Data return findings
During 2024, the Inspectorate gathered data through the annual data return which is a statutory reporting requirement of local authorities.
The quality of water supplies from private water supplies in Wales
Under regulation 16 of the Regulations, local authorities must keep records of every private supply in its area and submit these to the Inspectorate by 31 January each year.
For the reporting year 2024, a submission was received from all 22 local authorities in Wales.
In 2024, local authorities carried out 24,512 analyses of private water supplies samples (not including colony counts). This is an increase of 4.90% on the previous year (23,366) indicating a recent improvement in overall sample numbers but this is still some 5% less than the numbers taken in 2019 and pre-pandemic. Figure 2.1 shows a general increase in samples taken since 2020, with 2020 representing an exceptional year due to the impact of the pandemic. This increase may indicate that more resources are being allocated to private water supplies sampling and testing.
Figure 2.1 Number of tests from 2014 to 2024 (without colony counts)
When analysing the test data to produce this report, the Inspectorate identified an anomaly. There was a much greater number of colony-count failures than previous years. The standard for colony counts (specifically three days at 22 degrees centigrade) which is an indicator parameter, is ‘no abnormal change’. There is therefore no ‘prescribed concentration or value’ (PCV). It is the responsibility of local authorities to determine what would be an abnormal change for any particular private water supply. An abnormal change may indicate a problem with the quality of the supply. In public supplies, it is not uncommon to see colony counts between tens and hundreds, which are a stable background levels for a supply or perhaps an asset such as a treated water reservoir within distribution. Investigating this anomaly further, the Inspectorate found that a problem with the validation macros of the data return Excel template meant it would not allow a colony count value of greater than zero (in effect a count of one or more) to be entered as a ‘pass’. Unfortunately, many local authorities decided the best course of action was to change these results to a ‘fail’ to eliminate the spreadsheet error. This was unnecessary, as the database validation routine was working correctly, and would have accepted these results as passing tests, had the Excel template been uploaded. The Inspectorate will investigate options to rectify this incorrect data with local authorities affected, during 2025. Due to this data error, colony count test data and information have been removed from charts and figures in this report.
At an aggregate scale, it is clear that private water supplies are not being tested at the frequencies required by the Regulations. It is vitally important for the regulatory testing regime for a supply to be determined and then met each year, or every five years, as appropriate. This validation monitoring provides a baseline to measure the overall quality of those supplies and the effectiveness of mitigation measures. Other sampling which could be called ‘investigatory’, can be recorded as such on the annual data return. Figure 2.2 illustrates the issue of incorrect testing frequencies for one parameter, E. coli. At regulation 9 supplies, E. coli should always be tested at the Group A frequency stipulated in the schedules to the Regulations. Samples taken as group A must be recorded in the data return as such.
In 2024, 67% of regulation 9 supplies were tested at the correct frequency for E. coli according to the volume supplied which is a slight improvement on 64.7% in 2023, with a further 9% of supplies sampled for E. coli at a greater frequency than the Group A parameter frequency required by the Regulations. However, the remaining 24% had fewer tests than required by the Regulations, 2% being under-sampled and 22% not being sampled at all. In 2023, 19.2% of regulation 9 supplies were not sampled at all for E. coli, showing a reduction in regulation 9 supplies testing in 2024.
Figure 2.2 Regulation 9 supplies tested for E. coli at Group A frequency
Third party risk assessments and sampling
The Inspectorate attends various regular meetings and events each year. A common point of discussion at these meetings is risk assessments and test data from third parties. The Inspectorate can confirm that it does not receive data or any other information on private water supplies from third parties. The only data submission process in place is the one managed directly with local authorities.
If local authorities are using third parties to conduct private water supplies sampling and testing, the requirement of schedule 4 part 1(5) of the Regulations for an agreement (contract) to be in place must be met. This contract between the local authority and the provider must ensure that the requirements of the Regulations are met, and that any breach of those requirements is reported to the local authority within 28 days.
While there is no similar direct provision in the Regulations for local authorities to use third parties to conduct risk assessments, the very nature of conducting a holistic risk assessment means drawing on the information and knowledge of others who are informed about the supply. Though, third parties may be major contributors to the risk assessment process, local authorities still carry the responsibility for the risk assessment overall, and are the only entity empowered to act in response to the risk assessment’s conclusions.
Local authorities bear the sole obligation for reporting data and information to Welsh Ministers under the requirements of regulation 14 and schedule 4 of the Regulations, therefore the Inspectorate will not accept data or information from third parties, where no arrangement is in place.
