Is it acceptable to operate two pumps in parallel but not in the same speed and how can i determine then the QH diagram of the system of the two pumps?
fundamental principle - pumps operate at the intersection of the system resistance curve and the pump curve.
insufficient information is being provided.
are the pumps identical?
all other variables being equal, other than speed, the head for both pumps should be the same. exercise caution when operating pump at lower speeds so as to not operate below pump minimum flow.
perhaps the best recommendation is to obtain performance curve data from the pump mfg. for the desired speed. then cwith identical head, the pump's capacity can be determined.
Of course you can use two different pumps parallel.
But do you want to?
Probably not.
Design and Specification time mushroom into the project of the year as you try to make sure the controls encompass every last possible eventuality. Then you have that sucker hanging around your neck for five or ten more years as they call and ask you questions. Then, after ten years you finally get the chance to fix that booger by total replacement.
Controls and adjustments become tricky. Adjustments become "hypersensitive", Einstein (you) better be there to make adjustments, or at least duct tape phone to ear.
AND
Sometimes it just don't and cannot work no matter what, so you then wonder why so many people who are so dumb manage to get away with this crap all the time, while you (who knows a lot and pays due diligence) get caught with your underwear down around your ankles.
The above assumes we are not looking at a jockey pump.
pmover is correct about system curve & pump curve. But in your system the two unmatched pumps make that pump curve a tricky thing, so nows your chance to start on that project of the year.
It is rather easy to operate two pumps in parallel if the head is same (flowrate can be different).
I will definitely go with PUMPDESIGNER's suggestion if the pumps are totally different.
PS: Once a news paper reporter went to Einstein and said "Mr Einstein, is it (The General Theory of Relativity) so difficult that people say only two persons could understand it?" Einstein questioned back politely "Who is the other person?" Brilliance+Confidence personified as Einstein. No other metaphors either way.
what you intend is done all the time. If possible it is best to have both pumps with similar shutoff heads otherwise extra care needs to be taken to avoid having one pump deadheading (zero flow) the other.
To prepare the combined QH curve, I plot the QH curve of both pumps on the same graph. Then for each head point I add the flow from the first pump to that of the second. This gives a much flatter curve, sometimes with a hump at the start if the pump shutoff heads are not the same.
When this is done, I check the NSPH, efficiency/power and minimum and maximum flow limit for each pump to pick likely trouble points. Then I overlay the system resistance curve (or system curve envelope in variable demand/pressure scenarios) and make sure I am not close to any trouble points. In some cases it may be necessary to turn off one pump as part of the control strategy/system.
It is nice to have matched pumps in a system with a single system curve, but with multiple system curves and no VSDs somtimes different sized/speed pumps are better. The jockey pump is an extreme example. At other times it is simply necessary to use what is available,and this may result in using odd sized pumps.
Impeller trimming sometimes helps overcome a particular problem, as does a speed change by pulley or motor change.
As everyone has already noted, it is possible but not recommended to run two pumps in parallel at different speeds. The amount of trouble that this will cause you depends on the individual curves of the two pumps. First, you need to generate accurate curves for each pump at the speed that they are actually running. This can be done by hand using the affinity laws or with any of the readily available spreadsheet programs. Since pumps running in parallel have a combined curve that is additive in the direction of flow, you add collumns to the spreadsheet that add the flows produced by each pump at each head value. I would then plot all three of these curves on one plot: each pump's individual curve and the combined curve. Run the pumps and get a performance measurement. Read the suction and discharge pressure and total flow. Calculate the head and flow and mark that on the combined curve on the plot. Then following the line of constant head from the combined curve back to the two individual curves. That will tell you what each pump is contributing to the total. Check each of those points to verify that each pump is running above it's minimum flow. Check vibration on the pumps running this way to make sure that they are not in distress. Now that you know how, don't do it. It will almost certainly be trouble.
It is certainly possible to operate multiple types of pumps in parallel, and the above responses rightly suggest that the benefits are not likely to be worth the problems.
I recall a system that had just such an arrangement pumping lake water. Maintenance costs were absurdly high, and it was important to be careful when touching the pump casings. Sometimes the flow would stop in one or more of the pumps (the foot valves prevented flow reversal), and the power input then went to making hot water. The pump could become very hot very quickly. Do you really want to be responsible for such a mess?
