An Integrated Approach to Lubrication Excellence at BHP Steel

Dr. Jian Ding, BHP Research & Technology Development
Tags: industrial lubricants

Lubricated equipment at BHP Steel is massive in quantity and critical to production. It covers millions of bearings, gearboxes, heavy-duty diesel engines, and hydraulic systems throughout hundreds of BHP plants worldwide. An investigation conducted by BHP Research & Technology Development in 1997 revealed that the potential savings system-wide in the material and repair costs of lubricated components (including lubricant consumption but excluding downtime loss) is as much as $30 million per year.

DuPont found that the contribution of maintenance improvement to equipment uptime is 10 times that of the maintenance cost savings, to further raise the stakes. Due to its enormous cost reduction and profit-producing potential, a new program called the Lubrication Maintenance Improve-ment (LMI) has been targeted as a strategic business imperative at BHP. As such, a lubrication condition assessment project was initiated to focus upon extending the life of lubricated equipment. This article details the results of the lubrication condition assessment project including two successful applications where lubrication contamination problems were effectively rectified.

Assessing Lubrication Condition
In order to understand and quantify lubrication problems, a comprehensive lubrication condition assessment for critical equipment was conducted.

Lubrication conditions were assessed using three condition indices, 1) Oil Condition Index, 2) Solid Particle Contamination Index and 3) Machine Wear Index. The three indices are described further below:

• Oil Condition Index - This is an integrated indicator representing changes in viscosity, water (%), oxidation, additive depletion, etc., used to identify the lubrication problems such as incorrect lubricants, over-extended oil drain intervals, fluid contamination, etc.

• Solid Particle Contamination Index - This index is measured by automatic particle counter and is represented by the ISO Cleanliness Code.

• Machine Wear Index - This is an integrated assessment based on comprehensive analysis consisting of:

• SOA (spectrometric oil analysis, i.e., wear metals) for detecting polishing, rubbing and corrosive wear problems

• PQ and/or Direct Reading Ferrograph (DR) for quantitative wear measurements of larger ferrous particles

• Microscopic filtergram and/or ferrogram wear particle analysis to confirm machine wear modes and wear severity levels

A key step in the assessment is the process of setting appropriate target levels for various lubricated machines based on their structure, criticality, age, operation and environment, etc. For example, the target cleanliness level of a gearbox might range between ISO 14/11 and 20/18 based on its downtime and replacement costs, speed and load, component types, filtration effectiveness and risk for contamination ingress. After establishing target levels, three lubrication condition severity levels, On-target, Alert and Warning, were specified based on a comparison of actual conditions to the target conditions.

On-target - The condition is within the target limits;

Alert - Outside the target limits by as much as 1-3 levels/grades. It is usually an "abnormal" condition but no immediate actions to be taken; and

Warning - Outside the target limits by greater than 3 levels/grades, a "severe" condition that can lead to significant life reduction and failure.

To differentiate machine wear severity levels consistently, a wear severity code system developed by Lubrosoft was deployed based on wear particle concentration, morphology and size distribution.
An assessment of 66 critical machines was conducted. The results of that assessment are summarized in Figure 1. From this assessment the team reached the following conclusions:

• A total of 82% of oil samples are outside the targeted cleanliness levels, highlighting the need for a significant effort for cost effective contamination control.

• The high level of abrasive particle contamination in lubricating oils is causing severe machine wear. This high rate of wear is reducing the life and increasing the failure rate of components, escalating maintenance costs and undermining reliability efforts.

Case I - Life Extension of the Tilt Table Gearbox - Hot Strip Mill

The tilt table gearbox is one of 25 gear units on the centralized lubrication system at the hot strip mill. It consists of a number of helical gears, shafts and rolling element bearings. The average life of the gearbox is currently 14 months. The total replacement costs, including re-builds and shutdown costs, are as much as $125,000 per year. A typical failure of the tilt table gearbox is characterized by Figures 2-1 to 2-4.

The team's previous efforts to improve reliability of the gearboxes were focused on improving sealing of the interface between the shafts and the case to prevent ingress of mill scale, dust and water into the lubricating oil, and on increasing the surface hardness of the gears to reduce wear. However, these actions failed to extend the life of this gearbox.

The new approach focused upon identifying the failure root causes through systematic investigation and integrated measurements as described below:

Participation of overhaul and inspection of the failed gearboxes
The failed components were visually inspected rather than just reading inspection reports and interviewing the maintenance people (the past practice). This step revealed that the excessive wear and scratching of both shaft and case holes is the result of the excessive wear and failure of the rolling element bearings, rather than being caused by water and scale particle ingression. This finding significantly altered the team's attention to the premature failure of the rolling element bearings.

Microscopic examination of the worn surface morphologies of the bearing components
SEM examination revealed that all surfaces of the inner and outer raceways, rollers and cages were characterized by particle-induced wear mechanisms. The results were as follows:

• The inner raceway shows a typical rubbing wear polished by tiny solid particles in the lubricating oil (Figure 3-1).

• Abrasive wear was apparent on the cages (Figure 3-2).

• The outer raceway indicated very early pitting wear, appearing as small surface indentations caused by hard, "clearance size" solid particles in the lubricating oil (Figure 3-3)

• Micro-cracks originated and propagated from the indents (Figure 3-4).

