Informa Markets

Author Bio ▼

Safety and Health Practitioner (SHP) is first for independent health and safety news.
August 31, 2011

Get the SHP newsletter

Daily health and safety news, job alerts and resources

Shift work – Exhausted options

The organisation of shift work and management of its potential effects on staff are becoming issues for more and more employers as the 24/7 economy grows, so John Wilkinson and Dick Rudd describe what needs to be done, using two case studies to illustrate how shift work should and should not be managed.

Employers of those who regularly engage in shift work have, of course, a duty of care towards those employees under general health and safety law, and certain regulations, such as those governing working time, but there is no specific definition of shift work in law. However, it usually means a work activity scheduled outside standard daytime hours, where there may be a handover of duty from one individual or work group to another; or a pattern of work where one employee replaces another on the same job within a 24-hour period. Standard daytime hours are considered as a work schedule involving an activity during the day, commonly for a period of eight hours between 7am and 7pm.1

While particular industries and organisations, like transportation, utilities, the emergency services and hospitals, have long implemented shift patterns in order to provide 24-hour operation, the growth in the service economy over the past decade or so has meant the likes of retail outlets, catering establishments, and call centres are increasingly employing them, too.

As a result, shift work – its design, implementation and potential effect on employees – needs to be carefully managed. The case studies outlined below illustrate both the consequences of failing to do so and the benefits of a proactive approach, but before getting into the specifics of those two situations, the following general points must be emphasised:

  • Managers should spend a little less time looking for a new, or ‘perfect’ shift system and more time managing and monitoring what they’ve already got. That doesn’t mean you shouldn’t review – rather, you need to spend more time making sure that what you’ve got is actually working for you, not against you, and establishing a baseline; 
  • Make sure you have enough staff, and in the right places, for the work to be done and to cover foreseeable peaks such as upsets, emergencies, absence, illness, training and project cover; 
  • The actual shift pattern you adopt, though important, is only one input. All shift patterns are a compromise – human beings aren’t wired for shift work – and it is important to take account of other factors, including social and local ones; 
  • Predictability of the pattern they work is a major issue for shift staff.2 This allows individuals to plan ahead, and make longer-term social commitments for themselves, their families and others. Managing the shift system to run as planned is therefore very important (but is something that current guidance overlooks);
  • Predictability is equally important for managers and supervisors. If the system runs as it should it means a substantial workload and nagging operational problems are taken off their shoulders, as they are not running around trying to fill gaps every week;
  • Choose different numbers of shift teams so that you can build in spare capacity for training, process optimisation, absence/illness cover, and so on – in other words, an ‘extra shift’. The optimum for most businesses is four shifts but major-hazard sites may warrant five to provide this extra capacity;
  • If you are going to change your system, don’t do it too quickly. Reviews can – and should – take some time and actively involve key staff. You may also need to provide support and training to enable good decision-making; and finally
  • Beware specialist fatigue/shift-work businesses selling you their services and products – get a grip on what you’ve got first, then engage your staff, then decide what you want to do.

So, what are the costs of, on the one hand, ignoring the above and, on the other, taking it into account and reviewing your shift-work patterns? The following tale of two refineries will illustrate.

The bad

In his excellent article in last month’s SHP on what went wrong at Buncefield to cause the huge explosion at the oil depot in December 2005 Steven Brooker focused mainly on systems and communication failures.3 But another major strand of the complex investigation looked at the human-factors elements – including shift work and handover, fatigue, human/computer interfaces, competence and training.

It was some time before the investigating team was able to interview supervisors and other key witnesses so, initially, it had to rely on what documents and other evidence could be recovered from the debris-strewn control room. A wall chart showing the supervisors’ roster provided some early clues that fatigue – through poorly-managed shift work and overtime – might be an issue. The interviewing team was thus able to shape questions for the witnesses, a process that took several months of painstaking work. 

