Accidents database FACTS

To verify the developed concepts underlying the structured 7-stage protocol in a reactive way, they were applied to an analysis of recent accidents in the Dutch chemical process industry. Despite the limitations in the information available from the accident database, it could be deduced that all accidents were preceded by precursors, and even that similar precursors had led to similar accidents, implying that companies had failed to learn from these re-occurring deviations which were in fact pre-warning signs of impeding accidents. 

The amount of detail on accidents recorded on the data base varies and the different levels of detail present, are indicated by a star ( ) rating code. The higher the star ( ) rating, the higher the richness of information present inside the database, as shown in Table 4 from the instruction manual of the FACTS database (FACTS, 2002). 

Like the accidents from Chapter 2 and in the previous Section, the accidents used in this Section are retrieved from the FACTS accident database. As in the other studies, in this case the quality of information is also an important figure of merit for analysing underlying causes of previous situations. Therefore, in this case the... 

The Eahure and Accidents Technical information System (FACTS) is an accident database, which contains information on more than 25200 (industrial) accidents (incidents) involving hazardous materials or dangerous goods that have happened all over the world during the past 90 years. 

A direct comparison with [GOL 04] results should take into account differences in units and the fact that his accident database also includes property... 

In order to define an accident database, we have to specify the types of facts or data elements to be included in each accident or near-accident record. We also have to specify a data type for each element. This tells us whether the data is coded according to a nominal, ordinal, interval or ratio scale or stored in free text. Both these specifications are included in the so-called database definition. Each data element and its associated data-type specification are... 

The Facts database consists of the coded data and for most accidents an additional extended abstract, which briefly describes what actually happened. The coded data shows in more detail ... 

The FACTS database does not show detailed accident analysis information, nor does it show in great detail which factors caused the accident. It mainly describes what happened and what were the consequences together with some general remarks about the causes. 

The chain of events in the operational process, leading to the accident is briefly described in the Facts database. However, whether these events were present before the accident occurred is not indicated. So, to retrieve possible deviations and their... 

In order to demonstrate the use of the analysis flow scheme of Figure 35, one of the accidents from the FACTS database is described and analysed according to this analysis flow scheme. The accident description in the FACTS database is very detailed and only a summary is given here. 

The information used in this research is based on the information that can be retrieved from the FACTS database. It is possible and even likely that more information about precursors was available before the real accident happened, implying that the amount... 

For all 17 accidents of this study, precursors could be identified. None of the 17 accidents could be classified as unforeseen . A total number of 39 precursors were identified in these 17 accidents which seems extremely low compared to normal accident analysis experience. Detailed accident analysis normally retrieves dozens of near misses and deviations leading to the final accident (e.g. van der Schaaf (Schaaf van der, 1992)), implying the existence of many more precursors. However, the limited amount of detailed information present in the FACTS database is the restricting factor here. The FACTS database reflects the kind of accident information companies and government agencies collect. It demonstrates clearly that detailed information about the period before the accident is not often collected. 

During the examination of the information from the FACTS database, it became clear that most accidents could be linked to changes or alterations in the company. These include not only technical and organizational changes, but also alterations in human staffing. 

The first purpose is unique to near miss reporting, while the other two could in principle be also achieved by accident reporting. However, for the second purpose the advantage of near miss reporting lies in the fact that a large database may be built within a relatively short timeframe. The absence of any real adverse effects in the case of near misses may be considered as an... 

Two kinds of causes are taken into account in the database the general cause and the specific cause. Of the 413 accidents studied in this survey, the cause was known only in 252 cases (61% of total). This can be explained by the fact that it can be very difficult to establish the cause after the accident. The various general causes are summarized in Table 41.11. Because one accident can be due to more than one cause, the sum of the percentages is again greater than 100. 

A database is a collection of related facts or data that has been designed, built and populated with data for a specific purpose (Elmasri and Navathe, 1994). We are here concerned with computerised databases on accidents and near accidents. These are electronic archives, containing a record of the essential facts about each unwanted event at the company. The purpose of storing the data electronically rather than in a paper-based archive is to make it easier to retrieve, analyse and present the data. Computer support also offers new possibilities for data analysis and presentation. 

The users of the database have to handle the following two activities (1) updating the database with new facts on accidents and near accidents and (2) querying the database by retrieving relevant data and generating reports that present the results of the queries. 

---