Drinking Water PTs
Recommendation of the Federal Environmental Agency
The German Federal Environmental Agency (UBA) has published "Recommendations for the execution of proficiency tests in the drinking water sector ( German text courtesy of Springer-Verlag) in the Bundesgesundheitsblatt . For the chemical sector this publication was in November 2003. Amongst other things the following key points are given (shortened here):
- The authorities responsible according to §15 clause 4 of the German Drinking Water Ordinance have to check during the review of compliance with the requirements of clause 4, whether the drinking water laboratories in their state are regularly and successfully participating in proficiency tests.
- The PTs have to be provided by competent PT providers that fufill the requirement auf DIN 38402-45:2003 "Ringversuche zur externen Qualitätskontrolle von Laboratorien (A 45)".
- The PTs for quality and proficiency check of drinking water laboratories have to fulfill the requirements of DIN 38402-45 with regard to planning, execution and evaluation.
- If possible the PT samples should be based on real drinking water sample that might be spiked with the relevant analyte
- Because the German Drinking Water Ordinance does not prescribe analytical methods for the parameters relevant for this recommendation the PT provider also should not prescribe methods in the PT.
- All chemical and indicator parameters contained in the German Drinking Water Ordinance 2001 that can be checked in PTs should be provided by one or more PT providers within a time period of 2 to 3 years (PT cycle)
- It is required that all drinking water laboratories provide evidence for successful participation within one PT cycle (2-3 years) for all parameters they are or want to be accredited for drinking water examination. If a laboratory fails in a PT for a parameter, it has to repeat this PT as soon as possible, if necessary with another PT provider.
- Evaluation of raw data of the drinking water PT has to be made according to the requirements of clause 10 of DIN 38402-45. If possible the evaluation procedure should cover and display all concentration levels distributed in the PT all together in a common way regarding to recovery and dispersion.
- The primary criterion for the assessment of the performance of the drinking water laboratory should be the so-called zU-score.
- Acceptable result: |zU| ≤ 2
- Successful determined parameter: Based on the assessment and the exclusion criterion |zU| ≤ 2 more than half of the results (concentration levels) should be analysed correctly.
- There is only an assessment for single parameters, no overall assessment for the total PT.
- Participants should receive a dated and meaningful certificate, which should contain the successful as well as the non-successful analysed parameters.
Cooperation with other PT providers
To facilitate compliance with these requirements of the UBA recommendation for the laboratories the PT providers working in
this field decided to provides the PTs in an harmonised way.
Now there are two PT schemes operating in an adjusted way. One of these schemes is located in North Rhine-Westphalia ( Landesamt für Natur, Umwelt und Verbrauchgerschutz Nordrhein-Westfalen in cooperation with IWW Water Centre), the other is operated by AQS Baden-Württemberg in cooperation with Institut für Hygiene und Umwelt - Behörde für Soziales, Familie, Gesundheit und Verbraucherschutz (BSG) der Freien und Hansestadt Hamburg.
The parameters to be analysed according to the German Drinking Water Ordinance are distributed between 10 different PT rounds:
- A 1. Anions: bromate, fluoride, nitrate, nitrite, chloride, sulfate, phosphorous, cyanide, turbidity
- A 2. Trace elements: chromium, copper, lead, cadmium, nickel, chromium(VI), silicate
- A 3. Cations, part 1: aluminium, iron, manganese, sodium, potassium, color quantitative (SACK436)
- A 4. Other inorganic parameters: conductivity, oxidisability, TOC, selenium, antimony, arsenic
- A 5. Cations, part 2: boron, calcium, magnesium, mercury, ammonium, pH-value
- O 1. Pesticides 1: N- and P-pesticides (triazines, phenylurea herbicides)
- O 2. VOC 1,2-dichloroethane, tetrachloroethene, trichloroethene trichloromethane, bromodichloromethane, dibromochloromethane, tribromomethane, benzene
- O 3. PAH: benzo(a)pyrene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(ghi)perylene, indeno(1,2,3-cd)pyrene
- O 4. Pesticides 2: phenoxyalkane carboxylic acids
- O 5. Special organic parameters: e.g. glyphosate/AMPA, dalapon/TCA, chloropesticides, epichlorohydrin or others
Both PT schemes provide the complete PT programme shifted in time for 1 year. Each scheme provides 4 PT round per year, so that the complete programme is finished within 2.5 years. For the AQS-BW/BSG-scheme it was agreed that AQS-BW provides 3 PT rounds per year and BSG one. With this procedure it is possible for the laboratories to participate in one of the schemes and therewith to fulfill the requirements of the UBA recommendation. If they fail with a parameter in one PT round they can repeat this in the other scheme after 1.5 years at the latest. On the other side it is also possible to finish the whole programme within 1.5 years. This is especially favourable for laboratories that are new on the market.
An overview on the harmonisation of the drinking water PTs in Germany is given in an flyer (in German only).
Evaluation and assessment
The evaluation and assessment of these PTs fulfill the requirements of DIN 38402 - A45 and the recommendation of UBA.
The assigned value usually is defined on the basis of the robust mean of all participants' results. In particular cases the PT coordinator may use reference values from weighings as assigned values. In this case the content of analyte in the matrix used has to be determined from the results of participants or from measurement results of reference laboratories and has to be considered when calculating the assigned value. The standard deviation necessary to calculate the zU-scores are usually also calculated from participants' results. Whenever possible a common evaluation over all concentration levels with the variance function described in DIN 38402 - A45 is used. The coordinator defines upper and lower limits for the standard deviation to prevent quality requirements from being too harsch or to weak.
After finishing the PT round all participating laboratories get an evaluation data sheet where all results are marked with "+" or "-" to show the assessment of this value. In addition for each parameter it is documented whether this parameter was analysed "successful" or "not successful". "Successful" means that more then 50% of the results for this parameter were within the tolerance limits. Parameters that were not analysed are marked with "not participated".
In addition the laboratories get a certificate for participating in this PT round, which also states the parameters in the PT round and also the assessment as "successful", "not successful" or "not participated". Of course all participants also get a comprehensive evaluation report, stating the results of all participating laboratories in an anonymuous way together with many tables and figures.
These reports are also available from the AQS-BW website.