2 Positive Pumps

I have a project on the horizon where I will be controlling the speed of 2 positive pumps in series. The first pump is a homogenizer and the second pump in series in the pipe is a booster pump to pump the product a long distance. The pumps are going to be controlled based on flow, with the obvious problem of needing to NEVER drop below the flash point on the inlet of the 2nd pump in series. This means we'll have to maintain about 40psi at the inlet.

I don't have any more details as of yet on this project. My concern and question is that my co-worker assures me that controlling flow rate with 2 positive pumps in series is extremely challenging. Can someone enlighten me on this? What is the difficult part, other that assuring you don't slow down the first pump too much and cavitate the 2nd pump?

I will be controlling the pumps with drives, my control options are wide open for whatever I'd like to use to control this flow loop......I'm sorry I don't have any further details as of yet on the project. I know the flow rate will need to be adjusted via the HMI I'm also integrating into the system......Can someone outline what the complexities of this type of control loop is?

TIA,

Greg

FIRST what are you pumping?
SECOND it is a little more compicated because in pump 1 your are mixing product A and Product B -- you said it homogonized which I assume (maybe incorrectly) as mixing.
THIRD Is this stuff flammable You used term flash point -- in my mind that means a flammable substance --
FOURTH What is temperature of the material. How are you going to keep it from vaporizing if you get above satuation temp and or below satuaration pressure. How are you going to control temperature going into the pumps?

With all thet the problem is simple ensure that the flow throughout the system is constant at all points. AND THAT IS NOT SIMPLE TO DO.

Recommendation
put a surge tank in between pump one and two.
Have overflow from surge tank to # 1 suction.
Run # 1 with VFD if you want at say 30 Hz
Run # 2 with a level sensor on surge tank such tht it speed is proprotional to level.
Put level switch in tank to prevent # 2 from running dry and to ensure #2 has NPSH.

Dan Bentler

Answers to your questions:

1.) The product will be pudding.

2.) Pump 1 is a homogenizer, but it's really just acting as a timing pump.....not really homogenizing anything, the second pump is a booster pump.

3.) This stuff is not flammable, but it's very hot and if pressure isn't maintained it will flash into steam.

4.) I haven't looked at the steam tables yet for this product, but yes, that is part of the problem the pressure will have to be maintained high enough between the two pumps to keep it from flashing off.

Surge tank is not an option here.....this is a quick and dirty means of getting product to a location it wasn't origionally intended to go....


Please ask more questions, and I'll answer them....

Thanks again,

Greg

Hmmmm
Food for thought??
BAD BAD joke?

Gotta think on this one

Dan

I agree with your co-worker, it will be very challenging. I've never did anything like it, but from other applications, the following comes to mind.
If the rate from pump 1 is greater than the rate from pump 2, you will overpressure the line and lift a relief, if pump 1 is less than pump 2, you will have flashing. So you will have to keep the flow rates from both pumps at or near the exact same rate. Very difficult.

The master loop will be the flow rate of pump 2, but what other instrumentation will you have?
I know that hygene is always an issue in the food industry, but would installing a pressurized bladder, between the pumps be possible? This could help when the two pumps are slightly out of sync.

I don't see how both pumps can be controlled on flow. How about putting a PT on the first pump discharge, and regulating the pressure. The second pump is controlled by flow, and the first maintains the suction pressure. You might still need a pulsation damper of some sort, even if it's only one of those crappy acceleration tubes they use with Cat pumps.

I would go along with Leigh's basic idea but I would use the pressure control on the second pump as I think this would provide a better response to process changes. If the pumps are positive displacement then flow is directly proportional to speed so you will only need to scale the speed setpoint for the first pump to give you the flow you require. The only difficult part I can see for your particular application is finding a suitable pressure sensor for the "pudding".

Hey thanks for the replys so far. That's exactly what we're doing, there is a PT after the first pump and the idea is to control pressure between the two pumps, and overall flowrate I think should be controlled by the second pump. A surge tank between the two pumps is out of the question, this is a sterile line so the surge tank would have to be an aseptic surge system.....thanks for the input thus far. Any other ideas?

