Draft Screening Assessment

Petroleum Sector Stream Approach

Heavy Fuel Oils
[Industry-Restricted]

Chemical Abstracts Service Registry Numbers
64741-75-9
68783-08-4
70592-76-6
70592-77-7
70592-78-8

Environment Canada
Health Canada

April 2012

---

(PDF Version - 592 KB)

Table of Contents

• Synopsis
• Introduction
• Substance Identity
• Physical and Chemical Properties
• Sources
• Uses
• Releases to the Environment
• Environmental Fate
• Persistence and Bioaccumulation Potential
• Potential to Cause Ecological Harm
• Potential to Cause Harm to Human Health
• Conclusion
• References
• Appendix

Synopsis

The Ministers of the Environment and of Health have conducted a screening assessment of the following industry-restricted heavy fuel oils (HFOs):

CAS RN[a]	Domestic Substances List Name
64741-75-9	Residues (petroleum), hydrocracked
68783-08-4	Gas oils (petroleum), heavy atmospheric
70592-76-6	Distillates (petroleum), intermediate vacuum
70592-77-7	Distillates (petroleum), light vacuum
70592-78-8	Distillates (petroleum), vacuum

These substances were identified as high priorities for action during the categorization of the Domestic Substances List (DSL) as they were determined to present greatest potential or intermediate potential for exposure of individuals in Canada, and were considered to present a high hazard to human health. All of these substances met the ecological categorization criteria for persistence or bioaccumulation potential and inherent toxicity to aquatic organisms. These substances were included in the Petroleum Sector Stream Approach (PSSA) because they are related to the petroleum sector and are all complex mixtures.

HFOs are a group of complex petroleum mixtures that serve as blending stocks in final heavy fuel products or as intermediate products of distillation or residue derived from refinery distillation or cracking units. The final fuel products usually consist of a mixture of HFOs as well as higher-quality hydrocarbons as diluents. HFOs are composed of aromatic, aliphatic and cycloalkane hydrocarbons, primarily in the carbon range of C₁₄–C₅₀ (C₇ is the smallest hydrocarbon found in the group), and have a typical boiling point range of 121°C to 600°C. As such, HFOs are considered to be of Unknown or Variable composition, Complex reaction products or Biological materials (UVCBs). In order to predict the overall behaviour of these complex substances for the purposes of assessing the potential for ecological effects, representative structures have been selected from each chemical class in the mixtures.

The HFOs considered in this screening assessment have been identified as industry restricted (i.e., they are a subset of HFOs that may leave a petroleum sector facility and be transported to other industrial facilities). According to information submitted under section 71 of the Canadian Environmental Protection Act, 1999 (CEPA 1999), and other sources of information, these HFOs are transported in large volumes from refinery or upgrader facilities to other industrial facilities by pipelines, ships, trains and trucks; therefore, exposure of the environment is expected.

Based on the available information, all HFOs assessed in this report likely contain significant amounts of components (C₁₀–C₅₀) that meet the criteria for persistence in soil, water and sediment as defined in the Persistence and Bioaccumulation Regulations of CEPA 1999. Based on the combined evidence of empirical and modelled bioaccumulation potential data, all HFOs assessed in this report likely contain large proportions of C₁₄–C₂₀components that meet the criteria for bioaccumulation potential as defined in the Persistence and Bioaccumulation Regulations.Some of the components of these HFOs (C₁₅ dicycloalkanes, C₁₄ and C₂₂polycycloalkanes, C₁₅-C₂₀ cycloalkane monoaromatics, C₂₀ cycloalkane diaromatics and C₂₀ three-ring aromatics) were found to meet the criteria for both persistence and bioaccumulation potential as defined in the Regulations.

Based on results of comparison of levels expected to cause harm to organisms with estimated exposure levels and the relatively low expected frequency of spills to water and soil during loading/unloading and transport operations, these five HFOs are not expected to cause harm to aquatic or soil organisms. The estimated releases to marine waters are also not expected to endanger seabirds.

Based on the information presented in this screening assessment, it is proposed that the industry-restricted HFOs (Chemical Abstracts Service Registry Numbers [CAS RNs] 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8) are not entering the environment in a quantity or concentration or under conditions that have or may have an immediate or long-term harmful effect on the environment or its biological diversity or that constitute or may constitute a danger to the environment on which life depends.

A critical effect for the initial categorization of industry-restricted HFO substances was carcinogenicity, based primarily on classifications by international agencies. Several cancer studies conducted in laboratory animals resulted in the development of skin tumours following repeated dermal application of HFO substances. It is unknown whether HFOs are carcinogenic via the inhalation route. HFOs demonstrated genotoxicity in in vivo and in vitro assays and may also adversely affect reproduction and development in laboratory animals.

General population exposure to industry-restricted HFOs results primarily from inhalation of ambient air containing HFO vapours due to evaporative loss during transportation. Due to the relatively low volatility of the HFO substances, as defined by their physical-chemical properties, losses into the air are expected to be minimal. The margins between the upper-bounding estimate of exposure, the maximum air concentration of HFOs (1.28 µg/m³), and the critical inhalation effect levels are considered to be highly conservative and adequately protective to account for data gaps and uncertainties in the human health assessment for both cancer and non-cancer effects. In general, the likelihood of inhalation exposure of the general population is considered to be low; thus, the risk to human health is likewise considered to be low.

General population exposure to industry-restricted HFOs via the oral and dermal routes is not expected; therefore, risk to human health from the industry-restricted HFOs via these routes is not expected.

Based on the information presented in this screening assessment, it is proposed that the industry-restricted HFOs (CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8) are not entering the environment in a quantity or concentration or under conditions that constitute or may constitute a danger in Canada to human life or health.

It is therefore proposed that the five industry-restricted HFOs listed under CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8 do not meet any of the criteria set out in section 64 of CEPA 1999.

As substances listed on the DSL, their import and manufacture in Canada are not subject to notification under subsection 81(1) of CEPA 1999. Given the potentially hazardous properties of these substances, there is concern that new activities that have not been identified or assessed could lead to these substances meeting the criteria set out in section 64 of the Act. Therefore, application of the Significant New Activity provisions of the Act to these substances is being considered. This would require that any proposed new manufacture, import use or transport be subject to further assessment, to determine if the new activity requires further risk management consideration.

Introduction

The Canadian Environmental Protection Act, 1999 (CEPA 1999) (Canada 1999) requires the Minister of the Environment and the Minister of Health to conduct screening assessments of substances that have met the categorization criteria set out in the Act to determine whether these substances present or may present a risk to the environment or to human health.

Based on the information obtained through the categorization process, the Ministers identified a number of substances as high priorities for action. These include substances that:

• met all of the ecological categorization criteria, including persistence (P), bioaccumulation potential (B) and inherent toxicity to aquatic organisms (iT), and were believed to be in commerce in Canada; and/or
• met the categorization criteria for greatest potential for exposure (GPE) or intermediate potential for exposure (IPE) and had been identified as posing a high hazard to human health based on classifications by other national or international agencies for carcinogenicity, genotoxicity, developmental toxicity or reproductive toxicity.

A key element of the Government of Canada’s Chemicals Management Plan is the Petroleum Sector Stream Approach (PSSA), which involves the assessment of approximately 160 petroleum substances that are considered high priorities for action. These substances are primarily related to the petroleum sector and are considered to be of Unknown or Variable composition, Complex reaction products or Biological materials (UVCBs).

Screening assessments focus on information critical to determining whether a substance meets the criteria set out in section 64 of CEPA 1999. Screening assessments examine scientific information and develop conclusions by incorporating a weight-of-evidence approach and precaution^[1].

Grouping of Petroleum Substances

The high-priority petroleum substances fall into nine groups of substances (Table A1.1 in Appendix 1) based on similarities in production, toxicity and physical-chemical properties. In order to conduct the screening assessments, each high-priority petroleum substance was placed into one of five categories (“streams”) depending on its production and uses in Canada:

1. substances concluded not to be relevant to the petroleum sector and/or not in commerce;
2. site-restricted substances, which are substances that are not expected to be transported off refinery, upgrader or natural gas processing facility sites^[2];
3. industry-restricted substances, which are substances that may leave a petroleum sector facility and be transported to other industrial facilities (e.g., for use as a feedstock, fuel or blending component), but do not reach the public market in the form originally acquired;
4. substances that are primarily used by industries and consumers as fuels;
5. substances that may be present in products available to the consumer.

An analysis of the available data determined that 16 petroleum substances are evaluated under Stream 2, as described above^[3]. These occur within five of the nine substance groupings: heavy fuel oils (HFOs), gas oils, petroleum and refinery gases, low boiling point naphthas and crude oils.

This screening assessment addresses five industry-restricted HFO substances described under Chemical Abstracts Service Registry Numbers (CAS RNs) 64741-75-9, 68783­08­4, 70592-76-6, 70592-77-7 and 70592-78-8. These substances were identified as GPE or IPE during the categorization exercise based on production volumes reported in the Domestic Substances List (DSL), and were considered to present a high hazard to human health. All of these substances met the ecological categorization criteria for persistence or bioaccumulation potential and inherent toxicity to aquatic organisms and were believed to be in commerce in Canada during the categorization of the DSL. According to information submitted under section 71 of CEPA 1999 (Environment Canada 2008, 2009), these substances can be consumed on-site or transported from refineries and upgraders to other industrial facilities, but they are not sold directly to consumers. These substances were included in the PSSA because they are related to the petroleum sector and are all complex mixtures.

Seven site-restricted HFOs were previously assessed under Stream 1, and the remaining HFOs (nine different CAS RNs) will be assessed separately, as they belong to Streams 3 and 4 (as described above). The health effects of the industry-restricted HFOs were assessed using toxicological data pooled across all high-priority HFOs due to insufficient data specific to the industry-restricted HFOs.

Included in this Stream 2 screening assessment is the consideration of information on chemical properties, hazards, uses and exposure, including the additional information submitted under section 71 of CEPA 1999. Data relevant to the screening assessment of these substances were identified in original literature, review and assessment documents, and stakeholder research reports and from recent literature searches, up to March 2010 for the human exposure and environmental sections of the document and up to September 2011 for the health effects section of the document. Key studies were critically evaluated, and modelling results were used to reach conclusions.

Characterizing risk to the environment involves the consideration of data relevant to environmental behaviour, persistence, bioaccumulation and toxicity, combined with an estimation of exposure of potentially affected non-human organisms from the major sources of release to the environment. To predict the overall environmental behaviour and properties of complex substances such as these industry-restricted HFOs, representative structures were selected from each chemical class contained within the substances. Conclusions regarding risk to the environment are based on an estimation of environmental concentrations resulting from releases and the potential for these concentrations to have a negative impact on non-human organisms. As well, other lines of evidence including fate, temporal/spatial presence in the environment, and hazardous properties are taken into account. The ecological portion of the screening assessment summarizes the most pertinent data on environmental behaviour and effects and does not represent an exhaustive or critical review of all available data. Environmental models and comparisons with similar petroleum mixtures may have assisted in the assessment.

Evaluation of risk to human health involves consideration of data relevant to estimation of exposure (non-occupational) of the general population, as well as information on hazards (based principally on the weight-of-evidence assessments of other agencies that were used for prioritization of the substances). Health effects were assessed using toxicological data pooled across high priority HFO substances. Decisions for human health are based on the nature of the critical effect and/or margins between conservative effect levels and estimates of exposure, taking into account confidence in the completeness of the identified databases on both exposure and effects, within a screening context. The screening assessment does not represent an exhaustive or critical review of all available data. Rather, it presents a summary of the critical information upon which the proposed conclusion is based.

This screening assessment was prepared by staff in the Existing Substances Programs at Health Canada and Environment Canada and incorporates input from other programs within these departments. The human health and ecological portions of this assessment have undergone external written peer review/consultation. Comments on the technical portions relevant to human health were received from scientific experts selected and directed by Toxicology Excellence for Risk Assessment (TERA), including Dr. Michael Dourson (TERA), Dr. Stephen Embso-Mattingly (NewFields Environmental Forensics Practice, LLC), Dr. Susan Griffin (United States Environmental Protection Agency [U.S. EPA]) and Dr. Donna Vorhees (Science Collaborative). Although external comments were taken into consideration, the final content and outcome of the screening assessment remain the responsibility of Health Canada and Environment Canada.

The critical information and considerations upon which the draft screening assessment is based are summarized below.

Substance Identity

HFOs are a group of complex petroleum mixtures that serve as blending constituents in final fuel products or as intermediate products of distillate or residue derived from refinery distillation or cracking units (Table A2.1 in Appendix 2) (CONCAWE 1998). The final fuel product usually consists of a blend of HFOs and high-quality hydrocarbons that have been produced in the refinery or upgrader facilities. HFOs are complex mixtures of varying properties, with a typical boiling point range of 121°C to 600°C (Table A2.2 in Appendix 2; API 2004), and are composed of aromatic, aliphatic and cycloalkane hydrocarbons, mainly in the carbon range of C₂₀–C₅₀ (CONCAWE 1998). The ratio of aliphatic to aromatic hydrocarbons is important for estimating environmental behaviour; however, very few data exist for these five CAS RNs, so a ratio of 50:50 has been assumed. This ratio will not bias results and is within the range of other types of HFOs (50–79% aromatics) (ATSDR 1999; API 2004). CAS RN 68783-08-4 contains constituents with a carbon number as low as C₇ (CONCAWE 1998) (Table A2.1 in Appendix 2).

Physical and Chemical Properties

The composition and physical-chemical properties of HFOs vary depending upon the sources of crude oils or bitumen and the processing steps involved. A summary of experimental data on the physical-chemical properties of industry-restricted HFOs is presented in Table 1.

Table 1. General Experimental Physical-chemical Properties of Industry-restricted HFOs

Property	Value	Temperature (°C)	Reference
Pour point (°C)	< 30		API 2004
Boiling point (°C)	121–600		API 2004
Density (kg/m3)	900–1100	20	API 2004; MSDS 2007
Vapour pressure (Pa)	282.6–3519.6	21	Rhodes and Risher 1995
Log Koc (dimensionless)	3.0–6.7		Rhodes and Risher 1995
Log Kow (dimensionless)	2.7–6.0	20	API 2004
Water solubility (mg/L)	< 100	20	API 2004

The theoretical vapour pressures of individual substances comprising HFOs are low to moderate due to their high molecular weights. However, the actual vapour pressures will be influenced by the substance composition of the HFO mixture in which they occur. Water solubilities of all HFOs are low, and octanol–water partition coefficient estimations vary considerably, probably due to the complex nature of these mixtures.

To predict the environmental behaviour and fate of complex petroleum products such as these HFOs, representative structures were selected from each chemical class contained within the mixture. Forty-seven structures were selected from a database in PETROTOX (2009) based on boiling point ranges for each HFO (Table A2.3 in Appendix 2), the number of data on each structure and the middle of the boiling point range of similar structures. As the compositions of most HFOs are not well defined and are indeed variable, representative structures could not be chosen based on their proportion in the mixture. This resulted in the selection of representative structures for alkanes, isoalkanes, one-ring cycloalkanes, two-ring cycloalkanes, polycycloalkanes, cycloalkane monoaromatics, cycloalkane diaromatics and one-, two-, three-, four-, five- and six-ring aromatics ranging from C₉ to C₅₀ (Table A2.4 in Appendix 2). Physical-chemical data for each representative structure were assembled from scientific literature and from the group of environmental models included in the U.S. EPA’s Estimation Programs Interface Suite (EPIsuite 2008) (Table A2.4 in Appendix 2).

Sources

Industry-restricted HFOs are produced in Canadian refineries and upgraders. The CAS RN descriptions (NCI 2006) and typical process flow diagrams (Hopkinson 2008) indicate the origin of these HFOs. Information submitted under section 71 of CEPA 1999 shows that these substances can be intermediate streams consumed within a facility or be transported off-site by pipeline, truck, train and ship for use as a feedstock in other industrial facilities or for disposal (Environment Canada 2008, 2009).

CAS RN 64741-75-9 is a residual fraction from distillation of hydrocracking effluents in both refineries and upgraders.

CAS RN 68783-08-4 is a general description of distillates from atmospheric distillation of crude oil in refineries, primarily ranging from C₇ to C₃₅.

CAS RNs 70592-76-6, 70592-77-7 and 70592-78-8 have slight differences in their dominant carbon range, but they all refer to distillates from vacuum fractionation of the residue produced from atmospheric distillation of crude oil.

Uses

According to the information collected through the Notice with respect to certain high priority petroleum substancespublished under section 71 of CEPA 1999 (Environment Canada 2008) and the Notice with respect to potentially industry-limited high priority petroleum substances (Environment Canada 2009), these industry-restricted HFO substances have been identified as being consumed at the facility or transferred to another industrial facility for use as feedstock or for disposal. Although these substances were identified by multiple use codes established during the development of the DSL, it has been determined from information submitted under section 71 of CEPA 1999, voluntary submissions from industry, an in-depth literature review and a search of material safety data sheets that these industry-restricted HFOs (i.e., the CAS RNs identified in this screening assessment) may leave a refinery or an upgrading facility and be transported to another industrial facility for use as a feedstock, fuel or blending component or for disposal, but do not reach the public market in the form originally acquired.

Releases to the Environment

Potential releases of industry-restricted HFOs consist of releases within facilities from activities associated with processing these substances, as well as releases related to transportation of these substances between industrial facilities.

Due to the complex nature of the petroleum industry and transportation industry, as well as the ambiguity in the literature in the use of the terminology that is critical to the understanding of the Stream 2 PSSA assessments, it is important that the definitions specific to the assessment of the industry-restricted petroleum substances are well understood. Table 2 lists the terminology specific to the present assessment.

