Ktm.hu

Promulgation of Methods and Standards for Derivation of Test Thresholds and
Measures Thresholds pursuant to the Federal Ordinance on Soil Protection and
Contaminated Sites (Bundes-Bodenschutz- und Altlastenverordnung
(BBodSchV))
Promulgation of Methods and Standards for Derivation of Test Thresholds and
Measures Thresholds pursuant to the Federal Ordinance on Soil Protection and
Contaminated Sites (Bundes-Bodenschutz- und Altlastenverordnung
(BBodSchV))
Pursuant to Art. 4 (5) 2nd sentence of the Federal Ordinance on Soil Protection and Contaminated Sites (Bundes-Bodenschutz- und Altlastenverordnung (BBodSchV)), the methods and standards for derivation of the test thresholds and measures thresholds pursuant to Art. 8 of the Federal Soil Protection Act are promulgated. The sole purpose of such promulgation is to ensure proper execution of the Federal Soil Protection Act and of the Federal Ordinance on Soil Protection and Contaminated Sites. The methods and standards, and their bases, are not necessarily directly applicable to other legal areas in the field of environmental and health protection. The methods and standards for derivation ˉ support proper and standardized application of the test thresholds and measures thresholds of the Federal Ordinance on Soil Protection and Contaminated Sites in individual cases and ˉ ensure that equally reliable decisions are made, in individual cases, regarding substances for which the Federal Ordinance on Soil Protection and Contaminated Sites contains no test thresholds or measures thresholds. Departures from these methods and standards are permitted only as reliable new scientific findings become available. The derivation standards shall be adapted to new scientific findings at the appropriate time. Bonn, 18 June 1999 Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU) Contents
1 Background . 8 Test and measures thresholds, pursuant to the Federal Ordinance on Soil Protection and Contaminated Sites, for children's play areas, residential areas, parks and recreational facilities and for industrial and commercial properties 10 2.1 Differentiation Measures thresholds of Annex 2 Number 1.1 for direct resorption of dioxins/furanes on children's play areas, residential areas, parks and recreational areas and industrial and commercial properties . 12 Assessment standards from human toxicology . 13 2.3.1 Derivation of assessment standards from human toxicology, within the framework of Art. 8 Federal Soil Protection Act and Art. 4 Federal Ordinance on Soil Protection and Contaminated Sites . 14 2.3.1.1 Definition . 14 2.3.1.2 Database. 15 2.3.1.3 Protection Criteria for adverse effects . 17 Assessment scale for carcinogenic substances. 18 Assumptions regarding resorption . 19 Extrapolation, use of safety factors . 20 Calculation of body doses (LOAEL, NOAEL), from animal-experiment 2.3.1.10 Procedure for assessment standards for carcinogenic substances . 25 2.3.1.11 The issue of higher sensitivity of children to carcinogenic substances . 28 2.3.1.12 Validation of TRD derivations. 31 2.3.2 Hazards reference. 31 2.3.2.1 Meaning of a hazards-oriented dose. 31 2.3.2.2 Substances with effects threshold. 31 2.3.2.3 Substances with no effects threshold. 34 2.3.3 Background exposure/usage quota. 35 2.3.4 List of available human-toxicological assessment standards and justifications Excepted types of exposure within the framework of derivation of test 2.4.1 An overview of usage scenarios. 39 2.4.1.1 Ingestion of soil . 40 2.4.1.1.1 Exposure factors . 40 2.4.1.1.2 Direct soil ingestion in connection with carcinogenic substances . 43 2.4.1.1.3 Formulae for calculating soil ingestion . 43 2.4.1.2 Inhalative soil intake for the scenarios "children's play areas", "residential areas" and "parks and recreational facilities" . 46 2.4.1.2.1 Exposure factors . 46 2.4.1.2.2 Calculation formulae . 47 2.4.1.3 Inhalative soil intake in the "industrial and commercial facilities" scenario49 2.4.1.3.1 Exposure factors . 49 2.4.1.3.2 Calculation formulae . 52 2.4.1.4 Dermal soil contact and percutaneous resorption . 54 2.4.1.4.1 Database and method. 54 2.4.1.4.2 Additional factors that influence percutaneous resorption. 59 2.4.1.4.2.1 Influence of the thickness of the soil layer on the skin . 59 2.4.1.4.2.2 Influence of the exposure time . 60 2.4.1.4.2.3 Soil properties . 60 2.4.1.4.2.4 Animal model . 60 2.4.1.4.3 Consequences for consideration of percutaneous resorption in derivation of test thresholds. 61 2.4.1.4.4 Procedure for consideration of percutaneous resorption of pentachlorophenol . 61 2.4.1.5 Consideration of one-time high resorption of substances with high acute Criteria for plausibility consideration of calculations in derivation of test Test and measures thresholds pursuant to Annex 2 number 2 Federal Ordinance on Soil Protection and Contaminated Sites for soils used for cultivation, horticulture, home vegetable/fruit gardens and grassland . 67 3.2 Differentiation orientation . 69 3.4 Procedure. 70 Determination of the highest permissible heavy-metal concentrations in plants Derived maximum permissible pollutant concentrations in plants . 72 Soil/plant heavy-metal transfer. 73 4 Literature . 78 List of tables
Table 1: Overview of the safety factors relevant for TRD values . 21 Table 2: Species-specific parameters for calculation of body doses. 23 Table 3: Noxa with epidemiological or experimental indications of higher sensitivity in infantile organisms (sources listed in Schneider, 1999) . 29 Table 4: Substances for which human-toxicological assessment standards (TRD values) are available . 37 Table 5: Exposure pathways, considered in connection with usages . 39 Table 6: Results of in-vivo animal experiments studying percutaneous resorption of soil-bound pollutants (the skin-water permeability coefficients are also listed for each substance) . 57 Table 7: Classification of substances in accordance with McKone's criteria, on the basis of their physio-chemical properties . 58 Table 8: Sample guidelines for pollutants in plant foods (BGVV, 1997) – here, for lead and cadmium – expressed in mg/kg fresh weight in the form sold (edible parts) . 71 Table 9: Maximum permissible levels for Cd and Pb pursuant to the Feedstuffs Ordinance (FMVO, 1992), supplemented by VDI guidelines for feeds (VDI, 1991, 1992), expressed in mg/kg feedstuffs with 88% dry substance. 72 Table 10: Maximum permissible pollutant content of plants [mg/kg dry weight] used for derivation of soil test thresholds, for Cd and Pb (as examples), calculated from 2 times the ZEBS standards and 1 times the FMVO standards, and converted to dry weight, taking the water content [WC %] of edible parts into account (nutrition tables, Souci et al., 1986) and using 12% WC for grassland cover: . 73 Table 11: Results of analysis of the TRANSFER database for farm cultivation, commercial vegetable cultivation, small/home gardens; the soil figures calculated for Cd and Pb are in µg/kg; AN = ammonium-nitrate extract; KW = aqua regia extract; B = certainty measure; summarized from Knoche et Table 12: Results of analysis of the TRANSFER database for grassland and feedstuffs cultivation; the soil figures for Cd and Pb are in mg/kg; grassland cover includes 3% addition to take soiling into account; KW = aqua regia extract; B = certainty measure; summarized from Knoche et al. 1 Background
The Federal Ordinance on Soil Protection and Contaminated Sites defines test thresholds and measures thresholds pursuant to Art. 8 (1) Federal Soil Protection Act. The fol owing definitions apply: Test thresholds: Values which, when exceeded, mean that case-oriented testing must be carried out, taking into account the type of soil use in question, to determine whether a harmful soil change has occurred or the site is contaminated; Measures thresholds: Impacts or pollution values which, when exceeded, normally mean, taking into account the type of soil use in question, that a harmful soil change has occurred or the site is contaminated and that relevant measures are In Art. 4, the Federal Ordinance on Soil Protection and Contaminated Sites mandates that findings of studies carried out pursuant to this Ordinance, taking the circumstances of the specific case into account, must, in particular, be assessed on the basis of test and measures thresholds. Where the Federal Ordinance on Soil Protection and Contaminated Sites specifies no test or measures thresholds for a given pollutant, the methods and standards used in Annex 2 of the Federal Ordinance on Soil Protection and Contaminated Sites to derive the relevant values are to be used to evaluate study findings in specific cases. Derivation of test and measures thresholds is oriented to Art. 8 (1) Federal Soil Protection Act and to its reference to fulfillment of obligations, set forth in Art. 4, for prevention of hazards in connection with existing harmful soil changes or contaminated sites. Such derivation begins largely with soil functions, and their significance for • people who come into direct contact with the soil, • maintenance of the purity of food and feed plants, and • water that leaches into the soil and is destined to become groundwater. The assets that must be protected include human health, the quality of food plants and feeds, and water that leaches into the soil and is destined to become groundwater. These assets are differentiated in a specific way in derivation of test and measures thresholds. This does not rule out the need, in individual cases of harmful soil changes or contaminated sites, and taking all relevant circumstances into account, for assessing additional assets, such as the soils' habitat function; relevant methods and standards are still being developed. To permit integration of relevant enforcement experience of the Länder, work within the framework of the Federal/Länder working group on soil protection (LABO), the Länder working groups on waste (LAGA) and water (LAWA) and of the working group of the supreme state health agencies (AOLG, formerly AGLMB) have been integrated into preparation of basic data and the technical bases for derivation of test and measures thresholds (especial y LABO/LAGA, 1996). In addition to the work with the aforementioned Länder bodies, a number of technical and harmonization discussions were held to provide additional scientific expertise for preparation of test standards (for details, see: BMU-Umwelt, 1998). Derivation of test and measures thresholds takes the fol owing into account in connection with exposure: • substance properties that influence spreading of substances and, possibly, their availability for resorption, • soil properties that affect substance compounds and their environmental behavior, • different types of human behaviour (play, work; different resorption pathways and relevant presence durations), and • the quality and number of the available data (statistical data, epidemiological For derivation of test thresholds, the exposure is chosen so that "in an unfavourable exposure case", the presence of a danger for the relevant asset must be assumed. The extent of the possible impairment of the asset in question must also be considered. Depending on the reliability and extent of the data available for assessing the exposure, the "unfavourable case" is assumed to include a high percentage of the possible exposure conditions. Test thresholds for protection of health of humans in direct contact with soils are listed as the total content of the pollutant in question. Measures thresholds are not listed in Annex 2 Number 1 of the Federal Ordinance on Soil Protection and Contaminated Sites – except for dioxin – because the technical bases and methods are still lacking that would be needed to express a given measures threshold as the amount of a soil pollutant that would be available for human resorption. Measurement of the portion that is available for resorption, of the total amount of a given pollutant in the soil, is considered to be an important methodological prerequisite for introduction of measures thresholds. The Federal Environmental Agency has published a guide to application of methods and standards for calculation of test thresholds, for each substance listed in Annex 2 of the Federal Ordinance on Soil Protection and Contaminated Sites (Federal Environmental Agency, 1999). For other substances and substance properties, especial y volatile substances and nitroaromates, other derivation standards have to be applied; such alternatives are also listed in the Federal Environmental Agency's publication (1999). Test and measures thresholds, pursuant to the Federal
Ordinance on Soil Protection and Contaminated Sites, for
children's play areas, residential areas, parks and recreational
facilities and for industrial and commercial properties
Differentiation of usages
The usage orientation of the test and measures thresholds means that the values set forth in Annex 2 of the Federal Ordinance on Soil Protection and Contaminated Sites must be allocated to certain forms of usage. If parts of a suspect site or site suspected of being contaminated have been used in a different, more sensitive manner than that predominating for the site overall, then such site parts must be assessed in accordance with the standards defined for such other usage. For values pursuant to Annex 2 Number 1 Federal Ordinance on Soil Protection and Contaminated Sites, the fol owing usages are differentiated (see also number 1.1 in Annex 2 of the Federal Ordinance on Soil Protection and Contaminated Sites): Children's play areas This includes areas set aside for children and that are normally used as play areas, but does not include the sand in sandboxes, which is normally subject to special regulations. As this definition indicates, the areas actually used for play are the issue. Areas that have been special y designed for children's play (children's playgrounds) fal into this category. In the framework of the public sector's responsibility for human safety, officially designated children's playgrounds enjoy special public care; as a result, such playgrounds must also be assessed in keeping with the standards used by the public health-care The focus here is on residential areas, including yards and small gardens and other, similarly used gardens, and also where such areas have not been included or categorized under planning law, within the meaning of the Federal Land Utilization Ordinance. Under "residential areas", the Federal Land Utilization Ordinance includes small developments, residential-only areas and general residential areas and village areas. This does not include parks and recreational areas, which are assessed as a separate usage category. Where unpaved areas in residential areas are used as children's play areas, they must be assessed as such. This distinction from the first type of usage mentioned above makes it possible to assess sub-areas that are used in a different, more sensitive manner than that predominating for the relevant site overall in accordance with the standards defined for such other usage. Where home gardens, which are also referred to in the Federal Land Utilization Ordinance, are used for raising vegetables for home consumption, then such usage must be individually reviewed to determine whether it should be assessed in keeping with the criteria prescribed for the soil-plant effects Parks and recreational facilities Parks and recreational facilities are understood to be facilities reserved for social, health-related and sports-related purposes, especial y public and private parks and unpaved areas that are regularly accessible to the public. Regular accessibility is a condition oriented to the children's usage that is assumed in deriving the relevant thresholds. Industrial and commercial properties This refers to unpaved portions of work and production areas that are used only during working hours and that are not themselves the focus of work. As a rule, militarily used areas are placed in this category. Usages may be specified that are relevant to assessment of other substances for which the Federal Ordinance on Soil Protection and Contaminated Sites does not provided any test and measures thresholds (Federal Environmental Agency, 1999). Measures thresholds of Annex 2 Number 1.1 for direct resorption of
dioxins/furanes on children's play areas, residential areas, parks and
recreational areas and industrial and commercial properties
In general, derivation standards for these measures thresholds should be oriented to the pollutant components in the soil that are available for human resorption. With respect to the asset "human health", various methods for determining the resorbable portions of pollutants in the soil have been developed and tested. Studies with various extraction agents (Hack, Kraft, Selenka, 1997) have shown that matrix effects can become very significant, one example being lead in different soil materials. In addition, the heterogeneity of the material in question may have to be considered. Standardized, and thus scientifically founded, methods for determining availability for resorption are not yet available; research and standardization is currently being carried out relative to selection of the "correct" physiologically accurate elution In addition, derivation of a measures threshold may also be defined, as a recourse, in terms of determination of the total content of a substance in the soil, if definition of a measures threshold seems more suitable, with respect to enforcement, than definition of a test threshold, in light of the disproportional testing complication that a test threshold might require. This is the case for dioxins/furanes. Results of calculations similar to approach described below, for test thresholds, confirm the soil guideline established for children's play areas by the Federal/Länder working group on "dioxins". For this reason, the values listed by the Federal/Länder working group on "dioxins" for children's play areas, for residential areas and for parks and residential areas have been adopted here. Additional testing in cases where these values are exceeded, i.e. additional measurements and (for example) human biomonitoring studies, are normally not appropriate and would create disproportionate costs. The values listed have thus been defined as measures thresholds. For application of these measures thresholds, it has been established that where dioxin-containing alkaline residues from copper schist ("pebble red") are found, measures thresholds, in light of the low human resorption involved, are applied not directly to protect human health, but in the interest of sustainable, i.e. precautionary danger prevention. This reflects the current findings, which are considered largely assured, that dioxins/furanes in pebble red have only low resorption availability. Assessment standards from human toxicology
The methods and standards used for derivation of test thresholds must be in keeping with current findings. Assessment standards from human toxicology are applied in accordance with standardized methodology. In addition to oral ingestion of pollutants from the soil, inhalative and even dermal intake must be considered. To this end, tolerable resorbed dosages (TRD) are used as standards for assessment of internal impacts (Kalberlah, Hassauer, Schneider, 1998), standards that, in terms of definition, derivation methods and protection levels correspond largely to the values introduced, in a similar context, by the World Health Organization (WHO) and by other organizations such as the U.S. Environmental Protection Agency (EPA). The TRD values may deviate from such other values in cases in which the cited bodies fail to present standardized findings or in which more recent studies necessitate a reassessment. By definition, the TRD values refer to daily exposure levels that are not likely to impair the health of even sensitive persons who experience them over the course of a lifetime. The assessment standards from human toxicology are derived and justified as scientific assessments of data that may also contain empirical, plausibly justified extrapolations to the realm of the asset "human health". The tolerable resorbed dosages (TRD), as assessment standards, represent a possible basis for justifying test thresholds for the soil. In principle, other assessment standards from human toxicology could also be applied, as long as they fulfill the requirements arising from the explanations below (LABO/LAGA, 1996). In actual fact, the test-threshold calculations carried out for the individual substances mentioned in the Federal Ordinance on Soil Protection and Contaminated Sites, including those for lead and arsenic, for example, show that additional assessment standards from human toxicology have also been used (Federal Environmental Agency, 1999). 2.3.1 Derivation of assessment standards from human toxicology, within the
framework of Art. 8 Federal Soil Protection Act and Art. 4 Federal
Ordinance on Soil Protection and Contaminated Sites
Definition
Tolerable resorbed doses (TRD) are defined as tolerable, daily-resorbed bodily doses at which, with sufficient probability when individual substances are considered in accordance with current findings, no negative impacts on human health are expected and which are expected to have only low probability of causing illness. Combined impacts are not taken into account. The TRD value refers to the maximum daily internal impact, resulting exclusively via the pathway under consideration, that can still just be tolerated. TRD values are available for the substances referred to in the Federal Ordinance on Soil Protection and Contaminated Sites, Annex 2 Number 1, for inhalative and oral pathways. Where necessary, dermal substance resorption is also considered. These values are listed as the daily resorbed pollutant amounts per kg body weight (mg/kg x d). Ideally, a TRD value is based on findings concerning impacts on the most sensitive members of the population. Frequently, such data (human data) are either not available or are inadequate. In such cases, the needed TRD value is extrapolated, with the help of certain factors, from data gained in animal experiments or from inadequate human data. For the inhalative loads, media-oriented values are given – for example, as airborne concentrations in mg/m3 – in cases in which it makes little sense to determine a bodily dosage, such as in cases of local respiratory tract impacts. Since the doses concerned here are not resorbed doses, these concentration values are referred to not as TRD values but as reference concentrations (RC).
