View all publications

3 Derivation of Environmental Guideline Values

The process of deriving environmental guideline values is highly complex, but an understanding of at least the general principles is necessary before any value is applied to a specific site. Most guideline values (especially for soil) are derived using standard default assumptions, and these may or may not reflect actual site conditions. Also, the purpose of the guideline values derived for soil (eg, whether for site investigation or remediation) will influence the protectiveness of the derived values in different countries. Similarly, the use or potential use of water or groundwater (eg, as drinking water or for irrigation) will influence the protectiveness of the derived values.

The guideline values from the reference documents (Table 1) that are included in the EGV database are listed in Table 2. This section provides an overview of how these guideline values were derived. You should refer to the reference documents themselves for specific details and assumptions about how the individual guideline values were derived.

Environmental guideline values for soil, groundwater and water are the most commonly used in contaminated site assessments. The fundamental difference between guideline values for a given environmental medium is the basis of protection – whether for human health or ecological receptors.

Environmental guideline values can be risk-based or threshold values. Risk-based values are derived from a given exposure scenario (protection of human health) or the protection of a nominal proportion of species in an ecosystem. Threshold values may be derived from toxicological data where insufficient data is available to calculate risk-based values. Guideline values may also be classified as threshold values where insufficient information on their derivation is provided (eg, Lead Guidelines, Ministry of Health, 1998). The level of protection afforded by threshold values is unable to be determined.

Table 2 provides a summary of the name, purpose (what action exceedance of the guideline value initiates, or how the values are used within the specified reference document), and basis of protection of the guideline value used in each reference document and included in the EGV database. Further discussion on the methods used to derive criteria for the protection of human health and ecological receptors is provided below. This discussion focuses on the derivation of guideline values for soil. Discussion on the derivation of guideline values for surface water and groundwater is provided in sections 3.4 and 3.5, respectively.

Table 2: Name, purpose,* number, and basis of protection of guideline value in reference documents listed in Table 1 and included in the EGV database

Country   Name Purpose* Basis# No. of guideline values Source
New Zealand Timber treatment Acceptance criteria Site investigation HH/P 7 MfE and MoH (1997)
Gasworks Acceptance criteria Site investigation HH 19 MfE (1997)
Oil industry Acceptance criteria Site investigation HH 10 MfE (1999)
Sheep-dip Soil guideline values Site investigation HH 19 MfE (2006)
Drinking-water standards Maximum acceptable values (MAV) Drinking water HH ~130 MoH (2000)
Australia Contaminated sites NEPM Health investigation levels (HIL) Site investigation HH 26 NEPC (1999)
Environmental investigation levels (EIL) Site investigation Eco 11
ANZECC water quality Water quality guidelines Sustainable water quality Eco 81 ANZECC/ ARMCANZ (2000)
Sediment quality guidelines Investigation Eco 34
USA Federal Soil-screening level (SSL) Site investigation HH 110 US EPA (2001)
Eco-SSLs Eco 21 US EPA (2003)
Regional screening levels Human Health Medium-specific screening levels Site investigation HH ~600 US EPA (see current website version)
Preliminary remediation goals (PRG) Remediation goal HH ~460 US EPA (see current website version)
Canada   Soil quality guideline values Remediation goal Integrated 29 CCME (see current website version
Water quality guideline values Sustainable water quality, drinking water Eco
HH
94 CCME (see current version on website)
UK   Soil guideline values Site investigation HH 10 DEFRA (see current website version )
Netherlands   Intervention value Remediation urgency assessment Integrated 75 VROM1 (see current website version)
Target value Sustainable soil quality Eco 75 VROM1 (see current website version
Long et al (1995) Effects range low Concentration at which 10% of studies in a database observed an effect Eco 28 Long et al (1995)
Effects range median Concentration at which 50% of studies in a database observed an effect Eco 28 Long et al (1995)

* What action exceedance of the guideline value initiates, or how the values are used within the specified reference document.

# Integrated = integration of human health and ecotoxicological data; Eco = ecotoxicological data only; HH = human health data only; P = phytotoxicity.

1 VROM: Former Ministry of Housing, Spatial Planning and the Environment; now Ministry of Infrastructure and the Environment.

3.1 Guideline values for the protection of human health

The toxicological basis for deriving human-health guideline values is either:

  • tolerable daily intakes (TDI) for contaminants that have a threshold concentration, which needs to be exceeded for toxic effects to be manifested (threshold contaminants), or
  • the excess cancer risk for contaminants that have the potential to cause detrimental effects at all concentrations (non-threshold contaminants).