Table 2. Definitions of Terms Specific to the PSSA Assessments of Industry-restricted Petroleum Substances

Terminology	Definition
Release	A generic term to define a leak, spill, vent, or other release of a gaseous or liquid substance, including controlled release and unintentional release, as defined below, but not including catastrophic events.
Controlled release	Any planned release into a closed system for safety or maintenance purposes that is considered part of routine operations and occurs under controlled conditions.
Unintentional release	Any unplanned release of a petroleum substance. Causes can include equipment failure, poor maintenance, lack of proper operating practices, adverse weather-related events or other unforeseen factors, but can also be a routine part of normal operations. The following two categories are included under unintentional releases: (1) unintentional leaks or spills that occur from processing, handling and transport of a petroleum substance; such leaks or spills can be reduced or controlled by the industry; and (2) accidental releases that may not be controllable by the industry. Only unintentional leaks or spills (category 1 defined the above) are considered in the assessment of the potential of industry-restricted petroleum substances to cause ecological harm.
Fugitive release	A specific type of unintentional release. It refers to an unintentional release, which occurs under normal operating conditions, of a gaseous substance into ambient air and which may occur on a routine basis. Fugitive releases can be reduced but may not be entirely preventable due to the substance’s physical-chemical properties, equipment design, and operating conditions. Evaporative loss during the transportation of petroleum substances is a fugitive release and is considered in the human exposure analysis for purposes of assessing the potential of the substance to cause harm to human health.

Potential On-site Releases

Potential releases of HFO substances from refineries or upgraders can be characterized as either controlled or unintentional releases. Controlled releases are planned releases from pressure relief valves, venting valves and drain systems for safety purposes or maintenance. Unintentional releases are typically characterized as unplanned releases due to spills or leaks from various equipment, valves, piping or flanges. Refinery and upgrader operations are highly regulated, and regulatory requirements are established under various jurisdictions. As well, voluntary non-regulatory measures implemented by the petroleum industry are in place to manage these releases (SENES 2009).

Controlled Releases

The industry-restricted HFOs considered in this screening assessment originate from distillation columns in a refinery or an upgrader, either as a residue (bottom product) or as a distillate. Thus, the potential locations for the controlled release of these HFOs include relief valves, venting valves and drain valves on the piping or vessels where these streams are generated.

Under typical operating conditions, controlled releases of industry-restricted HFOs under the CAS RNs identified in this screening assessment would be captured in a closed system^[4], according to defined procedures, and returned to the processing facility or to the facility’s wastewater treatment plant. In both cases, exposure of the general population or the environment is not expected.

Unintentional Releases

Unintentional releases (including fugitive releases) occur from equipment (e.g., pumps, storage tanks), seals, valves, piping, flanges, etc. during processing and handling of petroleum substances and can be greater in situations of poor maintenance or operating practices. Regulatory and non-regulatory measures are in place to reduce these events at petroleum refineries and upgraders (see Appendix 3) (SENES 2009). Rather than being specific to one substance, these measures are developed in a more generic way in order to reduce unintentional releases of all substances in the petroleum sector.

Conclusion for Potential On-site Releases

Based on the information presented in this screening assessment and previously in the screening assessment of the Stream 1 (site-restricted) HFOs, exposure of the general population or the environment to the on-site releases (controlled or unintentional) of industry-restricted HFOs is not expected.

Potential Releases from Transportation

As these industry-restricted HFOs can be transported between facilities, releases may also occur during transportation. In general, three operating procedures are involved during the process of transportation: loading, transit and unloading.

The on-site handling of petroleum substances for transportation is often regulated at the federal and provincial/territorial levels with legislation covering loading and unloading (see Appendix 3).

Storage of industry-restricted HFOs may be required before they are transported off-site. Releases of HFO vapours from the storage tanks into the air are expected to be small because the HFOs have low volatility. All relevant releases from storage, including leaks, spills and breathing loss (expulsion of vapour due to changes in temperature and pressure), will be similar to the aforementioned potential on-site releases and will be managed under the relevant legislation currently in place.

Tanks or containers for transferring petroleum substances are typically dedicated vessels; thus, washing or cleaning is not required on a routine basis (U.S. EPA 2008a; OECD 2009). As such, exposure of the general population and the environment to the HFOs considered in this screening assessment from tank cleaning is not expected.

Release Estimation

Information on the transportation quantities and relevant transportation modes was collected under section 71 of CEPA 1999 (Environment Canada 2009) with respect to each CAS RN considered in this screening assessment. Four modes of transportation--ships, pipelines, trucks and trains--were identified as being involved in moving industry­restricted HFOs to other industrial facilities. The total transport quantity of the five HFOs considered in this report is less than 3 million tonnes (year 2006). Further details on volumes and modes of transport for each CAS RN are considered to be confidential business information, but have been taken into consideration in conducting this assessment.

Two types of potential releases occur during transportation and are considered in this screening assessment. These are evaporative losses and unintentional releases (e.g., spills or leaks) during the handling and transit processes.

Evaporative loss is similar to breathing loss of organic substances from storage tanks. The quantity lost depends on the volatility of the substances, temperature or pressure changes that occur during transportation, and tightness of transport vessels and settings of valves. Ambient air is the receiving medium for evaporative losses.

Evaporative losses to the environment were considered in transportation by ships, trucks and trains and were estimated based on empirical equations from the U.S. EPA (2008a), physical-chemical properties (e.g., vapour pressure, molecular weight and density of vapours) of these HFO substances, and the annual transported quantities. No evaporative losses are considered for pipeline systems, as typical releases are generated as a result of leaks through seals, flanges and valves and are defined as unintentional releases.

Unintentional releases of the HFO substances due to spills generally enter water or soil, depending on the modes of transportation involved. Due to the relatively low volatility of the HFOs, as defined by their physical-chemical properties, losses into the air from spills would occur in a lower proportion compared with the proportions entering water or soil.

Potential releases associated with the transport of these HFOs to marine, freshwater and soil environments were assessed through analysis of historical spill data (2000–2009) from the Environment Canada Spill Line database (Environment Canada 2011). There was no spills category for HFOs; spills of Bunker C fuel oil were therefore used. The releases labelled as Bunker C fuel oil (fuel oil No. 6) would also include these industry­restricted HFOs. There were also a small number of releases that were generically labelled as just “Bunker,” and there was no indication as to what specific type of Bunker was released. Thus, all releases labelled as “Bunker” were also considered to be Bunker C fuel oil. Bunker C is considered a heavy fuel oil but is not industry­restricted and has a wider distribution. Thus, it is expected that the actual number and volume of industry-restricted HFO spills are considerably lower than those of Bunker C fuel oil spills, but this could not be reliably determined. Of note was the large-volume spill of 730 000 L in 2005, which is known to be a Bunker C fuel oil spill into Lake Wabamun, Alberta; it was not included in the release estimate, as it was known not to be an industry-restricted HFO spill. As well, any extremely large spills with no known origin were not included, as these were likely from environmental emergencies training exercises, which are not differentiated from actual events in the Environment Canada Spill Line database (Environment Canada 2011). Spills where collisions, poor road conditions and/or adverse weather-related events were listed as a source, reason or cause of spill were not included in the release estimate, as they are not considered preventable with regard to loading/unloading and transport of these HFOs.

Many of the individual reports had no estimate of the volume released into the environment. In order to account for the underestimation of the volume released, the estimated total volumes were extrapolated by assuming that the distribution of reported volumes released was representative of all releases (Table A4.1 in Appendix 4). From 2000 to 2009, the extrapolated total of spills of HFOs to all media (soil, salt water and fresh water) was 2.4 million litres from 339 spills (Table A4.1 in Appendix 4).

The historical spill data were also separated into the specific compartment affected, so that the estimated average release quantity per spill to each compartment could be determined. Within each compartment, a similar extrapolation was conducted to account for reported spills with no associated volumes. The estimated average quantities of these HFOs released per spill to fresh water and salt water from ship transport are shown in Table 3. Because these HFOs are handled the same as Bunker C fuel oil for loading/unloading and ship transport, it is assumed that an average spill volume of these CAS RNs would be the same as for Bunker C. There is no distinction in the database as to whether the spills occur during loading, transport or unloading. Thus, the average spill volume will be used for each of the scenarios.

Table 3. Average Release Quantities Per Spill of Industry-restricted HFOs to Various Compartments Based on Historical Bunker C Spill Data from the Environment Canada Spill Line Database (2000–2009)

Compartment Affected	Average Release Quantities Per Spill
	kg[a]	L[b]
Marine (salt water)	13 754	13 225
Fresh water	15 262	14 675
Soil	5105	4909

The largest fraction of HFO spills documented by Environment Canada from 2000 to 2009 affected land (130 incidents), followed by 108 releases to sea water and 53 releases to fresh water. For some reported spills, the compartment affected was not documented, whereas for others, multiple compartments were included; thus, this total does not equate to the total reported spills shown in Table A4.1 in Appendix 4. These numbers are considered to be a low estimate of actual releases, as not all provinces were reporting their spills to Environment Canada for all years, and some provinces have minimum reportable spill quantities. Releases to groundwater were not included in the analysis.

The Environment Canada (2011) Spill Line database provides three columns of data (sources, causes and reasons) for many releases of Bunker C fuel oil. The data in these columns were analyzed to determine how and why the majority of HFO releases occur (Tables A4.2a–c in Appendix 4).

The industrial areas where the majority of HFO releases occurred (see Table A4.2a in Appendix 4) were other watercraft (25% of the volume), pipelines (20% of the volume) and marine tankers (20% of the volume). Releases at storage depots and facilities accounted for about 2% of the volume, refineries accounted for 2%, tank and transport trucks accounted for 3%, trains accounted for 4% and “other” sources accounted for 9%. The majority of truck releases were in New Brunswick (50%), and the rest were reported in Newfoundland and Labrador, Nova Scotia, Quebec, Prince Edward Island, Ontario and British Columbia.

The Environment Canada Spill Line data were also analyzed for causes of HFO leaks (Table A4.2b in Appendix 4). It was found that pipe leaks accounted for 38% of the volume released, which is consistent with pipelines being a major source of Bunker C releases (see Table A4.2a). Likewise, sinking and grounding of vessels accounted for 13% and 6% of the total volume, respectively, which is also consistent with the high total spill volume by watercraft as a source. Twenty-five percent of the volume spilled was due to unknown causes, and 8% was due to “other” causes.

Analyzing reasons for releases, the data (Table A4.2c in Appendix 4) identified material failure as a major cause of releases, accounting for 16% of the volume released. Unknown reasons accounted for 43% of the volume, human error and negligence accounted for 18%, and fire and explosion accounted for 6% of the volume (from a single spill). The remaining 17% was divided over a wide variety of reasons.

For purposes of assessing the potential exposure of the environment from the transportation of industry-restricted HFOs, the ecological assessment focuses on unintentional releases to water, soil and air due to spills, although releases to water contributed significantly greater volumes than releases to soil and air. In comparison, assessment of potential exposure of the general population from transportation of industry-restricted HFOs focuses on evaporative loss, which occurs during regular operation activities. Although spills occur during transit and in loading or unloading operations, such releases are considered to occur on a non-routine or unpredictable basis in distinct locations and are therefore not considered in the assessment of exposure of the general population.

In addition, as relevant legislation and best practices are in place for on-site handling of these industry-restricted HFOs (see Appendix 3), non-occupational human exposure as a result of loading and unloading is not expected and is not considered in the human exposure assessment.

Environmental Fate

When petroleum products are released into the environment, four major fate processes will take place: dissolution in water, volatilization, biodegradation and adsorption. These processes will cause changes in the composition of these UVCB substances. In the case of spills on land or water surfaces, another fate process, photodegradation, can also be significant.

The rates of dissolution in water or volatilization of individual petroleum components are retarded by the complex nature of these petroleum mixtures. The solubility and volatility of individual components in mixtures are proportional to the solubility or volatility of the components in its pure state and its concentration in the mixture. Solubility and volatility of a component decrease when the component is present in a mixture (Banerjee 1984; Potter and Simmons 1998).

Each of the fate processes affects hydrocarbon families differently. Aromatics tend to be more water soluble than aliphatics of the same carbon number, whereas aliphatics tend to be more volatile (Gustafson et al. 1997). Thus, when a petroleum mixture is released into the environment, the principal water contaminants are likely to be aromatics while aliphatics will be the principal air contaminants (Potter and Simmons 1998). The trend in volatility by component class is as follows: alkenes ≈ alkanes > aromatics ≈ cycloalkanes. The most soluble and volatile components have the lowest molecular weight; thus, there is a general shift to higher molecular weight components in residual materials.

Biodegradation is almost always operative when petroleum mixtures are released into the environment. It has been widely demonstrated that nearly all soils and sediments have populations of bacteria and other organisms capable of degrading petroleum hydrocarbons (Pancirov and Brown 1975). Degradation occurs both in the presence and absence of oxygen. Two key factors that determine degradation rates are oxygen supply and molecular structure. In general, degradation is more rapid under aerobic conditions. Decreasing trends in degradation rates according to structure are

(1) n-alkanes, especially in the C₁₀ to C₂₅ range are degraded readily;

(2) isoalkanes;

(3) alkenes;

(4) benzene, toluene, ethylbenzene, xylenes (BTEX) (when present in concentrations that are not toxic to the micro-organisms);

(5) monoaromatics;

(6) polynuclear (polycyclic) aromatic hydrocarbons (PAHs); and

(7) higher molecular weight cycloalkanes, which may be very slow to degrade (Pancirov and Brown 1975).

These trends typically result in the depletion of the more readily degradable components and the accumulation of the most resistant in residues.

Level III fugacity modelling of representative hydrocarbons contained in the HFO group of substances was performed using EQC (2003) (Table A5.1 in Appendix 5) based on their physical-chemical properties as given in Table A2.4 of Appendix 2.

If released solely to air, all C₉ to C₁₅representative structures will remain in air. With an increase in molecular size, the proportion remaining in air declines. Some of the C₂₀ components will also remain primarily in air, except for alkanes, polycycloalkanes, cycloalkane monoaromatics, and four-, five- and six–ring PAHs. Moderate amounts of the C₃₀ isoalkanes (70%) will also remain in air with the same pattern of decreasing atmospheric retention with increasing molecular size (Table A5.1 in Appendix 5). Aside from the C₃₀ isoalkanes, the C₃₀ and C₅₀representative structures of HFOs will partition almost entirely to soil.

If released solely to water, most C₉ representative structures will remain in water, with the exception of alkanes, which will partition almost equally between sediment and water. The C₁₅ one- to three-ring aromatics will also undergo significant partitioning between sediments and water (12 to 49% into water), while all other representative structures will partition largely to sediment. Volatilization from water surfaces is not expected to be an important fate process despite the presence of some representative structures with moderate to very high estimated Henry’s Law constants. Thus, if water is a receiving medium, all HFOs are expected to have a large proportion of the mixture partitioning to sediment (Table A5.1 in Appendix 5). It is likely, with a release situation into water where the HFO is not immediately in contact with sediments or suspended matter, that the moderate to high Henry’s Law constants will drive the C₉ to C₂₀ representative structures out of the water. The tendencies for evaporation and sorption are competing and the exact nature of the release would dictate how the HFO behaves.

If released to soil, all representative structures of HFOs are expected to have high adsorptivity to soil (i.e., expected to be immobile with > 99% remaining in the soil). Competing with this tendency are evaporative forces. Volatilization from moist soil surfaces may be an important fate process based upon estimated Henry’s Law constant values of 5.1 to 1.3 × 10⁶Pa·m³/mol. Lower molecular weight representative structures of HFOs (alkanes, isoalkanes, cycloalkanes and one-ring aromatics) may slightly to substantially volatilize from dry soil surfaces based upon their moderate vapour pressures (Table A5.1 in Appendix 5).

Fugacity estimations in soil do not take into account situations where large quantities of a hydrocarbon mixture enter the soil compartment. Under these situations soil adsorption sites can be fully bound and the petroleum hydrocarbons create a saturated zone within which even heavy mixtures of hydrocarbons can move as a non-aqueous phase liquid (NAPL). Petroleum hydrocarbons in soil at over 10 000 mg/kg almost always form NAPLs, and some that are liquid at room temperature will form at 100 mg/kg (Charbeneau et al. 1998).

Persistence and Bioaccumulation Potential

Environmental Persistence

In water, hydrolysis half-lives could not be predicted for hydrocarbons using the HYDROWIN (2008) model. Alkanes, alkenes, benzenes, biphenyls, PAHs and heterocyclic PAHs are all known to be resistant to hydrolysis (Lyman et al. 1990).

Since no empirical data were available on the degradation of these HFOs as complex mixtures, a QSAR-based weight-of-evidence approach (Environment Canada 2007) was applied using the BIOHCWIN (2008), BIOWIN 3,4,5,6 (2009), CATABOL (c2004-2008) and TOPKAT (2004) biodegradation models (Table A5.2 in Appendix 5).

Primary biodegradation (estimated with BIOHCWIN and BIOWIN 4) is the transformation of a parent compound to an initial metabolite. Ultimate biodegradation (estimated with BIOWIN 3, 5 and 6, CATABOL and TOPKAT) is the transformation of a parent compound to carbon dioxide and water, mineral oxides of any other elements present in the test compound and new cell material (EPIsuite 2008). BIOHCWIN (2008) is a biodegradation model specific to petroleum hydrocarbons. Model results are in domain for all MITI-based models (BIOWIN 5 and 6).