The toxicity equivalent concept may be applied in suitable cases. 2.3.1.2 Database
The fol owing sources may be used as secondary sources to support comprehensive documentation of the toxic properties of the substances in question: ˉ Air Quality Guidelines of the WHO (World Health Organization) ˉ WHO Guidelines for Drinking Water Quality ˉ Assessments, by WHO/FAO (Food and Agricultural Organization) bodies of experts, of food additives or contaminants ˉ Reports within the framework of the International Programme on Chemical Safety (IPCS) of the WHO (Environment Health Criteria) ˉ Monographs of the International Agency for Research on Cancer (IARC) of the ˉ Air Quality Criteria of the Federal Environmental Agency (individual publications) ˉ Risk Reduction Monographs of the Environment Directorate of the European ˉ Substance reports of the advisory group on environmentally relevant old substances (Beratergremium Umweltrelevante Altstoffe – BUA) ˉ Medical-toxicological justifications of MAK values, and relevant publications on workplace standards in other countries, such as the workplace standards justifications of the ACGIH (American Conference of Governmental Industrial Hygienists, U.S.) ˉ Toxicological assessments of the chemical industry's professional association ˉ Integrated Risk Information System (IRIS) of the U.S. Environmental Protection ˉ Substance reports of the Agency for Toxic Substances and Disease Registry ˉ Health Effects Assessment Summary Tables of the EPA ˉ Health Effects Assessment Documents, Ambient Water Quality Criteria and Drinking Water Health Advisories of the EPA ˉ Priority Substance List Assessment Reports of the Canadian government In individual cases, materials such as the "toxicity reviews" of the English "Health and Safety Executive", of the U.S. "National Institute for Occupational Safety and Health" (NIOSH), and publications of similar organizations in the Netherlands and in Sweden, can be important sources of data. The fol owing on-line databases contain relevant publications on toxic effects in the low-dosage range: ˉ Assessment standards for limiting cancer risk from air pollution, prepared by the working group on "cancer risk from air pollution", of the Länder committee for immission protection (LAI), ˉ Assessments of the German Cancer Research Center in Heidelberg, ˉ Air Quality Guidelines of the WHO and ˉ the Integrated Risk Information System (IRIS) of the EPA (U.S.) may be used in description and assessment of cancer risk figures. For purposes of derivation of soil test thresholds, no separate classifications of cancer risks have been carried out; the existing cancer-risk figures have been adopted. 2.3.1.3 Protection
Derivation of TRD figures begins with observations of humans (for example, in the workplace) or with animal experiments. The so-called "LOAEL" and "NOAEL" are determined for both short-term and long-term exposure and for the inhalative, oral and dermal resorption pathways: LOAEL: lowest observed adverse effect level = the lowest hazardous substance dosage or concentration at which (in the present study) adverse effects were NOAEL: no observed adverse effect level = the highest hazardous substance dosage or concentration at which no more adverse effects were observed. The most sensitive end points for the most sensitive animal species are used as a basis, if there are unfounded indications of non-transferability to humans. Where a LOAEL, but no NOAEL, is reported in the literature, the latter is estimated with the help of factors (see below). This is in line with standard practice, even if such (estimated) NOAELs are no longer based on observation ("observed effect") (an imprecision within the definition). No NOAELs are to be determined with respect to carcinogenic effects. Criteria for adverse effects
Assessment of adversity of effects is carried out by analogy to the WHO definition. The WHO (1994) defines "adverse effect" as fol ows: a "change in morphology, physiology, growth, development or life span of an organism which results in impairment of functional capacity or impairment of capacity to compensate for additional stress or increase in susceptibility to the harmful effects of other environmental influences". Specifically, adverse effects are understood to include, in addition to histopathological and clinically detectable changes, such effects as ˉ serious losses of body weight (> 10%) ˉ enzymatic changes, where such changes are indicative of commencing pathological processes (especial y in connection with dosage/effects correlations), ˉ significant behavioral changes and neurophysiologically detectable deviations. On the other hand, slight effects on body weight, and enzymatic changes with no correlation to the organ damage documented for higher dosages, are not considered to be adverse. Reversible effects may also be assessed as adverse, depending on their types and extent, and their different qualitative significance compared with irreversible effects is taken into account, if necessary, in extrapolation (for example, by means of a lower factor for the distance between LOAEL and NOAEL for irritation Assessment scale for carcinogenic substances
No TRD values are derived for carcinogenic effects, because it is fundamentally impossible to speak of a tolerable substance dose for such effects. Instead, for carcinogenic substances a resorbed body dose is assumed that corresponds to an additional, single-substance-oriented mathematical risk of 1:100,000 (1 x 10-5) of contracting cancer through lifelong exposure to the relevant hazardous substance. This risk definition is based on a vote of the Council of Environmental Advisors (SRU, 1993). In the explanatory remarks for its decision of 17/18 November 1994 on the value of quantitative risk assessments in environmentally oriented health protection, the Conference of Ministers of Health concurred with the SRU in this regard. Accordingly, a risk of 10-5 could be, for individual substances, the target for phased reduction of concentration values. It is in this sense that the mathematical risk of 10-5 for carcinogenic effects is to be applied here. It has been chosen to provide a protection level equivalent to that of the TRD value. This definition corresponds to 1/40 of the assessment standard used by the Länder Committee for Immission Protection (LAI) for multiple-substance stresses (risk of 1:2,500 or 40 x 10-5, LAI, 1992), a standard it is seeking for a first step for minimizing the risk from all carcinogenic air pollution. The basis for the substance-specific dose that corresponds to the mathematical risk is the "unit risk", which has been presented in publications of other competent organizations and institutions, including the LAI, German Cancer Research Center (DKFZ) in Heidelberg, the WHO and the U.S. EPA. The assessment standard for carcinogenic substances isolates, from the "background noise" of ubiquitous effects of pollutants distributed throughout the environment, the sufficient probability of a harmful effect resulting from additional exposure caused by soil contamination/contaminated sites. Below this assessment standard, the additional exposure for humans resulting from soil contamination/contaminated sites can normally hardly be measured or classified. Nonetheless, non-measurability certainly does not in itself mean that the presence of carcinogenic substances in the soil and the environment is not a matter of concern. The "acceptable" increase of tumor incidence, by one case in 100,000 exposed persons, takes into account only the statistical criterion of incidence, for all types of tumors. The significance attached to incidences of cancer varies greatly, however, in light of early warning symptoms, malignity and tendency toward metastasis, curability, treatment costs and, especially, the disease's progress and its impact on quality of life (see the Council of Environmental Advisors (SRU), 1995, Box in Tz. 86). Such differentiated consideration has not been carried out for the "Derivation standards .", however. This option should be reviewed again in the future if, in keeping with the requirement posed by the Council of Environmental Advisors, an overarching concept for assessing the different types of health impairments caused by cancer disorders can be found. Assumptions regarding resorption
In certain exposure scenarios, several different resorption pathways (often, both inhalative and oral) may be simultaneously involved. To permit appropriate consideration of pathway-specific components, the human-toxicological assessment standards are listed as resorbed doses. For determination of total internal exposure, the relevant pathway-specific resorbed pollutant amount is multiplied by the resorption quota, in order to obtain the proportional internal exposure. For each relevant substance (Federal Environmental Agency, 1999), it is noted what resorption quota is used to derive the TRD value from the animal-experiment data and what quota is used for calculation back to the resorbed pollutant amounts in humans. For many substances, no studies on resorption quotas in humans or animals have been carried out. Where this is the case, resorption is assumed to be 100%, if qualitative consideration of the substance properties points to good bio-availability. Extrapolation, use of safety factors
Wherever possible, derivation is based on reliable human data that have been reported in detail. Where such data are not available, data from animal experiments is used and extrapolated, using certain factors, to sensitive persons (groups of persons). The safety factors used in this connection correspond largely to those used by the WHO and the EPA. Other (safety) factors that are also used by the WHO and the EPA and that contain uncertain risks or reflect inadequate data are not included. The various safety factors are only partly based on conventions; for the most part, such conventions have to do with the protection level desired and statistical reliability. They are also partly based on biologically plausible assumptions on variabilities and sensitivity differences between humans and animals and, in this sense, represent extrapolation rather than inclusion of a margin of safety (Kalberlah and Schneider, 1998). On the basis of the documented effect dose and concentration, the fol owing safety factors (SF) are used to derive the TRD values: Table 1: Overview of the safety factors relevant for TRD values
Type of extrapolation: For estimating a chronic NOAEL by Difference between long-term extrapolating from subchronic to chronic and short-term exposure, in exposure duration (is not used if evaluable humans or study animals chronic experimental or epidemiological study data are available) For estimating a NOAELTV 1) from an experimental LOAELTV (is not used if evaluable [sub]chronic epidemiological data are available), using the convention NOAELTV = For estimating a NOAELE 2) from an Shape of the dose-effects curve epidemiologically determined LOAELE, using for humans and/or study animals the convention NOAELE = LOAELE : SFb (is not E or LOAELe or NOAELe is For estimating a NOAELe from a LOAELe, using the convention NOAELe = LOAELe : SFb (is not used if NOAELe is known) For bridging the interspecies variance between Interspecies variance between humans and study animals, using the humans and study animals convention LOAELE = LOAELTV : SFc or NOAELE = NOAELTV : SFc (is not used if evaluable [sub]chronic epidemiological data are available) For bridging the intraspecies variance in Intraspecies variance in humans humans, if NOAELE has been derived from an animal experiment, as a substitute, using the convention NOAELe = NOAELE : SFd For covering the intraspecies variance in humans, if NOAELE has been epidemiologically determined, using the convention NOAELe = NOAELE : SFd (is not used if LOAELe or NOAELe is known) 1) TV = from animal experiments 2) E = for the healthy adult population 3) e = for groups of sensitive persons Normally, each of the factors is set at a default value (= set value) of 10 (Kalberlah et al., 1998). When these factors are linked through multiplication, a maximum total safety factor of 10,000 results. Such a large total safety factor must be seen as indicative of very large data uncertainties. In general, safety factors should be replaced with better assessments, whenever the relevant information for this is available. The fol owing reasons, which are presented as examples, could lead to changes (normally, reductions) in safety factors: Data exist that indicate that only a low effects progredience (effects amplification) is to be expected in the transition from subchronic to chronic exposure. The effects in the observed LOAEL were already marginal and/or the course of the dose/effects relationship indicates that the presumed NOAEL is near There are wel -founded indications (for example, from model calculations) that point to low relevant species differences between animals and humans (relative to toxico-dynamics or toxico-kinetics). The underlying study already covers effects in particularly sensitive population groups (for example, epidemiological studies of large population groups that include sensitive persons). If assessment is required of a harmful soil change caused by substances with high data uncertainty (total safety factor > 3,000), then certain substance-specific and effects-specific conventions are to be chosen as standards, justified and disclosed – in addition to use of the scientific bases of assessment criteria of the TRD values. 2.3.1.8 Time
reference
Different organizations differentiate between short-term (acute and sub-acute) and long-term (subchronic and chronic) exposure in different ways. With this is mind, the fol owing time frames are used in connection with the "rodent" study-animal model: Short-term exposure is defined as exposure lasting up to four weeks (in reference to both animal-experiment conditions and to the scope of application of the TRD value for short-term exposure for humans), while TRD values for long-term exposure should provide adequate protection even for lifetime exposure. In general, results of short-term studies are not used to derive long-term TRD values. Short-term effects are understood to include observations from animal experiments in which fetus-harming effects (fetotoxic and/or teratogenic) occurred fol owing substance exposure during pregnancy. In such effects, the time of impact (day of gravidity), i.e. the fetal development phase in question, often plays a larger role than the exposure duration per se. Calculation of body doses (LOAEL, NOAEL), from animal-experiment
In animal experiments, as wel as in studies involving humans, concentration data (amount of a hazardous substance in food or in the air) often have to be converted into body doses and/or values for intermittent exposure have to be converted into continuous-exposure values. Furthermore, when animal-oriented inhalative data is transferred to humans, different lung volumes have to be taken into account. Where such conversions have not been supplied by the relevant study's author himself, with specific data, the calculations are carried out in accordance with a general scheme, using the fol owing parameters: The fol owing parameters are used for species-specific data: Table 2: Species-specific parameters for calculation of body doses
Weight (kg) Lung volume Water 2.0 0.35 0.223 0.049 0.05 Hamster 2.4 0.14 0.13 Human 70.0 70.0 20.0 2.0 0.028 1) Feed factor: the figures in mg/kg feed are to be multiplied by this value, to produce the body doses in mg/kg x d. Where different figures are used in the literature, such figures are applied. Animal-to-human conversions are carried out via the weight. b) Calculation of body doses from air concentrations With human data, the fol owing procedure is applied (KG = body weight): With animal-experiment data, the procedure is as fol ows – for example, for a rat: • resorption quota kgKG • d Calculation of the airborne concentration from the body dose (i.e. the calculation in the opposite direction), for human beings, is carried out in a similar manner, using the standard assumptions for human beings: 70 kg body weight, 20 m3/d lung volume and the relevant resorption quota. Any special aspects arising from local irritation effects must be considered separately. c) Conversion of intermittent exposure to continuous exposure, for inhalative Where the data are based on a study with non-continuous exposure, a linear conversion to continuous exposure was carried out (to 24 hours per day and 7 days per week). Such conversion is not permitted in cases in which, for example, the absolute dose ("bolus dose") or peak airborne concentrations play a decisive role in the effect. In such cases, no conversion is carried out. 2.3.1.10 Procedure for assessment standards for carcinogenic substances
The carcinogenicity of substances is assessed on the basis of two criteria: 1. Does the substance exhibit carcinogenic effects? (Does the substance have carcinogenic potential?) 2. How must the strength of the substance's carcinogenic effects in human beings be assessed? (How potent is the substance carcinogenically?) Assessment of the carcinogenic potential – carcinogenicity classes
Designation of substances' carcinogenic potential is based on the relevant classifications of the European Union, of the Committee on hazardous substances (Ausschuss für Gefahrstoffe), of the Senate commission for testing of health- hazardous substances (of the German Research Foundation – DFG), of the WHO's International Agency for Research on Cancer (IARC) and/or the U.S. EPA. The European Union's classifications pursuant to Art. 4 of the Ordinance on Hazardous Substances is accepted as binding for Germany; alternatively, the classification system of the AGS (Technical Regulations for Hazardous Substances, TRGS 905) is also used. Where no EU or AGS classification is available, the classification of the German Research Foundation's (DFG's) Senate Commission is used. Where no such DFG classification is available, classifications of the other organizations are used. In each case, the date of the database used or the year of classification are ignored. Assessment of carcinogenic potency
The carcinogenic potential is estimated using a cancer-risk calculation pursuant to the unit-risk concept; for a given dose, this indicates the cancer risk for resorption of Quality criteria for unit-risk estimates
