Learning from microbiological failures

The following tables show all the compliance failures within the year, the full list is published within Annex B.

Parameter	Standard	Total number of tests	Number of tests not meeting the standard	Percentage compliance (3 DP)
Clostridium Perfringens (sulphite-reducing clostridia) (confirmed)	0 number/100 ml	24,840	8	99.968%
Turbidity – indicator	1 NTU	171,438	19	99.989%
E. coli (confirmed)	0 number/100 ml	171,591	3	99.998%
Total coliforms (confirmed)	0 number/100 ml	171,598	80	99.953%
Total		539,467	110	99.980%

Table 6 – Failures at treatment works in 2025

Microbiological failures

E. coli at treatment works

E. coli is an indicator of faecal contamination, and robust contingency plans must be in place to mitigate the impact of any detections and to protect consumers from water that may be considered unwholesome or unfit for consumption.

In 2025, there were three detections of E. coli in samples taken from treatment works out of 171,550 tests (Table 7). A further four detections were identified from 182,499 tests at service reservoirs, and 26 detections from 156,949 tests at consumer properties.

In total, there were 33 E. coli detections in 2025. This represents a continued year-on-year reduction from over 40 detections in 2024 and 44 in 2023, indicating consistently high and improving compliance for this parameter across the industry.

Company code	Site name	DWI action
BRL	Shipton Moyne	Satisfactory investigation, did not identify cause
NES	Coldfair Green	Recommendations made
TMS	Lullingstone	Recommendations made

Table 7 – treatment works with E. coli detections

Bristol Water – Shipton Moyne

In July 2025, a single E. coli detection was reported in a regulatory sample taken at Bristol Water’s Shipton Moyne works. The detection was also reported as an event.

The company’s investigation established that the raw water quality remained within normal operating parameters and they identified no issues with the works treatment processes. Company checks confirmed that the disinfection systems, both UV and chlorination, were operating correctly at the time of the detection. No deficiencies were identified in the sampling arrangements at the works, nor in sampling or laboratory practices. All samples taken in response to the breach returned satisfactory results.

Due to elevated seasonal demand, Bristol Water reported a short delay in removing the works from supply to allow the inspection of the treated water tanks. To enable isolation of the site, a temporary pumping station was installed which was operational one week after the initial E. coli detection. Inspection of the treated water tanks found no structural defects, no evidence of ingress, and no conditions that could account for the detection.

As an interim risk mitigation measure, pending removal of the works from supply, Bristol Water increased the chorine setpoint and undertook enhanced monitoring at the works. This continued once the works was returned to supply with no further detections reported providing assurance on the quality of the supply.

The company conducted a comprehensive investigation, the root cause of the E. coli detection could not be determined.

Northumbrian, Essex and Suffolk Water – Coldfair Green

Coldfair Green works is a pristine groundwater site located in Suffolk. The works had an E. coli (and coliform) detection in a final water sample on 30 June. The company’s investigation sampling found turbidity measured in a raw water sample on 2 July was 43 NTU, which was significantly elevated compared with historical data for this borehole; raw water sampling data from 2022 indicated a maximum previously recorded turbidity of 27 NTU, with typical values considerably lower. In addition, Clostridium perfringens was detected in an investigation sample collected on 2 July from the downstream asset, Leiston tower, two days after the E. coli detection from the works. The company’s investigation neither addressed nor linked the two results and a root cause analysis found no cause for either microbiological detection. Contact tank inspections conducted subsequently did not identify any ingress or any water quality impacting issues and all site process trends presented as satisfactory.

The Inspectorate recommended the company undertake timeseries sampling at all individual raw water boreholes supplying Coldfair Green works to assess incoming water quality and review the capability of the treatment process to manage elevated raw water challenges, in accordance with regulation 18. The company did not complete this exercise and instead advised they would enhance the frequency of raw water samples taken for bacteriological parameters and turbidity commencing in 2026. The company’s explanation for the elevated turbidity was that it was likely due to oxidation of iron in the sample, although there was no additional iron analysis carried out to evidence this working theory.

The Inspectorate expects that all instances of microbiological contamination, especially by faecal indicators, are appropriately investigated in order to understand and mitigate the root cause as per the requirements of regulations 4, 18 and 27.

