Subject: Pump partnering
With all the current talk about seal partnering, can pump
partnering be far behind? As with mechanical seals the concept is
simple; call in several pump companies, talk to them about a lot of
lofty ideals and then learn which of them will give you the best
price on a standard ANSI (American National Standards Institute) pump
if you promise to give them all of your business.
This is often called convergence and is recognized as a clearly
defined trend in industry..
If you are getting good life out of your present pump, and if
every one was selling the same thing, then that would be a good idea,
but that is not the way it is. The fact is that the ANSI (American
National Standards Institute) pump is not giving reasonable trouble
free life and to standardize on it would be foolish if you were
looking for long service.
There are two major problems with this pump:
- You will get poor mechanical seal life. A seal should run
until the carbon face wears down and in better than 90% of the
cases the seal will begin to leak with plenty of carbon face left
- The bearings have a L10 rated life of from 100 to
300 years and you know that is not happening.
The problem simply stated is that the ANSI (American National
Standards Institute) pump was made for conventional packing and you
are trying to use a mechanical seal because leakage is no longer
acceptable for a variety of reasons that include:
- Pollution problems cause hassles with the government and you
do not need these problems.
- Expensive products cost too much to have them leaking on the
- Fugitive emission laws legislate against all leakage.
- The packing is using too much water flush.
Let's look at the cause of the seal problem:
- Your inital seals were supplied by the pump manufacturer who,
in many instances, was forced to use Teflon® and other
non-elastomers in the seals because the pump was shipped to a
distributor that did not know where the pump was going to be used
other than in a product that was compatible with the pump
metallurgy. The Teflon® caused fretting problems (damage) on
the shaft so the shaft diameter was reduced and a sleeve was
installed to accept the damage and the resultant diameter
reduction weakened the shaft.
- The shaft was made long to accommodate at least five rings of
packing, a lantern ring or seal cage, a follower to tighten the
packing and enough room had to be left to fit your hands into the
pump to install this hardware. To provide all of this room the
impeller was moved too far away from the bearings.
- When the packing was replaced with the mechanical seal, the
shaft became unstable causing excessive radial movement.
- The long shaft additionally added a lot of unnecessary cost to
- The narrow stuffing box made sense when packing was installed
but made no sense when the packing was replaced with an expensive
mechanical seal. There is not enough clearance between the seal
outside diameter and the stuffing box inside diameter to provide
proper cooling and allow solids to settle out when the pump is
stopped. It took the pump companies forever to make the large
stuffing box available as an extra cost accessory.
- The double volute design was always the answer to shaft
deflection caused by operating off the pump's best efficiency
point (BEP). This feature was just about eliminated in the smaller
impeller sizes (less than 14 inch or 355 millimeters) to
accommodate the consumer's request for more efficiency. The pump
manufacturer should have explained to the consumer that the small
decrease in efficiency would be offset by a major decrease in seal
failure, but the pump companies kept silent. The whole subject has
proven to be academic because over 90% of the present pumps in
industry currently run throttled as a result of safety factors
imposed when the pump was sized and purchased requiring the
manufacturer to supply an oversized pump.
- The pump should have come equipped with a "C or D" frame
adapter to eliminate the need for making an alignment between the
pump and the motor. Unlike the packing, the seal is very sensitive
to misalignment problems and most mechanics complain because there
is never time to do a proper alignment.
- A vortex impeller would have reduced the amount of solids
sticking on the impeller in heavy slurry applications. The
attached solids destroyed the dynamic balance of the rotating
assembly, causing seal problems. Again the desire to appear
efficient took priority over reliability.
- With the exception of oil refineries most pumps sold in the
United States are of the open impeller type. This means that the
impeller has to be adjusted to either the wet end volute or back
plate to maintain the pump's efficiency. In all cases the ANSI
(American National Standards Institute) pump adjusts the impeller
from the power end causing the rotating part of the mechanical
seal to move every time an impeller adjustment is made. The open
impeller clearance should have been adjustable from the power end
and this problem never would have occurred. To adjust the open or
semi-open impeller without disturbing the seal face load requires
that an expensive cartridge version of the seal be purchased that
will consume most of the small radial clearance that is available
in the stuffing box.
- The mechanical seal should be designed to be installed in the
space between the stuffing box and the bearing case where there is
plenty of radial room and the seal is physically located closer to
the bearings. Pusher gland bolts could be used to eliminate the
problems caused by the present gland bolt's variety of bolt circle
The overall problem is that the pump manufacturer did not want to
alter his packed pump design to accommodate a mechanical seal. Since
he held all the power over which seal design was going into his pump,
he insisted on a set of criteria that reduced his cost and guaranteed
premature seal failure. The consumer who possessed little pump
knowledge and even less knowledge about mechanical seals went along
with a "child like faith" that all was well because he was dealing
with reputable manufacturers.
The ball or roller bearing problems are just as easy to
- There is not enough oil in the bearing case sump causing the
oil to get too warm. There should be enough room to accommodate at
least two liters of oil when the lubricant is at the proper level
of half way through the bottom ball when the pump is at rest. Oil
has a useful life of 30 years at 30°C (86°F). The life
of oil is cut in half for every 10°C (18°F) rise in
temperature. This means that oil has a useful life of only 90 days
at 100°C (212°F) because of "coking" problems.
- Grease fitting make no sense for grease lubricated bearings.
The fitting will guarantee that the bearings will be over
lubricated causing a heat problem. The bearing must be hand packed
and that is a real inconvenience.
- Grease or lip seals should never be approved for use in
centrifugal pumps or any other rotating equipment. The pump
company provides them as a standard
- These seals have a designed life of less than 2000 hours (83
days at 24 hours per day)
- They will cut (frett) the shaft because they remove the
corrosion resistant shaft's protective oxide coating.
- Labyrinth or face seal are a much better choice. The problem
is they cost more than grease seals.
- During operation, the shaft will thrust towards the pump
volute. Too often the radial bearing is being retained by a simple
snap ring that can bend or loosen as the snap ring groove
There are other features that can be incorporated into the
standard pump that would eliminate many of the current seal and
- A centerline design would compensate for normal thermal growth
of the volute causing the shaft to run non-concentric with the
- A low L3/D4 shaft would resist much of
the deflection caused by:
- Operating off the best efficiency point (BEP).
- Misalignment between the pump and driver.
- A shaft that is not dynamically balanced.
- Severe vibration. Cavitation as an example.
- The manufacturer should make various specific speed number
impellers available to the consumer. Too often the impeller sold
is a compromise, because the correct specific speed number was not
- An impeller inducer should be available if needed. Some times
it is the most practical way to get the NPSH you need.
- A "C" or "D" frame adapter would eliminate the argument that
"we do not have time to do an alignment".
Equally as important as the pump design is the knowledge you need
to troubleshoot piping and installation problems. A very high
percentage of troubleshooting time is spent on those two subjects. If
the consumer is going to save the cost of this service in the form of
a pump distributor discount (pump manufacturers seldom get involved
in seal and bearing failures other than to blame some one in
operation or maintenance as the cause of the premature pump failure)
then the consumer is going to have to provide his own service in
In this age of multi-craft mechanics and non-specialized
mechanical engineering, I'm not very confident about the probability
of their success. Think about it this way:
Two people coming towards each other from opposite directions will
meet somewhere in the middle as they converge, but they are clearly
going in opposite directions.
For information about my
CD with over 600 Seal & Pump Subjects
Link to Mc Nally home page