Benchmarking maintenance practices at North American paper mills.
Mill profiles of survey respondents
The General Mill Information section of the survey focused on mill profiles (e.g., grades, production volume, number of employees, etc.) and select maintenance metrics (e.g., maintenance spending, scheduled and unscheduled downtime, preventive and predictive maintenance, etc.) Table 1 lists the types of mills that responded to the survey. Many of the responding mills could be grouped by a single product line or grade. However, 39 of the 141 mills either reported producing multiple grades or did not match any of the 10 identified types. Mills ranged in tonnages from less than 10 tons/year to 1 million tons/year, and from less than 100 to greater than 1200 employees.
Maintenance budget and organization
Figure 1 depicts the total maintenance budgets for the responding mills. Maintenance budgets at 44% of the mills amounted to US$ 5 million per year or less. The remainder of the responding mills (56%) reported maintenance budgets exceeding US$ 5 million per year. As expected, labor expenditures represented a significant portion of the maintenance budget. Figure 2 shows that labor expenditures equaled 30% to 40% of total maintenance budgets at more than two-thirds of responding mills.
[FIGURE 1-2 OMITTED]
Respondents reported that the number of maintenance employees equaled an average of about 20% of total mill employment. This percentage dropped slightly from the 23% and 22% reported respectively for North American mills in 1986 and 1987 surveys conducted by Smith and Carpenter  and by Smith . Three-fourths of the respondents indicated the maintenance superintendent reported through an engineering manager, maintenance manager, or the combined function of manager of engineering and maintenance.
Preventive and predictive maintenance, CMMS
Since earlier paper industry surveys, mills have made substantial progress in the areas of preventive maintenance (PM) and predictive maintenance (PdM). A decade ago, preventive maintenance was becoming a recognized best-practice, and predictive maintenance was seen by most as a "nice theory." In fact, the earlier surveys did not even mention PdM. However, with advances in sensor technologies and reduced costs of measuring and diagnostic equipment, PdM has become practical and cost-effective. In 1986 and 1987, preventive maintenance ranged from 9% to 22% of the total maintenance budget. In contrast, a few of the recently surveyed mills indicated budgets for both PM and PdM covered up to 100% of the total maintenance budget, with an average of nearly 4(1% (Fig. 3). The industry average for combined PM and PdM expenditures in 1997 was 52%, as reported by Gulati . According to the Marshall Institute , world-class organizations spend about 70% on PM and PdM initiatives.
[FIGURE 3 OMITTED]
Today's maintenance organizations rely on Computerized Maintenance Management Systems (CMMS) to track machine repair histories, schedule PM tasks, and provide a means of cost control. A survey conducted by Harrison  reported about 85% of mills have CMMS systems; the 1997 TAPPI survey showed this level to be unchanged. The Plant Maintenance Resource Center  reported industry averages of 90%.
Work order completion
A commonly used key performance indicator (KPI) for an effective CMMS system is the number of scheduled work orders completed on schedule. According to Smith and Black , for an organization with "best maintenance practices," this KPI should be 90%. As shown in Figure 4, very few mills reported such high values. Of the respondents, 83% said they documented findings when scheduled PM and PdM were completed and 88% of mills maintained repair histories for critical equipment. For an organization to be considered world-class, these values need to be near 100%.
[FIGURE 4 OMITTED]
Tracking and controlling unscheduled downtime
Most mills ranked quite high in their ability to control unscheduled downtime, with 72% of responding mills having less than 5% unscheduled downtime and 51% of the total reporting less than 3% (Fig. 5). Not illustrated in the charts, 9% of mills indicated the amount of unscheduled downtime was increasing, whereas the vast majority (91%) reported it to be declining. Mills have been tracking unscheduled downtime for a long time, with 34% having more than 10 years of records. Surprisingly, most mills were seemingly lax in follow-up analysis. Only 26% of the responding mills had a system in place to analyze the causes of downtime and initiate corrective actions. Only 12% of mills indicated the amount of unscheduled downtime was increasing, whereas the vast majority (88%) reported it to he declining.