Thames Water – Lullingstone treatment works

In July 2025 Thames Water detected E. coli in a compliance sample from Lullingstone works. The detection was also reported as an event. A second compliance sample collected at the works later the same day, and a compliance sample collected the next morning, were satisfactory for microbiological parameters. The works was removed from supply on notification of the E.coli detection. Investigational in-process samples through the works were satisfactory for coliform bacteria including E. coli. Samples from the downstream service reservoir and at consumers’ properties were satisfactory for a wider range of microbiological tests.

The detection was ultimately attributed to the compliance sampling kiosk environment and failure to follow the sampling procedure. A new-style plastic sampling kiosk had been installed. However, the kiosk had limited ventilation which led to a significant build-up of condensation on the sampling pipework. The company subsequently installed additional meshed ventilation at this site and amended the company asset standard for sampling kiosks.

The compliance sample tap at Lullingstone works runs constantly. The sampler had attempted to disinfect the sample tap with water running – which was allowed in the company sampling procedure manual. A recommendation had previously been made by the Inspectorate to amend the sampling procedure manual. Following this detection, the company issued guidance to the sampling teams to turn off the tap and flame sterilise the tap before sampling. Subsequently the company updated and reissued the sampling procedures manual.

E. coli at reservoirs

South East Water – Broader lane

Analysis of a compliance sample collected in May 2025 from South East Water’s Broader Lane No. 1 Service Reservoir returned a result of one E.coli and one total coliform bacterium per 100 ml. A compliance sample collected approximately 20 minutes earlier from the adjacent Broader Lane No. 2 Service Reservoir was satisfactory.

The detection of E. coli indicated the potential for faecal contamination within the supply system. The company removed Broader Lane No. 1 Service Reservoir from service approximately three hours after the result was reported by the laboratory.

The company undertook an investigation to identify the source and extent of contamination. Sampling was carried out at both Broader Lane service reservoirs, upstream and downstream assets and at downstream consumer properties. All investigative samples were satisfactory and did not indicate evidence of wider contamination within the associated supply system.

Meteorological conditions in the week preceding the failure were mixed, comprising two days of rainfall interspersed with predominantly dry conditions. No clear linkage between the weather conditions and the bacteriological failure was established.

An external inspection of the service reservoir was undertaken which did not identify any deficiencies that could reasonably explain the failure. Access hatches were found to be dry and secure, and the sample tap was reported to be in good condition.

Further investigation was undertaken by the use of a remotely operated vehicle while Broader Lane No. 1 Service Reservoir was being drained to waste. This inspection identified no obvious pathway or source of contamination and found the internal condition of the reservoir to be generally good. A subsequent full internal inspection and clean were completed again without identifying any defect or ingress route that could be positively attributed to the failure. The reservoir had also undergone a full inspection in April 2025, at which time no significant deficiencies were identified.

Despite the company carrying out a comprehensive investigation, no definitive root cause for the presence of E. coli was established. Following completion of the inspection and cleaning programme the company returned Broader Lane No. 1 Service Reservoir to service in July 2025.

Following return to supply, the company undertook additional large-volume sampling at Broader Lane No. 1 Service Reservoir for a period of three months, all following enhanced monitoring samples were satisfactory.

Yorkshire Water – Somerset Avenue

A regulatory sample taken at Somerset Avenue Service Reservoir, Baildon, in May 2025 contained 19 E. coli per 100 mL. Subsequent investigatory samples from the downstream distribution system and upstream treatment works were satisfactory, and an external inspection of the structure found no immediate operational deficiencies. As a precautionary measure, Yorkshire Water isolated both reservoir compartments, supplied the area directly from Graincliffe Water Treatment Works, issued a boil water notice to affected consumers, and undertook a detailed investigation. Internal reservoir inspections found no defects that could explain the contamination, while a root cause analysis was unable to identify a definitive source of the E. coli detection.

The company’s mitigation actions included isolating the reservoir, supplying customers directly from the treatment works, carrying out extensive sampling, inspecting reservoir infrastructure, checking upstream assets, flushing the distribution network, and maintaining the restriction of use until three consecutive satisfactory sample sets had been obtained; the boil water notice was withdrawn on 26 May 2025 and the reservoir returned to service on 2 June 2025. Key lessons included improving investigation and asset management processes: confirming adequate vermin protection on service reservoir overflows and washouts, strengthening root cause analysis templates to explicitly consider all contamination pathways, ensuring air valves are maintained and recorded within inspection programmes, considering chemical dosing assets during investigations, and reducing contamination risks around regulatory sampling points.

