RISK ASSESSMENT GUIDANCE FOR USE IN ENVIRONMENTAL IMPACT ASSESSMENT AND MONGOLIAN RISK REGULATION

Transcription

1 RISK ASSESSMENT GUIDANCE FOR USE IN ENVIRONMENTAL IMPACT ASSESSMENT AND MONGOLIAN RISK REGULATION Ministry of Nature Environment and Tourism Government Building No. 2 United Nations Street 5/11 Ulaanbaatar, Mongolia DISTRIBUTION: 1 Copy - MNET (Electronic Copy) August 2010

2 August i - MNET Risk Assessment Guide TABLE OF CONTENTS SECTION PAGE 1.0 INTRODUCTION The NEMO II Project Integration of Risk Assessment with Existing Environmental Legislation Use of Different Forms of Risk Assessment Use and Structure of this Document FUNDAMENTAL CONCEPTS OF RISK Distinguishing Between Risk, Effects and Impacts Objectives of Environmental Risk Assessment for Impact Assessment Projects with Chemical Use or Chemical Production Projects with Hazardous System Components FRAMEWORK AND METHODOLOGY FOR CONTAMINANT HEALTH RISK ASSESSMENT Framework for HHERA Problem Formulation Methods Human Health Problem Formulation Ecological Problem Formulation Chemical Screening Exposure Assessment Human Health Exposure Assessment Ecological Exposure Assessment Toxicity Assessment Human Health Toxicity Assessment Ecological Toxicity Assessment Risk Characterization Human Health Risk Estimation and Description Ecological Health Risk Estimation and Description Attributes Of A High Quality HHRA and ERA Risk Assessment Report28 LIST OF TABLES Table 1 Example Summary Table of Wildlife Species Selected as Key Receptors Table 2 Core Key Questions for Failure Mode Risk Assessment as a Component of Environmental Impact Assessment Table 3 Example Summary Table of Wildlife Species Selected as Key Receptors Table 4 Summary of Human Health Exposure Equations

3 August ii - MNET Risk Assessment Guide LIST OF FIGURES Figure 1 Human Health Risk Assessment Framework Figure 2 Ecological Risk Assessment Framework Figure 3 Schematic Illustration - Conceptual Relationship of the Components of Human Health Risk Screened during Problem Formulation Title LIST OF APPENDICES Appendix 1 Mongolian Regulation on Risk Assessment Appendix 2 Approaches to Predicting Future Environmental Concentrations in Different Media

4 August MNET Risk Assessment Guide 1.0 INTRODUCTION 1.1 The NEMO II Project NEMO II is an institutional strengthening project for the Ministry of Nature, Environment and Tourism (MNET) of Mongolia, which is partnered with World Bank and the Netherlands. As part of the NEMO II initiative, this document discusses and provides basic guidance on the process of Environmental Risk Assessment and Risk Management to support activities required under the Mongolian Regulation on Risk Assessment (Appendix 1). Such activities may be triggered primarily by the Law on Environmental Impact Assessment or the Law on Toxic and Hazardous Chemicals, and possibly other Laws cited herein. This NEMO II initiative strengthens the overall Mongolian environmental stewardship practices towards chemical management by the regulatory and industry sectors. 1.2 Integration of Risk Assessment with Existing Environmental Legislation Various laws and policies currently exist within the Mongolian regulatory regime to protect the environment and its biota, and to review newly proposed projects to avoid environmental impacts. The Regulation on Risk Assessment augments these Laws by providing a framework and defining the process and requirements to assess potential impacts in the form of risks to community/worker health, the ecological health and property, due to hazards mediated through the environment. In the context of the Law on EIA, environmental risk is just one impact assessment endpoint which may be required for evaluation to achieve regulatory EIA approval (e.g., other impacts for example may include habitat loss, impairment of ecology, and impacts to human social parameters). The following Laws are associated with the Law on EIA (i.e., cite a requirement for EIA) and, depending on site-specifc circumstances, may elict the need for an environmental risk assessment: 1. Environmental Protection Law 2. Law on Toxic and Hazardous Chemicals 3. Law on Water 4. Law on Buffer Zones 5. Law on Air 6. Law on Minerals 7. Law on Disaster Prevention Legislation, other than the Law on EIA, may employ this risk assesment process, modified appropriately within the context of the risk assessment framework, for casespecific needs. This may include, but not be limited to, situations involving:

