Wrench time: what counts, measuring it with a work sampling study, where the shift goes, and how planning and kitting raise it
Wrench time, or tool time, explained: what counts as direct work and what does not, how to measure it with a work sampling study and how many observations are needed, where a maintenance shift goes, the losses that eat it and what removes each, and how to use the result without turning it into a target. With a drawing of a shift before and after planning and kitting.
Key takeaways
- Wrench time is the share of a tradesperson's paid time spent working on the job, as against travelling, waiting or looking for things. Most studies put it at 25 to 35 per cent where work is largely unplanned.
- It is measured by work sampling: many instant observations at random times, each recorded in one category, so the share of observations estimates the share of time.
- About 900 random observations give an answer within three percentage points, which is enough to act on. Names are never recorded, because the study measures the system, not the people.
- Almost all of the loss is caused by the work system: parts not kitted, permits raised on the day, scope missing, support not booked. Planning, scheduling and kitting remove it. Pushing people harder does not.
- A rise from about 30 to 50 per cent adds the equivalent of more than half again to the crew's productive hours without hiring anyone.
A maintenance crew of twelve is paid for about 1,200 hours a week. A supervisor watching a single morning sees two fitters at the store counter, one at the permit office, three waiting for a crane, and the rest working. Nobody is idle by choice, and everyone is busy, but less than half of the hours are reaching the equipment. That share is wrench time, and it is usually lower than anyone on site would guess.
This guide covers what wrench time is and what counts, how to measure it with a work sampling study and how many observations are needed, where the time goes, and how planning, scheduling and kitting raise it. Wrench time is also called tool time, which is the term in the site's glossary. The figures are commonly reported ranges and illustrations, not a benchmark for any one site.
Where the shift goes
Wrench time is the hours spent working on the job divided by the hours paid. A shift breaks into direct work and everything else: travel, waiting for parts, permits and other people, meetings, breaks, and the slow start and early finish when no job is ready. Some of that is necessary. Breaks and a short pre-start are part of a safe shift. Most of the rest is waste caused by how the work is organised.
The gain from raising it is larger than it looks. A crew at 31 per cent that reaches 50 per cent has increased its hours on the job by about 60 per cent, the equivalent of hiring seven more people into a crew of twelve, without anyone working harder. That is why wrench time is one of the maintenance KPIs worth measuring once or twice a year, even though it is too costly to measure every week.
Low wrench time is almost never people choosing not to work. It is people waiting for something the system did not have ready.
What counts as wrench time
A study is only as good as its categories. They are agreed before the first observation, written down with examples, and kept the same in every repeat so the results can be compared. Preparation done at the job, such as reading the job pack or setting out tools, is usually counted as direct work. Fetching the tools from the workshop is travel.
| Category | What it covers | Treated as |
|---|---|---|
| Direct work | Hands on the job: dismantling, repairing, adjusting, testing, inspecting, and the preparation done at the job itself | Counts as wrench time |
| Travel | Walking or driving to the job, the store, the workshop or the permit office | Loss |
| Waiting for parts or tools | At the store counter, searching for a part, or waiting for one to be delivered | Loss |
| Waiting for permits or isolation | At the permit office, or at the job waiting for an isolation to be applied or verified | Loss |
| Waiting for others | For instructions, another trade, a crane, scaffold, an operator or access to the plant | Loss |
| Meetings and toolbox | Pre-start, toolbox talks, training and other meetings | Necessary, kept short |
| Breaks | Scheduled breaks | Necessary, not a target |
| Late start and early finish | Time at the start and end of the shift with no job under way | Loss |
Measuring it with work sampling
Following one person with a stopwatch is slow, changes how they work, and measures one person on one day. Work sampling measures the crew. An observer makes rounds at random times and records what each tradesperson is doing at the instant they are seen. Each observation is a single category, and the share of observations in each category estimates the share of time spent in it.
The number of observations sets the precision. At 95 per cent confidence, the observations needed are 1.96² × p × (1 − p) ÷ e², where p is the expected wrench time and e is the precision wanted. With an expected 30 per cent, the numbers are below. Plus or minus three points is enough to act on and to compare with a later study.
| Precision | Observations | For a crew of 12 |
|---|---|---|
| ± 5 percentage points | 323 | About 3 days for a crew of 12 at 10 rounds a day |
| ± 3 percentage points | 897 | About 8 days for a crew of 12 at 10 rounds a day |
| ± 2 percentage points | 2,017 | About 17 days, rarely worth it |
Running a study
A wrench time study goes wrong when it feels like surveillance. People who think they are being judged work differently while the observer is in sight, and the result is both wrong and resented. Telling the crew in advance, recording no names, and sharing the results with them first is what keeps it honest, and the crew usually know better than anyone where their time goes.
