Steel Mill Wrench Time KPI Software: Craft Utilisation Guide

Wrench time and craft utilisation answer different questions

Teams often use the two terms as if they were the same. They are related but they expose different problems, so measure both.

A simple example: 230 hands-on minutes in a 480 minute shift gives 47.9 percent wrench time.

Benchmarks you can compare against

Published figures vary by source and by how time is counted, so treat them as ranges to calibrate against, not as targets to copy.

Scale runs from 0 to 100 percent. Ranges are widely cited industry figures, not Oxmaint customer results.

Why the gap is worth chasing

Moving from 30 to 55 percent nearly doubles the productive labor available from the same headcount. That capacity can clear backlog, complete preventive tasks, and reduce emergency calls.

Where steel mill hours disappear

Steel plants add losses that a typical factory does not have. Large footprints, high heat, and heavy lifting all take time away from the job.

Travel across the plant

Walking or driving between a workshop, store, and the work site on a large site.

Permits and isolation

Waiting for hot work, confined space, or lockout approvals before work can begin.

Crane and access waits

Waiting for overhead crane time, scaffolding, or mobile equipment.

Cool-down and release

Waiting for equipment to cool or for operations to release it to maintenance.

Parts and tools

Searching for spares, tools, or special fixtures after the job starts.

Job clarity

Unclear scope, missing drawings, or no history of the last repair.

Paperwork

Re-entering notes, closing orders late, or filling separate permit forms.

Interruptions

Pulled off a planned task for a breakdown elsewhere in the mill.

How to measure wrench time without a stopwatch war

The goal is a fair baseline, not a ranking of individuals. Combine two methods so each checks the other.

1. Work samplingAn observer records what each technician is doing at fixed intervals, such as every 15 minutes, and classifies it as wrench time or not. A few shifts per crew give a first baseline.
2. Work order labor loggingTechnicians book hours and delay reasons against work orders. Over weeks this shows which delay codes dominate.
3. Compare and reconcileWhere sampling and booked hours disagree, find out why. Missing bookings usually explain the gap.

Delay codes make the data useful

A short list of consistent delay reasons turns vague complaints into numbers. Use codes such as awaiting permit, awaiting parts, awaiting crane, awaiting operations, and unclear scope.

• Keep the list short enough that technicians actually use it
• Use the same codes on every shift and every area
• Review the top three codes weekly with supervisors
• Explain clearly that data is used to fix the system, not to blame crews

Seven levers that raise wrench time

One job, two ways: a gearbox bearing replacement

This illustrative sequence shows how the same job loses or keeps time depending on planning. Durations are not benchmarks.

What the capacity gain looks like in hours

Percentages feel abstract until they are converted to labor hours. This example uses round numbers and is for illustration only.

That is 24 extra hands-on hours per shift, or 50 percent more productive labor, without a single new hire. The same logic applies to overtime and contractor spend.

Use the hours, do not just count them

Recovered time only helps if the schedule is ready to absorb it. Keep a backlog of approved, kitted jobs so crews can start valuable preventive work the moment a window opens.

Wrench time drags by steel plant area

Each production area loses time in its own way. Tailor the first improvement to the dominant drag in that area.

Zone and route the work

Grouping jobs by location cuts travel, which is one of the cheapest wins in a large plant. A scheduled route for inspections and small repairs also removes the need to send crews back to the same area twice in a shift.

Shutdown work follows different rules

During a planned outage, labor is often high but congestion, permits, and sequencing dominate the day. Measure shutdown wrench time separately so it does not distort the daily picture.

• Load jobs by area so crews do not queue for the same access point
• Prepare permits and isolations before the outage starts
• Stage parts and tools at the job site, not in a central store
• Track delay reasons daily and clear the top blocker at the next briefing

Review the outage after it ends

Compare planned hours with booked hours for each shutdown job, and record why the biggest gaps happened. These lessons make the next turnaround plan more realistic and give planners better estimates for similar work.

A weekly routine that keeps the number moving

Wrench time improves through small repeated fixes. A short weekly review is enough if it follows the same pattern.

1. Review the ParetoLook at last week's top delay codes and the hours attached to each.
2. Pick one causeChoose the largest controllable delay and name an owner.
3. Test and checkTry a fix, such as earlier permits or kitting, then see if the code count falls.

Communicate the purpose

Tell crews plainly that the goal is to remove obstacles. When technicians see permits arrive sooner and parts waiting at the job, they report delays more honestly and the data improves.

Companion KPIs to read with wrench time

Wrench time alone can mislead. Read it next to these measures.

How Oxmaint supports craft utilisation reporting

Oxmaint organizes the work, labor, and parts data that wrench time analysis depends on.

Start with the data you already have

Even a few weeks of labor hours and delay reasons will show whether the main problem is permits, parts, travel, or planning. Fix the largest one first.

Data quality: making craft utilisation numbers trustworthy

If hours are booked loosely, craft utilisation will look worse than reality and nobody will trust the report. A few simple rules fix most problems.

• Book time to a work order every shift, including small jobs and inspections
• Use a standard category for training, meetings, and permit time so they are not hidden
• Close work orders within the same shift while the details are fresh
• Require a delay reason whenever a job takes much longer than planned
• Audit a sample of closed orders monthly and feed corrections back to crews

Separate reactive and planned labor

Tag each work order as planned, reactive, or emergency. Without that split, a rising craft utilisation figure can simply mean the crew is busy fixing breakdowns, which is not the outcome you want.

Trends changing how mills manage maintenance labor

Steel plants are tightening their maintenance labor management for practical reasons. Experienced craftspeople are hard to hire, and every hour of skilled time is worth protecting.

• Mobile work orders reduce trips back to the office for instructions and sign-off
• Condition-based triggers shift effort from emergency repair to planned intervention
• Dashboards put backlog, schedule compliance, and delay reasons in front of supervisors daily
• Standard job plans capture the knowledge of senior technicians for newer staff

Capture knowledge while you plan

Each well-written job plan is a reusable asset. When a senior technician retires, the steps, tools, and cautions stay in the system and the next crew starts with a better first hour.

Pitfalls to avoid

• Treating wrench time as a score for individuals rather than a system measure
• Comparing plants that count time differently
• Chasing a high number by skipping safe permit and isolation steps
• Ignoring partial bookings that make craft utilisation look worse than it is
• Setting a target before measuring a baseline

Safety steps are not waste. Permits, isolation, and cool-down must stay in the process even when they take time.

Frequently asked questions