Bypassing a Lot to Move a Little 2

[PuffJr]

I plan to use a centrifugal pump to mix the contents of a storage tank by recycling 590 gpm. I am also considering using this same pump to deliver 40 gpm to a process. To enable this, I would bypass 560 gpm back into the tank. This approach would save a pump purchase, some piping, and an electrical install. Besides consuming more energy than necessary to move 40 gpm, are there obvious disadvantages to what I propose?

 

[BigInch]

Sounds OK at first glance, but the arrangement would depend more on the head required to do both and also on if you mix constantly or flow to the process constantly.

Obviously, if you're only mixing for a short time per day and your process needs a constant flow, you'll be way oversized with the pump. Pumping all day recycling to the tank, just to get a small process stream pressure may overmix your product or heat it up, or heat up the pump. In any case, it will probably waste too much energy.

If you have to mix all the time, and if similar heads are required for both, just offload a bit to your process through a valve that you can regulate pressure and flows to meet your process condions.

If there is a timing mismatch, ie process is continuous, mixing is relatively short time, I really think you might be trying to do too much with one pump and your systems would be better served using two task specific pumps.

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[PuffJr]

BigInch, I think that you are right on with the heating aspect. I thought of that as soon as I posted, and I don't think that I want to go there.

I'm not sure I understand your point about the head being similar for the two processes. The flows I mentioned are based on hydraulic analysis of the proposed piping system. The tank is atmospheric, and so is the other process. I calculated the flows assuming equal pressure drop along the two pipelines, analogous to parallel resistances across the same voltage drop. I calculated the flow through each pipeline for various head values, then summed the flows and plotted the result against head on the pump curve. The flows I cited occur where my system curve intersects the pump curve. Any issue with my analysis?

 

[BigInch]

Its just much easier for both lines if the pressure drop across any offtake isn't too much, so if your process was a hi press, you'd need a hi head pump to go there, but only minimal to get into the atmos tank and you'd have a big pressure cut to do that, or visa versa. If the head is about the same for both lines, no big pressure cuts to either side and no problems either. Just set up something to control the split, because if its lo press heads, as the tank fills up, it might send more flow to your process without some form of controls (valve etc.)on one or both of the lines.

With a low head pump, probably not too much heating will ever happen, so perhaps not to worry about that either, but if you were running hi head, low flow to one side or the other, that could be an issue. If its low head, probably not to worry about it, as long as you keep the flows above 20% BEP, which you should want to be doing anyway (at least 50% I'd hope).

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[BigInch]

The only problem would come if you had to run the pump all the time just to get into your process when you didn't have to mix, as that might be a big waste of energy that some bean counter might hang on you (or hang you) later. And a VFD isn't the solution, as 40 gpm is way too low below BEP and you'd most likely lose all of the even low head you have now running at such a low speed. That's why I said 2 pumps might be better (if you don't need to mix all (or most) of the time, but do need to flow into the process all (or most) of the time.

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[SteveWag]

The prior comments are all well founded. BUT, I have seen this done a lot in domestic sewage service and never does any good result. If this is sewage service, expect to either have short pump life or pay a large amount for a wet end. With repumping the same piece of grit over and over or the same string, expect the grit to wear and the string to reweave and clog.
Steve

 

[nationalpump]

Assuming this is water service and the pump runs constantly and the right controls are installed (i.e. valving etc.) would this set-up be a problem? As long as there is enough head right?

 

[BigInch]

If done right, it will work, just might be expensive to run.

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[PuffJr]

nationalpump, hydraulically, I think it works fine. Energy usage and product heating issues are nudging me toward two dedicated pumps, though. In the one pump concept, I know that the pump would recycle while the other process sits idle quite a bit. I see that as a big waste of energy, and the resultant heating might be considerable, especially when the tank approaches empty.

A 2 hp pump for the 40 gpm process will only cost me about $1100.

 

[BigInch]

You did the right thing.

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[rcooper]

I looked at something similar a while ago. It was an automated system, and I needed the 40gpm (I'll use your numbers for convenience) flow to be accurate, so I needed a modulating actuated valve and flow meter on the line into the process. By the time I had added an actuated valve, flow meter and all the necessary controls, the installation costs were higher than a fixed speed progressive cavity pump, which is what I ended up installing.