Miljøprojekt, 912; Teknologiudviklingsprogrammet for jord- og grundvandsforurening – Diffus jordforurening og kulturlag

Diffuse Soil Pollution and Historic Fill

Summary and Conclusions

This data report presents the physical trials of an investigation strategy to map diffuse soil pollution in urban areas (historic fill) at test sites in Copenhagen and Ringsted. The data report is one of seven reports prepared in connection with a project initiated and supervised by the Agency of Environmental Protection in the City of Copenhagen under the Danish EPA Technology Program. The overall objective is to prepare methods to optimize and simplify technical investigations by the environmental authorities in connection with mapping of diffuse soil pollution at the legislative Knowledge Level 2.

Investigation of historic fill (contribution model)
The investigation strategy for historic fill in urban areas is tested on 10 different housing areas in Copenhagen and in Ringsted, a provincial town. Samples are also taken from a reference area (recreational area) in the old town in Copenhagen. No samples are taken in areas where there is information or suspicion about point sources, filled areas or waste landfills.

The term historic fill is used to describe the soil that has been affected by anthropogenic activities, whereby small random contributions of pollutants and materials are spread to the historic fill over decades or centuries. The diffuse soil pollution is expected to constitute a variable and random load in the top soil that is assumed to decrease with depth.

The aim of the project is to investigate the soil contaminant levels in housing areas of different age as well as to improve the basis for characterisation of diffuse soil pollution in urban areas.

Characterisation of diffuse soil pollution
The conceptual model
Generally, it is concluded that the pollution patterns in the 10 test areas are well described by a contribution model,

Typically, a random contaminant pattern is seen in areas with a long history of development and with many changes of land use including housing. In this type of area, many smaller areas with low or high contaminant levels are seen. These areas are assumed to represent point sources of pollution or areas where the soil has been excavated.

Heavy metals, oil and PAH
Soil quality criteria (JKK) for lead and sum of polycyclic aromatic hydrocarbons (PAH) including benzo(a)pyrene (BaP) are exceeded in the top soil layer (0 - 0.3 m) In old city areas, soil intervention levels are exceeded for lead and BaP. The contaminant patterns for lead and BaP differ, and most probably the pollution originates from different sources. On some test areas, the concentration levels for total hydrocarbons (oil), cadmium, mercury and dibenz(ah)anthracen (DiBahA) exceed the JKK criteria. In urban areas, elevated concentrations of copper and zinc are noted in comparison with background levels in rural areas.

Generally, the concentration levels of arsenic, nickel and chromium are not elevated.

PCB
Polychlorinated biphenyls (PCB) are generally not found in soil samples from areas with urban housing (PCB is not detected in 90% of the samples). There is no Danish soil criterion for PCB, but the criterion in the Netherlands is 20 µg/kg DW for seven PCB congeners.

Phthalates
Low contents of phthalates are measured, but levels are greatly below the Danish soil quality criterion (JKK) of 250 mg/kg DW (phthalates are not detected in 30% of the samples, and only one sample had a content of more than 1/250 of JKK).

Dioxins
All the analysed soil samples had a low content of dioxins (1 - 20 ng international toxic equivalents (ITE)/kg DW). There is no Danish soil criterion, and therefore the toxicological aspects cannot be assessed. However, in Germany, the background levels in rural and urban areas are 1 - 5 ng ITE/kg DW and 10 – 30 ng ITE/kg DW respectively.

Pesticides
The pesticide content (only persistent pesticides) was analysed in 10 soil samples, but only low levels of DDT and parathion were found in four of the 10 samples. Soil quality criteria JKK (0.5 and 0.1 mg/kg DW respectively) are not exceeded.

Housing age
To illustrate the effect of housing age, the median values for all pollutant parameters from the 10 test areas as well as from former investigations are compared. It is important to note that the individual measurements are below or above the median values, and a large range of measured values was noted.

The contaminant levels for lead, copper, mercury, zinc and BaP are significantly higher for areas of greater age and a long history of development.

