The UK water sector has made genuine progress on safety over the past decade. Investment in monitoring technology has improved, reporting processes have matured, and many organisations have built safety cultures that would be unrecognisable compared to where they were twenty years ago. That progress deserves acknowledgement.
And yet the regulatory picture tells a more complex story. The October 2025 government proposals to give the Environment Agency powers to issue automatic fines for permit breaches, alongside executive accountability provisions and a £104 billion infrastructure investment commitment, did not emerge from nowhere. Between 2015 and 2025, the Environment Agency opened 11,474 investigations into water companies, of which only 58 resulted in prosecutions. Over 600,000 raw sewage discharges were recorded in UK waterways in 2023 alone. These are not figures that point to a lack of effort. They point to something more specific: a persistent pattern of the same categories of failure appearing across different organisations, different sites, and different years.
As said by our CEO, Mark Rushton,
"When a pattern repeats at that scale, it is worth asking whether the way we investigate incidents is giving us the full picture, or whether something important is being missed."
Human error is where most investigations end. It should be where they begin.
COMET's Head of Investigations, Alan Smith, spent his earlier career as a criminal detective before moving into industry investigation. The distinction he draws between the two disciplines is one of the most useful frames we have encountered for thinking about why incident investigation so often fails to prevent recurrence.
In criminal investigation, the objective is to establish who was responsible and prove what they did. Understanding why the individual acted as they did is largely irrelevant to the outcome. In industry investigation, the why is everything. Without it, you cannot design a fix that actually holds.
What Alan observed when he moved into industry was that organisations genuinely committed to blame-free culture were still, in practice, defaulting to worker-focused corrective actions. More training. More detailed procedures. Increased supervision. These responses have their place, but they share a common limitation: they target the individual at the point of failure rather than the conditions that made the failure likely. Alan draws a comparison with the prison system, where attempts to fix individuals through rehabilitation carry an estimated success rate of less than 3 per cent. Organisations that respond to every incident by targeting the person involved tend to see similarly limited returns.
The more instructive question, when a worker makes an error, is not what they did wrong but what made that error likely. At the moment a human error occurs, the person involved almost certainly believed they were doing the right thing. They were working with the information, tools, time, and organisational context available to them. As one researcher on human factors has put it, attributing an incident to human error is about as useful as attributing a fall to gravity.
The pattern is not unique to water. According to the FAA and NTSB, pilot error accounts for between 60 and 80 per cent of aviation accidents. Allianz Global Corporate and Specialty, analysing nearly 15,000 marine liability insurance claims, found that human error was a primary factor in 75 per cent of the value of all claims, and that between 75 and 96 per cent of marine accidents involve human error in some form. Across every high-risk industry, the conclusion is the same: human error is ubiquitous precisely because the conditions that produce it are rarely examined.
Why the water sector faces this challenge in a specific way
The operational environment in water and wastewater creates a set of conditions that make human and organisational factors particularly hard to surface through older generation investigation methods.
A significant proportion of the workforce operates remotely, often alone or in small teams without direct supervision. Routine tasks, which make up the majority of operational activity, carry a particular risk: familiarity creates the conditions for normalised deviation, where shortcuts and adaptations accumulate quietly over time until something goes wrong. Procedures can often get shortened not through carelessness but through the entirely rational process of adapting to a working reality that the written procedure did not anticipate.
The Performance Influencing Factors that shape behaviour in these environments fall into three broad categories, each of which carries specific weight in water operations:
•Task factors: the clarity and practicality of procedures, time available to complete tasks safely, the quality of tools and equipment, and the degree to which attention is divided across competing demands.
•Individual factors: fatigue, stress, experience gaps, and the mental models workers carry about how a system is supposed to behave, including assumptions built up over years that may no longer reflect current operational reality.
•Organisational factors: the quality of supervision and support, the pressure to meet operational targets, the culture around near-miss reporting, and whether the organisation has demonstrated, through its actions, that it genuinely learns from previous events.
A purely technical investigation will surface equipment failures and procedural deviations. It will not surface the shift pattern that correlates with higher incident rates, the monitoring system whose alerts have been discounted through overexposure, or the maintenance schedule that looks adequate on paper but has been quietly stretched by resource pressure. These factors are not hidden. They are simply invisible to an investigation approach that stops at the person.
One practical tool for testing this is the substitution test: if someone else with the same skills and experience had been placed in the same situation, could they have made the same error? In most cases, the answer is yes, and when it is, the investigation has confirmed that this is a systemic issue, not an individual one. That confirmation is where the real investigation work begins.
What a different approach makes possible
The organisations making the most meaningful progress on prevention share a common characteristic: they have stopped treating investigations as a compliance output and started treating them as a cumulative data asset.
This requires a consistent investigation framework applied across every failure event, one that captures not just the immediate cause but the communications, operating environment, management decisions, equipment condition and training context that surrounded it. Each investigation conducted in this way produces codeable, comparable data. Over time, that dataset reveals what no individual investigation could: the human factors that cluster in particular operational contexts, the organisational conditions that keep appearing in serious events, and the early warning patterns that precede those more significant incidents.
COMET's investigation methodology is built around this principle. Our AI Assistant, purpose-built for incident investigation and root cause analysis, supports investigators directly within the workflow, helping surface barrier failures, identify contributing factors, and ensure human factors are properly analysed rather than reduced to a "human error" conclusion. The design principle throughout is that the AI supports, and the investigator decides. For safety managers and senior leaders, or those keen to find out more about what might be hidden behind the "big data", our Insight AI capability allows the entire investigation portfolio to be interrogated through natural language, making it possible to ask which human and organisational factors are concentrating across the business and which interventions are correlating with actual incident reduction.
Several organisations in the water and utilities industry use COMET to improve their incident investigation and RCA capabilities. Across this client base, the consistent outcomes are improved visibility into previously unseen risks, greater consistency in the investigation process across sites, and actions that address why things went wrong rather than simply recording that they did.
The independent validation of the COMET methodology and system for incident investigations & RCA is also growing. In December 2025, Verdantix published its Buyer's Guide to Water Management Software, covering 18 vendors assessed for the water and wastewater sector. In its profile of Intelex, Verdantix noted that the company expanded its investigation and root cause analysis capabilities through a partnership with COMET in October 2025, recognising investigation depth as a component of complete water management capability.
Learn more about our partnership with Intelex
Download the Verdantix Buyer's Guide: Water Management Software 2025.
The opportunity in front of the sector
The water industry is operating under greater scrutiny than at any point in recent memory, and some of that scrutiny is fair. But the organisations best placed to respond are not those that improve their compliance reporting or tighten their disciplinary processes. They are those that use the current moment as a catalyst to genuinely improve the quality of their investigation and prevention work.
That means treating human error not as a finding to be recorded but as a signal to be read. It means building investigation practices that look honestly at the organisational and environmental conditions that shape behaviour, not just the behaviour itself. And it means using the data that already exists inside most organisations to identify patterns before they become incidents rather than after.
The sector has the capability and the experience to do this well. The question is whether the investigation approach currently in place is designed to make it possible.
Find out more
If this piece has prompted questions about how your organisation approaches investigation and root cause analysis, we would be glad to help you explore them.
• Explore COMET for Utilities and Waste Management
• Learn about COMET for investigations and RCA
• Book a discovery call to see if COMET can help your organisation
Sources
Environment Agency/Parliament. Water Companies: Fines. Hansard, House of Lords, February 2025
Surfers Against Sewage. Sewage discharge data 2023.
Allianz Global Corporate and Specialty. Safety and Shipping: Casualties report.
Smith, A. Human error: the start of your investigation, not the end. COMET, November 2024.