5 August MNET Risk Assessment Guide accidents and disasters related to toxic and hazardous chemicals; unintended toxic and hazardous chemical wastes; intentional use of toxic and hazardous chemicals that may potentially cause unintended impact to human health, livestock, aquatic biota, wildlife, environment and property (in accordance with Article 12 of Law of Disaster Protection); renewal of a permit authorizing discharge of waste water effluent or air emission; assessment of existing contaminated sites for the purposes of remediation and risk management; and derivation of risk-based environmental standards, or remediation clean-up targets. The risk assessment process and this regulation are meant to be flexible to allow risk assessment principles to be applied and integrated with diverse scenarios and needs, consistent with the science and practices of experts in the field. In all cases, scientific defensibility is a necessary aspect of risk assessment. 1.3 Use of Different Forms of Risk Assessment Pursuant to Articles and of the Law on EIA, which form the requirement for risk assessment as part of a DEIA, the Terms of Reference (ToR) for Risk Assessment shall be developed per Article 3.2 of the Regulation on Risk Assessment (Appendix I). The ToR for Risk Assessment shall clearly state whether or not the fundamental approach to risk assessment shall be (i) contaminant human health risk assessment (HHRA), and/or (ii) contaminant ecological health risk assessment (ERA), and/or (iii) system failure mode (accident mode), effects, and criticality analysis (FMECA). Accordingly, this manual describes general methods and principles associated with the different forms of risk assessment. In brief, HHRA and ERA (hereafter collectively referred to as HHERA) are fundamentally similar; they assess how contaminants and either human or eco-receptors interact through exposure pathways, and estimate the potential for a health impact. The third method of risk assessment (FMECA) is notably different; it quantifies the likelihood of hazardous events and the severity of their adverse consequences to humans, the environment, and property. Risk assessment methods other than HHERA and FMECA may also be appropriate and if desired, should be discussed with regulatory staff to resolve the acceptability of the approach to meet EIA and regulatory objectives.

6 August MNET Risk Assessment Guide 1.4 Use and Structure of this Document The science and techniques associated with various steps of risk assessment are constantly evolving, and draw from a wide array of disciplines. Risk assessments are invariably conducted by a team with expertise in risk assessment and various additional disciplines (e.g,, field sampling, statistical analyses, engineering process design and equipment reliability assessment, surface and groundwater contaminant fate and transport modelling, air emission and dispersion modelling, ecology, eco-toxicology and human health toxicology). Accordingly, the present document is limited to fundamentals of risk assessment. Users who are new to risk assessment will need to develop further expertise by consulting the public literature and familiarizing with commercial numerical models or developing their own spreadsheet models. Foreign entities undertaking project development in Mongolia may refer to this document and the Risk Regulation to better understand and requirements to be fulfilled in completing a risk assessment as part of an EIA submission in Mongolia. The document is structured as follows: Section 2; Introduces the most fundamental concepts of risk, the two basic forms of risk assessment preferred by MNET, and the different objectives required to be fulfilled by these forms of risk assessment. Section 3: Describes the framework and basic methodological steps for human health risk assessment (HHRA) and ecological risk assessment (ERA). Because the HHRA and ERA frameworks are virtually the same, the steps for HHRA and ERA are discussed in juxtaposition within each stage of the risk assessment framework. Section 4: Describes the framework and basic methodological steps and requirements for Failure Mode Effects and Criticality Analysis (FMECA). Throughout the document reference is made to specific procedural requirements cited within the Risk Regulation, however the cross referencing is not exhaustive and users should consult the Risk Regulation for the complete requirements (Appendix 1).

7 August MNET Risk Assessment Guide 2.0 FUNDAMENTAL CONCEPTS OF RISK 2.1 Distinguishing Between Risk, Effects and Impacts Risk is a mathematical concept describing a set of possible conditions, it is therefore abstract in nature. It is predicated on a set of factors (risk factors) which may (or may not) evoke an event and an adverse consequence. The International Organization for Standardization (ISO 2009) notes risk is often expressed in terms of a combination of the consequences of an event (including changes in circumstances) and the associated likelihood of occurrence. Building on this concept, the working definition of risk for the Mongolian Regulation on Risk Assessment and this manual is: Risk - a quantitative or descriptive expression of the likelihood and severity of adverse effect (e.g., property loss, injury or health effect to people or ecological receptors) arising from a specific event or operation (risk scenario); risk may be further defined as being acceptable or unacceptable, based on regulatory or societal tolerance of the combination of the likelihood and severity of the adverse effect. This definition satisfies both contaminant health risk assessment (HHRA and ERA) and failure mode analysis (FMECA). Note that the expression of risk may be numerical (quantitative) or descriptive (i.e., qualitative); in both cases however the expressions of risk must relate to a meaningful benchmark. As described later, the numerical expression of risk used for contaminant health risk is benchmarked to an index with meaningful effects endpoint to allow interpretation of the severity of risk. Descriptive expressions of risk (e.g., low, high, extreme) require clear descriptive definitions to allow meaningful use and clarity of the relative differences between the terms. It is useful to clarify the terms effects and impacts and to distinguish their meaning versus the term risk: Effects should be viewed as the specific adverse/undesired outcomes that are, or could be, realized by a receptor due to an event. Effects are either real (i.e. they exist and can be seen, touched or otherwise experienced), or hypothetical (i.e., potential effects which might occur in the future, such as in the case of a proposed project). Effects may pertain to: i. Local or regional environment (e.g., various adverse ecological or human health effects),

