Storm pump discharge piping banging

[acdecal]

We have installed 30 HP storm pumps with about 500 lineal feet of 6" discharge piping. A check valve is installed in the pit near the pump discharge. The pump lifts the water approx 40 feet along the route (inside and outside of the building) with quite a few offsets. The pipe is hung from a trestle and the building steel. The final point of piping discharge has a 10' drop into a pit. A soft start/stop was installed to control the pumps. When the pump stops the pipe bangs violently at some offsets about 160' upstream from the discharge point. This is also near the highpoint of the system.

There is no noise or movement when the pump starts. The check valve slamming after the pump stops could be causing a shock wave but there is no reaction anywhere near the beginning of the run. Maybe the tail end of the system trying to drain the final leg while air is trying to move upward is causing the problem. We are thinking of adding a air relief valve at the highpoint. Any thoughts or experience with this problem?
Thanks in advance.

 

[gerhardl]

I think you are on the right track of the problem.

You observe the banging at pump-stop (soft-stop included) only. Questions for detailed diagnosis (see causes below): One bang at one location or at several locations? One bang or several? If several damped in a sinus-curve lessening way?. Banging near top-point of system away from or near check-valves?

Banging in a system is (almost) always caused by water-hammering. The waterhammering is a sudden rush/stop of the water, resulting in a pressure peak in the fluid, with a changing of pressure direction.

My guess for possible causes (could be several in combination):

a) Pipeline from top-point down will act as a vacuum pump and create an underpressure. When normal air-pressure re-enters this will cause a larger reversing pressure. Pressure will be stopped at the checkvalve (and/or bends).

b)It might, but less likely, be caused by air pockets trapped in the system, and giving back-pressure when the pump stops. (More likely to occur at start)

c) The third possible cause is a wrong type/dimension and/or placement of the checkvalve. In this last case the checkvalve will close suddenly by and after the fluid has started flowing in the reverse direction, and will often be diagnosed by bangs at the valve in a repeating sinus-curve.

d) The checkvalve might be of a type reacting at the lesser fluid at soft-stopo flow by starting gulping and slam shut by reverting fluid stream between the gulps. (Valve is too large to be kept open at the lower flow, but is then more likely to also occur at start up, but not necessarily))

Suggestions for improvement, not giving priority to the single points.
1. Check piping layout and dimensioning, including pump and checkvalves.
2. Change checkvalve to soft-closing type (non-slam nozzle check valves)
3. Checkvalve, soft closing, at outlet because of large length of pipeline.
4. Air outlet/inlet valves at highpoints.
5. Securing/clamping etc of the pipeline to be improved?

 

[LittleInch]

Can you sketch a system diagram and profile for this line. I think if your system is open ended at the drain point then you will pull a vacuum once you stop pumping and all sorts of strange things will happen at that point.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[cvg]

without significantly more information, really hard to come to a conclusion. flow rate, velocity, pipe profile and type of hangers are a few nuggets that might shed some light on the problem.

as you indicate, if you have a pipe leg draining to the sump then an air valve (possibly a goose neck) at the high point might be useful

 

[acdecal]

See attached sketch. Pipe is open ended at discharge point. Hangers are U-Bolt clamps and clevis types. The piping is moving/banging approx. at the point between the 80' horizontal sections when the pump stops. Thanks for the input thus far.

 

[bimr]

You have left out some key data, in particular the pump output in gpm.

However, it seems that your pump system is designed poorly. The 6-Inch pipe for this application should have been designed for a velocity of approximately 5 ft/sec.

The 30 Hp pump will probably put out around 900 gpm which is around 10 ft/sec, which makes the design of the piping system more difficult.

You are getting a water hammer when your system shuts down.

Your options are:

1. Throttle the pump back as much as possible.

2. Install a smaller pump.

3. Install larger pipe.

4. Install a vacuum breaker at the top. Note that you can expect some leakage from this device at startup.

 

[cvg]

is the check valve non-slam type?
are the hangers free to swing such that the pipe hits the structure? appears you may need some restraint in both the x and y directions at the 10' offset.
does the soft start ramp down properly?

 

[acdecal]

OK here is what we did. Added a vacuum relief at the top of the 10" drop pipe. Seems to have resolved the problem. We are adding additional restraints but this looks like it did the trick. Our next stem would have been a silent check valve at pump discharge.

Thanks to all for the help.

 

[bimr]

Agree with cvg's comment. Too many simple pipe hangers in succession results in an unstable line when control valves operate and during pump start-up and shut-down.

Above ground piping systems may be designed as restrained or unrestrained. Selection of the design method is dependent on variables such as operating temperature, flow rates, pressures, and piping layout. The high velocity that you have pushes you into the restrained category.

The restrained system is referred to as an anchor and guided design. Anchors are employed to restrain axial movement as well as to provide vertical support. It will help if you install anchors in the middle of the horizontal runs, each time your pipe changes direction.