Threshold and non-threshold terminology is used throughout this report in line with those countries (including New Zealand) that differentiate between genotoxic and non-genotoxic carcinogens. In this case, non-genotoxic carcinogens are considered threshold contaminants and the values are derived accordingly. Typically, the most sensitive end-point is used to set guideline values.

The TDI may also be expressed as a hazard quotient, which is the ratio of exposure to the tolerable daily intake. For non-threshold contaminants (genotoxic carcinogens) the individual excess cancer risk is expressed as the number of permissible or acceptable excess cancers allowable in a population exposed to the contaminant of concern. For example, an acceptable risk level of 1x10-4 indicates that one additional cancer in every 10,000 people in an exposed population is allowable. A risk level of 1x10-5 and 1x10-6 represents one additional cancer in 100,000 and 1,000,000 people, respectively. A risk level of 1x10-5 is used in New Zealand. If an overseas value has been derived with some other risk level, the guideline value should be adjusted up or down accordingly.

For soil, guideline values are predominantly risk based, in that they are typically derived using designated exposure scenarios that relate to different land uses. Table 3 lists the exposure scenarios used in the derivation of soil guideline values in the reference documents included in this guideline and the EGV database, and the acceptable cancer risk level used in different guidelines. For each exposure scenario, selected pathways of exposure are used to derive guideline values. These pathways typically include soil ingestion, inhalation of particulates and volatiles, and dermal absorption. For residential and agricultural exposure scenarios (where considered), produce consumption is used as an exposure pathway in guidelines from all countries except the US and Australia. The original documents should be consulted to ascertain the specific details and assumptions on which the individual guideline values are based.

It should be noted that for the human health soil guidelines referenced in this document, the soil ingestion rates are typically higher than those used in deriving the values in the Methodology.  This means the soil guideline values in this document are often more conservative (lower) than if they had been derived using the methods detailed in the Methodology.  On the other hand, some of the residential values cited in this document do not allow for the home-grown produce consumption pathway, which, for some contaminants, may result in values that are insufficiently conservative compared with values derived using the Methodology.       

Table 3: Designated exposure scenarios for guideline values

Country   Acceptable risk level (non-threshold contaminants) Land use
New Zealand Timber Treatment Guidelines 10-5

Agricultural (100% produce consumption)

Residential (10%, 50% produce consumption)

Industrial – paved, unpaved

Maintenance

Gasworks Guidelines 10-5

Agricultural/horticultural (100% produce consumption)

Standard residential (10%, 50% produce consumption)

High-density residential (no produce consumption)

Commercial/industrial

Maintenance

Parkland/recreational

Oil Industry Guidelines 10-5

Agricultural (100% produce consumption)

Residential (10%, 50% produce consumption)

Commercial/industrial

Maintenance

Sheep-dip Guidelines 10-5

Lifestyle block (50% produce consumption)

Standard residential (10% produce consumption)

High-density residential (no produce consumption)

Parks/recreation

Commercial/industrial (unpaved)

Drinking Water Standards 10-5b Potable water
Australia Contaminated Sites NEPM None specified

Standard residentialc

Residential with minimal soil contactd

Parks, recreation, open space

Commercial/industrial

ANZECC Water Quality Guidelines None specified Potable water
US SSL 10-6

Residential (no produce consumption)

Industrial - indoor worker, outdoor worker

Construction

Regional Screening Levels 10-6

Residential (no produce consumption)

Industrial - indoor worker, outdoor worker

Regional Screening Levels 10-6

Residential (no produce consumption)

Industrial

Canada   10-6

Agricultural

Residential/parkland

Commercial

Industrial

UK   None specified

Allotments

Residential with and without produce consumption

Commercial/industrial

Netherlands   10-4 Residential (10% produce consumption)

a The various scenarios may look similar, but they generally have differences in assumptions and they are all different to the similarly named scenarios in the Methodology (Ministry for the Environment, 2011).

b Arsenic in drinking water has been calculated at a different cancer risk rate. The Drinking Water Standards for New Zealand make the comment, "For excess lifetime skin cancer risk of 6 x 10-4. P[rovisional] MAV, because of analytical difficulties".

c Includes children's day-care centres, kindergartens, preschools, and primary schools.

d Includes dwellings with fully and permanently paved yard space (eg, high-rise apartments and flats).

