Wrench time

June 30, 2006
Industry experts have for many years pointed to the low productivity levels in maintenance departments of most companies around the world. They cite anywhere from 30% to 50% as an average for “wrench time,” the productive time technicians spend actually repairing or replacing equipment, as opposed to walking to the job, receiving instructions, waiting for parts and other productive or non-productive activities.

Industry experts have for many years pointed to the low productivity levels in maintenance departments of most companies around the world. They cite anywhere from 30% to 50% as an average for “wrench time,” the productive time technicians spend actually repairing or replacing equipment, as opposed to walking to the job, receiving instructions, waiting for parts and other productive or non-productive activities. Do you know the average wrench time for your maintenance department and how it compares to others within and outside your industry?

View related content on PlantServices.com

There are, of course, those senior managers who say they would be thrilled to hear that their maintenance folks are sitting idle most of the time. This “Maytag repairman” image of a maintenance department is analogous to a fire department, where the fewer fires there are, the better. But a maintenance department can be far more productive in so many ways than putting out fires. Maintenance could be implementing preventive and condition-based maintenance programs, participating in process and quality improvement projects, and working on capital improvement initiatives. Maintenance workers can upgrade their skills, or train others such as junior staff and even operators.

In my view, it’s imperative for plant professionals to be aware of the productivity level in your maintenance department and to understand which areas are strong and which require improvement. You can derive this information and ensure that productivity is optimized in a variety of ways.

The most obvious source of information is the CMMS, as most have a wealth of reporting tools available. For example, a CMMS can track actual hours charged to a work order versus planned hours, or report on actual hours versus budgeted hours for a given cost center. Through mobile technology, the CMMS can monitor progress of individuals working on various jobs, or even their physical location using GPS devices.

But the picture isn’t quite complete with these reporting tools. For example, if eight hours are charged to a given work order, how many hours were spent actually working the tools? This level of detail is hard to capture using a CMMS, and it isn’t easily verifiable. Work measurement techniques, such as predetermined or video-based time and motion study, can get at the specifics. However, these tools can be quite impractical for maintenance workers that move around the facility so much and do so many non-repetitive tasks. Work sampling is one tool that is effective in these situations.

Work sampling — what is it?

The theory of work sampling is derived from the fundamental laws of statistics and probability. If you take enough random observations of work over a somewhat typical time period, the number of observations of a given activity or type of activity divided by the total number of observations should approximate the percentage of time that activity or activity type actually occurs. This is more easily understood with a simple example.

Suppose an analyst makes 1,000 observations at random intervals over several weeks. The analyst might observe that a given maintenance crew was working productively “on the tools” in 300 of the observations and, in 700 instances, the crew wasn’t “on the tools” for miscellaneous reasons. Then, it would be reasonably certain that average wrench time for the crew was 30%.

It also should be noted that the more observations taken, the greater the accuracy of a given activity breakdown at a given confidence level. Doubling the number of observations gives you a greater certainty that the observed wrench time through work sampling reflects the true wrench time.

Similarly, if you increase the number of productive and non-productive activity types that are distinguished upon observation, more observations are required to maintain the same level of accuracy at a given confidence level. In the previous example, if you were to distinguish between different types of wrench time (eg, time to diagnose a problem, time to fix the problem), and different categories of “non-wrench time” (eg, preparing a job site, data entry), you would need many more observations to maintain the same certainty for any activity.

Perform your own work sampling study

The first step is to determine your objectives, scope and resource requirements. As with any project, it’s critical to clarify:

  • What you hope to accomplish (determine the level of productivity of the maintenance department and identify nonproductive activities that can be curtailed).
  • What is in and out of scope (physical locations, individual people and equipment, activities to be observed, and relevant time periods).
  • What internal or external resources are available that have the skills to conduct work sampling.

First, it may be useful to conduct a pilot work sampling study of a few hours spread randomly over a few days. This gives you a feel for the most appropriate categories of productive and nonproductive work; the problems that might be encountered, such as the difficulty in finding people at the designated random time; the reaction of people being observed; and so on.

Second, determine the number and frequency of observations required. Once you’ve determined the number of work categories to be analyzed, use the standard statistical tables, charts and formulae to determine the correct number of observations required for a given level of certainty. If time is of the essence and the number of observations is large, consider throwing more analysts at the problem.

For example, suppose you need the results of the work sampling in a month (ie, 20 working days). Furthermore, suppose you find you need to take 10,000 random observations to achieve your desired accuracy. A single analyst would need to make 500 observations per day, a rate that is likely too difficult to maintain. It might be necessary to let the deadline slip, increase the number of analysts, reduce the number of work categories studied, ease off on the accuracy required, or find some way to increase the number of observations one analyst can make each day.

Third, prepare a communication plan. One of the most important steps is to ensure those being observed are properly informed as to the objectives, scope, methodology and timing of the work sampling exercise. Also, the analysts should be introduced to ensure familiarity when entering the study area. Of prime importance is that anyone participating in the study should understand the cost/benefit of the study from their perspective, not just what’s in it for management.

Fourth, make your observations and analyze your data. You can use a simple tick sheet to conduct the study, where each work category has its own column and each row represents a pre-established random time that an observation is to be made. For each designated time, the analyst puts a tick in the appropriate column depending on the activity observed. The ticks are then added for each column, and percentages are calculated. If, for example, 200 of the 1,000 total ticks were in the column marked “preparation for shift start or end,” we know that on average 20% of the time is spent on that activity.

Lastly, identify and implement your improvements. Use the percent breakdown of work to start a discussion with management and those who were observed during the project. Look for opportunities to reduce the non-value-added activities. You’re now in position to prepare and execute your implementation plan.

E-mail Contributing Editor David Berger, P.Eng., at [email protected]

Sponsored Recommendations

Reduce engineering time by 50%

March 28, 2024
Learn how smart value chain applications are made possible by moving from manually-intensive CAD-based drafting packages to modern CAE software.

Filter Monitoring with Rittal's Blue e Air Conditioner

March 28, 2024
Steve Sullivan, Training Supervisor for Rittal North America, provides an overview of the filter monitoring capabilities of the Blue e line of industrial air conditioners.

Limitations of MERV Ratings for Dust Collector Filters

Feb. 23, 2024
It can be complicated and confusing to select the safest and most efficient dust collector filters for your facility. For the HVAC industry, MERV ratings are king. But MERV ratings...

The Importance of Air-To-Cloth Ratio when Selecting Dust Collector Filters

Feb. 23, 2024
Selecting the right filter cartridges for your application can be complicated. There are a lot of things to evaluate and consider...like air-to-cloth ratio. When your filters ...