Regulation 8 supplies would be the most common supply type that may involve a third party. Some of these onward supplies serve a large number of consumers, with the third-party distributors often being large organisations or companies. The Inspectorate has been involved in informal discussions when local authorities have cited that these third parties either refuse to share their test data or assure the local authority that test data is being shared with the Inspectorate. As the Inspectorate does not receive any data or information from third parties, this indicates a potentially large data gap which is likely due to missing third-party data.
Table 1 A summary of the percentage of samples failing for various microbiological and chemical parameters
Microbiological failures
The detection of specific indicator micro-organisms means that a supply is contaminated. When E. coli, Enterococci and to a lesser extent Clostridium perfringens are found, this suggests that the contamination may be faecal in origin. Faeces often carry micro-organisms including bacteria, viruses and parasites which are harmful to health and when a faecal indicator is found, this water should not be consumed. Table 1 shows that during 2024, in Wales, one in 11 private water supplies may be unfit for consumption and pose a risk to health containing E. coli. Whilst coliforms are not always a direct indicator of faecal contamination, they still indicate that there is a route for contamination to enter the supply and that contamination has not been removed by treatment. This was found in over one in eight tests. Protection of supplies from contamination is critical to protecting public health. Should a supply be found to contain the presence of faecal matter, the local authority is obligated to investigate in accordance with regulation 18 of the Regulations.
Other failures
Taste and odour failures could be caused by the quality of the source water or develop as water passes though the distribution system. How consumers describe the taste or odour can help identify the cause, with descriptions such as ‘earthy’ or ‘musty’ pointing to possible algal problems in the source water and ‘woody/pencil shavings’ suggesting the presence of black alkathene pipework, to use a couple of examples. In 2024, the percentage of tests failing the standards for taste and odour were 1.43% and 1.82% respectively.
Lead and nickel were detected above the standards in 2.80% and 5.80% of tests respectively. For lead this is a decrease compared with 3.93% in 2023, 3.8% in 2022, 3.3% in 2021 and 2.1% in 2020. Lead is a neurotoxin that particularly affects children and can cause health effects in adults including chronic kidney disease, raised blood pressure and cardiovascular disease. Where lead is detected above the standard, local authorities must serve a regulation 20 notice to secure actions to protect the health of the consumers. The most robust long-term solution is the removal of any lead pipework, lead-containing fittings, and lead solder. The World Health Organization has recently published their refreshed technical brief on lead which can be found on its website.
Nickel is largely detected because of nickel plated domestic fittings such as taps. Notably, the percentage of failures is on an increasing trend which is also seen in public supplies as fittings containing nickel appear to be becoming more prevalent in properties. Sometimes the source is not obvious and could be under-sink mixers and temperature regulators. Consumers should be advised to replace these fittings where failures of the standard occur.
Iron, manganese and aluminium can be seen to have failed standards in 4.03%, 6.4% and 4.16% of tests respectively. All these metals can be found naturally occurring in source waters. Local authorities should consult with Public Health Wales to determine whether the concentration of these metals pose an immediate danger to health, and if so, serve a regulation 20 notice. Should the presence of these metals not be deemed a potential danger to health, local authorities can still act under section 80 to compel the relevant persons to make the supply wholesome and acceptable for all domestic purposes which includes washing and laundry.
Turbidity and colour are seen to have been found in concentrations above the standards in 2.19% and 2.29% of tests respectively. Turbidity can reduce the effectiveness of disinfection so a detection of turbidity in excess of the standard should trigger an investigation to determine the cause, followed by the appropriate course of action under regulation 20 or section 80 of the Water Industry Act 1991 (the Act) to carry out improvement actions. Colour detections are usually caused by compounds which arise from the catchment of the source waters and can be removed by suitable treatment processes. A short-term increase in colour and/or turbidity can be a result of high rainfall mobilising humic substances in source waters. If these short-term occurrences become more frequent, this may be a result of long-term changes in weather patterns.
