pump bypass location

[bsfc9]

Hello, I am hoping for some good advice in an area I only know a little about. I am installing a 15hp centrifigual pump, 2inch inlet; 1 1/2inch outlet. I will be installing a spring loaded relief valve in the discharge line to regulate the pressure. Should relief valve discharge into the supply tank or the pump inlet? Thanks in advance for you help.

 

[JRLAKE]

Personally I prefer the suction tank to minimize velocity and turbulence in the pump suction. Even returning the line to the tank requires some consideration such as air entrainment and velocities near tank outlet.

However, I have bypassed to the pump suction in the past without problems, depending on the system and the pump design.

 

[EngAbdelkader]

hell bsfc9,
actually I don't get the point with your post!
specially for the sentence "I will be installing a spring loaded relief valve in the discharge line to regulate the pressure"!!! keep in mind that we install relief valves for emergency cases as they are called elsewhere PSV for pressure safety valves! and NOT to regulate the pressure of the discharge line!

so please re-post with the complete data (that way we can help you better):

- liquid pumped
- viscosity, temperature and pressure of the liquid
- the height of the tank you are pumping from
- the height of the one that you'll pump to
- piping type (material) and length to compute the pressure losses.

kader

 

[JJPellin]

It is generally preferred to route the spill-back line to the tank in most instances as previously noted. This is especially true if the pump will be required to run with zero outgoing flow on full spill-back for any period of time.

The terminology is sometimes confusing. But we commonly use relief valves specified as full-flow relief valves for pressure regulation on small systems. This is very common on small lube oil skids. A full-flow relief valve differs from a typical pop valve (PSV). The full-flow valve will crack open at a certain pressure and gradually open further as the pressure increases. It often has scored seats so it passes some flow even when fully closed. It does not "pop" open at a pre-determined set point. This makes it impossible to test and adjust using equipment commonly used to test PSV's. In most cases, the full-flow relief valve can only be field adjusted in the system at normal operating conditions.


Johnny Pellin

 

[bsfc9]

Thanks for the replies everyone. I'm sorry my terminology may be incorrect. The "relief" valve is a spring loaded manually adjusted valve that will be used to bleed off pressure in the discharge line until the needed working pressure is achieved. The application is a water brake dynomometer. The supply tank--500gallons of water, sit 10ft above the test cell. The total amount of piping is about 15ft total. Galvanized steel pipe. The dyno control valve requires 90psi and the pump will produce over 100, so we need to bleed off some pressure. I think we will bypas back into the supply tank. Thanks again!

 

[JRLAKE]

Just make sure your bypass line empties below liquid level and in an area opposite of the pump suction side. That's my only advice.

 

[EngAbdelkader]

Hi again guys!
bsfc9 why don't you think about a much more simple solution?
I mean why don't you think to throttle the discharge line by installing a valve (a pressure control valve I would say).
Knowing the deltaP=10psi (which will be the pressure drop through the valve) that you are wondering, the viscosity of you fluid (1 cp for water) and the flow rate it would be "easy" to find out the valve that would do the job!

This doesn't answer your question but we'll open an another view or solution to your problem.

To answer your question "Should relief valve discharge into the supply tank or the pump inlet?"
I would say that both of the solutions are possible the difference would be only the pressure drop due to the piping (from tank to pump) that you'll avoid recycling directly to the pump inlet.
I advise you to read this article: "controlling centrifugal pumps" from

http://www.driedger.ca/ce1_cp/CE1_CP.html

it treats the 2 points:
-throttling control
-recycling control

good luck!

 

[djack77494]

I have two comments regarding this post. My terminology differs some from that suggested by Mr. Pellin. I would refer to his recirculation valve as a regulating relief valve, or I would name it by its function as a pressure control valve.

Secondly, regarding kader007's post, I do not agree with his statement,
"Should relief valve discharge into the supply tank or the pump inlet?" I would say that both of the solutions are possible the difference would be only the pressure drop due to the piping (from tank to pump) that you'll avoid recycling directly to the pump inlet."
Recycling back to the pump suction results in a much more turbulent/less developed pump suction, and it fails to provide for any removal of heat. For a small inventory of liquid experiencing full recycle, this could result in rapid overheating with resulting pump problems. In my opinion, it is a thousand times better to return to the suction vessel if at all possible.

 

[pumpking]

Another consideration would be to installa Frequency Inverter (VSD) controlled by pressure sensors in the system - this way the pump can be run at lower speed (and provide big energy savings) to generate exactly what your system requires ?? We use these often in ring main systems very successfully, and save many pounds in energy costs.

Just a thought...

Ash Fenn

http://www.cdrpumps.co.uk

 

[BigInch]

While it is possible, I don't agree that VSD save energy in many systems. They should only be used where flowrate variations are required, the system head can be reached by the pump running at reduced speed and losses due to flow are approximately proportional to the head produced by the pump at the reduced speeds. Low speed settings cannot always reach the required head to produce flows needed in many systems, thus their use and possible energy savings is highly dependent on the exact characteristics of each specific system. It may certainly be worth looking at, just depends.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain