John,

With regard to your question on measuring wrench time, I look at such measurement not as a single task, but as one data point that needs to be validated. It is much too easy for people to manipulate a study if you only use a sampling approach. I much prefer using a combination of sampling and measures of performance that are related to wrench time. I prefer this method because, let’s say, your workforce knows you will be doing a wrench time study. They may attempt to select or prepare for certain work tasks, and they may have everything organized for that job to a much higher degree than normal. This skews the data and will return a much higher wrench time value than when the study is conducted without forewarning.

When I say I prefer the use of a combination of sampling and measures, I am referring to calculation of measures that can be used to paint a picture of whether the wrench time numbers make sense or not. For instance, if your wrench time study resulted in a wrench time of 55% (very high) but you were at something like 70% unplanned maintenance, I would say that the numbers for the wrench time study are not likely to be valid. Examples of measures I would consider to be good cross references to wrench time include:

• Percent of planned and unplanned labor hours versus total available labor hours
• Number of man-weeks of backlog maintenance that is planned and ready to schedule
• PM/PdM completion percentage
• Production unplanned downtime
• Percent of re-work (jobs repeated on the same equipment within a period of time, such as 6 months or a year)

So with the understanding of issue of having corroborating data, then a wrench time or barrier study can be conducted. I prefer to call these barrier studies because the point you want to make with the foremen and workforce is that the study is not concerned with how they personally carry out their job tasks. I don’t care if it takes a worker 20 minutes or 2 hours to remove a coupling between a motor and pump assembly. Rather, the study is being conducted to see what barriers there are that keep the workforce from doing their job efficiently.  

The idea is to work smarter, not harder. Most workers don’t like travelling across the refinery to get to the job site only to find that the job plan did not identify the actual work that needed to be done, that they didn’t bring the right tools for the job, that the repair parts were not correct in type or quantity, etc. They understand that not having the job well planned will lead to their having to go back to the shop or stores counter repeatedly, and most likely, they will be disrupted by some other higher priority work.

A barrier study can be conducted by several different methods. There are vendors who can supply portable devices that require the worker to select the activity they are doing at random times during the day. Over time, you get a statistical average on where time is being spent. I don’t like this approach because it is too easy to skew the data if the worker doesn’t want to cooperate with the study intent. I prefer “attaching yourself to a work order”. Be in the shop when the work is handed to the craftsman, then walk around with the craftsman for the morning or the whole day to observe the things that get in the way of them doing the work. Write these barriers down, including the time it took away from their getting the work done. Repeat the process for at least a dozen work orders and with a variety of workers. The more you can get done, the better the statistics. Categorize what the worker is doing into five to eight categories; i.e. doing value added work, getting instructions/clarifications, travelling, break/lunch, waiting for parts, waiting for tools, waiting for operations to turn over equipment, etc. Then, graph the categories; value-added work is wrench time.

Barrier studies can create problems if your organization does not have a lot of trust between the workforce and management. Spend some time with labor leaders and those in the workforce who have influence over the rest of the workforce. Explain what you’re doing and why. Share the data with the worker that is being followed so they see what you’re writing; this will allow them to be less anxious about the study.  

Also, have a plan for what you intend to do if you identify and resolve inefficiencies. The workforce will always believe you are looking at efficiency gains as a means to reduce head count. In my experience working across refineries, chemical plants, aerospace manufacturing, mining, facilities and a number of other industries, organizations have usually reduced their head count more than they should have, given their performance level. I usually coach my client leadership teams that they should use the benefits from efficiency to reinvest in proactive reliability activities; PM/PdM optimization, FMECA or RCM, root cause snalysis, etc. These continuous improvement activities will have a much higher ROI than cutting head count.

Hope this helps, and if you need support in performing these activities, please keep Alidade MER, Inc. in mind. We would be happy to support your efforts.

-Tom Moriarty

Read the article: Practical labor rates - Tom Moriarty, P.E., CMRP, contributing editor, says they exceed base pay rates and, therefore, should be controlled.