Difference between Copenhagen and Ringsted
Contamination with heavy metals (lead, copper, mercury and zinc) are generally lower in Ringsted than in Copenhagen, and only few soil samples exceed JKK in areas where housing was established after 1940 in Ringsted. In the oldest housing areas in Copenhagen established before 1900, the median values for lead exceed soil intervention levels (ASK). This is not the case in Ringsted.

BaP levels are, however, high in the oldest housing areas in both Copenhagen and in Ringsted. Furthermore, the median value for BaP in housing areas in Copenhagen and Ringsted established before 1920 often exceed the soil intervention limits (ASK). JKK is exceeded in all housing areas in Copenhagen, but the median value is not exceeded in housing areas established after 1950 in Ringsted.

Depth in the soil profile
In Copenhagen, the lead contaminated historic fill is found to a depth of at least 1 m, except in newer housing areas (established after 1915). In Ringsted, the lead contaminated historic fill is only found to a depth of about 50 cm in the older housing areas (established before 1920).

In Copenhagen and Ringsted, the BaP contaminated historic fill is found to a depth of at least 0.5 -1 m in the oldest housing areas (established before 1950), but is typically found to a lower depth in the newer housing areas (established after 1950).

Experience with application of the investigation strategies
Sampling and size of area
The data treatment has indicated that a certain minimum of results from individual positions within an area with a common pollution pattern is necessary if a geostatistical analysis is to be performed.

Based on experience from the investigations, the range for the geostatistical analysis in urban areas is typically of a magnitude of 200 – 400 m.

The number of data pairs and the range of variation between them is important for the data analysis, and not necessarily the size of the area – however, the area must have a size at least 0.2 km². More data are necessary if the "noise level" - the variance (nugget) between positions close to each other - is high. The area must be large enough to avoid boundary effects and the data density sufficient in relation to distance (range) between the correlated data pairs.

At least 40 sampling position from an area of 0.2 km² should be applied, corresponding to a sampling density of 200 sampling positions/ km².

Composite or single samples
For a geostatistical experimental plan, the use of composite samples is not recommended, since the data treatment requires measurements of samples taken close to each other. It is assessed that the advantages of composite samples (which reduce variation in the sample point) do not equate with the work involved and the risk of inadequate mixing.

Field measurements
It is more advantageous to use laboratory analyses such as ICP or EDXRF instead of field measurements such as x-ray fluorescence (EDXRF). Analysis by ICP or AAS gives lower detection limits and ensures comparison with other investigations (no bias). Furthermore, a better analytical accuracy can most probably be achieved, as it is assessed that EDXRF is very sensitive to micro heterogeneity of pollutants in the soil media.

Analytical parameters
The important pollutants in diffuse soil pollution in urban areas are lead and benzo(a)pyrene (BaP). Supplementary analyses for other heavy metals (cadmium, copper, and zinc) and sum of PAH will lead to a better description and documentation that the contaminant levels across an area lie within the expected range and are uniform.

Data treatment
As part of the project, various forms of data presentation have been assessed with a view to providing a visual impression of contaminant levels as well as to comparing data from different areas. If more than seven data values are to be analysed, it is recommended that the following parameters are calculated and presented:

number of data
minimum, median, maximum
quantiles (e.g. 10, 25, 50 75 and 90%)
average

Geostatistics has been used to assess the spatial tendencies and to estimate concentration levels and uncertainty for the estimate across an area.

Geostatistics is based on the assessment that measurements made close to each other are more alike than measurements made at a distance. Measurements are said to be correlated with in a certain distance.

Similarly, the probability that soil at a given position in the area exceeds the JKK can be calculated. Furthermore, the applied geostatistical data treatment can be used to assess the probability of exceeding any desired concentration limit, e.g. the soil intervention limit or another limit value.

Not all areas are suitable for a geostatistical analysis. Areas without a common historical link, i.e. different pollutants loads and large heterogeneity, require more detailed investigation, possibly at the level of the individual land plots. The general experience with the use of geostatistics in the 10 housing areas is summarised in a flow diagram showing the individual steps in the data treatment.