Pesticider og vandværker. Udredningsprojekt om BAM-forurening
Summary and conclusions
During the past years, an increasing number of pesticide pollutions have been encountered in Danish water supply wells. In order to investigate this problem further, the Danish Environmental Protection Agency has initiated the present project regarding pollution of BAM (2,6-dichlorbenzamide) in water supply wells. BAM was chosen among the pesticides and their degradation products as it is regarded as the most problematic pesticide-related compound for the water supplies.
The purpose of the project is to evaluate the geological conditions under which pollution with pesticides occur and, more specific, the types of groundwater wells and well constructions that are especially susceptible to pesticide pollutions. Furthermore, the aim is to describe the extent of BAM-pollution that has already infiltrated to groundwater aquifers in order to estimate the duration of the pollution after the herbicides Prefix and Casoron are no longer used. For this purpose, the transport ways of BAM from contaminant sources through groundwater aquifers and via leaky wells are investigated.
The project has been divided into the subsequent sections:
 | Selection and investigations of waterworks polluted with BAM |
| Investigations of well constructions and their significance for BAM-pollution |
| Investigations and mapping of BAM contaminant sources |
| Experimental studies of degradation and sorption of BAM and the parent-compound dichlobenil |
| Description and evaluation of different ways of transporting BAM to the groundwater aquifer and analysis of the durability of BAM-pollution based on modelling |
| Preparation of guidelines for handling and remediation of BAM-pollution at waterworks |
Data regarding waterworks polluted with BAM were collected from the Danish counties in order to establish an overview of the extent of the problem and to point out a number of localities suitable for further investigations. Twelve waterworks was selected for investigations, as they represent different geographical and geological conditions in Denmark as well as big and small waterworks and urban and agricultural catchment areas.
From the investigations knowledge was obtained about BAM contaminant sources and spreading and leaching of BAM from the sources. Furthermore, different types of well constructions were examined in order to evaluate their significance for the groundwater pollution.
Studies of degradation and sorption of dichlobenil and BAM were also part of the project, since data about BAM in the literature is generally limited. The experiments were carried out in the laboratory using sediment and groundwater with different contents of clay and organic matter under varying chemical conditions of the groundwater.
The extent and durability of BAM contamination were evaluated based on model simulations for 6 representative main aquifer types in Denmark. The evaluation was carried out by means of mathematical modelling based on data collected from the project regarding contaminant sources, field concentrations of dichlobenil and BAM and results from the degradation and sorption experiments.
BAM contaminant sources and spreading of BAM pollution
The pesticide (or pesticide degradation product) which has hitherto polluted most water extraction wells in Denmark is BAM. BAM derives from the herbicides Prefix and Casoron and is a degradation product of the active substances dichlobenil and chlorthiamide in these herbicides. The herbicide products were used in the period from 1965 to 1997 but are now prohibited.
The average annual consumption in Denmark of chlorthiamid and dichlobenil was 29,000 kg/year in the period 1965-1997. The application dose of dichlobenil was around 4-30 kg active substance per ha.
Prefix and Casoron were used widely by many different users in different areas (not in order of priorities):
| Courtyards for farm holdings |
| Gardening/market gardens (in greenhouses, courtyards and surrounding areas) |
| Plantations (under bushes and trees in fruit- and spruce plantations and on the courtyard) |
| Nursery gardens (in beds and walks) |
| Allotments and kitchen gardens (in beds and walks) |
| Apartments blocks (in beds, walks, parking areas, playgrounds) |
| Single-family houses (in beds, walks and driveways) |
| Parks/recreational areas (in beds, walks and playgrounds) |
| Churchyards (walks) |
| Sport grounds (walks, running tracks, tennis courts etc.) |
| Schools and child-care institutions (beds, walks, playgrounds) |
| Public buildings/hospitals etc. (beds, walks, parking areas) |
| Roads (road verges, sidewalks, footpaths) |
| Railroads/railroad stations |
| Waterwork sites (non-paved areas) |
| Surrounding areas around waterwork wells. |
Weed control using Casoron and Prefix occurred both in urban and agricultural groundwater catchment areas. The number of BAM contaminant sources, however, is typically much higher in urban areas and the contaminant sources closer compared to agricultural areas.