Unit-risk calculations normally require extrapolation from high concentrations to the low-dose range. Such extrapolations can involve considerable uncertainty. For this reason, the quality of cancer-risk calculations must be assessed on a substance-by- substance basis, and the decision as to whether carcinogenic potential is to be used as a quantitative standard must be made on a substance-by-substance basis. In this approach, assessment of carcinogenic potency and of carcinogenic potential are carried out separately. As a result, the unit risk of a substance that is considered clearly carcinogenic in human beings may be considered qualitatively inadequate if the lack of a dose-effects relationship does not permit a sufficiently reliable quantitative conclusion. Conversely, the unit risk of a hazardous substance that is carcinogenic for only one species (with dose-effects relationship), and thus – for example, within the meaning of the Ordinance on Hazardous Substances – is not considered clearly carcinogenic in animals, may be considered to be qualitatively good. The quality of existing unit-risk derivations is given using the fol owing categories (Kalberlah et al., 1999): Category UR++: "unit risk highly suited": In principle, linearized estimation of the additional cancer risk in the low-risk range (0 to 10% additional risk), as obtained from properly conducted animal experiments and/or wel -founded epidemiological studies, seems highly suited as a basis for additional risk considerations. Findings concerning the effective mechanism do not seem to preclude the selected extrapolation method. Substances for which a carcinogenic-effects threshold clearly exists and can be quantified should not be classified in Category UR++. Category UR+: "unit risk suited": In principle, linearized estimation of the additional cancer risk in the low-risk range (0 to 10% additional risk) seems suited as a basis for additional risk considerations. Findings concerning the effective mechanism do not seem to preclude the selected extrapolation method. The differentiation criteria for classification in the relevant category must be applied. Category UR-: "unit risk not suited": In principle, linearized estimation of the additional cancer risk in the low-risk range (0 to 10% additional risk) does not seem suited as a basis for additional risk considerations, or there are serious shortcomings in the available data (from animal experiments or epidemiological studies) and/or the extrapolation methods used, or findings concerning the effective mechanism seem to preclude the selected extrapolation method. Risk quantification procedure in connection with low-quality unit risk
assessments
While the risk extrapolations produced with highly suited and suited unit risks (Category UR++ and Category UR+) have been used, no nationally or internationally accepted concept is available for dealing with cases in which risk quantification is of inadequate quality (Category UR-). In such cases, the fol owing procedure is used: • In connection with substances that are carcinogenic in humans, and that have clear carcinogenic potency, risk extrapolation may be assessed as qualitatively unsuited (UR-) due to shortcomings in relevant studies, to inadequacy of exposure data, to the inadequate size of the study group, etc. This may apply to carcinogenic substances of EU Category 1 and carcinogenic substances of EU Categories 2 or 3 for which a unit risk has been derived on an epidemiological basis. Where this is the case, the unit risk estimate is retained, in spite of its inadequate quality, due to the high quality of the proof of carcinogenicity. • This also holds for genetically toxic substances, for which the UR- Category can arise when the relevant animal experiments are obviously poorly designed. The reason for this is such substances' presumed lack of an effects threshold. • In substances with presumed heterogeneous or unknown effects structure, such as carcinogenic substances for which species-specificity cannot be ruled out, and which are presumed to have a non-genetically toxic effective mechanism, or for which no dose-effectiveness relationship has been established, another procedure, in addition to unit risk estimation, is used to consider carcinogenic effects in connection with soil-value calculations (Konietzka, 1999): With respect to the lowest dose for which cancer can significantly be proven in animal experiments (cancerogenic effect level, CELmin), a sufficient safety margin should be retained that has an order of magnitude similar to that used for serious, but not carcinogenic effects. To simplify matters, in animal experiments a first detectable (significant) cancer risk may be assumed for at least 10% (risk 1: 10) of the animals studied. A risk of 1:100,000 must be used as a basis for deriving soil test thresholds from this simplification, however. For this reason, the CELmin/10,000 is determined, as a reference figure, from the carcinogenicity studies. In parallel with the TRD-based soil-test-threshold calculations for non- carcinogenic effects, this reference figure is used in a soil test threshold calculation for carcinogenic effects. In cases in which the calculation with CELmin/10,000 leads to a lower soil value, the carcinogenicity is to be considered a relevant toxicological end point. Within the framework of plausibility tests for the substance in question, discussion continues as to whether the reasons for categorization in UR- are so serious that the unit risk estimate is less significant or whether they must be used as additional arguments, in comparisons with the calculations produced with the TRD values and CELmin/10,000, for determining a plausible test threshold. Carcinogenic substances for which no unit risk has yet been derived are treated in a similar manner; that means that when the carcinogenicity classification has been carried out on the basis of animal-experiment data, the substances are treated as if only inadequate risk quantitatification (UR-) has been carried out to date. 2.3.1.11 The issue of higher sensitivity of children to carcinogenic substances
Outset situation
In derivation of test thresholds for the children's play area scenario, it is assumed that children ingest soil only in their first eight years of life. Assuming that children and adults react equally sensitively to the effects of carcinogenic substances, the cumulative dose connected with the tolerable risk may be distributed over eight years (see number 2.4.1.1.2). The assumption that children and adults have comparable sensitivity has been reviewed, using substance-specific data, for the substances listed in Annex 2 Number 1 Federal Ordinance on Soil Protection and Contaminated Sites. This review has provided no clear indications that children have greater sensitivity to these substances; nonetheless, they may have greater sensitivity to other substances (Schneider, 1999). The continuing lack of a general (model) review seems a fundamental deficit in this area. As to the effects of ionizing radiation, epidemiological studies have shown that cancer risks for exposure during childhood are higher than those for adult exposure. These results were obtained in studies of survivors of the atom-bomb explosions in Japan, of inhabitants of areas near a Pacific atom-bomb test area, and of tumor patients undergoing radiation therapy. The relevant data for chemical substances come predominantly from animal experiments. Vinyl chloride, and some nitrosamines and nitrosamides, have been wel studied in this respect. In addition, some studies have been carried out on benzo(a)pyren [B(a)P], on other polycyclic aromatic hydrocarbons (PCAH) such as dimethylbenzanthracene (DMBA) and on complex, PCAH-containing mixtures such as diesel soot. The various studies are designed in a number of different ways. All together, they show that exposure in early life, under otherwise comparable conditions, leads to greater tumor frequency than exposure in adult animals. Similar results, obtained from a range of different studies, are available for a number of additional substances. These results are listed below. The substances listed in Table 3 are considered to be genetically toxic carcinogenics. In actual fact, mechanistic studies have suggested that the higher tumor incidences have been caused by the substances' genetically toxic mechanisms in combination with intensive cell division in target organs of the growing organism. In general, comparable behavior can be expected in connection with other substances with genetically toxic effects. Table 3: Noxa with epidemiological or experimental indications of higher
sensitivity in infantile organisms (sources listed in Schneider, 1999)
Genetic toxicity Ionizing radiation Diethylnitrosamine yes animal ethylnitrosourea Nitrosomorpholine yes nitrosoguanidine Benzidine yes PCAH [B(a)P, tar-pitch aerosol, diesel soot, DMBA] 2-acetylaminofluorene yes Aflatoxin B1 yes methylazoxymethanol Urethane yes On the other hand, studies with substances for which other carcinogenic mechanisms are assumed have provided no clear proof of greater sensitivity in infantile organisms. In various test systems, the substances polybrominated biphenyls, ethylenethiourea and diphenylhydantoin have no genetically toxic effects. These substances have been tested in long-term studies. The studies showed no relevant differences in numbers of tumors in animal groups that were exposed, prenatally via the mother animal, postnatally via the mother's milk and, fol owing weaning, chronically through feed, in comparison with animals with exposure solely fol owing With saccharin, a substance that is bladder-carcinogenic in high doses but which has no genetically toxic activity, intensified effects were observed when the animals were exposed both before and after birth. In contrast to the studies with the aforementioned genetically toxic substances, long-term exposure during adulthood was required for tumor formation, however. The data on carcinogenic metals is inconclusive. One study with nickel acetate can provide a relevant indication, however. In this study, kidney tumors were caused in young fol owing transplacental exposure as a result of substance introduction in the abdominal cavities of the mother animals. A complete description is provided by Schneider (1999). The fol owing may be concluded: • For some genetically toxic carcinogenics, greater sensitivity in immature organisms has been documented. These substances, such as vinyl chloride, are not subjects of test threshold derivation within the framework of Annex 2 Number 1.4 Federal Ordinance on Soil Protection and Contaminated Sites, however. • Where, in individual cases of hazards assessment pursuant to Art. (5) Federal Ordinance on Soil Protection and Contaminated Sites, relevant assessments are required for these substances, the exposure period chosen must reflect greater sensitivity of children. If the toxicological data are inadequate for quantification, children's greater sensitivity can normally be taken into account by means of a default value (set value), normally 10, by which the risk-oriented dose must be reduced (i.e. divided). For benzo(a)pyrene and PCAH, the fol owing must also be taken into account: • For PCAH, there are epidemiological and experimental indications of greater sensitivity in infantile organisms – for both PCAH mixtures (DMBA, tar-pitch aerosols) and for B(a)p. 2.3.1.12 Validation of TRD derivations
Wherever possible, derivations of TRD values as human-toxicological assessment standards should be validated through relevant consideration by groups of experts. 2.3.2 Hazards
reference
2.3.2.1 Meaning of a hazards-oriented dose
Exposure to a hazardous substance, in doses up to the NOAELe, which is tailored to sensitive groups of the general population, is unlikely to lead to health impairment even in sensitive persons. Doses in excess of this "practically safe dose" do not necessarily pose a health hazard; there can be cause for concern, however, when the amount in excess is considerable and the exposure is prolonged. Exposure of the healthy adult population to the lowest dose with effects that would probably be assessed as adverse (LOAELE), on the other hand, already seems sure to impair the health of sensitive persons. The hazards concept within the meaning of the Federal Soil Protection Act is correlated only with the sufficient probability of an adverse effect, not with the certain occurrence of the hazard, however. The hazards-oriented dose GD lies between the NOAELe and the LOAELE. For sensitive individuals, the hazards orientation would be set between "unlikely" (NOAELe) and "very likely" (LOAELE), and sometimes at a level of "sufficiently likely". To determine the hazards-oriented dose, an interpolation must be carried out that begins with the NOAELe or TRD value and leads to a result that is considerably smaller than the LOAELE and that, if possible, corresponds to the presumed LOAELe (Konietzka and Dieter, 1998). 2.3.2.2 Substances with effects threshold
The fol owing remarks hold only for chronic/lifelong exposure, meaning that in general only extrapolated or observed data for chronic exposure may be used to estimate hazards-oriented doses. The hazards-oriented dose GD is found in the area between "total internal exposure greater than NOAELe" and "total internal exposure smaller than LOAELE". For derivation of the hazards-oriented dose, an estimate must be made, beginning with the NOAELe, and leading to a value "smaller than LOAELE", i.e. as close as possible to the presumed LOAELe. A meaningful and plausible result for a hazards-oriented dose is obtained by multiplying the TRD value with the appropriate geometric mean of the human-relevant safety factors (SF) that have been used to extrapolate the TRD value. This mean corresponds to the square root of the total extrapolation factor between the chronic NOAELe = TRD and the measured chronic NOAELTV (TV = from animal experiments), LOAELE or NOAELE. The reference to the human-relevant safety factors is justified through the need for their use, and the manner in which they are used: ˉ Use of the factors SFb and SFd has the purpose of taking the critical dose-effects curve for humans (E and e) into account and assessing its toxicological relevance. The size of the factors is selected in keeping with the presumed steepness of the dose-effects curve. SFc has the purpose of bridging the toxico-kinetic and toxico-dynamic differences between study animals and humans. The size of this factor depends not only on systematic differences (the effects intensity depends on the concentration/time integral in the target organ which, in turn, differs in keeping with the basic metabolisms of the various species), but also on the presumed or known sensitivity variability between species. If such extrapolations justify definition of a tolerable dose, in terms of protection of human health, then the reduction of this (human-oriented) total safety factor leads to an increase of this dose within a range between the effects threshold for sensitive persons and the LOAELE. Since the exact location of the hazards-oriented dose is unknown and normally cannot be determined experimentally, at a reasonable cost, a suitable surrogate must be defined. Selection of the geometric mean of the human- relevant SF as a multiplier for derivation of this desired surrogate from the tolerable resorbed dose seems plausible, for the fol owing reasons: • The link to the human-relevant SF provides substance-specificity; each substance is allocated a specific surrogate LOAELe. • Attention to the size of the safety factors creates a dependency on findings, since substance-specific findings regarding the dose-effects curve enter into the extrapolation or the size of the factor. • The conceivable results are plausible; the maximum value can attain, but not exceed, the theoretical LOAELe. The geometric-mean calculation is in keeping with the various human-relevant safety factors' independence of each other. The interpolation of a hazards-oriented dose for substances with an effects threshold is carried out with the help of a suitable hazards factor F(Gef), taking the relevant underlying substance-specific database into account. The fol owing procedure is to be used: 1. database LOAELE (F(Gef)1 = √SFb • SFd): hazards-oriented dose GD = TRD value • √SFb • SFd 2. database NOAELE (F(Gef)2 = √SFd): hazards-oriented dose GD = TRD value • √SFd 3. database LOAELe (F(Gef)3 = SFb: hazards-oriented dose GD = TRD value • SFb 4. database NOAELTV or LOAELTV, without knowledge of the steepness of the dose- effects curves for study animals and humans (F(Gef)4 = √SFc • SFd): hazards-oriented dose GD = TRD value • √SFc • SFd 5. database LOAELTV, with proven same steepness of the dose-effects curves for study animals and humans (F(Gef)5 = √SFb • SFc • SFd): hazards-oriented dose GD = TRD value • √SFb • SFc • Fd In all cases, the TRD value is given as the "provided dose" (not as the resorbed portion) and is included in calculation of the hazards-oriented dose. The relevant documentation of the Federal Environmental Agency provides substance-specific figures (Federal Environmental Agency, 1999). For substances with effects thresholds, the fol owing results: • For substances that have been less wel characterized toxicologically, the distance between the TRD value and the LOAELTV from animal experiments, as formed through the product of the human-relevant safety factors, is relatively large. The hazards-oriented dose GD, estimated by multiplying the TRD value with the square root of these safety factors, is thus also considerably smaller than this LOAELTV, and thus exceeds the TRD value only relatively slightly. Thanks to the GD, only a relatively small part of the safety margin is used up, in spite of the absolutely high multiplication involved. • For substances that have been wel characterized toxicologically, the human- relevant total safety factor between the TRD value and the LOAEL, which usually has been determined epidemiologically or with other human data, is relatively small. In the above-formulated procedure, the hazards-oriented dose GD, therefore, lies relatively slightly below the epidemiologically determined LOAEL, i.e. relatively far above the TRD value. As a result, the multiplication uses up a higher portion of the safety margin, even though the multiplier is lower than it is for an "animal experiments" database. 2.3.2.3 Substances with no effects threshold
