|
Improving pump reliability
through robust designs and pump selections
Stan Knecht, ITT Goulds
Over the course of the past ten years, pump users have made much progress
towards improving pump reliability. These activities started in the late
80’s and early 90’s mostly stemming from Total Quality initiatives
where-by tools like root-cause failure analysis, statistical analysis
techniques and a heightened awareness. Education through “Best Practice”
forums have allowed users to identify some of the elementary issues which
accounted for a significant portion of pump failures and downtime.
Since that time most users have implemented or are implementing some form
of Reliability Improvement programs. These typically involve some or all
of the following activities which allow them to identify and implement
corrective actions long before equipment failure occurs. Result, increased
Mean-Time-Between-Failure (MTBF) intervals:
Typical reliability improvement initiatives
1. Vibration monitoring and trending
2. Lubrication oil analysis and sampling
3. Improved pump alignment and pump installation practices
4. Improved rotor balance
The collective result of these initiatives have been impressive. Surveys
of pump user in the North American chemical industry have shown typical
improvements in MTBF from 15 months to 24 months. While improvements like
these have helped many companies to improve profitability and to maintain
their competitiveness, the everincreasing pressures of today’s marketplace
has many users now looking for additional measures to further increase
MTBF to maintain their competitive edge. When you analyze the actions
taken to date one can see that the focus has been primarily on improving
mechanical aspects of equipment reliability. In essence we likely have
harvested the low hanging fruit, yet to begin to harvest the higher level
fruit, one will need to take a more holistic approach and begin to gain a
better understanding of how the equipment itself functions within the
system.
Robust pump selections
During the design process of a pumping system designers and engineers
must consider many variables. One area to consider which can improve
equipment reliability is to look at the criteria which is used to select
and hydraulically size pumps. Over the years there have been many
guidelines and “un-written rules” established to help engineers specify
pump. Basic selection criteria often include objectives such as “selecting
a point to operate at its Best Efficiency Point (BEP),” “slower speeds are
better,” or to “provide adequate Net Positive Suction Head (NPSH).” While
these all represent sound engineering practices, often in “real world”
applications it is not practical or possible to meet these all of these
criteria and the engineer is left to make a subjective judgement as to
what pumps is best for the given service. Three common factors, which are
pertinent to most pump selections, are operating speed, impeller diameter
and operating point. In situations where traditional “ideal selection
criterion” cannot be satisfied one method proposed by Bloch [1] to help
engineers predict and compare the projected reliability of one hydraulic
selection to another is the concept of Reliability Factors. These factors
are simply non-dimensional numbers used to provide a relative index
ranging from 0 – 1 of one attribute as compared to the ideal for that
given attribute. A rating of one (1.0) does not mean infinite reliability,
rather that this would be the best selection possible; conversely a rating
of zero (0.0) does not imply zero reliability, rather that this would not
be recommended for this application. These factors proposed by Bloch have
been further confirmed via laboratory testing as reported by Erickson, et
al.
Operating speed factor (FR)
The effects of operating speed or RPM on pump reliability are likely
easiest for all to understand. It is likely best expressed by a phrase
used in the sports world, “Speed kills.” Operating speed affect pump
reliability through rubbing contact primarily at the faces of mechanical
seals in addition to having a significant impact on reduced bearing life
though increased cycling, lubricant degradations and reduced viscosity due
to increased temperature. Operating speed also has obvious negative
impacts on pump component (impellers, casing, etc) wear especially in
services where the pumpage is abrasive. Lastly speed has an inherent
impact on a pumps suction performance which can ultimately lead to
decreased reliability brought on by the increased susceptibility to
cavitation problems due to high NPSHr requirements. Figure 1 provides a
graph to predict the Operating Speed Factor (FR). In practice a given pump
designed with a maximum operating speed of 3500 RPM would be assigned FR=
0.2 when applied on a service at 3500 RPM. However, the same pump operated
at 1750 rpm, or a ratio of 50%, would be assigned an Operating Speed
Factor FR= 0.6.
