Readers of the DWI's annual reports have tired of hearing quite how close water companies have come to meeting 100% compliance with quality standards. In 1992, the figure was 98.65% for England and Wales; it rose to 99.88% in 2003 (ENDS Report 354, p 19 ). In the latest report for 2004 everything changes.
Another change is that the reports for England and Wales are now separate - which makes for complications because national boundaries do not coincide with water company supply areas.
Water quality is now reported by six geographical regions rather than by water company.
Launching the report, chief inspector Professor Jeni Colbourne explained: "We have got to the position where failures only reflect local circumstances. The report needs to be in context of those consumers affected."
The DWI wrote to 379 local authorities in January seeking their requirements for information on drinking water quality. Local authorities have a statutory responsibility to keep informed of the adequacy and quality of drinking water supplies.
They replied that they wanted local results and to hear about consumer complaints, compliance with new parameters and the outcome of water quality incidents.
The focus on local reporting should bring the Inspectorate closer to the public.
"Our report shows that the consumers' experience of tap water quality does indeed vary across the country," Professor Colbourne said. "Consumer expectations are higher than ever and those who intermittently receive discoloured or bad tasting water rightly want something done about it."
The solution, the Inspectorate believes, is smaller reports targeted at the regions where local quality issues can be discussed and details of improvement programmes provided.
Maurice Terry, chair of the consumer watchdog WaterVoice, congratulated the DWI: "You have put consumers at the heart of the report and highlighted areas of concern. Companies may say they don't like it but it is supporting them in helping to eliminate small but significant incidents."
However, the new format leaves the report short on national compliance data. One exception is a new indicator that the Inspectorate has devised known as mean zonal compliance (MZC). This measures compliance only in water supply zones, not the results of monitoring at water works and reservoirs.
Overall MZC in 2004 was 99.94%, but there are no figures from previous years for comparison. Individual company performance varied from 99.76% to 100%, although the larger companies lie well within this range.
The calculation of MZC is statistically complex because risk-based monitoring means that larger zones are sampled more frequently. Pesticide monitoring requirements also vary depending on the water source.
The lowest scores were for nickel at 99.77%, iron at 99.60% and lead at 99.53%.
Nickel, the chief inspector revealed, is a problem which is local to East Anglia. There were 37 failures, of which 31 were in two supply zones in Suffolk and Norfolk. The maximum level recorded was 37.5 micrograms per litre, compared with the 20µg/l limit. The exceedance is of little health significance in the short term and Anglian Water is installing pipelines to allow blending with other sources.
Another national comparator is an operational performance indicator based on three key parameters linked to discoloured water problems: turbidity, iron and manganese. The measure, known as OPI(TIM), ranks companies' compliance on the three parameters.
Those companies which still have substantial work to do to improve old water mains have the lowest levels of compliance on OPI(TIM). The national mean is 99.79% - but Welsh, South West, United Utilities are below average
OPI(TIM) has been calculated for previous years and the figures reveal an improving trend from 99.45% in 1997 to 99.79% in 2003. However, there was no improvement in 2004.
The DWI notes that many companies have further investment planned before 2010 in replacing or relining water mains and it expects further improvement as a result.
The only other enforcement action was a formal caution for Three Valleys Water in October. This followed the disruption of water supply to several villages in Hertfordshire after a water tower ran dry. Many residents received discoloured water.
The report reveals that two prosecutions for supplying unfit water are under consideration in the Southern region, one in Thames region and four in the Northern region.
For most companies the norm is one or two calls per 1,000 customers. However, South West Water's figures were "extremely high" at 14 per 1,000. United Utilities, Northumbrian and Yorkshire Water were at five per 1,000, largely due to discoloration complaints.
While the problem of rusty brown water is well known, the report draws attention to white water - a particular issue in the Thames and Anglian regions. Although often due to entrained air bubbles, it was the cause of more than a third of Thames Water's consumer contacts, and over a quarter of Anglian's.
The report notes: "The relatively high resource effort focused on United Utilities and South West Water reflects other factors such as consumer complaints, water quality incidents and responding to standards not being met."
For most companies, the number of zones failing the standard were in single figures. However, Thames Water found 64 failures and United Utilities 24.
The large number of Thames failures is due to the company's larger sampling programme, the DWI considers. In only two cases was there found to be a company-owned lead connection pipe to the affected property. In other instances, the failure appeared to be due to the customer's lead pipe, or more rarely lead solder.
Nine of United Utilities' failures may have been linked to company-owned lead pipes. However, the company also has 15 outstanding schemes to install or improve plumbosolvency treatment and one outstanding scheme to remove naturally occurring lead.
For the first time, the DWI acknowledges that the illegal use of lead-based solders on domestic copper pipes is a problem. In the past, only Scotland has highlighted this as a significant cause of failure (ENDS Report 346, p 16 ).