For many of the C₉ to C₂₀ components, both the primary and ultimate biodegradation models in BIOWIN (2009) and BIOHCWIN agree that these compounds would degrade quickly and would not likely be persistent (Table A5.2 in Appendix 5). The following show persistence (half-life ≥ 182 days based on criteria in the Persistence and Bioaccumulation Regulations [Canada 2000]) in the environment: C₃₀–C₅₀ isoalkanes, C₃₀–C₅₀ one-ring cycloalkanes, C₁₅–C₅₀ two-ring cycloalkanes, C₁₄–C₂₂ polycycloalkanes, C₃₀–C₅₀ one-ring aromatics, C₁₀–C₂₀ cycloalkane monoaromatics, C₁₅–C₅₀ two-ring aromatics, C₁₂–C₂₀ cycloalkane diaromatics, C₂₀–C₅₀ three-ring aromatics, C₁₆–C₂₀ four-ring aromatics, C₂₀–C₃₀ five-ring aromatics and C₂₂ six-ring aromatics. Many of the C₅₀components were found to have extrapolated half-lives ≤ 182 days; however, BIOHCWIN (2008) indicates that these components do not degrade easily, with half-lives ≥ 182 days. Thus, these C₅₀ components are expected to be persistent based on primary degradation results from BIOHCWIN, as it is specific to petroleum hydrocarbons. The potential presence of these persistent components in each CAS RN is shown in Table A5.3 (Appendix 5).

Using an extrapolation ratio of 1:1:4 for a water : soil : sediment biodegradation half-life (Boethling et al. 1995), the half-life in soil for most heavy (> C₂₀) representative structures is also ≥ 182 days and the half-life in sediments is ≥ 365 days.

Based on a read-across approach of these HFOs to Fuel Oil No. 6 (Table A5.4, Appendix 5), the average weight percent of components that would be expected to be persistent ranges from 30 to 60%, based on samples of Canadian Fuel Oil No. 6 (Table A5.3, Appendix 5 and Table 4; Fuhr 2008).

Table 4. Proportion of Components in Each HFO that are Expected to be Persistent (Based on Fuhr 2008)

CAS RN	Average Weight % of Components that are Persistent[a]
64741-75-9	60
68789-08-4	30
70592-76-6	30
70592-77-7	30
70592-78-8	53

AOPWIN (2008) is a model that calculates atmospheric oxidation half-lives of compounds in contact with hydroxyl radicals in the troposphere under the influence of sunlight. Atmospheric oxidation rates were calculated for all of the representative structures. Although the low vapour pressures of these representative structures indicate that volatilization may not be a very significant fate process, oxidation half-lives of less than 1 day (Table A5.5 in Appendix 5) indicate that this would be a relatively rapid removal process if these substances were introduced into the atmosphere (Atkinson 1990; API 2004).

Persistence Conclusion

Based on results from AOPWIN (2008), there would be a relatively rapid removal process if these HFOs were introduced into the atmosphere, based on oxidation half-lives of less than 1 day. With regard to the primary and ultimate biodegradation modelling, the C₃₀–C₅₀ isoalkanes, C₃₀–C₅₀ one-ring cycloalkanes, C₁₅–C₅₀ two-ring cycloalkanes, C₁₄–C₂₂ polycycloalkanes, C₃₀–C₅₀ one-ring aromatics, C₁₀–C₂₀ cycloalkane monoaromatics, C₁₅–C₅₀ two-ring aromatics, C₁₂–C₂₀ cycloalkane diaromatics, C₂₀–C₅₀ three-ring aromatics, C₁₆–C₂₀ four-ring aromatics, C₂₀–C₃₀ five-ring aromatics and C₂₂ six-ring aromatics in these HFOs meet or exceed the criteria for persistence (half-life in soil and water ≥ 182 days and half-life in sediment ≥ 365 days). These HFOs are expected to contain a large proportion (about 30–60%) of components (C₁₀–C₅₀) that meet or exceed the persistence criteria as defined in the Persistence and Bioaccumulation Regulations (Canada 2000).

Potential for Bioaccumulation

Bioconcentration Factors (BCF) and Bioaccumulation Factors (BAF)

Experimental Studies

Since no experimental bioaccumulation or bioconcentration data for these HFOs as mixtures were available, empirical data on representative structures found in Fuel Oil No. 6 and other lighter hydrocarbon mixtures in a read across approach, and a predictive approach using a bioconcentration/bioaccumulation factor (BAF) model, were applied (Arnot and Gobas 2003, 2004). According to the Persistence and Bioaccumulation Regulations (Canada 2000) a substance is bioaccumulative if its BCF or BAF is ≥ 5000; however, measures of BAF are the preferred metric for assessing the bioaccumulation potential of substances. This is because BCF may not adequately account for the bioaccumulation potential of substances via the diet, which predominates for substances with log K_ow > ~4.5 (Arnot and Gobas 2003).

Neff et al. (1976) exposed clams (Rangia cuneata), oysters (Crassostrea virginica) and fish (Fundulus similus) to the water-soluble fraction of Fuel Oil No. 2 (2 ppm total naphthalenes) for 2 hours, followed by depuration of hydrocarbons for 366 hours. All organs examined showed rapid accumulation of naphthalenes within the 2-hour exposure period, with the gallbladder and brain of fish accumulating the highest concentrations. BAFs of naphthalenes in clams ranged from 2.3 to 26.7 L/kg (Table A5.6 in Appendix 5). Release of naphthalenes by fish began immediately following transfer to fresh water, reaching undetectable levels after 366 hours (~ 15 days).

Burkhard and Lukasewycz (2000) compiled data on uptake of polycyclic aromatic hydrocarbons (PAHs) in lake trout (Salvelinus namaycush) from three published works and derived BAFs. Measured BAFs for PAHs in these fish were 87, 1550 and 3990 L/kg for phenanthrene, fluoranthene and chrysene/triphenylene, respectively (Table A5.6 in Appendix 5).

Hardy et al. (1974) carried out an experiment giving cod (Gadus morhua) single doses of hexadecane (a C₁₆ alkane) in the diet and tracked metabolites. Entirely unchanged hexadecane was found in the liver. Hardy et al. (1974) suggest that such findings do not support high metabolic conversion of hexadecane in the liver of cod, and n-alkanes were preferentially deposited in liver over flesh of cod. However, the liver is the major site of chemical biotransformation, so higher concentrations in liver would be expected. Cravedi and Tulliez (1981) dosed rainbow trout with dodecyl cyclohexane (a C₁₈ alkyl cycloalkane) and studied its elimination and metabolism from the fish. Approximately 75% of the dose was absorbed. A major source of unmodified substance elimination was through the gills, with considerable amounts metabolized to a fatty acid and distributed throughout the body, with 14% excreted in urine (Cravedi and Tulliez 1981).

Cravedi and Tulliez (1983) also studied the dietary uptake of 1% C₁₃–C₂₂ n-alkanes in rainbow trout for 7 months. Trout were dosed with 10 000 ppm total alkanes in feed, and showed preferential fixation of C₁₃-C₁₄ n-alkanes in the adipose tissue. The mean accumulated mass of n-alkanes was 958 ppm per fish, so that a calculated BCF (diet) was 0.1. N-alkanes longer than C₁₆ were well retained (over 60% of accumulated n-alkanes remained after 8 weeks of depuration), while short-chain (< C₁₆) n-alkane concentrations decreased more rapidly (only 20–50% remained after 8 weeks of depuration).

Colombo et al. (2007) studied the bioaccumulation dynamics of C_12-C₂₅ n-alkanes and unresolved aliphatic hydrocarbons (UCM) in a detritivorous fish (Prochilodus lineatus) collected from the sewage-impacted Buenos Aires coastal area. Fish muscles contained large amounts of C₁₂–C₂₅ n-alkanes and UCM, reflecting the chronic bioaccumulation of fossil fuels from sewage particulates. The hydrocarbon composition in fish muscles was enriched in C₁₅–C₁₇ n-alkanes relative to fresh crude oil and settling particulates. The bioaccumulation factors (BAFs: 0.4–6.4 dry weight or 0.07–0.94 lipid-organic carbon) plotted against K_ow showed a parabolic pattern maximizing at C₁₄–C₁₈.

McCain et al. (1978) reported that 1- and 2-methyl naphthalene and 1,2,3,4-tetramethyl benzene were accumulated to a greater extent than other oil components in English sole (Parophrys vetulus) from oil-contaminated sediments. Tissue burdens of hydrocarbons decreased with increasing exposure time, such that after 27 days of exposure, only the liver had a detectable hydrocarbon burden. McCain et al. (1978) suggested that induction of the aryl hydrocarbon hydroxylase enzyme system eventually resulted in hydrocarbon removal.

Weinstein and Oris (1999) found that 4-day-old fathead minnows (Pimephales promelas) bioconcentrated fluoranthene (BCF 9054 L/kg) with only 24 hours exposure. They observed that the age of the fish likely impacted the ability to depurate fluoranthene and that older, more mature fish would be unlikely to bioacumulate PAHs. As well, this study may not have been long enough for fish to significantly degrade or depurate fluoranthene. De Maagd (1996) found a BCF of 3388 for fluoranthene in fathead minnows.

Guppies (Poecilia reticulata) bioconcentrated pyrene, producing BCFs in the range of 4786 to 11 300 (depending on type of test) after 48 hours of exposure, while lighter­weight PAHs had lower BCFs (1050–2238 for fluorene and 4550–7244 for anthracene) (De Voogt et al. 1991). However, the fish were capable of depurating pyrene completely within 160 hours of cessation of exposure; while anthracene was 70% depurated within 200 hours and fluorene was 20% depurated within 200 hours.

Jonsson et al. (2004) used a long-term (36-day) study to determine the bioconcentration of pyrene in sheepshead minnows (Cyprinodon variegatus). Fish reached a steady state after 4 to 7 days of exposure. The BCF was 2700, which was likely due to biotransformation of PAHs by the fish.

Unlike fish and some crustaceans, molluscs are unable to rapidly metabolize aromatic hydrocarbons. Accumulation can occur in stable tissue compartments with low hydrocarbon turnover and that are not readily exchangeable (Stegeman and Teal 1973; Neff et al. 1976).

Mollusc studies have typically found high potentials for the bioconcentration of PAHs. This may be caused by the relatively slow rates of depuration when compared to fish studies; however, the uptake of PAHs appears to occur fairly rapidly. Other works have shown that BCFs for PAHs in molluscs and some crustaceans are considerably higher than in fish (Table A5.7 in Appendix 5).

The zebra mussel (Dreissena polymorpha) showed fast uptake of benzo[a]pyrene and pyrene over 6 hours exposure which led to high BCF values (Bruner et al. 1994; Gossiaux et al. 1996). After 3 days of depuration, body concentrations of benzo[a]pyrene (B[a]P) had dropped to less than 50%, and after 2 weeks, the concentrations had been reduced to 5–20% (Gossiaux et al. 1996). Pyrene elimination was highly temperature dependent, with depuration occurring more rapidly at higher temperatures and occurring very slowly at colder temperatures (Gossiaux et al. 1996). Lipid content was also important to the bioconcentration values, with higher lipid contents accumulating PAHs more readily, whereas body size did not affect the BCF values (Bruner et al. 1994).

McLeese and Burridge (1987) studied the bioaccumulation potential of a number of saltwater invertebrates from PAH-spiked water or PAH-contaminated sediments. When PAHs were dissolved in water, fluoranthene, pyrene, triphenylene and perylene produced high BCF values in mussels (Mytilus edulis) and clams (Mya arenaria) after short (96-hr) exposures. However, when PAHs are present in the sediment, only mussels have a high potential for bioconcentration. All of these substances can be depurated from molluscs given time, but heavier PAHs (triphenylene and perylene) depurate more slowly than lighter PAHs (phenanthrene, fluoranthene and pyrene). Shrimps and polychaetes did not readily bioaccumulate PAHs.

Although some crustaceans can readily bioaccumulate higher-weight PAHs, they can also rapidly depurate PAHs. After 6 hours of exposure to B[a]P, the amphipod Pontoporeia hoyi and the freshwater shrimp (Mysis relicta) showed rapid uptake (Evans and Landrum 1989), but also had rapid depuration over 10–26 days (Evans and Landrum 1989). In Daphnia magna exposed to PAHs for 24 hours, bioconcentration was observed in 11 higher-weight PAHs, ranging from 6100 for chrysene to 50 000 for dibenz[ah]anthracene (Newsted and Giesy 1987). Depuration was not studied.

Invertebrates have also been shown to bioaccumulate certain petroleum hydrocarbons. Muijs and Jonker (2010) studied the bioaccumulation potential of the aquatic worm, Lumbriculus variegatus, after exposure to a series of 14 field-contaminated sediments (0.2 L of sediment in 0.7 L of tap water) with a known history of oil pollution for 49 days. A subsequent 28-day bioaccumulation factor was determined for all sediments. After 28 days of exposure, 70–90% of equilibrium was reached and uptake kinetics became slower with increasing boiling point of components. Equilibrium for the C₁₁–C₁₆ fraction reached a maximum after 14 days, but then decreased, while hydrocarbon fractions beyond C₃₄ may take > 90 days to reach equilibrium. Characterization of the accumulated hydrocarbons was not determined; however, alkanes from C₁₀ to C₃₄were identified in the aquatic worms. Muijs and Jonker (2010) suggest this may be due to the ingestion of organic matter to which the chemicals are sorbed. Depuration was not studied.

BCF values determined for various PAHs (Table A5.7 in Appendix 5) were highly variable, ranging from 180 to over 28 000. The majority of BCF studies on PAHs have found that bioconcentration can occur after short exposure times but that the majority of organisms also exhibit rapid depuration once the contaminant is removed. If the exposure to PAHs is limited in time, it is unlikely that bioaccumulation is actually occurring and that depuration rates will be the determining factor in PAH bioaccumulation potential. In the case of HFO spills in water, PAH release is not expected to be constant or continuous. However, some components have been shown to meet or exceed the persistence criteria. Despite the persistence of some HFO components, depuration is expected to occur in molluscs and crustaceans (Koyama et al. 2004).

In general, BCFs for PAHs in fish are low, in part because PAHs are metabolized by fish, resulting in very low or nondetectable concentrations of the parent PAHs in fish tissues (Varanasi et al. 1989). As PAHs tend to accumulate in sediments, benthic organisms may be continuously exposed to the contaminants. However, sediment-sorbed PAHs have only limited bioavailability to marine organisms (Salazar-Coria et al. 2007).

Three studies on BAFs of PAHs in aquatic organisms were found. Hence, experimental values of BAFs from the work of Neff et al. (1976), Zhou et al. (1997) and Burkhard and Lukasewyez (2000) were compiled for comparison with modelled data (Arnot and Gobas 2003) (Table A5.6 in Appendix 5). In general, the modelled values approximate the measured (Table A5.8 in Appendix 5) for the selected PAHs. None of the measured and modelled values were shown to be bioaccumulative according to the criteria (BAF ≥ 5000) in the Persistence and Bioaccumulation Regulations (Canada 2000), with the exception of the substituted PAH iso-heptyl fluorene (see Table A5.8 in Appendix 5).

Characterizing BCF/BAF

In characterizing bioaccumulation, the derivation of a BAF is preferred over a BCF since chemical exposure through the diet is not included in the latter (Barron 1990). BCFs are typically derived under laboratory controlled conditions. According to Arnot and Gobas (2006), the BCF is a poor descriptor of biomagnification in food webs because it is derived from laboratory experiments and does not include dietary exposure. Thus, BCFs based on laboratory studies have been shown to underestimate bioaccumulation potential or biomagnification of chemicals in the food web, as predators consume prey containing lipophilic compounds (U.S. EPA 1995). For very hydrophobic chemicals, dietary uptake is likely to be more important than absorption from water. Furthermore, laboratory BCFs have been shown to overestimate bioaccumulation potential when a chemical is bound or tightly adsorbed to sediment, i.e., less bioavailable.

Due to the scarcity of measured BAF values (Table A5.6 in Appendix 5), BCFs from various published works were compiled (Table A5.9a in Appendix 5) and used to help verify measured and modelled BAF values. In contrast to the few available experimental BAFs on PAHs, a suite of BCFs for components of HFOs were found, including alkanes, isoalkanes, two-ring cycloalkanes, one-ring aromatics, cycloalkane monoaromatics, cycloalkane diaromatics and polyaromatics (Table A5.9a in Appendix 5). Model estimates of these BCFs were also produced using a kinetic mass-balance model (Arnot and Gobas 2003) to fit the model kinetic elimination constants to agree with the observed BCF data in order to generate BAF predictions that reflect the known elimination rates.

A kinetic mass-balance model is, in principle, considered to provide the most reliable prediction method for determining bioaccumulation potential because it allows for correction of the kinetic rate constants and bioavailability parameters, when possible. BCF and BAF model predictions are considered “in domain” for this hydrocarbon assessment because it is based on first principles. As long as the mechanistic domain (passive diffusion), global parameter domain (range of empirical log K_ow and molecular weight) as well as metabolism domain (corrected k_M) are satisfied, predictions are considered valid (Arnot and Gobas 2003, 2006). The kinetic mass-balance model developed by Arnot and Gobas (2003, 2004) was employed using metabolic rate constants normalized to both conditions of the study and a representative middle trophic level fish as outlined in Arnot et al. (2008a, b) when the BCF or growth-corrected elimination rate constant is known. Both BCF and biomagnification factor (BMF) empirical data were used to correct default model uptake and elimination parameters, which are summarized in Appendix 5, Table A5.9b.

In Table A5.9b (Appendix 5), some metabolic rate constants calculated from the empirical BCF data were negative, suggesting that the metabolic rate is essentially zero and that other routes of elimination are more important. Accordingly, no metabolic rate correction was used when predicting the BCF and BAF for these structures. Gut and tissue metabolism is generally not regarded as an important elimination process for chemicals with log K_ow less than ~4.5 (Arnot et al. 2008a, b; Arnot and Gobas 2006), but this can depend on the size and lipid content of fish used in testing.