3.2 Guideline values for the protection of ecological receptors

How guideline values are derived for the protection of ecological receptors depends on the type (number of species and end-points) and amount of data. Standardised end-points, typically either the no-observed-effect concentrations (NOECs) or the lowest-observable-effect concentrations (LOECs), are used. Guideline values based on ecotoxicological data are based on a hierarchy of methods dependent on data availability. Risk-based methods are preferred, which use statistical extrapolation procedures or calculation of the geometric mean of data. These methods require a significant quantity of high-quality data.

For example, the Canadian protocol uses a 'weight of evidence' approach based on that of Long and Morgan (1990) and requires at least 10 LOEC data points from three studies, including a minimum of two soil invertebrate and two crop/plant data points (CCME, 1996). The Dutch protocol uses the statistical extrapolation method of Aldenberg and Jaworska (2000) and requires that at least four NOECs for four different taxonomic groups are available (de Bruijn et al, 1999). The US protocol uses the geometric mean of all toxicity values from the highest preference level to establish Eco-SSLs for plants and soil invertebrates. The preference level of a study is determined during the process of review, where studies are scored against nine criteria (maximum score is 2), including toxicological end-points and contaminant bioavailability. Risk-based methodologies are typically used to derive guideline values for aquatic ecosystems. Long et al (1995) used a weight-of-evidence approach to derive sediment quality criteria.

Where insufficient data are available for the preferred methodologies, guideline values are derived using factors that extrapolate available data to the desired ecotoxicological end-point (eg, LC50 to NOEC), and/or to take into consideration the limited amount of data. These extrapolation factors may range from 2 to 1000 (Cavanagh and O'Halloran, 2002 and 2003). Due to the lack of available soil ecotoxicity data, extrapolation factor methods are predominantly used to derive soil guideline values. If no data are available, the Dutch guidelines use quantitative structure activity relationships (QSAR) to extrapolate toxicity data from structurally similar compounds which have the same mode of action. Equilibrium partitioning (EqP) methods may also be used to derive values for soil toxicity by extrapolation from aquatic toxicity data. Values derived by these methods are termed 'threshold values'.

The interim ecological investigation levels (EIL) provided in the Australian Guideline on the Investigation Levels for Soil and Groundwater (NEPC, 1999) are also considered threshold values, as there is no information on their derivation. These values have been collated from ANZECC B-levels, which originate from Environment Canada (1988) and Richardson (1985), and soil survey data from four Australian cities.

A summary of the ecotoxicological end-points, methods and level of protection used in different reference documents is given in Table 4.

Table 4: Ecotoxicological end-points, method and level of protection

Country Endpoint Method Level of protection*

Netherlands

NOEC

Statistical extrapolation

Extrapolation factor

EqP

QSAR

95% (TV), 50% (IV)

NAD

NAD

NAD

Canada

LOEC

Weight-of-evidence

Extrapolation factor

75%

NAD

Australia

(ANZECC Water Quality Guidelines)

NOEC

Statistical extrapolation

Extrapolation factor

95%

NAD

US Eco-SSL

NOEC

Geometric mean

~50%

* Level of protection expressed as a percentage of species in an ecosystem.

Notes: TV = target value; IV = intervention value; EqP = equilibrium partitioning; QSAR = quantitative structure activity relationship; NAD = not able to be determined.

There is a growing body of information on guideline values for ecological receptors, including ecotoxicity data and how these data are used to derive criteria suitable for use by regulators. The document A Critical Review of Methods for Developing Ecological Soil Quality Guidelines and Criteria (American Petroleum Institute Biomonitoring Task Force, 1999, produced as US EPA, 2000, Exhibit 1-1) also provides useful background information on the derivation of ecological guideline values in a variety of countries. This document can be found at: http://rais.ornl.gov.

Additional criteria may also be found at the URLs listed in the Appendix. Note that any criterion sourced from these locations should be independently verified with an experienced ecotoxicologist before applying it.

3.3 Integrated guideline values

The Canadian Environmental Quality Guidelines for soil and the Dutch intervention and target values are the only criteria that are based on protection of human health and ecological receptors. For the soil quality guidelines separate values are derived for the protection of human health and ecological receptors, as described above. These values are compared and the lowest is selected as the final guideline value.

The Dutch intervention values are similarly derived, but their target values are based on the assumption that organisms in ecosystems are probably more exposed to compounds in water, sediment, and soil than are humans, so these are solely based on ecotoxicological data (de Bruijn et al, 1999).