8 August MNET Risk Assessment Guide ii. iii. Corporate operations (e.g., losses in productivity, reputation, financial performance), or Local or regional economy (e.g., sustainable livelihoods and commerce). Unlike risk, there is no concept of likelihood or probability connected to the definition of effect; the latter term simply refers to the adverse outcome that the receptor experiences, or could experience sometime in the future. For this reason, effect and risk should not be used synonymously. In contrast to the physical experience of effects, one does not physically experience risk (other than perhaps mental stress that may cause); the experience of risk is simply the day-to-day living or coexistence with a set of factors that might evoke an event or condition leading to physiological, physical or financial adverse effect(s). Impact is essentially synonymous with effect, but impact tends to be used in a more generic sense and embraces a wide array of non-specific effects, in contrast to specific effects (e,g., cancer or reproductive effects of a chemical, or financial loss effects of an earthquake). For this reason, the terms impact and risk should also not be used synonymously for reasons stated above. However, in the context of a regulatory EIA approval process, it is appropriate to consider unacceptable risks as impacts and to proactively manage these risk to an acceptable level to ideally avert future adverse effects. Uncertainty is an inherent part of risk, and arises because of (i) natural variability (also known as stochasticity) of the system or factors that govern the risk scenario, and/or (ii) the imprecise information about the risk factors used to numerically estimate the risk. Elimination of uncertainty from risk estimation is impractical if not impossible, however it can be reduced through iterative improvements of the risk model and risk variables. Reduction of uncertainty to improve risk estimates should only be done to a degree that yields adequate confidence when interpreting the risk estimates to making risk management decisions. 2.2 Objectives of Environmental Risk Assessment for Impact Assessment Under the Mongolian Law on An Environmental and Social Impact Assessment (ESIA), assessors quantify potential impacts of proposed new projects or major project upgrades in order to: i. identify and mitigate impacts at the design stage, and ii. obtain regulatory approval and permitting for the project.

9 August MNET Risk Assessment Guide Within this context, the Law provides for the use of risk assessment as one of the analytical approaches to assess impacts. The overall objective of the risk assessment component of an EIA should align with those of the EIA, and therefore should be: To assess the proposed project at the design stage to determine if project (i) construction, (ii) operation, (iii) decommissioning and (iv) post-decommissioned state would pose unacceptable risks to receptors. This involves defining the conditions (risk factors) and receptors that are integral components of the various risk scenarios that currently exist or are anticipated in the future, and then to define the probability and severity of adverse effects that could arise. In many environmental risk assessments, assessors do not actually derive or express the risk as separate components of probability and severity of effect, but instead may use a hybridized index that blends consideration of likelihood and severity of effect into either a single numerical or descriptive expression (illustrated in subsequent chapters). To achieve the overall objective, risk assessment requires thoroughness in defining meaningful and relevant risk scenarios, and then assessing those risk scenarios using realistic and somewhat conservative values for the input variables, so as to avoid underestimating the true risk. As previously noted, two forms of risk assessment are recommended for use, contaminant health risk assessment (HHERA) and failure mode effects and criticality analysis (FMECA). The selection of either or both should be based on the type of project hazards (e.g., significant/insignificant chemical use), and discretion of the risk assessor and Mongolian regulators. More definitive objectives are also pursued within the risk assessment, and these are dependent in part by the type of risk assessment implemented (HHERA versus FMECA discussed in the following section), and in part by the sitespecific nature of the project and stakeholder input during public consultation Projects with Chemical Use or Chemical Production If the project in question is a facility or operation that involves chemical use or chemical production, the potential for an uncontained release of chemicals to the environment is among the potential project hazards that could cause a future impact. This may be due to normal operations (e.g., effluent discharge to a river, stack emissions to atmosphere), or due to accidental conditions (e.g., equipment failure, human error). It is appropriate therefore to use the process of contaminant health risk assessment to assess potential future impacts to human and ecological receptors for situations where chemicals may be potential hazards. The chemicals in question do not need to be highly toxic to justify the use of HHERA; substances with relatively low toxicity, but nevertheless released in high quantity may still be intolerable (e.g., effluent with high salinity causing osmotic stress and shock to aquatic biota). Therefore, one should avoid the premature exclusion of

10 August MNET Risk Assessment Guide contaminant health risk assessment within an ESIA; the risk assessment process itself provides an early step to rationalize whether certain chemicals should or should not be considered in the risk assessment. When implementing HHERA within an EIA, a set of core key questions must be addressed as part of the objectives which requirements under the Risk Regulation (Table 1). Note that these objectives must integrate with the overall risk assessment objective, and therefore be applied to each of the project phases (i.e., construction, operation, decommissioning and post decommissioning) Table 1. Core Key Questions for Human and Ecological Contaminant Health Risk Assessment as a Component of Environmental Impact Assessment HHKQ1: What effect will project releases have on water quality and subsequently human health? HHKQ2: What effect will project releases have on air quality and subsequently human health? HHKQ3: What effect will project releases have on soil quality and subsequently human health? HHKQ4: What effect will project releases have on food quality and subsequently human health? HHKQ5: What will be the collective effect of changes to water, air, soil and food on human health? Notes: HH human health; ER ecological risk; KQ Key Questions ERKQ1: What effect will project releases have on water quality and subsequently ecological health? ERKQ2: What effect will project air releases, and particulate deposition to soil have on ecological health? ERKQ3: What effect will project releases have on soil quality and subsequently ecological health? ERKQ4: What effect will project releases have on food resource quality and ecological health? ERKQ5: What will be the collective effect of changes to water, air, soil and food on ecological health? Projects with Hazardous System Components. If the project in question is a facility or operation involving extensive physical structures or industrial processing equipment, these could present health and safety hazards to workers or possibly a local community under normal or potential upset conditions. Upset condition may arise from equipment failure, human error or through external factors such as weather or seismic activity. To understand the risk to health, safety, or property loss it is appropriate to use the process of Failure Mode Effects and Criticality Analysis (FMECA). When implementing FMECA within an EIA, a set of core key questions must be addressed as part of the objectives which are requirements under the Risk Regulation (Table 2). As for HHERA, these objectives must integrate with the overall risk