Impeller Diameter Factor (FD)
Many people might assume that selection of an impeller trim at the
maximum diameter is the best selections as it is at this point where the
geometry of the impeller and casing are best match as engineered during
the hydraulic design of the pump. While this is true if the pump is in
fact operating at its Best Efficiency Point (BEP) or design point, however
off BEP operation with max trim impellers can result in some undesirable
effects that are detrimental to pump reliability. Specifically when a pump
is operating at an off BEP condition the exit angle of the fluid leaving
the impeller is miss-matched with the angle of the “tongue” of the casing
which results in a pressure pulsation that ultimate leads to increased
shaft deflection which will decrease the life of the mechanical seal. It
is important to note that at reduced speeds the effects of impeller
diameter trim is somewhat reduced although still present. To alleviate
this phenomenon it has been found that if the clearance between the OD of
the impeller and the tongue of the casing can be increased that the
magnitude of the pressure pulsation can be decreased resulting in
increased pumping reliability. Based on the chart shown in Figure 2, which
has been confirmed by lab testing, it can be seen that pump selections
where the impeller trim is selected to be 60 – 80% of the total trim will
result in the best pump life. Thus, a pump which has a maximum diameter of
10” and a minimum trim diameter of 6” would be best applied at a trim of
9” where the Impeller Diameter Factor FD= 1.0.
Operating point factor (FQ)
As mentioned earlier centrifugal pumps are typically designed toward
achieving a single flow and head at a given speed. This point can is
identified as the BEP on a pump given curve. At this point all the
geometry of the pumps hydraulic design are matched and the pumps will
typically operate at its highest level of performance, efficiency and have
the lowest hydraulic loading. Like with the impeller trim, when this same
pump is applied for a service for off-BEP rating operating points
hydraulic loading and other pumps performances aspects are not optimized
which typically results in decreased pump reliability. In practice this
condition has been found to be less detrimental for smaller pumps, thus
Operating Point Factor is shown to be contingent pump design size. So
while the use of each of these factors allows one to better understand the
ramification of a given pump selection as compared to another for a given
aspect, most real world pump application will involve a compromise of all
these factors. One needs not to worry as an overall assessment can be made
simply by multiplying all of these factors together to create a
Reliability Index as follows:
RI = FR x FD x FQ.
The resultant from this calculation will give an engineer a quantifiable
ratio that can be used to help decide the merits of one pump selection as
compared to the other to determine which might offer best performance with
regards to reliability. Like all the individual factors again a RI = 1.0
does not imply infinite reliability, likewise a RI = 0.0 does not imply
zero reliability rather they simply are and indication that one selections
might be better suited than the other. It is also important to note that
this methodology must not be used to compare one pump design to another as
the mechanical designs of these two pumps are likely to be different which
can also influence pump reliability.
Robust Pump Designs
As stated above in addition to hydraulic selection, further
improvements in pump reliability can be obtained by selecting pumps that
have robust mechanical designs. One of the benefits of the TQ movements
mentioned earlier was that a lot of data was collected and analyzed, from
which it was found that the majority of pump failures were actually the
result of seal failures. Using root cause analysis techniques it has been
found that these failure were not the result of poor seal designs, rather
they were the result of poor sealing environments. When you look at the
basics of a mechanical seal design in principle they are a contacting
design for which supply of a clean and cool sealing environment with
adequate lubrication for the seal faces is critical to extending seal
reliability. Based on significant testing conducted by both mechanical
seal vendors as well as pump manufacturers it has been found that seal
chambers designed with enlarged and tapered cavities similar to those
shown in Figure 4 will for most applications provide the best sealing
environment to promote extended seal life. As a further improvement to
these designs some pump manufacturers have made these seal chamber designs
even more robust by incorporating devices into them which better controls
the flow pattern within the seal chamber cavity. These devices (Figure 4)
function to keep solids and grits, often present in process streams, away
from the seal faces and spring mechanism of the mechanical seal to
eliminate premature wear and failure of these components. The second most
common cause of pump failures was found to be the result of bearing
failures. Like mechanical seals, as bearings are a wearing component a
proper operating environment must be maintained for extended bearing life.
As with mechanical seal, the proper operating environment is one that
provides adequate cooling to dissipate the frictional heat generated and
one that maintains a clean environment free of contaminants. Research
conducted by ITT Industries – Goulds Pumps has shown that the use of a
large capacity oil sump will significantly extend pump bearing life and
reliability [3]. Reasons for this are as follows:
1. Larger Oils sumps have greater radiating surfaces to dissipate more
heat.
2. The larger the oil sump the longer the oil stays in the sump allowing
more cooling.