3.4 Surface water

The toxicological basis for guideline values for human consumption of water is the same as that described above; that is, a TDI, or an acceptable risk level (10-5 in New Zealand). These guideline values are also termed 'risk-based' and are typically based on the consumption of 2 L of water per day by a 70 kg adult over a defined exposure period, which in New Zealand is 70 years. In New Zealand, a variable proportion of the TDI (often 10%) is allocated to exposure via drinking water.

Guideline values for aquatic ecosystems in the ANZECC Water Quality Guidelines and the Canadian Guidelines are derived as described above, and are risk-based when sufficient data are available. Guideline values for aquatic ecosystems in New Zealand Guidelines are primarily based on older Australian, Canadian and US data.

Guideline values for additional water uses such as irrigation, stock watering and recreational use provided in these documents are largely threshold based, and limited (if any) information is provided on their derivation.

3.5 Groundwater

Groundwater quality may be protected either by deriving a soil concentration that is protective of the groundwater resource, or by setting a water concentration that cannot be exceeded. Soil concentration guidelines are typically derived by back-calculation of the soil concentration that would exceed a given water quality standard, using equilibrium partitioning equations and taking into account any dilution that is expected to occur. Dilution may occur for a number of reasons, including the infiltration/recharge rate of the aquifer, the size of the aquifer, and the depth of the aquifer in relation to the contaminated soil.

The New Zealand Guidelines for Assessing and Managing Petroleum Hydrocarbon Contaminated Sites in New Zealand (Revised 2011) (Ministry for the Environment, 1999) and the US EPA guidelines provide soil concentrations that are protective of groundwater resources.  No specific dilution factor is provided in the New Zealand guidelines, although values were derived using parameters appropriate to New Zealand (Ministry for the Environment, 1999).  The US EPA guidelines use default dilution factors of 1 and 20.

Where a water concentration is set, the potential use (eg, drinking water, irrigation) of the groundwater is typically taken into consideration. The exceptions to this are the Dutch intervention and target values. Here, groundwater intervention values are derived from the soil intervention values using equilibrium partitioning and may have been corrected for human consumption of groundwater (ie, if the derived value was higher than that based on consumption of 2 L per day over a lifetime, the value was lowered), or the detection limit of the compound (ie, if the derived value was lower than the detection limit, the value was corrected upwards). For groundwater target values a distinction is made between 'shallow' (less than 10 metres) and 'deep' groundwater for metal contaminants to take into account background (naturally occurring) concentrations of metals in the Netherlands. Groundwater target values for organic contaminants are derived from a separate study, although no information is provided on their derivation in Ministry of Housing, Spatial Planning and the Environment, 2000.

3.6 Hazardous Substances and New Organisms Act

The Hazardous Substances and New Organisms (HSNO) Act 1996 was introduced into New Zealand with the aim of protecting people and the environment from the adverse effects of hazardous substances and new organisms.  The Act is administered by the Environmental Protection Authority (EPA) whose function under the Act is to make decisions on applications relating to the introduction of hazardous substances and new organisms into New Zealand.  This includes the re-assessment of previously approved substances or organisms.

The HSNO Act provides for the EPA to establish exposure limits for hazardous substances for the protection of human health and ecological receptors.  An exposure limit is defined as the maximum amount of a hazardous substance that can be legally present in a particular environmental medium such as air, water or soil, or deposited on a plant surface (such as plant foliage).

There are two types of exposure limits that may be set for hazardous substances.

  • The tolerable exposure limit (TEL) is designed to protect humans from the adverse effects of toxic substances. It is the concentration of a substance in an environmental medium that will present a low risk of a toxic effect occurring in people exposed to that substance.
  • The environmental exposure limit (EEL) is designed to protect organisms other than humans (including plants) from the adverse effects of ecotoxic substances. It is the concentration of a substance in an environmental medium that will present a low risk of adverse environmental effects in non-target areas.

TELs and EELs are set for 'new' toxic and ecotoxic substances that are assessed under the HSNO Act, and are also set for existing substances as they are transferred to the HSNO regime as required by the Hazardous Substances (Classes 6, 8, and 9 Controls) Regulations. Where a TEL or EEL has been set for a substance, it is an offence to use that substance in a way that causes the concentration to exceed the exposure limit set for that specific environmental medium. A limited number of TELs and EELs have been set and are available at www.epa.govt.nz.