11 August MNET Risk Assessment Guide assessment objective, and therefore be applied to each of the project phases (i.e., construction, operation, decommissioning and post decommissioning) Table 2. Core Key Questions for Failure Mode Risk Assessment as a Component of Environmental Impact Assessment FMKQ1: What are the plausible accidental events (scenarios) that are of prime concern to cause impact to (i) community health and safety, (ii) environment, (iii) workers, and (iv) plant operations? (consider human factors, engineered equipment/systems failure, and natural/environmental factors)? FMKQ2: What are the event probabilities and consequence severities for each of the scenarios considered in FMKQ1? FMKQ3: Which of the scenarios and impacts warrant risk mitigation? FMKQ4: How many people, livestock, and threatened species could potentially be impacted by the risk scenarios noted in FMKQ3? FMKQ5: What are the recommended mitigation measures for the risk scenarios identified in FMKQ3? FMKQ6: What would be the residual estimated risks if the recommended mitigation measures were implemented? Notes : FM failure mode; KQ Key Question Additional key questions may be warranted as objectives depending on the site, projectspecific details, and stakeholder feedback during the EIA public consultation. 3.0 FRAMEWORK AND METHODOLOGY FOR CONTAMINANT HEALTH RISK ASSESSMENT 3.1 Framework for HHERA The framework for methodology of HHRA an ERA are similar and illustrated in Figure 1 and 2. Each feature the fundamental stages of Problem Formulation, Exposure Analysis, Toxicity Analysis, and Risk Characterization, and reflect the typical framework that is widely practiced in Europe, North America, Australia and parts of South America. Although initially derived to guide risk assessment of existing contaminated sites, the frameworks are equally useful for use in EIA of proposed new projects

12 August MNET Risk Assessment Guide Questions typically addressed by the assessor within HHERA process for EIA are: 1. Who/what are the relevant human and ecological receptors relevant to future project scenarios? 2. Are there contaminants already present under baseline condition that pose potential human health and/or ecological concerns (e.g., at concentrations above applicable regulatory criteria for the protection of health and the environment)? 3. How might human or ecological receptors be exposed to the contaminants? 4. What types of adverse effects might result from exposure to the contaminants? 5. Based on a quantitative analysis of contaminant exposure and toxicity, what is the magnitude of health risks to human and ecological receptors within the local and regional study area? 6. How do current and future proposed activities influence the predicted risks? 7. Is the risk likely to remain stable, increase, or decrease with time (e.g., for a mine project could acid? 8. If warranted, where, and in what conceptual manner should risk reduction measures be implemented? 9. If warranted what methods should be used to ensure that the risk reduction measures are effective (e.g., monitoring program)? Addressing these questions provides insight for mitigating excessive risks, should they arise, during the project design stage while the EIA is in process; this is a requirement per articles 5.4 and 5.5 of the Risk Regulation: 5.4 Where significant unacceptable risks exist within the context of a proposed but unapproved Project, recommendations shall be provided by lead risk assessor to the Project Implementer to modify the proposed Project to mitigate risks, prior to submission of the DEIA. 5.5 When risk mitigation is warranted, Project Implementer shall re-evaluate project design for feasibility of modifications that will reduce risks to acceptable levels, prior to completion and submission of DEIA. If supplemental mitigation strategies are adopted, they should be described within the DEIA and the risk assessment should summarize the mitigation realized for the risk scenarios.

13 August MNET Risk Assessment Guide Figure 1. Human Health Risk Assessment Framework (Health Canada, unpublished, 1995)

14 August MNET Risk Assessment Guide Figure 2. Ecological Risk Assessment Framework (USEPA 1992) Note that although approaches are very similar, ecological risk assessment includes additional considerations during the Problem Formulation respecting and Analysis phases to address the assessment/measurement endpoints (i.e., the level of effect to tolerated) and the ecosystem/habitat which may be affected.