As in the considerations for substances with effects threshold, a hazards factor F(Gef) for a carcinogenic substance, when multiplied by the dose corresponding to the additional risk, must lead to a hazards-oriented dose that makes hazards within the meaning of the Federal Soil Protection Act seem sufficiently likely. In light of the effects of carcinogenic substances, it seems plausible to derive a hazards-oriented risk for carcinogens, from the additional acceptable risk, using a maximum factor that is normally smaller than the relevant values used for non- carcinogenic substances. For substances without effects thresholds, therefore, the "hazards-oriented" risk is derived on the basis of an extrapolation (quantification of the carcinogenic potency) for the acceptable additional cancer risk (ZRakz); this is accomplished by multiplying this additional mathematical risk by the selected factor hazards-oriented risk = ZRakz • 5 (ZRakz = additional and acceptable cancer risk) For substances with no effects threshold, the "hazards-oriented" risk would be mathematically reached for ZRakz = 10-5, for example, if of 100,000 persons exposed for a lifetime to the relevant substance dose an additional five would contract cancer as a result of such exposure (ZRakz • 5). For carcinogenically suspect substances that can be linked only experimentally to contraction of cancer and/or for which quantification of the carcinogenic potency is qualitatively inadequate, there is no sufficiently reliable scientific database for calculating an acceptable additional cancer risk (ZRakz). Due to the suspicion of carcinogenicity that remains in comparison to the non-carcinogens, it is possible, using a procedure similar to that of the WHO for drinking water guideline values, to use an additional safety factor of 10 (or a multiplier of 0.10) in estimating the hazards of such substances. For substances in which carcinogenic effects are simply possible, this multiplier can also be applied to the path-similar TRD value for long-term exposure, to produce a value that is 10 times smaller than the derived toxicological TRD value for the same substance. To determine a hazards-oriented dose, this value is multiplied, in the same way, by the database-dependent factor F(Gef), in the form defined for all substances and dependent on the data situation (Number 2.3.2.2 points 1-5): hazards-oriented dose(mkanz.) = TRD value • 0.1 • F(Gef) The decision as to which toxicological end point of a value derivation should be applied, especial y in light of not absolutely clearly determined carcinogenicity potentials (EU legal classification C 3, possible carcinogen), must be made in individual substance-specific cases of derivation. Background exposure/usage quota
The hazards-oriented body dose should not be exceeded for the relevant substance, including all exposure via all pathways to human beings. The total human exposure consists of exposure from harmful soil changes and of background exposure via food To take this background exposure into account, data on the real exposure to the relevant substance should be used wherever possible. Empirical data for the ("background") exposure via food and the air are available for only a few substances, however. A rough calculation for the substances arsenic, lead and cadmium has suggested that an 80% amount of the tolerable dose be assumed. For other substances as wel , it cannot be assumed that the actual exposure via food and the air considerably exceeds this value and thus has a negative effect on human beings. A background exposure of 80% of the TRD value is thus normally assumed. In cases of systemic effects, this also applies to consideration of inhalative exposure. For other substances for which background-exposure data are available (where possible, from duplicate studies), different assumptions – either higher or lower – may be applied. The hazards orientation is affected in that the actual or assumed background exposure of human beings is subtracted, as a corresponding subset of the TRD value, from the hazards-oriented body dose. The assumption regarding the amount of the background exposure normally also includes a slightly increased resorption via other effects pathways that begin with soil pollution (pathways not considered here, such as household dust originating in the soil, for which exact quantification is currently not possible). If considerable resorption via several pathways must be assumed, this should always be checked in specific individual cases. Since, in the case of carcinogenic effects, the additional cancer risk caused by
harmful soil changes is assessed, mathematical consideration of the background exposure is not included here. The total background exposure is also not considered for substances with local effects on the respiratory pathway. List of available human-toxicological assessment standards and
For the fol owing substances or substance groups, the Federal Environmental Agency has obtained human-toxicological assessments (where possible, in light of the database, using TRD values) and relevant justifications (see also Eikmann et al., Table 4: Substances for which human-toxicological assessment standards
(TRD values) are available
Number Name
Ammonium and compounds Antimony and compounds Arsenic and compounds Beryllium and compounds Lead and compounds Buthyl benzyl phthalate Cadmium and compounds Chromium (except for Cr. VI) 50-29-3 (p,p-DDT) Dibutyl phthalate Dichlorobenzene; m- Dichlorobenzene; o Dichlorobenzene; p Dichloroethane; 1,1- Dichloroethane; 1,2- Dichloroethene; 1,1- Dichloroethene; 1,2- Dichloropropane; 1,2- Diethyl hexyl phthalate Diethyl phthalate Dihydroxybenzene; 1,2- Dinitrophenol; 2,4- Dinitrotoluol; 2,4- Dinitrotoluol; 2,6- Dinitrotoluol; techn. Sodium fluosilicate 16893-85-9 (Na-) Hexachlorobenzene Cresol (mixed isomers) Copper and compounds Monochlorophenols Sodium vanadium oxide Nickel and compounds Pentachlorophenol Polychlor. biphenyls Polychlor. naphthalines Mercury (anorg.) Selenium and compounds Tetra methyl lead Tetrachloroethane; 1,1,2,2- Tetrachloroethene (PER) Tetrachloromethane Tetrachlorophenol 58-90-2 (2,3,4,6) Thallium and compounds Thiocyanate, anorg. Trichlorobenzene; 1,2,4- Trichloroethane; 1,1,1- Trichloroethane; 1,1,2- Trichlorophenols; 2,4,5- Trimethylbenzene; 1,3,5- Trinitrotoluol; 2,4,6- Vanadium pentoxide Zinc and compounds Excepted types of exposure within the framework of derivation of test
thresholds
An overview of usage scenarios
Chapter 2.1 described the usage scenarios that are relevant to calculation of test thresholds for direct transfers from soil to humans: ˉ Children's play areas, ˉ Parks and recreational facilities, ˉ Industrial and commercial facilities. It must be assumed that different scenarios have different relevant exposure pathways, in keeping with the characteristics of the potential y exposed groups of persons. Normally, the fol owing exposure pathways are considered: • Ingestion of soil, • Inhalative intake of soil (inhalation of contaminated dusts), • Dermal resorption from soil (percutaneous resorption). This means that the fol owing exposure pathways must be considered for the various Table 5: Exposure pathways, considered in connection with usages
Ingestion of soil Inhalative intake of Dermal resorption Residential areas commercial facilities The fol owing section describes calculation of relevant soil data for the aforementioned exposure pathways. 2.4.1.1 Ingestion of soil
2.4.1.1.1 Exposure factors
The ways in which pollutants in soils affect human health must be assessed not only on the basis of a standardized toxicological foundation, but also on the basis of uniformly applied assumptions and factors for describin the exposure circumstances. Normally, the way in which an area is used determines the relevant exposure pathway that must be considered and the affected groups of persons. It must be remembered that the soil-intake rates used in deriving relevant soil data have not been confirmed by rigorous scientific methods, through comprehensive tests. They may be based on empirical data, and should be so based, wherever possible. But often it is simply necessary to depend on conventions – conventions which of course must be plausibly justified with reference to empirical studies. The fol owing assumptions are used in estimating soil ingestion: Children: 10 kg of body weight; 0.5 g soil ingestion per day Calculation of the exposure conditions for the "children's play areas" is approached in a similar way for the usage categories "residential areas" and "parks and recreational facilities": for these categories, lower exposure rates are assumed, lower by a factor of 2 and of 5, respectively, as a result of the lower accessibility of (unpaved) soil material in these categories. The exposure factors for the "children's play areas" scenario are oriented to findings reported in the international literature. The fol owing reasoning is advanced for these assumptions: • Body weight: it is assumed that the "children's play areas" scenario is relevant for children from the ages of 1 to 8. In the literature, it is undisputed that children from the ages of 1 to 3 ingest considerably more soil, as a result of their natural behavior, than children from the ages of 4 to 8. A compilation of the AGLMB (BAGS, 1995) gives a mean body-weight range of 9.1 to 15 kg for children of ages 1 to 3, of 16.2 to 21.6 kg for ages 4 to 6, and of 23.27 to 29 for ages 7 to 9. The median weights for rare cases of unfavorable development are listed as 7.6, 13.54 and 18.7 kg, respectively. On the basis of these data, a body weight of 10 kg is assumed, to give greater consideration to the "unfavorable" cases, for derivation of test thresholds. • Soil ingestion/amounts: relevant measurements and observations concerning the amounts of soil that children regularly ingest during play have been provided primarily by researchers in the U.S. and the Netherlands. Any interpretation and transfer of their findings to standard German circumstances must take a number of methodological problems into account. For example, the key studies applied different study concepts to different target groups studied. The studies differ in such respects as whether, and to what extent, soil ingestion is correlated with dust intake; and whether, and to what extent, resorption of pollutants through food and other sources is considered. Furthermore, the study periods involved (in one case, two periods of four days each) are too short to permit conclusions about the amounts customarily ingested. In addition, problems are created by incomplete sampling of food, urine and stools, and by uncertainties regarding the heterogeneity of the substances in the soils studies. Finally, the social position of the children in question must also be taken into account (for example, in connection with university kindergardens in the U.S.). The U.S. EPA has published a revised edition of its "Exposure Factors Handbook" (EPA, 1997). It notes that the mean values given by the literature support the assumption that 200 mg of soil are ingested per day, while also conceding that higher figures could result if measurements were continued for longer periods of time. According to the EPA, the studies yield upper percentile figures (90th percentile) of 106 mg/day to 1,432 mg/day, with means of 383 (soil ingestion only) and 587 mg/day (soil and dust ingestion). A more recent study carried out in the U.S., involving 64 children 1 to 6 years of age who lived near an area contiminated with arsenic, reported mean soil ingestion of 117 mg/day and a 90th percentile of 277 mg/day. The maximum level was 899 mg/day (Walker and Griffin, 1998). These figures support the assumed value used here of 500 mg/day, in keeping with the principle that the most unfavorable case should be used in deriving test thresholds. • Soil ingestion/frequency: The assumption that soil ingestion takes place 365 days of the year seems implausible in light of central Europe's climate and its social circumstances. It must be assumed that other types of play take place during "bad weather". On the other hand, the aforementioned EPA handbook notes that although all of the exposure measurements were made in the summer, soil ingestion during winter months must not be considered zero. Frequency assumptions must also be adapted in keeping with the principle of basing assumptions on the most unfavorable case. No relevant studies of play behavior are available for Germany, however. A scenario frequency of 240 days per year seems to be a plausible assumption. This also seems consistent with exposure assumed for other scenarios (residential areas). Application of a general multiplication factor to assessment of industrial and commercial facilities and their predominating inhalative soil-resorption scenarios does not seem justified, due to the different resorption pathways involved. Different exposure pathways are concerned, along with often different types of exposure locations. Furthermore, resorption in the lungs may differ relevantly from gastro- intestinal resorption. As a result, the inhalative soil-intake rate must be quantified for Attempts to quantify soil ingestion by adults have led to the assumption that such ingestion is minimal. Very little relevant data are available. According to this assumption, while adults do ingest soil, the amounts involved are so small that they may be neglected. U.S. studies do not provide a sufficiently reliable data for estimating soil ingestion by adults in order to derive relevant soil standards. In the literature and in the "Exposure Factors Handbook" (EPA, 1997) referred to above, an assumption of 50 mg/day is used. This assumption is quite arbitrary, even in comparison with the considerations carried out above relative to children's soil ingestion; it is simply an attempt to signal that soil ingestion by adults has not been "forgotten". Efforts to take into account the variability of effects data and the range ([un]certainty) of exposure assumptions and data have led to the practice, primarily in the U.S., the Netherlands and the UK, of calculating statistical frequency distributions using so- called "Monte-Carlo" methods. Their purpose is to provide better justification for the conventions applied. Analysis of frequency distributions can be used, as long as the relevant statistical prerequisites relative to sample size, etc. are fulfil ed, to provide further light in individual cases. In particular, statistical analyses of frequency distributions can be used to check whether multiple links of "unfavorable cases" lead to disproportionately unlikely cases. 2.4.1.1.2 Direct soil ingestion in connection with carcinogenic substances
In the case of carcinogenic substances, the basic toxicological data include cancer- risk calculations that describe the additional cancer risk for lifelong resorption of a unit dose (see number 2.3.1). For carcinogenic substances, a mathematical cancer risk of 1 • 10-5 per individual substance is assumed to result from soil ingestion. Hazards relevance is seen when a population risk of 5 • 10-5 is exceeded. Consequently, the hazards-oriented body dose is 5 times the dose correlated with a cancer risk of 1 • 10-5 (see number 2.3.2). In calculations for carcinogenic substances, a proportional conversion is made, with reference to the lifetime risk, to the exposure period or the cumulative dose that must be assessed. This means that the amount resorbed over a lifetime, which corresponds to a mathematical risk of 1:100,000, must be distributed over the 8 years in which soil ingestion is relevant. Assuming a life span of 70 years and a soil- ingestion period of 8 years, therefore, a factor L of 8.75 (70 years/8 years) results. This approach is also based on the assumption that no further ingestion-based exposure to harmful soil change occurs after a person's 8th year of life and that there is a linear relationship between cancer risk and exposure time. As to the question of children's higher sensitivity to carcinogenic substances, experimental data is available for some substances. This have been summarized in number 2.3.1.11. 2.4.1.1.3 Formulae for calculating soil ingestion
Formula for calculating non-carcinogenic effects
With the aforementioned exposure factors • Body weight: 10 kg, • Daily (oral) soil ingestion: 500 mg/d, • Annual exposure period: 240 days/year a soil ingestion rate of 33 mg/kg • d in the "children's play areas" scenario results. For the scenarios "residential areas" and "parks and recreational facilities", and assuming that the amount of soil ingested is lower by factors of 2 and 5, respectively, the following soil ingestion rates result: • Residential areas: 16.5 mg/kg • d • Parks and recreational facilities: 6.6 mg/kg • d The test threshold for children's play areas is then calculated as fol ows: Test threshold [mg/kg] hazards oriented body dose soil ingestion rate received dose • (hazard factor F − std. value background soil ingestion rate  ng  received dose kg • d 33 kg • d hazards-oriented body dose = pursuant to number 2.3.2 soil ingestion rate = pursuant to number 2.4.1.1 received dose (tolerable resorbed dose converted to received dose) = pursuant to number 2.3.1 = pursuant to number 2.3.2 standard value background = pursuant to number 2.3.3 The calculations for residential areas and parks and recreational facilities are carried out in a similar way, using the relevant soil ingestion rate figures. Formula for calculating carcinogenic effects
With the above-described exposure assumptions, under which soil ingestion occurs only during the first 8 years of life, an exposure-time factor L of 8.75 results, as explained. The assessment basis is the dose that corresponds to a risk of 1:100,000 (= 1 • 10-5). The hazard factor, as described above, is 5 in the case of carcinogenic effects. With the above soil ingestion rates, the fol owing formula results for calculating oral exposure to carcinogenic substances in the "children's play areas" Test threshold [mg/kg] hazards oriented body dose ex time factor L soil ingestion rate dose for risk • hazard factor F soil ingestion rate dose for risk 10 5 33 kg • d exposure time factor = pursuant to number 2.4.1.1.2 in conjunction with number 2.3.1.11 dose for risk 10-5 = pursuant to number 2.3.1.5 The calculations for residential areas and parks and recreational facilities are carried out in a similar way, using the relevant soil-ingestion-rate figures. 2.4.1.2 Inhalative soil intake for the scenarios "children's play areas",
"residential areas" and "parks and recreational facilities"
2.4.1.2.1 Exposure factors
To complement the calculation of oral ingestion, inhalative intake of fine soil particles is considered especial y for substances in which inhalative intake is considerably more toxic than oral intake (other end points, other resorption). In the usage scenario "children's play areas", the value of 1 mg/m3 airborne dust is used. This value is based on a simulation of gardening during dry weather that was carried out in order to estimate the exposure of a child playing near the gardening site. The simulation is described in the report of the working group "risk estimate and assessment in environmental hygiene" of the Environmental hygiene committee (AUH) of the AGLMB (BAGS, 1995). A total of 6 mg/m3 of inhalable dust were measured, so the active amount assumed here seems plausible. The test threshold calculation for inhalative resorption must include a soil/dust enrichment factor. Concentration factor (A) Due to physical laws, the soil's fine-grain fraction exhibits higher pollutant concentrations in comparison with the coarse-grain fraction (and to the mass). To take this relatively higher pollutant concentration in the fine-grain fraction into account, an concentration factor of 5 is assumed for anorganic substances and a factor of 10 is assumed for organic substances. Where special materials such as coal dust are concerned in individual cases of soils for assessment, and where such substances are relevant to the assessment, the concentration factor may have to be The factor 5 for anorganic substances has been substantiated for lead in the study Dresch et al. (1976), and it is also supported by other studies (Reich and Frels, 1992). The literature provides no information regarding concentration of organic substances. Experience with dioxin measurements in pebble red have shown that different concentrations result as a function of the compared grain-size fractions. The factor of 10 seems sufficiently conservative, however. 2.4.1.2.2 Calculation formulae
