Référence:
GAUTHIER, Vincent, BESNER, Marie-Claude, TRÉPANIER, Martin, MILLETTE, Robert, CHAPLEAU, Robert, PRÉVOST, Michèle (1999). Understanding the microbial quality of drinking water
using distribution system structure information and hydraulic modeling, American Water Work Association, Tampa |
Type:
Conférence avec publication
Organisme:
Autres
Retour
Nouvelle recherche
|
Résumé
When analyzing specific water quality problems in distribution systems, and more particularly coliform occurrences, the most frequently-asked questions concern (a) historical quality data at the sampling point or in the surrounding area, (b) water itinerary during transport to the problematic area (including residence time, nature and age of pipes materials in contact with water
), (c) the occurrence of pipe repairs or other Utility interventions such as flushing during the previous weeks, etc. Such informations are seldomly available to scientists and water quality engineers, which limits the global understanding and prediction of microbial quality in distribution systems. Even if favorable conditions for coliform occurrences in distribution systems have been identified from statistics (Volk and Joret, 1994; LeChevallier et al., 1996), the origin (intrusion versus regrowth) of these microbes is still poorly understood. As long as coliforms origins are not identified, coliforms-positive events may only be limited through the maintenance of high disinfectant residuals (Haas, 1999) or the implementation of intensive flushing programs (Oliver and Jones, 1998).
The development of hydraulic models has put forward their potential coupling to deterministic water quality models to describe the microbiological parameters in distribution systems (Servais et al., 1994; Lu et al., 1995; Dukan et al., 1996; Laurent et al., 1997). To cope with the large number of phenomena affecting water quality in distribution systems, Skipworth et al. (1999) have recently proposed to use artificial neural networks which have the ability of learning complex relationships from example data sets, which they applied to the prediction of oxido-reduction potential. Nevertheless, the complexity of distribution system microbiology limits the capacity of these models to the gross estimation of attached and free-living populations of heterotrophic bacteria as a function of a few parameters such as temperature, hydraulic residence time, pipe diameter (surface/volume ratio), chlorine, and biodegradable organic matter. Some of the outputs from these models (residence times, chlorine levels) provide an indication on the bacterial regrowth potential at different locations in the distribution systems, which may then be compared to the dates and locations of coliforms occurrences. This may allow to define most at-risk zones for microbial contamination of water through intrusion or regrowth.
In this project, we aimed to identify the parameters governing microbial quality of water in the distribution system of the City of Montréal, Qc, Canada. To achieve this goal, we took advantage of the improvement of computer performances and of the development of softwares for databases handling. This approach allows to compile and compare data from different sources (Public Works Dept, Water Quality Laboratory, hydraulic modeling) in time and space, and led to the design and exploitation of an integrated visualization tool combining:
1) the structural description of the system (pipes, tanks, valves, pumps),
2) data collected from network management (tanks levels, pump operation), water quality data from routine analysis, pipe repairs etc.,
3) computed operational parameters (flow direction and velocity, water residence time).
Using the visualization tool, information is derived which identifies the causes for coliform events in some location, and leads to the proposition of remedial solutions. |