In general in such incidents, fatigue would rarely be the ‘smoking’ gun’ but it is nevertheless often a key issue: examples include Texas City, Three Mile Island, Chernobyl, the Herald of Free Enterprise, and Bhopal. At Buncefield, the shift system was unexceptional in some ways – a 24/7, 12-hour rotating pattern with an eight-week cycle and 7am-7pm start/finish times – but the wall charts showed that it was rarely worked as planned, and that there was a lot of overtime. 

Even if it had been worked as planned the shift pattern had some features that were less than good and could lead to accumulated fatigue, i.e. tiredness levels that would likely persist. However, there was a rest week in the cycle, which, if taken, would at least have allowed recovery at the end of the cycle.

The pattern was originally chosen to be worked by ten supervisors but was generally being worked by only eight, owing to recruitment and retention issues. Consequently, supervisors were working up to seven 12-hour shifts at a time, i.e. 84 hours, as well as in the rest week, so the chance of recovery at the end of the cycle was lost. So, over the normal eight-week cycle it was likely that cumulative fatigue would build up and persist. Regular added shifts and overtime exacerbated this, especially as the overtime was not evenly distributed. There were clear indications that those doing the most overtime had had enough, some time before the explosion.

As well as failures to manage the shift roster effectively, there was a failure to manage within-shift issues – for example, adequate rest and meal breaks, which are always essential but even more so in 12-hour shift systems. At Buncefield, supervisors generally took their meals in the control room, at their desks. There were no effective or formal arrangements in place to make sure breaks were taken, or even to specify them.

Regular short breaks help reduce fatigue, improve productivity and may reduce the risk of errors and accidents, especially when the work is demanding or monotonous.1 An active policy is best, taking breaks before fatigue sets in rather than to just allow recovery.

The fatigue effects on individuals would be more likely in control-room monitoring and other tasks carried out by the pipeline supervisors, which required vigilance, alertness and awareness. Fatigue would also be likely to affect their decision-making ability and recall. 

Several of the key events leading up to the blast occurred in the small hours of the morning and at the end of a 12-hour shift – known higher-risk periods for fatigue.

In summary, there was no fatigue management policy or procedure in place at Buncefield, nor any other effective arrangements to manage fatigue and shift work, leaving supervisors to manage the shift roster themselves, including cover for holidays and sickness. The designated stand-by man was often unavailable because of the reduced number of supervisors, and substantial amounts of overtime were also regularly worked to cover projects and training.

The good

Between 2006 and 2007 a large, complex UK refinery carried out a review of its shift arrangements. An employee consultation team led the review, with management support and guidance. At the time, around half the total staff worked on operations, on a 5×8-hour shift system. The representative shift-work group considered a range of options, including a 12-hour pattern. After evaluation – which included use of the HSE Fatigue and Risk Index Calculator (FRIC) – a shortlist of options was put to the shift staff for a vote, and the most popular one was implemented. To date, the new system has been generally regarded as a success.

A key – and novel – feature of the company’s approach is to monitor the new arrangements very closely and effectively, and to respond actively and promptly to the findings. This allows a real picture of how the shift system is working to be built up, and has led to greatly improved control of overtime, shift-swapping, workload and manning levels.

Specific tools are used to monitor and audit attendance and to accurately capture the actual shifts worked by individuals. These tools are themselves regularly audited, also. A weekly audit is made of weekly timesheets, looking for individuals whose hours have exceeded the agreed thresholds for daily hours, or successive shifts. A procedure – requiring a supervisor’s authority – is in place for holiday cover and shift-swaps to make sure there are good reasons for changes, that holidays are properly booked, and any impact on thresholds is considered. 

In addition, a monthly time-balance check is carried out to identify any staff whose holiday balance exceeds 10 days for the projected year-end. These are then brought to the attention of the relevant department or manager for action. This a very useful tool for checking that staff time is being effectively managed, year round, ensuring that there is no sudden or undue pressure to provide shift/holiday cover, or to swap shifts, etc. towards the end of the working year.
A further KPI tracks monthly staff time as a ‘health check’, including overtime and sickness. One check also shows the number of days in credit – over and above the annual holiday entitlement – and confirms that staff and supervisors are managing staff working hours. The balance is the number of days/shifts that the company owes individuals (i.e. based on the annualised hours quota).

