Pumps in Parallel - Duty Point and Power Consumption

Whether it be one centrifugal pump or several, the duty point is where the pump curve(s) and the system curve intersect.

If a pump is operating at zero flow, switch it off!

PWSlack is right, if your outlet manifold has more pressure than an additional pump can output, it needs to be shut down to protect the pump. This should be easy if you can install a flow detector downstream of every pump that will shut it down if a no flow condition occurs.

Even if you are getting a no flow condition, your pump will still be consuming electricity because it is doing work against an equal and opposite force; the pressure in the pipe pushes back just as hard as the pump is pushing. Work is still being done because the pump is creating pressure, just not flow. Problems arise because the energy needs to go somewhere, if the forces are equal and opposite, it is usually released as heat. Damage can result through cavitation or overheated parts (seals, bearings, motor).

Drew

Very true. Sewage would tend to render many of the flow detectors I am familiar with ineffective.

Drew

I suppose, God only knows why, that you speak of centrifugal pumps. Are you? Now, a running centrifugal with 0 Q (no discharge) is analogous to a pump running with it's discharge valve closed. This situation is well covered in most professional pumping textbooks I know of. As the pump is doing no work, the motor amperage is low and the pressure is low also, but the water that is in the volute cannot get out. It churns there rapidly and the pump temperature rises quickly. If the pump is not stopped in time it will be destroyed. As for measuring the power consumption in such a case, just measure your voltage, amperage and power factor and you got your watts. Piece'o cake, innit?

"I suppose, God only knows why, that you speak of centrifugal pumps"

That fact was revealed in the first sentence of the first post.

I agree with everything that's been said up to now.

Four equal pump-sets working next to each other in parallel will share the flow equally among them.

4 pumps will have to deliver more than 3 even if it just marginally so.

Your optimum usage will be the combination running at the highest efficiency.

This is only if they share an inlet manifold or tank. I believe the OP is using 4 pumps and 4 lift stations utilizing the same transfer line.

Drew

So I see now but :-

The pump does not pump but the sewage disappear?

Maybe the presumptions are incorrect.

Indeed "centrifugal" pump mentioned. Didn't notice. Sorry. No offence meant.

Happens all the time. No offence taken by anyone, I'm sure.

Guest

To satisfy my, your and our curiosity please supply some detailed information.

Pump rpm, curve, suction and delivery gauges. static hight , pipe sizes etc.

It is should be noted and well written down that you can only run pumps within the curve some where aorund BEP in order to insure long life of the pumps.If you run the pump with duty point towards left means to try to run pump near to its shut off.That mean less flow and high head your pump will induce heavy vibrations and pressuring your casing. So if you run the pump towards left of curve your pump will just run without pumping any more fluid

Now if you go with therotical calculation pump operating at Zero flow will require consume zero power This is Therotically correct.This means your pump requires no power to operate.But if you see pratical appilcation your driver which is not VFD will produce uniform power (You can refer to no-load condition curve provided from motor vendor) will produce X amount of power.

It is really impractical to ask if you want to run all pumps to end of curve in the end you will end up high flow and low head (P=Q*TDH*Den/36000) you will consume more power ..problems will occur and damaging the pumps.If you want to run the pump till extreme left of the curve please insure have sufficent NPSHr margin in order to insure that your pump doesnt lose prime at all times.

Every pump has a design capacity and head fixed for each size of impeller and casing design.you have to select a pump as per the requirements stating your requirements

Since the "challenged" pumping station is NOT overflowing, then the calculations must have some level of error- the pump is obviously meeting its design load.

If possible, measure the pressure at the point of each pump's connection to the main. You can also calculate the flow of each pump by turning off the pump for one minute during a normal waste outlet period and measuring the amount of water taken on during that time, then measure how long it takes the pump to bring the level down to the starting point. The answer can be used to calculate the pump's flow.

Since it is unlikely that there is constant sewage discharge to each of the four pumps, and since the pumps are likely controlled by some form of level switch, it is very likely that the station of concern is benefiting from the shut-off of one or more pumps when satisfied.

