Investigations

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

Report on drinking water quality incident at:
Gairloch Water Treatment Works in October 2006

Scottish Water Executive Summary

The water treatment works at Gairloch consists of simple disinfection only. Water from the loch source gravitates down to the WTW where it is disinfected prior to the clear water tank by dosing sodium hypochlorite which is flow proportionally controlled. There is an on-line chlorine instrument that monitors the treated water free chlorine residual before it enters the clear water tank and there is telemetry at this site.

A problem with achieving adequate flow proportional chlorine control was first reported at Gairloch WTW on the 16th September 2006. Despite a number of attempts to resolve the problem the disinfection failed completely at approximately 18:30 hours on Sunday 1st October. A low chlorine alarm was received via the telemetry and this was reported to the Network Service Operator (NSO) on stand-by at 20:00 hours. The NSO had checked the works via telemetry earlier in the day (following high chlorine alarms being reported) and had also been previously informed that there were problems with the chlorine control at the works. Consequently, he assumed that the low chlorine alarm was another spike and that it would soon return to normal. To avoid further nuisance calls he asked for the alarm to be suppressed until 08:00 hours the next day. At 08:01 hours on Monday 2nd October the low chlorine alarm was acknowledged (as the alarm suppression had been lifted), but it was not called out again to the Customer Operations team as it had already been reported. From the telemetry record of the treated water chlorine levels it can be seen that the sodium hypochlorite dosing pump had in fact failed at approximately 18:30 hours on 1st October.

The disinfection failure was not discovered until a routine site visit by the senior operator on Tuesday 3rd October at 13:00 hours. The situation was immediately reported by the senior operator to the Public Health Team and emergency procedures were carried out. On investigation it was discovered that the sodium hypochlorite dosing pump was not delivering a flow, so it was replaced by the operator and the dosing was reinstated. The CWT was shock dosed to achieve a residual of 1mg/l at the tank outlet and the chlorinated water was flushed through the network.

A bacteriological sample was taken at the works and two were taken from distribution once an adequate chlorine residual had been achieved and these were reported as having passed the following day.

The root cause of the disinfection failure at Gairloch WTW on the 1st October was the sodium hypochlorite dosing pump, although appearing to be active was not actually dosing any chemical. The length of time that undisinfected water entered into supply was prolonged by the Network Service Operator (NSO), on stand-by for the area, believing the alarms to be a continuation of a series of spikes, requesting for the chlorine alarms from the site to be suppressed.

The main actions as a result of this incident are:

• The Network Service Operator on stand-by during this incident will be formally interviewed about his actions.
• Instruction to be given to all stand-by staff that all P2 alarms must be responded to immediately.
• Upgrade the manual duty/stand-by sodium hypochlorite dosing pumps to auto change over.
• Carry out a full risk assessment of the disinfection system. Report on any actions required and ensure that measures are taken to minimise the risk of further disinfection failures.
• Improve communication between network and treatment teams and improve treatment knowledge of local network staff.

DWQR Assessment on the incident

There had been ongoing problems with the disinfection system at this works for at least a couple of weeks and most of these had been dealt with in a timely manner. However, a low chlorine alarm late in the afternoon of Sunday 1 October 2006 was suppressed by the local Network Service Operator (NSO) and due to other communication fails, the failure of the disinfection dosing pump was only discovered through a routine visit to the works and fixed around lunch time on Tuesday 3 October 2006. Undisinfected water was therefore being supplied for almost 2 days. However, DWQR is content with the actions being taken by Scottish Water which includes formally interviewing the NSO for his inactions, improving communications between network staff and treatment staff and upgrading the manual duty/standby disinfection dosing pumps to auto change over.