W.R. Banick and C. Wei
11750 Katy Freeway, Houston, TX 77079
The HAZOP method is commonly used in identifying causes and consequences of hazard scenarios and the required safeguards in confirming the engineering design for process safety. The LOPA method is then often used to provide a semi-quantitative evaluation of the adequacy of the identified safeguards to reduce risk to an acceptable level. Where gaps exist, this method provides initial input into determination of required risk reduction factors or safety integrity levels for safety instrumented systems. Integration of HAZOP and LOPA techniques into a single study has been accomplished and this provides the practical advantage of utilizing the same team of subject matter experts to complete the evaluation of hazard scenarios, to identify the need for further risk reduction, and to recommend the specific safeguards required and confirm that the resulting residual risk will be acceptable.
While practical and efficient, the one-step comprehensive integrated HAZOP/LOPA method can have pitfalls. For example, the LOPA method isn’t always appropriate for scenarios that may have high initiating event frequencies, such as repetitive manual or semi-automatic batch operations. In this case, use of standard initiating event frequencies for control failures or operator error could result in understating the risk. A human factors task analysis may be the more appropriate tool. Likewise, HAZOP/LOPA teams are usually instructed to provide concrete recommendations when additional safeguards are required, rather than just calling for further study. In some cases this could lead to costly overdesign. In particular, use of LOPA alone to confirm risk acceptability can lead to complex solutions to meet the relatively conservative order-of-magnitude threshold requirements for independent protection layers (IPLs). While LOPA is a good risk screening tool, it may not always be the most effective design tool.
This paper presents a risk assessment approach incorporating human factors task analysis and fault tree and/or event tree analysis following the traditional HAZOP/LOPA analysis. This approach allowed for a more complete evaluation of the engineering design and safe operation options. To illustrate this approach, an example is shown on the selection of appropriate engineering design options for cyclic regeneration of dense phase ethylene treaters due to their unique safety concerns. Using a combination of HAZOP scenario identification, LOPA, human factors task analysis and fault tree/event tree analysis, a cost effective design solution is developed to manage the unique risks associated with dense phase ethylene treater and regeneration systems. These techniques not only avoid the cost and complexity of implementing a SIL-3 instrumented safety system that are called for by HAZOP/LOPA, they also identify that the risks associated with routine opening of process equipment for blinding and un-blinding could be avoided as well.