Structure Critical Risk, Incident and Audit Data on Bowtie Barriers

Objectives/Scope:

To explain the ability to combine risk, audit and incident data onto bow tie
barriers for a threefold view of barrier management.

To offer solutions to engineers and managers who can suffer from information
paucity, overload, delayed delivery or indigestible and unrelated data.

Methods, Procedures, Process:

Using slides, I will demonstrate the sources of data and the inherent
problems interpreting them into meaningful data. I will then show how the data
can be designed around bow tie barriers as the common link.

The method is based on proprietary software.

Results, Observations, Conclusions

The output of the barrier linking will be demonstrated on visual barriers
and in tables and reports, with techniques to filter the data into specific
areas that need attention. The user can therefore target areas of concern and
create appropriate responses. I will relate the output to specific items on the
four pillars of Process Safety Management Systems.

Benefits of the Paper:

The safety and risk engineer can benefit from the presentation and technique
by understanding

- How to capture, organise and filter relevant information from the
cacophony of information available

- How to use these aspects of bow ties to assist compliance with the
objectives of RBPS

Abstract:

“In a risk management context, big data often refers to data which has a
direct relationship to the operation (e.g. percentage of downtime, maintenance
data, and sensor data like wind speed measuring), to risk analysis (e.g. risk
registers and fatigue studies), and to risk assurance data (e.g. incident
findings and audit results). There is much data available, but many
organisations struggle to effectively use the frequently-fragmented data. This
creates a challenge. How can we unify this risk critical data in a meaningful
way to assure operational excellence?

The solution needs to go beyond just bringing data together in one database.
All data needs to speak the same language to combine, weigh and compare.
Barrier-based risk management has proven to be a successful methodology
language to bring operational, risk analysis and risk assurance data together.

Many chemical organisations have already adopted barrier modelling by using
bowtie diagrams as their risk assessment methodology. Different data sources
can be plotted on these bowtie diagrams and operations can make risk-based
decisions based on these “live” bowtie diagrams.

Incident data as a source

The high volume of incident reports at large organisations is both a
blessing and a curse. On the one hand, it provides opportunities to learn about
issues and areas for improvement but it is difficult to detect meaningful
trends or prioritise problems.

There are two connected reasons for this. First, there is no intelligent
structure for reporting incidents, which causes reporters to report superficial
details and abstract categories without capturing the essence of the incident.
Second, because the input is lacking, the analysis of incident data is also
superficial at best, focusing on for instance graphs of high-level categories.

Audit data as a source

Traditional auditing often covers management system elements, but relating
this to the actual performance of barriers that are designed to avoid MAEs is
slow, difficult, or not even attempted.

We will show how incident reporting and audit results can be used as sources
of data that provide valuable information on the state of your barriers. Both
can be linked to specific bow tie barriers, providing a threefold view of
barrier management

Consolidating the data

CGE developed an approach that uses a bowtie risk assessment as the
underlying structure for incident reporting and/or audit surveys. Responses to
audit questions are automatically attached to barriers to provide instant
visual recognition of the state of the barrier, even on a daily basis if
required. Incident data is also mapped to barriers on a structure that is
well suited for trending, while at the same time containing much specific incident
information.

The approach was designed to allow trending of critical performance data
from many events and audits, and identify concrete actionable areas for
improvement. A case study with the Military Aviation Authority will be
discussed.

Simplified view on risk, incident and audit data combined onto one barrier.