[FIGURE 5 OMITTED]
The first line of defense for avoiding reactive maintenance is the operator. If properly trained and with adequate experience, operators can notice "out of the ordinary" changes in a machine and can perform minor maintenance operations (e.g., cleaning, oiling, tightening, and inspection). These operator-performed routine maintenance functions can mitigate system failures and postpone the need for major maintenance until it can be properly scheduled. Under the Total Productive Maintenance (TPM) method, operator involvement in the overall maintenance strategy is mandatory. Figure 6 shows that significant operator performed maintenance was practiced by only a small traction of the mills surveyed. These numbers can be compared to those reported by Harrison , the averages of which range from 6% to 29%. For a world-class organization, Connaughton 181 reports 30% of minor maintenance work is accomplished by production personnel.
[FIGURE 6 OMITTED]
The Mechanical section of the survey covered select programs and practices used by mills for maintaining mill equipment. Programs and practices included vibration analysis, lube oil analysis, vessel inspections, and alignment checks. The survey also asked mills to provide input on the sources of maintenance that caused repair problems in the mill.
Sources of repair work
Nearly two-thirds (62%) of those responding reported mechanical wear as the greatest source of maintenance (Fig. 7). Corrosion was the second leading cause of mechanical maintenance, cited by 38% of respondents. Abuse ranked third at 33%. Only 21% of mills cited lack of preventive or predictive maintenance procedures and 13% cited vibration. Fewer than 10% of respondents noted erosion, secondary maintenance, or impact as causes for mechanical maintenance.
[FIGURE 7 OMITTED]
Mechanical programs and practices
The survey requested comments on the use of specific mechanical maintenance programs and practices. Respondents selected from 19 different programs, including vibration analysis, noise inspections, lube oil analysis, alignment checks, etc. The survey also asked which programs provided the best value and the least value. Respondents could identify multiple programs used by their mills and could also select multiple programs as providing the best or least value. Table 2 lists the top five best value programs. Table 3 lists the five programs ranked by respondents as least valuable. There appeared to have been a correlation between use of a practice and perceived value of the practice. For example, of those mills reporting to use vibration analysis, 97% claimed the use of vibration analysis provided a best value and 1% claimed it provided a least value. Lube oil and wear analysis was the second best valued program at 64%. In contrast, only 5% of the mills that followed a documented foundation inspection program considered it a "best value." Coincidentally, it was also the program reported as providing the least value by 60% of mills following the practice.
The survey sought further details on vibration analysis programs. As expected, mills paid most attention to equipment in the paper machine building and stock preparation areas. An average of 1939 vibration points monitored in the paper machine area was reported by 90 respondents. However, the range in the number of points monitored varied greatly--from a low of 10 to a high of 11,730. Twelve mills reported monitoring 5000 or more vibration points in the paper machine and stock preparation area and 13 mills reported monitoring 100 or less vibration points. The survey asked respondents about the frequency of vibration monitoring. Figure 8 shows the results. Of 131 responding mills, 54 (41%) use a monthly routing interval, with 80% of mills monitoring vibration points on a monthly or shorter frequency.
[FIGURE 8 OMITTED]
Seventy-nine percent of mills reported they establish a signature baseline for new equipment, while 42% reported they outsource vibration analysis activities.
Pressure vessels are a critical safety and operating concern in all mills. Supporting this fact, the survey found that 84% of mills regularly performed some method of non-destructive examination (NDE). Figure 9 also shows the most frequently used NDE methods. The most favored of NDE--used by 78 (86%) of 91 responding mills--was ultrasonic testing. Other popular NDE methods were magnetic particle, performed by 34 mills (37%), and liquid penetrant, performed by 31 mills (34%). Fewer than 10% of mills performed radiography and acoustic emission testing. No mills reported using eddy current testing.
[FIGURE 9 OMITTED]
Lube oil sampling
Figure, 10 shows the frequency of lube oil sampling. Two-thirds (66%) of respondents reported performing lube oil sampling on at least a quarterly basis. More than one-third (37%) of respondents also reported using ferrography at their mills.
[FIGURE 10 OMITTED]
Drive coupling alignment
The survey asked mills about alignment of drive couplings. Most mills (74%) had migrated to laser alignment by 1997. The remaining mills (26%) indicated they used dial indicator and feeler gages as their method for aligning drive couplings. Mills were given four choices as their source of alignment specifications: OEM, local mill, consulting, or corporate engineering. Most mills (60%) used specifications generated at the local mill level. Thirty percent of mills used specifications provided by OEMs. Less than 10% of respondents indicated they used specifications provided by either consultants or corporate engineering.