Soil intake rates
For the exposure scenario "children's play areas" and ingestion soil resorption, and assuming the fol owing for children: • Body weight: 10 kg, • Breathing volume during moderate activity: 15 m3/d (=0.625 m3/h), • Playing time of 2 hours/day on 240 days/year, • An assumed dust concentration level of 1 mg/m3 air, a resorption rate of 0.082 mg/kg • d is then also assumed. For the scenarios "residential areas" and "parks and recreational facilities", the following soil resorption rates result, again under the assumption of daily resorption amounts that are lower by a factors of 2 and 5, respectively: • Residential areas: 0.041 mg/kg • d • Parks and recreational facilities: 0.016 mg/kg • d The soil value for inhalative dust resorption on children's play areas is then obtained received dose • (hazard factor F − std. value background Test threshold [mg/kg] = soil ingestion rate • concentration factor A  ng  received dose kg • d The calculations for "residential areas" and "parks and recreational facilities" are carried out in a similar way, using the relevant figures for the soil intake rate. Respiration-toxic substances
For substances in which local effects in respiratory pathways are relevant to the assessment, a "tolerable airborne concentration" is used, the reference concentration (RK) (see number 2.3.1). In this case, no body dose is calculated in obtaining the test threshold. In effects on respiratory pathways, it must be assumed that the local concentration at the exposure site, and not the dose resorbed by the body, is the decisive factor in the nature of effects. Therefore, equipotency of exposure concentration may be assumed for persons with different body weights. With the above assumptions regarding the presence time, a weighting factor G of 18.25 (24 hours/2 hours x 365 days/240 days) is obtained that takes the proportional presence time into account. As above, a dust concentration of 1 mg/m3 is assumed. reference concentration RK • F • weighting factor G Test threshold [mg/kg] = • concentration factor A  ng  Carcinogenic substances
In the case of substances with carcinogenic effects, a procedure similar to that for oral exposure is used. The assessment basis is the received body dose, which results from lifelong inhalation of the substance, at a concentration that corresponds to an additional cancer risk of 1:100,000. Once again, the hazard factor F(Gef) is 5. The assumption that exposure in keeping with the scenarios relevant here occurs only in a person's first eight years of life leads to an exposure time factor of 8.75. As with non-carcinogenic inhalative intake, concentration of pollutants in dust is taken into account by means of the concentration factor A (5 for anorganic substances and 10 for organic substances). dose for risk 10 5 • hazard factor F • ex. time factor L test threshold [mg/kg] = dust concentration • concentration factor A dose for risk 10 5 kg • d In keeping with the reasoning presented above for non-carcinogenic respiratory-toxic substances, calculation of a tolerable body dose is also not carried out in the case of carcinogenic substances that predominantly cause local tumors in respiratory pathways. The assessment basis is the airborne concentration that corresponds to an additional cancer risk of 1:100,000. As with the procedure for formula 4, the proportional exposure time of 2 hours per day and 240 days per year is included in the weighting factor G (24/2 x 365/240 = 18.25). Once again, a dust concentration of 1 mg/m3 is assumed. conc. for risk 10 5 • F • weighting factor G • ex. time factor L test threshold [mg/kg] = • concentration factor A −5  ng  conc. for risk 10 Inhalative soil intake in the "industrial and commercial facilities"
scenario
2.4.1.3.1 Exposure factors
The most important exposure pathway for derivation of test thresholds for the "industrial and commercial facilities" scenario is long-term inhalative intake. A number of exposure-dependent parameters must be considered in estimating inhalative exposure. In the case of particulate pollution, these factors include the breathing rate, the exposure frequency, determination of the fraction reaching the lungs (that portion of the total dust that is inhalable), deposition of inhaled dust-like particles, concentration (i.e. the contamination percentage) in the soil in relation to the dust and the dust concentration in the outside air. To derive test thresholds, it is not necessary to assess all factors needed to characterize the exposure scenario, however. For simplification, the fol owing exposure-dependent factors are defined: a) The assumed dust concentration (mg/m3) corresponds to the percentage of the dust that is resorbed, b) Prolonged strenuous bodily activity is ruled out, c) Exposure frequency (h, d), d) Concentration factor (percentage of fine-grain dust particles/pollutant concentration in the soil), e) Deposition: 100%. In a distinction from work-safety regulations, the fol owing is emphasized: The maximum workplace concentration (MAK) pursuant to the Ordinance on Hazardous Substances (GefStoffV) refers, for a given substance, to the maximum airborne concentration of the substance that can be attained without endangering workers'/employees' health. MAK values are established by the Committee on Hazardous Substances (AGS), at the proposal of the German Research Foundation's Senate commission for testing of health-hazardous industrial substances; they represent the maximum permissible workplace-air concentrations of industrial substances – as gas, steam or suspended particulates – that, in keeping with current findings, in general will not impair the health of employees, and not unreasonably annoy or irritate them, even in the case of repetitive, long-term, normally eight-hour (daily) exposure, within the framework of an average work week of 40 hours (in keeping with Art. 3 (5) GefStoffV). A technical guideline concentration (TRK) pursuant to the Ordinance on Hazardous Substances is the workplace-air concentration of a relevant substance that can be achieved using the best available technology (Art. 3 (7) GefStoffV). TRK standards are established by the AGS. Standards for workplace air are set forth in the Technical Regulations for Hazardous Substances (TRGS) (TRGS 900, Standards for workplace air – "air quality standards" Federal Labour Gazette (Bundesarbeitsblatt) 5/1998, p. 63). The TRGS provide standards only for industrial substances, expressed in ml/m3, for workplace air; they do provide standards for other media and other pollutants. In particular, they provide no data with respect to solid substances and soils. The key criteria for establishing these standards include (pursuant to: HVBG, 1994): • The possibility of analytically determining the substances concentrations within the range of the TRK value, • The current state of the art in process and ventilation technology, including technical developments expected for the near future, • Consideration of available experience/data in occupational medicine and As the above criteria show, toxicologically based derivation of soil standards may not be based on TRK values, which are also influenced by technical perspectives. The differences between the protection purposes involved (air in the workplace, soil in the environment of facilities and in unpaved areas of industrial properties) also indicate that the different regulated areas concerned do not overlap (work safety – environmentally oriented assessment of the hazards of harmful soil changes). Nonetheless, work-safety standards apply to work near and with contaminated soils. Since the basic toxicological data are oriented to permanent exposure (24 hours/day, 365 days/year), while exposure in industrial and commercial facilities takes place only during working hours, a weighting factor is introduced. This factor expresses the relationship between hours per year and hours of presence per year in contaminated industrial and commercial facilities. The fol owing assumptions were made: • Working time: 8 hours/day on 5 days/week and 45 weeks/year, • Soil moisture and other factors reduce the exposure to dust to 1/3 of the year. The exposure duration (D) is thus calculated as fol ows: D = 45 weeks/year x 5 days/week x 8 hours/day : 3 = 600 hours/year. The weighting factor Z is then obtained as fol ows: Z = 365 days/year x 24 hours/day : 600 hours/year = 14.6. The number of days on which soil material is blown in a manner relevant to the effects pathway can be estimated only with the help of simplifying assumptions. In the main, the fol owing two influencing factors have to be considered: b) Precipitation (as a factor influencing soil moisture, which prevents blowing). Wiesert et al. (1996) report that wind erosion plays a relevant role mainly at wind speeds of over 7 m/s (measured at a height of 10m) (there is little evidence for this figure; it does not take into account soil stirred up by the wind at ground level). Data of one weather station (Bocholt) indicate that wind speeds of more than 6 m/s (at a height of 10m) occur on 92-100 days of the year. In considering contamination of neighboring sites, it must be remembered that wind directions shift. The weather station mentioned reports at least 1 mm of rainfal on about one third of all days of the year. Needless to say, such figures vary widely by region. Following the conventions of the U.S. EPA (EPA, 1997), with 0.1 mg/m3 as a low dust concentration and 2 mg/m3 as a high concentration, and of the RIVM, with 0.165 mg/m3 as a mean concentration and 0.382 mg/m3 as the highest dust concentration measured in Dutch industrial areas, the derivation is based on dust exposure (E) of 1 mg/m3 for 2 hours/day and 0.1 mg/m3 for 6 hours/day. These assumptions lead to average dust concentration of 0.325 mg/m3 for 8 hours/day (using the working time as the exposure time). The concentration factor (A) for pollutants in the soil's fine-grain fraction is set at 5 for anorganic substances and at 10 for organic substances. The reasoning behind these factors was described above in number 2.3.4.3.1. 2.4.1.3.2 Calculation formulae
The basis for assessing non-carcinogenic effects is the air concentration that corresponds to the TRD value for long-term inhalative exposure or, in the case of substances with local impacts on the respiratory tract, the reference concentration The proportional presence time under typical workplace conditions is reflected by the above-described weighting factor Z: TRD − anal. conc. • (F − background • ) weighting factor Z Test threshold [mg/kg] = dust concentration • concentration factor A  ng  TRD − anal. conc.
(F − • A similar formula, with the reference concentration RK, results in the case of substances with local effects on the respiratory tract. On the other hand, the general background exposure is not considered in those cases. reference concentration RK • F • weighting factor Z Test threshold [mg/kg] = dust concentration •  ng  Carcinogenic substances
The procedure for deriving test thresholds for carcinogenic substances is largely similar to that for non-carcinogenic substances. The assessment basis is the airborne pollutant concentration that corresponds to an additional risk of 1:100,000. Once again, the hazard factor in the case of carcinogenic effects is 5. Since for carcinogenic substances the cumulative complete-lifetime exposure must be considered, and since cancer-risk calculations are based on continuous exposure over an entire lifetime (70 years), the relationship between lifetime and exposure time is calculated as fol ows: The relevant lifetime, expressed in hours, is as fol ows for an assumed mean life expectancy of 70 years: 70 years x 365 days/year x 24 hours/day = 613,200 hours. The exposure time in an unpaved industrial/commercial facility, applying the premises described in number 2.3.4.4.1, is 12,000 hours for a work period of 20 years and 24,000 hours for a work period of 40 years. The weighting factor (Z) is thus 51.6 for a work period of 20 years and 25.8 for a work period of 40 years. conc. for risk • weighting factor Z test threshold [mg/kg] = • concentration factor A −5  ng  conc. for risk 10 2.4.1.4 Dermal soil contact and percutaneous resorption
2.4.1.4.1 Database and method
Animal experiments concerning the percutaneous bio-availability of soil-associated pollutants have been carried out for only a few substances (see Table 8 and EPA, 1992). Relevant resorption through the skin is presumed to take place primarily with organic substances with amphiphilic behavior (i.e. with lipophilic properties and not overly low water solubility). The situation with anorganic substances is more difficult to assess. In general, it is assumed that percutaneous resorption from the soil can be neglected in the case of metal compounds. Due to the complex binding situation in the soil, and the lack of relevant indications (which contrasts with the situation with organic substances), it is assumed that dermal exposure is of little significance in the case of anorganic substances. Pentachlorophenol, one of the relevant organic substances in the present context, has been studied in rhesus monkeys, both in vitro and in vivo, by a working group under Wester and Maibach (Wester et al., 1993a) (see below). The same working group has also carried out similar studies, with rhesus monkeys, with DDT and benzo(a)pyrene (Wester et al., 1990) and with polychlorinated biphenyls (Wester et In addition, a group in New Jersey, U.S. has carried out relevant studies of benzene, xylene, toluene, and phenol (Skowronski et al., 1990; Abdel-Rahman et al., 1992; Abdel-Rahman et al., 1993; Skowronski et al., 1994). For all of these substances, practically complete resorption of the doses applied to the skin was observed. These in-vitro and in-vivo studies (rat model) were carried out with a soil suspension within the substances (concentration of the substances within the soil > 20%), however. They thus tend to represent the behavior of pure substances and do not permit conclusions relative to the situation in soils. What is more, the studies were set up in a manner that prevented the volatile substances from vaporizing. This leads to higher resorption quotas than would be encountered under real-life conditions. The soil was applied to the animals' skins first and then the substances were added in pure form, with the result that mixture and binding with the soil was not assured (see also the relevant discussion in EPA, 1992). For these reasons, the results of these studies are not used to assess the percutaneous resorption of soil-associated pollutants. When experimental data are lacking, modeling can be an aid in assessing the real- life situation. The fol owing modeling approaches have been developed (for an overview, see EPA, 1992). Assessment of the percutaneous resorption of substances fol owing dermal soil • on the basis of the skin-permeability coefficients for pollutants in water environments (skin-water permeability coefficient, Kp) (Reifenrath, 1994); • on the basis of octanol-water distribution coefficients and the Henry Constant Both approaches suffer from a lack of data for validation. Table 6 presents the main experimental in-vivo findings regarding the percutaneous resorption of pollutants from soils. Most of this work was carried out by the working group under Wester and Maibach. The results may be compared with the conclusions obtained from modeling. McKone (McKone, 1990, and a refinement in McKone and Howd, 1992) proposed a model for nonionic organic substances that describes percutaneous diffusion from the soil and thus permits assessment of the relevant percutaneously resorbed percentages. The main influencing factors are the octanol-water distribution coefficient PO/W and the Henry Constant Kh. It is not necessary to describe the model's mathematical details in the present context. According to McKone (1990), the model makes it possible to classify the substances in accordance with their physio-chemical properties: for substances with lipophilie within a PO/W-range of 10 to 106, resorption is predicted as fol ows, as a function of the Henry constant1: 0.001 < Kh < 0.01: 0.01 < Kh < 0.1: Table 7 presents a qualitative substance classification in accordance with these criteria. As it shows, some of the highly lipophilic substances (DDT, benzo(a)pyrene, PCB) slightly exceed the limit of 106 for the octanol-water distribution coefficients. As an approximation, these substances may be placed in Class 1, however, on the basis of their physio-chemical properties. Attempts to confirm these results on the basis of the experimental findings from 7 show that the model of McKone (1990) predicts overly high resorbed percentages for these substances. The model provides resorption quotas that greatly exceed the 1 The Henry Constant may be expressed either in dimensionless form (Kh) or with the unit Pa m3/mol, which referred to as "H": H = 101080 • R • T • Kh = 2430 Pa • m3/mol • Kh, with R = ideal gas constant (8.206 • 10-5 Pa • m3K/mol) and T = temperature (293 K). relevant experimental findings. The highest resorption level obtained in the in-vivo experiment was that for pentachlorophenol, at 24.4%. And it must be remembered that the experimental findings were obtained with relatively long exposure times (24 h), while the model predictions apply to a period of 12 hours. Correction to equivalent time frames would increase the discrepancy. Table 6: Results of in-vivo animal experiments studying percutaneous
resorption of soil-bound pollutants (the skin-water permeability
coefficients are also listed for each substance)