Investigations of the BAM contaminant sources show concentrations of typically up to 100 µg/kg of the parent pesticide dichlobenil in soils. The average concentration in the top 1 meter is at approx. 55 µg/kg. Dichlobenil is found in soil samples in more than half of the investigated sites, regardless of knowledge of dichlobenil-containing herbicides having been used at the sites or not. It is assumed that the herbicides have not been used since 1997, and that the observed concentrations of dichlobenil represent residual concentrations of dichlobenil that has not been degraded. Small amounts of BAM is also detected in the soil, which shows that dichlobenil is still degraded and that BAM continues to infiltrate from the surface. BAM is also detected in pore-water from vadose zone (in concentrations of up to 37 µg/l) as well in groundwater aquifers near the ground surface, which shows that BAM is still leaching through vadose zone to the groundwater. Often, 2.6-dichlorobenzoic acid also is detected in pore water samples from the vadose zone, indicating that BAM is degraded since 2.6-dichlorobenzoic acid in the literature is reported as being a metabolite from BAM.
Calculations show that the residuals of dichlobenil in soils from the upper few meters of the investigated sites can still cause groundwater contamination for many years to come. However, the main part of BAM is considered to have already infiltrated into deeper parts of vadose zone or into the groundwater aquifer.
Apart from the direct infiltration of BAM from the contaminant sources there might be a risk of spreading BAM via surface water (streams) and rainwater runoff from roads and paved surfaces. For instance, rainwater basins without firm bottom imply a risk of infiltration into the groundwater of BAM polluted rainwater.
Investigations of selected groundwater catchments have shown that the many BAM contaminant sources in urban and agricultural areas have caused widely spreaded BAM-pollution in the groundwater aquifers. The spreading of BAM-pollution is caused by leaching of BAM from all the contaminant sources resulting in many small pollution plumes which often seem to combine to larger coherent plumes due to the high mobility of BAM. The complex distribution of contaminant sources in the catchment areas combined with the spreading of BAM via surface water and runoff rainwater pattern makes it very difficult to identify preferential pathways from the contaminant source to the groundwater.
In the Danish groundwater aquifers BAM pollution is found in concentrations above the official limit value for drinking water in on average 10% of the analysed wells (data from the official Groundwater Monitoring Programme, 2000). However, the degree of BAM contamination varies from county to county from 1% in Frederiksborg County to 15% in Southern Jutland County. The difference in the frequence of BAM contamination may be due to different application doses of the herbicides, different land use, and different hydrogeological conditions. Most of the BAM findings originate from groundwater at a depth of up to 40 meter, but BAM has also been observed in groundwater reservoirs at a depth of 100 meter.
Transport via leaky wells
Groundwater BAM contamination caused by transport of pollution through leaky wells requires that the dichlobenil containing herbicides have been applicated close to the well.
Investigations in this project combined with experiences from other studies show that there is a potential risk of leaks through well constructions as a result of drilling techniques, well construction techniques and well construction materials. Evidence of defective well construction was recorded in 84 % of 172 wells investigated. In all wells constructed before 1980 there may be a risk of leaks due to one or more constructional errors. Many relatively newly established wells, however, are defective too.
Leakage along the casing will in most cases not cause BAM pollution of the groundwater above the limit value. At small waterworks with lower pumping rates leakage along a casing pipe may be a problem if there are high concentrations in the secondary groundwater. In water supply wells with a high continuous pumping rate, the contribution from leaky wells is nonessential
Leakage resulting from improper well closure may cause significant BAM transport and contamination problems, since there are a great number of abandoned inadequately closed wells nearby water supply wells that are in use.
Attenuation and degradation of dichlobenil and BAM
Degradation and sorption of dichlobenil and BAM are processes crucial for the resulting concentrations and duration of BAM-pollution in the groundwater. The experimental studies show that the sorption of dichlobenil depends on the content of organic matter and clay in the sediment. A very high sorption of dichlobenil has been observed for anaerobic clay. This means that the presence of a layer of anaerobic clay underneath the subsurface of the dichlobenil/BAM contaminant source minimize the risk for leaching of dichlobenil to the groundwater. Also, a high sorption of dichlobenil has been observed in the topsoil due to high contents of organic matter, which most probably inhibits the degradation of dichlobenil to BAM. This means that dichlobenil will remain in the top soil for many years to come. The most rapid degradation of dichlobenil to BAM has been observed at depths of 0.25-0.7 m b.s. with estimated half-live constants of 0.4-0.6 years. From a depth of 0.75 m the degradation rates decrease rapidly with depth, and at a depth of 4.0 m b.s. no significant degradation of dichlobenil is observed.
