Causes for Failed Backflow Forward-Flow Test?

9/29/2020

Recently had a job with a wet-pipe fire sprinkler system (smaller project) that has a 6-inch double check backflow. The area static and residual pressures are pretty consistent (had a flow test of ~40 psi static and ~38 psi residual at 1,300 gpm).

We ran a forward-flow test through the backflow preventer, needing a minimum flow of about 800 gpm, which includes the hose allowances for the project (no standpipes or interior hose connections).

The most flow we could get was about 450 gpm, even with three 2-1/2" hose outlets completely wide open. The static pressure following the test at the riser was about 40 psi. We used three 100 ft, 2-1/2" hoses with a 1-3/4" smooth nozzle and diffusers at the end of each hose, measuring about 3 psi pitot on each and full flow (which equates to roughly 450 gpm).

Do you have any tips or advice in why we were not able to get enough flow through the backflow, or where to start to troubleshoot?

Thanks in advance!

Submitted anonymously and posted for discussion. Discuss This | Submit Your Question | Subscribe

25 Comments

Dan Wilder

9/29/2020 08:39:45 am

First, if there are no interior hoses and if there are no hydrants between the backflow and the sprinkler system, there is no need for a hose stream allowance for this test (See NFPA 25-20' 13.7.2.1 & 12.7.2.2)

What was your residual with all the hose valves open (before & after the backflow)?

When was the backflow last inspected? What MFG & type for the backflow?

When were the gauges last calibrated? Are they specific to the testing equipment or just a standard gauge? Are you using any sort of remote reader (are the tubes clear and at the same elevation as the equipment during the test)?

What was the discharge with only 1 hose valve open? With 2?

You should see similar results from this forward flow test with the original flow test so my first guess is the backflow has an issue. Next step is to redo a flow test outside to verify current water supply conditions.

9/29/2020 08:47:42 am

All good points -

This test is for an acceptance test from NFPA 13, so I'm not positive that NFPA 25 applies here.

Pitot gauges were 2-3 psi on each of the hoses with all three open, 6 psi with only two hoses open, and 12 psi with one hose open.

Gauge calibration was done within the last 8 months and the certificates where checked. The pitot gauge was not fixed for the equipment, just a standard air/water gauge.

The backflow could have an issue. It's a brand new install with a new backflow preventer. That's the avenue we're pursuing now.

Colin Lusher

10/1/2020 01:07:36 pm

Any updated OP? Curious if you've figured out where the problem is.

10/1/2020 01:09:48 pm

Thanks Colin, not yet. The backflow had a specialist out and it was certified, so at this point we don't believe there's an issue with the backflow.

I've recommended use of playpipes with fixed certified pitot gauges, but the contractor doesn't have that onboard so they may need to rent some equipment to make that happen.

I'll post when we resolve it.

C.T. Carbonari

10/1/2020 02:08:37 pm

Please do post when resolved. Personally, I enjoy these MeyerFire daily posts and have learned a lot from the active community that responds to them.

With that said, this flow situation is interesting.

I'm going with "Key Valve @ Tap to City Supply not fully opened"

Without residual pressure readings from gauges at City Hydrants, BOR, and downstream of BFPA during flow it's all purely speculation from me.

Nick

9/29/2020 10:06:32 am

You need to take into account the losses through the 100' of 2.5" hose.

That is the issue.

9/29/2020 10:24:19 am

Yeah that's probably it. Maybe some elbows and playpipes right off the header would help.

Thanks for the tip Nick!

9/29/2020 10:49:02 am

Agree with Dan; no need to include hose flow through backflow if there will be no hose flow going through the backflow.

Agree with Nick, you're problem is likely the hoses. Get rid of the hoses and flow directly off the test header.

That said, you didn't indicate the residual pressure downstream of the backflow; this is the most important piece of information we need to make a final determination. If you have really low residual pressure downstream of your backflow during the test, then your backflow may be bad. However, if your residual pressure at the backflow is within the normal range, and you still have flow problems, then your problem is your hoses.

Keep in mind that you not only have friction loss through the hoses, but it's nearly impossible to avoid kinks, and even minor kinks can really choke down your flow through a hose. Get rid the hoses, and use a couple 45 eblows and aim your play pipes up in the air; the water will broadcast pretty wide and shouldn't cause any damage to surrounding landscape/structures.

Mike

9/29/2020 11:32:49 am

Isn't pitot flow pressure of the interior of the water stream flow? So hose stream friction loss is irrelevant. Pitot pressure measures the flow pressure. Friction loss of the hose would only be indicated on a pressure gauge perpendicular to the flow.

