Investigations

This section contains summaries of reports on drinking water
quality incidents and investigations.

Report on drinking water quality incident at:
Windyfield Rhynie Water Treatment Works in August 2007

Scottish Water Executive Summary

Windyfield Rhynie WTW (also known as Rhynie WTW) is located to the north of the Aberdeenshire village of Rhynie and supplies water to a population of approximately 400. The treatment works provides minimal treatment to a spring supply source in the form of chlorination by a sodium hypochlorite drip-feed system. The disinfected water passes through 2 storage tanks which supply the water to the distribution network.

On the 7th August 2007 a routine sample taken for Cryptosporidium provided a positive result, when 50 Oocysts were detected equating to a result of 0.371 Oocysts per 10 litres.

At the time of this result, improvement works required to deliver the Q&S3 drivers was underway. The construction works were being carried in a field to the immediate north of the disinfected water storage tanks.

As this was an unusual occurrence at this location, the subsequent investigation focussed on the ongoing construction work at the site at the request of the DWQR.

Investigations have highlighted that work to replace the catchment collection chamber lids may have caused surface water to collect around the chambers and allow ingress into the system.

In addition, the final water turbidity monitor was not functioning at the time of the event and this is to be urgently rectified.

As the treatment process at the site is simple disinfection it does not provide a barrier to cryptosporidium. Scottish Water is currently procuring a temporary filter to be installed at this site to address this in the short term.

DWQR Assessment on the incident

This incident occurred when a routine Cryptosporidium sample taken on 7 August 2007 was reported on 10 August as containing 0.371 oocysts per 10 litres. Samples from Rhynie WTW do not normally contain oocysts, therefore this was an exceptional result for the site. The source at Rhynie had been described to the Consultant in Public Health Medicine (CPHM) as a borehole, although in fact it is more closely aligned to a spring system. The Consultant in Public Health Medicine was concerned that the community had had no previous exposure to Cryptosporidium so they requested that Scottish Water issue advice to consumers in the area to boil their water. This boil notice was subsequently lifted on 16 August. Scottish Water's report into the incident is deficient in that it makes no mention of the imposition of the boil notice.

Investigation by Scottish Water showed that on the morning of 6 August, the chlorine residual in water leaving the treatment works fell away to zero. This triggered an alarm, and an operator attended the site to increase the chorine dose. The fall in chorine had been caused by an extremely sudden increase in water flow through the works.

Chlorine dosing at the site is currently not flow proportional, although at the time work was under way on site to provide a much improved dosing system. DWQR considers that the sudden loss of chlorine at the site should have triggered further investigation by operational staff, which could then have highlighted the sudden increase in flow through the works. This could only have been attributed to an ingress of surface water. The incident has highlighted that a turbidity monitor installed in 2006 at the site has never functioned correctly. This has not been rectified, despite being in clear breach of the Cryptosporidium (Scottish Water) Directions 2003. It is likely that the turbidity monitor trend would have shown a rise in turbidity associated with the surface water ingress and this could have prompted Scottish Water to take appropriate action at the time of the incident rather than four days later, once the Cryptosporidium sample results were known. Scottish Water have confirmed to DWQR that this turbidity monitor is now operational. DWQR has requested confirmation from Scottish Water that all final or treated water turbidity monitors are operating correctly at every site.

Scottish Water consider that the likeliest route by which surface water could have suddenly infiltrated the system was through the spring collection chambers. At the time work was ongoing to replace the chamber covers. Consequently, there was shuttering around the chambers and piles of excavation spoil in the vicinity, both of which could have acted to dam surface water and enable it to enter the collection system. There was exceptionally heavy rainfall on the night of 5/6 August. No water quality risk assessment was undertaken by Scottish Water prior to the commencement of the work, although Scottish Water say that they were monitoring the contractors closely and had been satisfied with the work undertaken. Although DWQR accepts the exceptional nature of the rainfall that occurred, it is possible that a formal risk / impact assessment process prior to commencing this type of work could have suggested an alternative approach which removed any impact the work could have on water quality. At the very least, it would enable Scottish Water to demonstrate that it had done its best to demonstrate due diligence. DWQR has requested that Scottish Water provides DWQR with a copy of its procedures for written water quality impact assessments where significant capital works are being carried out at a water treatment asset.

Scottish Water has identified three actions to be undertaken as a result of this incident. Work has been undertaken to ensure the turbidity monitor is operating correctly and the priority status of the low chlorine alarm has been reviewed and changed. Additionally, a temporary filtration system is under construction at the site to provide an effective barrier against Cryptosporidium. In the longer term, Scottish Water will be expected to work with DWQR in identifying the most appropriate solution for the site.

DWQR has requested a written response from Scottish Water on points raised concerning the functioning of turbidity monitors and the production of impact assessments by 30 November.