However, the KPI that has had one of the biggest impacts in identifying and reducing fatigue at the site is the disruption data. A specific process area raised a concern about the number of shift changes that were being made during a three-week turnaround period, so a KPI was developed that would track the number of “shift changes made for business need”. 

Part of the problem was that supervisors on this area knew that the number of 12-hour shifts worked on their area was being tracked. So, to reduce these numbers staff were moved to cover gaps on other shifts (eight hours), rather than work through on 12 hours – a classic ‘unintended consequence’ of well-intentioned change! With this data it was possible to track a more accurate level of fatigue/disruption across the site, with the target being one shift change per person, per month.

This disruption-tracking encouraged first-line supervisors to manage disruption on their shifts better, reducing the number of last-minute shift changes. Further analysis of the cause of these disruptions led to a number of changes and learning points, including:

  • employing replacement operators sooner, to reduce or eliminate time between operations staff retiring and their replacement learning their first job;
  • recognising that disruptions increase around unplanned events – by continuing to improve plant reliability these events can be reduced;
  • implementing an action plan for dealing with cover for employees off work due to sickness at an earlier stage; and
  • disqualifying the perception that areas with a smaller workforce would have a greater proportion of disruption per person – the data produced identified that this was not the case.

New systems

From this point, a number of other systems was developed to improve shift issues. An ‘Age and Experience’ database was set up and used to help balance shifts based on experience levels. This allowed improved control of shift competency. To improve resource management, a training plan was developed to capture known and projected sickness and light duties, priority training, turnarounds, day assignments and retirements. This allowed much-improved control of training and much less disruption on shifts when staff are released for training.

To improve attrition management, historic attrition data was factored into retirement plans. This was then used as the basis to predict attrition rates for the next three years. Using this information, re-adjustments have been made to established numbers between areas, and the base establishment figure was improved by employing new operators earlier to offset increase in retirements.

A shift-worker and family education programme was set up, led by Occupational Health, to help improve staff understanding of the risks from shift work, and the key control and personal measures that can be taken to reduce them.


All of these steps have helped reduce fatigue on site – not just in terms of the statistics and data but in the form of a real and noticeable improvement for the operations staff. Using the data gathered has allowed the company to better plan for future staffing requirements – a crucial effort, given that at previous attrition levels, there was a real risk of significant competency, succession and manning issues developing over the following years, with a likely peak ten years on.

Given that it takes up to five years to become a fully qualified operator at this site, the cost of not being ready for this increase would have had a major impact on the business – for example, through increased overtime and disruptions required to provide cover across the workplace.

Now, the new systems and arrangements help the company plan for the future, and deal with likely problem areas before they become real business issues. It’s clear that the future economic picture will continue to be tough, so these systems and arrangements continue to have a vital role.    

1    HSE (2006): Managing shift work – health and safety guidance (HSG265) –
2    Miller, James C (2006): Fundamentals of shift-work scheduling, Air Force Research Laboratory, April 2006 – search?query=fundamentals+of+shiftwork
3    Brooker, S (2011): ‘Precision parts’, in SHP August 2011, Vol.29 No.8 – -parts

John Wilkinson led the human-factors side of the HSE’s investigation into Buncefield and Dick Rudd is site operations training leader for a large UK refinery.

The Safety Conversation Podcast: Listen now!

The Safety Conversation with SHP (previously the Safety and Health Podcast) aims to bring you the latest news, insights and legislation updates in the form of interviews, discussions and panel debates from leading figures within the profession.

Find us on Apple Podcasts, Spotify and Google Podcasts, subscribe and join the conversation today!

Related Topics

Notify of

Inline Feedbacks
View all comments