15 August MNET Risk Assessment Guide Problem Formulation is the first and perhaps the most important stage of environmental health risk assessment. The objective is to screen contaminant sources, receptors and pathways to focus subsequent steps. Recognizing that both time and monetary resources are limited, these screening steps are designed to determine key exposure pathways and chemicals. Objectives of the risk management and therefore the risk assessment are also articulated at this early stage to promote proper design of the risk assessment. The result is a Conceptual Exposure Model which focuses the resources on those contaminants, receptors and pathways which are relevant to the project and risk management issues. Exposure Assessment is the process of estimating the dose rate (intake rate) or exposure concentration to which a human or ecological receptor is subjected. For virtually all animals (including humans) the primary routes of contaminant exposure are through inhalation, ingestion, and dermal (or trans-dermal) absorption. For fish, exposure is primarily via gill uptake, but ingestion may be important for chemicals which biomagnify through the food chain. The relative importance of each of these exposure routes will vary depending on the receptors selected. In the case of a newly proposed project which is not yet constructed, Various mathematical equations and fate and transport models are used to predict the future exposure concentrations and dose rates. In case of baseline conditions, the exposure concentrations are measured directly during the baseline investigation. Toxicity Assessment for human health risks involves classifying the contaminants in accordance with their potential toxic effects, and identifying the acceptable dose or concentration that can be received by a person without experiencing measurable adverse health effects (i.e. the exposure limit or toxicity reference value). The basic principles applied in human health toxicity assessments also apply to ecological toxicity assessments, however, carcinogenicity is rarely considered. In the case of ecological receptors, the selection of appropriate toxicity reference values relates to the desired level of protection that is to be given to ecological receptors (i.e in accordance with the assessment and measurement endpoints defined within the Problem Formulation). Risk Characterization involves numerical estimation of the effect magnitude (i.e., severity), an expression of its likelihood, and a descriptive interpretation of the estimated risks associated with exposure to contaminants of concern. Health risks are estimated by comparing the predicted exposure(s) to the acceptable toxicity reference values. For threshold-acting contaminants (i.e. substances where an exposure threshold must be achieved to cause and effect), the human and non-human risk expression is known as the Hazard Quotient (HQ). For human exposure to non-threshold-acting carcinogens (i.e. genotoxic substances which theoretically require no threshold to cause an effect), a numerical cancer risk estimate is calculated. The same mathematical principles used generate the risk estimate can be used to derive what is sometimes called a risk-based

16 August MNET Risk Assessment Guide criterion that defines a contaminant concentration (e.g., in soil) that would yield acceptable health risk to the receptor; this is particularly useful for risk reduction planning. When implementing this approach, it is common practice to make conservative exposure and toxicity assumptions in order to ensure calculated risks are not underestimated. This practice is called a screening level assessment and is typically carried out using a "deterministic" approach, in which single, conservative values are used as input parameters to the quantitative analysis. If the screening level assessment indicates that risks are acceptable, then a more detailed analysis is not necessary. If the screening level assessment indicates a potential problem, then the risk assessor should conduct additional iterations of the analysis with more realistic assumptions about exposure and toxicity to determine if the refined calculations support the original findings and warrant mitigation. The following sections provide additional insight in performing the fundamental steps of the above phases. 3.2 Problem Formulation Methods Human Health Problem Formulation Prior to Problem Formulation and the EIA baseline field investigations, the lead risk assessor must provide early Problem Formulation perspectives to the various leaders of the field sampling crews regarding baseline chemistry data that will needed to formally conduct the Problem Formulation; this is a requirement under Article 3.6 of the Risk Regulation in order assure high quality data is acquired for the risk assessment. Details shall be stated in a Technical Memorandum (Article 3.6) and should identify the desired environmental media to be sampled, sample locations, analytes of interest, required detection limits and other desired data quality objectives. Formal initiation of Problem Formulation should begin only after the baseline data has been acquired and the data quality validated. Problem formulation is primarily qualitative and identifies the key contaminants of potential concern, the key potential human or ecological receptors, and the significant exposure pathways that may combine to form risk scenarios (Figure 3).

17 August MNET Risk Assessment Guide Figure 3. Schematic Illustration - Conceptual Relationship of the Components of Human Health Risk Screened during Problem Formulation Chemical Hazard Receptor Pathway Risk During this stage, the main issues of the project and surrounding site are compiled and evaluated based on knowledge of the site, current, future and adjacent land use, baseline concentrations of chemical substances, presence and lifestyle of local residents, and potential for exposure pathways to exist between the future local/regional human receptors and chemicals of potential concern (COPC). Receptor Screening The receptor screening process should identify people who are currently living in, or using, areas in the vicinity of the project site. This is most easily assessed by also considering current and future landuse near the project site. Residents within the local and regional study areas are considered receptors, but those living on the project site and who are expected to be relocated are not considered receptors at that future location. Receptors should include both adults and children; children are considered to be more sensitive to some chemicals than adults and at certain ages have a greater intake rate to body weight ratio due to behaviour activities (e.g., playing in soil; Health Canada 2004a). Chemical Screening For assessment of baseline conditions, chemical screening should be based on comparison between measured concentrations of chemicals in air, water, soil, sediment and food (e.g., crops, livestock, fish, etc.) and relevant environmental quality guidelines. It is necessary to first compute descriptive statistics of the baseline chemical data. The data may need to be grouped according to media, climate, seasonality and geographical locations as relevant to the project and site. Typical parameters to be computed include mean (arithmetic and/or geometric mean as warranted), 95% upper confidence level of the mean (95UCLM), 90%iles and maximum values. These may be computed using