Substance In-vivo Resorption Application Coverage Source Classifica- tion according to McKone1) phenol DDT rhesus PO/W > 106 0.43 PO/W > 106 1.2 chlordane rhesus PO/W > 106 0.71 toluene benzo(a)- PO/W > 106 1.2 1) Classification of substances in accordance with criteria of McKone (1990); see 2) Where not indicated otherwise: taken from EPA (1992) 3) Taken from Reifenrath (1994) The model's relative conclusions, according to which strong reduction of resorption should occur in volatile substances with Henry Constants > 0.01 (classes 3 and 4), seems plausible, however. With volatile substances, "competition" occurs between entrance through the skin and vaporization into the air from the soil on the skin. McKone places the skin resorption for benzene (class 4) at a level 2 orders of magnitude lower than that for 2,4,6-trinitrotoluene (class 1). Table 7: Classification of substances in accordance with McKone's criteria,
on the basis of their physio-chemical properties
o-dichlorobenzene 1,2-dichloropropane chloroform hexachlorobenzene tetrachloroethane hexachlorocyclohexane pentachlorophenol trichlorobenzene tetrachloroethene polychlorinated biphenyls 1,1,1-trichloroethane trimethlybenzene Group 3 and 4 substances are substances with a higher Henry Constant (Kh = 0.01). Pentachlorophenol and phenol have a considerably lower Henry Constant, due to their polar properties and their lower volatility. This also holds for lipophilic chloro- organic compounds with low volatility and for benzo(a)pyrene. Model studies carried out by McKone (1990) predict low to moderate percutaneous resorption from the soil – just a few percent – for substances with a high Henry Constant (Kh > 0.01). Using the McKone model, Burmaster and Maxwell (1991) calculated a resorption quota on the skin, and with soil layers of usual thickness, of 1 to 2% for benzene, whose physio-chemical properties are similar to those of several other relevant substances (such as toluene and ethylbenzene). These results indicate that vaporization from the skin "competes" with percutaneous resorption and thus leads to a lower resorption. On the other hand, valid experimental studies that would permit review of the model have not been carried out for any substance within Reifenrath (1994) carried out experimental studies of percutaneous resorption of 2,4,6-trinitrotoluene in vitro and in vivo and compared his findings with model predictions on the basis of the skin-water permeability coefficient Kp. While the model predicted practically complete resorption, the in-vivo resorption quota was only about The skin-water permeability coefficients for the studied substances show no correlation with the resorption behavior from the soil. Consequently, theoretical assessments based on the skin-water permeability coefficient – at least in the present form – are subject to high levels of uncertainty. 2.4.1.4.2 Additional factors that influence percutaneous resorption
2.4.1.4.2.1 Influence of the thickness of the soil layer on the skin
Theoretical considerations using the McKone model (McKone, 1990; Burmaster and Maxwell, 1991) show that for some substances the thickness of the soil layer on the skin has a strong influence. This is true especially for high lipophilic substances with PO/W > 105. Yang et al. (1989) were able to confirm this for benzo(a)pyrene in in-vitro studies. The relative resorption was lower with coverage of 56 mg/cm2 than with 9 mg/cm2. On the other hand, these findings are contradicted by studies carried out by Wester et al. (1996) with 2,4-dichlorophenoxyacetic acid, studies which produced no dependency on the soil-layer's thickness – neither in vitro nor in vivo. The thickness of the soil layers covering the skin ranged from 1 to 40 mg/cm2. In assessment of exposure to substances originating from contaminated sites, the soil coverage of the skin is generally considered to be no more than 1.7 mg/cm2 (95th percentile) (BAGS, 1995, see below). This recommendation is based largely on the work of Finley et al. (1994). With these coverage levels, it may be assumed as an approximation that the largest portion of the relevant pollutant comes into contact with the skin. As a result, the thickness of the soil layer on the skin has a negligible influence on the resorbed portion of the entire amount of pollution on the skin (expressed in % of the total pollutant amount on the skin). The assumptions regarding skin coverage have been confirmed by more recent data (Kissel et al. 2.4.1.4.2.2 Influence of the exposure time
Normally, the resorbed amount of pollutant is assumed to depend linearly on exposure time, or a small retardation period (on the order of minutes) is assumed to be fol owed by such a linear relationship (EPA, 1992). Wester et al. (1996) found, in the case of 2,4-dichlorophenoxyacetic acid, that while application of the pure substance did produce such a linear dependence on the time, application in soil resulted in retardation of several hours. Since this observation has not been confirmed by other studies (for example, see the work on benzo(a)pyrene of Yang et al., 1989), the phenomenon may be substance- specific. The fol owing section assumes linear dependency and negligible retardation. 2.4.1.4.2.3 Soil properties
In addition to depending on the properties of the pollutants themselves, adsorption of pollutants in soil particles depends strongly on soil properties such as content of organic substances and clay minerals and ion-exchange capacity. Consequently, considerable uncertainty arises in transferring experimental findings obtained with a given special soil to other soils. Nonetheless, for purposes of test-threshold derivation such transfer represents an approximation relative to real location situations. 2.4.1.4.2.4 Animal model
The resorption found in the in-vivo model with rodent species is usually higher than the resorption in humans – either as estimated from in-vitro data or as observed in vivo (Watkin and Hul , 1991). The reasons for this are found in the skin's anatomic structure and in the denser hair cover on fur-bearing animals. Monkeys exhibit lower percutaneous pollutant resorption than rats and thus resemble humans in this respect more closely than they resemble rodents (Franklin et al., 1989). The rhesus monkeys used in the studies of the working group under Wester and Maibach (Wester et al., 1990; 1992; 1993a and b; 1996) are thus a suitable model for estimating the situation 2.4.1.4.3 Consequences for consideration of percutaneous resorption in
derivation of test thresholds
In general, percutaneous resorption upon direct skin contact is considered to be a relevant factor with organic substances. For the substances considered here – with the exceptions of pentachlorophenol, DDT, PCB and benzo(a)pyrene – no experimental studies have been conducted that would permit valid assessment, however. The highest resorption rate was found for PCP (24% in 24 hours). For PCB (14% in 24 hours), benzo(a)pyrene (13% in 24 hours) and DDT (3% in 24 hours), percutaneous resorption was considerably lower. Estimates based on the theoretical model of McKone (1990) permit tendency conclusions. Nonetheless, they lack possibilities for confirmation, especially for the group of highly volatile substances. Adequate experimental data are available for pentachlorophenol, but PCP, due to the large differences in its Henry Constant, cannot be used as a guideline substance for other volatile substances considered here. Only phenol's behavior may be considered comparable to that of PCP. 2.4.1.4.4 Procedure for consideration of percutaneous resorption of
The fol owing section presents an assessment of percutaneous resorption for pentachlorophenol, using data of Wester et al. (1993a). The authors applied 14C-marked pentachlorophenol to the skin of rhesus monkeys ˉ either as an acetonic solution ˉ or as soil contaminated with PCP. The concentration of PCP in the soil was 17 mg/kg, the soil coverage on the skin was 40 mg/cm2. The soil was characterized as fol ows: 26% sand, 26% clay, 48% silt, 0.9% organic substance. The resorption from the soil proved about as large as that from acetone: whereas resorption of 24.4% was found fol owing soil application, the resorption from the acetonic solution was 29.2%. In light of the low influence of soil adsorption on bio- availability, it is likely that the thickness of the soil layer on the skin played an insignificant role. For the estimated exposure time of 5 hours, and assuming a linear relationship between the exposure time and the resorbed amount, a resorption quota of 5 % may be used for pentachlorophenol, based on the experimentally observed resorption of approximately 24% over 24 hours. Basis for calculation of exposure with percutaneous resorption in the
"children's play areas" scenario
For calculation of the percutaneous resorption for the "children's play areas" scenario, various exposure-describing parameters have to be defined. In the main, the proposals of the Environmental hygiene committee (AUH) of the working group of managerial medical civil servants of the Länder AGLMB (BAGS, 1995) were Exposed group of persons: Children (2-3 years), with body weight of 10 kg 5 hours. A presence time of 2 hours per day is assumed. It must also be assumed, however, that soil is not always removed from the skin (i.e. through washing) immediately fol owing play. Therefore, the exposure time is set at 5 Body surface covered: 2,100 cm2 (pursuant to BAGS, 1995, 95th percentile, see below) Coverage of the skin with soil 1.7 mg/cm2 (pursuant to BAGS, 1995, 95th percentile, supported by more recent data: Kissel et al., 1996a and b) Resorption of PCP: According to Wester et al. (1993a): 24% in 24 hours. Assumption of a linear relationship: a resorption of 5% is assumed for the 5-hour Soil resorption rate = body surface soil layer thickness resorption body weight kg • d The calculation provides the resorption-relevant soil contact amount and the intensity of the pollutant contact. To preserve comparability, this is expressed in terms of the soil-resorption rate, as was done for oral exposure. In the case of PCP, this soil resorption rate for dermal contact may be compared directly with the soil resorption rate for oral soil contact, since a resorption of 100% is assumed for oral exposure. At 33 mg/kg • d, this rate is about twice as high as the dermal soil resorption rate calculated for PCP. By summing the soil resorption rates, one can obtain a test threshold for the case in which both exposure pathways are present. This calculation is carried out in derivation of the test threshold for pentachlorophenol (Federal Environmental Procedure with other substances
Highly volatile substances The conclusions made with the McKone model (1990) imply that resorption of substances with high Henry Constants is considerably below that of pentachlorophenol. Phenol is an exception. Due to the two substances' similar physio-chemical properties, findings relative to PCP may be transferred to phenol to provide an approximation. In the highly volatile organic substances, the physio-chemical data that are significant with respect to skin resorption (especial y the Henry Constant) differ strongly from relevant physio-chemical data for pentachlorophenol. As a result of PCP's lower vapor pressure, equilibrium for this substance is shifted more toward the watery phase. Vaporization, as a factor that competes with skin resorption, is less significant. For this reason, PCP should not be used as a guideline substance for highly volatile organic substances. A comparison of the relative significance of the various exposure pathways for pentachlorophenol shows that with this substance, under the assumptions made above, oral exposure plays a more significant role than dermal exposure. Percutaneous resorption is about 50% that of oral resorption. For highly volatile substances, it must thus be assumed that dermal soil contact is of considerably lower significance (<50% of oral soil resorption). Furthermore, the test threshold derivation indicates that vaporization and transport into living spaces, via soil air, may be the most important exposure pathway for these substances, a pathway that predominates over all others. For this reason, no further quantitative consideration is given to dermal resorption of these substances. DDT, polychlorinated biphenyls and benzo(a)pyrene and other lipophilic nonvolatile substances The Wester working group has provided experimental data for these substances and substance groups. The findings show that the resorption rates for all of these substances are considerably lower than those for PCP. In the case of PCB, in vivo studies with rhesus monkeys showed resorption of 14%, for a 24-hour exposure period. Review of the possible consequences for test threshold derivation, carried out in a manner similar to that described above for PCP, showed that inclusion of dermal exposure would have virtually no ramifications on test threshold determination. For DDT, dermal exposure is even less significant. The situation may be assumed to be similar for lipophilic substances from Table 7, Class 1 (aldrin, hexachlorocyclohexane, hexachlorobenzene). It thus seems justified, in light of current findings for these substances, to treat dermal exposure as a factor of low quantitative relevance and to neglect it. As to benzo(a)pyrene, this is a carcinogenic substance that functions locally. This means that it is absorbed into the skin, is metabolized within the skin and causes skin tumors. In this light, it seems justified to review the significance of this exposure pathway for this substance, within the framework of the pending derivation of a PCAH Consideration of one-time high resorption of substances with high
acute toxicity
Derivation of test thresholds is oriented to long-term resorption of low doses, in order to permit suitable prevention of chronic and subchronic toxic effects resulting from resorption of soil contaminants. And yet acute toxic effects can be caused by one- time resorption of high doses of contaminants. Such exposures occur only under extreme conditions, resulting from combinations of unfavorable influencing circumstances such as massive soil contamination, high substance resorption availability, pica behavior, etc. Pica behavior refers to intentional ingestions of large amounts (on the order of grams) of soil. Pica behavior also occurs among older Review of the relevance of acute toxic effects resulting from one-time large-scale soil ingestion is based on the fol owing considerations: • One-time ingestion of 10 g of soil, and a body of weight of 10 kg, are assumed • On the basis of the lowest reported acutely effective dose for humans, a distance of a factor of 10 to the first toxic reactions is taken as the toxicological assessment standard. In the case of cyanidin, the considerations relative to acute toxicity have consequences for the test threshold levels. For cyanide, the literature reports a lowest lethal dosis of 0.56 mg/kg. In combination with the above assumptions, this figure leads to a soil contamination figure of 56 mg/kg. This value is lower than the contamination figures calculated for all scenarios, for long-term exposure. Since one- time massive soil ingestion cannot be ruled out in the "children's play areas", "residential areas" and "parks and recreational facilities" scenarios, the test threshold for these scenarios must be set on the basis of considerations regarding acute In principle, one-time massive soil ingestion by children is also considered possible in the "industrial and commercial facilities" scenario. While under normal circumstances, regular children's play on industrial and commercial facilities can practically be ruled out, such areas are not always so wel secured that one-time entry by children would be impossible. On the other hand, older children would be the endangered group in such cases. Exposure to soil-born contaminants can result from activities such as moving earth, throwing dirt and rolling on the ground. It is difficult to estimate exposure under such circumstances. In all likelihood, exposure in extreme cases, occurring under very rare circumstances, will tend to be lower than that occurring through pica behavior of small children. To take this factor (which cannot be quantitatively specified) into account, the soil contamination figure calculated for the "children's play areas" scenario for consideration of acute toxicity for small children is multiplied by two for the "industrial and commercial facilities" scenario. Whenever the resulting test threshold is exceeded, possible exposure of children should be Criteria for plausibility consideration of calculations in derivation of
test thresholds
In plausibility considerations, all calculations are reviewed in light of any available epidemiological studies that provide confirmation or suggest modifications. For example, epidemiological studies may suggest that the resorption quota of a certain contaminant, in resorption from soils, differs from the corresponding resorption quota (i.e. for the same substance) for resorption from food and this is not adequately reflected by the underlying body-dose figures. In addition, calculations must also be compared with data for background soil contamination. While calculatory results that are less than or equal to background soil values can point to an undesirable toxicological load, they do not reflect hazards that could be correlated with soil changes and thus seem unsuited as test thresholds. It may also be postulated that in such cases intake of such contaminants via ingestion of soil (i.e. orally) should be interpreted as a hazard only when it differs from intake via food. The following section summarizes the scientifically founded criteria for plausibility • Comparison of results of calculations with data on background soil contamination, in light of the insight that a test threshold must be higher than the ubiquitous • Comparison of results of calculations with assessment-relevant human- toxicological contamination data (human biomonitoring), where such data permit conclusions relative to the relevant end points. The study and testing undertaken in cases of harmful soil changes may include human biomonitoring aimed at detecting any internal human contamination. • Selection of the test threshold from the lowest result from parallel calculations of inhalative and ingestive soil intake, for both carcinogenic and non-carcinogenic If the calculations show the same concentrations for both pathways, this result should be reflected in selection of a lower test threshold. • Consideration of the percutaneous resorption pathway, if necessary through summation with ingested contaminants. • Cross-checking against considerations regarding acute effects fol owing one-time massive soil ingestion, where the available data permit this. • Checking of calculations against findings regarding smell-threshold concentrations, in order to check possible significant contamination levels. • Review of indications of any typical effects resulting from combination with other • In plausibility checking regarding the carcinogenic potency, the standards from numbers 2.3.1.5 and 2.3.1.10 must be applied. • Review of the difference between the test threshold for children's play areas and that for industrial and commercial facilities. A difference is assumed to be justified if it is assured that compliance with test thresholds for industrial and commercial facilities will not result in secondary pollution on surrounding areas, caused by uncontrol ed, diffuse run-off or blowing of soil material from industrial and commercial facilities or by mudslides. Test and measures thresholds pursuant to Annex 2 number 2