In Table A5.9a (Appendix 5), only the C₁₅ isoalkane (2,6,10-trimethyl dodecane), C₈ one­ring cycloalkane (ethyl cyclohexane), and C₁₃ two-ring aromatics (2-isopropyl naphthalene) had measured and/or modelled BCFs or BAFs ≥ 5000. However, the measured diaromatic (2-isopropyl naphthalene) that was found to be highly bioaccumulative contains the isopropyl functional group that is considered atypical in petroleum and requires a more thorough appraisal of reasonableness of model predictions based on available experimental information (Lampi et al. 2010). As well, Neff et al. (1976) found that the C₁₂ and C₁₃ diaromatics (alkylated naphthalenes and biphenyls) were not highly bioaccumulative in clams upon exposure to an oil-in-water dispersion of Fuel Oil No. 2. Thus, the combined weight-of-evidence suggests that these C₁₂ and C₁₃ diaromatics are not likely to be highly bioaccumulative. For the C₈ cyclohexane (ethyl cyclohexane), the predicted BAF (Arnot and Gobas 2004) for the middle trophic level fish is 5495, which just exceeds the criterion (BAF ≥ 5000), suggesting that it is bioaccumulative when all routes of uptake are considered. This prediction, however, was generated with a metabolic rate equal to zero because of the potential error associated with the estimate of metabolism rates (see Table A5.9b, Appendix 5). Factoring in metabolism, it is expected that the BAF would be lower and likely below 5000. As well, the experimental BCF suggests this C₈ cycloalkane is not highly bioaccumulative (Table A5.9a, Appendix 5). Combining these lines of reasoning, this suggests that this C₈cycloalkane is also not likely to be bioaccumulative according to the Canadian criteria. For the C₁₅ isoalkane (2,6,10-trimethyl dodecane), two predicted BAFs are presented (575 and 47 863). The latter BAF of 47 863 is preferred, as the depuration rate constant from the study was available to calculate the metabolic rate constant. This higher predicted BAF value is also in agreement with the slow rate of metabolism. Combining these lines of reasoning, this suggests that this C₁₅isoalkane is likely bioaccumulative according to the Canadian criteria.

Most components >C₂₀ have an estimated log K_ow > 8 and were excluded from the modelling, as predictions may be highly uncertain due to limitations of the model (Arnot and Gobas 2003). In Arnot and Gobas (2006), at a log K_ow of 8.0, the empirical distribution of “acceptable” fish BCF data shows that there are very few chemicals with fish BCFs exceeding the Canadian criterion of BCF ≥ 5000. Examination of Environment Canada’s empirical BCF/BAF database for DSL and non-DSL chemicals developed by Arnot and Gobas (2003) and further by Arnot (2005, 2006) shows that these are all highly chlorinated substances (i.e., decachlorobiphenyl, nonachlorobiphenyl, heptachlorobiphenyl), which have BCFs in the 10⁵ range, noting that octachloro naphthalene has a measured BCF of < 1000, (Fox et al. 1994, Gobas et al. 1989, Oliver and Niimi 1988) and all have log K_ow values < 8.0. Therefore, the predicted BCF and BAF values with log K_ow > 8 were considered out of the parametric domain of the Arnot­Gobas model (2003) and considered highly uncertain and not reliable.

BCF and BAF model estimates were also generated for an additional 26 C₉ to C₂₂ linear and cyclic representative structures using the BCFBAF (2008) model (Table A5.8 in Appendix 5), as no empirical bioaccumulation data were identified for these substances. Metabolism and dietary assimilation efficiency kinetics were corrected for these predictions based on analogue BCF and BMF test data. From this analysis, only one C₁₄ polycycloalkane was predicted to have a BCF that suggested a high bioconcentration potential. However, one isoalkane, several polycycloalkanes, one-ring cycloalkanes and one-, two-, and three-ring PAHs were found to have a high bioaccumulation factors. The log K_ow for these structures suggests that dietary uptake can predominate (up to 87% of total uptake) but will not be the sole route of exposure as, some substances are expected to have a 90% bioavailable fraction in the water column. BAF is therefore considered the most appropriate metric for assessing the bioaccumulation potential of these structures and represents a comparison of whole-body burdens compared with concentrations in water. The BCF and BAF predictions for these fractions are within the parametric, mechanistic and metabolic domains of the model and so are considered reliable.

Biomagnification Factors (BMF) and Trophic Magnification Factors (TMFs)

BMF values from ExxonMobil Biomedical Sciences Inc. (EMBSI), used to derive kinetic information for 15 substances, are reported in Table A5.9a (Appendix 5) (Lampi et al. 2010). None of these analogues have BMFs > 1, suggesting that these hydrocarbons will not biomagnify when compared to the concentrations expected in food items. A combination of metabolism, low dietary assimilation efficiency and growth dilution appear to limit the biomagnification potential of these compounds (see Tables A5.9a and A5.9b in Appendix 5).

Lampi et al. (2010) also summarized TMFs for PAHs from three field studies. The TMFs for various PAHs are summarized in Table A5.10 (Appendix 5). Broman et al. (1990) studied TMFs for 19 PAHs in a marine food chain; seston to mussels (M. edulis) to ducks (Somateria mollissima), and did not find TMFs >1.

Field-based TMFs for the PAHs studied are mostly < 1, except fluorene and acenaphthalene, which are approximately 1. A combination of metabolism, low dietary assimilation efficiency and growth dilution appear to limit the trophic magnification potential of these compounds as well. Therefore, it is not likely that the linear, cyclic and aromatic constituents of HFOs will undergo biomagnification or trophic magnification.

Biota-Sediment Accumulation Factors (BSAFs)

Lampi et al. (2010) also summarized the available BSAF data for several PAHs from a database compiled by the US Environmental Protection Agency (U.S. EPA 2008b). A box plot of fish field biota-sediment accumulation factors (BSAFs) for several PAHs is presented in Figure A5.1 (Appendix 5). Box boundaries represent 25^th and 75^th percentiles, the median is indicated by a line within each box, and whiskers denote the 10^th and 90^th percentiles of the data. Individual PAH data points ranged from n = 10 to n = 53.

From the box plot (Figure A5.1 in Appendix 5), none of the PAHs have fish BSAFs greater than one. This is expected given the same rationale for low BMF and TMF values. However, data were not extracted for invertebrate BSAFs from the U.S. EPA database. In the case of invertebrates, these factors can be much greater than one because invertebrates do not have the same metabolic competency as fish (e.g., B(a)P) (Muijs and Jonker 2010; Stegeman and Teal 1973; Neff et al. 1976).

As previously noted, Muijs and Jonker (2010) studied the bioaccumulation of oil in the aquatic worm, L. variegatus.Resulting BSAFs varied from 0.01 to 2.3. The wide range is likely related to the differences in oil weathering status. The BSAF values for separate hydrocarbon blocks appeared to be relatively constant up to C₂₂, indicating that L. variegatus proportionally accumulated these fractions from sediment. Beyond C₂₂, BSAFs decreased for all sediments studied and is likely due to the reduced bioavailability of the higher boiling point fractions such as PAHs. Likewise, there may be enhanced sorption of PAHs to sediment and in some cases, the nonaqueous phase liquid (NAPL). Muijs and Jonker (2010) also suggest that the studied aquatic worm may even avoid NAPLs, which may also limit the bioaccumulation of the very hydrophobic fractions.

Bioaccumulation Conclusion

As noted previously, of the parameters that have prescribed Canadian regulatory criteria, BAF values are preferred over BCF values, as they represent a comparison of whole-body burdens compared to concentrations in water. Biomagnification (BMF), trophic or foodweb magnification (TMF) and biota-sediment accumulation factors (BSAF) are also considered.

In general, the majority of < C₁₅ components (alkanes, isoalkanes, cycloalkane monoaromatics, cycloalkane diaromatics and three-ring PAHs) were not found to meet or exceed the Persistence and Bioaccumulation Regulations (Canada 2000). This conclusion is based on consistencies found between available BCF and BAF experimental data and BCF and BAF kinetic mass-balance model predictions using the Arnot-Gobas (2003) three trophic level model.

The majority of components ≥ C₂₀ (alkanes, isoalkanes, one-ring cycloalkanes, two-ring cycloalkanes, one-ring aromatics, two-ring aromatics, three- and five-ring aromatics) have estimated log K_ows > 8 and were therefore excluded from modelling, as predictions may be highly uncertain due to limitations of the model (Arnot and Gobas 2003). Likewise, for these ≥ C₂₀ components, no experimental measured BCFs were found. In Arnot and Gobas (2006), at a log K_ow of 8.0, the empirical distribution of “acceptable” fish BCF data shows that there are very few chemicals with fish BCFs exceeding the Canadian criterion of BCF ≥ 5000. As well, examination of Environment Canada’s empirical BCF/BAF database for DSL and non-DSL chemicals developed by Arnot and Gobas (2003) and further by Arnot (2006) shows that the substances exceeding the BCF ≥ 5000 are all highly chlorinated substances (i.e., decachlorobiphenyl, nonachlorobiphenyl, heptachlorobiphenyl), which have BCFs in the 10⁵ range, noting that octachloro naphthalene has a measured BCF of < 1000 (Fox et al. 1994, Gobas et al. 1989, Oliver and Niimi 1988) and all have log K_ow values < 8.0. Therefore, the predicted BCF and BAF values with log K_ow > 8 were considered out of the parametric domain of the Arnot-Gobas model (2003) and considered uncertain and not reliable. Thus, the combined evidence suggests that these components are not likely to be bioaccumulative.

Results from the Arnot-Gobas (2003, 2004) three trophic level model show that the C₁₈ polycycloalkane (hydro-chrysene) did not exceed the criterion of BCF or BAF ≥ 5000. The metabolic rate constant (0.45/day) for this particular component suggests a quick rate of metabolism in comparison to the lower metabolic rate constants (0.01/day and 0.04/day) for the C₁₄ and C₂₂ polycycloalkanes, which were found to exceed the criterion of BCF or BAF ≥ 5000 based on the same model (Table A5.8 in Appendix 5). Study details from experimental evidence for a similar polycycloalkane could not be obtained to determine predicted BCFs and BAFs, thus the available evidence suggests that the C₁₈ polycycloalkane is not bioaccumulative based on modelled results alone.

The C₁₄ and C₂₂ polycycloalkanes, C₁₅ one-ring aromatics, C₁₅–C₂₀cycloalkane monoaromatics and C₂₀ cycloalkane diaromatics were found to meet or exceed the bioaccumulation criteria based on modelled results from the Arnot-Gobas (2004) three trophic level model. For these particular components, the metabolic rate constants range from 0.01 to 0.08 (day^-1), suggesting a slow rate of metabolism. In the case of C₁₄ and C₂₂ polycycloalkanes, C₁₅ one-ring aromatic and C₂₀ cycloalkane monoaromatic, only experimental BMFs for comparative analogues were available. The BMFs were all < 1, suggesting that these components will not biomagnify. In the case of the C₁₅cycloalkane monoaromatic, only an experimental BCF (3418) for a similar component (octahydro­phenanthrene) was found. However, considering the slow metabolic rate, there is the potential that predicted BCFs and BAFs could exceed the Canadian criteria, although this cannot be determined due to the lack of details from the relevant study. Lastly, the only analogue similar to the C₂₀ cycloalkane diaromatic (iso-heptyl fluorene) is fluorene, which has an experimental BCF of 1030. However, the presence of an iso­heptyl group may affect the bioaccumulation potential of fluorine, and the low K_M value suggests a slow rate of metabolism. Overall, the available evidence suggests that these components are likely to bioaccumulate based on available modelled and experimental results.

BMF values for 15 substances comprising some isoalkanes, one- and two-ring cycloalkanes, polycycloalkanes, one-ring aromatics, cycloalkane monoaromatics, cycloalkane diaromatics and three- and four-ring aromatics (see Table A5.9a in Appendix 5) show that no components have BMFs > 1. This suggests that these particular hydrocarbons will not biomagnify when compared to concentrations expected in food items. Thus, the available evidence suggests that there is limited biomagnification of petroleum hydrocarbons. While only BSAFs were found for some PAHs, it is possible that BSAFs will be > 1 for invertebrates as they do not have the same metabolic competency as fish, but BSAFs will likely decrease beyond C₂₂ due to reduced bioavailability of the higher boiling point fractions.

With regard to PAHs, field-based TMFs were mostly < 1, with the exception of fluorene and acenaphthalene, which are approximately 1. It appears that biomagnification and trophic magnification are mitigated by a combination of metabolism, low dietary assimilation efficiency and growth dilution. Fish BSAFs were also < 1, but were only available for certain PAHs. The available evidence suggests that there is limited biomagnification and trophic magnification for PAHs. The majority of BCF studies on PAHs indicate that bioconcentration can occur after short exposure times but that the majority of organisms also exhibit rapid depuration once the contaminant is removed. This is especially evident in molluscs and crustaceans. In the case of HFO spills in water, PAH release is not expected to be constant or continuous, although some components (C₁₅–C₅₀) have been shown to meet or exceed the persistence criteria. Despite the persistence of some HFO components, depuration has been shown to occur in molluscs and crustaceans. For fish, BCFs and BAFs for PAHs are low, in part because PAHs are metabolized by fish, resulting in very low or nondetectable concentrations of the parent PAHs in fish tissues (Varanasi et al. 1989). As PAHs tend to accumulate in sediments, benthic organisms may be continuously exposed to the contaminants. However, sediment-sorbed PAHs have only limited bioavailability to marine organisms (Salazar-Coria et al. 2007). Thus, bioaccumulation of PAHs is not likely an issue for aquatic organisms; the exception to this would be for substances that may be both persistent and bioaccumulative.

Overall, there is consistent empirical and predicted evidence to suggest that 10 representative structures (C₁₅isoalkane, C₁₅ one-ring cycloalkanes, C₁₅two-ring cycloalkane, C₁₄ and C₂₂polycycloalkane, C₁₅ one-ring aromatic, C₁₅–C₂₀ cycloalkane monoaromatics, C₂₀ cycloalkane diaromatic and C₂₀ three-ring aromatic) meet or exceed the bioaccumulation criteria as defined in the Persistence and Bioaccumulation Regulations (Canada 2000). These components are associated with a slow rate of metabolism and are highly lipophilic. Exposures from water and the diet, when combined, suggests that the rate of uptake would exceed that of the total elimination rate. However, these components are not expected to biomagnify in aquatic or terrestrial foodwebs largely because a combination of metabolism, low dietary assimilation efficiency and growth dilution allows the elimination rate to exceed the uptake rate from the diet.

Based on the boiling point ranges of each individual CAS RN (Table A2.4 in Appendix 2), the proportions of components that would be expected to be bioaccumulative are shown in Table 5. These proportions are based on Canadian samples of Fuel Oil No. 6, as the chemical characterization of these industry-restricted HFOs is unknown (Table A5.11 in Appendix 5). A more detailed analysis of how these bioaccumulative proportions were determined is shown in Table A5.11 in Appendix 5.

Table 5. Proportion of Components in Each Industry-restricted HFO that are Eexpected to be Bioaccumulative (Based on Fuhr 2008)

CAS RN	Average Weight % of Components that are Bioaccumulative[a]
64741-75-9	6
68789-08-4	21
70592-76-6	21
70592-77-7	21
70592-78-8	21

Thus, up to about 21% of components by weight of these industry-restricted HFOs may be bioaccumulative (Table 6; based on Fuhr 2008).

Potential to Cause Ecological Harm

Ecological Effects Assessment

Information relevant to the toxicity of HFOs to various organisms is provided below. As well, PAHs are components of HFOs and have been considered in a previous regulatory assessment. PAHs are on the List of Toxic Substances under Schedule 1 of CEPA 1999 (Environment Canada 2010c).

Evidence from field and laboratory studies using field samples indicates that biota are adversely affected at various Canadian sites contaminated by PAHs of different industrial origins (Canada 1994).

There are potential hazards associated with the metabolism of PAHs such as benzo[a]pyrene. This process may create metabolites that are potent mutagens. Under laboratory conditions, neoplastic and genotoxic effects have been associated with exposure to PAHs for both terrestrial and aquatic organisms. In field studies, preliminary stages of chemically induced carcinogenesis have been shown (Environment Canada 1994).

Aquatic Compartment

No experimental data were available for the aquatic toxicity of these industry-restricted HFOs; therefore, data from Fuel Oil No. 6 were used in a read-across approach to estimate the potential for aquatic toxicity. Other studies have shown that, with HFOs, variations in aquatic toxicity exist, in part, due to differences in boiling point ranges determining the composition of the HFOs (ECB 2000b).

Table A5.12 in Appendix 5 presents Fuel Oil No. 6 acute toxicity data. Aquatic median lethal concentration (LC₅₀) values range from 0.9 to > 10 000 mg/L. Oil-in-water dispersions (OWD) have been shown to be not nearly as hazardous to aquatic organisms as the water-soluble fraction. The lowest marine toxicity value of 0.9 mg/L was determined in a 48-hour acute LC₅₀ test using water-soluble fractions with Mysidopsis almyra (a mysid shrimp) (Neff and Anderson 1981). The same value was determined in a 96-hour LC₅₀ with Capitella capitata (a marine worm) (Rossi et al. 1976). The lowest freshwater value of 1.69 mg/L was determined by Anderson et al. (1974) in a 96-hour LC₅₀ test with the fish Fundulus similis.

Avian Effects

HFOs can have a wide variety of effects on birds, especially sea birds; heavy oils, including HFOs, can destroy the insulation provided by feathers, resulting in increased mortality due to exposure. HFOs are also directly toxic to birds through ingestion. The preening of feathers to clean them of oil, and the reduced insulation from oiled feathers increases metabolic requirements to the point where birds may starve to death while trying to keep warm.

As well, nesting birds that come into contact with fuel oils may transfer oil from their feathers and feet to their eggs during incubation. Toxicity to bird eggs via this route has been shown (Environment Canada 2010b; Michigan 2010). Fuel Oil No. 6 (FO6) is similar to four of the HFOs considered here (64741-75-9, 70592-76-6, 70592-77-7 and 70592-78-8) and can be used in a read-across approach for toxicity. Szaro (1979) found that 5 µL of FO6 applied to eggs significantly reduced hatching success to 36% and 6­day survival to 52% in Mallard Ducks (Anas platyrhyncos).

