Investigations

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

Report on drinking water quality incident at:
Mannofield Water Treatment Works in July 2006

Scottish Water Executive Summary

On Sunday, 9th July 2006 at approximately 2100hrs, a low level of ferric sulphate (coagulant) in the duty service tank resulted in the loss of the coagulation process at Mannofield WTW. Although no telemetry alarms were set up for such an event, the dosed water pH meter triggered a high pH alarm at 2155hrs, indicating a problem with the coagulation process. However, due to a possible fault with the satellite communications system between Outstation A on the site and Scottish Water's telemetry system, the alarm wasn't received by the Operations Management Centre (OMC) until 1244hrs the following day. By this time, site Operators had already been notified of a problem at the treatment works via a low level total chlorine alarm relating to the final water. The receipt of this alarm, triggered at 0411 hrs on 10th July and relayed via Outstation B, was also delayed which is presumed to have been associated with the satellite communications system as well. This delay of over 2 hours resulted in the standby Operator being notified at 0631hrs. Upon attending site at 0645hrs, it was apparent that the coagulation process had failed although the low level chlorine alarm was initially attributed to an instrument reagent running out. Following the reinstatement of the coagulation process at 0800hrs, site checks and replacement of the chlorine reagent confirmed the low chlorine residual which continued to fall until 1000hrs. This was due to the increased chlorine demand from the uncoagulated (but filtered) water passing through the works. The decreasing chlorine residual prompted a significant increase in the actual chlorine dose which resulted in the residual increasing to 1.00mg/l by 1330hrs. There were no water quality failures detected as a result of the coagulation failure.

Although no telemetry alarms were in place for the ferric sulphate service tanks at the time of the incident, which is to be addressed, the available volume was recorded on the morning of 9th July as part of the work scheduling daily checks. In the absence of the standby Operator during the preparation of this report due to annual leave, it is presumed that the available volume was considered to be sufficient for a further day's operation. However, as the ferric sulphate level decreased during the day, increased sedimentation in the bottom of the tank prevented the coagulant from being dosed. Procedures in place to address any build up of sediment in the service tanks below a volume of 3.5m3 were not implemented as it is presumed that the Operator judged this not be required.

Manual changes in tank duty up to now have been based on Operator experience with no documented guidance available on site regarding minimum working levels. Therefore, clearly visible indicators will be affixed to the tanks and supported by documented procedures in response to this incident.

Although a presumed problem with the satellite communication system prevented critical alarms being relayed from both outstations at Mannofield WTW, had communication channels been available a number of additional early warning alarms from other monitors could have been expected. However, one such monitor for clarified water turbidity was reading in error before and during the incident, although it is unknown whether the associated Emergency Action Level (EAL) was breached. Alarms for the filtered water turbidities may also have been expected but the EAL wasn't breached until 0700hrs on 10th July when Operators were already on site. However, alarms were not triggered as the alarm set points in telemetry did not correlate with the EAL, which has since been addressed. This has also prompted verification of EALs against telemetry alarm set points at all water treatment sites in the North East. Investigations into the presumed satellite communications error are still ongoing which includes the monitoring of all key signals from both outstations at Mannofield WTW.

DWQR Assessment on the incident

The primary cause of the incident was the failure to account for the loss of ferric sulphate due to sediment build-up in the storage tank; secondary issues contributing to the incident were the failure of the satellite telemetry system and the failure to align the Emergency Action Level alarms with the telemetry alarm set points. While DWQR has accepted the corrective actions identified by SW along with the corresponding target completion dates, concern remains over the robustness of the telemetry system and its ability to correctly and timeously relay operational information in time for action to be taken to prevent future incidents occurring. Any future drinking water incidents involving a failure of the telemetry system at this site will be subject to a formal investigation by DWQR and, where appropriate, enforcement action will be taken.