PFAS
The data return suggests that seven private supplies have had PFAS included in their regulation 6 risk assessment. This is an increase from three last year. Natural Resources Wales (NRW) has been writing to the owners and/or users of private supplies, when those sources, or nearby sources have been tested, and PFAS found. In parallel it writes to the appropriate local authority sharing those test results. If a local authority receives one of these communications, the relevant private water supply should be risk assessed and as appropriate, sampled and tested for PFAS. If the supply has any PFAS present in a concentration over 0.01 µg/L (micrograms per litre), the local authority should refer to the actions listed in the Inspectorate’s IL 3/2025 as a guide. If the supply is used for food production and PFAS levels are above 0.01 µg/L, the Food Standards Agency should be notified. The Inspectorate encourages local authorities to submit any PFAS data in annual data submissions. It is likely that most PFAS samples will be regarded as ‘investigatory’ samples. No test data was received from local authorities in Wales for 2024.
Interpreting test results
Water sample results provide insight into the quality of water at the time the sample was taken and can act as an indication that further investigation is needed.
Local authorities submit sample results to the Inspectorate annually. A small subset of results provided each year exceed the regulatory limits by a significant magnitude. Not all of these figures have a commentary provided with them in the data return, which would be an opportunity to explain exceptional circumstances.
One result over the limit of nitrite were reported at 10 times the limit at the consumer tap, which is 0.5 mg/L.
Nitrite is a health-based parameter as babies and younger children are susceptible to develop methemoglobinemia at concentrations of nitrite above the regulatory limit. This is because nitrite is an oxidant and so oxidises iron in haemoglobin (in blood) to methaemoglobin. This means that there is not enough iron in haemoglobin to transport oxygen in the body, resulting in symptoms such as a blue discolouration to the skin (known as cyanosis) and in severe cases, it can result in death.
Five results over 50 µg/L were reported for lead. There is no safe level of lead with the best solution to the presence of lead pipework being its removal. Lead is also present in brass fittings to varying levels. An investigation should be undertaken by the local authority for all lead results above the regulatory limit and should be considered at values below that.
Three results over 10 NTU were provided for turbidity. Turbidity indicates how much the light reflects within a sample of water. The higher the turbidity, the more difficult it is to fully disinfect the water. This includes with ultraviolet (UV) light. For supplies with high turbidity the risk assessment should consider the effectiveness of disinfection and the risk of disinfection byproducts. One pH result above the pH scale range (1-14) was also provided, which is clearly a typographical error as a result over 14 is not scientifically possible. Local authorities should ensure that data is sense-checked prior to being submitted to avoid situations like this.
Risk assessments
In 2024, the percentage of private supplies which have an in-date risk assessment are:
• Private distribution systems (regulation 8 supplies): 47.8%.
• Large, commercial and public use (regulation 9 supplies): 57.3%.
• Small supplies and those as part of a domestic tenancy (regulation 11 supplies): 31.8%
In total, 2,841 supplies require a risk assessment, and only 1,280 (45.1%) have one that has not expired. For all supply types other than single dwelling supplies, local authorities have been required to complete risk assessments since 2010. It is concerning that, almost 15 years after the introduction of this requirement, there is a significant proportion (42.1%) where the risk assessment has exceeded the requirement for a five yearly review and a critical 12.9% which have never had a risk assessment.
In addition, supplies to untenanted single dwellings (regulation 10 supplies) are only risk assessed upon the owner’s request and 11.1% have in-date risk assessments. Local authorities are reminded that once a supply to a single dwelling has been risk assessed the risk assessment must then be reviewed every five years.
In 2017, a change in the Regulations brought about the requirement for local authorities to provide a summary of the results of risk assessments to Ministers (in practice the Inspectorate), within 12 months of having carried out the assessment. During 2024, the Inspectorate received 53 risk assessment summaries from six local authorities in Wales, this is less than half of the total number of summaries received during 2023. 17 out of 22 local authorities in Wales that submitted data on private supplies did not meet the statutory reporting requirements regarding risk assessments in 2024. The Inspectorate would encourage local authorities to programme an annual exercise to submit the risk assessment summary pages to the Inspectorate, perhaps to coincide with the annual data return submission.
A review of the submitted risk assessment summaries suggests that some local authorities are only conducting reactive risk assessments following sample failures. Risk assessments and associated mitigation measures should be verified by sampling, and not the other way round. Data including sample results can be used to inform a risk-based prioritisation of risk assessments and it may be prudent to review risk assessments following sample failures. However, the risk assessment process must be a proactive one and not a process where risk assessments are only conducted where sample failures have occurred.
Protection of the source and the abstraction point were common root causes of risk identified as part of local authority risk assessments. The need for remedial actions to protect the source of a private supply from risk of contamination from livestock, slurry, pesticides, and septic tanks featured repeatedly on the summaries submitted to the Inspectorate. Similarly recurrent were actions required to secure spring chambers, headworks, and buildings from drainage issues, unauthorised access and vermin.