Also, maybe the underground tapping valve is partly closed.

Also, if you are using a Hose Monster they have a unique flow chart.

If the range of the gauge is too large, lower pressures could be false. Maybe try an oil fillef gauge to 10-25 psi where lower values could be more reliable.

Could also try a smaller nozzle to increase velocity through the tip.

Lastly, 800 seems within the range of a 6" device but might want to verify that on the data sheet.

My gut feeling is an obstruction of flow or false gauge readings.

9/29/2020 12:09:48 pm

Mike,

Friction loss through the hose is definitely not irrelevant. You're correct that the friction loss isn't measured on the Pitot gauge, but the losses through the hose have a MAJOR affect on the ability of water to flow through the hose, and therefore affects the stream pressure available at the end of the hose. As the water flows through the hose, it looses energy/pressure ("normal" pressure, perpendicular to the flow) due to the friction. When the flow leaves the end of the hose/pipe, all of the remaining "normal" (perpendicular) pressure is converted into velocity/hose stream pressure as there are no longer pipe walls to contain the stream. You are measuring the stream flow after it exists the end of the pipe, so any friction losses in the pipe translate to lower Pitot pressure at the exit of the pipe.

Conceptually, is there any question that flow directly from a hose valve will be much greater than the flow through a 100-ft hose connected to that very same valve, assuming all other variable remain the same?

9/29/2020 12:13:38 pm

Thanks Colin - I did go back and confirm that they were 50-ft hoses and not 100-ft, but the same principles apply. I'm wondering if we can't run a 90 and playpipes right off the test header like you had mentioned.

Ramon

9/29/2020 10:43:05 pm

So, how would you put back the friction loss from the hoses?

Is it ok by adding back that loss on the residual pressure?

9/30/2020 11:06:08 am

Ramon, I wouldn't recommend that. Because this is an "acceptance test", you can't really resolve the issue via theoretical calculation. Acceptance testing is meant to prove that the system functions in the real world, so if you're adjusting your results using physics theory, you're defeating the purpose of the testing.

Another note; while calculating the loss through a flexible hose is a simple formula in theory, a hose often has kinks and bends which may increase the friction loss, and may make those calculations inaccurate.

In theory, yes; you could use bernoulli's equation to balance energy within the hose, with the hose stream outside the hose. Just outside the end of the hose (where you'd take the pitot reading), the normal pressure energy and set to zero, and solve for the velocity pressure component.

9/30/2020 05:02:14 pm

Agreed. Maybe a better word to use was "negligible." 450 GPM/3 hoses = 150 GPM per hose. That friction loss seems quite low and rather incapable of restricting flow enough through a 2 1/2" hose.

Using the friction loss calculator from https://www.frictionlosscalculator.com/?hosesize=2.5&length=100&gpm=150

Loss was only 4.5 PSI

Maybe I am not understanding the question/comment of the OP.

Thanks for your comment.

9/30/2020 05:34:08 pm

Mike, yes, you're right; in theory it should be pretty low, but in practice, the bigger problem I've had with flowing through a long hose is that small kinks, twists, deformations, etc. in the hose (especially at lower pressures) can really restrict the flow and increase the losses over what the theoretical calcs show.

But based on the other new information here, it sounds like you're right, and the hoses may not be the problem; the true problem may be restricted waterway somewhere upstream. Can't really determine without the residual pressures above and below the backflow though.

Bob

9/29/2020 11:43:29 am

In addition to checking the tapping valve, you might check with the local water department. I sometimes ran into situations where part of the water grid was under repair or improvements affecting the system. I once ran into an issue where a water tower was down for maintenance which affected the supply a system was designed to. I also saw negative impacts from sectional valves closed due improvements in the nearby area.

9/29/2020 02:06:21 pm

(3) 2-1/2" Hoses should easily be able to achieve 750+ gpm flow from a 6" Underground Fireline even with a "low" pressure supply (~40 psi).

My gut instinct is that you are having restricted orifice issues most likely due to mechanicl blockage (a valve on the line not fully opened, damaged check valves in the Double Check). Here is what I would do to isolate the cause/location.

First - "Triple Check" testing equipment and methodology - Calibrated Gauges (that zero out), Proper Deviations, Proper psi pitot range readings (whether using hose monster or playpipe)

.Second - Install 3/4" x 0-6 nipple to ell and 1/4" bleed trees w/gauge on both the #1 and #4 Test Cocks of the Double Check Backflow Prevention Assembly (BFPA). If it is a Double Check Dector Assembly, then I would intall on the bypass Test Cocks and close the #1 SOV of the 3/4" Double Check.

