Pump outlet at same level as inlet, piping at elevated level 1

[sbnz]

We have to select a pump for transferring (lean) ash slurry from one pit to another, ~200m apart. The piping configuration requires that the pump lifts the fluid from the underground pit to over 10m height, flows through the pipeline at slight slope and then drops to the other pit. The final discharge in pit 2 is at about the same elevation as the pump CL.

From the pressure drop calculation for the overall system, the total pressure required to be developed by the pump practically is the line losses, as the difference in static head is 0. However, if I select a pump to meet the system resitance at nominal flow, it will clearly be incorrect as the pump may not develope the initial start-up head required to lift the fluid to the highest part of the piping and set the flow. What is the correct technique to specify/ select the pump in such situation? Should we consider a notional initial flow rate and ensure the pump can develope the head required to lift the fluid to the 10m level?

Must be a very elementary problem for experts in pump selection - would appreciate explaining in simple terms the confusion I am having?

thanks in advance,

 

[itdepends]

Depending on the pump you select- some may be able to self prime without adding any water, some will be able to if they have some water in the pumping bowl- others won't prime at all and you'll have to fill the pump (and possibly the suction line)to prime them.

That's not your biggest problem- your biggest problem is that you'll never get a pump to lift 10m. You can only lift by creating a vacuum. 10m of water = approx 1 atmosphere. i.e. if you create a perfect vacuum in the suction line you might- just - be able to lift 10m. In practical terms- you'll never manage it (even if you prime the pump first).

I'm assuming that it's >10m from the pump to the surface of the fluid in the ash pit.

Basically- for that sort of application you need to either get the pump closer to the liquid surface (by moving either) or use a submersible pump.

 

[sbnz]

Extremely sorry itdepends, the highest lift is 7.8m from pump CL. I don't know why did I casually mention 10m in the original post - my apologies.

My problem is selecting the pump which will have to operate at a much higher start-up head to lift upto this 7.8m, but once flow commences, the head drops as the final delivery point is same as pump CL.

I guess the pump is to be selected for a maximum head (to lift upto 7.8m) for a notional flow and then operate at the nominated duty point which is for the full flow and reduced head (for line friction loss only.

The problem is like pumping over a mountain and then bringing back to ground level again - what is the design philosophy?

regards,

 

[Artisi]

a little confused -
the suction lift is 7.8m, ie product level below the pump centreline - yes!

what is the highest point in the discharge line in relation to the pump c/line?

What is the height difference between the product level and the discharge point?

 

[katmar]

sbnz, usually (but not always) when the term "lift" is used with pumps it refers to the height that the fluid has to be raised from the source to the pump's inlet. The way I read your post is that you have a flooded suction, but you want to "lift" the fluid to 7,8m on the discharge side of the pump and your question is whether you can regain this 7,8m of head as the fluid comes down from that height into the discharge pit.

If my interpretation is correct then you must include the 7,8m head in the pump spec as the pump will have to overcome that head at startup. Theoretically you could recover some of that 7,8m head as the slurry flows down the last vertical leg, but my advice would be to disregard that recovery unless you have a massive flowrate and the pumping power is significant.

In order to recover that head the pipe size for the final leg has to be carefully selected to ensure that the pipe runs full, but that the friction losses are not too high. It is difficult to get this exactly right for even one specific flowrate, and if your flowrate can vary you have an impossible situation.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[sbnz]

The relative levels are as under:

Suction end level = RL (-)1.5m
Pump Cl = RL 0.0 m
1st elbow in discharge pipe = RL +6.3m
2nd elbow in discharge pipe = RL +6.3m
Discharge pipe end level = RL (-)1.5m

Thanks anyway and do not bother if it the configuration is still not clear.

 

[sbnz]

katmar - your interpretation is correct (almost, except it is not flooded suction though).

Your response is also clear. I realize that unless we get the syphon effect, the head may not be recovered in the discharge leg.

And we have all real life variations, the flow rate will not be constant etc etc.

We were developing the system resistance curve for various flow rates, using a software which analyses the pressure drop from the model geometry. When we model the complete geometry, it invariably considers the gain in static head and the total pressure drop merely becomes the line loss (velocity head constant throughout the constant pipe dia).

From what you said, it appears the logical modelling will be to leave the last discharge leg and calculate the total head required to deliver upto the elbow upstream of the discharge leg. (It means, however, the friction losses in the discharge leg will be ignored too.

I was originally going by the concept that once the flow commences, the 7.8m static head gain will indeed be recovered and the pump will actually operate at a lower duty point corresponding to full flow rate and low head (only to overcome the friction). However, the pump characteristic curve should cover the initial 7.8m start-up static head (plus the line losses), albeit for a lower flow rate. That is so long as the head developed by the pump for 0 flow is say, 10-12m and for the duty flow it is the head dictated by the system resistance, the system will work. You seem to disagree with that, and I can see your reasoning.
But tell me something, could it also be danegrous to select the pump for the maximum static head which means that if we do recover full or some of the static head in the discharge leg, the duty point will be thrown off the pump curve totally?

thanks heaps for the response.

 

[katmar]

sbnz, yes it certainly is possible to get the pump to run off its curve if there is some head recovery in the final down leg.

A way to ensure that this does not happen is to oversize the final leg (one pipe size should do it) so that it does not run full. To be even more sure that there is no recovery, use the larger size and instead of having an elbow at the end of the horizontal section at the top of this leg put a tee with a short open standpipe (say 2m) to guarantee that the pressure at the end of the horizontal section is atmospheric.

Regarding your other concern of ignoring the friction losses in the final down leg - this is usually fine because if the pipe runs full and there is some friction loss it will be less than the head recovery from the syphon effect.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[sbnz]

The idea of an open standpipe sounds excellent - will think a bit more about the option and might adopt in the design.

I agree with everything else you mentioned. Thanks so much for having such a fruitful discussion - much appreciated indeed.

 

[cb92]

Your pump has to move ash slurry (what's it's viscosity? is it like oil or like sludge?). It looks like you need a sewage lift pump to me. Make sure you design a good way to pull and service it. Consider double pumps with an alternating system, it all comes in a nice package. Just don't use the metal rails and plastic sliding pump connection, I've had that go bad on me. I would just as soon use a flange and a good strong wire reinforced hose.

 

[sbnz]

The slurry is very lean actually. Typically the solid (furnace bottom ash) addition in the slurry is less than 10% by weight - the slorry density is in the order of 1.05 t/m3.

The pump will, nevertheless, have the usual features for smooth operation and maintenance.

Thanks cb92.

 

[Artisi]

Another point you need to address besides the hydraulic considertion is that a 10% slury of furnace ash could be very abrasive so material choice for the pump is important as is operating in the right area on the pump curve.

 

[stanier]

I would suggest you look at an open box and gravitate to the last sump. Ash slurry tends to be at elevated temperatures and the "mist" carries with it particles of fly ash. These tend to solidify on the inside of vents and block them.

The benefit of the open box is that it can be observed and cleaned if necessary. It is does block it will overflow for all to see.

The gravity leg will be subjected to free surface flows and the velocity will be very high. The abrasion occurring will be extreme.

Can you not direction drill from the pump discharge to the receiving sump and overcome the challenge?

Another alternative is to add a discharge valve at the end of the pipe to create a friction head and make a selection of slurry pump in its range? I know this will be a source of constant maintenance as the wear will be high.