & bedding methods utilized, etc.)"


Number 1 was of course obvious. Number 2 was more what I was looking for but maybe with hard numbers/percentages.

Well I think I have formulated a reasonable answer for the client.

Thanks agian,
Walt