Study of selected machinery accidents for optimization of prevention

Status:

completed 06/2010

Aims:

The occupational accident statistics of the German Social Accident Insurance (DGUV, formerly the HVBG) are not sufficient to enable the causes of occupational accidents on machines and the conditions under which they occur to be reconstructed adequately. For example, it cannot be determined whether the machine on which an accident occurred had been tested and certified by one of the testing bodies of BG-PRÜFZERT. The proportion of tested machines among machines on which accidents occur could however yield important information on the quality of testing and certification, and consequently serve as an indicator in the context of the project "Quality in Prevention", conducted by the Institute for Work and Health (IAG, formerly the BGAG) of the DGUV.
The objective of this research project was to conduct a detailed analysis based upon the accident reports, with reference to selected accidents occurring on machines in the member companies of the Social Accident Insurance Institution for the woodworking industry. Besides examining the proportion of the tested machines, the project was also to develop a concept for the data from accidents on machinery which are absolutely essential for optimized prevention activity.

Activities/Methods:

If the occupational accidents on woodworking machines in the member companies of the Social Accident Insurance Institution for the woodworking industry are compared with accidents in member companies of all Social Accident Insurance Institutions, the following accident blackspots are apparent: moulding machines (3/4 of all new disability pensions are attributable to the institution for the woodworking industry); circular saws (2/5 of all new disability pensions are attributable to the institution for the woodworking industry); planing machines and combined or multistage machines. A more detailed analysis of the machine types generating high accident rates reveals that accidents on single spindle moulding machines, circular saw benches/dimension saws, pendulum/crosscut/radial arm circular sawing machines, surface planing machines and three- and four-sided planing machines were substantially responsible for new disability pensions in the years 2001 to 2003.
Single spindle moulding machines would appear particularly suitable for a study of the effect of testing and certification upon the safe design of machines, since modified safety concepts were introduced into European standards in recent years. A standardized questionnaire was therefore employed in this research project in order to record and evaluate the accidents on moulding machines in Germany over a little less than a year. Control studies were conducted at the same time to ascertain the proportion of spindle moulding machines permitted to bear the GS mark, and the working processes on moulding machines. Evaluation was to show whether this information could be used in the formulation of detailed requirements for prevention.

Results:

The present case-control study confirms that the accidents on moulding machines were reduced decisively by consistent use of the available safeguards. Inattention and poor instruction (the latter particularly in association with difficult or unusual working processes) are the main causes of accidents.
According to the case-control study, the newer machines (manufactured in 1998 or later) are substantially under-represented in the accidents. This is however not evident from the evaluation of 40 accidents from 2007.
A lower frequency of accidents on GS-tested or CE-certified machines could not be confirmed. However, since the study did not examine the frequency with which the machines were used, no conclusion can be drawn from the results regarding the relative risk.
The additional information recorded by the accident questionnaires appears suitable in principle for the evaluation of accidents for prevention purposes. Evaluation of the necessary free-text fields is however very time-consuming. The questionnaire should be modified in order to address this point.

Last Update: