Cryptosporidium is a minute protozoan pathogen which causes a severe gastrointestinal illness that can be fatal in immuno-compromised patients. The organism is resistant to chlorination, and preventing its spread in drinking water supplies requires efficient filtration at all times.
The new proposals are an attempt to plug a gap in the law which has prevented the prosecution of water companies alleged to have caused outbreaks of the disease through inadequate water treatment.
The legal loophole became apparent last autumn, when South West Water was acquitted of supplying water unfit for human consumption in the Torbay area in 1995 (ENDS Reports 272, pp 9-10 , and 275, pp 20-23 ). The judge ruled that epidemiological evidence linking the outbreak to the supply was inadmissible on a legal technicality. A case against Yorkshire Water was subsequently abandoned for the same reason (ENDS Report 276, pp 35-36 ).
The DETR has responded by proposing that water treatment works abstracting from rivers or from reservoirs with less than seven days storage should be required to monitor continuously for cryptosporidium. Companies would be required to pass a sample of at least 1,000 litres of treated water per day through a filter and analyse it for oocysts. A new legal limit of less than one oocyst in 10 litres of water on average would be set. Exceeding this would be a criminal offence.
Analytical results would have to be available within three days and placed on the public register. However, where a "significant increase" in turbidity - suggesting an increased risk of cryptosporidium breakthrough - was detected, analyses would have to be completed within a day. This is likely to be close to the practical limit for many companies - particularly if laboratory facilities are some distance from the water works.
The DETR believes that the proposals will increase confidence that water treatment is being operated properly. The extra monitoring will certainly provide new scope for investigating links between disease outbreaks and drinking water supplies. But it may also heighten public awareness of cryptosporidium risks.
The numerical limit for cryptosporidium is likely to prove controversial. It cannot be considered to be protective of human health since the infective dose of the organism is unknown. Its distribution in the water supply system is also uncertain because oocysts tend to stick together - increasing the chances of one individual obtaining a large and potentially infective dose at low concentrations.
A DETR spokesman described the limit as a "treatment limit" - presumably intended to be indicative of whether supplies had received adequate treatment. However, monitoring for cryptosporidium is a highly inefficient process. Recovery and detection of oocysts from filters may be at best 50% efficient and commonly as low as 20% (ENDS Report 237, pp 18-21 ). Levels of cryptosporidium may therefore be up to five times higher than monitoring suggests.
The unreliability of existing detection methods was one reason why routine monitoring was not suggested by two expert committees in 1990 and 1995 (ENDS Report 249, p 10 ). Costs were another reason. The consultation paper estimates total recurrent costs at £7.9 million per year. Welsh, South West and Northumbrian Water face the highest costs, at £1.8, £1.3 and £1 million, respectively.