| Forside | | Indhold | | Forrige | | Næste |
Reduced Costs of Analysing Pesticides in Small Water Supply Plants
Summary and conclusions
Introduction
By including several of the triazin-pesticides in the indicator package it is possible to increase the number of positive finds to
96.7% of the original number. More importantly, only a small percentage of the samples above the threshold of 0.1μg/l is not
identified.
A reduced pesticide package consisting of 10 indicator-pesticides (indicator package 4): 2,6-dichlorbenzamide (BAM),
desisopropylatrazine, desethylatrazine, atrazine, simazine, desethylterbutylazine, terbutylazine, diuron, bentazone, and AMPA
will be able to "catch" 97% of all the expected positive samples and, out of these samples, 99% of the samples where the
pesticide concentration exceeds the present threshold value of 0.1μg/l. The costs associated with this package using LC-GC/MS
will be approximately 60% of the costs associated with the analysis of all 26 pesticides.
An indicator package which only comprises BAM and the triazines will only catch 86% of the contaminated wells and
boreholes. At the same time, glyphosate and AMPA are not identified (the active component and the metabolite from the
common product ROUNDUP). This package will therefore not be able to contribute to securing the future supply of clean
drinking water.
By analysing exclusively for the pesticides that can be identified by using a method like GC/MS or LC/MS, it is possible to
catch approximately 92% of the positive samples. These methods were a few percentage points less successful than the
indicator packages in catching the samples above the threshold value.
The price index shows that up to 40% of the analysis costs can be saved by reducing the number of pesticides from the original
26 to five. The savings amount to over 50% if only the pesticides that can be analysed in one analysis batch are included.
Reduced specifications in relation to the analyses
A change of detection levels from the present 0.01μg/l to either 0.025μg/l or 0.05μg/l provides a reduction of analysis costs of
10-15%. The savings apply to a pesticide package comprising all 26 compounds. If the number of compounds is reduced the
savings will be correspondingly smaller and for the indicator packages comprising five compounds, the savings amount to
5-10%. The fall in costs is due to an easier data treatment process, since the methods, in the first instance, are not changed and
the methods still have to be accredited.
If a detection level of 0.025μg/l had been used in the project Pesticide-contaminated water in small water supply plants,, 15% of
the positive samples would not have been found. A further increase of the detection level to 0.05μg/l would mean that 25% of
the positive samples would not have been found. This excludes samples above the threshold value of 0.1μg/l, but much
information is lost when ¼ of the positive samples is not identified. A detection level of 0.05μg/l will, using the present
specifications in terms of quality, give problems in relation to a threshold level of 0.1μg/l. Doubts will easily be raised regarding
whether a pesticide concentration is above or below the present threshold level. The same disadvantages are not present using
a detection level of 0.025μg/l, but the savings are only 10% when the detection level is changed. The estimated savings are
tentative, and could even be significantly smaller. A detection level of 0.025μg/l is still so low that no significant changes can be
made to the analysis methods in the laboratory. It would definitely give certain problems for the analysis laboratories if the
detection levels for private wells and boreholes were to be increased, but not for standard drinking and groundwater samples.
Conclusion
Three different possibilities for reduction of costs in connection with pesticide analysis in water from private wells and boreholes
have been investigated: immune chemistry, indicator-pesticides, and increased threshold levels.
The most attractive solution for reduction of costs is to limit the pesticide numbers to a few selected indicator-pesticides. It
appears that it is possible to find over 95% of water samples above the threshold levels by reducing the number of pesticides
from the present 26 to five. By increasing the number of pesticides by a few compounds only a few percent of the positive
samples would not be identified. It will be possible to achieve savings of the costs for analysis up to 40% by switching to
selected indicator-pesticides. By compromising further with regard to the number of positive samples, savings could be in the
range of 50-60% with identification of 85-90% of positive samples.
The immune chemistry methods display many advantages, for instance they are quick and relatively cheap. Since the methods
can only be used for one compound at a time, it quickly becomes very expensive to analyse a water sample for several
pesticides. Furthermore, there a virtually no immune chemistry kits that can be used at the detection levels that are required in
connection with pesticide analyses. The commercial immune chemistry kits available only cover approx. 1/3 of the pesticides
that are to be identified in connection with drinking water monitoring. Furthermore, cross-reactions point towards immune
chemistry methods as screening tools for single compounds where positive results are verified using other methods. The
ambiguity in results achieved by using immune chemistry methods combined with the occurrence of pesticides in the uppermost
groundwater magazines will mean that almost all samples will have to be verified by using other methods. Conventional methods
will also most likely be quite competitive in terms of costs, if analyses are made only for a few selected compounds.
An increase of detection levels from the present 0.01μg/l to either 0.025μg/l or 0.05μg/l only provides minimal savings of 10-15%
of costs. An increase in detection levels could have the result that important information, e.g. if pesticide levels are rising or
falling, is being lost. A detection level too close to a threshold level is not optimal and can quickly lead to problems with
verification of potential exceedances. Since the savings are minimal by increasing detection levels it will be most sensible to
maintain a detection level at 1/10 of the threshold levels, where possible.
