Summary and conclusions, Miljøstyrelsen

Kortlægning af diffus jordforurening i byområder. Delrapport 2

Summary and conclusions

This report concerning practical trials of field screening techniques for the investigation of diffuse soil pollution is part 2 of a three-part report prepared under Phase I of a project on mapping of diffuse soil pollution in urban areas.

The project is instigated and supervised by Agency of Environmental Protection in Copenhagen under the Danish Environmental Protection Agency's technology development program. The overall objective is to prepare methods to optimise and simplify technical investigations by the environment authorities in connection with mapping of diffuse soil pollution at the legislative information level 2. Phase I has included a review of information concerning the sources of pollution, measurements from actual investigations, analytical and sampling techniques and investigation strategies

The objectives for the part 2 report are;

To clarify whether field-screening techniques can be used for investigations of diffuse soil contamination in connection with mapping of soil contamination at the legislative information level 2.

To clarify if field-screening techniques can be used to screen soil samples or to replace laboratory analyses for the documentation of the pollution levels.

Furthermore, the objective is to assess if field-screening techniques can improve the investigation strategy

Heavy metals and organics such as oil, PAH, PCB are identified in the part 1 report as relevant contaminants. Assessment of a range of field screening techniques with respect to measurement parameters, properties, detection limits and economic estimates for analysis of samples identified the following two field screening techniques as of interest for the practical trials.

EDXRF (Energy Dispersive X-ray Fluorescence) for metals.

Rapid Immunoassay for PAH and PCB.

The practical trials for these two field screening techniques involved respectively 9 and 24 soil samples from a depth of 10 cm from two test areas each comprising 100 m² within the city of Copenhagen. The two areas represent urban soils, but have not been exposed to known sources of pollution or to construction activities. On the basis of a former investigation, the two areas are defined as:

an area with an expected high level of diffuse pollution at the Center for Tegnsprog, formerly Blindeinstituttet, Kastelsvej 60, 2100 Copenhagen; a site that was developed in 1906

an area with an expected low level of soil pollution in Fælled Park, which was established in 1908 –1912 and which was formerly grazing pastures for cattle

The practical trials were designed to answer six questions:

What is the variation in soil sample concentration across the diffuse contaminated test areas due to the variable composition of contaminants in the soil?

Do composite soil samples give the same description of the soil contamination as measured in the individual soil samples?

Does pre-treatment of the soil samples affect the result as interpreted by comparison of the variation in soil sample composition for identical samples?

Does the extraction conditions for the immunoassay affect the analytical result (based on variation in soil sample composition for identical samples)?

How good is the analytical quality for the field screening techniques compared to the analytical quality achieved by the accredited laboratory methods?

How is the % recovery for certified soil standards when measured by the field screening techniques?

Ad 1. It is concluded that the contamination is uniform (homogenous), but different for the two test areas. The soil quality criteria are only exceeded for the content of lead and PAH. The pollutant levels for lead, copper and PAH are generally higher at the Center for Tegnsprog than at Fælled Park, while at Fælled Park, a higher chromium content is found in the soil samples. The EDXRF measurements of copper and zinc at the Center for Tegnsprog show more variation around the average (60 – 100%) than for the other metals (10- 30%) and compared to measurements at Fælled Park. The immunoassay measurements for PAH show large variation around the average of 80– 100%.

Ad 2. It is concluded that for the two test areas, which have uniform contaminant levels, there is no difference in the average content or in the variation coefficients for measurements on the individual samples as compared to the composite samples (each representing three individual samples).

Ad 3. It is concluded that pre-treatment of the soil samples by comminution, drying, crushing and sieving is inexpedient in that this has a significant effect on the results as compared to the laboratory results. The EDXRF measurements show significantly higher levels of metals if the analysis is performed on the fine soil fraction generated by the pre-treatment and sieving.

Ad 4. It is concluded that the longer extraction times and warming of the extract give an increased recovery of PAH from the soils.

Ad 5. It is concluded that for the actual investigation, the analytical results with EDXRF for lead, nickel, and chromium are directly comparable with the accredited laboratory results. The analytical results with EDXRF for zinc and copper are less accurate at the actual concentration levels, but indicate the concentration level. The analytical results with EDXRF for arsenic showed that the test areas had an arsenic level below the detection limit for the EDXRF method (<20 mg/kg), which was correct (actual level was 7-10 mg kg TS). According to the statistical tests, the analytical results for immunoassay for PAH and PCB were not found to be in agreement with the laboratory measurements, but they indicated the contaminant concentration level. The analytical quality for immunoassay for PCB is unreliable for the actual measurement interval for the practical tests (0,05- 1 mg/kg).

Ad 6. It is concluded that the concentrations for certified samples as measured by the field screening EDXRF technique can only be statistically shown to be identical with the true value for high concentrations of lead and zinc at 350 and 550 mg/kg respectively. For all other parameters measured by EDXRF and immunoassay, no statistical agreement can be demonstrated, but the measurements indicate the concentration level. The validation data (standard deviation, detection limits, precision and accuracy) for the EDXRF measurements for lead, chromium, copper, nickel, and zinc is calculated.

It is concluded that EDXRF can be used as an alternative to laboratory analysis to measure the content of zinc and lead in diffuse contaminated soils and is useful for screening of soil samples for arsenic, copper, nickel and chromium.

It is concluded that immunoassay can be used for screening of the general PAH levels in the soil, but that an individual result can not be attributed great value without confirmation by GC-MS-SIM analysis.

It is concluded that immunoassay for PCB is unreliable at low concentrations. The method can however be used to assess whether there is a PCB contamination present in the soil, and if present to allow identification of representative samples to be sent to analysis for PCB. It is assessed that the test kit can not be used to map the concentration levels in connection with diffuse soil contamination, since these levels are very low, but can be used to qualitatively identify a PCB problem.