Using Experience To Build a Safer Platform

Abstract

Summary This paper describes how Rogaland Research Inst. and Norsk Hydro A/S usedinjury records to identify hazards in the drilling module and to develop safetymeasures from the beginning of the design process. Database records, process. Database records, analyzed to identify injury concentrations within each modulesection, were used in a series of design review meetings to develop specificremedial measures. We also give examples of the hazards identified in the studyand recommendations made to combat them. "In 1986, the number of injuries suffered by offshore drilling personnelon personnel on Norwegian fixed platforms averaged 90 platforms averaged 90 per1,000 man-years. In per 1,000 man-years. In the same period, the average numberof injuries for all other activities on the same offshore platforms was 55 per1000 man-years." Introduction Drilling crews come in contact with a variety of hazards, including workingclose to heavy equipment in motion and being exposed to oil-based mud, chemicaldust, and poisonous and combustible gases. The pace poisonous and combustiblegases. The pace during much of the drilling operation is hectic, and the workinvolves a lot of heavy manual labor, with the attendant risk of straininjuries. The magnitude of the injury problem is illustrated by accidentstatistics. problem is illustrated by accident statistics. In 1986, the numberof injuries suffered by offshore drilling personnel on Norwegian fixedplatforms averaged 90 per 1,000 manyears. In the same period, the averagenumber of injuries for all other activities on the same offshore platforms was55 per 1,000 man-years. Since 1979, the overall trend has been toward a lower accident frequency. This decline is attributable in part to technical improvements and to greaterdrilling experience at all levels. The injury frequency in drilling from fixedplatforms has fallen from a 1979 level of 130 to 90 injuries per 1,000man-years. Because of this relatively high exposure to the risk of injury, focusing onthe danger early in the design process is important. In developing measures toimprove safety, the Norwegian oil industry now has a valuable tool at itsdisposal: information on all losttime accidents reported to the NorwegianRikstrygdeverket (RTV or Natl. Office of Social Security). This information hasbeen systematized into a database at Rogaland Research Inst. (RRI) for use inanalyzing injury patterns and identifying areas of high risk in the drillingindustry. RRI was engaged by Norsk Hydro A/S to analyze injury statistics fromdrilling operations on the Norwegian Continental Shelf and to use this analysisto identify injury concentrations (black spots) on the drilling modules ofmodern production platforms. These results, together with specific informationon the development project concerned, form the basis for recommendations forimproving safety in the design of drilling systems on new fixedinstallations. This paper discusses the method used to arrive at these recommendations andgives examples of the results obtained. We hope that this discussionillustrates how such a design review procedure could be incorporated intofuture development projects to emphasize safety at an earlier stage thanusually is the case. Using Accident Statistics In the Design Process Fig. 1 shows the procedure for identifying injury concentrations fromreported data and developing a list of measures to combat hazards. RRI Data Base. The RTV accident report form contains detailed information onthe person injured, the rig or platform the person injured, the rig or platformthe accident occurred on, the operations being performed at the time of theaccident, the performed at the time of the accident, the injuries incurred, andcomments from the injured person's safety representative. In the RRI data base, this information is encoded in a manner to preserve as completely as possiblean account of the circumstances possible an account of the circumstancessurrounding the accident. A drilling expert transfers the RTV-form informationto the data base by using the classifications shown in Fig. 2. This method isimportant to allow unrecorded operational details to be deduced whereverpossible and to avoid mistakes resulting from a lack of familiarity with thespecial terminology associated with drilling. When the work discussed here began, the RRI data base contained almost 3,000injury reports covering 7 years (to the end of 1986; and taken from fixedplatforms and mobile drilling rigs. Data Base Used in This Study. In preparing a design review of the drillingmodule preparing a design review of the drilling module on a new productionplatform, we could use only those injuries in the data base that had occurredon installations with certain minimum points of similarity with the projectedplatform. To avoid restricting the amount platform. To avoid restricting theamount of acceptable data to the point where it would be difficult to drawvalid or worthwhile conclusions, however, a compromise was necessary. JPT P. 264