Large pumping capacity

[TravisMack]

I am working on a project where there is a fairly large demand. The EOR has determined there are to be (3) 3000 gpm pumps. These are all installed parallel. One of the pumps is to be a backup pump. The design has not been finalized, but the system demand (overhead sprinklers and grate foam nozzles) is somewhere around 6000 gpm. That is why we have two 3000 gpm pumps in parallel to create a 6000 gpm pumping capacity. There were some errors in the EOR pump room design that led to a few other questions. We are getting conflicting information from the EOR and the pump guys. The arrangement is that the pumps share a common suction header and a common discharge header.

Per NFPA 20, a 3000 gpm pump needs a 12" discharge. The EOR showed 10" discharge piping and equipment, so this is what flagged some questions with the pump room design. The questions are now:

1 - Should the test header be sized to accommodate a 6000 gpm pump, or a 3000 gpm pump? The pump guy says that the common discharge header must be sized for the combined pumping capacity. The EOR states that it is only based on 1 pump flowing. I seem to agree with the pump guy, in that you need to test with both pumps flowing so you can test your water supply as well to show you can get the full 150% flow from the water supply. It is easy to size the piping so that the recommended velocities in NFPA 20 are not exceeded. It is just a question as to what is the basis.

2 - Should the common discharge header that the pumps tie in to be sized based on 3000 gpm pump or 6000 gpm pump? I believe that the common piping (after the pump discharge valve) only needs to be sized based on hydraulic calculations.

3 - These are diesel pumps. The individual relief valves are tied into a common discharge header. How should this be sized? Based on 3000 or 6000 gpm pumps?

We are awaiting input from the authority having jurisdiction as well on this topic. The biggest issue is the pump test header and relief valve discharge piping. There are more issues of now trying to fit 12" valves/pipe/fittings on the discharge side of a pump in a room that was tight to begin with and was designed for 10" discharge equipment.

Part of the issue is that NFPA 20 only goes to 5000 gpm tables for sizing pump room piping/equipment. The EOR is stating that the NICET tech is to correct anything in the pump room and provide it per NFPA 20. However, since we have a 6000 gpm total capacity, it is not possible to size things based on the summary table in NFPA 20 if it is to be based on 6000 gpm.

I am just interested as to the thoughts of the guys on this forum. I asked this on the AFSA forum as well.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[LCREP]

Travis

My 2cents from an insurance rep point of view

1 Go with the 6k or what ever the final flow for the fire protection and hose systems will be. NFPA 25 is going to require the test annually and any insurance guy that knows anything will too.

2. 6K see my answer above as to why.

3.Pipe each valve separate, what happens when they need to work on one relief valve and take it out of line, where will the water go for the other valves, perhaps flood the pump room. Yea I know if everyone does what they should the relief valves should never open...., I wish I did not have to pay claims because of stupid people!

Tom

 

[TravisMack]

The system is pre-designed. We just noticed some inconsistencies that brought these about.

The relief valves are piped back to the tanks which are a couple hundred feet away. I don't think the EOR will go for changing to separate relief valve lines, no matter how much sense it makes.

I have been informed that they are adamant that you only design the test header to support the flow for a single pump. I guess the purpose of a pump test is not necessarily to test the water supply. This is supplied by two large tanks, so there is not going to be an issue with water supply. It just seems odd that we really have a 6000 gpm pump "assembly," yet it is tested as a single 3000 gpm pump. But, as a tech, it is just for me to follow the criteria of the FPE. I brought the question. They shot it down and had their reason, so it is how it will be.

They are reviewing the requirements of the common relief valve discharge.

I'll let everyone know how it comes out.

This is another where the FPE does a beautiful layout (it really did look nice and well thought out), which is found to have some errors, and the initial response is the NICET tech is supposed to fix all/any of the FPE errors. I had to kick it back and say that since there is no provision for a 6000 gpm pump in NFPA 20, that I as a NICET tech can't fix the sizing issues. I can find ways to make the increased sizing of the valves/piping fit in the room, but they have to correct the sizing issues.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[LCREP]

Travis

So a year from now how will they be able to meet this section from NFPA 25, 2014 Ed if the pipe and outlets are designed for only 1 pump flowing?

8.3.7.7 The pump performance shall be evaluated using the unadjusted flow rates and pressures to ensure the pump can supply the system demand as supplied by the owner.

 

[stookeyfpe]

Travis:

What is the hazard?

 

[TravisMack]

Scott: This is an aircraft hangar. Military project

LCREP: Interesting question. I believe they will go back and say they are comparing the individual pump curves for the pumps independently. I don't like the proposed design for the test header. I think they have done several of these hangars in the past with this same configuration. I believe I am the first they have encountered that has asked this question.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[TravisMack]

We just found a section in NFPA 20 that describes this exact situation. It seems that my take on the issue has been correct all along:

14.2.5.2.3.1 Where simultaneous operation of multiple pumps is possible or required as part of a system design, the acceptance test shall include a flow test of all pumps operating simultaneously.

The EOR is a great guy. I really like the firm. However, it seems that they have been doing the test header and relief valve for 1 pump for a long time and on several projects. I believe I am the first to question this and they are now trying to save face on this job and all of the prior ones.

I always look at things like this: When you learn you have been doing wrong, you immediately stop and don't continue with the wrong methods.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[TravisMack]

There is another wrinkle to this that will help save the EOR. We have several test headers on the job for foam testing and sprinkler flow simulating. They can use the test header to test each pump individually. Then, they can open up several more valves on the other headers that will allow for flowing both pumps at full capacity. It just won't be all through the test header. Lot's of ways to skin a cat.

It would be great if there were hydrants on the pump discharge, but this is sprinklers only.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[fpst]

I'm curious, do you have to have a way to prevent all 3 pumps from running simultaneously in order to only flow test 2 at a time? (Since one is a backup but otherwise still "possible" to flow)

 

[TravisMack]

Yeah.. Close the valves It is in the way the controllers are set up for when they will turn on.

Travis Mack
MFP Design, LLC

http://www.mfpdesign.com

"Follow" us at

https://www.facebook.com/pages/MFP-Design-LLC/92218417692

 

[LCREP]

Easy turn the controller switch to OFF, works every time.