Measuring Reliability

Reliability of capital equipment is critical to the success of any manufacturing operation,

mobile fleet, facility operation or any combination thereof. How do you accurately go
about determining equipment reliability? Only when it is clearly understood that
achieving a high level of equipment reliability at a reasonable or competitive cost is
ultimately the job of everyone in the plant / facility; not just the maintenance department
– can accurate and effective reliability measurement schemes be developed. 

While it is true that maintenance performance should be measured in terms of reliable

capacity within predetermined cost performance targets and estimated annual and
monthly budgets respectively, isn’t this also true of operations? Yet most people still
believe that maintenance alone controls reliability.

When you consider that operations or production controls the equipment through their
operation of it, it becomes apparent that they in fact actually have a much greater
influence on equipment condition and therefore they also must be held accountable for
that equipment’s reliable performance.

The fact is that in actual practice when operations and maintenance are jointly
responsible for equipment reliability the opportunity to improve is significantly increased.

When measuring reliability, consideration must be given to the data contained in

management reports. Specific performance indicators must be carefully established,
measured and reported in order to provide operations and maintenance supervision
with pertinent and detailed information that they can effectively use to manage their
business unit.

These indicators must be consistently and accurately measured and reported.

Additionally, a higher level of detail, or summarization, of these indicators is required for
senior or executive management and should be consistently reported weekly or
monthly, with any supporting details available on request or during analysis.

Much of the data necessary to measure factors influencing reliability and budget
performance is available directly, or can be derived from, CMMS (Computerized
Maintenance Management System), assuming it is properly implemented and effectively
utilized. Reliable capacity data can be derived from operations records and electronic
manufacturing control systems.
Parts and materials usage data can be derived from inventory control systems and
purchasing systems in the case where these functions are not performed within, or
linked to, CMMS. Other useful data can be derived from accounting systems, quality
systems, time-payroll accounting and human resource systems.

Accurately establishing current baselines of performance is an important step to begin

with. Once baselines are established, identify both short and long-range targets for all
reliability measurements that are defined. You must know where you are and where you
want to go, or you may not recognize it when you get there.

“Distinguish Management Data and Reporting Information”

1. Budget performance
Annual budgets of anticipated reliability activities should be estimated for each
equipment item or asset each year and continuously monitored. A well-conceived
equipment hierarchy in the CMMS that breaks down the facility by category can
facilitate this process. Some recommended breakdowns are:

 by area
 by system
 by machine or asset
 by component

Cost data is then systematically and accurately collected for all labor, parts, materials
and subcontracts at the machine or asset level via the Work Order System. Whether
work is performed by operations or maintenance, it should be captured via the Work
Order System.

The CMMS databases must be completely and accurately developed to provide the
basis for this approach. This level of detail is critical for supervision to establish and
refine the proper level of budget control and identification of high cost equipment issues.
This information can be rolled up to a higher level by area, by system, or for the total
plant/facility for executive or senior management reporting.
 In proactive maintenance operations, a total planned cost estimate for each
maintenance job is prepared whenever possible. The estimate should include labor,
parts, materials, purchase requirements and subcontracts.

Comparisons of actual cost to estimated costs by job provides the basis to measure
performance of maintenance crews or operations personnel as well as a means to
control budgets and continuously refine future cost estimates.

2. Cost per Unit of Output

Measuring manufacturing and maintenance costs in terms of output of the facility can be
a very effective way of managing the business and monitoring budgets.

These measures can be benchmarked to an industry standard or an internal standard

and permit a concise and clear measure of performance, particularly if the data is
trended utilizing a thirteen month rolling average trending analysis. Some examples of
costs in terms of plant / facility output are:

 cost per pound

 cost per ton
 cost per case
 cost per square foot (facilities)
 cost per mile (mobile / fleet equipment)

3. Capacity Measurements
Typically, the negative influences on reliable capacity are time losses, speed losses,
scrap losses at start-up and quality losses caused by improper design, operation or
maintenance. Take note that other departments within the organization can influence
any of these parameters, either directly or indirectly.

