The Wrong Maintenance Priorities Threaten Corporations
Kathy | December 1, 2005
According to this well-known veteran of the reliability trenches, one of the best ways to hone the survival skills industry needs today and into the future is to pay attention to the basics. Taking time—making time—to read is a great way to start.
Many process plants assume that better maintenance strategies will lead to higher equipment reliability. Very often, the primary focus of these strategies is to avoid unnecessary oil changes or to optimize compressor overhauls and the pursuit of other preventive measures. Similarly, many of these strategies hope to help companies avoid equipment damage and costly production interruptions by doing appropriate maintenance “just-in-time.”
While these are commendable goals, they do not address the constraints that are built into vast quantities of equipment that incorporate less-than-optimized components. Nor do such strategies remedy the numerous random failures that strain the maintenance budgets throughout industry today.
Staffed by harried employees, shops frequently become adept only at replacing parts in kind. Likewise, relatively few companies position themselves to systematically implement maintenance avoidance measures. We know that, in existing plants and with few exceptions, failure avoidance would be far more profitable than implementing optimized maintenance timing on non-optimized equipment. In new equipment procurement situations, utilizing specifications that eliminate the very components that risk causing frequent maintenance and downtime would provide far greater returns on the incremental investment than fine-tuning an asset management or related program.
Take, for example, the many operating plants today with literally hundreds of thousands of pumps that were purchased from the lowest bidders. It is wistful thinking to expect that all components in these lowest-cost machines represent best-available technology. In the age of downsizing, rightsizing and outsourcing, how realistic is it to assume that all of the various equipment manufacturers and vendors employ seasoned, well-versed, well-read subject matter experts?
Suppose a manufacturer recently sold less-than-optimum equipment. Knowing that we live in a litigious environment, would we really expect this manufacturer to concede that he/she continues to make, sell or market non-optimized equipment or components? If the answer is no, then it is clear that the user/purchaser has to be the driver for identifying and implementing equipment upgrades.
Trends that lead nowhere vs. trends for best-of-class performers
At the risk of inviting irate responses from benchmarking companies, we contend that the trend towards increased benchmarking will, ultimately, add little value to many enterprises. A recently published article mentioned four so-called perspectives, labeling them Operations, Reliability, Work Management and Safety & Environmental. Goals were specified for each and 60 different key performance indicators (KPIs) were listed as useful for managing risk and improving profitability.
At best, each one of those 60 benchmarks may give plants an indication of where they are in the game. Yet, not any of these listings specified even one of the many precise steps that really represent lasting improvement. What good is it to tell a facility it is re-working too many pumps, if nobody is able to explain the root cause reasons for this “excessive re-working” at that plant? While it’s nice to point out the fact that “there must be a problem somewhere,” far more value would be derived by adequately describing the root causes and solutions.
Today, truly best-of-class performers use asset management and streamlined maintenance strategies as “icing on the cake.” They realize that these approaches add value only if the basics are in place and being practiced with consistency and forethought. As an example, best-of-class owner/purchasers are not likely to buy from the lowest bidder. They generally look at several competing offers and carefully examine which of them have “designed out” maintenance and failure risk. Best-of-class companies rarely, if ever, enter into lopsided alliances with suppliers. They will always use well-thought-out specifications that clearly describe and explain specific “upgraded” component materials, configurations, lubricant application methods, etc. Thus, the most important attribute of true best-in-class performers is their ability to provide authoritative answers to two questions:
1. Can a component be upgraded to resist failure?
2. If upgrading is feasible, is it also economically justified?
These are primary. . . these are the basics. Everything else is of lesser importance. Best-of-class performers know this to be a fact and are organized accordingly. Moreover, they are staffed so as to have a person—a designated and responsible individual—who can answer these two questions quickly and with great accuracy.
Why upgrading is often best
Unfortunately, even now, buying from the lowest bidder remains the predominant procurement mode. Equally disappointing is the fact that those responsible for shortsighted decisions are often the ones that block access to systematic failure-avoidance measures. Consequently, even the otherwise desirable life cycle costing (LCC) methodology is an academic exercise unless the person doing the comparison is in a position to answer the two previously-asked questions.
