How to avoid Ofwat pollution and sewer flooding penalties

Water companies entered AMP8 with more visibility and a larger investment programme. Monitored storm-overflow spills fell slightly in 2024, but total spill hours stayed flat at about 3.6m hours. Serious pollution incidents rose 60%. Ofwat reported about 6,000 internal sewer flooding incidents and more than 50,000 external incidents in 2023-24.

You cannot stop every incident on the day it happens. Pipes, overflows, rainfall, and asset condition matter. But the way a team handles each incident decides how much extra money the company pays. Report it quickly, collect good evidence early, and send crews to the right places, and the same incident can cost less.

That is the point for the water and wastewater sector. Monitoring has made more failures visible before the assets have made those failures less frequent. There is a lot a team can do with each incident to keep the penalty down. Bringing the number of incidents down is the slower job, and it takes money spent on the network.

1. Self-reporting speed

The first choice a team faces is whether to report a spill before they are sure what caused it. Report late, or let the regulator find it first, and the company loses marks the Environment Agency is already counting.

Self-reporting is a scored indicator in the Environment Agency's Environmental Performance Assessment, or EPA, the annual 1-4 star rating of water and sewerage companies. The Environment Agency expects at least 80% of all incidents to be self-reported, rising above 90% for treatment works and pumping stations. In 2024, the sector reached 85% and 92%, but four companies fell below target.

It is tempting to wait until the cause is clear before reporting a spill. Waiting usually makes things worse. The regulator is measuring whether the company recognised the incident quickly. The cause can be added later.

From the start of 2026, the Environment Agency's Water Industry Reporting Incidents guidance, known as WIRI, tightens what must be reported and how fast. That includes discharges where the pollutant came from another party, such as oil, milk, or paint.

So send the crew, but report the incident early as well, and write down what was known and when.

2. Evidence quality and cause

How bad an incident turns out to be is often decided in the first few hours, before the team knows the full story. Those hours are also when evidence gets lost: samples not taken, photos not captured, readings not saved.

Under the Common Incident Classification Scheme, incidents are classified from category 1 to category 4 based on evidenced impact. Category 1 is major, with serious, persistent, or extensive impact. Category 2 is significant, category 3 is minor, and category 4 has no substantiated impact.

The category matters because it decides how serious the consequences are. The Environment Agency attends category 1 and 2 incidents where possible, and serious incidents can carry civil sanctions, prosecution risk, and unlimited fines.

A good pollution incident file starts while the site is still live. It should hold samples, telemetry, the first known timeline, and what is still unknown. Waiting leaves the company reconstructing the record from partial logs and late photographs.

Cause still matters. The asset owner, pollutant source, and reporting duty may sit in different places. A company may need to report and evidence a pollution incident even when the pollutant came from another party. The file needs to show the environmental impact and the company's role in prevention, containment, and recovery.

The regulator needs a clear timeline and enough evidence to judge what happened.

3. Prioritising incidents in a storm

In a storm it is easy to work through whichever incidents are generating the most calls and alarms, rather than the ones that will do the most harm. Wet weather brings more alarms and more customer calls at once, with crews already committed elsewhere.

Sewer flooding has a direct human consequence. Internal sewer flooding means sewage entering a building or passing below a suspended floor. External sewer flooding affects gardens, driveways, and other land within the property boundary.

During a storm, a team needs to know more than the order the alarms came in. Vulnerable-customer information, flooding history, telemetry, and crew location all affect which incident should be handled first. If the queue follows the alarms, households that need urgent help may be left waiting.

DG5, the historical Ofwat register of properties at risk of internal sewer flooding, still matters because many teams used it to decide which properties needed attention first. The live commitments now sit around internal and external sewer flooding performance, and Ofwat required companies to review risk registers through RD 23/08.

Good triage means reaching customers fast, sorting internal flooding from external, merging duplicate reports, and noting why each crew was sent where it was. That protects people and premises and leaves a record for later review.

4. The annual sewer-flooding target

Ofwat does not fine a company over a single flooded home. It looks at how many flooding incidents there were across the whole year, set against a target. Customers feel the harm one property at a time, but Ofwat calculates the payment from annual performance.

An ODI is an outcome delivery incentive: Ofwat applies rewards or underperformance payments against agreed performance commitments. For sewer flooding, performance is commonly measured per 10,000 sewer connections.

A single flooding event does not come with a standard Ofwat bill attached to it. The annual number against target determines the underperformance payment, and those payments are returned to customers with a roughly two-year lag.

Ofwat pollution and flooding penalties do not follow one tariff. ODIs are company-specific underperformance payments tied to performance commitments. Wider regulatory penalties can be much larger when Ofwat finds serious breaches. In May 2025, Ofwat imposed a £122.7m penalty on Thames Water, including £104.5m for wastewater failures and £18.2m for a dividend-rule breach.

Sewer flooding teams need to record each event accurately. The record needs to show where the flooding happened, whether it was internal or external, whether duplicate reports have been merged, and what the team did.

Better handling lowers the cost, not the number of incidents

The honest limit is that better handling makes each incident cheaper. It does not by itself reduce the number of incidents.

The sector can now see much more of what is happening. By the end of 2023, every storm overflow had event duration monitoring installed. Near-real-time data has been mandatory since 1 January 2025. In 2024, monitored storm overflows still recorded 450,398 spills and about 3.6m hours.

Visibility helps companies find incidents, report them, and explain them. It also shows where teams decided too slowly or kept weak evidence.

The test is a wet day when alarms, calls, and crews arrive at once. The company needs to report early, start the evidence file while the site is active, and send crews to the cases where delay will make the harm worse. The sewer-flooding record then has to hold up at year end.

That work matters now because asset improvement takes time. AMP8 allows about £104bn of investment for 2025-30, including roughly £12bn for storm-overflow reduction. That spending is what can reduce spills, floods, and serious incidents over time.

Until that investment brings the number of incidents down, handling each one well is the part a company can control. Do it well, and the same incident costs less.