Fuel Oil No. 2 is similar to the light HFO (CAS RN 68783-08-4) and can be used as a toxicity surrogate. In tests on Mallard Duck eggs, lowest-observed-effect concentrations (LOECs) were found at 1 µL/egg (20% reduction in hatchability with a 28% reduction in duckling survival post-hatch) (Albers and Szaro 1978; Szaro et al. 1978). Coon et al. (1979) determined that a 5-µL/egg treatment reduced hatchability by 28% compared with controls with eggs of Great Black-backed Gull (Larus marinus). Common Eider duck (Somateria mollissima), Louisiana Heron (Hydranassa tricolour), Laughing Gull (Larus atricilla) and Sandwich Tern (Sterna sandvicensis) eggs experienced from 20% to 81% mortality at 20 µL/egg (Albers and Szaro 1978; White et al. 1979).

Terrestrial Compartment

The Canada-Wide Standards for Petroleum Hydrocarbons in Soil (CCME 2008) were used as a data source for effects of HFOs on terrestrial ecosystems. This standard was developed based on consideration of four fractions of total petroleum hydrocarbons (TPHs): F1 (C₆ to C₁₀), F2 (> C₁₀ to C₁₆), F3 (> C₁₆ to C₃₄) and F4 (> C₃₄). Fraction 3 (F3) is most like HFOs. This system also uses four land-use classes (agricultural, residential, commercial, industrial) and two soil types (coarse grained and fine grained). The most sensitive land-use and soil type is typically agricultural coarse-grained soils. The F3 soil quality guideline for soil contact by non-human organisms is 300 mg/kg (CCME 2008).

Estrogenic Effects

Exposure to contaminants with estrogenic effects can lead to both physiological and molecular changes in organisms, including impaired reproductive capacity and skewed sex ratios. Studies have shown that certain alkylphenols and PAHs have been shown to have estrogenic effects and are believed to be responsible for the estrogenic effects of petroleum products. However, while the effects of individual compounds have been studied, the additive, antagonistic and synergistic effects that can occur in mixtures have not been fully clarified; therefore, the extent of the estrogenic effects of petroleum products is not well understood.

In vitro assays suggest that heavy fuel oils have an estrogenic effect. Vrabie et al. (2010) conducted a study using recombinant yeast and a human cell line (Vrabie et al. 2011) to determine if various petroleum products, including two Bunker fuels, were able to induce a response in human estrogen receptor-a (ERa) and estrogen receptor-b (ERb). The two Bunker fuels were able to induce the expression of ERa at levels comparable to 17b­estradiol in the mammalian assay, though not in the yeast assay. The two Bunker fuels were also able to induce the expression of ERb in both yeast and mammalian assays, though in the yeast assay the level of expression exceeded the induction caused by 17b-estradiol while in the mammalian assay the level of expression was reduced. The differences between the two assays suggest that the metabolism of components of the heavy fuel oils in the mammalian cells plays a role in the estrogenicity of the fuels. Furthermore, it was found that potency of the Bunker fuels, determined by comparing the lowest effect concentrations of the fuels and 17b-estradiol, was several orders of magnitude lower than the standard. Because the fuels were able to induce response in the estrogen receptors in a manner comparable to the standard, however, the authors hypothesize that the Bunker fuels may contain extremely potent components that are present at a reduced concentration.

Kizu et al. (1999) investigated the estrogenic activity of two heavy oils by using an in vitro assay measuring the level of the cellular progesterone receptor (PgR) in a human cell line. PgR levels increase in response to 17b-estradiol. While in the absence of 17b­estradiol it was found that the heavy oils had a weak estrogenic effect, the addition of both 17b-estradiol and the heavy oils showed that the heavy oils act as 17b-estradiol antagonists.

In vitro studies also suggest that heavy fuel oils may have an androgenic effect. Kizu et al. (2000) investigated the antiandrogenic effect of crude extract of C-heavy oil in androgen-responsive mouse cell line. Results demonstrated that C-heavy oil contains antiandrogenic compounds. The antiandrogenic effect of the two heavy oil extracts was considered to be due in part to PAHs.

There have not been many in vivo studies on the estrogenic effect of heavy fuel oils, but those that exist have similar results to in vitro studies.

Mussels (Mytilus galloprovincialis) along the Galician and Bay of Biscay coast were studied from 2003 to 2006 after the release of heavy fuel oil from the Prestige oil tanker to determine what, if any, changes had occurred in their reproductive parameters (Ortiz-Zarragoitia et al. 2011). In April 2003, female mussels showed a high prevalence of oocyte atresia, which may decrease gamete quality. Hemocytic infiltration of the follicles was also observed in both sexes. In April 2004, no atresia was found, but both male and female mussels had reduced follicle size. The effect of reduced follicle size on mussel reproductive ability is unknown. In April 2006, female mussels from certain sites had necrotic gametes. An alkali labile phosphate (ALP) assay was also conducted. The ALP assay is used as an indirect measure of vitellogenin, a major precursor of egg-yolk proteins and a biomarker of estrogenic effects. Vitellogenin-like proteins have been identified in mussels, and ALP levels can be used as a measure of estrogenic effect (Blaise et al. 1999). Male mussel ALP levels were low in all sampled timepoints, suggesting that the xenoestrogenic effect of the fuel oil was insufficient to induce the expression of the vitellogenin-like protein (Ortiz-Zarragoitia et al. 2011). It was found, however, that ALP levels in females in 2004 were highly variable across the different sampling sites and did not correlate with gonad development. In subsequent years, ALP levels were more homogenous across sample sites and correlated more strongly with gonad development, suggesting an alteration in mussel reproduction in 2004.

Bilboa et al. (2010) exposed thicklip grey mullet (Chelon labrosus) to heavy fuel oil mixed into sediment and found that the expression of the aromatase gene cyp19a2 increased in the brain after the second day of exposure, though by day 16 the level of expression was similar to the control. Aromatases are key enzymes in the synthesis of estrogen and can be regulated by estrogen receptors.

Juvenile turbot (Scophthalmus maximus) were fed food pellets contaminated with heavy fuel oil for 42 days to determine if there were changes to plasma levels of testosterone and 17b-estradiol (Martin-Skilton et al. 2008). A heavy fuel oil concentration of 2.5mg/g of food was sufficient to cause a two-fold decrease in testorone levels; 17b-estradiol levels showed a 35% decrease when heavy fuel oil contamination was at a concentration of 50 mg/g of food.

Nishimoto et al. (2009) studied mice that ingested the water-soluble fraction (WSF) of heavy oil over the course of 28 days. After seven days, 80% of female mice that ingested the WSF developed cystoma-like formations around their ovaries; by the fourteenth day, all female mice had developed these formations. Some clinical reports have shown that hypertropic cystoma can progress to ovarian tumors, but this was not seen in the mice studied. Male mice that ingested the WSF had reduced prostrate glands. Using a human cell line that was estrogen responsive, Nishimoto et al. confirmed that the WSF contained compounds that had an estrogenic effect. The authors speculate that PAHs may be involved in the abnormalities of the genital organs and in the estrogen response, though further study is needed to conclusively identify the chemicals involved.

Current evidence indicates that heavy fuel oil does have estrogenic effects that potentially can have ramifications on reproduction, plasma steroid levels and gene expression. Further research would be needed, however, to understand the extent of the estrogenic effects.

Ecological Exposure Assessment

Estimations of releases of these HFOs were made using data included in responses to a notice published under section 71 of CEPA 1999 (Environment Canada 2009), estimations of losses to the sea on Canada’s east coast by Risk Management Research Institute (RMRI 2007) and Environment Canada’s Spill Line database (Environment Canada 2011).

Scenarios were developed for releases to marine and fresh water from ships during loading/transport/unloading. The other sources of release, including from other modes of transportation, contributed significantly less volume of HFOs to an environmental medium of concern for the ecological assessment.

Aquatic Compartment

To determine the predicted environmental concentration (PEC) arising from ship transport, the volume of water predicted to be in contact with spilled oil was provided by a report conducted by RMRI (2007). The area of a slick created within hazard zones around Newfoundland and Labrador was estimated for specific volume ranges of oil using ocean spill dispersion models, and then the volume of contacted water was estimated by multiplying the area by 10 to represent the top 10 metres of water. This assumes that all of the water is equally contacted by the petroleum product spilled.

For loading and unloading scenarios for ships, the volume of water in contact with oil is from Hazard Zone 1, as this region includes loading and off-loading operations at Whiffen Head and the Come By Chance refinery. For the ship transport scenario, the estimated volume of water in contact with oil is the average volume of water from Hazard Zone 2 (outer Placentia Bay). This work was originally done for crude oil, but it can be applied to HFOs as they have a similar density.

In the case of marine loading and unloading of HFOs by ship, an estimated 13 754 kg of fuel oil on average could be lost in one event to salt water (Table 3). At an average density of 1.04 kg/L (API 2004) this is equivalent to 83 barrels of fuel oil and is therefore expected to be in contact with 150 x 10⁹litres of water (Table A5.13 in Appendix 5). This volume is estimated from the enclosed waters found at wharves and loading terminals. The resulting concentration in water would be 0.09 mg/L (1.38 x 10¹⁰ mg/150 x 10⁹ litres), which is considered the marine PEC for ship loading and unloading.

In the case of the marine transportation of HFOs by ship, an estimated 13 754 kg of fuel oil on average could be lost in one event to salt water (Table 3). At an average density of 1.04 kg/L (API 2004) this is equivalent to 83 barrels of fuel oil and is therefore expected to be in contact with 6250 x 10⁹litres of water (Table A5.13 in Appendix 5). This volume is estimated from the open ocean of Placentia Bay. The resulting concentration in water would be 0.002 mg/L (1.38 x 10¹⁰mg/6 250 x 10⁹ litres), which is considered the marine PEC for ship transport.

In the case of the freshwater loading and unloading of HFOs by ship, an estimated 15 262 kg of fuel oil could be lost in one event to fresh water (Table 3). At an average density of 1.04 kg/L (API 2004) this is equivalent to 92 barrels of fuel oil and is therefore expected to be in contact with 150 x 10⁹litres of water (Table A5.13 in Appendix 5). This volume is estimated from the enclosed waters found at wharves and loading terminals. The resulting concentration in water would be 0.1 mg/L (1.53 x 10¹⁰ mg/150 x 10⁹ litres), which is considered the freshwater PEC for ship loading and unloading.

In the case of the freshwater transportation of HFOs by ship, an estimated 15 262 kg of fuel oil could be lost in one event to fresh water (Table 3). At an average density of 1.04 kg/L (API 2004) this is equivalent to 92 barrels of fuel oil and is therefore expected to be in contact with 6250 x 10⁹ litres of water (Table A5.13 in Appendix 5). This volume is estimated from the open ocean of Placentia Bay. The resulting concentration in water would be 0.002 mg/L (1.53 x 10¹⁰ mg/6 250 x 10⁹ litres), which is considered the freshwater PEC for ship transport.

Terrestrial Compartment

Less than one release event per year for pipeline transport of HFOs is predicted based on the short distance of transport determined from information submitted under section 71 of CEPA 1999 (Environment Canada 2009) and the average spill rate per length of pipeline (1 spill per 11 100 km of pipeline, as found in NEB 2008). Likewise, only two of the five industry-restricted HFOs are transported by pipeline. From the historical Canadian data from the Spill Line database for Bunker C (Environment Canada 2011), only 13 spills of HFOs from pipelines were reported over 10 years (2000–2009). Thus, less than 1 release event per year is expected for pipeline loading, transport and unloading for industry­restricted HFOs.

It is estimated that there will be ≤ 1 release event per year each for train and truck loading, unloading and transport based on historical release information from the Spill Line database (Environment Canada 2011). Spill events are expected to generally occur at industrial facilities for industry-restricted HFOs. It was additionally considered that these infrequent releases would likely occur on a hard surface and not on soil; therefore releases from truck and train are not considered to be of high importance under these circumstances. It is expected that the actual release frequency for these industry-restricted HFOs is lower, as the Spill Line database release information was for Bunker C.

Characterization of Ecological Risk

The approach taken in this ecological screening assessment was to examine available scientific information and develop conclusions based on a weight-of-evidence approach as required under CEPA 1999. For each endpoint organism, an estimate of the potential to cause adverse effects and predicted no-effect concentration (PNEC) was determined. Also, a PEC was determined from the aquatic exposure scenario. The PNEC is the lowest critical toxicity value (CTV) for the organism of interest divided by an appropriate application factor. A risk quotient (PEC/PNEC) was calculated for each endpoint organism and is an important line of evidence in evaluating the potential risk to the environment.

Since a read-across approach can be used with Fuel Oil No. 6, the CTVs for this assessment are selected from empirical data available for Fuel Oil No. 6 (Table A5.12, Appendix 5). For the marine scenarios, a CTV of 0.9 mg/L is selected based on the LC₅₀ value for Mysidopsis almyra. For the freshwater exposure scenarios, the CTV is selected from the 96-hour LC₅₀ of 1.69 mg/L for Fundulus similis (Table A5.12, Appendix 5). An application factor of 10 is used to account for the extrapolation of modelled data to field effects. Table 6 is the summary of the risk quotients for the industry-restricted HFOs.

Table 6. Risk Quotients Calculated for Industry-restricted HFOs

Compartment Affected	Organism	PEC	CTV	Application Factor	PNEC	Risk Quotient
Fresh water (loading/ unloading)	Fundulus similis	0.1 mg/L	1.69 mg/L	10	0.17 mg/L	0.6
Freshwater (transport)	Fundulus similis	0.002 mg/L	1.69 mg/L	10	0.17 mg/L	0.01
Marine (loading/ unloading)	Mysidopsis almyra	0.09 mg/L	0.9 mg/L	10	0.09 mg/L	1
Marine (transport)	Mysidopsis almyra	0.002 mg/L	0.9 mg/L	10	0.09 mg/L	0.02

For all aquatic spill scenarios, the critical spill volume for HFOs required to obtain a risk quotient equal to one and the frequency of spills above that threshold was determined from the Environment Canada Spill Line database (Environment Canada 2011) (see Table 7).

Table 7. Spill Volumes Required to Create Harmful Conditions to Aquatic Organisms and the Proportion of Reported Spills of HFOs Above this Threshold Volume^[a]

Compartment Affected	Spill Volume Required to Obtain Risk Quotient = 1 (Threshold Volume) (L)	Proportion of Reported Spills Above the Threshold Volume	Number of Spills Per Year Expected to be Above the Threshold Volume
Fresh water (loading/unloading)	24 550	11%	< 1
Fresh water (transport)	1 570 000	0%	0
Marine (loading/unloading)	13 000	15%	1.6
Marine (transport)	835 000	0%	0

For the marine and freshwater scenarios during ship transport, a spill volume of 835 000 L and 1 570 000 L of fuel oil, respectively, is needed to obtain a risk quotient (RQ) of 1 to aquatic organisms (Table 7) based on toxicity estimations and spill dispersion models of the volume of water affected. None of the reported spills from 2000–2009 were greater than these threshold volumes during ship transport and therefore, the expected number of spills per year above this volume is 0. As well, the whole dataset from the Environment Canada Spill Line database for Bunker C was used as a surrogate for these HFOs. The amount of HFOs released is unknown, but is certainly less than the total volume of Bunker C. Thus, spills to fresh water and salt water during transport are not considered harmful to aquatic organisms.

For the scenarios for ship loading/unloading, a spill volume of 13 000 L and 24 550 L of fuel oil is needed to obtain an RQ of 1 to aquatic organisms in marine and fresh waters, respectively (Table 7). There are some reported spills above these threshold volumes during the loading/unloading of ships in marine (15% of spills) and fresh water (11% of spills). However, these frequencies equate to an expected less than 2 spills per year above the threshold volume in marine waters and less than one spill per year in fresh water. These frequencies are based on the entire dataset from the Environment Canada Spill Line database for spills of Bunker C, which are expected to be more frequent than spills of industry-restricted HFOs. The RQ for marine loading was 1, and for fresh water loading the risk quotient was below 1, based on average spill volumes. Thus, based on the RQs and relatively low expected number of spills per year, spills of these HFOs to water during loading and unloading are considered to be infrequent and likely not harmful to aquatic organisms.

As previously mentioned, this assessment does not include illegal releases of fuel oil at sea in Canadian jurisdictions. Transport Canada has in place a National Aerial Surveillance Program to monitor and deter such releases (Transport Canada 2010).

Aquatic birds have not been included in the risk quotient analysis due to the lack of specific toxicity information on these HFOs to birds. However, both field reports and experiments have shown that commercial blends of HFOs are toxic to aquatic birds through ingestion (CONCAWE 1998; Environment Canada 2010b; Michigan 2010), contact with feathers (Environment Canada 2010b) and contact with eggs (Albers and Szaro 1978; Coon et al. 1979; CONCAWE 1998). The negative effects of oil on feathers, however, are not specific to HFOs and are primarily based on Bunker C fuel oil. Indeed, average spills to marine water for these industry-restricted HFOs are based on the Environment Canada Spill Line data for Bunker C fuel oil (Environment Canada 2011). Thus, releases to marine waters are not expected to cause harm to sea birds through direct toxicity and indirect effects. The four HFOs transported by ship (CAS RNs 64741-75-9, 70592-76-6, 70592-77-7 and 70592-78-8) are unlikely to cause harm to avian organisms due to a lack of exposure.

Based on the estimated < 1 HFO release event per year for pipeline transport, HFOs are unlikely to cause harm to terrestrial non-human organisms.

With regard to truck and train releases, a risk quotient was not determined. Release frequency and volumes from trains are less certain due to a lack of definitive data. The Spill Line database reports small numbers of Bunker C spills via train (11 spills) and truck (32 spills) from 2000 to 2009. Considering the cause and reason of spill, it was determined that for each scenario of loading, transport and unloading of trains, less than 1 spill per year is expected. By the same analysis, ≤ 1 spill per year each for loading, transport and unloading by truck is expected. The response times for these types of spills will be shorter, as the vehicle operator typically is present when the spill occurs. Thus, terrestrial impacts from train and truck transport of HFOs are unlikely to cause harm due to their low frequency (less than 1 spill per year for loading/unloading and transport). Likewise, the estimated spills from truck loading and unloading were not considered to be of high importance, as they would likely occur on a hard surface and not on soil.

