Untying the bowtie to understand probability
When it comes to calculating risk, it’s pointless dealing in absolutes. Much of the work we do involves locking down possible scenarios so that a shared understanding can be achieved, and from that, optimal plans devised for management. Fundamental to this process is to understand probability – the likelihood that an event occurs or is likely to occur.
The Bowtie is often misjudged to be just a pretty picture, not serious enough to compete with the rigour of mathematical modelling techniques. But in fact, it is the love child of two well accepted probabilistic modelling tools: Fault and Event Trees, and retains large portions of their capabilities.
In this article, we’re going to look a little closer at the two halves of the Bowtie and how they can help us improve our understanding of risk by coming together.
The left hand side of a Bowtie is essentially a simplified version of what is known as a Fault Tree – a deductive failure analysis method that examines a system from the top down. As found in Bowties, Fault Tree methodology models events (normally represented by causes) passing through boolean logic gates (simplified down to preventative controls, escalation and enabling events, etc.), which are designed to reduce the probability (that is, likelihood) of loss of control (loss of containment, etc.) and determine the optimal ways to minimise risk.
As calculating event probabilities isn’t always straightforward or possible, the Bowtie method as supported in Meercat RiskView, simplifies the process by overlaying qualitative terms (typically sourced from the risk matrix) to make the Fault Tree methodology easier to understand and, therefore, more powerful.
At the centre of every Bowtie is the Top Event, the point at which control is lost and the context switches from prevention to mitigation, or when the Fault Tree joins on to the Event Tree.
After the Top Event is a simplified version of the Event Tree. In this case, the system examines the possible outcomes (consequences) from the Top Event, assessing probability along every potential pathway, including measuring the success or failure (of mitigating controls) to bring the system back under control and therefore avert specific (more significant or catastrophic) consequences.
Layer of Protection Analysis (LOPA) is a specialised usage of Event Tree analysis and is widely used to assess the effectiveness of a combination of controls – think of a “lines of defence” model – to avoid serious loss.
Bringing the trees together
The story goes that Fault and Event Trees were first brought together back in the 1970s around ICI, but the mathematical theory of probability can be traced to the 16th century and gambling. In the succeeding years, technology has helped engage the power of these approaches to help improve the probability that billions of people around the world get to live with the successes of the modern world without suffering the consequences of the failures.
So, to bring all this together, you could say that tools like Meercat RiskView help companies capture workforce knowledge and insight through a range of established risk methodologies and modelling tools to deliver defendable and actionable decisions that have a high probability of increasing commitment and investment for a better world.
To learn more about Bowtie Analysis click here.