A facility which assumes that improvement initiatives spring forth from the original equipment manufacturer, or OEM, often will be disappointed. When, in 1986, a representative of a prominent pump manufacturer was asked why its designers didn’t engineer better pumps, the answer was that most customers selected pumps primarily based on cost and schedule. Accordingly, sales success was linked to cost and schedule, not long-term quality.
More recently, at a symposium in Houston, another pump manufacturer claimed that general-purpose pumps were designed to be overhauled or repaired every 18 months. To keep costs low, two pump manufacturers said they couldn’t afford to upgrade their pumps.
And, just last spring, at the 2005 NPRA Maintenance and Reliability Conference in New Orleans, several panel members touted key performance indicators that were largely based on not having production interruptions. To this day, a large number of managers and reliability engineers seem to be unconcerned if their pumps fail far more frequently than those at a competitor’s facility. The thought was even expressed that keeping pump failures at (relatively) high levels was one of the “safeguards” preventing upper-level managers from cutting the maintenance budget.
At the same NPRA conference we met with a presenter of asset management strategies. We attempted to argue the monetary merit of failure reductions by selective upgrading. When the speaker suggested that his organization was very effective in identifying and recommending the various upgrade options, we challenged his claims. We have yet to find asset management consulting companies that identify the needed upgrade measures to the degree of detail urgently needed by industry.
In support of our beliefs, we cited lube application in pumps as one of the many examples of industry not even being made aware of tangible reliability risks. This example deals with the use of oil rings in literally millions of equipment bearing housings, most of them in centrifugal pumps. Recall that the entire issue centers on our contention that industry is losing knowledge and application of the basics. Changing or fine-tuning management approaches will not bear the promised fruits unless the approaches are interwoven with systematic upgrade efforts. The following cases illustrate the type of dilemma with which industry is wrestling.
Case #1: Lifting oil with bicycle chains
While working with a client to determine the root causes of sludge in an oil sump and bearing failure in a pump, an experienced consultant (who was formerly employed as director of new pumping machinery development for two noted manufacturers) found a bicycle chain in the bearing housing. Its purpose, of course, was to feed lube oil to the bearings. Chances are that the bearing housing was simply too narrow to accommodate oil rings or similar means of lube application—a serious reliability risk.
When the consultant questioned the appropriateness of using a bicycle chain in this manner, the pump manufacturer objected to the criticism and claimed “that’s the way we generally do it. . .we hear no complaints.” Basic science, or the most elementary application of engineering formulas, though, would show that the chain would have no chance of moving at the peripheral speed of the shaft at anything other than—for process pumps—unusually slow speeds.
In most instances, the bicycle chain would slip relative to the shaft surface and, by virtue of the total downward-acting weight of the heavy chain, the side plates of the links would rub on the shaft. Wear-related oil contamination would almost certainly result, as was found and documented by the consultant. All of this begs the question: Would your asset management consultants have the basic knowledge to alert you to this? Or would your consultants limit their contribution to the rather obvious, i.e. telling you that you’re spending too much money on maintenance, and that you have “X% more” or “Y% less” shop backlog than the industry-recommended average? That would be nice to know, but where’s the real solution?
Case #2: The limitations of oil rings
Pump bearings in best-of-class U.S. oil refineries fail—on average—every 10 years. In certain other U.S. oil refineries, the failure rate is three times higher, with the average pump mean-time-between failures (MTBF) closer to three years. Let’s re-state our earlier point: To really add value, asset management consulting firms will have to authoritatively advise and advocate specific component upgrades. These firms must know, and must tell, the user-client, that oil rings (Fig. 1) impose a key limitation on the MTBF of many pumps.