Some components of these HFOs, particularly PAHs, have been the subject of previous risk assessments. PAHs are included in the List of Toxic Substances under Schedule 1 of CEPA 1999 (Canada 1994). Evidence from field studies and from laboratory studies using field samples indicate that biota are adversely affected at various Canadian sites contaminated by PAHs of different industrial origins (Canada 1994).

Based on results from AOPWIN (2008), there would be a relatively rapid removal process if these HFOs are introduced into the atmosphere, based on oxidation half-lives of less than 1 day. With regard to the primary and ultimate biodegradation modelling, the C₃₀–C₅₀ isoalkanes, C₃₀–C₅₀ one-ring cycloalkanes, C₁₅–C₅₀ two-ring cycloalkanes, C₁₄–C₂₂ polycycloalkanes, C₃₀–C₅₀ one-ring aromatics, C₁₀–C₂₀ cycloalkane monoaromatics, C₁₅–C₅₀ two-ring aromatics, C₁₂–C₂₀ cycloalkane diaromatics, C₂₀–C₅₀ three-ring aromatics, C₁₆–C₂₀ four-ring aromatics, C₂₀–C₃₀ five-ring aromatics and C₂₂ six-ring aromatics in these HFOs meet or exceed the criteria for persistence (half-life in soil and water ≥ 182 days and half-life in sediment ≥ 365 days) defined in the Persistence and Bioaccumulation Regulations (Canada 2000). Based on the available predicted information, these HFOs contain a large proportion (30–60% by weight) of components that may persist sufficiently in soil, water and sediment to meet the regulatory criteria.

Based on the combined evidence of empirical data and predicted analysis of BCFs, BAFs, BMFs, TMFs and BSAFs, the HFOs assessed in this report may contain a large proportion of components (up to 21%) that are bioaccumulated from water and dietary sources, but are not likely biomagnified in food webs. Both empirical and predicted BCF and predicted BAFs are ≥ 5000 for isoalkanes, cycloalkanes and some aromatic substances. There is consistent steady-state and kinetic evidence to suggest that these components do not metabolize very quickly and have sufficient dietary assimilation efficiency, that when tissue levels are compared with the bioavailable fraction in water, accumulation factors are expected to be high.

As shown in Table A5.14 in Appendix 5, some components may meet both the persistence and bioaccumulation criteria in the Persistence and Bioaccumulation Regulations. These include the C₁₅ dicycloalkanes, C₁₄ and C₂₂ polycycloalkanes, C₁₅–C₂₀cycloalkane monoaromatics, C₂₀ cycloalkane diaromatics and C₂₀ three-ring aromatics. Of these, there were only two measured BCF values--for the C₁₄ and C₁₈cycloalkane monoaromatics (octahydro-phenanthrene and dodecahydro-chrysene, respectively)--available for comparison. Based on these values, the C₁₈ compound was found to bioconcentrate in fish close to but not exceeding a BCF of 5000 (Lampi et al. 2010). Comparable measured BMF values from experimental studies were not > 1, indicating that these particular hydrocarbons will not biomagnify when compared to the concentrations expected in food items. Some PAHs (C₁₆–C₂₀ 4-ring, C₂₀ 5–ring and C₂₂ 6-ring) were also identified as being potentially persistent and bioaccumulative based on available empirical data (see Table A5.14 in Appendix 5). However, a majority of organisms exhibit rapid depuration, and exposure to PAHs is limited in time for HFO spills. Thus PAHs from these HFOs are not likely to have a bioaccumulative concern.

Based on the available data, it is proposed that these industry-restricted HFOs (CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8) are not entering the environment in a quantity or concentration or under conditions that may have an immediate or long-term harmful effect on the environment or its biological diversity.

Uncertainties in Evaluation of Ecological Risk

This analysis addresses uncertainty associated with each component of the current assessment, including but not limited to selection of representative structures and quantification, exposure estimation, effects estimation, and risk characterization.

All modelling of the substance’s physical-chemical properties, as well as persistence, bioaccumulation and toxicity characteristics, is based on chemical structures. As these industry-restricted HFO are UVCBs, they cannot be represented by a single, discrete chemical structure. The specific chemical compositions of these HFOs are not well defined. HFO streams under the same CAS RNs can vary significantly in the number, identity and proportion of constituent compounds, depending on operating conditions, feedstocks and processing units. Therefore, for the purposes of modelling, a suite of representative structures that provide average estimates for the entire range of components likely present was identified. Specifically, these structures were used to assess the fate and hazard properties of HFOs. Given that more than one representative structure may be used for the same carbon range and type of component, it is recognized that structure-related uncertainties exist for these substances. The physical-chemical properties of 48 representative structures were used to estimate the overall behaviour of these HFOs, in order to represent the expected range in physical-chemical characteristics. Given the large number of potential permutations of the type and percentages of the structures in HFOs, there is uncertainty in the results associated with modelling.

Uncertainty arises from the non-uniformity of spill data. The available data on spills generally do not report values for each specific transported substance by CAS RN. For example, spill or leak data for pipeline transportation are only divided into liquid lines or gas lines, and are not further characterized. For marine transportation, Environment Canada and Transport Canada reported spills data for products similar to these heavy fuel oils, specifically Bunker C fuel oil. Spill data specific to these industry-restricted HFOs are not available for each mode of transportation. The use of a generic loss fraction factor, derived from the available data, introduces uncertainty in the estimation of transportation releases.

Similarly, historical spills data classified as Bunker C fuel oil from the Emergencies Spill Line Database (Environment Canada 2011) were used in the ship, truck and train transport release scenarios for these industry-restricted HFOs. The amount of HFOs released is unknown, but is certainly less than the total volume of Bunker C. There is uncertainty in the estimation of the actual HFO loading, transport and unloading releases.

The fate, food chain interactions, and toxicity of a number of petroleum hydrocarbons depend to a large extent upon their chemical form. As such, conservative assumptions about chemical form, bioavailability, and absorption through the digestive tract were generally carried forward in the risk assessment. HFO representative structures were assessd with the conservative assumption that all of them were bioavailable.

This assessment involves the prediction of effects on biota using measured inputs and modelled accumulation or exposures. The process typically relies on modelled exposures for organisms at higher trophic levels. However, all models are simplifications of natural systems or processes, and therefore, rely on a number of assumptions. These, in turn, create uncertainties in the outcomes.

The BAF model calculations were derived from a large database of measured BAF values from the Great Lakes for chemicals that are poorly metabolized (e.g., PCBs). With metabolic biotransformation, the BAF model predictions are in general agreement with measured BAFs in fish. The model may not adequately capture biotransformation at the first trophic level for chemicals that are readily biotransformed in invertebrates and plankton. Many petroleum hydrocarbons are readily metabolized, somewhat by invertebrates and at much higher levels in fish.

The significance and impact of bioaccumulation is species specific and is dependent on a range of factors such as species, size and the environmental conditions. At present, there are no field data on the study of bioaccumulation of industry-restricted HFOs as a class; therefore, predicting effects is based on modelling their BAFs based on laboratory­acquired partitioning data.

Potential to Cause Harm to Human Health

Exposure Assessment

HFO substances are a group of heavy petroleum streams produced in oil refineries from crude oil. Due to the physical-chemical properties of HFOs, the dermal route is an important route of occupational exposure. In a recent study to quantify such workplace exposures, Yvette et al. (2011) found that dermal exposures were generally low. However, the authors indicated that the presence of HFO components with some degree of carcinogenic potential identified in all of the HFO blends they investigated requires that control measures to maintain low dermal exposure levels should be strictly adhered to, and additional means of reducing HFO exposure even further should continue to be sought.

Due to the relatively low volatility of the industry-restricted HFOs (see Table 1) and relevant regulations in place to limit potential releases during handling of petroleum substances, general population exposure to these substances by ingestion and inhalation during loading and unloading is expected to be negligible and will not be considered further.

Significant concentrations of hydrogen sulfide are known to accumulate in the headspaces of storage tanks that contain HFOs. Heating of such tanks may cause the decomposition of some of the sulfur-containing compounds, which release hydrogen sulfide. There is also evidence that vapours of light hydrocarbons accumulate in the headspaces of HFO tanks (CONCAWE 1998).

The human health assessment of industry-restricted petroleum substances focuses on the fugitive releases that occur when petroleum substances escape into ambient air. These include evaporative losses from tanks during the various modes of transportation of petroleum substances. The unintentional release (leaks or spills) data used in the ecological assessment are, for the purposes of assessing the potential to cause harm to human health, considered to refer to releases that occur on a non-routine or unpredictable basis in specific geographical locations. These unintentional releases (leaks or spills) typically do not contribute to the potential for exposure of the general population in Canada.

Evaporative losses of the industry-restricted HFO substances during transit will enter ambient air. As such, inhalation is the primary potential route of bystander exposure, which may occur as the substances are being transported between facilities, and is therefore the focus of the current human health exposure assessment.

Inhalation from Ambient Air

As monitoring data on HFOs in the environment are not available, the HFO vapour level in ambient air was estimated using SCREEN3 (1996), a screening-level Gaussian air dispersion model based on the Industrial Source Complex (ISC) model (for assessing pollutant concentrations from various sources in an industry complex). The driver for air dispersion in the SCREEN3 model is wind. The maximum calculated exposure concentration is selected based on a built-in meteorological data matrix of different combinations of meteorological conditions, including wind speed, turbulence and humidity. This model directly predicts concentrations resulting from point, area and volume source releases. SCREEN3 estimates the maximum concentrations of a chemical at chosen receptor heights and at various distances for a given population in the vicinity of the release source in the direction downwind from the prevalent wind 1 hour after a given release event. During a 24-hour period, for point emission sources, the maximum 1-hour exposure as assessed by the ISC Version 3, is multiplied by a factor of 0.4 to account for variable wind directions. This gives maximum concentration within 24-hour exposure (U.S. EPA 1992). Similarly, for exposure events happening over the span of a year, it can be expected that the direction of the prevalent winds will be even more variable and uncorrelated to the wind direction for a single event; thus, the maximum exposure concentration for one year is determined by multiplying the maximum 1-hour exposure by a factor of 0.08. Such scaling factors are not required for non-point source emissions. However, to prevent overestimation of the exposures, we use a scaling factor of 0.2 to obtain the yearly exposure concentration from the value of the maximum 1-hour exposure determined from SCREEN3 calculations. Detailed input parameters for SCREEN3 are listed in Table A6.1 in Appendix 6. As a conservative estimate, all the regular evaporative loss during 1 day of transit is assumed to originate from a defined area rather than a moving line source; as such, actual levels are expected to be lower, considering that the release source is typically moving.

Estimated regular evaporative loss to air during transit of industry-restricted HFOs is presented in Table A6.2 in Appendix 6 as a range to cover the losses from the various transportation modes involved. Formulas for evaporative losses of HFOs from truck and train transit of HFOs are not available in the AP 42 guidelines (U.S. EPA 2008a). A conservative estimate for these transit losses may be calculated by using stationary storage tank formulas adapted to typical dimensions of truck and train tanks. Even at this level of conservatism, due to the low volatility of the HFOs, the evaporative losses from truck and train transit are small. The upper value in the range is related to evaporative loss from ship transit. Emission rates in grams per second per square metre (g/s·m²) are derived based on the loss quantity of kilograms per day (Table A6.2 in Appendix 6) and the estimated emission areas for different transportation modes (Table A6.1 in Appendix 6). This emission rate (g/s·m²) was used for determining the concentration of the HFO vapours in ambient air by SCREEN3 (1996).

As evaporative loss quantities are different for various transportation modes, for those industry-restricted HFOs with more than one mode of transportation, the maximum concentrations of ambient HFO vapours during 24 hours are presented as a range in Table A6.3 of Appendix 6. A conservative estimate of exposure was chosen by using the maximum concentrations at 50 m (for transportation by trains), as these were the highest exposure values obtained, compared with those at distances farther from the release sources. The upper-bounding estimate of the maximum concentration in ambient air at 50 m was 1.28 µg/m³. This value was used for predicting the maximum daily intake of HFO vapours from evaporative losses by the general population. This estimation is conservative, as it is based on releases from a stationary point source. The actual concentration of the HFO vapours in the vicinity of the moving release source, for any given location, will be considerably lower than that represented by the total daily release quantity from a stationary point release source.

Potential daily intakes of industry-restricted HFOs from inhalation of ambient air were estimated in order to determine the age group receiving the highest internal dose. The maximum concentration of HFOs within 24 hours in ambient air at 50 m from the release source, the time spent outdoors per day for each age group, and inhalation rate and body weight (bw) for each age group were used to estimate the potential daily intakes (Health Canada 1998). A distance of 50 m was selected to represent bystanders in the vicinity of transportation as an upper-bounding estimation. The upper-bounding estimate of daily intake by inhalation was identified to be 0.1 µg/kg-bw per day for children 0.5–4 years of age (Table 8). This was based on a concentration of 1.28 µg/m³ (as above), multiplied by an average outdoor time of 3 hours/day and an average inhalation rate of 9.3 m³/day for this age group of children, divided by the average body weight of 15.5 kg (Health Canada 1998). To calculate a margin of exposure (MOE) for characterization of risk to human health, the highly conservative projected concentration of 1.28 µg/m³ is used.

Table 8. Estimated Maximum Daily Intake of Industry-restricted HFOs Via Inhalation of Ambient Air for Different Age Groups at 50 m from the Release Source (by Train Transportation)^[a]

Substances	Daily Intake Level (µg/kg-bw per day)
	0–0.5 years	0.5–4 years	5–11 years	12–19 years	20–59 years	60+ years
Industry-restricted HFOs	0.04	0.1	0.07	0.04	0.04	0.03

It should be noted that the estimated air concentrations of HFOs are considered to be conservative, as SCREEN3 is, by design, a conservative screening-level tool used as a rapid approach to estimate the air dispersion of various chemicals. Another consideration is that the releases of the industry-restricted HFO vapours that occur during the transit process occur continuously from a moving source (a line source) rather than from a fixed point source. As such, the actual concentration of the HFO vapours around a moving line source, for any given location, will be considerably lower than that represented by the total daily release quantity from a point release source, as was used in this assessment. Thus, the assumption of total daily evaporative loss within one defined area is considered to be a conservative estimate of the actual substance concentration in ambient air. Placing the receptor at 50 m from the release source is also conservative, as most Canadians do not reside within 50 m of HFO transport.

Health Effects Assessment

Given the limited number of studies available that specifically evaluate the health effects of the industry-restricted HFO substances, an adequately representative toxicological data set unique to these substances could not be obtained. Therefore, to characterize the toxicity of these HFOs, additional HFOs in the PSSA that are similar from both a process and a physical-chemical perspective were also evaluated for their toxicological effects. As both the industry-restricted and the additional HFO substances have similar physical-chemical and toxicological properties, the toxicological data across the industry-restricted HFOs were used to construct a toxicological profile to represent all HFOs. Accordingly, the toxicity of HFOs is represented as a group, not by individual CAS RNs.

Appendix 7 contains a summary of available health effects information on HFOs in laboratory animals. A summary of key studies selected to represent the toxicity of industry-restricted HFOs follows. The HFO category of petroleum mixtures represented in Appendix 7 includes both residual fuels from distillation or cracking units and blended products. It consists of aromatic, aliphatic and cycloalkane hydrocarbons. Heavy fuels may also contain hydrogen sulfide, as well as a broad range of chemicals that are tumorigenic (e.g., PAHs), and the quantities present in HFOs can vary (CONCAWE 1998; Yvette et al. 2011).

HFOs have low acute toxicity. Inhalation exposure resulted in an LC₅₀ of > 3700 mg/m³ in rats. Oral exposure resulted in a median lethal dose (LD₅₀) of > 2000 to > 25 000 mg/kg-bw in rats. Dermal exposure resulted in an LD₅₀ of > 2000 to > 5350 mg/kg-bw in rabbits and > 2000 mg/kg-bw in rats (CONCAWE 1998; ECB 2000a; API 2004; U.S. EPA 2005). Minimal to moderate skin irritation was observed for acute dermal exposure. Available data indicate that HFOs and HFO components are not eye irritants (CONCAWE 1998).

In an acute oral toxicity study conducted for CAS RN 64741-62-4, a single dose of 2000 mg/kg-bw or a single dose of 125, 500 or 2000 mg/kg-bw was administered to pregnant Sprague-Dawley rats on one of gestation days 11–15 or on gestation day 12, respectively. Decreased maternal body weight gain and thymus weights were reported, regardless of treatment day, for the gestation day segment of the study. Dose-related decreased maternal body weight gain and thymus weights were reported for the dose­response segment of the study (Feuston et al. 1989; Feuston and Mackerer 1996).

One short-term inhalation study was conducted for CAS RN 64742-90-1. A lowest­observed-adverse-effect concentration (LOAEC) of 540 mg/m³ was observed for decreased body weight (concentration-related) and increased liver weight in Fischer 344 rats following administration of 540 or 2000 mg/m³, 6 hours/day for 9 days (Gordon 1983).

