Investigations

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

Report on drinking water quality incident at:
Oykel Bridge Water Treatment Works in June 2007

Scottish Water Executive Summary

Oykel Bridge water treatment works is a small borehole works which supplies 6 houses and produces an average of 1.2 m3 of treated water a day.

Water is abstracted from two deep boreholes adjacent to the river Einig. The water is then dosed with sodium hypochlorite prior to filtration through four sand filters for the removal of naturally occurring manganese. The sodium hypochlorite is dosed flow proportionally and residually trimmed for final disinfection purposes, but also to aid the removal of manganese on the filters. There is both telemetry and power to the site.

On the morning of Monday 11th June 2007 the works operator carried out a routine visit to the works and found nothing untoward other than the chlorine analysers which were drifting out of calibration slightly and this was reported to the Scottish Water Instrumentation and Calibration Team. Shortly after the operator left site the mains power was lost for approximately 50 minutes. The works is designed to automatically come back on line after a power cut but on the restoration of power to the site the inlet flow meter and raw water turbidity meter circuit breakers tripped. As the sodium hypochlorite dosing is flow proportionally controlled, when the works started up on low clear water tank (CWT) level the dosing did not start as there was no signal from the flow meter. Low treated and final (supply) water chlorine alarms were generated but alarm response errors resulted in appropriate corrective action not being taken.

The failure was not discovered until Friday 15th June when a routine site visit was carried out. The works was reset, to restart the sodium hypochlorite dosing and the treated water storage tanks were dosed with crushed chlorine tablet to achieve a residual of 0.82mg/l. Bacteriological samples were taken after corrective action had been taken and these were reported as having passed the next day.

Undisinfected water entered into supply for in excess of three and a half days as a result of the loss of signal from the flow meter to the sodium hypochlorite dosing system and alarm handling errors.

Actions being undertaken as a result of this incident include:

• Formal Fact Finding interviews with Customer Operational staff involved with the Oykel Bridge WTW incident.
• Modifications to be done on the chlorine dosing system and telemetry configuration to prevent misleading alarms being generated when the works is shut down.
• An assessment into the feasibility of reconfiguring the works to automatically shut down in the event of a disinfection failure.

DWQR Assessment on the incident

The DWQR's assessment of this disinfection failure incident is that it was caused by a power failure on Monday 11 June 2007 which caused the circuit breaker for the inlet flow meter to trip. When the power came back on the flow restarted but the trip for the inlet flow meter had not been re-set, so the disinfection pump did not detect any flow so did not start to pump. Thus undisinfected water was going into supply. The operator had been out earlier in the day to investigate an earlier power fail and had carried out various tasks to restore the works. However, a second power fail occurred after the operator had left the site and although a low chlorine alarm alerted Scottish Water's (SW) Operational Management Centre (OMC), the OMC wrongly assumed that the operator was still on site so did not contact him again to investigate this second power fail. A further low chlorine alarm later in the day was generated and passed to the operator, but no action appears to have been taken. Another low chlorine alarm very early the next morning was passed to the operator who took no action assuming that it was another faulty reading as there had been a few in the weeks leading up to the incident. A further low chlorine alarm about an hour later was passed out to the operator who cannot recall receiving it. Since the site is only visited twice a week, all of this went without investigation until the next routine visit on the Friday 15 June 2007, so undisinfected water was going into supply for almost 4 days. Furthermore it was not until the situation had been resolved by mid afternoon on the Friday that SW's Public Health Team was informed and appropriate samples taken. However, by this time, the works was running properly again and the samples all passed. However, even this task was not uneventful as it was delayed due to sample bottles not being held on site and the operator having to go to another site to pick them up.

DWQR notes the actions being taken by SW, in particular the interviews with staff about the alarm handling errors, reconfiguring the telemetry to prevent all treated water chlorine alarms when the works is shut down, looking at the feasibility of reconfiguring the works to automatically shut down in the event of a disinfection system failure or flow meter/signal failure and investigating the extent of the problem with power fails.