Investigations

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

Report on drinking water quality incident at:
Boardhouse Water Treatment Works in April 2006

Scottish Executive Summary

Boardhouse WTW is located on the north-west of mainland Orkney, supplying 4853 customers with 5ML of water per day. Treatment consists of coagulation with aluminium sulphate, clarifiers and RGF filtration followed by disinfection. The final water is pumped up to Ravie Hill SR to gravitate into distribution. This site is not currently connected to Scottish Water s telemetry system.

On 30th March at 22.20 hrs the OSEC electro chlorination unit which converts sodium chloride salt into a chlorine solution for disinfection suffered a cell pump failure. There is no automatic change to the standby dosing system and the works shut down on an electro chlorination unit failure. Upon works shut down an auto dialler called the duty treatment operative who went to site. The works was then manually switched to the standby disinfection dosing system which consists of 2 in number Prominent type pumps, dosing 15% Sodium hypochlorite. The Prominent pumps are manually operated with no automatic change over between them. The plant was left running on the standby dosing system whilst awaiting repair of the electro chlorination unit.

On 1st April at 11.45 hrs standby Prominent pump number 1 failed to deliver adequate chemical. The chlorine residual dropped below the Low Low residual set point of 0.5mg/l. This shut the plant down and the auto dialler called out the treatment operative. Maintenance was carried out to this pump and to the dosing line. It was then returned to service. The operative left the site at 14.00 hrs after raising the dosed chlorine residual to 1.5mg/l.

On 3rd April at 16.10 Prominent pump number 1 again failed to deliver adequate chemical. The chlorine residual dropped below the Low Low residual set point of 0.5mg/l. This again shut the plant down and the auto dialler called the treatment operative, who was still onsite at the time. Standby pump number 2 was then put into service at 17.00 hrs and the operator left site at 18.10 hrs having raised the dosed residual to 1.35mg/l. At 18.45 Prominent pump number 2 also failed to deliver adequate chemical and the disinfection dropped below the 0.5mg/l Low Low set point. On this occasion the plant did not shut down as the alarm had not been reset following the earlier shutdown. This meant that the auto dialler did not call the treatment operative.

On April 4th at 08.00 hrs the treatment operative discovered the chlorine residual at zero. Prominent pump number 2 was running but delivering no chemical. The operator restored disinfection using the electro chlorination dosing pump and the chlorine solution held in the tank of the electro chlorination unit.

Team leader was informed, and the operator was dispatched to check the chlorine residual at Ravie Hill SR. This was found to be 0.04mg/l free chlorine. Process optimisation and the public health team were informed as part of the incident plan. Chlorine levels were elevated at the works and samples were taken of the WTW final water and from Quholm SR and Stromness SR in distribution for bacteriological analysis. A new electro chlorination unit cell pump was delivered to site and repairs were affected to all pumps.

Chlorine residuals remained elevated from the works until residual leaving Ravie Hill SR reached its set point of 0.5 to 0.6 mg/l Total Chlorine.

All samples passed with 0 total coliforms and 0 E.coli.

Work is now being carried out to the dosing system to provide an automatic duty/standby back up system. A new telemetry outstation is being sourced to provide a full suite of independent telemetry alarms through the new Open Enterprise telemetry system.

DWQR Assessment on the incident

DWQR's assessment of this incident is that the works shut down on three occasions (30 March, 1 April and 3 April 2006) due to problems with the chlorination pumps and low chlorine. The problem was compounded on the third occasion since the alarm had not been re-set following the second occasion. Furthermore, although Scottish Water's Public Health Team were informed on 4 April after the third occasion on 3 April and samples taken from the works and two service reservoirs, this was too late to take any meaningful samples. The Public Health Team should have been informed at the very latest as soon as possible after the second shut down was investigated by the operator at 11.45 on 1 April. Meaningful samples could then have been taken at the works and in distribution. DWQR has asked Scottish Water to look into this failure to communicate within its own organisation to ensure that the risk of this happening again is reduced as far as possible.

DWQR is however pleased to see that work is being carried out to upgrade the disinfection system from manual to auto duty / standby change over by 31 May 2006 and that the treatment works will be connected to Scottish Water's main telemetry system by 12 October 2006. Until then, Scottish Water will rely on the existing system at the works which should call the operator out if an alarm reaches its trigger level.