Yorkshire Water – Rudston

E. coli was detected in a compliance sample taken at Rudston Service Reservoir in August 2025, with a result of 2 cfu per 100 mL. The sample was taken from the reservoir tank outlet and had an adequate chlorine residual at the time of sampling (0.50 mg/L free chlorine and 0.54 mg/L total chlorine). A desktop root cause investigation involving operational and public health teams was undertaken, but no likely cause for the detection could be identified. Subsequent investigations, including an internal reservoir inspection under flood conditions, air valve inspections on the inlet main, and pressure monitoring, found no water quality related defects or operational issues that could reasonably explain the positive sample result.

To mitigate the potential risk and investigate the sample result, Yorkshire Water drained and inspected the service reservoir, inspected all inlet air valves, deployed pressure logging on the inlet main, and undertook enhanced bacteriological monitoring following the reservoir’s return to service in September 2025. Pressure monitoring demonstrated that the inlet main remained under positive pressure throughout operation, reducing the likelihood of contaminant ingress. In addition to routine compliance sampling, the company implemented enhanced monitoring, taking 1-litre bacteriological samples. No further detections of coliform bacteria have been identified in any subsequent samples, and the company’s investigations concluded that there was insufficient evidence to determine a definitive cause for the original E. coli detection.

Yorkshire Water – Drighlington

E. coli was detected (44 per 100 mL) in a routine compliance sample from Drighlington Service Reservoir in October 2025. Yorkshire Water carried out a comprehensive investigation, including repeat sampling, Cryptosporidium monitoring, reservoir inspections and checks of upstream treatment processes. All follow-up samples were satisfactory and no defects or evidence of contamination were found, meaning no root cause could be identified. The Inspectorate concluded that potentially unwholesome water may have been supplied, but there was no further evidence of contamination within the reservoir or distribution network.

The company immediately isolated and drained one reservoir compartment, undertook inspections and introduced enhanced monitoring. However, the second compartment remained in supply for a prolonged period because operational constraints prevented full isolation of the reservoir. The Inspectorate suggested that the company review and improve their reservoir bypass and resilience arrangements to enable rapid isolation during future events and recommended that the company ensure operational decisions are informed by contemporaneous water quality data, considering additional public health protections when full isolation is not possible.

Coliforms at reservoirs

Testing for coliform bacteria gives reassurance that water entering supply was treated adequately to remove bacterial and viral pathogens. Repeated occurrences of coliform bacteria in samples from the same works in one year are thus of concern and require action to be taken. There were 75 coliform failures from 182,576 tests undertaken at reservoirs. In 2025 three reservoirs failed the 95% wholesomeness test with repeated coliform detections.

All companies are expected to complete internal inspections as part of a risk based process, where the history of ingress and the age of the asset are taken into consideration to ensure public health remains protected.

Company	Reservoir name	Parameter	Samples taken	Samples failed	Percentage compliance (2 DP)
YKS	Scar House Sct	Total coliforms (confirmed)	51	3	94.12%
SEW	Cuckfields reservoir	Total coliforms (confirmed)	6	1	83.33%
AFW	Gunters Hill tower	Total coliforms (confirmed)	13	1	92.31%

Table 8 – reservoirs failing the 95% wholesomeness requirement

Yorkshire Water – Scar House suction tank

Coliform bacteria were initially detected in separate regulatory samples collected from the Scar House suction tank in January and June, and were appropriately notified to the Inspectorate as events and compliance failures. Subsequent sampling identified further coliform detections, both in regulatory samples and in investigatory large-volume samples, as well as at consumer properties.

Between August and October, additional coliform detections were identified upstream of the Scar House suction tank, including within the network downstream of Irongate Bridge Water pumping station, at Darley Low Green service reservoir, and at consumer taps. During this period, there were also isolated detections of E. coli and Clostridium perfringens, indicating a broader microbiological risk within the system.

In response to the initial detections, the company implemented a comprehensive sampling programme and, where appropriate, issued health protection advice during the June, August, September, and October events. A large number of water fittings inspections were also undertaken at rural properties to assess backflow protection arrangements and the influence of private supplies within the area.

The investigation identified several contributing factors, including long residence times within a sparsely populated rural network, the historic use of chloramination with low and inconsistently maintained disinfectant residuals, biofilm development within the system, and evidence of ingress at Darley Low Green service reservoir. To address these risks, a secondary chlorination booster system was installed at Eagle Hall service reservoir inlet, upstream of the Scar House suction tank, to improve chlorine residuals and control suspected biofilm growth.

