Pum System Curves - i need advice here!!

[MDStanleyPE]

We are working on a new elevated storage tank and pump station for a local municipality to be built on an undeveloped tract of land. We have determined the capacity of the tank to be 1 MG with a height of 200' to the TCL. We will be tying in to the existing water system and running the new yard piping to and from said system to our new tank and pump station. We need to have a small package pump station to fill the tank which is to be located about 50 feet away from the tank on the same property. i reiterate, the pump's sole purpose is to fill the tank, not to pump into the system directly. My dilemma is in generating the system curves...my manager and I are having a difference of opinion and he wants me to sign the plans.

We have modeled the hydraulic network, yet he is having our EIT generate crude system curves by an application of the energy equation. He has basically told her that the pump is only pumping against the static head from the pump station up to the TCL and the friction losses in the discharge piping ONLY. He is basically disregarding the suction side.

So far he has told the EIT the following:
- since we are tying into a water network, suction friction head loss is zero and all minor losses are negligible;
- take the energy equation between the new package station and the TCL of the new tank and disregard the 4 miles of suction pipeline losses and static head contributed by any upstream tanks . there are also various interconnects in the suction line which will contribute and take away head, but these would not be revealed in our "analysis".
- he wants her to solve for a single pressure at the pump station by plugging in a flow rate and solving the energy equation for p1 (i think there is too many unknowns since we do not know the head the pumps must pump against either)...this part is really bothering me. This is how he wants her to find the NPSHa.
- He basically does not want to select the pumps using system curves, just by using a single head-flow value determined from his rinky-dink energy equation application. (THIS IS KILLING ME TO LISTEN TO!!)

Now he has more experience than me and has worked on several pump-tank related projects. i have done one project where i selected pumps and have been involved as a sub-consultant where pumps were selected and EVERY TIME there has been a set of system curves over a range of static head values and friction values. Am i wrong or does his approach sound extremely questionable???? to me it sounds very back-o-the-napkin.

Thank you!!
Miles S.

 

[bimr]

Sounds questionable.

What I understand (a sketch would be helpful) is that you have a pump station 4 miles away which is pumping at an unknown flow and pressure from an unknown elevation.

You are adding a booster pump to pump into another water system where the discharge static head requirement is known.

With what you have presented, there is no method that enables you to select a pump that will be operating near the BEP.

 

[MDStanleyPE]

See attached for sketch. My take on the energy equation is in blue and my manager's is in green.

bear in mind, the 4 miles of pipeline on the suction side is part of a water distribution network, it is not a single run of pipe. Hence my apprehension in using this method rather than modeling it.

i hope this clarifies.

Thanks!!

 

[MDStanleyPE]

Oh, there is no pump station at the upstream EST, the water is just being fed by the pressure head provided by the tank.

Thanks!

 

[bimr]

You will be taking a large risk with the napkin writings. If your pipes were sized for 5 ft/sec, you would have a headloss from 190 to 420 ft across the 4 miles to your pump, depending on the pipe materials (which is unknown). And you only have 185 feet of available head. If you have a fire in the area, you will not be able to maintain pressure.

To properly design this system, you need to run a model with inputs of all of the piping materials, flows, etc. You also need to input potential fire flows.

You also need to design the system with pressure zones. With the system layout that you have described, you may have tank 2 overflowing during the night without some type of zone control of pressure.

 

[Compositepro]

The charge pumps should only be used at times of off-peak flow. The tanks are designed for supplying peak flow for a considerable time. If water is being drawn from the water towers while the charge pump is on, the water is simply recirculating through the pump. My guess is that the charging rate is a fraction of the supply piping capacity, and that is why your boss is taking that approach. But only you and he can know for sure.

 

[bimr]

That is not a normal practice for water systems. It is common to run a smaller supply pump for most of the day. It is not feasible to fill a water system in just the off-peak hours. The equipment to operate off-peak would have to be oversized and would not be economical.

When you pump into a water system with a supply pump, the excess water over the demand causes the water level in the storage tank to rise. The drawing is showing a 40' daily operational storage. The operational storage does not include the fire water storage. Note that a significant amount of the stored water is dedicated for fire flow.

The poster is in effect transferring water from one pressure zone to another pressure zone. He needs to set up a control system to control the pressure in the zones. One would try to install the water towers at the same elevation. If that is not possible because of the terrain, different pressure zones are necessary.

Otherwise, you will have the lowest level water storage tanks overflowing, or the highest level water storage tanks can not be completely filled to capacity.

 

[LittleInch]

As ever there is a mix of things going on here.

The system curve can be looked at separately from the pump discharge to the tank as you have stated originally.

The issue now is that the pump curve will have several curves depending on what you take as the inlet head into the pump.

I think you need two extremes - one at minimum flow with no other users all the way from your U/S tank.
The other full flow into the tank with all other outflows you anticipate.

The energy equation thing doesn't work for me, there are too many simplifications and assumptions going on to make it any use.

If you're doing this to find out what the inlet pressure into the pump is without looking at the whole U/S system then you might as well just guess, especially if you don't have a variable flow element or a decent control system. Far too many things going on to do a "simple" equation like this. Some idea of operational controls would be good as I assume this system has much more than you've sketched? - e.g. level controls/transmitters, control valves, flow meters etc

PS, Please try to avoid abbreviations without spelling it out at least once. I don't really know what MG (million gallons?) TCL and EIT are. Maybe everyone else does, but I don't.



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