(126a) Problems with Hazops and How to Correct Them

Duhon and Sutton (2010) identified multiple reasons why we don’t learn all we should from HAZOPs.  The paper discussed some of the cognitive, social, organizational and procedural factors that limit the effectiveness of projects in general and of HAZOPs in particular.  Ideas for improving the HAZOP process were suggested. 

Duhon (2011) suggested modifications to the HAZOP process to correct some of those identiifed issues.  

Duhon and Cronin (in progress -2015) addressed problems specific to risk assessment in HAZOPs.  Here insights from LOPA methods are used to simplify and improve hazard consequence and frequency assessment.  

This proposed paper will combine the observations from these previous three papers into an overall suggested HAZOP technique.

 One of the main problems with HAZOPs is that small equipment-based nodes create tunnel vision.  The suggested improvement is to define stream-based nodes for the flow deviation discussions.  A stream-based node may run through several equipment items that modify the stream in some way.  For example:  a gas product stream may be considered a single stream from its inception through multiple stages of compression, dehydration, metering, and discharge to the gas pipeline (and even through the pipeline to the customer).  The flow deviation discussion is dramatically improved and streamlined – a change in flow anywhere in the stream-based node affects flow everywhere or at least everywhere downstream.

 Stream-based nodes facilitate a discussion of operation.  The objective of most operating procedures is to establish or stop a flow.

 Other deviations (pressure, temperature, level) require smaller nodes, but these discussions are greatly facilitated by starting with the flow discussion at a higher level.

 A problem with most HAZOPs is fragmented reporting schemes.  Where a flow deviation in node 3 results in a pressure deviation in Node 6, this Node 6 pressure deviation may be reported under flow deviation in Node 3.  This results in reports that are too long, too confusing, and rarely used.

 The risk matrix approach used for risk ranking in most HAZOPs is ambiguous.  It is readily improved and simplified by applying some LOPA insights.

 The HAZOP method is probably the most commonly applied hazard analysis method in the process industries.  But HAZOPs are not as effective as they should be.  The proposed method, developed via lessons learned from multiple HAZOPs over the past 15 years is a significant improvement to the process.

References:

 Duhon, Sutton, (2010), “Why We Don’t Learn What WE Should From HAZOPS”, SPE-120735-PA, SPE Projects, Facilities and Cosntruction, June 2010

 Duhon (2011), “Stream-based HAZOP, A More Effective HAZOP Method”, Presented at SPE Americas E&P Health, Safety, Security and Environmental Conference, Houston, TX, USA, 21-23 March 2011

 Duhon, Cronin (2015), “Risk Assessment in HAZOPs”, Proposed for presentation at the SPE Americas E&P Health, Safety, Security and Environmental Conference, Denver, CO, USA, 16-18 March 2015