If you want to add supports, you should work that out with the structural engineer that designed the roof. Many of the roof designs are very economical and do not have an allowance for additional loads.

The least cost alternative is probably to just throttle the pump back.

 

[gerhardl]

acdcal: interesting problem. Seems that the problem is covered (partially) by my suggestion marked a): weight of fluid drawing down, air for pressure-equalizing pressing up, solved by letting air come in on top.

In addition, of course, the ever-present question: why wasn't it equipped according to the rules first time around, with an air inlet/outlet valve at top-point as normally prescribed. (Rethorical question, no answer needed )

 

[LittleInch]

That's certainly one big pump you've got there for an equivalent length of pipe of around 160m. I suspect you have a velocity close to 5 m/sec or more, but it would be interesting to know. Glad the vacuum valve solved it - sounds like the open end just kept moving and probably pulled a vacuum for a short while before the water rushed back and collapsed it - those sort of events can lead to some huge pressure spikes - your Victaulic joints did well to survive. Now air fills the gap but I think you've been lucky. You might not be so lucky next time so try and keep the velocity down to around 3m / 10 ft/ sec or lower to reduce this effect.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[acdecal]

Yes, big pump but that's what we were given to work with. A throttling valve is probably a good idea. Thanks

 

[chicopee]

Another thing that could have been done is to have the open drain outlet discharge into a partially filled pit with discharge waste so that when the pumping has terminated, no air would have rushed back into the piping by passing any need for an air relief valve. The same principle can be visualized with the inverted water bucket use to water chickens.; when the water supply within the trough at the bottom of the bucket is at the proper level, no air gets into the bucket needed to refill the trough eventhough the water supply in the bucket has a greater head than the water level in the trough.

 

[bimr]

Chicopee, don't think there is any concern about air in the pipe needing an air relief valve. The fluid velocity should be high enough such that if air was present, it would be flushed out. The poster has not mentioned any problems with the force of fluid entering the pit, but one would have to expect some splashing.

The problem seems to be that the momentum of the fluid when the pump shuts down is causing a vacuum situation to occur. Installation of a vacuum relief valve seems to have cured the problem.

The only question is how the poster obtained the vacuum relief valve so quickly.

 

[chicopee]

Bimr, so with the soft start/stop controller, was momentum still an issue?

 

[bimr]

The poster has not supplied all the details, but one may surmise what is occurring.

The soft/start on the pump will eliminate some of the water hammer, but because of the extreme flow rate, will not eliminate all of the water hammer. Surely, the situation would be worse without the soft start.

Here is a description:

As a further example of inertia, Fig. 3.1-a shows a pump discharge pipe. At a very small moment of inertia of pump and motor, the failing pump comes to a sudden standstill, which has the same effect as a suddenly closing gate valve, only this time on the downstream side of the gate valve. If mass inertia causes the fluid flow on the downstream side of the pump to collapse into separate columns, a cavity containing a mixture of water vapour and air coming out of solution will be formed. As the separate liquid columns subsequently move backward and recombine with a hammerlike impact, high pressures develop. The phenomenon is referred as liquid column separation" or macro-cavitation4.

http://www.ksb.com/linkableblob/ksb-pk/80892-193671/data/Druckstoss_Know-how_Band_1_en-data.pdf

 

[bimr]

Here is a better picture.

 

[hydrae]

I would suggest a fully detailed surge analysis, water hammer can do some major damage in both pressure and vacuum; particularly with the high velocities you have. The valves, endpoints, elbows, and offsets can create reflection points that can stack the pressure waves on top of each other which multiply the stresses. Sometimes these stacked waves will show up several minutes after pump shutdown. A rough rule of thumb is there is 50 psi pressure spike for 1 ft/sec change of velocity that happens quicker than the speed of sound in water over the length of the pipe, this assumes rigid pipe (cast iron). So for your 10ft/sec assumed velocity that is 500 psi which can be doubled or tripled.

Without the surge analysis, I would caution on adding restraints and supports, as making the pipe more rigid can change how the pressure waves propagate and increase the damage.

If macro cavitation is still forming, you will get destruction of the equipment over time even though the sounds have lessened.

The vacuum relief may have completely solved the problem, but without analysis you can never be sure.
Other solutions may be needed such as: a flywheel on the pump, multiple air relief/vacuum relief, surge anticipator valve and/or surge tank(s).

Hydrae

 

[rconner]

While grooves for grooved-end piping can be specified either "rigid" or flexible, in reality grooved end pipng systems may not really be quite as "rigid" as one joined by welding or traditional flanged joints. The original poster has also advised some or all of the piping is "hung" from a trestle.
While I believe flexibility can be very good in some applications or with good lateral support, and I have not seen enough sure details of flow or the pump etc. to know exactly what is going on here (with transients and/or system resonance etc.?), if we are talking about grooved end steel piping on e.g. simple rod hangers and with pretty high flow (with no effective lateral stability provided, as pipes strapped down etc.?), I wonder if this could also be a factor in why system is a little lively?