Comprehensive oil & wear particle analysis
An oil sample was taken from the lubricant that remained after overhauling the tilt table gearbox. Another sample was taken from the centralized lubricating system of the finishing mill. Both ferrogram and filtergram analysis suggested that over 95% of solid particles in the gearbox oil sample were ferrous rubbing wear particles less than 10 microns in size (Figure 4 and Figure 5). This finding supports the conclusion that the excessive wear and premature failure of the rolling bearings is caused by the high concentration of abrasive solid particles in the oil.

Comparing the two oil samples' contamination levels
The solid particle concentration in the sample taken for the gearbox sump is 10 times higher than the sample drawn from the centralized system. Likewise, the water content of the sump sample is 28 times that of the centralized lubrication system. Such high particle and water contamination levels can reduce the life of rolling bearings by 80 to 90%. Therefore, the real troublemaker leading to premature failure of the tilt table gearbox was found to be extremely contaminated lubricating oil. Further digging revealed that the root cause for the vast difference in contamination level between the two samples is tied to poor oil exchange efficiency between this gearbox and centralized lubrication system.

Based on the failure analysis and root cause identification, the team recommended that target cleanliness levels in the tilt table gearboxes be strictly maintained. Doing so will extend the life of the system from a current mean-time-between-failure of 14 months to an estimated 36 to 48 months. The team has high confidence in meeting or exceeding this life extension.

Case II - Reducing Failures with Lubricant Contamination Control - The No-Twist Mill
The driving system of the No-Twist Finishing Mill, Newcastle Steelworks operates at very high speeds of over 12,000 rpm and contains a large number of lubricated components including 63 gear-sets, 99 rolling element bearings and 43 Clevite bearings. Due to the system's complex structure and propensity to fail with little warning (due to its high speed), preventive maintenance is often limited to detecting and isolating component failures. To combat the high costs of failure, a proactive maintenance strategy featuring effective lubricant contamination control has been implemented to promote reliability.

Since 1986, failures in the no-twist finishing mill have been significantly reduced due to continuous improvements in lubricant contamination control. The practice has worked in unison with the efforts to rationalize mill operation and maintenance practices, utilization of better seals and various condition monitoring and inspection techniques. Updating the machines filtration systems was a major component of the effort. Before the undertaking, the system was equipped with >30 micron filters. As the system filters were upgraded first to ß12 = 75 (12 micron) in 1986 then ß6 = 75 (six micron) in 1988, the failure rates of the critical Clevite bearings were reduced from 12 failures in 1985, to just one failure in 1988. At the same time, the mill speed was significantly increased from 46 M/S in 1985 to 120 M/S in 1988, placing additional strain on the components (Figure 6).
After achieving world-class performance in productivity and equipment reliability, attention at the rod mill focused on further maintenance cost reduction and failure reduction of the lubricated components in the no-twist finishing mill. A Lubrication Maintenance Improvement (LMI) task was established to identify opportunities to reduce component failure rates and replacement costs for filter elements through further optimization of the existing lubricant filtration systems. The process included the following steps:

• Analysis of the effect of the particles on the various critical components.
• Rationalization of the target cleanliness levels.
• Optimization of filter location selection.

The investigation revealed that most of the particles were rolling scale greater than 10 microns that entered through defective seals with the cooling water. The particles being generated were determined to be very small, posing little threat to the systems. Therefore, filtration was focused on the removal of the ingested mill scale. Based upon analysis, by-pass filtration was deemed most suitable for achieving the organization's goals.

Conclusions
• Lubrication Maintenance Improve-ment (LMI) at BHP has been recognized as a strategic business area that contributes significantly to bottom-line performance through reduced maintenance costs and lost production due to poor reliability of lubricated equipment.

• Solid particle contamination of has been identified as the major cause for premature failure of the gearboxes. The key component in cost effective solid particle contamination control of lubrication systems is to initiate Target Cleanliness Level (TCL) to guide contamination monitoring activities, and to rationalize contamination control systems (filters, breathers and seals etc.).

• An integrated approach plays an important role in understanding lubrication related failures, eliminating the root causes that lead to failure and in developing cost effective solutions for specific plant lubrication problems.

Acknowledgements
The work is supported by Mr. Ron Whitely at the Rod Mill, Newcastle Steelworks, BHP; Mr. Bob King at HSM, Pt Kembla, BHP; Dr. Chao Liu, Department of Mechanical Engineering, Monash University, and Pall Australia.

References
J. G. Ding and J. Rucinski, Cost Effective Maintenance of Lubricated Equipment in BHP Steel, BHP Internal Report, Aug., 1997.
V. J. Flynn, Maintenance Benchmarking and the Evolution of DuPont's Corporate Maintenance Leadership Team, 1997.
J. G. Ding, A Computerized Wear Particle Atlas for Ferrogram and Filtergram Analysis, Technology Showcase 1998, Proceedings of JOAP International Condition Monitoring Conference, Mobile, Alabama, April 20-24, 1998.
J. C. Fitch, Diagnetics, Inc., Oil Analysis and Proactive Maintenance Seminar - A Complete Users Guide for Maintenance Professionals, 1995, USA


About the Author
Create your own user feedback survey