Obviously- if none of the stations is overflowing, then all of the pumps are capable of supporting their required waste stream volumes. Each may be pumping , say, 100 GPM while the average waste stream flow may only be 60 GPM. It is very possible that- if all pumps were actually pumping 100 GPM the resulting pressure would overload at least one of them (so the actual flow would only be 300 GPM) but is it also very likely that NONE of the pumps is actually operating at "rated" flow.

I would take some actual operating conditions (including watching average flow at each station when the pump is turned OFF) and use the average flows from all stations to determine likely pump operations (or at least the range of flows and pressures that will likely be observed). Since they are all supporting their respective loads- this exercise would be just to get a better understanding of potential operations in case one of the stations is ultimately, finally overloaded due to more than current loads.

Thanks guys for all the feedback, really appreciate it.

Drew K assumed correctly. The sumps and their pumps are situated at entirely different locations and serve different residential zones, but they all eventually share the same rising main up to the gravity manhole. See picture

http://pumpcurves.blogspot.com/2010/03/system-characteristics.html

Just for interest sake, the duty points were determined as prescribed in "PUMPS, principles and practice". This constitutes reducing each pump curve up to the connection point by subtracting friction and secondary losses as well as the required static lift from the pump curve. Then these "reduced curves" were added like you would with pumps in parallel. Where the system curve of the common transmission line intersects this combined curve, a horizontal line back to the "reduced curves" reveals the actual flow for each pump.

Hendrik here is a schematic drawing for the system http://pumpcurves.blogspot.com/2010/03/system-characteristics.html , as well as some info on the pumps. I provided an equation for each pump curve (dis makliker so)

And on the point of the sewage "disappearing", I think that it keeps on pumping until one of the other pumps shuts down, only then does it produce any flow (its sump dimensions are unknown, but it must be big enough). When looking at the pump curves it is evident that there is actually 2 of the 4 pumps that experiences this same problem.

I agree with DrewK and Loki that the pump needs to be shut down, the question is now, what is the cheapest and most effective way? Some kind of pressure switch that allows it to pump only when the pressure drops?

PWslack can you elaborate on your suggestion pls? will it not keep on tripping again and again? isn't that bad?

I have to admit that either the theory or my calcs might be flawed, but it might just be spot on (revealing highly ineffiecient and destructive operating conditions). I will have to follow energygd's advice and gain some information on the real operating conditions.

On the point of power consumtion, Loki's suggestion seems logical will give it a try.

Thanks guys

I think I'm beginning to see the light.

Having analysed several of these types of systems, we usually found that sewage pump stations had random periods of pumping. This usually led to windows being available for each pump station. A little statsitical analysis on the amount of hours each pump runs each day and the number of cycles will allow you to figure out the true odds of all four stations running at the same time.

Hopefully each pump has a check valve immediately down stream of the pump. A simple position sensor on the check valve swing arm and a timer will let you know if a particular pump station is producing no flow. As long as you have the storage, shutting the station down for 5-10 minutes will usually do the trick.

Alternately, you could place a pressure sensor on the forcemain (downstream of the check valve) at the station having the most trouble and delay start if you are reading pressures showing that the other stations are running.

Good day guest

If you feel like it go to http:/www.gwis.co.za and download a demo. You can get a good simulation and try a what if approach.

On the question of how to stop the pump suffering low flow.

Low flow equals low load on the motor. Simply trip the pump on low amperage.

Just calculate the amps required for about 30% (or whatever you choose) of BEP flow from the curve and set this as your trip.

This is a common (if not preferred) method of low flow protection.

The pressure sensor option has to be the best way to go. Set the pump start (hi level) sensor a little lower than it was to allow reserve overfill in case the other pumps are running and the "challenged" pump is inhibited.

Chas

This is exactly what I was going to suggest, but capblanc got there first.

This will work exactly like you want with minimal cost or system disruption. GA

Thanks alot guys, I think I'm good now.

I'll give it a go and let you know how it went.