Ultraviolet (UV) systems are commonly used for disinfection in private water supply systems. The efficiency of UV systems used for disinfection are affected by the water quality and the flow rate through the system. High turbidity can shield pathogens from the UV dose and an improper unit for the required flow rate may result in an insufficient UV dose applied. Local authority assessors noted risks from a lack of maintenance, resilience and control in some systems. Common mitigations required the need for intensity monitors, alarms and shutdowns in case of unit/bulb failures, power interruptions, or high turbidity in the influent water which may inhibit UV disinfection. Such controls would reduce the risk of partially or non-disinfected water being available for consumption.
The absence of regular inspection, cleaning, and maintenance of water storage tanks was frequently flagged as a risk in the risk assessment summaries. For comparison, treated water tanks (storage reservoirs) are an integral part of public water supply networks. Water companies conduct external inspections annually to identify risks from ingress and contamination. Furthermore, water companies must conduct internal inspection, cleaning, and remedial maintenance at a frequency dependent on the risk and the quality of the influent water. Assets such as water tanks are not in stasis and require regular intervention to maintain operability.
Enforcement
Local authorities are required to send notices served under the legislation to Welsh Ministers (in practice the Inspectorate) in accordance with regulation 20 where supplies are a potential danger to human health. They may also serve notices under section 80 of The Water Industry Act 1991 where supplies are unwholesome and/or insufficient.
In 2024, the Inspectorate received 103 notices served under regulation 20 of the Regulations for supplies that were considered a potential danger to human health. The data return indicated that 145 were served, the number now not being sent to the Inspectorate is reducing. Two notices were served under section 80 of the Act according to the data return. Both were received by the Inspectorate.
Figure 2.3 Regulation 20 notices served versus notices submitted to the Welsh Ministers
Reasons for serving notices
Most notices are historically served in response to a failure of a microbiological standard, with a very small minority for failures of other standards. In 2024, 94 out of the 103 notices submitted were served for failures of microbiological standards.
Eight were served in response to failures of the standard for lead. One notice was associated with a supply that was deemed a ‘risk to health’ after completing a risk assessment.
The tenet of the Regulations is one of proactive risk assessment to prevent failures and a risk to health from ever being realised. Local authorities still appear to be predominantly reacting to sample results, rather than proactively eliminating the problems that would result in a sample failure. The use of a proactive risk assessment, and if necessary, a notice in this context, is to protect users before an incident occurs so they are not unsuspecting, and they can be responsible for protecting their own health. Table 2 shows the number and driver for 103 notices returned to the Inspectorate.
Table 2 Notices served
Reason for serving the notice | Number |
Bacteria | 94 |
Lead | 8 |
Risk assessment | 1 |
Total | 103 |
Notices served for microbiological drivers, numbered almost 100. The majority of these required treatment to either be installed or better maintained. However, 15 of notices provided short term measures only, such as boil water advice. It is likely that further measures would be needed in these supplies to ensure long term compliance with the Regulations.
Of the two section 80 notices served in 2024, one related to bacteria and the other related to multiple parameters of metals and pH.
For the notice relating to bacteria, the notice document indicated that positive samples for enterococci and E. coli had been recorded in the supply. Enterococci and E. coli are faecal indicators. Local authorities should ensure correct liaison with health authorities when determining what constitutes a risk to human health to make sure that the correct notice is issued.
Microbiological parameters | ||
|
2023 (%) |
2024 (%) | |
E. coli | 7.12 | 8.97 |
Coliform bacteria | 13.62 | 12.52 |
Enterococci | 7.73 | 7.39 |
Clostridium perfringens | 4.98 | 5.25 |
Chemical parameters | ||
Odour | 0.52 | 1.82 |
Taste | 3.92 | 1.43 |
Manganese | 6.65 | 6.4 |
Iron | 3.19 | 4.03 |
Aluminium | 1.68 | 1.55 |
Turbidity | 1.51 | 2.30 |
Colour | 1.86 | 2.29 |
Lead | 3.93 | 2.80 |
Nickel | 5.16 | 5.80 |
Pesticides | 0.00 | 0.00 |
Fluoride | 0.0 | 0.00 |
Others | 2.87 | 3.78 |
Nitrate | 0.95 | 0.86 |
Nitrite | 0.00 | 0.42 |