| Step | What it involves | In practice |
|---|---|---|
| Explain it | Tell the crew and their supervisors what is being measured and why: the system, not the people | No names recorded, results shared with the crew first |
| Set the categories | Agree the categories and what counts in each before the first observation | The eight categories above, with examples written down |
| Size the sample | Choose the precision needed and work out the observations from it | About 900 observations for ± 3 points |
| Randomise the rounds | Random times and routes, so observations are not tied to the start of the shift or a favourite area | Ten rounds a day at random times, across all shifts worked |
| Observe | Record the category of each person at the instant they are seen, not what they did a minute ago | One observer, or several trained to the same categories |
| Analyse and act | Share of observations by category, by trade and by shift, and the reasons behind the biggest losses | Repeat in six to twelve months to see whether the changes worked |
What eats wrench time, and what removes it
The losses fall into a few groups, and each has a fix in the work system. Most of the fixes are the steps of maintenance planning and scheduling: a complete job pack, parts reserved and kitted, permits prepared the shift before, support booked in the weekly schedule. The permit to work system is the one to look at first on many sites, because a crew queueing at a single issuing officer each morning loses the first hour of the shift.
| Loss | Usual cause | What removes it |
|---|---|---|
| Looking for parts | No reservation, parts not kitted, store far from the work | Parts reserved at planning and kitted to the job before the day |
| Waiting for a permit | Permits raised on the morning of the job, one issuing officer for the whole crew | Permits and isolations prepared the shift before for scheduled work |
| Not knowing the job | Scope, steps, drawings or settings missing | A complete job pack with the steps, drawings, limits and history |
| Waiting for others | Cranes, scaffold, operators or another trade not booked | Support booked in the weekly schedule, with the crew sequence agreed |
| Travel | Work spread across the site, tools and consumables in the workshop | Work grouped by area, tools and consumables taken to the job |
| Late start, early finish | No job ready to start after the pre-start, nothing assigned for the last hour | The day's jobs assigned the day before, with a short job ready for the end of the shift |
Using the result
Wrench time is a diagnostic, not a target. Setting a wrench time target invites people to look busy, and the next study then measures the performance rather than the work. The useful output is the split by category, trade and shift, which shows where the waiting is. A crew losing an hour a day at the store counter needs kitting, and a crew losing it at the permit office needs permits prepared earlier.
Repeat the study six to twelve months after the changes, with the same categories and a similar number of observations. Between studies, the weekly measures show whether the changes are holding: the planned share of work, schedule compliance and break-in, and the ready backlog. A rising wrench time with falling schedule compliance usually means the crew is busy on the wrong work.
Common mistakes
- Following individuals with a stopwatch instead of sampling the crew.
- Observing at fixed times, such as on the hour, which the crew soon learns.
- Changing the categories between studies, so the results cannot be compared.
- Recording names, which turns a study of the system into a review of people.
- Setting a wrench time target, which rewards looking busy.
- Measuring once and acting on nothing, so the next study finds the same losses.
- Comparing with figures from other sites that defined the categories differently.
Common questions
What is wrench time?
Wrench time, also called tool time, is the share of a maintenance tradesperson's paid time spent actively working on the job: dismantling, repairing, adjusting, testing and inspecting. Time spent travelling, waiting for parts, permits or other people, in late starts and early finishes is not wrench time. It is a measure of the work system around the tradesperson rather than of the person.
What is a good wrench time percentage?
Studies of sites with largely unplanned work commonly find 25 to 35 per cent. Sites with working planning, scheduling and kitting are reported at 40 to 50 per cent, and figures above that are rare and worth checking, since breaks, meetings and some travel are unavoidable. The useful comparison is with the site's own last study, by trade and by shift.
How do you measure wrench time?
With a work sampling study. An observer makes rounds at random times and records what each tradesperson is doing at the instant they are seen, in one of a set of agreed categories such as direct work, travel, waiting for parts or waiting for a permit. Over several hundred observations the share in each category estimates the share of time. About 900 observations give a result within plus or minus three percentage points at 95 per cent confidence.
How do you improve wrench time?
By removing the causes of waiting and travel, which sit in the work system. Plan jobs with complete job packs, reserve and kit parts before the day, prepare permits and isolations the shift before, book cranes, scaffold and operators in the weekly schedule, group work by area, and assign the day's jobs before the shift starts. Each change targets a loss category the study showed to be large.
Is wrench time the same as productivity?
No. Wrench time measures how much of the shift reaches the job, not how well or how fast the job is done, or whether it was the right job. A high wrench time on unnecessary work is no gain. It is best read alongside schedule compliance, the planned work share and the reliability of the plant.
Key terms
Plain-language definitions from our glossary for the concepts this article leans on.
Standards and further reading
- Maintenance Planning and Scheduling Handbook, 4th edition (Doc Palmer, McGraw-Hill)
- Best practices, metrics and guidelines for maintenance and reliability (SMRP)
- Operations and maintenance best practices guide (US Department of Energy, FEMP)
Related case studies and tools
Related reading
Working through something like this?
See how we approach these initiatives, or tell us what you are dealing with.