Then I would run the flow test making note of the pressure drop accross the BFPA (should be same as the Mfg flow curve). Anything above a 10 psi drop @ rated flow or lower is indicator that the check valves are not working properly. Also, the #1 TC gauge will give you a clear indicator of your pressure loss due to friction from the "City Supply" to the BFPA inlet.

A gauge located near BOR (and where I'm assumining you are making your POC for the Hose Test Header) will indicate how much prssure loss due to friction your getting from downstream of BFPA to BOR.

I'm assuming that the DC is located in a vault near property line, but similar concept applied if it is inside the building.

The Pressure drop due to friction in the 2-1/2" hoses only becomes relevant if there is not enough psi to drive the flow @ POC. For 50' of 2-1/2" Hose, Pressure loss due to friction should be ~ 8psi range. (There should be at least 20 psi @ BOR available).

Finally, account for the losses due to elevation. That could make a difference if significant (I've assumed they're less than 5')

Knowing the exact pipe routing, configuration, size, lengths,material, and test equipment locations would significantly help to "dial in" addressing where you're losing psi.

Hope this helps and Good Luck!

9/29/2020 02:17:20 pm

It does, thank you! There is a backflow specialist that is taking a look at the backflow specifically today for any concerns/issues.

The backflow is vertically oriented, in the building, and is just above grade. It's a double check. There's no significant elevation differences on the site.

Jonathan Joseph

9/29/2020 02:12:20 pm

No need to add a hose allowance with no hose connections anywhere. I'm curious to what your sprinkler demand is at the base of the riser.

Your Flow Test does not add up. How do you have 40 static and 38 residual and able to flow 1300 GPM?

What was your pitot reading off the flow test?

9/29/2020 02:16:02 pm

Fair points; looked at running worst-case with the hose allowances since that's the way I read NFPA 13, but I see now there's mostly different opinion on that.

I flowed two 2-1/2" outlets on a hydrant and got a pitot reading of 16 psi with a 0.90 hydrant coefficient. There's a pretty sizeable grid network, but it's extremely flat with several water towers that aren't particularly tall.

9/29/2020 03:20:18 pm

Q=29.84(Cf)(ID)^2(P)^0.5 where Cf = 0.90 ; ID = 2.55 ; P = 16

Q= 698 gpm (per Outlet)

Q = 1396 gpm Total

It would seem that you're able to get the Volume Flow you require from the City Supply.

When you run your Interior flow test, having a gauge @ hydrant near POC of Fireline with City (at both hydrants on each side for a grid) and the gauge on the #1 TC of the BFPA - You will be able to determine how much pressure you're losing due to friction.

For 6" C900 DR 14 (thickest wall) @ Q = 750 gpm
Pressure drop will be ~1.64 psi per 100 ft.

I agree with you @ running worst case by adding hose allowance inside even though there is no physical application for it. The inside hose allocation (or elimination) ultimately always comes down to the AHJ (I've seen it go both ways regardless of the reality of the ability to connect inside)

In any regard, getting 750 gpm from (3) 2-1/2" Hoses should be easily achievable unless there is a flow restriction somewhere.

9/29/2020 02:40:02 pm

If you have 16 to the 1/2 power = 4
4x5.625x29.83 Q=671.175

James Kyle

10/3/2020 05:51:05 pm

I have found numerous backflows with foreign debris stuck inside keeping the 2nd check from fully opening as designed. And the Backflows pass the annual test with no problem. Isn't this exactly why we are installing Forward Flow Connections.

10/8/2020 12:11:36 pm

Thanks everyone for their input on this one - the verdict here was that the tap valve at the street was only partially-open.

The contractor had the water utility out to close and reopen the valve fully... once that was done we retested and even with each hose valve halfway open we had ~10 psi on each 1.75" pitot to get at least 900 gpm flowing through the backflow.

Thanks for the input and happy to share that the issue is resolved.

10/8/2020 08:36:08 pm

I'm happy to hear that it was a minor "boo boo" with a simple fix!

Causes for Failed Backflow Forward-Flow Test?

Leave a Reply.

ALL-ACCESS

SUBSCRIBE

COMMUNITY

YOUR POST

PE EXAM

FILTERS

ARCHIVES

PE PREP SERIES