|
InvestigationsThis section contains summaries of reports on drinking water Report on drinking water quality incident at: Scottish Water Executive Summary n 29th March 2006, a pump failure on a waste drilling mud facility operated by Shetland Oil Tools Ltd resulted in mains water being introduced to maintain a dust suppression system. Upon restarting the pump, the mains water valve remained open for an unconfirmed period of 2 minutes. Although a flap valve system was in place to act as a non-return valve (a Byelaws contravention), it is believed that this failed to operate and resulted in an estimated 290 litres of process water being pumped into the public water supply. Although Scottish Water first received a complaint associated with the contamination at 0512hrs on 30th March 2006, the details provided suggested a problem of discolouration. No evidence of contamination was apparent at this stage and, as a result, an appointment to investigate the complaint was deferred to normal working hours. Following repeat complaints and the receipt of additional information from the same customer, the complaint was escalated to the local Network Service Operator (NSO) by the Contact Centre. Although a second complainant, Enviroco Ltd (owners and co-occupiers of the site occupied by Shetland Oil Tools), claim that a call to Scottish Water was made at 0730hrs, this cannot be confirmed by Scottish Water systems and is still subject to an internal investigation. However, a call logged at 1036hrs prompted a combined investigation of the premises of both complainants located on adjacent industrial estates. This suggested the presence of a possible contaminant in the public water supply within the area of concern. However, full escalation of the problem was deferred until the most likely source and nature of the suspected contaminant had been identified (as process water from Shetland Oils Tools). Following involvement from Scottish Water's Public Health Team, the presence of the contamination was confirmed in a nearby fast food restaurant at 1500hrs. This resulted in attempts to contact key personnel within the North East Operations team to escalate the problem within the business. Following an internal Business Alert being issued at 1544hrs, central and local Strategic Incident Control Teams (SICT) were implemented and the potential area affected was confirmed by Networks. Prior to contacting the Consultant in Public Health Medicine (CPHM), further clarification of the situation was obtained by the Public Health Team. However, actions were already in place on the presumption that a 'Do Not Use the Water' notice would be issued, which was subsequently confirmed by the CPHM at 1725hrs. Although 'Do Not Use the Water' leaflets were available for distribution to customers shortly after 1815hrs, the SICTs agreed to issue explanatory letters and bottled water at the same time. Despite these not being issued for a further two hours, media statements were issued in the interim and recovery actions, such as mains flushing, were already underway. Notification to customers and distribution of the first batch of bottled water was completed by midnight on 30th March. Further alternative supplies in the form of static tanks were deployed by 0500hrs on 31st March. Scottish Water's Contact Centre received a significant number of calls over the period of the incident which was exacerbated by concurrent incidents in Edinburgh. Despite additional resources being utilised, which included the provision of a Messagelink service specifically for callers from Peterhead, a number of customers had to wait for some time before speaking to a Customer Advisor. This issue has been highlighted by customers, the local MSP and other external agencies. The response to the internal Business Alert was particularly swift with complete dedication from all those involved to restore normal supplies to customers as quickly as possible. This included initial steps to prevent an immediate reoccurrence of the problem. However, had the incident been escalated further once it became clear that part of the public water supply had been contaminated, it is possible that the overall response could have been brought forward by 1 to 2 hours. Furthermore, had the delivery of leaflets started 1 to 2 hours earlier, the overall direct notification of customers may have been achieved 3 to 4 hours sooner. However, it should be noted that recovery actions and notification to customers still commenced within 5 hours of full business escalation. DWQR Assessment on the incident Main Findings 1. Recording of information from members of the public There are several issues surrounding the recording of information by Scottish Water during the first few hours of the incident, namely: Scottish Water maintain that there is no record of a call having been received from the second complainant at approximately 07:30 on 30 March 2006 on their "Promise" System. Following a subsequent request for additional information from Scottish Water, Scottish Water have conceded that a call from the Peterhead area was received at 07:32 with a duration of 58.7 seconds but their investigations cannot relate this to a call from the second complainant. Investigations by DWQR with the second complainant have established, from the complainant's telephone records, that a call was placed from their account on 30 March 2006 at 07:32 to telephone number 08456008855 which lasted for 59 seconds. DWQR has verified that the number called was Scottish Water's call centre. Despite Scottish Water's assertions that the call duration was insufficient to convey the information the second complainant claimed to have been passed on, the fact remains that DWQR has established that the second complainant did place a call to Scottish Water at around 07:30 on the morning of 30 March 2006. It should also be noted that, in Scottish Water's own report on the incident, it was acknowledge that had this information been received and correctly acted upon, Scottish Water's response would have been brought forward by an estimated 2 hours. Additional information provided by Scottish Water indicated that call centre staff were dealing with over 2,000 calls relating to incidents affecting Scottish Water on 30 March 2006. Scottish Water should, as a matter of urgency, undertake and implement the results from the review they identified in recommendation (3) in the Peterhead Water Contamination Incident Report of April 2006 (Review decision making process for Customer Advisors). 