18 August MNET Risk Assessment Guide spread sheet software programs, or by using the USEPA program ProUCL (available for download from the USEPA Technical Support Center for Monitoring and Site Characterization). Chemicals with concentrations exceeding a relevant regulatory guideline (either as the mean, multiple individual sample exceedances, or as the 95% upper confidence limit of the mean) can be considered a chemical of potential concern (COPC) and retained for further risk analysis. Where only one or two exceedance occur and the remaining samples are within guidelines it is unlikely that the substance is a COPC, and it is prudent to verify the sample concentrations where exceedance were noted. A suggested approach for chemical screening applicable to future project scenario(s) is as follows: Step 1: Compare of the predicted concentrations (future chemical levels during the Project construction, operations, decommisiong and/or post-decommisioning phases) to measured baseline concentrations. If the predicted concentrations are the same or very similar (within 10%) to the baseline concentrations, then the project is unlikely to have a significant incremental impact on human health. Step 2: Compare the predicted concentrations of any chemicals that significantly exceed the baseline concentrations, as determined by Step 1, to Mongolian environmental quality guidelines; where Mongolian guidelines are not available, use those form other reputable agencies such as WHO, European Union, Environment Canada, Health Canada, and USEPA). Only chemicals with predicted concentrations more than 10% above baseline levels and above environmental guidelines need to be considered as COPCs and carried forward in the analysis. Further details on estimation of future concentrations for various environmental media are provided in Appendix 2. Exposure Pathway Screening The objective of the exposure pathway screening process is to identify potential routes by which people could be exposed to chemicals and the relative significance of these pathways to the total exposure. A chemical represents a potential health risk only if it can reach receptors through an exposure pathway at a concentration that could potentially lead to adverse effects. If there is no pathway for a chemical to reach a receptor, then there cannot be a risk, regardless of the chemical concentration (Figure 3). Potential exposure pathways include ingestion, transdermal absorption and inhalation.

19 August MNET Risk Assessment Guide Screening human exposure pathways simply requires methodical consideration of (i) expected receptor behaviour (based partly on landuse), (ii) physical-chemical properties of the contaminant to determine its transport behaviour., and (iii) the media in which the contaminant will reside. Using this insight the risk assessor must decide whether the receptor and contaminant are likely to interact through an operable pathway; if so then the pathway is included in the subsequent risk assessment steps Ecological Problem Formulation Similar to HHRA, Problem Formulation for ERA focuses on the chemicals, receptors and exposure pathways of greatest concern. Therefore similar screening methods and principles are employed as described above. A key difference however, is that ecological risk assessment establishes assessment endpoints which provide the basis for judging resultant risks against a desired ecological state. Assessment endpoints are statements that reflect the level of protection to be conferred to the ecological receptors (e.g., no loss of species populations due to degradation of water quality). The concept of variable protection levels does not apply to HHRA because complete protection of individual humans is mandatory. Because assessment endpoints are often difficult to quantify due to their descriptive nature, measurement endpoints are used to provide a numerical and more practical basis to define and interpret whether the goal of the assessment endpoint is being achieved. The measurement endpoint is a statement if the effect level that will be tolerated within the realm of risk outcomes. For example, studies may show that a species is able to tolerate a 15% reduction in growth or reproduction with no apparent impact to overall population dynamics. In such a case, the dose or exposure concentration causing a 15% effect in toxicity studies may then be used as the safe toxicity reference value (TRV) that is employed to compute the final risk estimate. In accordance with Article of the Risk Regulation, the lead risk assessor will define within the Problem Formulation the acceptable risk levels, and therefore for ecological risk assessment there must be clear statements of both assessment and measurement endpoints. As general guidance for ERA undertakings, MNET prefers measurement endpoints be limited to inhibitory effect levels of 0 to 15% for growth or reproduction, and justification should be provided for inhibitory effect levels that are greater than zero. Where the ecological receptor is an endangered species, the inhibitory effect level must be zero.

20 August MNET Risk Assessment Guide Ecological Receptor Screening Representative receptors are those that have the greatest potential for exposure, that play a key role in the food web and that have sufficient characterization data to facilitate calculations of exposure and health risks.. In accordance with Article of the Risk Regulation, ecological receptors shall be selected to satisfy the following considerations: native species representative of other species in the same feeding guild; important component of the ecological food web in the local study area; economically important (e.g., livestock, agricultural crop); rare or endangered species, or a suitable surrogate for such species; and at least one representative receptor from each major or relevant feeding guild in the local food web (e.g., herbivore, omnivore, insectivore, carnivore). To initiate ecological receptor screening, an ecologist and risk assessor should first consider the type of biogeoclimatic zone in which the project site exists. Then, based on the habitat defined in the biogeoclimatic zone, a list of resident species should be compiled using existing publications and/or site reconnaissance. Once the species list is compiled, the entries are compared to the selection criteria listed above and appropriate selection are made and summarized as exemplified in Table 3. Table 3: Example Summary Table of Wildlife Species Selected as Key Receptors Type of receptor Common name (Latin name) Diet Conservation status e.g., Aquatic mammal Species A diverse carnivorous diet, including of crustaceans, mollusks, fish, frogs, and other small vertebrates and invertebrates found in the shallow waters shoreline e,g, International Union for Conservation of Nature (IUCN) Red List Category Least Concern e.g., Terrestrial mammal e.g., Terrestrial mammal Species B Species C mainly insectivorous, but it has a diverse diet omnivorous, generally feeding mainly on insects, fruits, small vertebrates, carrion, and plant material e.g.,. IUCN Red List: Category - Least Concern e.g., IUCN Red List Category Least Concern e.g., Terrestrial bird Species D insectivorous IUCN Red List Category Least Concern