The process of deriving environmental guideline values is highly complex, but an understanding of at least the general principles is necessary before any value is applied to a specific site. Most guideline values (especially for soil) are derived using standard default assumptions, and these may or may not reflect actual site conditions. Also, the purpose of the guideline values derived for soil (eg, whether for site investigation or remediation) will influence the protectiveness of the derived values in different countries. Similarly, the use or potential use of water or groundwater (eg, as drinking water or for irrigation) will influence the protectiveness of the derived values.

The guideline values from the reference documents (Table 1) that are included in the EGV database are listed in Table 2. This section provides an overview of how these guideline values were derived. You should refer to the reference documents themselves for specific details and assumptions about how the individual guideline values were derived.

Environmental guideline values for soil, groundwater and water are the most commonly used in contaminated site assessments. The fundamental difference between guideline values for a given environmental medium is the basis of protection – whether for human health or ecological receptors.

Environmental guideline values can be risk-based or threshold values. Risk-based values are derived from a given exposure scenario (protection of human health) or the protection of a nominal proportion of species in an ecosystem. Threshold values may be derived from toxicological data where insufficient data is available to calculate risk-based values. Guideline values may also be classified as threshold values where insufficient information on their derivation is provided (eg, Lead Guidelines, Ministry of Health, 1998). The level of protection afforded by threshold values is unable to be determined.

Table 2 provides a summary of the name, purpose (what action exceedance of the guideline value initiates, or how the values are used within the specified reference document), and basis of protection of the guideline value used in each reference document and included in the EGV database. Further discussion on the methods used to derive criteria for the protection of human health and ecological receptors is provided below. This discussion focuses on the derivation of guideline values for soil. Discussion on the derivation of guideline values for surface water and groundwater is provided in sections 3.4 and 3.5, respectively.

Table 2: Name, purpose,* number, and basis of protection of guideline value in reference documents listed in Table 1 and included in the EGV database

3.1 Guideline values for the protection of human health

The toxicological basis for deriving human-health guideline values is either:

  • tolerable daily intakes (TDI) for contaminants that have a threshold concentration, which needs to be exceeded for toxic effects to be manifested (threshold contaminants), or
  • the excess cancer risk for contaminants that have the potential to cause detrimental effects at all concentrations (non-threshold contaminants).

Threshold and non-threshold terminology is used throughout this report in line with those countries (including New Zealand) that differentiate between genotoxic and non-genotoxic carcinogens. In this case, non-genotoxic carcinogens are considered threshold contaminants and the values are derived accordingly. Typically, the most sensitive end-point is used to set guideline values.

The TDI may also be expressed as a hazard quotient, which is the ratio of exposure to the tolerable daily intake. For non-threshold contaminants (genotoxic carcinogens) the individual excess cancer risk is expressed as the number of permissible or acceptable excess cancers allowable in a population exposed to the contaminant of concern. For example, an acceptable risk level of 1x10-4 indicates that one additional cancer in every 10,000 people in an exposed population is allowable. A risk level of 1x10-5 and 1x10-6 represents one additional cancer in 100,000 and 1,000,000 people, respectively. A risk level of 1x10-5 is used in New Zealand. If an overseas value has been derived with some other risk level, the guideline value should be adjusted up or down accordingly.

For soil, guideline values are predominantly risk based, in that they are typically derived using designated exposure scenarios that relate to different land uses. Table 3 lists the exposure scenarios used in the derivation of soil guideline values in the reference documents included in this guideline and the EGV database, and the acceptable cancer risk level used in different guidelines. For each exposure scenario, selected pathways of exposure are used to derive guideline values. These pathways typically include soil ingestion, inhalation of particulates and volatiles, and dermal absorption. For residential and agricultural exposure scenarios (where considered), produce consumption is used as an exposure pathway in guidelines from all countries except the US and Australia. The original documents should be consulted to ascertain the specific details and assumptions on which the individual guideline values are based.

It should be noted that for the human health soil guidelines referenced in this document, the soil ingestion rates are typically higher than those used in deriving the values in the Methodology.  This means the soil guideline values in this document are often more conservative (lower) than if they had been derived using the methods detailed in the Methodology.  On the other hand, some of the residential values cited in this document do not allow for the home-grown produce consumption pathway, which, for some contaminants, may result in values that are insufficiently conservative compared with values derived using the Methodology.       

