An underground mining operation assumed the answer to fatigue incidents was moving off 12-hour shifts. The diagnostic told a different story — the shift length wasn’t the issue; the rotation pattern was.
Mining · UndergroundAn underground mining operation working a 12-hour rotating shift pattern had seen fatigue-related incidents climb over an 18-month period and was preparing to move the workforce to an 8-hour pattern at significant cost and operational disruption. The diagnostic identified the actual drivers of the fatigue trend — backward rotation, inadequate recovery windows between blocks, and an unpredictable weekend distribution — none of which required a shift-length change. The redesign kept the 12-hour envelope and changed the rotation mechanics. Fatigue-related incidents fell substantially within the first year.
A North American underground mining operation producing base metals from a deep-shaft mine. Approximately 620 underground workers organized into four crews on a rotating 12-hour shift pattern, plus surface crews on separate schedules. The 12-hour pattern had been in place for over a decade and was an established part of the workforce contract under a long-tenured union agreement. Production ran continuously across all four crews with planned weekly maintenance windows.
Over the prior 18 months, fatigue-related incident reports had climbed roughly 60% against the operation’s historical baseline. The reports included a mix of near-miss events, equipment damage events, and a small number of recordable injuries with fatigue identified as a contributing factor. The internal safety review had pointed at 12-hour shift length as a probable cause, and a proposal was on the table to move the underground workforce to a three-shift, eight-hour pattern — a change that would require contract reopening, a substantial increase in headcount, and operational re-engineering of crew handoffs underground.
The proposed move to 8-hour shifts carried real costs beyond the contract negotiation. Underground mining handoffs are operationally expensive — each one consumes 20 to 40 minutes of productive time at the working face as crews travel to and from the location. Three handoffs per day rather than two would meaningfully reduce face time, and the headcount increase required to staff three crews at the existing crew size was substantial. The 8-hour solution was viable but expensive, and leadership wanted certainty that it would address the actual cause of the fatigue trend before committing to it.
Phase 1 · Business Assessment
We started with the incident data itself — not the summary view, but the underlying records. For each fatigue-related incident in the prior 18 months, we mapped the rotation position of the affected worker: which day of the block the incident occurred on, which direction the worker was rotating, how recently the worker had transitioned between day and night, and what the recovery time had been between the prior block and the current one. We also examined the rotation mechanics across the four crews: rotation direction, block length, turnaround windows, and weekend distribution.
The incident clustering was not uniform across shift hours, which is what would have been expected if 12-hour length itself were the driver. Instead, incidents clustered heavily in three specific situations: the first night shift after a day-block-to-night-block transition, the final shift of any block exceeding four consecutive days, and any block following a short turnaround — defined as less than 48 hours of recovery between the end of one block and the start of the next. The existing rotation was backward (day → night → evening), which fights natural circadian drift and produces sleep disruption that takes multiple shifts to recover from. Block lengths ranged from four to six consecutive shifts, with the longest blocks correlating most strongly with incidents. Turnaround windows were inconsistent, with some sequences delivering less than 36 hours between blocks. None of these were shift-length problems — they were rotation-pattern problems operating inside a 12-hour envelope.
The shift-length question is almost always the first question asked when fatigue incidents climb — and almost never the question that produces the answer. Rotation direction, block length, and recovery windows usually carry more weight than the length of any single shift.
— Ethan Franklin, Senior Partner, Shiftwork Solutions LLC
Phase 2 · Workforce Assessment
We surveyed the underground workforce on sleep, recovery, and shift experience — with specific attention to the situations the business assessment had flagged. The survey results aligned closely with the incident clustering. Workers consistently reported the most difficulty sleeping during the first night shift after a day-to-night transition. The five-day and six-day blocks were experienced as substantially more fatiguing than four-day blocks. Short turnarounds were widely identified as the worst part of the existing pattern. We also asked about preference for forward versus backward rotation — a question many workers had not encountered before — and after explanation, preference shifted strongly toward forward rotation among those who tried it during a small structured trial we ran with two volunteer crews during the assessment phase.
Phase 3 · Solution Design
The redesigned schedule kept the 12-hour shift length and the four-crew structure, but changed three elements. Rotation direction moved from backward to forward (day → evening → night), aligning with natural circadian drift. Block lengths were standardized at four consecutive shifts maximum, eliminating the five-day and six-day blocks that had concentrated the late-block incidents. Turnaround windows were standardized at a minimum of 48 hours between any two blocks. Weekend distribution was rebuilt to fall in predictable positions, allowing workers to plan recovery activities around a known schedule rather than chasing weekend rest opportunistically.
Phase 4 · Implementation Preparation and Rollout
The implementation manual was prepared in close coordination with the union leadership, recognizing that the rotation change was a material modification to working conditions under the contract. The transition was timed carefully to avoid putting crews through a rotation change during the operation’s annual production push, and the first 90 days of the new pattern included structured fatigue monitoring through a third-party reporting channel that would alert the operation to any drift back toward the prior incident pattern. Rollout took ten weeks from manual approval through full operation on the new rotation.
Measured against the client’s stated objective:
| Metric | Before | After (12 months post) |
|---|---|---|
| Shift length | 12 hours | 12 hours (unchanged) |
| Rotation direction | Backward | Forward |
| Maximum block length | 6 consecutive shifts | 4 consecutive shifts |
| Minimum turnaround between blocks | ~36 hours | 48 hours |
| Fatigue-related incident frequency | Baseline + 60% | ~35% below baseline |
| 8-hour migration cost (projected) | ~$8.4M ongoing | $0 (plan avoided) |
The third-party fatigue monitoring channel reported a sharp drop in self-reported fatigue scores within the first three months and sustained improvement through the twelve-month review. The union leadership, which had initially viewed the proposal cautiously, endorsed the new pattern publicly at the next contract review and the change was incorporated into the contract without dispute. The avoided 8-hour migration also preserved face-time at the working area — production rates held steady through the transition rather than dropping as they would have under a three-handoff model.
The Design Principle: Shift length is one variable in a rotation pattern, not the variable. When a fatigue trend emerges, the diagnostic question is which mechanics are driving the trend — rotation direction, block length, recovery windows, weekend distribution — and addressing those usually produces the durable improvement. Defaulting to the shift-length lever first often costs more than the underlying problem ever did.
The pattern in this engagement repeats wherever a 12-hour rotation has been in place long enough that the original design rationale has been lost. The fatigue trend appears, the explanation reaches for the most visible variable (shift length), and the proposed remedy is a shift-length change. But underneath the visible variable, several rotation mechanics are usually operating below the threshold of routine attention — and those mechanics carry more explanatory weight than the shift length they sit inside.
A second pattern, particularly relevant in underground mining: handoff cost is asymmetric. Moving from two handoffs per day (12-hour pattern) to three (8-hour pattern) is not a 50% increase in handoff cost — the productivity loss compounds with the travel time underground, the safety protocols at the working face transition, and the supervisor time required to manage each transition. The implicit cost of an 8-hour migration is almost always larger than the operation budgets for it, which makes a 12-hour redesign — when the diagnostic supports it — the more favorable path.
If your team is looking at fatigue incidents and a shift-length change is being considered, the most useful first step is a granular look at when the incidents are occurring in the rotation pattern — not just which shift, but which day of the block, which direction the worker is rotating, and what the recovery window before the block looked like. That analysis usually changes the conversation, and often changes the answer.
Shiftwork Solutions LLC has guided hundreds of engagements across mining, refining, chemical processing, and other 24/7 industrial operations over more than three decades. Visit shift-work.com to start a conversation.
