Pesticides in tap water: how are controls carried out in France?

In September 2022, the magazine Complément d'enquête (France 2) and the daily newspaper Le Monde revealed that by 2021, 12 million people in France had been affected by exceedances of drinking water quality thresholds for pesticides and their "metabolites" - the by-products of pesticides, the result of their evolution over time.

Julie Mendret, University of Montpellier and Alice Schmitt, University of Montpellier

AdobeStock_529005836 © AIDa.videophoto - stock.adobe.com

On February 15, 2023, Anses announced that it was initiating a market withdrawal procedure for S-metolachlor, a herbicide whose metabolites are responsible for major groundwater pollution in France. Its metabolites, notably metolachlor-ESA, metolachlor-OXA and metolachlor-NOA, are believed to be present in French groundwater in concentrations exceeding the quality limit set by European legislation.

This worrying situation calls for a review of the way drinking water is controlled in France.

Mainly used for agricultural purposes in rural areas, pesticides, once spread, diffuse into the soil and can contaminate underground resources. They can also enter surface waters (rivers, lakes, etc.) through soil erosion or runoff during rainfall.

Yet the water used to produce our tap water is drawn from these same resources. It is therefore potentially polluted by pesticides. For this reason, tap water in France is subject to regular health monitoring.

Water monitored on several levels

This monitoring consists of surveillance, carried out by the person morally responsible for producing or distributing the water, and sanitary control, carried out by the regional health agencies (ARS), in application of the provisions of the French Public Health Code and European Directive 98/83/EC.

The ARS oversees the entire healthcare system at regional level. Public players, local authorities, healthcare professionals and associations all interact with the ARS. It is the reference point for decisions taken by the prefecture in the field of health, and carries out sanitary water controls on its behalf.

Samples are taken at the resource level (i.e. at groundwater or river level, at the point of capture before treatment), at the treatment plant outlet and at the distribution point (i.e. at the tap).

The parameters measured and the frequency of sampling are specified by decree and depend on the control points. At the distribution point, for example, nitrate and iron levels are checked, as well as the presence ofE. Coli. Medicines are not currently subject to these controls.

Sampling and analysis are carried out by laboratories approved by the French Ministry of Health to control water quality. These laboratories must comply with certain analysis methods and achieve a certain level of "analytical performance", i.e. present regular and reliable results.

Indeed, the validity of the data produced is closely linked to the reliability of the analyses, and the development of a specific analysis technique for each substance can be long and laborious. In particular, a compound's detection limit must be below the regulatory value in order to interpret its analysis. This requires particularly precise analytical instruments.

What's more, other compounds present in the water sample may also interfere with the analysis of the parameter in question. For this reason, analyses should only be carried out by laboratories with proven expertise.

The cost of analysis means that choices have to be made

As far as pesticides are concerned, over 1,000 substances are currently authorized and used on crops in France. What's more, some pesticides and their "metabolites" can persist in the environment for several decades, and must therefore continue to be monitored even long after they have been banned.

Although the number of pesticides and metabolites tested has been increased since the end of 2020, it is not currently possible to test all these substances, for reasons of cost and resources.

[Nearly 80,000 readers trust The Conversation newsletter to better understand the world's major issues. Subscribe today]

For this reason, the French health authority (Direction Générale de la Santé) specifies that pesticide testing in drinking water should be targeted according to the likelihood of their presence and the risks to human health.

The ARS selects the molecules to be tested on the basis of local agricultural activities, cultivated areas and quantities of pesticides sold. On average, 170 pesticides and metabolites are researched, with considerable variations from one region to another.

In the Grand Est region and Brittany, around 200 pesticides and metabolites have been tested since the beginning of 2021.

Inspection frequencies depend on the flow rate of the catchment and the size of the population served, with variations ranging from 0 annual inspections to over 800, as indicated by the data provided by the AIDA website. The Ministry of Health then makes available the results of the ARS health checks.

How are pesticide regulatory values established?

The "quality limit" refers to the maximum concentration of pesticides in drinking water permitted by law under normal circumstances. It is 0.1 micrograms per liter per individual substance, and 0.5 µg/l for the sum of pesticides and metabolites measured, with the exception of molecules considered more dangerous, for which the quality limit is set at 0.03 µg/L: aldrin, dieldrin, heptachlor and heptachlorpoxide.

Classified as "persistent organic pollutants" and banned in France since the 1990s, these molecules continue to pollute certain soils and water resources.

Unlike the standards set for other pollutants, this regulatory standard, defined by a European directive published in 1998, is not based on toxicological criteria, but on the detection limits of analytical equipment at the time.

More specifically, with regard to pesticides and their metabolites, the European Commission follows the so-called ALARA principle (for as low as reasonnably achievable): it suggests keeping their concentrations as low as reasonably achievable in the interests of resource protection.