The purpose of the project is to investigate the possibilities for reducing costs in connection with analyses of pesticides in water
samples from private wells and boreholes.
Three different avenues, immune chemistry, indicator pesticides, and raised detection levels, are investigated. Therefore, the
project is divided into three main sections. In the section about immune chemistry a literature survey of kits for immune
chemistry is provided, together with advantages and disadvantages associated with this approach. Furthermore, prices and
specifications for commercially available immune chemistry kits sold in Denmark or overseas are provided. In appendix D the
theory behind the immune chemistry methods is described.
In the section about indicator pesticides it is investigated whether the number of pesticides can be reduced, thus reducing the
costs without the number of positive samples being reduced. In the last section the possibility for increasing the detection levels
is addressed in terms of the consequences this would have in relation to costs and the actual results of the analyses.
Immune chemistry methods
Immune chemistry methods display, as do conventional methods, a number of advantages and disadvantages. Disadvantages
include cross-reactions and single-compound-identification, but the main advantage is a quick method which requires a
relatively small sample.
The main handicap in connection with immune chemistry methods is cross-reaction, which leads to both overestimation and
false positive results. Each positive result, achieved using immune chemistry methods, will include the risk that the result
originates from another compound than the target compound. Therefore almost all authors who have applied immune chemistry
methods, advise their use as a screening method after which the results should be verified using conventional GC or LC
methods. Due to the general absence of false negatives, their use as a screening tool is evident. Unfortunately, only screening
can only be made for one compound at a time, but some immune chemistry kits, like the traizin kits, are group-specific and will
therefore be useful for identifying several compounds.
It will be possible to assess whether particular kits will be useable as indicators for several compounds by checking the kits
individually. The Danish requirements in relation to detection levels are problematic, as there are only two commercial immune
chemistry kits that provide a detection level which honours the value of 0.01g/l.
It has been shown that the use of more precise equipment can provide a significant lowering of detection levels. A thorough
method-optimisation and control of individual kits will probably indicate that more will actually be able to honour the required
detection levels. By using immune chemistry methods there is a clear possibility for increased sample flow due to the absence of
cumbersome pre-treatment and very small sample amounts. There is practically no problem waste involved, and the equipment
requires minimal service. In order to obtain consistent and precise results, experienced and well-trained personnel is required,
especially in connection with samples with low concentrations. This is not always clearly stated in the advertisements for immune
chemistry methods.
Expenses in connection with the use of commercial immune chemistry kits can be divided into the purchase of the kits,
equipment, and salaries. By recalculating the purchase price into a price per sample (see Appendix A) it can be seen that the
costs for the materials range from just below 100 DKK/sample and above, which is roughly equivalent to the price when using
conventional methods.
The costs for the equipment will be significantly smaller and is evaluated at roughly 1/10 of the corresponding costs using
conventional methods. The biggest difference probably lies in the fact that more samples can be analysed per time unit than
using conventional methods, and therefore salary costs can be kept low. The problem is that only one compound can be
identified at a time. If all the compounds specified in the Statutory Order on Drinking Water are to be identified it will require
the use of over 20 different immune chemistry kits, which probably would result in a flow that was worse than when using
conventional methods. Usage of immune chemistry methods will at the moment only be economically advantageous if one or
two compounds are to be identified. In the longer term, conventional methods will most likely be able to carry out the analyses
at a similar cost if this involves one compound.
Indicator-pesticides
An analysis of the results of the project Pesticide-contaminated water in small water supply plants ("pesticidforurenet vand i
små vandforsyningsanlæg") shows that a reduction in the analysis program would not have had a major influence on the number
of positive finds of pesticides.
By reducing the analysis program from the original 26 to an indicator package comprising only five different pesticides, it would
have been possible to find 91.7% of the positive samples.
In the table below the most interesting packages are displayed.
|
Number |
AMPA |
Positive fund i % |
|
Indicator package |
Pesticides |
Included |
all |
≥0,1μg/l |
Price index |
All 26 pesticides |
26 |
X |
100 |
100 |
100 |
|
|
|
|
|
|
All compounds minus glyphosat and AMPA |
24 |
|
93.5 |
93.4 |
69.9 |
Only glyphosat and AMPA |
2 |
X |
21.3 |
11.7 |
30.1 |
|
|
|
|
|
|
Indicator package 1 |
5 |
X |
91.7 |
95.9 |
60.7 |
Indicator package 2 |
5 |
X |
91.4 |
97.0 |
60.7 |
Indicator package 3 |
6 |
X |
93.5 |
97.5 |
61.4 |
Indicator package 4 |
10 |
X |
96.7 |
99.0 |
63.6 |
|
|
|
|
|
|
Indicator package GC/MS |
18 |
|
92.0 |
92.4 |
40.8 |
Indicator package LC/MS |
13 |
|
89.9 |
92.9 |
37.2 |
|
|
|
|
|
|
The five most common |
5 |
|
84.0 |
89.3 |
31.3 |
Only BAM |
1 |
|
68.9 |
79.7 |
28.4 |
Table. Overview of different indicator packages with specification of effectiveness in spotting positive samples, and samples
above the threshold value of 0.1μg/l. A price index for individual packages is also provided.
