Investigations

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

Report on drinking water quality incident at:
Savalbeg Water Treatment Works in July 2007

Scottish Water Executive Summary

Lairg Water Treatment Works (WTW) is a chemical coagulation works with primary and secondary Dynasand filtration. A Dynasand is a continuous filter, in which the filtering flows upwards counter current to the downward movement of the filtration media (sand), the downward movement is induced by a central air-lift which transports the dirtiest sand to the top of the filter vessel and washes it on the way. Lairg WTW is designed to operate with a maximum flow of 0.72 mega litres per day (MLD) supplying a population of approximately 844.

The raw water gravitates to the works from a loch source and is dosed with alum and soda ash before feeding into the first Dynasand filter. The primary filtered water is then pH corrected again with soda ash, for optimum aluminium removal before, feeding into the second Dynasand filter. The secondary filtered water is then dosed with soda ash and sodium hypochlorite for final pH correction and disinfection. There is on-line monitoring for pH throughout the process and treated water monitoring for aluminium, turbidity and free residual chlorine. All treated water instruments are on telemetry and a high priority alarm will be generated if the water quality is not within the required specification. The treated water feeds on to a clear water tank (CWT) with approximately 11 hours storage. The WTW final sample point is as the water leaves this CWT and enters the distribution network.

On Sun 01/07/07 a alarm was generated for low treated water chlorine at 21:30. However due to a fault with the Remote Telemetry Unit (RTU) the alarms did not appear on the screen at the OMC Flight desk until polled at midnight (00:00). The flight desk acknowledged the alarm at 00:09 but did not pass it to the standby operator until 01:42, who on arrival at site found the suction pipe on the inlet of the chlorine dosing pump blocked with crystallised chlorine. The Operator cleared the blockage in the pipe using hot water and clean wire. The pump was then rebuilt and put back into service and normal disinfection resumed by 03:05.

Whilst no chlorine was dosed into the works for approximately 4 hours, manual sampling of the at the CWT outlet indicated that the chorine residual in the Clear Water Tank (CWT) did not drop below normal operating levels and no disinfection fails were recorded through statutory samples in the network.

The route cause of the incident was due to a blocked chlorine suction hose, which would have been repaired sooner had the RTU not failed and the flight desk staff had passed forward the alarm within allocated 15min. Investigations show that the RTU restarted after polling at midnight and has caused no further problems since.

Delays in informing the DWQR of this event was due to internal miss communications between different departments within Scottish Water.

Recommendations as a result of this incident include:

1. Review internal communications between Ness Operational staff/ process scientists and PHT.
2. Formal fact finding interview with Operations staff involved.
3. OMC Flight desk staff to comply with alarm protocols.
4. Investigate why the RTU failed to pass forward alarms on that occasion.

DWQR Assessment on the incident

The DWQR assessment of this incident is that it was caused by crystallisation of the chlorine in the suction line to the chlorine pump. A low chlorine alarm was generated at 21:30 but this was not passed on to Scottish Water's (SW) alarm monitoring centre until midnight (00:00), some two and half hours later due to a fault with the remote telemetry unit at the works. The alarm monitoring centre acknowledged the alarm at 00:09, but did not pass this on to the local operator until 01.42 which caused another delay of an hour and a half. There was thus a total delay in response to the alarm of about 4 hours. The operator arrived on site about an hour after the alarm had been dispatched to him which is not unreasonable given the time of day and the travel time to the works. Once at the works, the operator took appropriate remedial action to resolve the problem but failed to inform his team leader and SW's Public Health Team. SW states that no formal samples were taken as the chlorine residual at the outlet of the clear water tank was always within target levels and did not breach any Emergency Action Level parameter. However, there is no evidence in the report to support this view. Since no samples were taken it is not possible to say if public health was compromised or not.

DWQR notes the actions being taken by SW which includes:-

1. Holding an incident report workshop to clarify roles, responsibilities and procedures between functions.
2. Formal clear written communication to be given to all alarm handing staff about the handling of alarms, especially those relating to disinfection failures. DWQR is concerned that such a procedure is not already in place. DWQR notes that, in this case, the low chlorine alarm was a priority 2 but is aware of ongoing work to rationalise all low chlorine alarms as priority 1. DWQR has therefore asked SW to ensure that re-prioritisation of the Savalbeg WTW alarms is captured as part of the ongoing work.
3. Formal fact finding interview with the relevant operator to investigate the apparent non-conformance with SW's policies and procedures.