21 August MNET Risk Assessment Guide Chemical Screening Statistical summaries of exposure concentrations and the two-step screening process described for the human health assessment (i.e., comparison of predicted concentrations to baseline conditions as well as Mongolian and international environmental guidelines) are also appropriate for ecological chemical screening. Estimation of future concentrations for various media were previously noted in Appendix 1. The key difference between the human and ecological chemical screening processes are the selection of environmental quality guidelines against which to judge the ecological relevance (i.e., water criteria to protect aquatic biota, etc) 3.3 Exposure Assessment Human Health Exposure Assessment Exposure assessment for human health risk assessment may be based on various scientifically valid intake equations which quantify the average daily dose (intake rate) normalized to body mass unit. The approach and equations for different intake pathways described by Health Canada 2004(a) are presented in Table 4 below as an example of acceptable methods. The approach derives an estimate of the average daily intake of contaminant per unit body mass (i.e., units of mg/kg-day). There are two variations in the calculation of the average daily intake and this is to accommodate the difference in risk computations required for non-carcinogenic versus carcinogenic substances. For non-carcinogens requiring threshold exposures to evoke toxic effects, the process simply evaluates the number of hours, days and weeks along with pathway-specific intake rates and exposure concentration to estimate the total average daily intake rate from all relevant pathways. If the receptor were to move away from the project site and exposure ceased, then the intake rate and the risk would also cease (i.e., exposure becomes zero and risk becomes zero). There would be no consideration for subsequent hours, days or weeks where exposure is zero or below the threshold that causes an effect, because it will be irrelevant to the risk due to lack of effect. In contrast, for carcinogenic substances that cause mutations and self-propagating genotoxic injury, the termination of exposure does not cause cessation of the selfpropagating genotoxic injury. The risk of genotoxic toxic injury and risk of cancer continues as cells divide to make more mutated cells (note: the rate of injury and risk are likely less due to cessation of external exposure). Accordingly, the estimation of average daily intake rate and cancer risk for genotoxic substances requires averaging the exposure

22 August MNET Risk Assessment Guide period over the human receptor s lifetime to allow computation of the lifetime-averaged daily intake rate. This exposure parameter amortized over a lifetime mirrors the exposure parameter used in experimental animal studies which relate lifetime-averaged daily intake rate to the observed incidence of cancer within the animal test population. The parameter that relates the dose rate and cancer incidence is cancer slope factor (SF) and is used with the human lifetime-averaged daily intake rate to estimate incremental lifetime-averaged cancer risk (ILCR). Examination of the equations below will illustrate two parameters in the equations which enable lifetime-averaging of the daily intake rate (parameters D3, and LE). If the scenario does not involve a genotoxic carcinogen, then these variables are NOT zeroed, but rather eliminated from the equation (see notes in Table 4). Table 4. Summary of Human Health Exposure Equations (Health Canada 2004a) Pathway Water Ingestion Soil Ingestion Food (vegetables and fish) Ingestion D water IR x C = water Equation and Equation Parameters AF GIT BW LE D1 D2 D3 D water = administrated dose due to ingestion of water (mg chemical/kg body weight-day) IR = ingestion rate (L/day) C water =chemical concentration in water (mg/l) AF GIT = Absorption factor from the gastrointestinal tract D1 = Days per week exposed/7 days D2 = Weeks per year/52 weeks D3 = Total years exposed to site (carcinogenic assessment only) BW = receptor body weight (kg) LE = Life expectancy (carcinogenic assessment only) D soil IR C = soil AF GIT BW LE D1 D2 D3 D soil = administrated dose due to ingestion of soil (mg chemical/kg body weight-day) IR = ingestion rate (kg/day) C soil =chemical concentration in soil (mg/kg) AF GIT = Absorption factor from the gastrointestinal tract D1 = Days per week exposed/7 days D2 = Weeks per year/52 weeks D3 = Total years exposed to site (carcinogenic assessment only) BW = receptor body weight (kg) LE = Life expectancy (carcinogenic assessment only) D food IR C = food AFGIT D BW LE c D

23 August MNET Risk Assessment Guide Pathway Equation and Equation Parameters D food = administrated dose due to ingestion of food (mg chemical/kg body weight-day) IR = ingestion rate (L/day) C food =chemical concentration in food (mg/kg) AF GIT = Absorption factor from the gastrointestinal tract D c = Days per year exposed D = Total years exposed to site (carcinogenic assessment only) BW = receptor body weight (kg) Inhalation Of Contaminated Soil Particles Dermal Contact With Soil LE = Life expectancy (carcinogenic assessment only) D inhal_soil Cs P = Air IR A AFinh D1 D2 D3 D4 BW LE D inhal_soil = administrated dose due to inhalation of contaminated soil particles (mg chemical/kg body weight-day) Csoil = chemical concentration in soil (mg/kg) P air = particulate concentration in air (kg/m 3 ) IR A = inhalation rate (m 3 /h) AF Inh = Inhalation absorption factor D1 = hours per day exposed (h/days) D2 = Days per week exposed/7 days D3 = Weeks per year/52 weeks D4 = Total years exposed to site (carcinogenic assessment only) BW = receptor body weight (kg) LE = Life expectancy (carcinogenic assessment only) D dsoil Cs SA = H SL H AF skin BW LE EF D1 D2 D3 D d_soil = dose due to dermal contact with soil (mg chemical/kg body weight-day) SA H= skin surface area exposed (cm 2 ) SL H= soil loading to exposed skin (kg/cm 2 -event) AF skin = dermal absorption factor (unitless) EF = exposure frequency (events/year) D1 = Days per week exposed/7 days D2 = Weeks per year/52 weeks D3 = Total years exposed to site (carcinogenic assessment only) BW = receptor body weight (kg) LE = Life expectancy (carcinogenic assessment only) To the extent possible, values for input variables should reflect site and project conditions; for example contaminant exposure concentrations in different media should be based on meaningful predictive models as exemplified in Appendix 1. Additionally,