By including several of the triazin-pesticides in the indicator package it is possible to increase the number of positive finds to
96.7% of the original number. More importantly, only a small percentage of the samples above the threshold of 0.1μg/l is not
identified.
A reduced pesticide package consisting of 10 indicator-pesticides (indicator package 4): 2,6-dichlorbenzamide (BAM),
desisopropylatrazine, desethylatrazine, atrazine, simazine, desethylterbutylazine, terbutylazine, diuron, bentazone, and AMPA
will be able to "catch" 97% of all the expected positive samples and, out of these samples, 99% of the samples where the
pesticide concentration exceeds the present threshold value of 0.1μg/l. The costs associated with this package using LC-GC/MS
will be approximately 60% of the costs associated with the analysis of all 26 pesticides.
An indicator package which only comprises BAM and the triazines will only catch 86% of the contaminated wells and
boreholes. At the same time, glyphosate and AMPA are not identified (the active component and the metabolite from the
common product ROUNDUP). This package will therefore not be able to contribute to securing the future supply of clean
drinking water.
By analysing exclusively for the pesticides that can be identified by using a method like GC/MS or LC/MS, it is possible to
catch approximately 92% of the positive samples. These methods were a few percentage points less successful than the
indicator packages in catching the samples above the threshold value.
The price index shows that up to 40% of the analysis costs can be saved by reducing the number of pesticides from the original
26 to five. The savings amount to over 50% if only the pesticides that can be analysed in one analysis batch are included.
Reduced specifications in relation to the analyses
A change of detection levels from the present 0.01μg/l to either 0.025μg/l or 0.05μg/l provides a reduction of analysis costs of
10-15%. The savings apply to a pesticide package comprising all 26 compounds. If the number of compounds is reduced the
savings will be correspondingly smaller and for the indicator packages comprising five compounds, the savings amount to
5-10%. The fall in costs is due to an easier data treatment process, since the methods, in the first instance, are not changed and
the methods still have to be accredited.
If a detection level of 0.025μg/l had been used in the project Pesticide-contaminated water in small water supply plants,, 15% of
the positive samples would not have been found. A further increase of the detection level to 0.05μg/l would mean that 25% of
the positive samples would not have been found. This excludes samples above the threshold value of 0.1μg/l, but much
information is lost when ¼ of the positive samples is not identified. A detection level of 0.05μg/l will, using the present
specifications in terms of quality, give problems in relation to a threshold level of 0.1μg/l. Doubts will easily be raised regarding
whether a pesticide concentration is above or below the present threshold level. The same disadvantages are not present using
a detection level of 0.025μg/l, but the savings are only 10% when the detection level is changed. The estimated savings are
tentative, and could even be significantly smaller. A detection level of 0.025μg/l is still so low that no significant changes can be
made to the analysis methods in the laboratory. It would definitely give certain problems for the analysis laboratories if the
detection levels for private wells and boreholes were to be increased, but not for standard drinking and groundwater samples.
Conclusion
Three different possibilities for reduction of costs in connection with pesticide analysis in water from private wells and boreholes
have been investigated: immune chemistry, indicator-pesticides, and increased threshold levels.
The most attractive solution for reduction of costs is to limit the pesticide numbers to a few selected indicator-pesticides. It
appears that it is possible to find over 95% of water samples above the threshold levels by reducing the number of pesticides
from the present 26 to five. By increasing the number of pesticides by a few compounds only a few percent of the positive
samples would not be identified. It will be possible to achieve savings of the costs for analysis up to 40% by switching to
selected indicator-pesticides. By compromising further with regard to the number of positive samples, savings could be in the
range of 50-60% with identification of 85-90% of positive samples.
The immune chemistry methods display many advantages, for instance they are quick and relatively cheap. Since the methods
can only be used for one compound at a time, it quickly becomes very expensive to analyse a water sample for several
pesticides. Furthermore, there a virtually no immune chemistry kits that can be used at the detection levels that are required in
connection with pesticide analyses. The commercial immune chemistry kits available only cover approx. 1/3 of the pesticides
that are to be identified in connection with drinking water monitoring. Furthermore, cross-reactions point towards immune
chemistry methods as screening tools for single compounds where positive results are verified using other methods. The
ambiguity in results achieved by using immune chemistry methods combined with the occurrence of pesticides in the uppermost
groundwater magazines will mean that almost all samples will have to be verified by using other methods. Conventional methods
will also most likely be quite competitive in terms of costs, if analyses are made only for a few selected compounds.
An increase of detection levels from the present 0.01μg/l to either 0.025μg/l or 0.05μg/l only provides minimal savings of 10-15%
of costs. An increase in detection levels could have the result that important information, e.g. if pesticide levels are rising or
falling, is being lost. A detection level too close to a threshold level is not optimal and can quickly lead to problems with
verification of potential exceedances. Since the savings are minimal by increasing detection levels it will be most sensible to
maintain a detection level at 1/10 of the threshold levels, where possible.
| Forside | | Indhold | | Forrige | | Næste | | Top |
Version 1.0 Juli 2004, © Miljøstyrelsen.
|