Water Quality Event Detection/CANARY
Rapid and accurate detection of contamination incidents in drinking water is critical for notifying consumers of threats and risks to public health and for making remediation and recovery decisions. Sandia National Laboratories and the United States Environmental Protection Agency (EPA) developed the CANARY Event Detection software to enable online contaminant event detection for time-critical decision making in both routine and emergency water quality assessments. As a free software tool, CANARY is available to drinking water utilities of all sizes (worldwide) who are striving to provide the best quality water to their customers.
CANARY is a software package that performs on-line, multi-variate, event detection from networked sensor data. Employing statistical forecasting and classification algorithms, CANARY continuously analyzes time signals for anomalous conditions. The software allows for:
- The use of a standard data format for input and output of water quality and operations data with the ability to directly connect to existing utility SCADA systems.
- The ability to select different detection algorithms (CANARY contains three different mathematical approaches for analyzing the data).
- Custom configuration of detection algorithms to suit the background water quality variation of any monitoring station.
- Both an online operations mode and an off-line evaluation mode.
- The ability to generate data needed to establish performance metrics (e.g., false alarm rates).
One example of a water utility currently using CANARY software is Singapore’s national water authority, (PUB), where the use of this tool has enabled a “quantum leap” in the utility’s practice. Historically Singapore’s water utility has depended on preset limits of three water characteristics to determine water quality. With the implementation of CANARY, relative changes in the patterns of the three parameters can be used to uncover water quality events, even if each individual parameter lies within the alarm limits. This dramatically improves PUB’s ability to respond to water quality changes, and allows PUB to arrest poor quality water before it reaches consumers.
CANARY is compatible with any sensor technology running on any information technology platform and can be easily modified for specific applications. In contrast to proprietary systems, CANARY provides the end-used with transparency in algorithms and their parameterization, which is important for utility and location-specific customization.
Through an open-source licensing approach, CANARY allows all utilities access to state-of-the art event detection capabilities that can leverage their existing investments in water quality sensors.
For more information on Sandia National Laboratories’ R&D 100 Award winning software, CANARY, click on the link to the right.
Optimal Sensor Placement/SPOT
Early Warnings Enhance Water Security and Provide Faster Crisis Response
Several elements of Sandia’s research revolve around the optimal design and management of Contamination Warning Systems (CWSs). A well-designed CWS protects drinking water systems by identifying a significant range of contaminants early enough to reduce public health and economic consequences of a contamination event. In conjunction with other tools such as physical monitoring, CWSs use sensors to detect drinking water contaminants. Sensor placement optimization is an important component of CWSs, particularly for large water distribution systems that can include thousands of miles of mostly underground pipe.
Sandia, in partnership with the U.S. Environmental Protection Agency, Argonne National Laboratory, and the University of Cincinnati, provides support for CWS networks with its Sensor Placement Optimization Toolkit (SPOT). SPOT facilitates the creation of sensor networks that can provide early warnings for multiple contaminants within a given water distribution system.
The toolkit provides discrete optimization tools that ascertain the number and optimal locations for contaminant sensors while minimizing monitoring cost and response time. SPOT also facilitates improvements to distribution system management and can be used to simulate the effects of different response times to a contamination incident.
Proven Results and Unique Advantages in Protecting Water Supplies
In a 2006 Sensor Network Optimization, SPOT was used to design sensor placements for eight large U.S. cities and execute actual placement in four of those. The estimated reduction on fatalities from high consequence attacks on drinking water was 48%, with a $19B median reduction in estimated value of lives lost due to high consequence attacks.
SPOT also incorporates attributes for each water distribution system, including utility-specific system network models, and permits the user to examine trade-offs among different network designs by integrating customized performance objectives such as population-based health measures, volume of contaminated water consumed, and time to detection. The toolkit also provides optimization techniques that can function on limited computer memory, addressing runtime restrictions associated with end-user computer systems.
SPOT has additional application possibilities, including protecting air networks in sensitive buildings, detecting intruders in road networks, and providing physical site security protection.
Image to the right: Containment Warning System (would be particularly effective if sample sensor network locations could be overlaid on the system)
Response/Recovery/Restoration (WST)
The Goal of the Water incident Response Management (WST) tool is to integrate real-time models and sensor data to manage contaminant incidents. WST is uniquely designed with a dynamic response strategy that is able to:
- Predict the source of the contaminant
- Plan manual samples
- Collect grab samples
- Initiate flushing
- Hyperchlorinate pipes
- Redirect flows with valves
- Isolate the contaminated sub-network
- Provide continuous iteration until contaminant is identified