Table 3: Designated exposure scenarios for guideline values

3.2 Guideline values for the protection of ecological receptors

How guideline values are derived for the protection of ecological receptors depends on the type (number of species and end-points) and amount of data. Standardised end-points, typically either the no-observed-effect concentrations (NOECs) or the lowest-observable-effect concentrations (LOECs), are used. Guideline values based on ecotoxicological data are based on a hierarchy of methods dependent on data availability. Risk-based methods are preferred, which use statistical extrapolation procedures or calculation of the geometric mean of data. These methods require a significant quantity of high-quality data.

For example, the Canadian protocol uses a 'weight of evidence' approach based on that of Long and Morgan (1990) and requires at least 10 LOEC data points from three studies, including a minimum of two soil invertebrate and two crop/plant data points (CCME, 1996). The Dutch protocol uses the statistical extrapolation method of Aldenberg and Jaworska (2000) and requires that at least four NOECs for four different taxonomic groups are available (de Bruijn et al, 1999). The US protocol uses the geometric mean of all toxicity values from the highest preference level to establish Eco-SSLs for plants and soil invertebrates. The preference level of a study is determined during the process of review, where studies are scored against nine criteria (maximum score is 2), including toxicological end-points and contaminant bioavailability. Risk-based methodologies are typically used to derive guideline values for aquatic ecosystems. Long et al (1995) used a weight-of-evidence approach to derive sediment quality criteria.

Where insufficient data are available for the preferred methodologies, guideline values are derived using factors that extrapolate available data to the desired ecotoxicological end-point (eg, LC50 to NOEC), and/or to take into consideration the limited amount of data. These extrapolation factors may range from 2 to 1000 (Cavanagh and O'Halloran, 2002 and 2003). Due to the lack of available soil ecotoxicity data, extrapolation factor methods are predominantly used to derive soil guideline values. If no data are available, the Dutch guidelines use quantitative structure activity relationships (QSAR) to extrapolate toxicity data from structurally similar compounds which have the same mode of action. Equilibrium partitioning (EqP) methods may also be used to derive values for soil toxicity by extrapolation from aquatic toxicity data. Values derived by these methods are termed 'threshold values'.

The interim ecological investigation levels (EIL) provided in the Australian Guideline on the Investigation Levels for Soil and Groundwater (NEPC, 1999) are also considered threshold values, as there is no information on their derivation. These values have been collated from ANZECC B-levels, which originate from Environment Canada (1988) and Richardson (1985), and soil survey data from four Australian cities.

A summary of the ecotoxicological end-points, methods and level of protection used in different reference documents is given in Table 4.

Table 4: Ecotoxicological end-points, method and level of protection

Country Endpoint Method Level of protection*

Netherlands

NOEC

Statistical extrapolation

Extrapolation factor

EqP

QSAR

95% (TV), 50% (IV)

NAD

NAD

NAD

Canada

LOEC

Weight-of-evidence

Extrapolation factor

75%

NAD

Australia

(ANZECC Water Quality Guidelines)

NOEC

Statistical extrapolation

Extrapolation factor

95%

NAD

US Eco-SSL

NOEC

Geometric mean

~50%

* Level of protection expressed as a percentage of species in an ecosystem.

Notes: TV = target value; IV = intervention value; EqP = equilibrium partitioning; QSAR = quantitative structure activity relationship; NAD = not able to be determined.

There is a growing body of information on guideline values for ecological receptors, including ecotoxicity data and how these data are used to derive criteria suitable for use by regulators. The document A Critical Review of Methods for Developing Ecological Soil Quality Guidelines and Criteria (American Petroleum Institute Biomonitoring Task Force, 1999, produced as US EPA, 2000, Exhibit 1-1) also provides useful background information on the derivation of ecological guideline values in a variety of countries. This document can be found at: http://rais.ornl.gov.

Additional criteria may also be found at the URLs listed in the Appendix. Note that any criterion sourced from these locations should be independently verified with an experienced ecotoxicologist before applying it.

3.3 Integrated guideline values

The Canadian Environmental Quality Guidelines for soil and the Dutch intervention and target values are the only criteria that are based on protection of human health and ecological receptors. For the soil quality guidelines separate values are derived for the protection of human health and ecological receptors, as described above. These values are compared and the lowest is selected as the final guideline value.

The Dutch intervention values are similarly derived, but their target values are based on the assumption that organisms in ecosystems are probably more exposed to compounds in water, sediment, and soil than are humans, so these are solely based on ecotoxicological data (de Bruijn et al, 1999).

