Office of Councillor Jeff Leiper, Kitchissippi Ward, Ottawa | (613) 580-2485  | jeff@kitchissippiward.ca
Responsive image

Learn something new every day: turbidity

You are here

One of the most fascinating parts of this job is learning how the City works beneath our feet. Most of us don't think very often about the civil works and engineering that mostly just hum along.

Recently, a small item in the 2015 water operations report caught my eye. Since Walkerton, City Councillors in Ontario have had a personal obligation for the safety of their cities' water supply. It's a higher level of accountability on us that for most other City functions. When I saw that in early 2015 there was a failure of a server responsible for monitoring filter operations, I followed up with staff to see what the implications of that were and how we would address it in future.

The answer they provided back both provided me significant comfort that the issue is well-understood and being addressed. But, it was also fascinating as a lay-person to get this insight into a part of our City's operations that I would otherwise never have the opportunity to learn. I always maintain that Ottawa has the best tap-water in the world. I want to thank City staff for the thorough explanation, and hope at least some of our residents will find it as interesting as I did.

***

Thank you for your question on this incident and please excuse my delay in responding. Let me start by explaining turbidity:
 
Turbidity is a means of determining the amount of particulates in a liquid. By comparing turbidity before and after a treatment process, we can understand the particle removal effectiveness of that process. Particle removal at our Britannia and Lemieux Island Water Purification Plants (WPPs), which is the backbone of our water treatment processes, is provided by the processes of coagulation/flocculation/sedimentation and filtration. The measurement itself is one of shining a light through a flowing sample of water and determining how much of the light is scattered while passing through the liquid. More light scattered indicates more particles in the liquid; less light scattered, fewer particles in the liquid. A lower number is better. We measure incoming/raw water turbidity, post sedimentation turbidity and post filtering turbidity, both as individual filter performance as well as a combined filter discharge.
 
In accordance with Ontario drinking water regulations, a filter from a surface water plant, like those installed at the City’s Lemieux and Britannia Plants, must automatically shut down if its effluent turbidity exceeds 1 NTU. In the City of Ottawa, our plant’s filters are automatically shut down, by Programmable Logic Controllers, should a filter’s effluent turbidity exceed 0.1 NTU for 5 minutes. Provided that all other water treatment processes and barriers are functional, the health risk associated with a single filter being operating above 0.1 NTU, even for an hour or so, is almost negligible. Nevertheless, filter operation must always be carefully monitored as it is a critical water treatment process.
 
As to the events reported in the Lemieux Island Summary Report, our explanation follows below:
 
Drinking water treatment is fundamentally based upon a “multiple barrier” approach that requires designs and operational approaches to be such that no single process failure or impairment results in an overall service delivery reduction. This approach extends beyond treatment process and equipment design and includes the monitoring of water treatment processes as well.
 
Field instruments, such as the turbidity analyzer, communicate their results to the local Programmable Logic Controller (PLC) that then transmits information to the Plant’s Operator Computer Terminals via the communications server.  The local PLC is the point of water treatment process control.
 
On January 26, 2015, the communications server failed, preventing operational data from being reported from the local PLC to the Operators Console. In this case, all of the filter turbidity analyzers remained fully functional and continued to monitor and control filter performance, however, the data was not being reported to the Plant’s Operator Computer Terminals.  If the data is not reported to the computer terminal, then the system loses an ability to trigger an alarm in case of an exceedance.
 
In response to this server outage, Plant Operating staff monitored filter operation by periodically checking the local display of each filter’s effluent turbidity analyzer that remained on-line throughout. As an additional precaution, operating staff cross-checked the results of the on-line turbidity analyzer using bench top lab equipment.  If a filter had exceeded its performance target (our alarms are set well within the regulatory standards), the filter would have been shut down right away. The communication server was repaired within two hours and the system restored to normal operating conditions.
 
It is also important to note that this single point of turbidity monitoring is duplicated with an additional turbidity monitoring instrument that monitors total Plant production. This monitor remained on-line and operating throughout the communication failure and this second turbidity monitoring analyzer provided confirmation that treated water quality remained high throughout the incident.
 
Finally, as part of a larger and previously planned life-cycle replacement project of Supervisory Control and Data Acquisition system equipment, all of Lemieux’s plant servers were replaced by year end 2015. As an important reliability improvement of that upgrade, redundant servers have been provided.
 
City staff have taken a precautionary approach in reporting this incident as a non-compliance event, despite the fact that the filter continued to be continuously monitored throughout the incident by the individual turbidimeter itself, by an operator on a precautionary basis and collectively as a part of the production turbidity monitoring. We are reviewing this situation with the MOE CC to determine if our interpretation is consistent with the MOE CC’s interpretation.
 

Posted May 16, 2016