Establishment of Hours of Work Risk Factors Relevant to Coal Mining

The objective of this project was to determine the principal variations in accident and incident risk in relation to roster design within the Australian Coal Mining Industry.   These were determined by analysing the accident and incident statistics for a wide range of coalmine operating conditions. This analysis enabled the likelihood of an accident or incident occurring for different roster patterns to be characterised. This will enhance the application of the risk management framework developed for shiftwork by including these coal mining specific relative risk factors.

From the general literature it would be expected that:

• For the same type of work, night shift should have a higher incidence rate than day shift.
• As the number of shifts increases, the incidence rate should increase.
• As the hours of work per shift increases beyond 8 hours, the incidence rate should increase.

However, there is some evidence from previous mining studies that this may not be true in the Australian coal mining industry.  Factors that may complicate this relationship are:

• Whether the rate of work is self paced or limited by machinery through-put.
• Whether the hours of work are rostered hours or overtime.
• The organisation and impact of rest breaks within shifts.
• The organisation and impact of breaks between shifts.
• The effect of fatigue management policies on accidents and incidents.

In general the data analysed in this project did not indicate any strong correlation between hours of work and the number of incidents or injuries.  The patterns found in other industries and used as the basis for regulations and guidelines have not been found. This lack of relationship between hours of work and incidents and accidents may reflect effective fatigue management policies at mine sites.

From the data analysis, the highest number of incidents occurred in the first four hours of a shift.   Day shift had more incidents than night shift on average.  For the longer shifts the number of incidents settled to a relatively constant value after the initial peak, in hours 2 – 5 of the shift, decayed.

There was no detectable change in the relative number of severe injuries between shift types.

There was no discernable effect of roster design on the incident occurrence patterns for extended shifts (more than 8 hours).  Shift rosters designed to operate over a seven day work week had the highest number of incidents on the first shift back after a break, normally a day shift. Analysis of number of incidents per shift for successive shifts found no evidence of any change in pattern as the number of shifts increased.

On one roster cycle type there was some evidence of an increase in incidence with increasing number of night shifts following a period of day shifts, however the number of incidents on the nights shifts was less than the number of incidents on the preceding day shifts.

The pattern of number of incidents per shift versus hour of shift for open cut mines is slightly different to that obtained for underground mines, the decay after the 2- 4 hour peak is less pronounced with more incidents occurring later in the shifts in open cut mines than in underground mines.

It was not possible to compare different rosters to see if one roster pattern had a lower incidence than another because there appears to be no uniform definition or application of definitions in reporting incidents between mines.    It was not possible to assess the incident rates during overtime hours of work as the hours worked as overtime were not tracked in sufficient detail.

If mines want to monitor the variation in incidents with time on shift or number of successive shifts then they must ensure that they maintain high data collection quality and collect all the relevant information including day of roster cycle and roster cycle details.  This is particularly true for those incidents relating to contractors. It was not possible to investigate the hours of work behaviour of contractors other than those working on mines for long periods of time.  The issue of hours of work and incident rates for short term contractors merits further exploration.