Optimization of drinking water quality

Drinking water production plants

 

The treatment chain of a modern plant includes several stages such as membrane filtration, adsorption on activated carbon, ozonation and chlorination. The interactions between the functional parameters of the treatment stages and the physico-chemical quality of the water are complex. Modeling these interactions to optimize process control leads to concrete measurable results in terms of produced water quality and operating costs.

 

Quality of the drinking water produced

 

Thanks to an excellent application of state-of-the-art techniques, the risk of catching an infectious disease by drinking tap water is almost zero. There is no problem.

 

However, the steady increase in the number and amount of anthropogenic chemicals in water has become a cause for concern. In response to this development, investments in the order of hundreds of millions of francs are planned in the canton of Vaud. In addition, regular sampling and analysis are used to verify compliance with limit values ​​for a large number of phytosanitary products and micropollutants.

 

The chemical requirements for drinking water described in Annex 2 of the DFI Ordinance on Drinking Water and Water for Publicly Accessible Bathing and Showering Facilities (OPBD, RS 817.022.11) includes a list of 57 substances. Among the substances on this list are disinfection by-products.

 

Disinfection by-products are molecules produced during the treatment of water by chemical oxidation technologies such as chlorination. Chlorination transforms organic molecules into new molecules that are sometimes toxic to humans. This problem has been known for more than 40 years and remains a subject of study and debate at the international level. Three examples of the many recent scientific publications are:

 

• Chavez et. al. (2019). Review : Hazard and mode of action of disinfection by-products (DBPs) in water for human consumption: Evidences and research priorities. Comparative Biochemistry and Physiology Part C: Toxicology & Pharmacology. Volume 223, September 2019, Pages 53-61. https://doi.org/10.1016/j.cbpc.2019.05.015

• von Gunten, U. (2018). Oxidation processes in water treatment: are we on track? Environmental Science and Technology, 52(9), 5062-5075. https://doi.org/10.1021/acs.est.8b00586

• Tang et. al. (2020). Bibliometric review of research trends on disinfection by-products in drinking water during 1975–2018. Separation and Purification Technology. Volume 241, 15 June 2020, 116741. https://doi.org/10.1016/j.seppur.2020.116741.

Proposed drinking water production plant – Saint Sulpice II

 

The Lausanne municipality plans to build a new drinking water production plant with a capacity to treat 1,5 m³.s^-1 of Lake Geneva water. Considering the quality of Lake Geneva water, it would be prudent to study alternative water supply sources and technologies.

Comments about Saint Sulpice II

Alternative water supply sources and technologies

 

The Saint Sulpice water treatment plant is located near the discharge point of the largest wastewater treatment plant and the mouth of one of the most important river pollution sources on Lake Geneva. Moreover, the source water quality is degraded by micropollutants, plastic litter, and by the presence of algae. Alternative water supply sources in the canton of Vaud exist in the Jura (e.g. le lac de joux), the Jorat, and the pre-Alpes (e.g. le lac de l'hongrin).

 

The current estimation of the construction cost of the new Saint Sulpice water treatment plant (SS II) is CHF 125 million (source: Préavis de la Municipalité de Lausanne Nº 2022 / 09 du 7 avril 2022). Logically, the cost of upgrading the Lutry plant to the level of SS II should be added to this amount. Additionally, the costs of energy, operation and new distribution pipelines should be included in the economic evaluation. Given the high costs of the project, current and future energy scarcity, the relatively poor quality of the source water, and emerging concerns about disinfection biproducts, it seems prudent to consecrate millions of Francs to study and compare alternative water sources and treatment technologies.

 

When introduced 125 years ago, chlorination solved many problems related to microbial contamination of drinking water. More recently, it has been demonstrated that UV disinfection can be used to maintain microbiologically safe water without chlorination. For example, the association intercommunal d'amenée d'eau d'Echallens et environs (AIAE) provides safe water without chlorination (1). The city of Zurich sources 70% of its water from the lake and generally renounces to chlorination (2). Membrane filtration effectively removes many pollutants and microbial contaminants. The energy cost to pressurize the water depends on the type of membrane. The energy cost of Nanofiltration (NF) is approximately 10 times that of Ultrafiltration (UF). Considering these observations, a possible treatment sequence to investigate is UF -> activated carbon -> UV disinfection (at multiple points in the distribution network).

 

The decisions taken regarding the SS II plant will set the canton on a water supply path that will last hundreds of years.  Now, is the time to evaluate the long-term water supply strategy with consideration of a cantonal network, source quality, energy consumption, human health, and customer preferences.

 

1. https://aiae.ch/wp-content/uploads/2022/04/Article-de-lEcho-du-Gros-de-Vaud-du-14-avril-2022.pdf

2. https://www.stadt-zuerich.ch/dib/de/index/wasserversorgung/trinkwasser.html

Optimization of the quality of the water produced

 

A numerical model establishes the relationships between input parameters in a water treatment system and output parameters. For example, a model would predict the quality of the outgoing water based on the variable physico-chemical quality of the incoming water and the electromechanical parameters of the treatment plant. Numerical models are essential tools for process optimization. The so-called “artificial neural network” numerical modeling strategy has already demonstrated advantages for the identification of complex and non-linear systems such as the drinking water production process. For example, Godo-Pla et. al.1 have developed a predictive neural network model which is useful to operators for oxidant dosing. Compared to existing modeling strategies, this strategy could lead to improvements in energy efficiency, chemical consumption, and the quality of outgoing water from treatment plants and in water distribution networks.

 

1. Lluís Godo-Pla, Pere Emiliano, Fernando Valero, Manel Poch, Gürkan Sin, Hèctor Monclús. Predicting the oxidant demand in full-scale drinking water treatment using an artificial neural network: Uncertainty and sensitivity analysis, Process Safety and Environmental Protection, Volume 125, 2019, Pages 317-327.