Short-term and subchronic dermal toxicity studies conducted over periods of 3 days to 13 weeks are available for HFO substances, including one industry-restricted substance (CAS RN 68783-08-4). Slight to severe skin irritation was observed in several studies; the lowest dose reported for skin irritation was 8 mg/kg-bw per day (Mobil 1994a, b). Selected systemic effects observed in these studies included decreased body weight gain and body weight, decreased thymus weights, increased liver weights and changes in hematological parameters (e.g., platelets, hemoglobin, red blood cells) and serum chemistry (i.e., liver enzymes and other indicators of liver toxicity) (API 1983; Mobil 1988, 1990, 1992, 1994a, b; UBTL 1990, 1994; Feuston et al. 1994, 1997). A lowest-observed-adverse-effect level (LOAEL) of 1 mg/kg-bw per day was reported for maternal toxicity following dermal exposure of pregnant CD rats to CAS RN 64741-62-4 at doses of 0.05, 1, 10, 50 or 250 mg/kg-bw per day from gestation days 0 to 19. Effects observed at the LOAEL included significantly decreased body weight gain, body weight and feed consumption, as well as decreased gravid uterine weight and the occurrence of red vaginal exudates (Hoberman et al. 1995). For subchronic exposure, a LOAEL of 8 mg/kg-bw per day was established following dermal exposure of male and female rats to CAS RN 64741-62-4 or 64741-81-7 at doses of 8, 30, 125, 500 or 2000 mg/kg-bw per day for 13 weeks. Effects noted at the LOAEL included decreased platelet counts and increased liver weights, as well as dose-related skin irritation (Mobil 1988, 1992, 1994b; Feuston et al. 1994, 1997). Lack of testing at doses lower than 8 mg/kg-bw per day lowers confidence in the LOAEL.

The genotoxicity of HFOs has been evaluated in in vivoand in vitro assays. Results from in vivogenotoxicity testing of three HFO substances were mixed (i.e., both positive and negative results were obtained for the same assay and endpoint). Positive results were observed in mice and rats for micronuclei induction, sister chromatid exchange and unscheduled deoxyribonucleic acid (DNA) synthesis when HFOs were administered orally or by intraperitoneal injection (Khan and Goode 1984; API 1985a, b). Negative results were also observed for micronuclei induction, as well as for chromosomal aberrations (API 1985c; Mobil 1987a).

In vitro assays evaluating the genotoxicity of HFOs also exhibited mixed results. Positive results were obtained in the Ames test battery and mouse lymphoma assays, as well as for cell transformation and unscheduled DNA synthesis (Brecher and Goode 1983, 1984; Blackburn et al. 1984, 1986; API 1985c,d, 1986a; Mobil 1985; Feuston et al. 1994). Regarding CAS RN 68553-00-4, negative results were obtained in the Ames and mouse lymphoma assays, as well as for forward mutations and sister chromatid exchange (Farrow et al. 1983; Vandermeulen et al. 1985; Vandermeulen and Lee 1986). Additional negative results were observed only for one cytogenetic assay and one forward mutation assay for CAS RNs 64741-57-7 and 64741-62-4, respectively (API 1985e; Mobil 1987b). Equivocal results were observed in one forward mutation assay and one sister chromatid exchange assay and for cell transformation (Papciak and Goode 1984; API 1985f, 1986b).

The overall genotoxicity database indicates that although the results varied depending on the substance tested and the assay used, HFOs do exhibit genotoxic potential.

The European Commission has classified industry-restricted HFOs as Category 2 carcinogens (R45: may cause cancer) (European Commission 1994; ESIS 2008). The United Nations’ Globally Harmonized System of Classification and Labelling of Chemicals has classified these substances as Category 1B carcinogens (H350: may cause cancer) (European Commission 2008a). The International Agency for Research on Cancer (IARC) has classified residual (heavy) fuel oils as Group 2B carcinogens (possibly carcinogenic to humans) (IARC 1989a).

A number of skin-painting studies were conducted in laboratory animals to investigate the dermal carcinogenicity of HFOs using both chronic and initiation/promotion methodologies. Skin tumours, including both malignant carcinomas and benign papillomas, were frequently observed in mice, rabbits and monkeys (Smith et al. 1951; Shubik and Saffiotti 1955; Shapiro and Getmanets 1962; Saffiotti and Shubik 1963; Getmanets 1967; Weil and Condra 1977; Bingham and Barkley 1979; Sun Petroleum Products Co. 1979; Bingham et al. 1980; Lewis 1983; Blackburn et al. 1984, 1986; API 1989a, b; McKee et al. 1990). Exposure durations for the chronic studies ranged from 25 weeks to lifetime, with reported tumour latency periods ranging from 8 to 113 weeks. In several studies, however, the durations of exposures and latencies were not specified. In a chronic study, male mice were dermally treated with CAS RN 64741-62-4 at doses of 8.4, 16.8, 42.0, 83.8 or 167.6 mg/kg-bw, 3 times per week for a lifetime. Significant skin tumour formation was observed at all doses in a dose-response fashion (McKee et al. 1990). In the one initiation study that was identified, male mice were dermally treated with CAS RN 64741-62-4 at a dose of 16.8 mg/kg-bw once per day for 5 consecutive days. Significant skin tumour formation was observed at this dose. In the corresponding promotion study, no increase in histologically confirmed tumour incidence was observed. A statistically significant increase in the number of mice with gross masses (and shortened latency periods) was observed, however, indicating possible weak promoting activity (API 1989a).

Regarding the tumorigenicity of HFOs, it is recognized that these substances may contain appreciable concentrations of constituents that are tumorigenic, such as PAHs, and the quantity of this fraction can vary depending on the nature and amount of diluent fractions and whether the residue component is cracked or uncracked.

The Government of Canada has previously completed a human health risk assessment of five PAHs, including a critical review of relevant data, under the Priority Substances Program. Based primarily on the results of carcinogenicity bioassays in animal models, these PAHs were classified as probably carcinogenic to humans: substances for which there is believed to be some chance of adverse effects at any level of exposure (Canada 1994). Evaluating the contribution of minor constituents to HFO tumorigenicity is beyond the scope of the current assessment.

HFOs have also been investigated for their reproductive and developmental effects. A LOAEL of 1 mg/kg-bw per day was identified for reproductive toxicity after dermal exposure of pregnant rat dams to CAS RN 64741-62-4 during gestation days 0–19 (the no-observed-adverse-effect level [NOAEL] was 0.05 mg/kg-bw per day). Reproductive effects included decreased number of live fetuses, increased incidences of resorptions and early resorptions and increased percentage of dead or resorbed conceptuses per litter. Fetal developmental variations were also observed in this study but were determined by the authors not to be treatment related (Hoberman et al. 1995). A LOAEL of 8 mg/kg-bw per day for treatment-related developmental toxicity was determined in another study, based on an increased incidence of fetal external abnormalities, including cleft palate, micrognathia (shortened lower jaw) and kinked tail, when catalytically cracked clarified oil was applied dermally to pregnant rats (Mobil 1987c; Feuston et al. 1989). These effects were noted to occur at low incidences. Reproductive toxicity and further developmental effects were observed at 30 mg/kg-bw per day. Reproductive effects included an increased incidence of resorptions and a decreased number of viable fetuses at and above 30 mg/kg-bw per day. At 250 mg/kg-bw per day, no viable offspring were produced (Mobil 1987c; Feuston et al. 1989). In another study, various HFO substances were applied dermally to rats. Substance-dependent LOAELs ranged from 30 to 500 mg/kg-bw per day based on fetal resorption rates ranging from 35.1% to 78.0% (Feuston et al. 1994).

Only one oral reproductive and developmental study was identified for any HFO substance. A LOAEL of ≥ 125 mg/kg-bw was established based on a dose-related increase in resorptions (concomitant decrease in litter size), decreased fetal body weight and increased incidences of skeletal malformations in this acute study that exposed pregnant Sprague-Dawley rats to CAS RN 64741-62-4 during gestation (Feuston and Mackerer 1996). No reproductive or developmental toxicity studies were identified for any HFO substance via the inhalation route of exposure.

Although results varied depending on the substance tested, the overall weight-of-evidence suggests that HFOs exhibit reproductive and developmental toxicity in laboratory animals. The most sensitive LOAEL is 1 mg/kg-bw per day for reproductive and developmental effects.

Epidemiological data were not available for consideration in the human health effects evaluation of HFO substances.

Characterization of Risk to Human Health

Industry-restricted HFOs were identified as high priorities for action because they were considered to present a high hazard to human health and were determined to present greatest potential or intermediate potential for exposure of individuals in Canada during categorization of the DSL. A critical effect for the initial categorization of industry­restricted HFO substances was carcinogenicity, based primarily on the classification by international agencies. These substances are classified as Category 2 carcinogens by the European Commission (European Commission 1994; ESIS 2008), Category 1B carcinogens using the Globally Harmonized System (European Commission 2008a) and Group 2B carcinogens by IARC (1989a). Several cancer studies conducted in laboratory animals resulted in the development of skin tumours following repeated dermal application of HFO substances (API 1989a; McKee et al. 1990). Skin carcinomas and papillomas developed in 100% of mice tested after 36 weeks of dermal exposure to an HFO substance at 167.6 mg/kg-bw per day, whereas tumours developed in 18% of the mice exposed to the lowest dose of 8.4 mg/kg-bw per day (McKee et al. 1990). HFO substances appear to exhibit genotoxic potential in in vivo and in vitro assays, and a mode of action for the induction of tumours involving direct interaction with genetic material cannot be precluded. There are no carcinogenicity studies by the inhalation route to inform the carcinogenic potential of these substances in the general population following inhalation exposure.

Given that the potential for general population exposure to the industry-restricted HFOs results primarily from inhalation of ambient air containing HFO vapours due to evaporative losses during transportation and that the estimated maximum air concentration (1.28 µg/m³) is considered to be low, the risk to human health is likewise considered to be low. The conservative nature of the ambient air concentration estimated is highlighted by a bystander being placed at 50 m and the assumption of total daily evaporative losses occurring within a defined geographic area from a stationary point source (under normal operating conditions, evaporative losses occur predominantly from a moving source; thus, the releases are diluted across a large geographic area).

General population exposure to industry-restricted HFOs via the dermal and oral routes is not expected; therefore, risk to human health from these routes of exposure is not expected.

With respect to non-cancer effects, decreased body weights and increased liver weights in rats were the primary adverse effects observed following a short-term repeated inhalation exposure of 6 hours/day for 9 days. A critical LOAEC of 540 mg/m³ was reported in the single available inhalation study. Comparison of this critical effect level for inhalation exposure in rats with the estimated maximum daily exposure concentration of 1.28 µg/m³ in ambient air results in an MOE of approximately 420 000. The margin is considered more than adequately protective to account for the uncertainties in the data set for the human health risk assessment for both cancer and non-cancer effects, especially in light of the highly conservative nature of the estimated exposures.

Uncertainties in Evaluation of Human Health Risk

The PSSA screening assessments evaluate substances that are complex mixtures (UVCBs) composed of a number of substances in various proportions due to the source of the crude oil or bitumen and its subsequent processing. Monitoring information or provincial release limits from petroleum facilities target broad releases, such as releases of oils and grease, to water or air. These widely encompassing release categories do not allow for the detection of individual complex mixtures or production streams. As such, the monitoring of broad releases cannot provide sufficient data to associate a detected release with a specific substance identified by a CAS RN, nor can the proportion of releases attributed to individual CAS RNs be defined.

Uncertainty exists by using empirical equations for the estimation of evaporative losses. It is noted that the transit evaporative losses also vary with physical conditions, such as the tightness of transport vessels or the valve settings. The screening estimation of evaporative losses does not account for these, nor did industry provide any information relevant to these considerations.

There is uncertainty regarding the conservative estimation of human exposure because of the lack of monitoring data of HFOs in the ambient air and the use of modelling. SCREEN3 modelling of the dispersion profile of HFO vapours requires limited input parameters and non-site-specific meteorological data. These assumptions will introduce more uncertainty compared with other complex dispersion models (Tables A6.1 and A6.2 in Appendix 6).

Because the relative differences in absorption of HFOs through the inhalation, dermal and oral routes of exposure are not well documented, a conservative assumption of 100% absorption was made. Thus, the internal (systemic) doses were considered to be equivalent to the external doses that were used for treatment of the laboratory animals.

As the industry-restricted HFOs are UVCBs, their specific compositions are not well defined. HFO streams under the same CAS RN can vary significantly in the number, identity and proportion of constituent compounds. Consequently, it is difficult to obtain a single representative toxicological data set for HFOs. For this reason, all available toxicological data on HFO substances were pooled across CAS RNs to develop a comprehensive toxicity profile. More research by the scientific community or the petroleum sector to elucidate the compositions of petroleum substances would allow for better characterization of the potential health risks associated with possible exposure to these substances.

Uncertainty also exists due to the paucity of data available regarding the physical-chemical properties of certain HFOs. The densities of the specific CAS RNs were not provided in the toxicity studies; thus, these properties were often obtained from alternative sources. However, because each sample of a particular CAS RN can be slightly different in its composition (as stated previously), these properties may not be entirely representative of a specific sample tested in any one study.

Uncertainty also exists because certain details of the laboratory animals (i.e., sex, strain, body weight and minute volume) were often not stated in the toxicity studies and were obtained from laboratory standard data. Thus, these data may not be entirely representative of the physical features of the actual test animals used in the studies.

Conclusion

The five HFOs considered in this report likely contain significant amounts of components that meet or exceed the criteria for persistence in soil, water and sediment as defined in the Persistence and Bioaccumulation Regulations of CEPA 1999. Based on the available data, the HFOs considered in this report likely contain large proportions of components that meet the criteria for bioaccumulation potential as defined in the Persistence and Bioaccumulation Regulations. Some components of these HFOs were found to meet the criteria for both persistence and bioaccumulation potential as defined in the Peristence and Bioaccumulation Regulations.

Based on comparison of levels expected to cause harm to organisms with estimated exposure levels, these HFOs have some potential to cause harm to aquatic life in the confined marine waters around loading wharfs. However, the estimated frequency of--and, hence, exposure to the environment from--unintentional spills of these HFOs during ship loading is low.

Based on the information presented in this screening assessment, it is proposed that the five industry-restricted HFOs (CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592­77-7 and 70592-78-8) are not entering the environment in a quantity or concentration or under conditions that have or may have an immediate or long-term harmful effect on the environment or its biological diversity or that constitute or may constitute a danger to the environment on which life depends.

Based on the information presented in this screening assessment, the critical effect for the initial categorization of risk to human health was carcinogenicity. However, because the estimates of exposure indicate that the potential exposure of the general population to industry-restricted HFOs from ambient air is expected to be very low, resulting in an extraordinarily large MOE (approximately 420 000), the likelihood of inhalation exposure of Canadians is considered to be very low. Exposure of the general population to industry-restricted HFOs via the dermal and oral routes is not expected. Therefore, based on the adequacy of the margins between estimated exposure to industry-restricted HFO substances and critical effect levels, it is proposed that the five industry-restricted HFOs (CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8) are not entering the environment in a quantity or concentration or under conditions that constitute or may constitute a danger in Canada to human life or health.

It is therefore proposed that these five industry-restricted HFOs listed under CAS RNs 64741-75-9, 68783-08-4, 70592-76-6, 70592-77-7 and 70592-78-8 do not meet any of the criteria set out in section 64 of CEPA 1999.

As substances listed on the DSL, their import and manufacture in Canada are not subject to notification under subsection 81(1) of CEPA 1999. Given the potentially hazardous properties of these substances, there is concern that new activities that have not been identified or assessed could lead to these substances meeting the criteria set out in section 64 of the Act. Therefore, application of the Significant New Activity provisions of the Act to these substances is being considered. This would require that any proposed new manufacture, import use or transport be subject to further assessment, to determine if the new activity requires further risk management consideration.

References

Albers PH, Szaro RC. 1978. Effects of no. 2 fuel oil on common eider eggs. Mar Pollut Bull 9:138–139.

Anderson JW, Neff JM, Cox BA, Tatem HE, Hightower GM. 1974. Characteristics of dispersions and water-soluble extracts of crude and refined oils and their toxicity to estuarine crustaceans and fish. Mar Biol 27:75–88.

[AOPWIN] Atmospheric Oxidation Program for Microsoft Windows [Exposure Assessment Tools and Models]. 2008. Version 1.92a. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

[API] American Petroleum Institute. 1983. A 28-day dermal toxicity study of catalytic cracked clarified oil, API sample 81-15. Study conducted by Borriston Laboratories Inc. Washington (DC): API. API Medical Research Publication No.: 30-32854. [cited in CONCAWE 1998; ECB 2000a, b].

[API] American Petroleum Institute. 1985a. In-vivosister chromatid exchange (SCE) assay. API sample 81-15 catalytically cracked clarified oil (CAS 64741-62-4). Study conducted by Microbiological Associates, Inc. Washington (DC): API. API Health and Environmental Sciences Department Report No.: 32-32254. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1985b. An evaluation of the potential of RO-1, 81-15 and PS8-76D5-SAT to induce unscheduled DNA synthesis (UDS) in the in-vivo/in-vitro hepatocyte DNA repair assay. Study conducted by SRI International. Washington (DC): API. API Medical Research Publication No.: 32-32406. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1985c. Mutagenicity evaluation studies of catalytically cracked clarified oil API sample 81-15 in the rat bone marrow cytogenetic assay and in the mouse lymphoma forward mutation assay. Study conducted by Litton Bionetics, Inc. Washington (DC): API. API Medical Research Publication No.: 32-30534. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1985d. An evaluation of the potential of RO-1, 81-15 and PS8-76D5-SAT to induce unscheduled DNA synthesis (UDS) in primary rat hepatocyte cultures. Study conducted by SRI International. Washington (DC): API. API Medical Research Publication No.: 32-32407. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1985e. CHO/HGPRT (Chinese hamster ovary/hypoxanthine-guanine phosphoribosyl transferase) mammalian cell forward gene mutation assay of API sample 81-15 catalytic cracked clarified oil (CAS 64741-62-4). Study conducted by Pharmakon Research International, Inc. Washington (DC): API. API Medical Research Publication No.: 32-32118. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1985f. Sister chromatid exchange (SCE) assay in Chinese hamster ovary (CHO) cells. API sample 81-15 catalytically cracked clarified oil (CAS 64741-62-4). Study conducted by Microbiological Associates, Inc. Washington (DC): API. API Health and Environmental Sciences Department Report No.: 32-32750. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1986a. Salmonella/mammalian-microsome plate incorporation mutagenicity assay (Ames test). Study conducted by Microbiological Associates, Inc. Washington (DC): API. API Medical Research Publication No.: 33-30599. [cited in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1986b. Morphological transformation of BALB/3T3 mouse embryo cells. API sample 81-15 catalytically cracked clarified oil (CAS 64741-62-4). Study conducted by Microbiological Associates, Inc. Washington (DC): API. API Health and Environmental Sciences Department Report No.: 33-32638. [cited in in CONCAWE 1998; ECB 2000a; API 2004].