2. Delay responding to complaint and escalation of response While Scottish Water has recognised that there was a delay in the Network Services Operator (NSO) attending the premises of the first call and has identified steps to be taken to address this in Recommendation 1 (NSOs should carry their laptops in their vans when away from their depots), of more serious concern is the time taken to escalate the response. While Recommendation 5 (Refresh SHOUT campaign or implement alternative method of reminding operational staff to escalate potential incidents) goes some way to address this, Recommendation 6 (Review communication arrangements for the escalation of suspected incidents) has no specified outcome. A more committed position than just to review - such as to have the review externally assessed and act on the conclusions identified by internal/external review - would be more appropriate. Of particular concern is the failure to escalate the incident by the NSO, the Network Team Leader (TL) and the Process Scientist. Further concern arises through the failure of the Process Scientist (or the other staff involved) to alert the Public Health Team. Scottish Water's report acknowledges these shortcomings (Section 3.2.3). Scottish Water concede in their Incident Report that a more expeditious response to the investigation and escalation could have advanced their response by 2 hours - a further 2 hours over and above the response time lost through the initial confusion over the calls being recorded. If 13:15 is taken as the time when the Public Health Team was advised of the situation then Scottish Water "lost" 4 hours as subsequent responses could have occurred from around 09:15 that morning. 3. Internal Telecommunications Issues Scottish Water acknowledge in their Incident Report that problems with internal communication systems delayed the overall response process by an hour. However, had the potential time line been followed i.e. 09:15 for advising Public Health Team, then these issues may not have come into play in this incident. 4. Internal Communications with Public Health Team The Incident Report (section 3.2.3) suggests that when the Process Scientist was contacted about water quality issues at Forehill WTW (11:50 according to section 3.1) they did not escalate the incident by notifying the Public Health Team.. The Process Scientist advised at 13:15 that the Public Health Team (and hence, the rest of Scottish Water) should be advised of the developing situation). The report suggests that the Public Health Team had become aware of the situation through other means (not specified) at around 13:30. It should not be necessary to rely on informal methods of communication to inform the business about developing incidents. Public Health Team should have been involved from an earlier point in the developing situation. It is arguable that Networks TL and NSO should have contacted Public Health Team directly once they had established the situation on site. Scottish Water's Incident Report (section 3.2.3) makes this point directly although this is perhaps too harsh a response given the apparent lack of formal training of the Networks TL in relation to the Scottish Waterborne Hazard Plan (SWHP). Nevertheless, the issue of front line staff having the appropriate training and confidence to take the initiative to escalate when the next line in the chain of command is missing should be reviewed and revised to prevent a similar situation occurring in future. Frontline staff should be formally trained in the SWHP so that their roles and responsibilities and the procedures for escalation are understood and this training should be formally recorded as part of their training record 5. Initial Classification of Incident Networks confidence regarding the affected area seems to be somewhat misplaced since the boundary values were incorrectly foamed and tagged and Burnhaven School seems to have been overlooked. Reliance on "local knowledge" about boundary valve locations and state is not acceptable when such knowledge should be captured by corporate systems such as the Geographic Information System (GIS). Scottish Water stated in the Incident Report that "although the Networks team had concluded that the affected area was confined to 200 properties and possibly the prison and oil supply base, the Senior Scientist sought confirmation from the Networks Manager" and in a follow-up response (7 July 2006), "although Networks were confident of the area affected (based on local knowledge), further verification of this was required before ruling out the prison etc. This was quickly verified from area plans and GIS data and subsequently confirmed by the boundary valve checks". This suggests a lack of confidence in the information held by the local Networks team. Scottish Water's Incident Report (28/4/06) did not identify any issues relating to boundary valves although subsequent boundary valve checks undertaken once the incident was running highlighted the fact that the wrong valves had been foamed and tagged as boundary valves. Containment of the incident to the area that was ultimately affected could therefore be considered to be good fortune and the whole issue of incorrect foaming and tagging of boundary valves and the failure to verify the correct valves on the GIS considered as a "near miss" within the incident. 6. External Communications Scottish Water's incident report identified a 2 hour delay in informing customers of the Consultant in Public Health Medicine's decision that the water should not be consumed (via "Do Not Drink" notices). In the incident report there is a recommendation that the preparation of such letters should be speeded up. It is disappointing to note that the issues of public notification have arisen in a number of incidents ranging from small-scale (Strathyre) to large scale (Glasgow, 2002) but the lessons learned identified in previous reports do not appear to have been acted upon. The incident report details the problems that arose on the day with the distribution of bottled water. No further comment on these events is necessary other than to endorse SW's internal investigation into these matters. Of greater concern is the issuing of the limited stock of bottled water to business customers, apparently on the sole initiative of the of the Networks TL. This appears to be contrary to the requirements of the SWHP. The Networks TL should not be making such decisions particularly as it was known that the stocks available in the immediate area were low and appropriate contact with the Emergency Planning Officer for the Council had not been established so the precise number and nature of vulnerable customers was not know. SW should ensure that business customers are not be given priority over vulnerable customers in contravention of the policies and procedures laid down in the SWHP. 7. Byelaws inspections In Scottish Water's incident report, it is acknowledged that, prior to the incident, there had been no byelaws inspection. In the incident report SW suggest that they are preparing a policy on Byelaws inspection which would result in a prioritised system of inspections and the inspection cycle would be at least once every 5 years. Future incidents of this nature that implicate premises where water byelaws have not been conducted within the 5 year period may be considered as a failure of due diligence and could result in DWQR preparing a report to the Procurator Fiscal. | |||||||||||||||||
| © Crown Copyright 2007 | regulator@dwqr.org.uk | Website: www.stephenrasmussen.com | |||||||||||||||||