Impact of PM and PdM on unscheduled downtime
Most respondents indicated that PM and PdM programs had some positive impact in decreasing unscheduled downtime. PM and PdM programs and activities significantly decreased unscheduled maintenance downtime for 111 of 137 respondents (81%). TWenty-four of 137 respondents (18%) reported that PM and PdM programs and activities had only a marginal impact on decreasing unscheduled maintenance downtime. Only 1% of the mills reported that mechanical preventive and predictive maintenance procedures had no effect on decreasing unscheduled maintenance downtime.
The Electrical section of the survey queried mills on their use and perceived value of twenty-two testing and diagnostic procedures. It also captured the frequency of certain procedures, the age of mill electrical systems, and the perceived impact of PM and PdM on unscheduled mill downtime.
Electrical programs and practices
Table 4 lists the top five best value electrical maintenance programs. Walk-down inspections was the next best value program (43%). Table 5 lists the five programs that respondents felt gave the least value. As observed with mechanical maintenance practices, the increased use of a specific practice generally correlated with its perception as a "best value."
Age of mill electrical systems
The average age of a responding mill's electrical system was 20 years. Although most mills (84%) reported the average age of their electrical systems at 30 years or less, 62% of mills reported having at least one electrical system greater than 30 years old. Nine mills reported having at least one system greater than 70 years old.
Frequency of electrical system testing and diagnostics
Table 6 captures the frequency of use of some of the best-used practices. Eighty-six percent of mills reported using infrared testing at least on an annual basis. Although load testing of switchgear was used by 47% of mills, nearly 70% of mills indicated a time period ranging from one to five years for load testing main switchgear. (This inconsistency may be partially attributable to respondents' interpretation of the definition of "regularly"). More than two-thirds (70%) of mills using calibration of protection relays reported performing it at two- to five-year intervals. Over one-half of mills using transformer oil analysis reported performing the analysis on an annual basis.
Of the responding mills, 90% were using some level of infrared testing as a preventive maintenance technique. Of those 127 respondents, almost all used this practice in motor control centers (98%) and on high voltage switchgear (91%). While these numbers varied by equipment type, only 44% of respondents actually used infrared testing on "all" of the listed categories (mainly due to the low percentage of mills using this technique on powered equipment). Most of the mills (84%) had this service performed by an outside contractor.
Impact of electrical preventive and predictive maintenance on unscheduled downtime
The results mirror those found for mechanical-related preventive and predictive maintenance activities. Most respondents indicated that PM and PdM programs had some positive impact in decreasing unscheduled downtime. PM and PdM programs and activities significantly decreased unscheduled maintenance downtime for 92 of 131 respondents (70%); 36 of 131 respondents (28%) said such programs and activities only marginally decreased unscheduled maintenance downtime. Only 2% of the mills reported that electrical PM and PdM procedures had no effect on decreasing unscheduled maintenance downtime.
Mill managers and maintenance managers can use these results to broadly assess their mill maintenance practices in certain areas. They can also compare results to international surveys reported by Arommaa and Klarin . While this survey's results do not suggest specific performance measures, managers may want to look into areas where significant gaps exist between their individual mill maintenance practices and the survey results. However, caution and judgment are required when comparing a single mill's performance or practices with the survey results.
It is difficult to determine a common denominator for comparing individual mill practices with the results of a broad survey such as this. Several additional attributes not included in this survey would likely influence an individual mill's maintenance program, such as age of the mill, timing and extent of last mill modernization, number of pieces of equipment, degree of vertical integration, frequency of grade changes, profit margin of each grade, maintenance philosophy, etc.
