Pressure Boosting Problems

re-size the pumps. a good pump manufactorer will get you in the ball park. you'll need to know rise, needed volumes and needed pressures at each pump stage.

All of this was provided when the pump designer manuf. the skid. Pumps are:

600gpm

187TDH

What I am wondering is whether or not the demand was miscalculated or if there is a problem with pipe sizing upstream, that causes too much of a pressure drop for the pumps to handle. As I stated, there are over 600 prison cells being fed off of a 5" pipe.

I ran some numbers using Hunters Curve and calculated a potential demand of over 700gpm. But my underdstanding is that Hunters Curve is meant for more intermittent use, than we have seen in a correctional setting.

Maybe I didn't catch all the details, and at the risk of sounding like an idiot, I'm going to say this anyway.

You said you have 10 stories and each story is 20 feet in height right? That's 200' of liquid head that you have to overcome (I guess maybe 180-190 since you don't pump to the top of the last floor). Without knowing the suction pressure of the pump, it sounds like you're probably undersized.

I had read the initial description differently. From the 4th floor there are risers to service the higher floors. These risers are 20' high each.

[At least that's how I saw it.]

Sorry for the con fusion. The actual height of each floor is approx. twenty feet, so the risers are about 100 feet high each

You are correct on the total developed head. Overall it's 187 feet from base to top outlet.

The suction pressure at the pump is city pressure at approx. 60psi.

Hi & welcome to CR4 (noticed the Post#1). Hope you can find assistance here.

If the problem is water hammer, there are ways to manage it.

My understanding of water hammer is that it is the change in momentum of the water. ie the moving water suddenly stops, or the stopped water suddenly starts. In either case, the sudden change in energy, is dissipated in the pipework as water hammer, either as pipework movement or a pressure wave bouncing up and down the line. Both effects are very serious and can cause real damage (pressure spike can cause line failures even).

In my experience the key to this is managing the change in momentum of the water.

Firstly the velocity of the water has big impact. If the velocity is greater than 2.5 m/sec, water hammer problems will be likely. At flows less than this much less of a problem. Can you checkthe velocity of water in each section of the network?

Secondly the start stop speeds ofthe water is also critical.

The opening and closing speeds of isolations valves should be checked, depending on the velocity and the diameter will determine the opening and closing speed. slow down the speed to smooth out the change.

The ramp up and shutdown speeds ofthe pumps should be checked. DOL pumps are notorious for this. Consider using soft starters or VSD to control both the ramp up and shutdown of the pumps. I use a soft starter for pumps of 15kW or greater is to manage this problem, or even a VFD/VSD if it is justified by the control uses of the pump. For VSD's ramp up over 20-30 seconds, ramp down over no less than 15 seconds.

So recommend the following to track down the source of the problem;

- Checking the velocity in each section of your network.

- Check the opening and closing speeds of valves

- Check the ramp up and shutdown speeds of pumps.

Thanks you for your response. My understanding of water hammer is the same. What I a wondering is how velocity plays a role in hammer. I intend to try and measure the velocity of the water and have found a formula that should work:

v = 0.4085 q / d² (1)

where

v = velocity (ft/s)

q = volume flow (US gal. /min)

d = pipe inside diameter (inches)

The only thing I'm not sure of regarding the above calculation is....which pipe diameter do I use. If the main piping is 3" and I measure flow from a 3/4" outlet off the main, do I use the pipe diameter of 3" or 3/4"?????

As far as opening and closing speed of valves, the only fast acting valves are the diaphragms on the flush valves. These types of valves are not rated to open based on time, but rather on consumption, for instance, a flush valve will open until it has reached 1.6 gallons.

The ramp up and down speeds are controlled through the VFD, but being that the building has such a large quantity of high demand fixtures, with sporadic use, I think that by the time the transducer senses the pressure drop, the pump system is already being told to hurry up and boost the pressure, but since this pressure drop is related to a rapid closing flush valve, the pump boosts too late, and overpressurizes the system.

I hope i'm not confusing.

The only thing I'm not sure of regarding the above calculation is....which pipe diameter do I use..... 3" or 3/4"?????

I'm not sure about the formulae used to calculate but sure about which pipe diameter. Use the ID of the pipe size where want to know the velocity. If you want to know what is the velocity in 3" pipe, use ID of 3" pipe (from which pipe you are measuring the flow is immaterial). Do not forget to use correct units.

Thank you for the clarification. This should allow e to pull a reading from anywhere along a horizontal run and determine the velocity.

A soft-starter will do nothing for the step-change in pressure and flow when the pump starts. All a soft-start does is soften the impact on the local electrical system of the pump starting.

Sorry for second post, just read in alittle more detail.

So the water hammer was with the original 3 x 40 hp pumps (2 x duty, 1 x standby), and now you have replaced with 3 x 15 hp pumps (assuming all are running).

Even with 3 of the new pumps running, you have replaced 2 x 40 hp = 80 hp, with 3 x 15 hp = 45 hp (almost half as much energy - is this correct?).

You mentioned the new pumps are on VSD, what pump speed are they running at when they trip? Or is it just one pump in VSD and 2 x DOL (that come in and out based on pressure setpoints)?

Before the pumps trip are they making the required pressure? Then does the pressure drop rapidly when one of the pumps drops out ? If so consider a Soft Starter to ramp up and shutdown the pump in a smoother way.

Or are the pump stuggling to make the pressure?

Sorry more questions than answers?

Actually, when we started the new triplex system...one single 15hp pump sustained the building pressure at 145psi. This pump was running at a full 60hz and pulling about 20amps. Then, as needed, a second pump would start....run for a little while and full 60hz and then time out. What we noticed though, was that the pressure wold not stay consistent. The pressure would go up to 145psi, then drop to 120psi, a second pump would kick on and spike the pressure up to 160psi and then the second pup would drop off.

Do/did you have air release valves at the high points?

Your cause might be a case of organized "crime" by inmates organizing to flush toilets at the same time and thereby causing an overload that possibly resulted in privileges or diversions.

LOL. Inmates actually organize flushing at the same time as a means of passing contraband through the sewer system. And they flush and flush and flush, until the cell or cells below them can retrieve the goods fro the sewer pipe.

As far as air relief valves on the high points....we have not installed any, but have been discussing this option with the facility.

That is illegal water use. Rather introduce a courier service and save water.

Welcome to the forum.

Do you have any pressure accumulator vessels in the system? (Normally near the pumps)

These may help your issue with the standby pump. Our water reticulation system uses this principal in many locations, even with VSD.

Is this not a triggered resonance problem? I like the idea of 10 gallon diaphragm-style water well accumulators, maybe at the ends of the lines, plumbed big. They'd be capacitors in an electrical power supply. Those toilet valves are not soft stop.

I searched the entire piping network and found no pressure accumulation vessel. In the U.S. we call these expansion tanks or hydropnuematic tanks (for wells). I think one of these will definitely help compensate the surges in pressure and stabalize the system.

Still have the issue regarding the low system shut-down.