3.4 Surface water

The toxicological basis for guideline values for human consumption of water is the same as that described above; that is, a TDI, or an acceptable risk level (10-5 in New Zealand). These guideline values are also termed 'risk-based' and are typically based on the consumption of 2 L of water per day by a 70 kg adult over a defined exposure period, which in New Zealand is 70 years. In New Zealand, a variable proportion of the TDI (often 10%) is allocated to exposure via drinking water.

Guideline values for aquatic ecosystems in the ANZECC Water Quality Guidelines and the Canadian Guidelines are derived as described above, and are risk-based when sufficient data are available. Guideline values for aquatic ecosystems in New Zealand Guidelines are primarily based on older Australian, Canadian and US data.

Guideline values for additional water uses such as irrigation, stock watering and recreational use provided in these documents are largely threshold based, and limited (if any) information is provided on their derivation.

3.5 Groundwater

Groundwater quality may be protected either by deriving a soil concentration that is protective of the groundwater resource, or by setting a water concentration that cannot be exceeded. Soil concentration guidelines are typically derived by back-calculation of the soil concentration that would exceed a given water quality standard, using equilibrium partitioning equations and taking into account any dilution that is expected to occur. Dilution may occur for a number of reasons, including the infiltration/recharge rate of the aquifer, the size of the aquifer, and the depth of the aquifer in relation to the contaminated soil.

The New Zealand Guidelines for Assessing and Managing Petroleum Hydrocarbon Contaminated Sites in New Zealand (Revised 2011) (Ministry for the Environment, 1999) and the US EPA guidelines provide soil concentrations that are protective of groundwater resources.  No specific dilution factor is provided in the New Zealand guidelines, although values were derived using parameters appropriate to New Zealand (Ministry for the Environment, 1999).  The US EPA guidelines use default dilution factors of 1 and 20.

Where a water concentration is set, the potential use (eg, drinking water, irrigation) of the groundwater is typically taken into consideration. The exceptions to this are the Dutch intervention and target values. Here, groundwater intervention values are derived from the soil intervention values using equilibrium partitioning and may have been corrected for human consumption of groundwater (ie, if the derived value was higher than that based on consumption of 2 L per day over a lifetime, the value was lowered), or the detection limit of the compound (ie, if the derived value was lower than the detection limit, the value was corrected upwards). For groundwater target values a distinction is made between 'shallow' (less than 10 metres) and 'deep' groundwater for metal contaminants to take into account background (naturally occurring) concentrations of metals in the Netherlands. Groundwater target values for organic contaminants are derived from a separate study, although no information is provided on their derivation in Ministry of Housing, Spatial Planning and the Environment, 2000.

3.6 Hazardous Substances and New Organisms Act

The Hazardous Substances and New Organisms (HSNO) Act 1996 was introduced into New Zealand with the aim of protecting people and the environment from the adverse effects of hazardous substances and new organisms.  The Act is administered by the Environmental Protection Authority (EPA) whose function under the Act is to make decisions on applications relating to the introduction of hazardous substances and new organisms into New Zealand.  This includes the re-assessment of previously approved substances or organisms.

The HSNO Act provides for the EPA to establish exposure limits for hazardous substances for the protection of human health and ecological receptors.  An exposure limit is defined as the maximum amount of a hazardous substance that can be legally present in a particular environmental medium such as air, water or soil, or deposited on a plant surface (such as plant foliage).

There are two types of exposure limits that may be set for hazardous substances.

  • The tolerable exposure limit (TEL) is designed to protect humans from the adverse effects of toxic substances. It is the concentration of a substance in an environmental medium that will present a low risk of a toxic effect occurring in people exposed to that substance.
  • The environmental exposure limit (EEL) is designed to protect organisms other than humans (including plants) from the adverse effects of ecotoxic substances. It is the concentration of a substance in an environmental medium that will present a low risk of adverse environmental effects in non-target areas.

TELs and EELs are set for 'new' toxic and ecotoxic substances that are assessed under the HSNO Act, and are also set for existing substances as they are transferred to the HSNO regime as required by the Hazardous Substances (Classes 6, 8, and 9 Controls) Regulations. Where a TEL or EEL has been set for a substance, it is an offence to use that substance in a way that causes the concentration to exceed the exposure limit set for that specific environmental medium. A limited number of TELs and EELs have been set and are available at www.epa.govt.nz.