Upload
ismael-khalil
View
215
Download
0
Embed Size (px)
Citation preview
8/19/2019 cr0701ch12
http://slidepdf.com/reader/full/cr0701ch12 1/4
0-8493-????-?/97/$0.00+$.50© 1997 by CRC Press LLC
12
©1999 CRC Press LLC
Case History #3 —
Progressing CavityPump Failures
As mentioned earlier, progressing cavity pumps (PC pumps) are known to operate
very well in many particularly tough environments where abrasion and wear make
other types of pumps impractical. In the case of this example, a medium-size PC
pump was installed at the wastewater plant in the last stage of the treatment process.
At that stage, all bacterial and other contamination impurities are already removed
by the upstream processes (chemical additives, settling tanks, etc.). Now, the pump-
age had to be transferred, as a final sludge of very high viscosity (500,000 cp), into
large containers for trucks, to be subsequently hauled as fertilizer to the farm fields.
The problem reported was that the pump would stop pumping in less then oneweek. Rotors would wear badly, and increased leakage path would result, losing all
of the pumped flow in slip. It was first suspected that the initial fit between the rotor
and a stator was too tight, which, in the presence of a very dry viscous pumpage,
would cause high friction and wear.
The relationship between the rotor/stator clearance and slip is not linear (see
Figure 73). A good thing about the fit being tight is that there is practically no slip,
especially if the number of stages is large at the same time. PC pumps are typically
designed for approximately 75 psi per stage. A 300 psi application is typical for PC
pumps, although there are extremes in both directions, with pressures as high as2000 psi. For 300 psi, a 4–5 stage design would keep slip to a minimum. A bad
thing, however, about a tight fit is high mechanical friction, resulting in high running,
as well as starting, torque.
If fit is reduced, but still tight, the pump flow would remain relatively constant,
and would not change much until the value of –0.015" diametral interference is
reached (‘–’ indicates interferences, and ‘+’ means clearance). Eventually, slip begins
to take place, and increases almost linearly with change in interference, until about
line-to-line fit is reached. After that, slip increases exponentially and very rapidly,
and at +0.015"/+0.020" most of the flow could be recirculating back to suction, (i.e.,fully slipped). The volumetric efficiency follows the trend as explained previously.
(Note: the values are approximate and depend on other factors, some of which are
explained subsequently).
8/19/2019 cr0701ch12
http://slidepdf.com/reader/full/cr0701ch12 2/4
©1999 CRC Press LLC
The starting torque follows the reverse relationship as compared to slip. At first,
with tight interference, the starting torque is extremely high — many field problems
can be traced to this, and many motors have suffered when started in these (essen-
tially locked rotor) conditions.
The trick is to know the correct interference or clearance, to satisfy both through-
flow requirements with minimum slip, and to ensure that the pump starting ability
is not impacted. The temperature is another factor to consider, since the elastomer
(rubber) expands, changing the rotor/stator fit. It is probably better to err on the safeside, and end up with somewhat more slip, but longer life — the life of the rotor
and stator decrease exponentially with fit.
Working with the customer, we did not find anything obviously wrong with the
pump or the system; no blatant mistakes were made in the application. The factory
test data and the design parameters did not point out at any obvious reasons for
concern either. The pumpage was not abrasive, and all hard particulate inclusions
had been removed way upstream before the pump. The pump was running at about
450 RPM.
After a full day of work, rather disappointed with lack of porgies, I stood somedistance away from the pump, thinking about other possible causes of the problem.
Absentmindedly, I picked up a handful of lime, which had seeped out of the bag
laying nearby. Mauling it around in my hand, and thinking about the pump, my hand
felt moisture apparently emanating from inside the lime. I could also sense sharp
particles, like little stones, in my hand.
Lime is added to the pumpage in small quantities to “soften” the pumpage or
to reduce the high viscosity of the sludge. Added at the pump suction, the lime
produces a mild chemical reaction resulting in some water, which is helpful to
lubricate the stator/rotor interference, to reduce friction (with the obvious intent to
increase the pump life), as well as save energy cost via better efficiency.
We picked more lime, placed it in a container, and washed it out. The lime
“melted away” and washed, but a few stones remained at the bottom of the container.
The label on the bag indicated the lime contained only a very small percentage of
FIGURE 73 Slip and torque, as functions of rotor/stator fit.
torque slip
-60 -40 -20 0 20 40 60 80
- interface + clearance
8/19/2019 cr0701ch12
http://slidepdf.com/reader/full/cr0701ch12 3/4
©1999 CRC Press LLC
impurities. Although the process sludge is not abrasive, the addition of the hard
particulates with lime changed what was entering the pump suction.
Generally, PC pumps work well on abrasive applications, provided that the
rotational speed is selected properly. When abrasives pass throughout the PC pump,it is the unique characteristics of the elastomers, such as rubber, that save the day:
instead of fighting it, the rubber actually catches the particles, deflecting into itself,
like a trampoline. This deflection results in elastic energy, which is then released to
catapult (i.e., to eject) the particles out. Very little damage to the rubber results, but
it can be spotted upon the examination as many small, often microscopic, cuts (see
Figure 74).
When the clearance between the rotor and a stator is sufficient, there is enough
space for the particles to be released to; otherwise, some of them get “jammed” just
when the catapulting action is about to release them. If this natural swing-likemechanism of embedding-and-release is interrupted, or gets out of synchronism
(similar to a swing abruptly stopped), the elastic energy is lost, and the particle
remains embedded in the rubber. Depending on rotor/stator fit and particle size and
shape, enough particles may become embedded in the elastomer, creating a very
abrasive surface (see Figure 75). The wear that it causes is then obvious. Even if
this lime was purchased at a premium with lower residual limestone, it would not
take long to create the abrasive surface; perhaps one week failure rate could be
stretched to two, but no more. The only way to handle these particulates is to remove
them. Working together, we devised a simple cyclone separator, sort of a trap, tokeep the particles from jumping over the wall of the “stopper” device we designed
for the suction line (see Figure 76). Periodically (we calculated once a day), a hatch
in the bottom of the trap would have to be opened and cleaned of particles — either
manually or automatically.
The solution may be good for this relatively small pump size, but would it work
for larger sizes? The inlet trap may or may not work for the larger sizes because the
settling velocity of the injected lime could be different (if too high, it would carry
the particles over the wall of the barrier), and again, some experimentation could
be required. Often a simpler and possibly better solution, from the beginning, wouldbe to apply a larger pump size, running at slower speed, perhaps 150 to 200 RPM.
The slower speed would be more forgiving to abrasives. The somewhat higher initial
cost could have been paid up in repairs, parts, and process downtime. The fit between
the rotor and stator should also be loosened.
FIGURE 74 Resistance to wear, due to rubber elastic energy to receive/eject particles.
Particleejected
Particleembedded
StressesRubber
Particleenters
Flow
8/19/2019 cr0701ch12
http://slidepdf.com/reader/full/cr0701ch12 4/4
©1999 CRC Press LLC
Overall, a combination and impact of several variables would need to be inves-
tigated and sorted out, perhaps with some testing, to devise a reliable, trouble-free,
economical, and lasting pump operation.
Those who tried know that a solution to pump problems is often not obvious,
and, just like the cases we discussed in this book, must be approached carefully and
with consideration of all — even seemingly unimportant — factors.
FIGURE 75 A very abrasive surface.
FIGURE 76 Lime stone “stopper” device. Height “h” depends on particles’ size, compressor
air supply, and percent of allowable stone to the pump.
Rubber
Particles
CompressorPump
PureLime
Lime & Stones
Bagh