Miljøprojekt, 1033 – Pesticider i dansk grundvand - Grumo- og boringskontroldata

Pesticider i dansk grundvand - Grumo- og boringskontroldata

Summary and conclusions

The recent 10-15 years of research has shown that many pesticides are degradable in groundwater environments. However, the degradation among other factors depends on the groundwater chemistry, especially the redox conditions. Since 1993, Danish groundwater has been systematically monitored for pesticides. Groundwater monitoring also includes a number of naturally occurring compounds (oxygen, nitrate, sulfate etc.) which characterize the redox conditions in the groundwater. The scope of the present project has been to investigate, whether the collected data shows any correlation between the redox conditions in the groundwater and the presence of pesticides.

Once a year, the Danish counties report water quality data from a network of groundwater monitoring wells and supply wells to GEUS (Geological Survey of Denmark and Greenland). The data is very diverse in the different regions of the country. The whole country, however, is well covered by monitored wells, but there is a larger concentration of wells around the bigger cities than in the open land.

The initial work of this project has consisted of a systematic screening of data. For GRUMO and BK two separate databases have been produced. From these data bases a number of extracts has been made, from which it has been possible to combine data of pesticides and redox sensitive compounds, such as oxygen, nitrate, sulfate, etc. Also information on well depths, type of aquifer and geology has been studied. For the pesticides, extracts describing the individual pesticides as well as groups of pesticides have been made. The pesticides have been grouped in: Phenoxy acids and their degradation products, Triazines and degradation products, Dichlobenil and degradation products (especially BAM), Glyphosate and AMPA, yellow compounds (DNOC, Dinoseb), Others (Isoproturon and Bentazon).

For the preparation of redox specific plots as well as for the statistically interpretation of data, a division of wells into two groups has been made. One group consisting of filters in which pesticides have been detected ("Detects"), and one group consisting of filters where pesticides has never been detected ("Non Detects").

Finally, a number of model simulations have been conducted in order to estimate degradation rates of the individual pesticide.

The phenoxy acids and degradation products, triazines and degradation products, and Dichlobenil and degradation products (especially BAM) has been studied. For Glyphosate and AMPA, yellow compounds (DNOC, Dinoseb), and others (Isoproturon and Bentazon) the data has been too limited.

Concerning the phenoxy acids, the following conclusions can be drawn:

Phenoxy acids are more often detected in reduced (i.e. without oxygen and nitrate) than in oxidized (with nitrate) groundwater. This correlation indicates a degradation of the phenoxy acids in the oxidized groundwater.
Indications of further degradation of the phenoxy acid metabolites (chloro phenols) under oxidized conditions are seen. The data is not sufficient to determine whether degradation under reduced conditions are taking place.
Model simulations have shown degradation rates of minimum 0.02 d-1 under oxidized conditions and minimum 8 x 10-4 d-1 under reduced conditions.
Statistically, it has not been possible to distinguish between "Detects" and "Non Detects" and the redox conditions in the groundwater. However, data indicate strongly that aquifers consisting of a non saturated zone followed by a wide oxidized zone, e.g. the sandy aquifers in Western Jutland, is less vulnerable to pesticides than aquifers with reduced conditions close to the surface.
Model simulations indicate that the point sources are important factors in the contamination of the supply wells. The simulations also indicate that it is the point sources and not the general agricultural use on the fields, which are seen in the GRUMO wells.

For the triazines, the following conclusions can be drawn:

The most intensive occurrence of triazines is found under oxidized conditions/young water. I.e. the occurrence is age related (dependent on transport time) and not redox dependent. Different use patterns between the different areas of the country can, however, also influence the groundwater contamination.
Under oxidized conditions, only weak indications of degradation of triazines are seen. Most likely, the majority of degradation is taking place in the top soil.
Less "Detects" of triazines under reduced conditions in the artesian aquifers than in the non confined aquifers are seen.
Degradation rates are difficult to estimate, since the source concentration and the use pattern is not sufficiently known.

For Dichlobenil and BAM the following conclusions can be drawn:

Only very few "Detects" of Dichlobenil in groundwater. This indicates that the compound is degraded in the upper part of the soil.
The most intensive occurrence of BAM is found under oxidized conditions/young water. I.e., the occurrence is age related (dependent on transport time) and not redox dependent.
No indications of degradation of BAM in the groundwater aquifers are observed.
Less "Detects" of BAM under reduced conditions in the artesian aquifers than in the non confined aquifers are seen.
There is a mass balance problem for BAM, since the source concentration is probably significantly less than expected.
Source tracking is difficult

It has been examined, when groundwater aquifers are vulnerable, and to which pesticides. However, among the groups of "Detects" and "Non Detects" no systematic exists that can designate areas, where the drinking water supply will be secure to pesticide contamination in the future.

The examination of data from GRUMO and BK has to a large extent confirmed the existing presumption of the correlation between redox conditions and degradation of pesticides.