Mini-Case Study - OHS & Maintenance Reliability

Seemingly unrelated incidents like fatalities, equipment failure and premature bearing failures may all have related root causes. 

 

One of the problems upon which REASON was used during early testing was experienced by a leading oil drilling company that was suffering a rash of premature bearing failures within a large inventory of new diesel engines.

 

To provide a constant level of power to the drilling bit, the company had installed four new diesel engines on each of their portable drilling rigs. The engines were setup in a computerized, slave/master configuration to provide constant power regardless of the conditions encountered by the bit during drilling. This inventory of new diesel engines for over 60 portable drilling rigs represented a major investment. Each engine was state of the art for the early eighties, with a full array of protective devices including high and low temperature, high and low pressure, and autoclave units that centrifugally forced any metal particles in the engine’s oil against a magnetic screen that triggered alarms and controlled shut downs. Oil testing and usual safeguards had not stopped the problem nor revealed its source.

 

A team of engineers from the engine manufacturer flew to the drilling company’s headquarters, where one Saturday morning the company’s maintenance personnel and the team of engineers produced a causal model with the REASON Root Cause Analysis method and tools. From that exercise came visibility of many controllable root causes within both the company and the manufacturer.

 

The list of root causes read like a design for failure: sensing devices that were so sensitive that alarms were continually sounding and shutting down the rig, alarm systems being disconnected to keep the rig on line, failure of the manufacturer to provide key maintenance change bulletins to the company, and numerous other root causes.

 

Root cause analysis details from that maintenance study were soon to combine with data from other loss events to reveal a critical risk factor that existed broadly within the company, and that was producing losses routinely throughout the company’s operations. Within a three-week period, two other significant loss events occurred in that company, and were analysed with REASON Root Cause Analysis.

 

Because the company did not yet view such data as company knowledge to be managed as a resource, the information from the loss sites would normally never have come together for analysis. The cases appeared to be different problems, and would usually have been treated as such. One event involved the failure of a new man lift when the lift cable failed; the other involved a fatality when the stabilisers on a mobile crane shifted in the soil and caused a steel structure that it was supporting to swing against a worker.

 

Pervasive Root Causes

Because REASON Root Cause Analysis was being tested under a development agreement, the two other loss events were also investigated and analysed with REASON. The three models that were produced by the root cause analysis process found their way to the Operations Director who had insisted upon being personally involved in the test. Although each root cause analysis had produced several different root causes and prevention options, the Operations Director immediately perceived within the three REASON models one root cause that was shared in common between the seemingly dissimilar loss events: an equipment maintenance problem, a field design problem, and a safety problem.

 

Under normal circumstances, the data from these different cases would have come together only in a loss report column. But with the visibility of all root causes provided by the unique REASON Root Cause Analysis data, the Operations Director was able to correct that one shared root cause with a phone call to the corporate director of purchasing. He was instructed that in the future no purchasing clerk would have the authority to substitute a product that was "equal to or better than" without a criterion for comparison, or a clearance from the requisitioner. It seems that three different purchasing agents in three different offices elected to substitute "better than" products that resulted in losses to the company: a premature fatigue failure of a wire rope cable on a man lift that achieved faster acceleration by reducing the diameter of a pulley wheel, the rental of a bigger and heavier duty mobile crane with more lifting capacity that sank into the soil at a critical time, and more efficient oil filters that worked so well that they filtered out essential synthetic lubricants from the oil of those new diesel engines.

 

Systemic Causes: The Primary REASON we Conduct Root Cause Analyses

Root cause analysis should find all of the systemic causes of a problem. The point of relating the circumstances of these early tests is to emphasise to you the importance of adopting a root cause analysis method that identifies all root causes when analysing your problems. REASON Root Cause Analysis data provides decision-makers with an understanding of the entire system. It details all options, and provides the opportunity to analyse for causes that are shared in common between systems, as illustrated in this early test example. The REASON Root Cause Analysis methodology provides an objective, standard operating procedure that teaches you how to ask the right questions at the right time, to get the right answers.

 

All other root cause approaches rely upon someone’s assumptions of what should be done. These other approaches use gimmicks like checklists and cause lists to focus upon only a small portion of the causal system.

 

The costly reality check comes when root causes and key prevention options go undetected and unreported, when prevention opportunities are lost, and when managers discover that their analysis system has induced them to make an unsound decision by filtering key data away from their view.

 

Without the opportunity to view the entire problem with all root causes identified, the likelihood of a manager finding shared causal factors existing in common among multiple events, as described above, is remote. It should be noted that these shared causal factors are the "systemic" root causes that are the primary target and goal of professional root cause analyses.

 

 

Links to MORE REASON Root Cause Analysis …

Full Detailed Case Study - Maintenance and OHS Safety

Mini Case Study - Front line operations staff using REASON for RCA’s

Case Study – Air Safety Investigators Conference Investigation Paper

Case Study - Hanford Plutonium Production Complex (includes slides and presentation PDF)

Criteria for evaluating root cause analysis systems

 

NB: Many national organisations participated in the REASON Beta testing, including Texas Instruments, Brown and Root, Delta USA, and Marathon LeTourneau. The testing environments ranged from the manufacture of computer chips to the pouring of molten steel.