Compared to dichlobenil the sorption of BAM to sediments is much lower. However, the sorption of BAM still seem significant for most of the investigated sediments. The sorption of BAM is dependent on the content of organic matter and clay like the sorption of dichlobenil. As expected, in most cases BAM is not degradable. Significant degradation has only been observed in the topsoil (0-0.75 m b.s.) after an experimental period of 436 days. However, for one of the study sites a significant but low degradation was observed in a sediment from a depth of 5 m b.s. just above the groundwater level, but no significant degradation has been observed below the groundwater level. The experimental data indicate that the transportation rate of BAM in groundwater aquifers is controlled by the low but significant sorption to the sediments.
Durability of BAM contamination in groundwater
The extent and durability of BAM pollution in the groundwater and water extraction wells depend on a complex interaction between the following main factors:
| The BAM contaminant sources and their location |
| The source strength (application dose) |
| The geological conditions (type of aquifer, composition and thickness of overlaying aquitards) |
| The hydrological conditions (precipitation, surface run-off, infiltration, groundwater flow etc.) |
| The groundwater catchment (extraction and yield) |
| Leakage along inadequately or abandoned wells |
| The ability of the soil and aquifer to attenuate the pollution (degradation and sorption of dichlobenil and BAM). |
The evaluation was carried out by means of mathematical modelling using a linked aquitard- and groundwater model FRAC3dvs and MODFLOW/MT3D, which include the above mentioned factors of importance for the durability of the pollution. Model simulations were carried out for geologically different types of aquifer in order to describe the extent and durability of BAM under various typical Danish hydrogeological conditions. The model evaluation therefore provides an overall picture of the average extent and durability of BAM in groundwater aquifers with rough intervals for concentrations and duration of contamination.
The modelling shows how BAM-pollutions appear and develop in the groundwater, and the future evolution is highly varying and dependent on e.g. the thickness of the overlaying clay aquitard and of the water balance.
The modelling indicates that the maximum BAM concentrations will occur in groundwater aquifers with only a thin or no protecting clay aquitard above. The highest concentrations are expected in approximately year 2000 and the concentrations will decrease to below 0.1 µg/l after a period of 20-50 years.
The modelling indicates that groundwater with thick protecting clay aquitards above will have a breakthrough of BAM pollution after many years. For groundwater aquifers with 30 m of protecting clay aquitard above BAM contamination is only expected after approximately 50 years and hereafter the BAM concentration will slowly increase until the middle next century. However, the modelling shows that the lowest maximum concentrations in groundwater occur where the thick inter-layered clay layers are without fractures.
The modelling moreover indicates that the groundwater will be more exposed to BAM pollution underneath aquitards of glacially disturbed clay and heterogenous clay types containing abundant thin sand layers. In such areas the model indicates that, other things being equal, there is or there will soon be a breakthrough of BAM to aquifers and following long duration of BAM contamination.
The highest BAM concentrations appear in groundwater under the urban areas of the model catchment, while the BAM concentrations in the agricultural areas in general are lower. Lower BAM-concentrations and longer distances between the BAM contaminant sources in the agricultural catchment areas mean a lower frequency of BAM findings and shorter durability compared to the urban catchment areas.
It shall be emphasized that the results of the model are very sensitive and dependent on the degradation potential of BAM. In the laboratory studies no degradation of BAM was observed in experiments representing groundwater aquifers. However, if BAM is degraded in groundwater, the degradation rate must be very low since the model calibrations with "no degradation" seem to fit the actual findings of BAM in the groundwater. Nonetheless, the predicted durabilties of BAM pollution are very long, and even a very slow degradation of BAM (e.g. half life of 10 years) in the groundwater would reduce the predicted extend and duration of BAM pollution dramatically.