Federal Ordinance on Soil Protection and Contaminated
Sites for soils used for cultivation, horticulture, home
vegetable/fruit gardens and grassland
3.1 Preliminary
The soil's function in crop cultivation is central to derivation of test and measures thresholds pursuant to Annex 2 number 2 Federal Ordinance on Soil Protection and Contaminated Sites. With respect to the plants involved, various types of cases and valuable resources must be differentiated: • Prevention of human-toxicological effects fol owing consumption of plant foods, especial y wheat, potatoes and vegetables, • Assuring the saleability of food plants, as food, from farms and professional • Safety of feed crops and grassland cover for use as feed. In addition, phytotoxic effects on food and feed crops, resulting from harmful soil changes, must be prevented. Where substances are regulated by food or feed guidelines or regulations (standards for maximum permitted amounts, guidelines of the Central cataloguing and assessment agency (Zentrale Erfassungs- und Bewertungsstelle – ZEBS) for metals in foods, standards for feeds), the relevant standards are used as a basis for deriving the contaminant levels that may not be exceeded in plants. Another important derivation standard results from the substance-specific components of soil-borne substances that can be systemically resorbed by crop plants or that contribute to the percentages of soil-borne substances that stick to plants and that contribute to contamination of feeds. Differentiation of usages
Because the test and measures thresholds are usage oriented, the values set forth in Annex 2 must be allocated to specific usages. For test thresholds pursuant to number 2, soil usages for cultivation, horticulture, home vegetable/fruit gardens and grassland are differentiated. For soil usages for cultivation, horticulture and home vegetable/fruit gardens, standardized test and measures thresholds are useful – especial y because they are more easily understood by soil users. It is assumed that when relevant concentrations are below the maximum permissible plant concentrations, as derived from the food guidelines, then private consumption of fruit and vegetables cultivated in private gardens is safe – even from a toxicological perspective (see also below). Since fields used for cultivation of feed grasses must be assumed to exhibit the same soil/plant transfer relationships as grasslands, the former are assessed as if they were grasslands. On fields used for cultivation of silage corn, compliance with standards for grassland normally ensures compliance with requirements of the Feedstuffs Ordinance. For this reason, the standards for grassland are also applied to such fields. 3.3 Protection
orientation
In assessment of hazards in the soil/plant pathway, the fol owing different types of cases must be differentiated, in keeping with the soil usage involved: a) Sale of food plants from cultivation and commercial vegetable gardens, b) Sale/use of feed plants from cultivated fields and from grasslands, c) Consumption of home-grown fruit and vegetables from private gardens (home/small gardens). Therefore, the fol owing must be protected: • The saleability/useability of foods and fees (cases a and b above), • Human health (case c; indirectly, in cases a and b). Plant health (phytotoxic effects of soil contamination on plants) may also be Two different impacts pathways play the key roles in heavy-metal contamination of plants on contaminated soils. Firstly, heavy metals in the soil solution are absorbed by plant roots and then transported within the plant (systemic resorption pathway). Secondly, external soiling with contaminated soil material causes heavy-metal contamination of plants. In the case of food plants, such contamination cannot be completely removed even by washing and kitchen preparation (pollution pathway). From the perspective of agricultural-chemistry/soil science, predictions regarding heavy-metal soil/plant transfer via the systemic resorption pathway require a soil-test method that provides a suitable estimate of the heavy-metal fractions that are significant in resorption via plant roots. All experience to date has shown that the best method for this is soil extraction using diluted salt solutions (such as ammonium- nitrate extraction pursuant to DIN 19730). On the other hand, such heavy-metal fractions in the soil that are relevant for systemic resorption may also be estimated indirectly, using a soil-study method that characterizes the approximate total heavy- metal content (such as aqua regia extraction pursuant to DIN 38414, Part 7), taking availability-influencing soil parameters (such as pH, Corg-, clay content) into account. The heavy-metal fraction available to the plant plays a less significant role in heavy- metal soil/plant transfer via soiling (i.e. external application) than does the total heavy-metal content. Even if both resorption pathways, in general, are always effective, they do differ considerably in terms of their relative importance for plants' heavy-metal content; these differences are manifested with respect to specific elements, plant species and plant organs. In extreme cases, systemic resorption alone can be the determining factor (for example, relative to the Cd content of wheat grains); in other cases (such as the Pb content of grassland cover), they may play an insignificant role in comparison with the pollution pathway. In the area of grassland use, aqua regia extraction, within the framework of
evaluation of the TRANSFER database, is more conclusive than ammonium-nitrate soil extraction. Since, in grassland use, some feed soiling must be assumed to be unavoidable (in grazing, also through animals' soil ingestion), and since such pollution plays the key role, even at relatively low soil concentrations, in grazing animals' heavy-metal resorption, it seems more appropriate to assess the relevant hazards on the basis of the aqua-regia-soluble soil substances. 3.4 Procedure
The test and measures thresholds are derived by means of the fol owing steps • Determination of the highest permissible pollutant concentrations in plants (preparation of the plant-oriented assessment standard), • Description of heavy-metal transfer from the soil into plants, fol owed by calculatory derivation of the highest soil concentration that will still ensure compliance with the highest permissible plant concentration, • Checking of the calculated soil values for plausibility, including estimation of the toxicological load from vegetables growing in gardens polluted with heavy metals, especial y for cadmium (Delschen, Th. & Leisner-Saaber, 1998), • Definition of test or measures thresholds. Determination of the highest permissible heavy-metal concentrations in
To ensure the "saleability of food", the valid food guidelines of the ZEBS (BGVV, 1997), must be used as the basis (Table 8). Table 8: Sample guidelines for pollutants in plant foods (BGVV, 1997) – here,
for lead and cadmium – expressed in mg/kg fresh weight in the form
sold (edible parts)
Fruit with peels Leafy vegetables (except for parsley leaves, kitchen herbs and spinach) Parsley leaves Sprout vegetables Fruit vegetables Root vegetables (except for celeriac) Pomaceous fruits Stone-fruits 0.50 Fruits and rhubarb The ZEBS guidelines are not legal y binding; they are intended for administrative use and to provide orientation. Their purpose is to indicate when undesirably high pollutant concentrations are present in foods. In the interest of precautionary consumer protection, peak pollutant loads should be recognized and eliminated whenever possible. In general, the guidelines reflect empirical distributions – they are not food safety standards that have been obtained through rigorous human- toxicological methods. Derivation of the maximum permissible levels of heavy metals in feeds must be based on the Feedstuffs Ordinance (Futtermittelverordnung – FMVO, 1992) (Table 9). It includes maximum permissible levels for Cd and Pb, so-called "undesirable substances". Table 9: Maximum permissible levels for Cd and Pb pursuant to the
Feedstuffs Ordinance (FMVO, 1992), supplemented by VDI guidelines
for feeds (VDI, 1991, 1992), expressed in mg/kg feedstuffs with 88%
dry substance
Complete feed of plant origin, complete-nutrition feed for cattle, sheep and goats (except for calves and lambs) Fresh green feed, pasture grass, beet leaves, green-feed silage, hay Other complete-nutrition feeds Complete-nutrition feeds for calves, lambs and baby goats Other complete-nutrition feeds for cattle, sheep and goats Other complete-nutrition feeds Derived maximum permissible pollutant concentrations in plants
These specifications are used to derive maximum permissible pollutant concentrations in plants (some examples are listed below), which are then used to establish soil-test or measures thresholds. Here, both the double ZEBS standards and the simple FMVO standards have been used as a basis. In food safety monitoring practice, the double ZEBS standards are used as indicators that show "real" overruns of the guidelines. The simple FMVO values apply to sale of feeds; when farmers use feeds on their own farms and mix them with pollution-free feeds, the FMVO permits concentrations up to 2.5 times the standards. For conversion of fresh weight to dry weight, the water contents for edible plants parts were taken from nutrition tables (Souci et al., 1986); for grassland, they were taken from FMVO (88% dry substance). Table 10: Maximum permissible pollutant content of plants [mg/kg dry weight]
used for derivation of soil test thresholds, for Cd and Pb (as
examples), calculated from 2 times the ZEBS standards and 1 times
the FMVO standards, and converted to dry weight, taking the water
content [WC %] of edible parts into account (nutrition tables, Souci et
al., 1986) and using 12% WC for grassland cover:
Leafy vegetables (except for parsley leaves, kitchen herbs and spinach) Parsley leaves Cabbage lettuce and other types of lettuce Root vegetables (except celeriac) Sprout vegetables Fruit vegetables Fruit and rhubarb Fruit (with peels) Grassland, silo corn 1) Complete-nutrition feed (cattle) Soil/plant heavy-metal transfer
The database used for evaluating soil/plant heavy-metal transfer is the Federal Environmental Agency's "TRANSFER" database, to which extensive additional Länder data have now been added as part of the work of the federal/Länder working group on soil protection (including data from the 1995 inter-Länder study programme [LABO-ad-hoc working group on soil/plant heavy-metal transfer, 1995]). The TRANSFER database currently contains some 320,000 soil/plant2 data pairs, resulting from combination of about 120 plant species and parts and various soil- 2 Including some 61,000 data pairs for aqua regia extraction and about 21,000 data pairs for ammonium-nitrate extraction, all from free-range studies. extraction substances. In some cases, several pertinent soil study findings are available for one plant sample (use of different extraction substances with the same The database has been analyzed under commission to the Federal Environmental Agency; LABO provided technical support (Knoche et al., 1997). Aqua regia is included as a soil-extraction method, because it has been used as an extraction substance, in many areas, for study of heavy-metal soil pollution, and thus the TRANSFER database contains an extensive relevant data set. In addition, it is useful, especial y for plant species/parts in which significant pollutant concentrations are expected to result from simple contact with polluted soil (especial y grassland cover), to carry out transfer studies on the basis of substances that can be extracted with aqua regia3. Furthermore, in transfer studies based on aqua-regia extraction, the influence of mobility-determining soil parameters (pH, Corg-, clay content) must also be taken into account. Ammonium-nitrate extraction is used because it shows the heavy-metal fractions in the soil that are easily available to plants, and thus is of greater use than aqua-regia- extracted heavy-metal content figures in estimating the soil/plant heavy-metal transfer (systemic resorption via the roots). Additional perspectives regarding applicability of ammonium-nitrate extraction are found in DIN 19730. From the database used to derive test or measures thresholds, all data records are removed that were obtained in tests with artificial pollution of soil (especial y by adding water-soluble heavy-metal salts) and/or from experiments with vascular structures. The evaluation employs only data from realistic free-range studies. It is carried out via regression calculations. The dependent variable in each case is the heavy-metal concentration in the plant (dry weight), and the independent variable is the heavy-metal concentration in the soil. The evaluation includes only values that are greater than the relevant listed testing thresholds. In general, the data 3 Aqua regia extraction reveals the mobile heavy-metal content, and the majority of the non-mobile heavy-metal content – but not the entire heavy-metal content, however. For the sake of simplicity, the heavy-metal content that can be extracted with aqua regia is nonetheless referred to as the "total heavy-metal content". calculations are carried out in logarithmic form. In the evaluations relative to grassland use, the regression calculations are carried out with both the original plant data and with mathematically adjusted plant data. The reason for this is the insight that field-test and field-survey data on heavy-metal pollution of grassland cover tend to result in underestimation of the soil/plant/farm animal heavy-metal transfer that occurs in actual use. In reality, this heavy-metal transfer is not inconsiderably increased by unavoidable soiling of feed during harvest and by grazing animals' direct soil ingestion. A survey of the literature (Böcker et al., 1995) indicates that the unavoidable soiling and soil ingestion is at least 2-4% of the relevant dry feed weight. The mathematically adjusted data are thus obtained by adding 3% of the aqua-regia-soluble heavy-metal concentration in the relevant soil to actually measured plant concentrations. The regression equation shows the statistical relationship between soil and plant concentrations. The 60% confidence interval between the individual values (soil/plant data pairs) is also calculated as an estimation interval (this means that about 20% of the values lie above the upper confidence interval and about 20% of the values lie below the lower confidence interval). These figures are then used to estimate (predict), for a given maximum permissble plant concentration, at which soil concentration 20, 50 or 80% of plants would exceed the permissible concentration. More conclusive pictures of soil/plant transfer can be provided with the calculated regression relationships than with the usually listed transfer factors, because the transfer factors also depend strongly on the soil concentration (very low soil concentration and low plant concentration = high transfer factor; very high soil concentration and high plant concentration = low transfer factor); they do not exhibit constant, linear behavior. Therefore, various studies tend to produce very different transfer factors, depending on the concentration range concerned. The fol owing tables summarize the results of statistical analysis that have been taken into account in deriving test or measures thresholds. The individual results of the evaluations have been published separately (Knoche et al., 1997). Table 11: Results of analysis of the TRANSFER database for farm cultivation,
commercial vegetable cultivation, small/home gardens; the soil
figures calculated for Cd and Pb are in µg/kg; AN = ammonium-
nitrate extract; KW = aqua regia extract; B = certainty measure;
summarized from Knoche et al. (1997)
Vegetables with 40 moderate concentration behavior Vegetables with 1,180 moderate concentration behavior Table 12: Results of analysis of the TRANSFER database for grassland and
feedstuffs cultivation; the soil figures for Cd and Pb are in mg/kg;
grassland cover includes 3% addition to take soiling into account;
KW = aqua regia extract; B = certainty measure; summarized from
Knoche et al. (1997)
1) No figures for grassland cover, since the intercorrelation between plant and soil data is achieved through mathematical inclusion of 3% soiling percentage in the plant dry weight Explanation of the derivation standards, illustrated with the examples of lead
and cadmium / usage: farming (pursuant to LABO, 1997)
1. Lead: When transfer of mobilized lead (extraction with ammonium nitrate) from the soil to plants is compared with the ZEBS values for the relevant plant type, wheat, as a farm crop, and carrots, as a garden crop, prove to be the most sensitive crops in their relevant usage categories. Analysis of the TRANSFER database's lead data produced only low certainty measures for the regression equations (0.26 for carrots and 0.12 for wheat); the results of the regression equations are thus of limited suitability for derivation of soil values. For carrots, values in excess of twice the ZEBS value were found in 19% of cases studied, with soil readings < 100 µg/kg; with soil readings > 100 µg/kg, they were found in 36% of all cases studied. For wheat, values in excess of twice the ZEBS value were found in 9% of cases studied, with soil readings < 100 µg/kg; with soil readings between 300 and 1,000 µg/kg, they were found in 27% of all cases studied. On this basis, the LABO's ad hoc working group on soil/plant heavy- metal transfer has proposed a test threshold of 100 µg/kg ammonium-nitrate-extractable lead. Vegetables with more moderate concentration behavior exceed the simple ZEBS value in 27 % of all cases at 500 µg/kg. Therefore, when soil concentrations are much higher than 100 µg/kg, consumption of home-grown vegetables must be expected to deliver toxicologically relevant additional pollution 2. Cadmium: cadmium is the heavy metal with the closest relationships between soil and plant concentrations. As a result, predictability of expected plant concentrations, on the basis of soil testing, is comparatively high. The database underlying the derivation is also relatively large (for example, for wheat, n = 401). Wheat is the derivation basis for the measures threshold of 40 µg/kg. At a certainty measure of 66%, the transfer relationship is very close. The probability that wheat grown on soil with this concentration will exceed twice the food guideline standard is between 50% and 80%. At soil readings above 40 µg/kg, the simple ZEBS value is exceeded in all soil relevant samples in the database. With high reliability (probability of 50-80%), this value predicts that readings will exceed twice the ZEBS value. At soil concentrations above 40 µg/kg Cd, a total of 91% of the available data exceed twice the ZEBS value; in only 9% of all cases is the plant concentration between the simple ZEBS value and twice the ZEBS value. Even at readings below 40 µg/kg, the simple ZEBS value is exceeded in 25% of all cases and twice the ZEBS value is exceeded in over 20% of all cases. This means that soil readings below 40 µg/kg cannot generally be considered unproblematic. Below the measures threshold, it is highly probable that human consumption of home-grown vegetables will deliver no toxicologically relevant additional pollution For cadmium, hazards assessment on the basis of the aqua regia extract is also permissible, due to the substance's high availability to plants. Analysis of cadmium data obtained with ammonium-nitrate and aqua regia extraction shows that the measures threshold for average soils in the average range of pH-value and clay-content fluctuations corresponds to a value of 2 mg/kg dry weight obtained with an aqua regia extract. 4 Literature