24 August MNET Risk Assessment Guide intake equations should reflect receptor exposure variables (e.g., frequency of exposure, body weight, location of exposure, diet, source of water, etc.). Where site-specific receptor variables cannot be adequately characterized, professional judgment and/or suggested values described in Health Canada (2004a) may be used with discussion of the uncertainty and applicability of these value. The exposure assessment and reporting component must satisfy Article of the Risk Regulation: statistical metrics of exposure concentrations should include the mean, 95% upper confidence limit of the mean (95UCLM), and 90 th percentile; daily intake rates shall be computed for each COPC and exposure pathway; the exposure analysis shall consider all relevant exposure media (e.g., surface water, groundwater, soil, air, fish, native plants, crops, and livestock), and relevant ecological food chains; mathematical models used to assist in the exposure analysis (including contaminant release, fate and transport, receptor contact and receptor intake) shall be clearly described, and scientifically defensible for the application; exposure variables shall, to the extent practical, be site-specific in nature; and where exposure assumptions are based on expert judgment, the rationale and an estimate of the margin of error introduced to the risk estimate shall be reported in a section on uncertainty analysis Ecological Exposure Assessment Exposure Assessment of Fish, Free Swimming Invertebrates and Aquatic Plants Exposure assessment for free swimming aquatic biota and aquatic plants located in the water column is typically based on direct exposure concentrations anticipated in the dissolved phase of the water column. The dissolved contaminant concentration rather than total concentration generally provides a more realistic measure of bioavailable exposure. Final exposure concentrations should be based on best descriptive statistics (arithmetic or geometric mean, 95UCLM, and 90 th %ile) derived by appropriately segregating baseline field data according to variation in season, catchment areas, and meaningful geographic locations. For the future risk scenarios, predicted water columns concentrations shall be based on baseline values plus relevant modelled incremental mass inputs attributed to the project (e.g., via groundwater, surface water and atmospheric particulate deposition). Discussion of professional judgment shall be provided when is it is a key consideration in resolving predicted water concentrations, per Article of the Risk Regulation.

25 August MNET Risk Assessment Guide Dose Estimation for Terrestrial and Aquatic Wildlife The following equations illustrate the principles for total dose estimation by terrestrial and aquatic wildlife from water, soil/sediment and food. These equations illustrate the basic process which is suitable for a screening level risk estimate in EIA. The procedure can be made more sophisticated by including additional variables to increase realism of exposure, as was done in human exposure assessment using the absorption factor term (AF GIT) for incomplete contaminant absorption across the gastrointestinal tract (provided that the toxicity reference term is also adjusted for bioavailability. Additional instruction, input values and equations for extrapolating to different animal groups may be derived from Sample et al. (1997) For the terrestrial receptors: For aquatic wildlife receptors: D = D + D total soil food D = D + D total water food Where: Unit Definition D total mg/kg-day = Total dose D soil mg/kg-day = Dose via soil ingestion D water mg/kg-day = Dose via water ingestion D food mg/kg-day = Dose via food ingestion For estimation of dose via water ingestion: Where: D = C I water water water Unit Definition Values used D water mg/kg-day = Dose of a chemical via water ingestion C water mg/l = Concentration in water Measured (baseline assessment) and predicted (future scenarios) in water I water L/kg body weight-day = Water ingestion rate Estimated (Sample et al. 1997) Water ingestion rate (I water ) may be estimated as (Sample et al. 1997):

26 August MNET Risk Assessment Guide Where: 0, 0,099( BW ) = (mammals) BW I water 90 Unit Definition Value used I water L/kgbody weight-day = Water ingestion rate BW Kg = Body weight Literature For estimation of dose via food ingestion (for each food item): Where: D food = C I F food food food Unit Definition Values used D food mg/kg-day = Dose via food ingestion C food mg/kg dry weight = Concentration in food (i.e., fish) Measured (baseline assessment) and predicted (future I food Kg food [dry weight]/body weightday scenarios) in fish = Food ingestion rate Estimated based on body weight F food - = Fraction of the diet composed by a specific food item Assumed based on diet description In the absence of species-specific food ingestion rates, this parameter may be scaled according to Sample et al., 1997: 0, 0,0687( BW ) = (placental mammals) BW I food 822 0, 0,0582( BW ) = (birds) BW I food 651 Where: Unit Definition Values used I food Kg food [dry weight]/body weight-day = Food ingestion rate Estimated based on body weight BW Kg = Body weight Literature