Why is Fire Pump Pulling the Underground Dry?

3/6/2024

Had a review comment come back and needed to clarify.

I have a 14-inch city main with an 8-inch tap to a backflow in a pit. It then runs to four hydrants, and then turns into the building supplying a fire pump.

The flow test at the hydrant near the tap (#1 on image below) was 49 psi static, 47 psi residual at 920 gpm.

The flow test at a hydrant nearest the building (#2 on image below) was 49 psi static, 39 psi residual at 750 gpm.

The 95 psi @ 1,500 gpm pump in the building is running the underground dry. The plot curve shows 20 psi at 2,100 gpm, however the pump rep said he barely got 500 gpm at 9 pitot and had to shut it down as the gauges went below 20 psi and air was starting to come out of the 2-1/2" hose valves.

This is a mystery to us - we have five different flow tests all ranging from 750 gpm to 1,060 gpm at 40-32 psi residual.

Why would the pump be pulling the underground so low?

They checked all the valves and rebuilt the backflow in the pit. Water meter is good as well (per the utility department). Looking for suggestions. Thanks in advance.

Sent in anonymously for discussion. Click Title to View | Submit Your Question | Subscribe

19 Comments

Pete H

3/6/2024 06:34:00 am

Random question:

Is the 14" street main you're connecting to part of a circulating main or is it the dead end off a circulating main?

Dan Wilder

3/6/2024 06:49:13 am

1. Are those flow tests being done with just a single 2½" outlet? You're not getting the entire picture of the supply when the test results are only providing 50% of the requirements. Retest with multiple outlets (either 2-2½" outlet, the pumper connection, all 3 together, or additional outlets on additional hydrants). Also, try to get as close to 2250GPM as possible, at a minimum the rating of the pump or the system demand whichever is greater.

2. Stopped the test at 20PSI? Why? If air is coming out, yes good reason to stop so the pump doesn't overspeed/cavitate then get slammed while the UG burps. This would need to be resolved via opening the remote hydrant and let it run until the air has been purged. Taking a test below 20PSI (unless forbidden by the AHJ) is acceptable.

3. Where are the AHJ pumps in relation to this location? Are they down, did they activate, were two supposed to be activated and only one triggered? They AHJ should have good logs of the status and pressures at their locations.

Size of the main doesn't matter if it's got nothing with muscle behind the water (water tanks on hills are notorious for this, having a good static but the residual drops off like a rock with any decent flow demands, same with larger water districts when near the higher end of the system). This may be an issue of needing a break tank to overcome the initial draw until the booster pumps kick in or worst case a dedicated tank if the water supply isn't robust enough to take the demand. I'm wondering how the Fire Flow was established, what that GPM was, and how it was proven/verified.

James Phifer

3/6/2024 08:12:15 am

We had a similar experience once. Turns out there was a 2x4 stuck in an underground elbow, restricting the flow. We ran a video camera thru the pipe to discover this.

Jesse

3/6/2024 08:13:30 am

This one's perplexing, and truth be told; I love mysteries like this.

Dan's on the right track here. I'm wondering if this is a dead-end line or simply a gravity fed system.

Greg

3/6/2024 08:15:54 am

Air...air coming out of the hose lines in the middle of the testing, after you achieved flow ?

There's the problem. If air is being pulled into the water pipe?

From where though?!? Is there some pipe that is allowing air to enter that is part of the fire pump piping ? Is one of the four fire hydrants open ?

Air and the words 'fire pump test' are polar opposite.

Chris

3/6/2024 08:31:16 am

Test #1 as identified above doesn't seem to be pulling enough water (as Dan said above) to be worthwhile data.

Where were the residual measurements taken? At the test hydrant or the flow hydrant? Was the same hydrant used for both flow and residual readings?

I'd recommend reviewing NFPA 291 with regards on how to perform a flow test.

My guess would be that there is a valve partially closed or an obstruction between hydrants #1 and #2. You could narrow down the location of the issue by flowing out of the intermediate hydrants to see if you have similar conditions.

Craig V link

3/6/2024 09:31:14 am

Based upon the flow test results from the hydrant nearest the building, calculations indicate that you could expect a flow rate of 1,332.8 gpm at residual pressure of 20 psi. In comparison, looking at the flow test results from the hydrant near the tap, calculations indicate that you could expect a flow rate of 3,898.8 gpm at a residual pressure of 20 psi. The reduction in available flow to the downstream fire hydrant indicates why you are unable to achieve the flow necessary for the peak load of the fire pump. In fact, if you were to run the test all the way down to a residual pressure of only 5 psi in the underground main, you would only achieve a flow rate of 1,669.3 gpm, well below the 2,250 gpm that you are hoping for. If each valve indicates that it is open, I recommend using a camera to explore the internal condition of the underground main to check for obstructions as James described above.

Jack G

3/6/2024 09:44:12 am

Not being able to read the plan is a problem.
1. Flow test doesn’t flow enough water to determine if the fire pump is too big. Flow. Or too much pressure that requires a relief valve to relieve excess pressure ( above 175 psi- which may be max system pressure)
2. Is there a relief valve ? —- the relief may be sized - relieving excess psi—so that the relief valve is discharging enormous Quantities of water, thru a 6 inch relief valve . ( delta psi to the .54 exponential times 150 will approximate the gallons thru the relief valve ) formula I developed and tested over last 40 years ( works — metered the relief water) . If the pump is oversized and needs to relieve 40 psi— the gallons relieved would be over 1100 gpm. This could cause cavitation- if water supply is below 2250 plus 1100 gpm) main sucking dry and spitting. A better clearer sketch would help determine this so sizes and components can be seen.
A. Need a flow test at 2250 gpm—- thru the test header, pump off. Relief valve closed.
B. Pump test with relief valve and systems off( if exceeds 175 psi )
C. Pump test normal relief valve on.
This should lead you in the correct direction.
D. If on a hill, there may be an air pocket in the main. Clay valve makes an air relief valve that attaches to the underground at the high point that takes care of this.
I had a similar situation — solving a similar problem at a TJ Max warehouse in NE Philadelphia for a FPE friend. Turned out to be the relief valve and pump being too large ( PSI ) wise. 2500 gpm pump. Relieved 75 psi. Engineer choice. ( thought it was ok to size big and relieve excess pressure)
Nfpa 20 doesn’t allow this anymore.
Opinion.

