Human Health Risk Assessment - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 227

independent of the site, region or country, the exceptions being sorption character-
istics (which could be expressed as a function of soil properties) and plant uptake
factors. Physiological factors such as body weight and skin surface area vary only
slightly between countries and general data often can be used in Risk Assessment.
Since there is no general guidance for dealing with foreign data, a sensible trade-
off must be made between the relevance of the data for the Human Health Risk
Assessment (e.g., in regard to climate and geographical characteristics for soil-
related input parameters or cultural differences for human behaviour-related input
parameters) and the number of data points available. In case of lack of data, it is
more attractive to include foreign data, even when they have been measured under
different conditions.
Exposure characteristics can be found in several documents, for example:

• Exposure Factors Handbook (US Environmental Protection Agency 1997 );

• ECETOC 2001 , with focus on UK data;

• Otte et al. ( 2001 );

• Child-specific Exposure Factors Handbook (US Environmental Protection

Agency 2008 );

• ExpoFacts database (Vuori et al. 2006 );

• Van Holderbeke et al. ( 2008 ).

When choices are made for input parameters for the model, whether it is at the
point of developing the model or during site-specific assessments, it is advisable to
report background information for the choice made such as the origin of the data,
the range of the data and the representativeness.

5.3.3.9 Reliability

Since the mid-1990s, several software packages have become available which
stimulate the wide use of exposure models. As a consequence, human expo-
sure models are in widespread use, both implicitly and explicitly. An example of
implicit use is the comparison of measured contaminant concentrations with Soil
Quality Standards derived from these exposure models (e.g., Swartjes ( 1999 )for
the derivation of Soil Quality Standards in the Netherlands; Provoost et al. ( 2004a )
in Flanders, Belgium; and Environment Agency and DEFRA ( 2002 ) in the UK).
Explicit use is decision-making based on site-specific exposure calculations. Insight
into the reliability of calculated exposure could be most directly addressed by per-
forming a validation study, that is, comparing calculated exposure with measured
exposure. However, measuring exposure in human beings is difficult or even pre-
cluded for technical and ethical reasons. It is possible to validate the calculation
of concentrations in contact media such as vegetables and indoor air. From these
validation studies it can be concluded that the reliability of these calculations and,
hence, of these exposure models is generally limited, because of uncertainties about
model algorithms and input parameters. Reliability decreases if contaminants are
more mobile and even more so when contaminants are more volatile. The reason for