Abdel-Rahman, M. S. U., Skowronski, G. A., Turkall, R. M., 1992: Effects of soil on the bioavailability of m-xylene after oral or dermal exposure. Journal of Soil Contamination, Vol. 1, 1992, p. 183-196 Abdel-Rahman, M. S., Skowronski, G. A., Turkall, R. M., 1993: A comparative study of the kinetics and bioavailability of pure and soil-adsorbed benzene, toluene, and m- xylene after dermal exposure. Health Risk Assessment, Dermal and Inhalation Exposure and Absorption of Toxicants, Eds. R. G. M. Wang, J. B. Knaak, H. I. Maibach, CRC Press, Boca Raton, USA, pp. 105-116 BAGS, Behörde für Arbeit, Gesundheit und Soziales, 1995: Bericht des Ausschusses für Umwelthygiene, Arbeitsgemeinschaft der leitenden Medizinalbeamtinnen und - beamten der Länder. Standards zur Expositionsabschätzung, Hamburg, 1995 Bartels, R., Scheffer, B., 1993: Reduzierung der Schwermetallaufnahme von Pflanzen aus kontaminierten Substraten durch Abdeckung mit unbelastetem Boden. In: Z. Kulturtechn. Landentwickl. 34, p. 303-310 Bergmann, W., 1989: Boden- und Umweltfaktoren, die die Mineralstoffaufnahme der Pflanzen beeinflussen – unter besonderer Berücksichtigung der Schwermetalle. In: Behrens, D., Wiesner, J. (ed.), 1989: Beurteilung von Schwermetallkontaminationen im Boden. p. 317-340, Dechema, Frankfürt a. M. BGA, Bundesgesundheitsamt (ed.), 1976: Richtwerte '76 für As-, Pb-, Cd- und Hg- Gehalte in Lebensmitteln. Bundesgesundheitsblatt 20, p. 20-21, Berlin BGVV, Bundesinstitut für gesundheitlichen Verbraucherschutz und Veterinärmedizin (ed.), 1997: Richtwerte für Schadstoffe in Lebensmitteln. Bundesgesundheitsblatt 40, Birke, C., Werner, W. 1991: Eignung chemischer Bodenextraktionsverfahren zur Prognose der Schwermetallgehalte in Pflanzen. In: Forschungszentrum Jülich (ed.), 1991: Auswirkungen von Siedlungsabfällen auf Böden, Bodenorganismen und Pflanzen. Berichte aus der ökologischen Forschung, Bd. 6, p. 224-288, Jülich Böcker, J., Meyer-Burgdorf, K.-H., Abel, H., 1995: Bodenaufnahme beim Weidegang oder bei der Fütterung von Rindern. In: Landesumweltamt NRW (ed.), 1995: Literaturstudien zum PCDD/F-Transfer vom Boden in die Nahrungskette. LUA- Materialien 11, p. 81-148, Essen Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit – AG WA I 5 / Umweltbundesamt, 1998: Basisdaten Toxikologie für umweltrelevante Stoffe zur Gefahrenbeurteilung bei Altlasten. Grundlage für untergesetzliche Regelungen zum Bundes-Bodenschutzgesetz. BMU – Umwelt Nr. 7/8 (1998), p. 346-348 Burmaster, D. E., Maxwell, N. I., 1991: Time- and loading-dependence in the McKone model for dermal uptake of organic chemicals from a soil matrix. Risk Analysis, Vol. 11, 1991, p. 491-497 Crößmann, G., 1994: Transfer und Akkumulation von Thallium bei Gemüse und Obstarten. VDLUFA-Schriftenreihe 38, Kongreßband Jena, p. 513-516, VDLUFA- Verlag, Darmstadt Delschen, Th., 1996: Bodenüberdeckung als Sanierungsmaßnahme für schwermetallbelastete Gärten: Ergebnisse eines Feldversuches. In: Pfaff-Schley, H. (ed.), 1996: Bodenschutz und Umgang mit kontaminierten Böden. p. 167-181, Springer-Verlag, Berlin Delschen, Th., Rück, F., 1997: Eckpunkte zur Gefahrenbeurteilung von schwermetallbelasteten Böden im Hinblick auf den Pfad Boden/Pflanze. In: Zs. Bodenschutz 3, Heft 4, p. 114-121 Delschen, Th., Leisner-Saaber, J., 1998: Selbstversorgung mit Gemüse aus schwermetallbelasteten Gärten: Eine Gefährdungsabschätzung auf toxikalogischer Basis. In: Zs. Bodenschutz 4, Heft 1, p.17-21 Delschen, Th., Leisner-Saaber, J., Jannusch, B., 1996: Untersuchungen zum Schwermetalltransfer Baden/Pflanze in Hausgärten der Stadt Lengerich. Herausgegeben vom Landesumweltamt NRW, Essen DIN (ed.), 1983: Aufschluß mit Königswasser zur nachfolgenden Bestimmung des säurelöslichen Anteils von Metallen. DIN 38414, Teil 7, Beuth Verlag, Berlin DIN (ed.), 1997: Extraktion von Spurenelementen mit Ammoniumnitratlösung. DIN 19730, Beuth Verlag, Berlin Dresch, W., Einbrodt, H. J., Schröder, A., 1976: Beurteilung einer möglichen Gesundheitsgefährdung durch bleihaltige Sportplatzbeläge. Sportarzt und Sportmedizin, Vol. 9, 1976, p. 216-220 Eikmann, T., Heinrich, U., Heinzow, B., Konietzka, R. (ed.), 1999: Gefährdungsabschätzung von Umweltschadstoffen, ergänzbares Handbuch toxikologischer Basisdaten und ihre Bewertung. Grundwerk, Erich Schmidt Verlag, EPA, 1997: Exposure Factors Handbook. Volume 1, General Factors, U.S. Environmental Protection Agency, Washington, DC EPA, 1992: Dermal Exposure Assessment: Principles and Applications. Interim Report. U.S. Environmental Protection Agency, Washington, DC, 1992 Finley, B. L., Scott, P., Mayhall, D. A., 1994: Development of a standard soil-to-skin adherence probability density funtion for use in Monte Carlo analyses of dermal exposure. Risk Analysis, Vol. 14, 1994, p. 555-569 FUVO, 1992: Futtermittelverordnung, Anlagen 1 bis 7, Anlageband zum Bundesgesetzblatt Teil I vom 21. November 1992 Franklin, C. A., Somers, D. A., Chu, I., 1989: Use of percutaneous absorption data in risk assessment. Journal of the American Col ege of Toxicology, Vol. 8, 1989, p. 815- Hack, A., Kraft, M., Selenka, F., 1997: Erfassung und Bewertung von Daten zur biologischen Verfügbarkeit altlastenrelevanter Stoffe, Teilprojekt Bioverfügbarkeit von organischen und anorganischen Schadstoffen aus dem Boden, Forschungsbericht zum F.+E.-Vorhaben Nr. 103 40 123, im Auftrag des Umweltbundesamtes, 1997 Hassauer, M., Kalberlah, F., 1999: Arsen und Verbindungen, in: Gefährdungsabschätzung von Umweltschadstoffen, ergänzbares Handbuch toxikologischer Basisdaten und ihre Bewertung, Eikmann, T., Heinrich, U., Heinzow, B., Konietzka, R. (ed.), Grundwerk, Kennziffer D 095, Erich Schmidt Verlag, Berlin, HVBG, Hauptverband der gewerblichen Berufsgenossenschaften (ed.), 1994: Gefahrenstoffliste, BIA-Report 1/1994, Essen Jahn, R., Kunold, W., 1997: Exkursionsführer für die Jahrestagung der Deutschen Bodenkundlichen Gesellschaft 1997 in Konstanz, Mitteilungen der Deutschen Bodenkundlichen Gesellschaft, 82, 1997 Kalberlah, F., Hassauer, M., Konietzka, R.; Schneider, K., 1999: Qualitätsbeurteilungen für unit risk Berechnungen. In: Gefährdungsabschätzung von Umweltschadstoffen, ergänzbares Handbuch toxikologischer Basisdaten und ihre Bewertung. Eikmann, T., Heinrich, U., Heinzow, B., Konietzka, R. (ed.), Grundwerk, Kennziffer B 010, Anhang 2, Erich Schmidt Verlag, Berlin, 1999 Kalberlah, F., Hassauer, M., Schneider, K., 1999: Basisdaten Toxikologie für umweltrelevante Stoffe zur Gefahrenbeurteilung bei Altlasten. Methodische Beschreibung: Tolerierbare resorbierte Dosen. In: Gefährdungsabschätzung von Umweltschadstoffen, ergänzbares Handbuch. toxikologischer Basisdaten und ihre Bewertung. Eikmann, T., Heinrich, U., Heinzow, B., Konietzka, R. (ed.), Grundwerk, Kennziffer B 010, Erich Schmidt Verlag, Berlin, 1999 Kalberlah, F., Schneider, K., 1998: Quantifizierung von Extrapolationsfaktoren, Endbericht des Forschungsvorhabens Nr. 116 06 113 des Umweltbundesamtes, Schriftenreihe der Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, Forschung, Fb 796, Verlag für neue Wissenschaft GmbH, Bremerhaven, 1998 Kissel, J. C., Richter, K. Y., Fenske, R. A., 1996a: Field measurement of dermal soil loading attributable to various activities: implications for exposure assessment. Risk Analysis, Vol. 16, 1996, p. 115-126 Kissel, J. C., Richter, K. Y., Fenske, R. A., 1996b: Factors affecting soil adherence to skin in handpress trials. Bulletin of Environmental Contamination and Toxicology, Vol. 56, 1996, p. 722-728 Knoche, H., Brandt, P., Viereck-Götte, L., Böken, H., 1997: Auswertung der aktualisierten Schwermetalldatenbank hinsichtlich der Ammoniumnitrat-Extraktion im Rahmen des UBA F+E Vorhabens Erarbeitung fachlicher Grundlagen zu untergesetzlichen Regelungen im Bodenschutz (Nr. 107 06 001/20). Fraunhofer- Institut für Umweltchemie und Ökotoxikologie, Schmallenberg, 06/1997 Knoche, H., Klein, M., Kördel, W., Klein, W., 1993: Auswertung von Länderdaten zu anorganischen Umweltchemikalien, Fraunhofer-Institut für Umweltchemie und Ökotoxikologie, UBA-FB-Nr. 107 03 007/14, Umweltbundesamt, Berlin Konietzka, R., 1999: Handlungsorientierte Vorgehensweise zur Berücksichtigung kanzerogener Wirkungen bei Substanzen mit für die weitere Risikobetrachtung nicht geeigneter Risikoquantifizierung. In: Gefährdungsabschätzung von Umweltschadstoffen, ergänzbares Handbuch toxikologischer Basisdaten und ihre Bewertung, Eikmann, T., Heinrich, U., Heinzow, B., Konietzka, R. (ed.), Grundwerk, Kennziffer B 010, Anhang 3, Erich Schmidt Verlag, Berlin, 1999 Konietzka, R., Dieter, H. H., 1998: Ermittlung gefahrenbezogener Schadstoffdosen zur Gefahrenabwehr beim Wirkungspfad Boden – Mensch. In: Rosenkranz, Einsele, Bachmann, Harreß (ed.), Handbuch Bodenschutz. 27. Lieferung, Kenn-Nr. 3530, Erich Schmidt Verlag, Berlin 1998 LABO (Bund/Länder-Arbeitsgemeinschaft Bodenschutz), 1997: Eckpunkte zur Gefahrenbeurteilung des Wirkungspfades Bodenverunreinigungen/Altlasten – Pflanze. In: Rosenkranz, Einsele, Bachmann, Harreß (ed.), Handbuch Bodenschutz, 1998, Kennziffer 9009 LABO (Bund/Länder-Arbeitsgemeinschaft Bodenschutz), 1998: Hintergrundwerte anorganischer und organischer Stoffe in Böden. In: Rosenkranz, Einsele, Bachmann, Harreß (ed.), Handbuch Bodenschutz, 1998, Kennziffer 9006 LABO/LAGA (Bund/Länder-Arbeitsgemeinschaft Bodenschutz/Länderarbeitsgemeinschaft Abfal , 1996: Eckpunkte zur Gefahrenbeurteilung des Wirkungspfades Bodenverunreinigungen/Altlasten-Mensch (Direkter Übergang), (unveröffentlicht) LABO-ad-hoc-AG Schwermetalltransfer Boden/Pflanze, 1995: Leitfaden für die Durchführung eines länderübergreifenden Untersuchungsprogrammes zum Schwermetalltransfer Boden/Pflanze. Bearbeitet von Th. Deischen und J. Leisner Saaber – Landesumweltamt NRW (unveröffentlicht) LAI, Länderausschuß für Immissionsschutz, 1992: Krebsrisiko durch Luftverunreinigungen. Ministerium für Umwelt, Raumordung und Landwirtschaft des Landes Nordrhein-Westfalen (ed.), TZ-D 161/92, 1992 Landesgesundheitsamt Baden-Württemberg, 1993: Prüfwerte für die Szenarios Kinderspielflächen, Siedlungsflächen, Gewerbeflächen (Stand 16. August 1993), Liebe, F., Welp, G., Brümmer, G. W., 1997: Mobilität anorganischer Schadstoffe in Böden Nordrhein-Westfalens. In: Landesumweltamt NRW (ed.): Materialien zur Altlastensanierung und zum Bodenschutz, Bd. 2, Essen McKone, T. E., 1990: Dermal uptake of organic chemicals from a soil matrix. Risk Analysis, Vol. 10, 1990, p. 407-419 McKone, T. E., Howd, R. A., 1992: Estimating dermal uptake of nonionic organic chemicals from water and soil: I. Unified fugacity-based models for risk assessments. Risk Analysis, Vol. 12, 1992, p. 543-558 Ministerium für Umwelt und Verkehr Baden-Württemberg (UVM-BW) (ed.), 1996: Schadstoff-transfer Boden-Pflanze. Landesweites Untersuchungsprogramm zum Transfer von Arsen, Quecksilber, Blei und Thallium aus Böden in Grünlandaufwuchs und ausgewählte Gemüsepflanzen im Vegetationsjahr 1995. Verfaßt von der UMEG Gesellschaft für Umweltmessungen und Umwelterhebungen mbH, Karlsruhe (unveröffentlicht) Mühlendahl, 1995: Vergiftungen im Kindesalter, Enke-Verlage Stuttgart 1995 Prüeß, A., 1992: Vorsorgewerte und Prüfwerte für mobile und mobilisierbare, potentiel ökotoxische Spurenelemente in Boden. Verlag Ulrich E. Grauer Reich, T., Freis, L., 1992: Bericht über Staubmessungen bei Gartenarbeit zur Abschätzung einer maximalen Belastung durch Schadstoffe im Boden (unveröffentlicht), 1992 Reifenrath, W. G., 1994: Assessment of Skin Penetration of Environmental Contaminants in Air and Bioremediated Soil Utilizing the Pig Skin Model. Reifenrath Consulting and Research, Rovato, CA, 1994 Robinson, A. L., Sextro, R. G., 1995: The influence of a subslab gravel layer and open area on soilgas and radon entry into two experimental basements. Health Physics, Vol. 69, 1995, p. 367-377 Rück, F., 1993: Standortspezifische Stickstoffmineralisierung, jahreszeitlicher Verlauf des Mineralstickstoffvorrates und der Nitratauswaschung in Böden des Wasserschutzgebietes Donauried. Hohenheimer Bodenkundliche Hefte, Heft 15, Universität Hohenheim Sachverständigenrat für Umweltfragen, SRU, 1993: Stellungnahme des Umweltrates zum Verordnungsentwurf nach § 40 Abs. 2 Bundes-Immissionsschutzgesetz (BImSchG), in: Umweltgutachten 1994, Verlag Metzler-Poeschel, Stuttgart Sachverständigenrat für Umweltfragen, SRU, 1995: Altlasten II, Sondergutachten, Februar 1995. Metzler-Poeschel, Stuttgart Sauerbeck, D., Lübben, S., 1991: BMFT-Verbundvorhaben ,,Auswirkungen von Siedlungsabfällen auf Böden, Bodenorganismen und Pflanzen", Teil I – Gesamtüberblick -. In: Forschungszentrum Jülich (ed.), 1991: Auswirkungen von Siedlungsabfällen auf Böden, Bodenorganismen und Pflanzen. Berichte aus der ökologischen Forschung, Bd. 6, p. 1-32, Jülich Sauerbeck, D., 1989: Der Transfer von Schwermetallen in die Pflanze. In: Behrens, D., Wiesner, J., (ed.): Beurteilung von Schwermetallkontaminationen im Boden. Dechema, p. 281-316, Frankfurt Schneider, K., 1999: Unterschiede in der Empfindlichkeit von Kindern gegenüber krebserzeugenden Stoffen im Vergleich zum Erwachsenen. Umweltmedizin in Forschung und Praxis, 4 (3), p. 155-162 Skowronski, G. A., Kadry, A. M., Turkall, R. M., Botrous, M. F., Abdel-Rahman, M. S., 1994: Soil decreases the dermal penetration of phenol in male pig in vitro. Journal of Toxicology and Environmental Health, Vol. 41, 1994, p. 467-479 Skowronski, G. A., Turkall, R. M., Kadry, A. R., Abdel-Rahman, M. S., 1990: Effects of soil on the dermal bioavailability of m-xylene in male rats. Environmental Research, Vol. 51, 1990, p. 182-193 Souci, p. W., Fachmann, W., Kraut, H., 1986: Die Zusammensetzung der Lebensmittel-Nährwert-Tabellen. Herausgegeben im Auftrag des Bundesministeriums für Ernährung, Landwirtschaft und Forsten von der Deutschen Forschungsanstalt für Lebensmittelchemie, München, 3. Auflage. Wissenschaftliche Verlagsgesellschaft Steubing, L., Vogel, A., Hammel, W., 1996: Bindung und Mobilität ökotoxischer Metalle und Metalloide im Bereich von ehemaligen Quecksilber-Bergwerken im Nordpfälzer Bergland – Bestimmung der mobilen Fraktion von Quecksilber, Arsen und Antimon in Böden mittels Ammoniumnitrat (NH4NO3). Forschungsvorhaben im Auftrag des Ministeriums für Umwelt Rheinland-Pfalz, Institut für Pflanzenökologie, J.-L.-Univ. Gießen (unveröffentlicht) Umweltbundesamt (ed.) 1999: Berechnung der Prüf- und Maßnahmenwerte der Bodenschutz- und Altlastenverordnung aufgrund der Ableitungsmaßstäbe für Prüf- und Maßnahmenwerte gemaß Bekanntmachung des BMU im Bundesanzeiger (in Umweltbundesamt (ed.), 1996: Umweltmedizinische Beurteilung der Exposition des Menschen durch altlastenbedingte Schadstoffe, Zwischenbericht 01.01.1995- 31.01.1996, Anhang 1.3: Fallbeispiel Pelzfabrik (Land Hessen), Arbeitsgemeinschaft Institut Fresenius GmbH und focon-Ingenieurgesellschaft mbH, Erlangen, Aachen, Umweltforschungsplan des Bundesministers für Umwelt, Naturschutz und Reaktorsicherheit, F+E-Vorhaben 203 40 107, im Auftrag des Umweltbundesamtes, Umweltministerium Baden-Württemberg (UM-BW), 1993: Dritte Verwaltungsvorschrift des Umweltministeriums zum Bodenschutzgesetz über die Ermittlung und Einstufung von Gehalten anorganischer Schadstoffe im Boden (VwV Anorganische Schadstoffe) Walker, S., Griffin, S., 1998: Site-specific data confirm arsenic exposure predicted by the U.S. Environmental Protection Agency. Environmental Health Perspectives, Vol. 106, 1998, p. 133-139 Watkin, G. E., Hull, R. W., 1991: Factors affecting the dermal bioavailability of hydrocarbons in soil: applicability to human health risk assessment. Hydrocarbon Contaminated Soils, Vol. 1, 1991, p. 541-554 Wester, R. C., Maibach, H. I., Bucks, D. A., Sedik, L., Melendres, J., Liao, C., Dizio, S., 1990: Percutaneous absorption of (14C)DDT and (14C)benzo(a)pyrene from soil. Fundamental and Applied Toxicology, Vol. 15, 1990, p. 510-516 Wester, R. C., Maibach, H. I., Sedik, L., Melendres, J., Liao, C. L., Dizio, S., 1992: Percutaneous absorption of (14C)chlordane from soil. Journal of Toxicology and Environmental Health, Vol. 35, 1992, p. 269-277 Wester, R. C., Maibach, H. I., Sedik, L., Melendres, J., Wade, M., Dizio, S., 1993a: Percutaneous absorption of pentachlorophenol from soil. Fundamental and Applied Toxicology, Vol. 20, 1993, p. 68-71 Wester, R. C., Maibach, H. I., Sedik, L., Melendres, J., Wade, M., 1993b: Percutaneous absorption of PCBs from soil: in vivo rhesus monkey, in vitro human skin, and binding to powered human stratum corneum. Journal of Toxicology and Environmental Health, Vol. 39, 1993, p. 375-382 Wester, R. G., Melendres, J., Logan, F., Hui, X., Maiback, H. I., Wade, M., et al., 1996: Percutaneous absorption of 2,4-dichlorophenoxyacetic acid from soil with respect to soil load and skin contact time: in vivo absorption in rhesus monkeys and in vitro absorption in human skin. Journal of Toxicology and Environmental Health, Vol. 47, 1996, p. 335-344 WHO, Weltgesundheitsorganisation, 1994: Assessing human health risk of chemicals. Derivation of guidance values for health based exposure limits. WHO, Wiesert, P., Rippen, G., Dörr, H., 1996: Die Verfrachtung von Schadstoffen durch Winderosion – Versuch einer Quantifizierung. altlasten spektrum, 2/96, 1996, p. 79- Yang, J. J., Roy, T. A., Krueger, A. J., Neil, W., Mackerer, C. R., 1989: In vitro and in vivo percutaneous absorption of benzo(a)pyrene from petroleum crude-fortified soil in the rat. Bulletin of Environmental Contamination and Toxicobogy, Vol. 43, 1989, p.

Source: http://www.ktm.hu/szakmai/karmentes/egyeb/karmentnemet/promul_en.pdf