Chapter 1
Critical Points Concerning Root Cause Analysis (RCA)

'Addiction is finding a quick and dirty solution to the symptom of the problem, which prevents or distracts one from the harder and longer-term task of solving the real problem'.

- Donella H. Meadows1

Immediate, Contributing and Root Causes

If someone asks, 'Why did the Titanic sink?', the reply, 'Because it hit an iceberg', is a reasonable first answer. Specifically, the immediate cause that led to the sinking of the Titanic was that it hit an iceberg whilst travelling at around 20 knots on 14 April 1912.

However, an immediate cause is not the same as a root cause.

An immediate cause is the event or trigger that sets off a chain of events that results in an adverse outcome: think of a spark that might light a fire.

A contributing cause is something that 'sets the stage' for an immediate cause to create an adverse impact: think of dry tinder on a forest floor or the limited number of lifeboats on the Titanic.

So, beyond immediate and contributing causes, root causes are the reasons why things didn't go as planned.

Root cause analysis (RCA) describes a range of tools and techniques that examine what, when, where and how something happened, and who was involved, but ultimately seek to understand why something happened. It's about properly diagnosing an illness and - if we can - finding a cure, not just offering 'sticking plaster' solutions.2 So, RCA can help you better understand what can be done to stop problems from arising in the first place, or if you can't prevent problems, be clear about what can be done to make any setbacks tolerable.

RCA Fundamentals: Facts, Timelines and Causality

The first cornerstone for an RCA, or inquiry, is gathering relevant facts and evidence in enough detail. Secondly, we need to establish a timeline of what happened; after all, one thing can't cause another without occurring before, or at the same time, as something else. And then we need to cross-check whether one thing is really causing the other or just loosely associated with it.3

Given that facts and evidence are central to an effective RCA or inquiry, the root causes of a specific setback, or disaster, will always depend on the precise facts and circumstances at the time. This is where forensic analysis comes in. And many of the crime and crash investigation programmes we can see on TV show us clearly how much can be involved to get to the truth. However, for this overview book, I want to concentrate on establishing root causes after this hard work has been done.

The Bowtie Diagram: Thinking through What Can Happen and What to Do about It

The Bowtie diagram was developed over 40 years ago in the chemical industry.4 It sets out how you can picture an event that has gone wrong or might go wrong. It is a very useful way of helping you think through some of the most essential ideas in RCA. An example of a Bowtie is provided in Diagram 1.1:

Diagram 1.1 Bowtie diagram (illustrative)

You read the diagram from left to right on the top row and then again on the bottom row, thinking about measures to ensure that incidents do not occur in the first place (e.g. through preventative and detective measures). Or, if they occur, they do not have significant adverse consequences (e.g. because of recovery measures that can be deployed after an incident).5

Let's consider travelling by air as a simple example. We can think about the ingredients that make up a Bowtie in this context:

• A key objective is to travel safely without damage to the plane or injury to any passengers or crew.
• The threats include the risk that an aircraft might hit something.
• Preventative measures to avoid encountering threats in the first place include flight plans for each aircraft, air traffic control with radar, a competent crew who can steer the aircraft away from danger, a plane that is capable of manoeuvring safely, etc.
• Detective measures to spot threats coming up include radar on aircraft, anti-collision devices, the crew who will look out for other aircraft, etc.
• Recovery measures include emergency procedures in case of an incident, trained crew capable of dealing with difficult events, air masks for passengers, etc.

Exactly what measures will be needed will depend on clearly understanding the threats the aircraft might face. But the overall message is that there should be, and are, multiple ways to keep an aircraft and its passengers safe, and this is one of the reasons flying is so safe because we are not relying on just one or two measures to protect us. Indeed, we have an entire system that is designed and operated, as much as possible, to support air travel and keep passengers safe.

When you think about air safety, prevention and detection are far preferable to recovery because there may be only so many things you can do after an incident occurs. Thus, the Bowtie reminds us to consider, in each context, the appropriate balance between prevention, detection and recovery measures.6

As well as using the Bowtie to think through aircraft and chemical plant safety, etc., it can be used for medical procedures and in a range of other domains.7 Of course, in daily life we don't usually explicitly refer to preventative, detective or recovery measures, but much of what goes on in practice seeks to achieve those ends.

Prevention, Detection and Recovery Measures

With the Bowtie perspective, we can start the process of RCA if we fail to achieve an objective. We can 'unpack' causes, in a step-by-step way:

• First, was the objective we had clear and did we understand the threats to achieving it?
• Second, what measures were planned, or designed, to prevent and/or detect things from things going off track, or recover afterwards?
• Third, were the measures implemented in practice and working effectively?

Considering how things are planned or designed provides an illustration of how RCA techniques can be used to anticipate and engineer away potential problems before they occur, increasing our chances of success. This takes us to High-Reliability Organisations (HROs).

High-Reliability Organisations and RCA

HROs can be found in aviation, oil and gas, nuclear, the chemical industry and other areas with very high stakes.

Consider a chemical plant where a pipeline starts to leak. You would expect the operators to be able to detect a pressure drop and then organise a rapid response to look at the leak. After that, you would expect them to repair or replace the pipe and clean up any pollution. The mindset of identifying, planning for and fixing problems is sometimes called 'single-loop' learning, which can be summarised in Diagram 1.2.8

However, imagine that a few months later, there is another leak in a pipe and then another leak. It would likely become clear that a series of relatively quick-fix solutions aren't good enough; you need to go deeper into the underlying reasons why similar issues are arising. For example, is the pressure in the pipes too high? Or are the lines being damaged by external factors, for example, falling branches?

Thinking about why problems might be recurring takes us to what is called 'double-loop' learning, as illustrated in Diagram 1.3.

Diagram 1.2 Single-loop learning

Diagram 1.3 Double-loop learning

'Quick fixes' have their place, but the smart thing to do is to recognise that seeing the same, or similar, problems repeatedly is likely to be a sign that we have missed some other factor that's causing problems. In the example given, maybe we missed opportunities to protect the pipeline adequately. Such a hypothesis could be investigated, and if it explained why the problem was recurring, you would work on this. And this way of thinking might encourage you to find even smarter ways to stop a leak in the first place. For example, you might reroute the pipeline or change the chemicals you are using.

So, HROs aim to maintain a high level of safety, quality, security, etc., over an extended period, by continuously improving operational activities beyond quick fixes. The key ingredients of an HRO are as follows:9

• A constant recognition that problems and failures might arise;
• Continuous monitoring and awareness of what is happening and what is changing;
• A reluctance to oversimplify things, meaning respect for expert views and insights;
• A commitment to resilience; designing 'layers of defence' that will stop or minimise the chances of minor problems turning into catastrophes.

An HRO perspective says, 'keep looking out for potential trouble' and 'it's better to be safe than sorry', even if there is an additional cost in the short term.

HROs have developed a range of tools and techniques to minimise the chances of problems arising by thinking about (i) design excellence, and (ii) excellence in operational delivery and monitoring, and...