Practice Makes Perfect: How Process Simulation Improves Operator Response to Emergencies

No one is perfect, so any safety measure that depends on operators responding perfectly is bound to disappoint. However, practice improves the likelihood that operator response in emergencies will be appropriate and effective.

The role that human error plays as an initiating cause has been widely studied and is routinely incorporated into LOPA studies. Each opportunity to perform a non-routine task under stress is typically taken to be an initiating cause. On the other hand, each opportunity to perform a procedure on which there is frequent training and into which independent checks are built is typically taken to an initiating cause only one time per thousand times the procedure is performed. Training vastly improves the success of procedures.

Operator response to emergency situations is a distinct issue and not as conducive to training and built-in checks. The effectiveness of an operator response is frequently judged solely on the basis of response time available, which must be adequate to provide time to detect, time to decide, time to act, and time for action to take effect, as well as sufficient buffer time. Training is typically taken as a given, but while current guidance suggests that training can be in the form of tabletop exercises, drills, and process simulations, they are silent on the subject of how much training is required and whether the response is adequate.

Routine tasks and procedures provide ample opportunity for training and auditing, and the values used in LOPAs are easy to justify. Emergency responses, on the other hand, are rare and by their very nature, not good training opportunities. While tabletop exercises and drills are valuable, there is increasing interest in using process simulations as the means to assure that operators are prepared to recognize and appropriately respond to abnormal conditions.

This paper discusses the role of human action in hazardous scenarios and compares the assumptions used to quantify the contribution of human action to initiating causes as well as to responses that act as independent layers of protection. It goes on to define the potential improvement in failure rate and probability of failure on demand that can result from using process simulations as part of the training regime for hazardous scenarios.