3/6/2024 09:57:10 am

If it turns out that high excess pressure is being relieved— try recirculating it back into the suction ( with a small relief valve- 100 gpm , to make sure relief line doesn’t overheat)
( if water supply is found to be strong— 2250 gpm)

David

3/6/2024 09:49:59 am

Based on test #2 you only have 1333 gpm available at 20 psi. I think the others are on the right track and you have a valve not fully open or an obstruction in your piping. I would also conduct another flow test, opening a sufficient number of outlets to verify you can get at least your rated pump capacity.

Eric R

3/6/2024 10:03:35 am

A useful way to check for closed valves or other obstructions is to perform a flowtest from the fire pump test header while flowing through the pump bypass assembly (I'm assuming you installed one) with the fire pump discharge control valve closed.

You then put pressure gauges on both hydrants and on the other system riser at the right corner of the building to simultaneously record the different residual pressures while flowing as much water as possible at the pump test header.

You can then compare your results to what your hydraulic calcs say you should have at each point based on the friction losses and any disparity between the calcs and reality should be quickly apparent.

3/6/2024 10:12:23 am

Another scenario is the water company rules and regulations.
Usually there is a monthly standby charge for the meter. There is a tendency to size the meter for just “ the fire flow” and not the pump test.
So a meter smaller than the main size could be installed in order to eliminate boo-cuu bucks a month. ( ie 4 inch meter on a 6 inch line 6 inch on a larger line)
At the main Stream light facility I found a 6 inch meter on 8 inch line and too small a line to the pump ( like sucking a milkshake thru a small straw )
Saved a grand a month……….but didn’t work.

3/6/2024 10:19:01 am

Also— flow tests are not accurate beyond the point of their measured flow. The flow test in this scenario should have been 2250 gpm plus domestic water added to it.

3/6/2024 10:30:20 am

Question — right side of sketch— 2 lines- one dark blue the other light blue- and red dots in those 2 lines and in loop line above and below it !
Red dots…check valves !
The 2 lines, one into a pump, the other back into the loop with increased pressure?
I hope not.
Need a cleaner pic of set up. And not in reverse image

3/6/2024 10:40:34 am

If the red dots are arrows, it would appear that you are recirculating the pump discharge into the underground loop than loop to suction . Just making a circle—- not a proper arrangement.

Patrick Spillane link

3/6/2024 10:39:22 am

I would place a gauge before the Backflow
Preventer and use that as your 20psi cut off. So you might get down to 10-15 on the pump, but a sling as the city is above 20psi, you are still in the best practices for pressure.

Also if the difference between the pump suction and pressure before Backflow, that can help narrow down if you have a lockable between the two. But if the pressure before the Backflow is just 7-10 psi higher, then that means it’s the supply coming into the Backflow.

Glenn Berger

3/6/2024 10:45:48 am

When conducting a fire hydrant flow test, you need to know what the anticipated facility demands will be and then test to ensure that the requirement can be met with actual flow test data. Never count of the extension of the plotted line as a guarantee of anticipated flow rate.

Ed Glynn link

3/7/2024 10:28:33 am

Let's start with the basics

What is the sprinkler system requirement in terms of pressure and flow?

That info will help to answer why you have a 95 psi @ 1500 GPM pump.

My very strong suspicion is that the fire pump is over sized for what the supply can deliver and if that is the case and the fire pump can't be swapped out you would need to add a tank based on calcs to compensate for the supply issues.
We experienced a very similar thing where a hospital had an oversized pump that had been installed previously and they could not test the pump without the supply lines being sucked dry.

Anthony

3/14/2024 09:47:21 am

Your hydrant tests make sense given your pipe sizes and
500 ft (guessing) 8'' ID 8.249 c factor 140 for 500 ft lose you apx 6.26 psi using Hazen Williams(.0125 psi loss per ft). Add in that back flow and an 8psi differential in the residuals makes sense.

Now let's look at the test at hydrant 2. You have 1333 gpm available at 20psi due to the loss factors in the 8'' pipe.

For a system requiring 1500 gpm I'd look at 10'' incoming. That would lose you 2.5~3 psi (over 500 ft, flowing 1500 gpm, id 10'' c 140) leaving a residual of apx 45 psi.
So with 10'' feeding hydrant 2 I'd suspect a flow test 49S 45R with a flow apx 1000gpm. That would get you 2915 gpm at 20psi.

A note on flow tests: you can develop a multi point curve taking tests at different flow rates to better understand the flow from the supply. It's possible to have a water supply that is not a straight line on an N^1.85 graph. We use the straight line because its easy and often we only get one test or one point in the system. As others have said test more.

Why is Fire Pump Pulling the Underground Dry?

Leave a Reply.

ALL-ACCESS

SUBSCRIBE

COMMUNITY

YOUR POST

PE EXAM

FILTERS

ARCHIVES

PE PREP SERIES