Investigations

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

Report on drinking water quality incident at:
Londornoch Water Treatment Works in May 2007

Scottish Water Executive Summary

Londornoch WTW serves a population of approximately 2182 residing in the town of Dornoch and surrounding area.

Water is abstracted from two upland lochs and treated using coagulation with aluminium sulphate, dissolved air flotation (DAF) clarification and filtration in one unit known as a Flofilter system, secondary filtration through rapid gravity filters (RGFs) and final dosing of sodium hypochlorite, lime and monosodium orthophosphate for disinfection, pH correction and plumbosolvency control.

The aluminium sulphate dose is manually set and the optimum dose is established by carrying out raw water jar tests when changes in the raw water quality are observed. The works is currently being visited on a daily basis following recent compliance issues.

There is telemetry at this site which provides visibility of flows, levels, mechanical equipment status and water quality.

Late in the afternoon on 30th May, 2007 in response to a slight deterioration in filtered water quality, the site operator changed the coagulation pH operating condition reducing it from 6.0 to 5.5 and then 5.3 pH units. He then carried out a backwash procedure on the Flofilter in an effort to improve filtered water quality and believing the situation was stable left the site at 19:15 hours. A number if site alarms, low coagulation pH, high filtered water turbidity and high filtered water aluminium were received later that evening and on into the next morning.

The operator visited the site at 05:00 hours on the morning of 31st May and found the filtered water aluminium level to be 3.00 mg/l and the final water leaving the clear water tank was 540 µg/l (the PCV is 200µg/l). The WTW flows were diverted away from the CWT to waste until the water quality was recovered at 09:00 hours.

The root cause of the event was the action taken by the operator to dramatically change the coagulation pH.

Ultimately a number of issues resulted in high aluminium levels being supplied:

• The operator making such a dramatic change to the coagulation pH towards the end of the working day and not waiting long enough to see the effect of these changes.
• The delay in going to site due to high priority water quality alarms either not being called out or responded to by stand-by operator.

Actions to address this incident include:

• Instruction given to site operators not to change WTW process set points without Process Scientist approval.
• The site alarms and telemetry have been reviewed and the required amendments have been made.
• An Aztec 2000 Coagulation Controller Installed May 2007. Full automatic controlling of coagulation process will be implemented on completion of front end remedial works mentioned below.
• Flofilter sand media is being replaced.

DWQR Assessment on the incident

The DWQR assessment of this aluminium failure incident is that it was essentially caused by the operator adjusting the pH control to deal with an increase in turbidity without first discussing the issue with the local process scientist. The situation was exacerbated by the changes having been made towards the end of the working day and the operator leaving the works without waiting to check the effects of the changes he had made. Furthermore, it appears that high priority alarms were suppressed since the operator claimed not to have received any alarm calls later that night or early the next morning. Even when the operator did acknowledge the aluminium alarm there was a delay before he arrived on site.

DWQR notes the actions being taken to improve the process at this works. However, more fundamentally, there is a real issue about operators not changing the process settings without first discussing these with the process scientist and also about delays in responding or not responding to pH, turbidity and aluminium alarms. DWQR notes that operators should by now have been instructed not to make changes to the process without discussing these first with the process scientist and that SW is investigating the delays in response to the alarms. Operators must understand the importance of reacting to alarms and that SW's due diligence can be brought into question if there is any failure on their part in this area.