3. The longer the oil stays in the oils sump the more opportunities there
are for contaminates to settle out to the bottom of the sump where they
can be collected on the magnetic tip of the drain plug
In addition to large capacity oil sumps one other robust pump design
feature to consider is labyrinth style oils seal. The primary benefit of a
labyrinth style seal is gained from the fact that sealing of the bearing
frame is achieved via a series of closely machined passages which are
arranged to form a tortuous path which prevents the ingress of contaminant
into the oil sump while at the same time retaining the lubricating oil
within the oil sump without contacting the rotating pumps shaft. As
compared to the more commonly used lip seal designs, which rely on the
physical contact of an elastomer with the rotating shaft significant
reliability improvements are gained for the following reasons:
1. Unlimited life and full time protections against contamination as
Labyrinth seal do not wear out.
2. The non-contacting design does not produce any frictional heat that
could be transmitted to the lubricating oil. This results in cooler oil
temperature which lead to extended bearing life.
“Smart” pumping solution
Yet even if one adopts all of the recommendations and best practices
mentioned here and know to the industry, it is unlikely that one will
achieve maximized reliability due to the fact that many failures
experienced in the industry today are the result of random system upsets
or operators errors. Failures which occur as the result of valves not
opening, etc which then cause pumps to be “dead headed” or “running off”
the pumps operating curve are difficult to predict and to prevent. However
in recent years some pump manufacturers have begun to offer “Smart Pumping
Solutions” that typically involve a robust pump design packaged with self
contained monitoring and control system which can continuously monitor
pump performance versus system demand. These “intelligent pumping
solutions” are akin to the electronic ignition systems used in most
automobiles today for which the package typically includes a pump, an
array of pressures, temperature and flow sensors, a variable speed drive
and “chip” which contains the pumps performance capabilities, DCS
communication and control software. In operation these systems act
continuously not only to provide diagnostic monitoring of the pumps
operations to detect and prevent against failures, yet they also
constantly regulate pump performance to match system demand resulting in
maximized performance often leading significant energy savings. A
documented example of this was a Georgia-based chemical manufacture that
had a problematic pump that was failing every 17 days due to cavitation
damage created by unavoidable process operating conditions. Many “fixes”
to this problem were attempted with limited success. This customer then
retrofitted his pump installations with a Goulds Pumps - PumpSmart™, smart
pumping solution and his pump has been running trouble free for the last
10 months without a single failure. At another installation at a Texas
industrial gas manufacturer, the PumpSmart systems was applied to a
cooling tower system which had variable demand and was able to reduce
energy consumption by 30% due to it’s ability to regulate pump demand
versus system performance. This system replaced a common fixed speed pump
which “burned off” pump performance across a control valve in order to
match pump performance to the needs of the system.
Summary
As industry seeks to elevate equipment reliability to the next higher
level one must take a holistic approach to identify measures which can
further improve reliability. Improvements can be found by implementing a
more robust approach\method as to how pumps are selected and sized for a
given system. The use of Reliability Factors and the Reliability Index
give users a quantitative method to assist them to help make a more
objective determination to compare one pumping selection to another to
maximize reliability. Further one must look to the pump designs themselves
to identify design features which will provide a more robust piece of
equipment which is better able to sustain itself in less than ideal
operating conditions. Features such as enlarged seal chamber, fitted with
flow control devices, as well as oversized oil sumps and labyrinth style
oil seals are a few concept which have been proven to provide results.
Lastly as we continue to prune the “low hanging” reliability improvements
it will be likely that we will increasingly be faced with random and/or
more complex problems as we pursue further additional opportunities. The
use of “Smart Pumping” technologies will likely constitute a key component
of the means which we will use to push Pumping Equipment reliability to it’s
maximum.
REFERENCES:
(1) Bloch, H.P. and Geitner, F.K., “An Introduction to Machinery
Reliability Assessment,” 2nd ed., Gulf Publishing Co., Houston, TX (1994).
(2) R. Barry Erickson, Eugene P. Sabini, Anthony Stavale, ITT Industries –
Goulds Pumps, “Hydraulic Selection to Minimize The Unscheduled Maintenance
Portion of Life Cycle Cost,” 1998
(3) Dr. Lev Nelik, ITT Industries – Goulds Pumps, “Bearing Life Extension
and Reliability features of Modern ANSI Pumps, The 2nd International
Conference on Improving Reliability In Petroleum Refinereies and Chemical
and Natural Gas Plants,” 1993
|
