Litterature and or theroy behind pumping speeds

aircooled81

As frequency drives are fairly frequent these days, i have come to notice not much theroy is implemented in their set-up and use. I am charged with the task of starting these up, for secondary piping systems often.
Though i have read some long pdfs from bell & gossett about system pumps and varying speed, i just dont see the big picture.
My questions are, is there some fantastic litterature that speaks to controlling hydronic loops with vfd's effectively and efficiently, anything regarding pumps in parrellel pros and cons about running lead lag vs similar twinned speeds, and minimum gpm to run a secondary loop with high efficiency boilers that require constant flow through their associated primary.

I am also open to someones theroy, or rule of thumbs.

For instance,
I read that as the speed is decreased by lets say 1/2, the hp is decreased by 1/4. This equates to great power savings, but when two pumps are running in parallel, they manage to slightly increase the gpm at the lower half of the pump curve, but not raise the head. As they both get toward full speed you maintain the same head but gain significant gpm. But i dont see the advantage if you have a fixed cv ratings on control valves in the loop, have too much pump, and throttle most of the loop back as the load decreases. Now you have 2 pumps running at their min speed so they dont burn up their motors, and no increase of flow through the loop because your head never got higher to shove a bit more water through some three way valves.

Jamie Hall

Best to start off your learning with constant speed pumps! They're bad enough.

For centrifugal pujmps, thepump has a characteristic curve of head vs. flow, and a related curve of power requirement and efficiency -- which varies with where on the curve the pump is operating.

With two pumps in series, the heads will add but flow will not change. With two pumps in parallel, the flows will add, but the head will not change.

There will also be a curve of head vs. flow for the system into which the pump is pumping. Where those two curves (the pump characteristic curve and the system characteristic curve) intesect is where the system will operate. Change any of the variables, and both the flow and the head will change, along with the power required and the efficiency.

Note that with two pumps either in series or parallel, the operating point -- intersection of the combined pump curves and the system characteristic curve -- will change. Thus for instance with two pumps in parallel, with very rare exceptions the flow will not be twice the flow of either pump alone into the same system, since the flow -- and hence head loss in the system -- will be greater; the combined flow will be less than the total flow of the two pumps operating alone. Similarly, with two pumps in series, the combined flow will be greater than either pump operating alone, as they can produce more head and thus move farther up the system characteristic curve -- but the total head will be less than the simple sum of the two pumps.

Different pump construction and geometry will result in different head vs. flow curves. Some pump designs have remarkably steep curves (the head changes greatly with flow); some are rather flat.

Some astonishingly evil things happen if pumps are mismatched, by the way.

I would also note that all centrifugals have a definite shutoff head, related to pump geometry and operating rpm, and a definite maximum flow which is related to geometry, but not rpm. The efficiency, of course, drops to zero at both maximum head and maximum flow.

The above comments emphatically do not apply to other types of pumps! Only centrifugals.

hot_rod

An excellent overview from Jamie.

One part that pump installers should understand is the system curve, how to establish it, and how it overlays or corresponds to the pump curve. Where the two curves intersect, the pump curve and the system curve, establishes the OP operating point. With VS and zoned systems the system curve moves across the pump curve.

A general rule is to run in the mid 1/3 of the pump curve. Small wet rotor circulators are not very officient, even in a sweet spot, maybe mid 20%..So running one way off curve, at either end is inefficient.

The good news is we have a much wider selection of pumps and ECM V/S offerings these days. It should be possible to find a single pump to cover the loads. Multiple pumps may be installed for back up or redundancy, or other special applications.

Here is some basic, understandable, reading on developing and using curve tables and hydronic applications in general.


http://www.caleffi.com/sites/default/files/coll_attach_file/idronics_16_na_0.pdf