TABLE 1: mill types represented in the survey. Mill type Number Market pulp 9 Boxboard 4 Recycled paperboard 26 Linerboard 13 Paperboard plus linerboard 9 Specialty 16 Tissue 5 Toweling & tissue 6 Free sheeting 7 Newsprint/directory/groundwood 7 Three or more product lines 15 Could not be grouped 24 Total 141 TABLE 2: Mechanical maintenance "best value" programs. Mechanical maintenance Percent of program of best value program users Vibration analysis 97% Lube oil & wear analysis 64% Walk-down inspections 61% Alignment checks 54% Temperature inspections 52% TABLE 3: Mechanical maintenance "least value" programs. Mechanical maintenance Percent of program of least value program users Foundation inspections 60% Noise inspections 29% Belt tension inspections 28% Pump dismantel inspections 25% Over-pressurization 17% TABLE 4: electrical maintenance "best value" programs. Electrical maintenance Percent of program of best value program users Infrared measurements 88% Transformer oil analysis 68% Calibration of protective relays 63% Insulation resistance testing 54% Switchgear load testing 50% TABLE 5: electrical maintenance "least value" programs. Electrical maintenance Percent of program of least value program users Insulation resistance of buried cables 45% Motor surge testing 27% Polarization Index 23% Power factor test--transformers 23% Lightning arrestor testing 23% TABLE 6: Frequency of selected electrical system testing and diagnostics. Procedure 6 mo 1 year 2 year 3 year Infrared electrical system 39 75 12 3 Load test main switchgear na 47 12 19 Calibrate protection relays na na 49 24 Transformers--Screen test insulating oil na 78 11 4 Transformers--Gas analysis on oil 3 67 9 6 Procedure 5 year Infrequent Never Infrared electrical system na 2 2 Load test main switchgear 20 19 19 Calibrate protection relays 20 25 13 Transformers--Screen test insulating oil 9 21 na Transformers--Gas analysis on oil 7 27 na
IN THIS ARTICLE, YOU WILL LEARN:
* Methods mills rely on most to assess their maintenance efforts
* Which mechanical end electrical programs mills consider the "best value."
* Where mills are spending their maintenance dollars.
* Enter "maintenance" as the search term for selected sections of TAPP's web site (www.tappi.org to find dozens of articles, conference presentations, TIPs, and other resources on this topic
[1.] Smith, K.E. and Carpenter, B.A., Pulp Paper 60(9): 60(1986).
[2.] Smith, K.E., Pulp Paper61(2): 97(1987).
[3.] Gulati, R, "Maintenance benchmarking: Best practices and survey results," MARCON 2000 Maintenance and Reliability Conference Proceedings, University of Tennessee, Knoxville, Tennessee, USA, 2000.
[4.] Marshall institute, Preventive/Predictive Maintenance, The Marshall Institute, Raleigh, North Carolina, USA, 1998
[5.] Harrison, A., Pulp Paper67(2): 43(1993).
[6.] Plant Maintenance Resource Center, "CMMS Implementation Survey Results-2000," avail able electronically at <http://www.plantmaintenance.com/artcles/CMMS_survey_2000.shtml>.
[7.] Smith, R. and Black, C.. "How to develop and implement multiskilled training," Proceedings of the 8th Annual Conference of the Society for Maintenance & Reliability Professionals, SMRP, Knoxville, Tennessee, LISA, 2000.
[8.] Connaughton, G. "The state of the art of maintenance in North America," Proceedings of the 8th Annual Conference of the Society for Maintenance & Reliability Professionals, SMRP, Knoxville, 2000.
[9.] Arommaa, J. and Klarin, A., eds., Materials, Corrosion Prevention and Maintenance, Fapet Oy, Helsinki, Finland, 1999.
The complete results of this survey were first presented by the authors at the TAPPI 2001 engineering/Finishing & Converting Conference.
About the authors: Mickey Bevis is with Smurfit-Stone Container Corporation, Panama City, Florida, USA; Michael Hayes is with Risk Consultants, Miamisburg, Ohio; Roland O'Brien-Bills is with Inspection Services (A/S), Inc., Gloversville, New York; Bill Rodrigue is with Fluor Corporation, Greenville, South Carolina; and Jerry Kahn, PE, CMRP, is with Siemens Energy & Automation, Inc., P.O. Box 2897, Peachtree City, GA 30269; email email@example.com
|Printer friendly Cite/link Email Feedback|
|Publication:||Solutions - for People, Processes and Paper|
|Date:||Aug 1, 2002|
|Previous Article:||Hot show hints industry is warming up: the mood at the SPCI pulp and paper exhibition often provides a guide to the industry's direction--and this...|
|Next Article:||How TAD has come of age.|