Many organizations focus solely on unscheduled downtime and proceed to measure it

inaccurately to make the reported numbers visually look good or only focus on
maintenance downtime, ignoring operations or scheduled downtime.
“The cost of downtime is the same, no matter what the cause”

A high level of unavailability without consideration of speed and quality losses can prove
quite costly. The ultimate capacity measurement method is monitoring “Overall
Equipment Effectiveness” (O.E.E.)

OEE = (Availability) x (Efficiency) x (Quality)

where:
Availability = total time minus delays divided by total time
Efficiency = actual speed divided by design or optimum speed
Quality = total output minus scrap or waste divided by total output.

This method lends itself to a manufacturing operation and it cannot effectively be

utilized in mobile fleets or facilities. This method of measuring capacity, when cost of
loss to the operation is analyzed for each parameter, can effectively be utilized to
evaluate the risk of not performing the optimum level of maintenance or operating the
equipment ineffectively as well as to justify the investment to improve performance.

4. Work Backlog
The total amount of work that has been requested or identified determines the
maintenance backlog. All categories of work that estimates have been prepared for and
trends of unplanned or emergency work and maintenance-performed projects define the
backlog. Backlog must be segregated into logical work types, and then understood and
managed in order to achieve effective resource utilization.

“Resources Must be Balanced with Work Load”

This information is very useful in determining the appropriate staffing level, overtime
decisions or subcontracting.

Work orders must be segregated into “ready to schedule” and (with status) “pending
action required” (i.e., waiting on approval, waiting on parts/materials, waiting on
shutdown, etc.). Including the status of pending categories allows continuous monitoring
and can eliminate bottlenecks in the work control process.

Total work backlog should not exceed 4-6 weeks and “Ready to Schedule” should not
exceed 2-4 weeks.

5. Schedule Compliance
Each job should be scheduled, usually in the week prior to the week of execution, for a
specific timeframe during the schedule period. Each job should also have a defined
scope of work; parts and materials should be identified and delivered to the job site, all
tools and equipment made available and any permits or special considerations
prepared.

All available labor should be scheduled each week (capacity scheduling) based upon an
analytical method of establishing the number of available labor hours. Supervisors, who
are responsible for tactical work execution, should be held accountable for schedule
adherence at the 85-90% level each week.

Professional Planning/Scheduling can provide the means to assign each individual

worker a full workload each week and the supporting documentation to hold them
accountable to achieve it. Schedule compliance = actual hours required divided by
scheduled hours.

6. P.M. Compliance
An effective Preventive Maintenance program is also critical for achieving a reliable
operation. The success of a P.M. program is dependent upon timely and consistent
execution, continuously evaluating the results and ensuring that a dynamic approach is
taken that focuses on changing the frequency or content of the P.M. when the desired
level of reliability is not attained.

Operations and Maintenance should be jointly engaged in executing the PM program.

Consideration should be given to optimizing the use of condition-based scheduling
techniques as opposed to traditional time-based scheduling. Most conditiondetermining
routines are performed with the equipment operating, which can significantly reduce
scheduled downtime to perform time-based tasks.

P.M. Compliance is simply a measure of how many of the scheduled P.M.s are actually
completed on time or P.M.s completed divided by the number of P.M.s scheduled.
Using the 10% rule when determining on time completion; is a good way to determine
timely completion.

The 10% rule is simply allowing a range of time for completion defined as: P.M. should
be completed +/- 10% of the scheduled frequency, i.e., monthly would be +/- 3 days,
daily would be +/- 2.4 hours, etc.

This is a high level measurement that should be reported to management on a routine

basis. P.M. Compliance should exceed 90%, and should not be confused with P.M
completion rate, meaning that the P.M. is completed, within schedule or not, should be
100%.

There are a great many performance measures that can be utilized. Use caution to
avoid the pitfall of most businesses. They measure too many performance indicators
and then have less time to spend on analysis and on taking action to improve
performance.

When utilizing any performance measure, a target must be established that defines best
practice or ideal performance and employees must be educated on what these targets
mean and how their performance influences the wellness of the business.

The measures discussed here are not intended to be an all-inclusive list, but a good
place to start in measuring – and improving – reliability in your plant or facility.