Conclusions
The following overall conclusions from the project can be made:
| The herbicides Prefix and Casoron containing dichlobenil og chlorthiamid, which is parent pesticides to BAM, have been applicated to many areas by a great number of different types of users in both urban and agricultural catchment areas. |
| BAM is the pesticide (or degradation product) which so far has polluted most water supply wells (24%) and in 10% of the wells the concentration is above the drinking water limit of 0.1 µg/l. |
| Many water supply wells are in a bad shape and can cause a rapid leakage of pesticide pollution into the groundwater. Evidence of defective well construction was recorded in 84 % of 172 wells investigated. However, BAM contamination of water supply wells via leakages in wells is probably often minor in comparison to BAM contamination caused by infiltration through the soil layers. At small waterworks with lower pumping rates preferential pathways via leaky wells may cause problems if high contaminant concentrations are present in secondary groundwater. |
| Approximately 90 % of the observed BAM contaminations are observed in shallow groundwater wells of 40 m or less. As expected, there is a tendency that the upper, young groundwater is more polluted than the deeper, old groundwater. |
| Despite the fact that application of Prefix and Casoron was abandoned in 1997, the parent pesticide dichlobenil is still detected in many top soils, and the residual pool of dichlobenil may cause leaching of BAM in the future. However, the main part of dichlobenil is degraded, and thus the main part of BAM has most probably already infiltrated from the topsoil. |
| The parent pesticide dichlobenil is strongly sorbed to soil and sediments, which to some extent inhibits the infiltration of dichlobenil to the groundwater, but the strong binding also inhibit the degradation of dichlobenil to BAM. Additionally, a low but significant sorption of BAM to sediment and soils is observed which may result in a slight attenuation of BAM. |
| No degradation of BAM is observed in the experimental studies representing groundwater conditions during an experimental period of 436 days. |
| A low but significant degradation of BAM is observed in top soils and in one sample from a depth of 5 m, with measured half live constants in the range of 3-16 years. |
| Modelling based on the main Danish types of geological conditions, typical patterns of contaminant sources and the expected sorption and degradation of dichlobenil and BAM shows that BAM pollution will appear with a high frequency and for a long time in water extraction wells in both urban and agricultural catchments. |
| The modelling indicates that the maximum BAM concentrations will occur in groundwater aquifers with only a thin or no protecting clay aquitard above. The longest durability will be expected in areas of thick clay aquitards. |
| There are a few remediation alternatives available in order to reduce the spreading of BAM pollution. Especially remedial pumping often seems to reduce the BAM problems at waterworks. Also, it is recommeded that renovation of leaky wells is initiated. Finally, as a last resort the groundwater can be filtrated through activated carbon. |
The project has covered a lot of questions, that motivates further initiatives and activities in order to ensure drinking water without BAM or other pesticides in the future. It is recommended that development projects are initiated in order to establish the necessary knowledge and basis for further initiatives. Such project may include:
| Further studies of the fate of BAM in soil and groundwater, e.g. detailed studies to clarify if BAM is degraded in the groundwater and if the adsorption of BAM is irreversible, both things important for the durability of BAM. Also, studies focusing on the importance of different ways of transporting BAM to the groundwater which is essential for the extent and durability of BAM. |
| Investigations focusing on a more precise mass balance of BAM for the overall water balance. |
| Evaluation of the consequences of the large number of leaky wells for groundwater pollution with pesticides and the possible remediation techniques towards BAM pollution, including the effect of using these remediation techniques. |
| Investigations and evaluation of the toxic effects of BAM for possible acceptance of temporary elevated BAM-values in the groundwater above the limit values. |
Furthermore, there is a need for cooperation between counties and water supplies in order to integrate further activities focusing on:
| Examination of the frequence of BAM contamination levels and spreading of BAM regionally |
| Monitoring of BAM contaminant sources |
| Monitoring of the groundwater |
| Reduction or possibly removal of the contaminant sources |
| Reduction or possibly removal of unnecessary pathways of contamination (e.g. leaky wells) |
Finally, the results from the project indicate that BAM pollution will appear at many waterworks sooner or later and therefore the water supply security will be minimized. There is thus a great need for guidelines for the handling of future pesticide pollutions, so that the waterworks can prepare themselves to handle the eventual appearances of pollution in their supply wells by e.g. establishing cooperations between several waterworks in order to solve the problems.