As part of water safety planning, annual submissions of raw (untreated) water data are made to the Inspectorate. Sampling focuses on identifying hazards, assessing their presence and severity. This data plays a crucial role in guiding catchment management efforts and shaping the design, operation, enhancement and optimisation of treatment processes.

For the past century, water quality stressors have been primarily anthropogenic, with increasing freshwater demand and the polluting consequences of the urban, industrial and agriculture sectors. With climate change driving the increase in frequency and severity of extreme events such as storms and floods, inland water quality is set to continue worsening. This will further strain waste and drinking water treatment works and threaten drinking water supplies, and thus, there is a need for a more comprehensive quantification of raw water quality, to understand past and current trends through data.

In 2024, the Inspectorate commissioned a research project in association with the University of Cambridge to analyse the raw water dataset held by it to see what insights could be generated. The dataset comprises:

  • Around 22.1 million test results.
    • 13.5 million are of groundwater samples.
    • 8.4 million are of surface water samples.
  • 764 parameters have been tested.
  • At 3,258 sites.

The project aims to deliver a tool which can initially be used for the interrogation of a small set of parameters to enable effective analysis with the potential to expand the number of parameters at a later stage.

A better understanding of the evolving distributions of different chemicals across water company catchments will facilitate data-driven prediction of contamination risks and potential for contaminated drinking water sources. Understanding how these contaminants behave through treatment works could enable the prevention of contamination events, enabling water suppliers to mitigate transient increases in chemical levels in the raw water. The Inspectorate has been collating raw water quality data collected by water companies since 2009 and is now extending its analysis of treated water compliance data to this growing set of raw water data. Analysis will support the Inspectorate in assessing the consequences of action or inaction in catchments and the relationship between the outputs of water safety plan risk assessments and the test results. It will also help identify hotspots and highly stressed areas that might begin to struggle to provide drinking water to regulatory standards in the future.

Poly and perfluoroalkyl substances (PFAS)

Given the increased identification of the risk of Poly- and perfluoroalkyl substances (PFAS) in UK and global waters in recent years, the Inspectorate has started reviewing this group of chemicals separately to the rest of the dataset. These are sometimes referred to as ‘forever chemicals’ and are byproducts of various industrial processes being practiced over the last 80 years. This project will facilitate a separate inspection of the PFAS subset.

The Environment Agency (EA) conducted the first extensive spatial analysis of 41 PFAS parameters in England from February to December 2023. They analysed 631 sites (275 ground water, 309 river and surface water, 47 coastal and estuarine river), and performed analysis per water source type. They identified a significant difference in PFAS detection in each raw water source, with river and surface waters having triple the detection frequency than groundwater across all measured parameters. This is likely due to the differing transport and fate dynamics, with attenuation processes in groundwater diffusing PFAS contamination. This study only pursued data collection for a spatial analysis of PFAS in England; temporal analysis has not yet been carried out for PFAS in England to date.

As of May 2025, the PFAS database has:

  • Around 1,789,000 test results.
    • 832,000 are raw water.
    • 957,000 are operational.
  • 48 parameters.
  • 2,275 sites.

The project continues into 2025 with the aim to complete the interactive tool focusing on the initial small number of parameters selected, with ideas around future work to expand this analysis to cover a more extensive range.