While perhaps representing one of the least expensive means of applying lube oil to bearings, oil rings are rarely a wise choice for the reliability-focused. From about 1840 until 1990, they were furnished in brass or bronze. More recently, and for reasons we wish to subsequently spell out, some manufacturers have experimented with plastic and aluminum rings. The results are mixed, at best. In any event, oil rings suffer from a number of limitations that are rarely recognized by equipment suppliers and users. Reliability-focused users avoid oil rings because these components represent an undue reliability risk. Here’s why:
- Even some of the most advanced laser-optic shaft alignment systems will not have provisions ensuring that the shaft centerlines are absolutely horizontal. Visualize, therefore, how oil rings installed on shaft systems that are not totally parallel with the true horizon will run downhill. Doing so, an oil ring will make frictional contact with either a groove machined in the shaft, or some stationary surfaces associated with the bearing housing. The oil ring now tends to slow down, feeding less oil into the bearing. Many observers have also seen oil rings that showed clear evidence of edge wear and metal loss. Needless to say, the lost metal shavings end up contaminating the lubricant—not a desirable condition by any measure.
- Oil-ring movement and circumferential speed are affected by the degree of immersion in the lubricant and by lubricant viscosity. Typical immersions are shown in Fig. 1, but recommendations may vary for different types of equipment. Clearly, a more deeply immersed oil ring or oil rings contacting an excessively viscous lubricant will not perform as intended. Also, for good tracking and to revolve with reasonable consistency, oil rings must be concentric within 0.002 inches (0.05 mm).
- Oil-ring operation is affected by shaft surface velocity. As an experience-based rule, authoritative texts (Refs. 3 and 4) caution that shaft velocities as low as 2,000 fpm (~10.16 m/s) might represent the safe, or practical, field-installed (non-laboratory) limit for many oil rings. At 3,600 rpm, this limit infers a maximum shaft diameter of approximately 2.125 inches (~55 mm). It represents a “DN” value of 7,650, where DN is the product of shaft diameter (inches) and speed (rpm).
- Reliability-focused users recommend flinger discs. Since flinger discs are secured to the shaft, they are not subject to the compounded influences of shaft horizontality, oil viscosity, depth of immersion and ring concentricity. They are a vast improvement over oil rings and are, in fact, available in many pump models presently marketed by U.S. and European suppliers. Ref. 1 contains an illustration from a 1960s-vintage catalog issued by a then prominent, major U.S. pump maker. The page shows the flinger discs furnished with this manufacturer’s pumps and states, rather pointedly, “anti-friction oil thrower (meaning flinger disc) ensures positive lubrication and eliminates the problems associated with oil rings.”
Indeed, oil rings were problematic in the 1960s, and, more than 40 years later, they are still causing problems in many field installations. Retrofit flinger discs are available as cost-effective upgrade and retrofit options. Made to oversized dimensions, they can be easily trimmed to the required diameter. Their elastomer will fold into an umbrella shape during insertion through a narrow bearing-housing bore and will then snap back into its regular disc shape.
In 2003 and 2004, thorough testing was done on a Viton¨ disc’sconfiguration at different speeds and with oils of different viscosities. Two results of this testing are shown in Fig. 3 and Fig. 4 for ISO Grade 32 and 68 lubricants at 3,600 rpm shaft speed.
In each case, with flinger discs installed, the oil and bearing temperatures were compared against operation with the flinger disc removed and lube oil reaching the center of the lowermost bearing ball. From the graphs, it can be seen that, at higher pump speeds, lowering the oil level and using the trimmable flinger disc will reduce oil temperatures. Reduced oil temperatures will slow the rate of oil oxidation and tend to more closely maintain lubricant viscosity. Incidentally, with premium synthetic lubricants and operation at typical process pump speeds, the rate of oxidation is extremely slow. In that case, concern over oxidation issues on hermetically closed pump bearing housings are of very academic interest.