Attempts to increase chlorine residuals within the network led to elevated chlorate concentrations, which the company attributed to a combination of dosing practices and potential backflow from a large consumer site in the area. This resulted in the issue of additional health protection advice in the form of a do not drink notice.

The Inspectorate’s assessment of these events identified concerns relating to the company’s chlorine management strategy, pressure resilience within the rural network, and the management of chlorate. Deficiencies were also noted in consumer communications during the incidents, although the timeliness of issuing initial health protection advice was recognised.

Yorkshire Water acknowledged the seriousness of the events and implemented a range of mitigating actions. These included ceasing chloramination in the affected network to improve chlorine residual stability, revising incident management guidance for senior scientists, and strengthening chlorate management through enhanced training, tighter stock control, improved monitoring, and the establishment of new chemical storage hubs. The company also committed to a wider review of similar low-turnover, chloraminated networks and improvements to customer communication during incidents.

Lessons for the wider water industry highlight that the use of chloramination in low-turnover rural systems may increase the risk of biofilm development and intermittent coliform detections. Interventions to increase free chlorine to manage microbiological risks must be carefully balanced, as they can lead to elevated disinfection by-products, such as trihalomethanes (THMs) and chlorate, and may obscure the underlying root cause. These competing risks should be anticipated, formally assessed, and managed proactively rather than reactively during incidents.

Public health risk assessments must draw on all available and emerging data, including uncertainties, to ensure that appropriate and timely health protection advice is provided. Additionally, even where prior inspections and safeguards (such as air gaps) are in place, large third-party premises with private supplies and on-site treatment must be re-evaluated during incidents. Finally, consumer communication must evolve in line with the incident, ensuring that accurate, address-level information and clear explanations of potential impacts are provided, particularly to vulnerable customers.

South East Water – Cuckfields reservoir

A compliance sample taken from South East Water’s Cuckfield number 1 reservoir on 15 July 2025 detected coliform bacteria. Samples collected from the neighbouring Cuckfield number 2 reservoir and from an upstream asset on the same day were satisfactory, as were investigatory samples collected in response to the failure. Cuckfield number 1 was removed from supply on 24 July 2025 for inspection.

Before this detection, Cuckfield number 1 had last been inspected in December 2024. That inspection identified points of ingress. Following remedial works, successful flood testing and reinspection, the reservoir was returned to supply in June 2025.

In response to the Inspectorate’s assessment of the compliance failure, the company confirmed that a further inspection planned for July 2025 had been postponed while it assessed whether additional works were required at the site. The company identified that, under the current configuration of the two reservoirs, pipework modifications were required to Cuckfield number 2. Inspection of Cuckfield number 1 was planned for completion by the end of November 2025, with return to supply in early December. Cuckfield number 2 would then be removed from supply to allow the required modifications to be completed.

On 4 December 2025, an operational sample collected from Cuckfield number 2 detected both coliform bacteria and Enterococci. The reservoir was removed from supply on 6 December following receipt of the presumptive Enterococci result. As Cuckfield number 1 had not yet been returned to supply following the July detection, the area was supplied using an alternative network configuration. Cuckfield number 1 was returned to supply on 25 March 2026. Cuckfield number 2 remains out of supply until the company completes the required pipework modifications.

Affinity Water Gunters Hill tower

In January 2025, Affinity Water detected a coliform in a regulatory sample collected from Gunters Hill water tower. All investigational samples subsequently collected from the reservoir, the supplying works, and consumer properties returned satisfactory results. The supplying works was confirmed to be operating satisfactorily in the period prior to the detection.

Upon notification of the presumptive coliform result, the company took precautionary action and isolated the tower from supply. The subsequent investigation identified ingress into the reservoir as the most likely root cause of the detection, alongside other contributory issues. The asset remained out of service until the necessary remedial works were completed.

A previous inspection conducted in November 2015 had also identified ingress at the tower. Although the interval between inspections remained within the maximum permissible period, greater application of risk-based methodology by taking into account the asset’s age (constructed in 1967) and its history of ingress may have prompted earlier proactive inspection and intervention ahead of the standard ten-year cycle.

All companies are expected to undertake internal inspections as part of a risk-based approach, considering factors such as asset condition, age, and prior defects, including ingress. This is essential to ensure that risks are appropriately managed and that public health protection is maintained.

Table of contents