AutomaticLeigh, I'm not familiar with the acceleration tubes you're talking about.....Can you clue me in a little on them?


Greg

Leigh's solution is the best.

I don't see how this will work perfectly with out some kind of surge tank.

It is not a problem keeping a relatively constant volume flow from the second stage pump. It is not a problem pacing the second stage pump to match the first stage, and "trimming" the using a pressure control. The problem is going to be with any tuning issues on the pressure control.

If you just have a momentary low speed on the first step because of hunting or response delays, the pressure will drop in the pipe between the two pumps, because the fluid is nearly incompressible. That means you will have a momentary "flash" or steam generation. It will not cause an explosion, because it will still be at the same pressure. A soon as the first pump catches up and the pressure increases the steam bubbles will collapse.

I expect it will cause erosion of the piping, similar to cavitation on a pump impeller.

Tuning this loop for fast respone, stability, and no hunting is going to be a challenge.

I can see two things that might help you. First, keep a big margin of extra pressure in the second pump inlet, so that pressure fluctuations will keep the minimum pressure above the vaporization point. Second, frequently inspect the intermediate piping for signs of erosion, and plan to replace it at some interval. You could possibly use a lined pipe to facilitate this.

Thanks for your response Tom, I was hoping you'd respond, with your pumping experience. Thanks to everyone else so far also.


I have to think about this one a bit.....A surge tank is tough, as I said before. We're not a retorting plant, we're aseptic. This means the product is sterile and the container is sterile when the 2 are put together. This is in contrast to retorting where you put 2 unsterile things together, say beans and a can, then heat or retort them. Having a surge tank means it needs to be aseptic, which isn't a problem, we have them on all of our processing lines, but it's not practical in this very small application. The cost of an aseptic surge tank is 10 times the cost of this entire retrofit...

This is a 2" line between the two pumps, about 100 feet or so between the pumps, but as Tom said with virtually no compression the pumps might has well be 2 feet apart.

I wish I had some way of modeling this better than in my head...


Greg

Greg
I'm not a pipework guy, but I think the acceleration tube is basically a short length of flexible tube, used on the suction side as a pulsation damper. The ones I have seen looked suspiciously like a bit of lay-flat hose (with a nice cover to justify the price). I guess if the feed pressure is OK, the change in volume as the tube collapses and inflates gives the damping effect.

Leigh, that seems like just the ticket. If Greg can get his tuning close that should provide the error margin he would need.

OK so we hve food processing FDA USDA or whomever setting purity and sanitation criteria.

I still think he is going to have to have some "slop factor" in between pumps 1 and 2. Maybe instead of surge tank let us think "accumulator". Could be a tank or a line T'd off line between 1 and 2 then pressurized witth nitrogen to #2 NPSH plus 10%. Set it up to provide easy access for cleaning.

Dan Bentler

I know the road of thinking you're going down with the accumulator or surge tank, whatever you want to call it. Yes, we're regulated by the FDA, but that has nothing to do with why our process systems function the way they do. On a production line, either we're sterile or we're not, based on time/temperature and flow parameters.

Any sort of unecessary surge system makes these variables very hard to maintain, without proper steam blocks etc.....when you say accumulator, that can really just be a larger section of pipe. If you have a vessel on the other hand, yes that would be the solution to this...AND it's how we do it on full fledged process systems here. We have a timing pump in series with product or filling pumps, there's always a surge between them.

This application is a logistics thing, trying to get product from one process line, after it's timing pump to another portion of the plant using the booster pump. Adding a surge system, again can be done, but literally WAY overcomplicates this whole process. Our surge systems alone have around 200 I/O, required for monitoring sterility critical control points. It's just not a simple tank, exposed to atmosphere, it's a highly controlled vessel under pressure.

I really appreciate all of this feedback...

Greg