It is therefore not a health standard and does not mean that water is unfit for consumption when this value is exceeded.

Exceeding regulatory standards

The limit above which exposure to the molecule is considered dangerous is defined by the maximum health value, known as Vmax.

Vmax is set by Anses in the following way: each compound is assigned 10% of a toxicological reference value via dietary water exposure (assuming that a 60 kg individual consumes 2 liters of water per day throughout his or her life).

The WHO considers that 10% of our chronic exposure to pesticides is linked to the water we drink (according to the Anses, this figure drops to 5% in France), while food accounts for 80% of this exposure; the rest comes from our environment, via the air we breathe, for example.

Vmax has no legal value and can sometimes be much higher than the regulatory value, for the reasons explained above: the ALARA principle is applied to protect the environment and as a precautionary principle for our health.

These Vmax have a limited use of nine years, during which time so-called remedial actions (e.g., improving the water quality of the resource, installing treatments) must be implemented.

Only if the Vmax is exceeded will tap water consumption be immediately restricted.

Compliance with Vmax therefore authorizes water distribution even when it exceeds the legal standard for pesticides. In the latter case - i.e., when the regulatory value is exceeded, but not the Vmax - local authorities apply for a derogation to continue distributing this non-compliant water, and new regulatory thresholds are then set for a period of 3 years, renewable once.

These derogatory values can be dozens of times higher than the initial regulatory values, and are set so as not to cut off the water supply to thousands of people.

Article D1321-104 of the French Public Health Code sets out how town halls are to inform the public: the town hall must post the results of water testing within two working days of receiving them, except in emergency situations where measures are taken to inform users as quickly as possible and to propose palliative solutions, such as the distribution of bottled water.

When the Vmax has not yet been established by the Anses, notably for lack of scientific data, restrictions on use are applied, or a transitional health value may be defined by the DGS and submitted for the opinion of the Haut Conseil de la santé publique (HCSP).

On June 15, 2022, an instruction from the French Ministry of Health set a transitional health value of 3µg/L in response to the HCSP's recommendations.

This is the value that will now be used to implement water consumption restrictions.

A lack of data on pesticide metabolites

Still from a regulatory point of view, pesticide metabolites present in drinking water are classified into two categories, according to their hazardousness: "relevant" and "irrelevant".

According to Anses, a pesticide is considered relevant for water intended for human consumption "when there is reason to consider that it (or its transformation products) could generate an unacceptable health risk for the consumer".

Depending on the category under consideration, limit values in drinking water differ. Relevant metabolites are subject to the same quality limit of 0.1 µg/l as pesticides, while non-relevant ones are subject to a limit of 0.9 µg/l.

Deciding whether a pesticide should be classified as "relevant" is far from easy, as this definition leaves considerable room for interpretation. Added to this is the lack of data and studies on these products, which makes it impossible to know their real impact on health.

This is why, in the autumn of 2022, S-metolachlor metabolites were already in the news: when the relevance of these metabolites was reassessed, their limit was raised to 0.9 µg/l, making water unfit for consumption in some French communes once again potable, although without any change in quality. These changes are a source of incomprehension for consumers, who may justifiably have doubts about the quality of their tap water.

How can we limit contamination?

A first option for limiting contamination of drinking water is logically to limit inputs at source, by ceasing the massive use of pesticides. There remains, however, the question of persistent organic pollutants, which can remain in the environment for years or even decades.

There is also a curative solution, which would involve adapting existing treatment processes. Indeed, the typical conventional process currently in place is generally not sufficient to eliminate all pesticides. For this, a multi-barrier approach must be adopted, adding several treatments to the conventional steps.

A good example is the combination of powdered activated carbon and ultrafiltration (a membrane technique), which seems to be particularly well-suited to the removal of pesticides and their metabolites.

Membranes could thus find their place in potabilization processes. In particular, nanofiltration can eliminate a large number of pesticides, as is the case at Mery-sur-Oise, where the potabilization unit has incorporated a nanofiltration process since 1999. Only membranes complying with thedecree of June 22, 2012, which guarantees their safety, can be used in the production of water intended for human consumption.

How should you behave?

While drinking bottled water should not be a systematic solution because of its catastrophic environmental impact, as well as possible health risks linked to the consumption of plastic and plasticizer micro-particles, it can nevertheless be a solution in the event of contamination of your water distribution network.

However, while bottled water is subject to sanitary control by the ARS, semi-croplastics and plasticizers are not on the list of parameters to be tested for...

Julie Mendret, Senior Lecturer, HDR, University of Montpellier and Alice Schmitt, Post-doctoral fellow in process engineering, European Membrane Institute, University of Montpellier

This article is republished from The Conversation under a Creative Commons license. Read theoriginal article.