Economic value explored
Upon close examination, and with competent failure analysis, many observers have reached the conclusion that a large percentage of oil rings show signs of severe abrasion. It is undisputed and well known that the resulting lube oil contamination is reflected in premature bearing failures. Based on these observations, it has been estimated that at least 5% of the centrifugal pumps installed in the average petrochemical plant suffer from oil-ring deficiencies of sufficient magnitude to reduce bearing life from an assumed achievable six years to typically only three years. Other pumps may experience oil-ring degradation that reduces bearing life from five years to four years, and so forth. The issue is so intuitively evident that, to date, no one appears to have seen fit to spend research funds on scientific studies. Accordingly, empirical observations will have to suffice.
In any event, expanding on this conservative estimate, we might be dealing with a plant comprising 600 pumps. Suppose that of these, 18 “suspect” pumps were being repaired every three years to the tune of $6,000 per incident. This would require an expenditure of $36,000 per year. If, using trimmable flinger discs, the MTBR (mean-time-between-repairs) could be extended to six years, this expenditure would drop to $18,000 per year for the affected 5% of the plant’s pump population. Needless to say, if one paid $50 per flinger disc, the 18 discs would have cost $900 and the investment would have had a payback of $18,000/$900 = 20:1. It is certainly no stretch to foresee greater savings and even more significant payback than demonstrated in this example after one or two years of operation.
Belaboring the point
The issue at hand is important enough to be highlighted again. Management often doesn’t seem to get it. Our view is simply that asset management and maintenance strategies are rather pointless if oil rings and flinger discs, the pitfalls of millions of inadequate old-style constant level lubricators and a veritable host of other basic issues are either not known or not addressed.
Much money is lost when the basics are not understood. If each of 10 important or failure-prone components, practices, commissions or omissions in a pump were to reduce its reliability by 10%, raise 0.9 to the tenth power and convince yourself that you get less than 35% overall reliability. Staying with vulnerable components and not upgrading is a very poor choice indeed. Before looking for “high tech” and whatever else might be “icing on the cake,” a reliability-focused organization will learn to view every repair event as an opportunity to upgrade!
Furthermore, if a manager is really serious about upgrading the knowledge base of a reliability workforce, he or she will cheerfully spend a few hundred dollars on solid textbooks that explain hundreds of these upgrade opportunities. He or she will know, or at least accept as fact, that implementing one or more of a number of highly cost-justified upgrade examples will definitely avoid failures. Since the average API pump failure event costs U.S. refineries in excess of $10,000 (Ref. 2), a single avoided failure represents a three-week payback for, say, a modest $600 spent on books.
A good manager will probably insist that his/her reliability staffers read 200 textbook pages per yearÐthis adds up to a single page per work day. A good manager will not tolerate any excuses.
Surely, a professional who has neither the time nor motivation to read a page a day will never help his employer move ahead.
In the words of Mark Twain: “A man who chooses not to read is just as ignorant as a man who cannot read.” To which we might add that managers who choose not to make their people learn would serve their stakeholders better by going on permanent vacation.
Before encouraging or allowing subordinates to simply “decorate the cake,” a good manager will see to it that the underlying foundation, that is, the cake itself, is edible. That implies that the basics are in place.
- Bloch, Heinz P.; “Slinger Rings Revisited,” Hydrocarbon Processing, August 2002
- Bloch, Heinz P. and Alan Budris; (2004) Pump User’s Handbook: Life Extension, The Fairmont Press, Inc., Lilburn, GA 30047, ISBN 0-88173-452-7
- Wilcock, Donald F., and Richard E. Booser, Bearing Design and Application, (1957), McGraw-Hill, New York, NY
- Bloch, Heinz P.; “Centrifugal Pump Cooling and Lubricant Application—A Technology Update,” International Pump User’s Symposium, Texas A&M University, Houston, TX, 2005
- Bloch, Heinz P., (1998) Improving Machinery Reliability, Third Edition, Gulf Publishing Company, Houston, TX, ISBN 0-88415-661-3
Heinz P. Bloch is a professional engineer with over 43 years of experience in reliability engineering and maintenance cost reduction. He has written 14 comprehensive books on these subjects and continues to advise process plants worldwide on reliability im-provement and maintenance cost-reduction opportunities.