How to Design a 24/7 Shift Schedule Without Hiding Relief Capacity

A 24/7 roster must cover 168 hours each week for every continuously staffed position. The rotation pattern changes how those hours are divided, but it does not remove the underlying capacity requirement.

8-hour versus 12-hour shifts

Three 8-hour shifts create more handovers each day. Two 12-hour shifts create fewer handovers but longer duties. Neither is universally better; choose the pattern that fits the operation and then test it against the rules and workforce that apply locally.

Common rotation patterns are starting points

2-2-3, 4-on-4-off and other named rotations are useful templates, but whether the roster can actually be covered still depends on headcount, skills, leave, availability, rest and weekly-hour capacity.

Check fairness separately from whether every shift can be covered

A roster can be technically feasible and still distribute nights or weekends poorly. Treat hard coverage and eligibility as one layer, then use soft objectives for preferences and fairness.

Compare the same demand before comparing patterns

Assume one continuously staffed role, 40 contracted hours per worker each week and no overlapping handovers in the initial estimate. Both an eight-hour pattern and a twelve-hour pattern need 168 staffed hours. Three eight-hour shifts produce 21 assignments per week; two twelve-hour shifts produce 14. Assignment count changes, but the basic staffing hours do not.

If each shift boundary requires a 15-minute overlap by the incoming and outgoing workers, an eight-hour pattern adds 21 × 0.25 = 5.25 staffed hours per week; the twelve-hour pattern adds 14 × 0.25 = 3.5. Model overlap explicitly instead of assuming that fewer handovers remove the underlying 24/7 demand.

A rotation label does not define capacity

A twelve-hour 2-2-3 pattern usually alternates five-shift and two-shift weeks for a worker, producing 60 and 24 hours and an average of 42 over 14 days. A twelve-hour 4-on-4-off pattern averages 48 ÷ 8 × 7 = 42 hours per week across its cycle. A 40-hour weekly hard limit would conflict with some weeks of these patterns. Check whether your rule is weekly, averaged over a defined period, or a payroll threshold before selecting a template.

Four crews covering a single day/night duty pattern may meet the basic rotation, but they do not automatically provide absence relief. With a 15% example absence assumption, 4.2 ÷ 0.85 = 4.94 FTE. That arithmetic and the staffing calculation belong alongside the pattern, not in a hidden reserve column.

Define a fairness objective you can actually measure

First require valid availability, qualifications, non-overlap and configured rest. Then measure nights, weekends and unpopular duties across the relevant rotation period. Suppose two equally eligible full-time workers receive six and two nights in a month. Their night-count gap is four; a valid redistribution to four each reduces that gap to zero. If one worker is available only for days, equal night counts across the whole workforce is an inappropriate objective.

Compare people within eligible groups and account for contracted hours. Record which preferences are soft goals and which restrictions are mandatory. A fair score that silently breaches a required qualification is not a usable roster. Keep exceptions visible and review fairness over the full cycle rather than judging only the first week.

Test changes and boundary conditions

1. Generate the full rotation period and include assignments immediately before and after it.
2. Remove one available worker to simulate leave or sickness.
3. Check whether relief covers the same role and location.
4. Recalculate hours, rest, nights and weekends after any swaps.
5. Review remaining gaps before publishing.

The UK HSE shift-design guidance is a reference for evaluating fatigue-related design choices, not a universal legal schedule template. Apply the local policy and work demands. Compare the 2-2-3 generator, 4-on-4-off guide and twelve-hour rotation guide using the same workforce inputs.