Domestic water loop feed scheme

[rdeagan]

On a number of industrial projects, I have laid out the aboveground domestic water serving plant restrooms, cooling towers, etc. with a grid, or loop, feed. The loop feed allows us to minimize the pipe sizes and velocities while meeting relatively large demands. The loop feed is of course common on public water mains installed underground.

On a recent project, we had an inordinate amount of lateral pipe movement (swaying side to side) in the system when any demand occurred (it seemed the smaller the demand, the more movement). Trying everything we could think of, we closed an isolation valve, effectively cutting the loop in two, and the movement stopped.

This same layout has been used on quite a few other projects with no observable problems.

Does anyone have experience with this phenomenon or does anyone have a theory as to why this would occur?

Thanks!
David

 

[LittleInch]

What was the feed mechanism into the loop.

what you describe is some sort of surging. I can only surmise that the system is fed by a pump where start /stop is triggered by high / low pressure switches. given that the outlet would not be in the middle point, the pressure surges would be received at different times and could therefore set up a pulsing motion.

sounds a bit odd and maybe all it needs is a bladder arrangement or a bigger one to prevent these pulses or a smaller pump for low flow. equally a few NRVs in the system might stop it.

Only you can see your system so any more data would be good.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[rdeagan]

Thanks for your response. The loop is fed with street pressure. I'll respond with more details - and a sketch if this forum allows them - a little later today.

David

 

[LittleInch]

The best way is to scan it and then go to the bottom of the posting box and click on "Click here to upload your file..."

It could just be a specific response and excitation frequency for your particular system.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[rdeagan]

Thanks for your responses, LittleInch. I've attached a drawing that shows the general arrangement of the piping. The branches to restrooms are oversimplified - there are probably a dozen or so such branches, many to fixtures like emergency showers.

The piping is hung with clevis hangers with about 10 feet of all-thread rod per hanger. The movement, which appears to be primarily side-to-side, is estimated at about 6 to 8" in each direction when it occurs. There is no lateral bracing.

As I posted earlier, the movement appeared to stop when the isolation valve within the loop itself was closed.

Thanks again!

David

 

[LittleInch]

I think the key is that prv. I suspect, but can't tell from here, that this valve is hunting at low flows and pulsing flow. As this runs around the system it generates a frequency which is close to the natural frequency of the hanging pipe. When you close the isolation, this alternating flow is reduced or eliminated.

If you can add a sensitive pressure guage you might be able to see if this is the root cause.

Either add a pressure bladder or replace/repair the prv or brace the pipe.

Let us know how it goes.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[rdeagan]

You're very perceptive! We brought in an outside consultant who came to the same conclusion (that the PRV was the problem, but he didn't offer any solutions), but I was convinced that there was an inherent problem with a loop feed and that closing the valve was the key.

We'll try the pressure bladder first since it's the least expensive alternative.

Thanks again and I'll let you know if that solves it.

David

 

[LittleInch]

Hmmm, defining the probable issue is only a quarter of a job....

From a distance You have a few options -
a smaller PRV,
two smaller PRVs with one set to a pressure say 10psi below the other to open on full flow demand only
a bladder arrangement to smooth out the probable pressure pulses from the PRV
brace the pipe
close the valve
service the PRV

I would check the PRV is actually doing what we think before you change anything too much.

Let us know what happens- feedback is always welcome.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[hydrae]

I had a similar problem when pressure testing a new underground water main, We did the whole project as a loop which followed up a hill and back down. Got the system under pressure and the gauge would swing back and forth, closed one valve and the cycle stopped. I believe there was still a small amount of air mixed with the water as with any startup.
In water systems, a 4" PRV would not be installed by itself unless the flows were known to be high enough all the time to be in the proper flow range of the valve. I would add a 3/4" direct acting PRV around the 4" unit. Lower the set point of the 4" by 5 psi below the set point of the 3/4". Is the 4" PRV direct acting or pilot operated?, in either case under low flows the valve barely cracks open which can vibrate and hum trying to allow the right amount of water into the system.
I would be sure to install the small PRV upstream and downstream of the maintenance isolation valves of the 4", that way either valve can be maintained without disruption of service.
I would avoid the bladder tank, PRV's provide a steady pressure such that the water inside the bladder tank gets really old, they are designed to work with a pump which cycles.

Hydrae

 

[hydrae]

PS A 3" PRV would be installed on a 4" pipe. There are some valves that are internally reduced such as the Cla Series 60 models

Hydrae

 

[rdeagan]

Hydae: the valve we installed is a Watts 115 ACV.

link [URL unfurl="true"]http://www.watts.com/pages/_products_details.asp?pid=2820
]Link[/url]

I thought the bypass was for lower flows, but looking at it again, it appears to be just for regulating the main valve.

Sorry for the poor quality of the link. Haven't figured out how to insert a hyperlink in this forum.

David

 

[hydrae]

Correct, that is a diaphragm operated pilot controlled pressure reducing valve.
If you look on page 3 of the spec sheet, it says minimum continuous flow 50 gpm, so anywhere below 50 gpm the valve has trouble regulating, it will also wear out certain parts faster at the low flows
do you know what is your expected maximum flow is?
Is it less than 210 gpm? if it is, you could just replace that valve with a 2" which has a minimum flow of 1 gpm but, you may have trouble returning the 4" for credit, in which case installing the 1" bypass direct acting PRV would be the choice. (I would go for the 1" over the 3/4" listed earlier with the min flow at 50 gpm, other brands have lower min flows)
The fix could even be done without taking the system out of service, couple of saddles with live taps and plumb up the new bypass.

fyi The control loop inside the valve will flow about 1 gpm through that 3/8 tubing

Hydrae

 

[rdeagan]

That makes sense. As for a maximum flow, it would exceed 200 gpm with the process loads, so I believe I'll leave the 4" in place and go with the 1" PRV bypass to handle low flows.

Thanks so much. I'll let you know how it comes out.

David

 

[PEDARRIN2]

I would talk to your Watts rep.

They do this all the time with two PRVs - one to handle lower flows and one to handle larger flows. They set the open/close settings to overlap somewhat so there is a smooth transition at the intermediate flow. They can assist in making sure you do not oversize the valves. Just because you have a 4" pipe, does not mean you need a 4" PRV.

 

[rdeagan]

I agree PEDarrin2, but in my defense, the Watts rep came out and went over the system with a fine-toothed comb and couldn't figure it out. We even changed out the PRV, thinking it might be bad. No luck.

In retrospect, I should have used a two-PRV arrangement, with more care taken in sizing the two.

Thanks for your reply.

 

[PEDARRIN2]

Sizing these is not an easy thing to do, unless you do it a lot.

Your Watts Rep should have known better, if he didn't catch this issue. His PRV was feeding a toilet room so it could be flowing 0.5 gpm for the lavatories or ~50 gpm, if two toilets flushed simultaneously. A 3" or 4" valve has a minimum flow of 15-16 gpm. He should have seen that problem right away. The valve was barely cracking open to let a lavatory flow.

 

[rdeagan]

You're right PEDarrin. BTW, I found a similar Watts valve that has a bypass sized to handle the flow too small for the main valve. For a 4" PRV, the bypass is 3/4". Spec sheet attached.

David

 

[PEDARRIN2]

3/4" seems too small. The flow/pressure requirements between the two may not overlap enough for efficient operation at the borderline flows.

When I have specified this type of arrangement, the valves are similar models, but different sizes. I would probably go with a 1.25" valve on the lesser flow path.

 

[LittleInch]

When you've swapped it out, let us know if it solves the problem.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.