WATERNET
Distributed Water Quality Monitoring
using Sensor Networks

Keywords:monitoring water quality, data fusion and validation, decision support system
Application site: Paris on the River Seine, F; Llobregat river basin, Barcelona, E
E-mail: abillington@citi.suez-lyonnaise-eaux. fr

Context
Since in many parts of Europe, river water is a major source of drinking water, treatment plant operators are faced with the more and more difficult task of determining water quality and deciding whether to use the available resources or not. Increasingly centralised monitoring is placing a major burden on operators who can no longer effectively oversee the quantity of data coming into the control rooms. Such difficulties may increase the risk of an incident which is detected too late or misinterpreted.

Objectives
The objective of WaterNet was to provide the users, namely drinking water authorities, with a suitable technology and it has demonstrated the usability and appropriateness of newly developed methods in real-life applications. By using such technology, the operators and authorities are helped by an intelligent aggregation and simplification of the vast amount of environmental measurement data.

Operators, monitoring large-scale processes such as drinking water production and wastewater treatment plants, have to manage large amounts of data, from on-line sensors and from laboratory analyses. At these installations there is a strong need for appropriate methods and tools to achieve a clear assessment of the current state of the water quality, to quickly detect major pollution events and to assess trends of the water quality, both short and long term. Current control and supervision systems do not fulfil these needs.

Results
The WaterNet project has delivered and validated, on real applications, a tool capable of overcoming the bottleneck in real time environmental and process monitoring system information processing. The WaterNet tool combines a number of useful methods which allow operators to effectively deal with large quantities of real time data by presenting high level assessment information in a form that is wished for.

WaterNet tool also implements other functionality for the diagnosis of potential events based on user defined rules describing expected behaviour during a specific foreseen event type. In the case of pollution being detected WaterNet also implements functionality estimate the time of arrival and the expected concentration at the downstream location. Information generated by the WaterNet tool is presented in a hierarchical manner design and suited to it role as a real time monitoring system. The project developed an application supporting a number of selected data assessment methods. The selection was based on the identified needs of the WaterNet target users. The methods fall into three groups:

Specifically these requirements led to the development of a tool embodying the following functionality and characteristics: This functionality was built into a system architecture based on distributed, self-contained computational units enabling multiple users with different needs to share and exchange data. The tool developed was validated at two application sites (in Paris and Barcelona) over a period of 4 months.

Technical characteristics
A tool designed on operate on a Windows NT platform and SQL Server was developed during the project programme. Specific methods for validation, assessment and classification were implemented, tested and validated at the application sites.

Low level validation, synchronisation and missing data gap filling
Low level or single parameter validation provides two methods for attributing a confidence value in the range 0 to 100% to each measurement based on its current value or dynamics. Low level validation is applied to all incoming data. Data synchronisation and missing data gap filling implements a simple method for dealing with missing data and for synchronising the asynchronous data. Synchronisation is necessary to allow subsequent application of multi-parameter assessment and classification methods.

Multi-parameter validation
The WaterNet tool implements three methods of multi-parameter validation. These provide further estimates of the validity of a time series based on related parameters from the same or different measurement locations.

Classification
The important aspect for operators dealing with large quantities of real time information is the ability to make assessments in real time on the condition of the process, in this case the river. The WaterNet tool supports these needs via a suite of situation classification methods. These methods allow the online assessment of multi-dimensional data to provide the user with an easily understood indication of the state of the process and or its behaviour.

Additional functionality and user interface:

The WaterNet tool also implements another functionality for the diagnosis of potential events, based on user defined rules describing expected behaviour during a specific foreseen event type. In the case of pollution being detected, WaterNet also implements a functionality to estimate the time of arrival and the expected concentration at the downstream location. Information generated by the WaterNet tool is presented in a hierarchical manner designed and suited to its role as a real time monitoring system.

Transferability
The application of the WATERNET tools to new sites may be constrained by the existing data collection infrastructure. However when this exists, the problems involved with setting up a new application are confined to linking to the existing database(s) and configuration. The difficulties of dealing with large quantities of process data in real time are common in many large process monitoring applications. For this reason not only the methods and techniques implemented in the tool but the tool itself is largely applicable to other large-scale process monitoring problems. Discussions with staff from water production plants in England have identified that the development is well suited for application to joint river and treatment plant monitoring processes.

 
Do you have a comment, reaction or opinion regarding what you've read? If so, please leave it at the online discussion board!

REC * ENWAP * VIRTUAL GALLERY

The Regional Environmental Center for Central and Eastern Europe

HOME PAGESEARCH