[API] American Petroleum Institute. 1989a. Short-term dermal tumorigenesis study of selected petroleum hydrocarbons in male CD-1 mice. Initiation and promotion phases. Study conducted by IIT Research Institute. Washington (DC): API. API Health and Environmental Sciences Department Report No.: 36-32643. [cited in CONCAWE 1998].

[API] American Petroleum Institute. 1989b. Lifetime dermal carcinogenesis/chronic toxicity screening bioassay of refinery streams in C3H/HeJ mice (AP-135r). Study conducted by Primate Research Institute, New Mexico State University. Washington (DC): API. API Health and Environmental Sciences Department Report No.: 36-31364. [cited in CONCAWE 1998; ECB 2000a].

[API] American Petroleum Institute. 2004. High Production Volume (HPV) Challenge Program. Robust summary of information on heavy fuel oils. Washington (DC): API.

Arnot JA. 2005. Bioconcentration factor and bioaccumulation factor assessments for organic chemicals on the Canadian domestic substances list: Task 1: Supplementation of Environment Canada’s BCF database. Report to Environment Canada, New Substances Branch. March.

Arnot JA. 2006. Bioconcentration factor and bioaccumulation factor assessments for organic chemicals on the Canadian domestic substances list: Database update. Report to Environment Canada, New Substances Branch. March.

Arnot J, Gobas F. 2003. A generic QSAR for assessing the bioaccumulation potential of organic chemicals in aquatic food webs. QSAR Comb Sci [Internet] [cited 2007 Mar 15]; 22(3):337–345 [restricted access].

Arnot J, Gobas F. 2004. A food web bioaccumulation model for organic chemicals in aquatic ecosystems. Environ Toxicol Chem 23:2343-2355.

Arnot J, Gobas F. 2006. A review of bioconcentration factor (BCF) and bioaccumulation factor (BAF) assessments for organic chemicals in aquatic systems. Environ Rev 14:257–297.

Arnot JA, Mackay D, Bonnell M. 2008a. Estimating metabolic biotransformation rates in fish from laboratory data. Environ Toxicol Chem 27(2):341–351.

Arnot JA, Mackay D, Parkerton TF, Bonnell M. 2008b. A database of fish biotransformation rates for organic chemicals. Environ Toxicol Chem 27(11):2263–2270.

Atkinson R. 1990. Gas-phase tropospheric chemistry of organic compounds: a review. Atmos Environ 24A:1–41.

[ATSDR] Agency for Toxic Substances and Disease Registry. 1999. Toxicological profile for total petroleum hydrocarbons (TPH). Atlanta (GA): U.S. Department of Health and Human Services, Public Health Service, ATSDR. [cited 2009 Mar].

Banerjee S. 1984. Solubility of organic mixtures in water. Environ Sci Technol 18:587–591.

Barron MG. 1990. Bioconcentration. Environ Sci Technol 24:1612–1618.

[BCFBAF] Bioconcentration Factor and Bioaccumulation Factor Estimation Program for Windows [Exposure Assessment Tools and Models]. 2008. Version 4.00. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

Bilboa E, Raingeard D, Diaz de Cerio O, Ortiz-Zarragoitia M, Ruiz P, Izagirre U, Orbea A, Marigómez I, Cajaraville MP, Cancio I. 2010. Effects of exposure to Prestige-like heavy fuel oil and to perfluorooctane sulfonate on conventional biomarkers and target gene transcription in the thicklip grey mullet Chelon labrosus. Aquat Toxicol 98:282–296.

Bingham E, Barkley W. 1979. Bioassay of complex mixtures derived from fossil fuels. Environ Health Perspect 30:157–163. [cited in IARC 1984].

Bingham E, Trosset RP, Warshawsky D. 1980. Carcinogenic potential of petroleum hydrocarbons: a critical review of the literature. J Environ Pathol Toxicol 3:483–563. [cited in IARC 1989a].

[BIOHCWIN] Biodegradation of Petroleum Hydrocarbons Estimation Program for Windows [Exposure Assessment Tools and Models]. 2008. Version 1.01a. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

[BIOWIN] Biodegradation Estimation Program for Windows [Exposure Assessment Tools and Models]. Version 4.10. 2009. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

Blackburn GR, Deitch RA, Schreiner CA, Mehlman MA, Mackerer CR. 1984. Estimation of the dermal carcinogenic activity of petroleum fractions using a modified Ames Assay. Cell Biol Toxicol 1:67–80.

Blackburn GR, Deitch RA, Schreiner CA, Mackerer CR. 1986. Predicting carcinogenicity of petroleum distillation fractions using a modified Salmonella mutagenicity assay. Cell Biol Toxicol 2:63–84.

Blaise C, Gagné F, Pellerin J, Hansen PD. 1999. Determination of vitellogenin-like properties in Mya arenaria hemolymph (Saguenay Fjord, Canada): a potential biomarker for endocrine disruption. Environ Toxicol 14:455–465.

Boethling RS, Howard PH, Beauman JA, Larosche ME. 1995. Factors for intermedia extrapolations in biodegradability assessment. Chemosphere 30(4):741–752.

Brecher S, Goode JW. 1983. BALB/3T3 transformation test. Aromatic pyrolysis oil. Project No. 2084. Prepared for Gulf Oil Chemicals Co., Houston, TX. Pittsburgh (PA): Gulf Life Sciences Center. [cited in U.S. EPA 2005].

Brecher S, Goode JW. 1984. Hepatocyte primary culture/DNA repair test of aromatic pyrolysis oil. Project No. 2083. Prepared for Gulf Oil Chemicals Co., Houston, TX. Pittsburgh (PA): Gulf Life Sciences Center. [cited in U.S. EPA 2005].

Broman DC, Lindberg NC, Zwbuhr Y. 1990. An in-situ study on the distribution, biotransformation and flux of polycyclic aromatic hydrocarbons (PAHs) in an aquatic food chain (seston-Mytilus edulis L.-Somateria mollissima) from the Baltic: an ecotoxicological perspective. Environ Toxicol Chem 9:429–442.

Bruner KA, Fisher SW, Landrum PF. 1994. The role of the zebra mussel, Dreissena polymorpha, in contaminant cycling: I. The effect of body size and lipid content on the bioconcentration of PCBs and PAHs. J Great Lakes Res 10(4):725–734.

Burkhard L, Lukasewycz M. 2000. Some bioaccumulation factors and biota–sediment accumulation factors for polycyclic aromatic hydrocarbons in lake trout. Environ Toxicol Chem 19(5):1427–1429.

Byrne CJ, Calder JA. 1977. Effect of the water-soluble fractions of crude, refined and waste oils on the embryonic and larval stages of the quahog clam Mercenaria sp. Mar Biol 40:225–231.

Canada. 1994. Priority substances list assessment report - Polycyclic aromatic hydrocarbons [Internet]. Ottawa (ON): Environment Canada; Health Canada.

Canada. 1999. Canadian Environmental Protection Act, 1999. S.C., 1999, c. 33. Canada Gazette, Part III: Vol. 22, No. 3.

Canada. 2000. Canadian Environmental Protection Act, 1999: Persistence and Bioaccumulation Regulations, P.C. 2000-348, 23 March 2000, SOR/2000-107. Canada Gazette, Part II: Vol. 134, No. 7.

Canada. 2001. Canada Shipping Act, 2001. S.C., 2001, c. 26. Canada Gazette, Part III: Vol. 24, No. 4.

Canada. 2010. Petroleum Refinery Liquid Effluent Regulations (C.R.C., c. 828) to the Fisheries Act. [cited 2010 Feb 19]

[CanLII] Canadian Legal Information Institute [CanLII Databases]. 2001– . Ottawa (ON): CanLII. [cited 2009].

Carlson RM, Oyler AR, Gerhart EH, Caple R, Welch KJ, Kopperman HL, Bodenner D, Swanson D. 1979. Implications to the aquatic environment of polynuclear aromatic hydrocarbons liberated from northern great plains coal. U.S. EPA Environmental Research Laboratory (EPA 600/3-79-093). [cited in Lampi and Parkerton 2009].

[CATABOL] Probabilistic assessment of biodegradability and metabolic pathways [Computer Model]. c2004-2008. Version 5.10.3. Bourgas (BG): Bourgas Prof. Assen Zlatarov University, OASIS Laboratory of Mathematical Chemistry.

[CCME] Canadian Council of Ministers of the Environment. 2008. Canada-wide standard for petroleum hydrocarbons (PHC) in soil. Canadian Environmental Quality Guidelines and Summary Table. Winnipeg (MB): Canadian Council of Ministers of the Environment. Report number PN-1398.

Charbeneau RJ, Linz DG, Newell CJ, Rixey WG, Ryan JA. 1998. Proceedings from the Pellston workshop on assessing contaminated soils: from soil-contaminant interactions to ecosystem management. Pellston (MI): Pellston Workshop on Assessing Contaminated Soils. p. 23–27.

[CITI] Chemicals Inspection and Testing Institute. 1992. Bioaccumulation and Biodegradation Data on Existing Chemicals Based on the CSCL Japan. Tokyo, Japan.

Colombo J, Cappelletti N, Migoya M, Speranza E. 2007. Bioaccumulation of anthropogenic contaminants by detritivorous fish in the Río de la Plata estuary: 1--Aliphatic hydrocarbons. Chemosphere 68:2128–2135.

[CONCAWE] CONservation of Clean Air and Water in Europe. 1998. Heavy fuel oils. Prepared by CONCAWE’s Petroleum Products and Health Management Groups. Brussels (BE): CONCAWE. Product Dossier No.: 98/109.

Coon NC, Albers PH, Szaro RC. 1979. No. 2 fuel oil decreases embryonic survival of great black-backed gulls. Bull Environ Contam Toxicol 21:152–156.

Cravedi J, Tulliez J. 1981. Distribution and elimination routes of a naphthenic hydrocarbon (dodecylcyclohexane) in rainbow trout (Salmo gairdneri). Bull Environ Contam Toxicol 26:337–344.

Cravedi J, Tulliez J. 1983. Hydrocarbon disposition, lipid content, and fatty acid composition in trout after long-term dietary exposure to n-alkanes. Environ Res 32(2):398–413.

Curry P, Kramer G, Newhook R, Sitwell J, Somers D, Tracy B, Oostdam JV. 1993. Reference values for Canadian populations. Prepared by the Environmental Health Directorate Working Group on Reference Values. Ottawa (ON): Health Canada. [update of 1988 version].

De Maagd PGJ. 1996. Polycyclic aromatic hydrocarbons: fate and effects in the aquatic environment. Ph.D. Thesis. Utrecht University. Utrecht, NL. [cited in European Commission 2008b].

De Voogt P, van Hattum B, Leonards P, Klamer JC, Govers H. 1991. Bioconcentration of polycyclic heteroaromatic hydrocarbons in the guppy (Poecilia reticula). Aquat Toxicol 20:169–194.

[ECB] European Chemicals Bureau. 2000a. IUCLID dataset for clarified oils (petroleum), catalytic cracked (CAS No. 64741-62-4).

[ECB] European Chemicals Bureau. 2000b. IUCLID dataset for fuel oil, No. 6. (CAS No. 68553-00-4).

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2004a. Fish, dietary bioaccumulation study, study no. 0400947. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2004b. Fish, aqueous bioaccumulation study, study no. 0409544. Annandale (NJ): ExxonMobil Biomedical Sciences Inc.

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2005a. Fish, dietary bioaccumulation study, study no. 0409547. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2005b. Fish, aqueous bioaccumulation test, study no. 0523644. Annandale (NJ): ExxonMobil Biomedical Sciences Inc.

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2005c. Fish, dietary bioaccumulation study, study no. 0523747. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2005d. Fish, dietary bioaccumulation study, study no. 0531047. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI Exxon Mobil Biomedical Sciences, Inc. 2006a. Fish, dietary bioaccumulation study, study no. 0674247. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2006b. Fish, dietary bioaccumulation test, study no. 0681647. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2007a. Fish, dietary bioaccumulation study, study no. 0796347T. ExxonMobil Biomedical Sciences Inc. Annandale, NJ. [cited in Lampi and Parkerton 2009].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2007b. Fish, dietary bioaccumulation study, study no. 0796347C. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi and Parkerton 2009].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2008a. Fish, dietary bioaccumulation study, study no. 0711047. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2008b. Fish, dietary bioaccumulation study, study no. 0821047. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2008c. Fish, dietary bioaccumulation study, study no. 0818447. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2009. Fish, dietary bioaccumulation study, study no. 0818447. Annandale (NJ): ExxonMobil Biomedical Sciences Inc. [cited in Lampi et al. 2010].

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2010a. Fish, dietary bioaccumulation study, study no. 0818447. Supplement. Annandale (NJ): ExxonMobil Biomedical Sciences Inc.

[EMBSI] Exxon Mobil Biomedical Sciences, Inc. 2010b. Fish, dietary bioaccumulation study, study no. 0681647. Annandale (NJ): ExxonMobil Biomedical Sciences Inc.

Environment Canada. 1994. Polycyclic aromatic hydrocarbons. ISBN 0-662-22209-1, Cat. No. En40-215/42E.

Environment Canada. 2007. Guidance for conducting ecological assessments under CEPA 1999. Science Resource Technical Series, Draft Module on QSARs. Draft document prepared by Environment Canada, Existing Substances Division.

Environment Canada. 2008. Data for petroleum sector stream substances collected under the Canadian Environmental Protection Act, 1999, section 71: Notice with respect to certain high priority petroleum substances. Data prepared by: Environment Canada, Oil, Gas, and Alternative Energy Division.

Environment Canada. 2009. Data for petroleum sector stream substances collected under the Canadian Environmental Protection Act, 1999, section 71: Notice with respect to potentially industry-limited high priority petroleum substances. Data prepared by: Environment Canada, Oil, Gas, and Alternative Energy Division.

Environment Canada. 2010a. Oil properties database.

Environment Canada. 2010b. Birds oiled at sea. [accessed 2010 Feb 17].

Environment Canada. 2010c. CEPA Environmental Registry. Toxic Substances List – Schedule 1.

Environment Canada. 2011. Spill line data report on petroleum product spills 2000–2009. Gatineau (QC): Environment Canada, Environmental Emergencies Division. Unpublished information.

[EPIsuite] Estimation Programs Interface Suite for Microsoft Windows [Exposure Assessment Tools and Models]. 2008. Version 3.4. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

[EQC] Equilibrium Criterion Model. 2003. Version 2.02. Peterborough (ON): Trent University, Centre for Environmental Modelling and Chemistry.

[ESIS] European Chemical Substances Information System [database on the Internet]. 2008. Database developed by the European Chemicals Bureau (ECB). [cited 2011 Aug 29].

European Commission. 1994. Commission Directive 94/69/EC of 19 December 1994. Annex II. Official Journal of the European Communities. 31.12.94. L381, vol. 37. European Commission. 21st ATP.

European Commission. 2008a. Details on substances classified in Annex VI to Regulation (EC) No 1272/2008 [European Chemical Substances Information System]. Database developed by the European Commission. [cited 2011 Aug 29].

European Commission. 2008b. European Union risk assessment report. Coal-tar pitch, high temperature. CAS No.: 65996-93-2. EINECS No.:266-028-2. Risk assessment. Environment. May 2008. Luxembourg (NL): Office for Official Publications of the European Communities.

Evans MS, Landrum PF. 1989. Toxicokinetics of DDE, benzo(a)pyrene, and 2,4,5,2’, 4’,5’-hexachlorobiphenyl in Pontoporeia hoyi and Mysis relicta. J Great Lakes Res 15:589–600.

Farrow MG, McCarroll N, Cortina T, Draus M, Munson A, Steinberg M, Kirwin C, Thomas W. 1983. In vitro mutagenicity and genotoxicity of fuels and paraffinic hydrocarbons in the Ames, sister chromatid exchange, and mouse lymphoma assays. Toxicologist 3:36 (Abstract No. 144). [cited in IARC 1989a].

Feuston MH, Mackerer CR. 1996. Developmental toxicity of clarified slurry oil, syntower bottoms, and distillate aromatic extract administered as a single oral dose to pregnant rats. J Toxicol Environ Health 49(1):45–66.

Feuston MH, Kerstetter SL, Singer EJ, Mehlman MA. 1989. Developmental toxicity of clarified slurry oil applied dermally to rats. Toxicol Ind Health 5(3):587–599.

Feuston MH, Low LK, Hamilton CE, Mackerer CR. 1994. Correlation of systemic and developmental toxicities with chemical component classes of refinery streams. Fundam Appl Toxicol 22:622–630.

Feuston MH, Hamilton CE, Mackerer CR. 1997. Systemic and developmental toxicity of dermally applied syntower bottoms in rats. Fundam Appl Toxicol 35:166–176.

Fox K, Zauke G, Butte W. 1994. Kinetics of bioconcentration and clearance of 28 polychlorinated biphenyl congeners in zebrafish (Brachydanio rerio). Ecotoxicol Environ Saf 28(1):99–109.

Fuhr B. 2008. Hydrocarbon composition of fuel products for risk assessment modelling. Final Report. Edmonton (AB): Alberta Research Council, Fuels and Lubricants Laboratory for Environment Canada..

Getmanets IY. 1967. Comparative evaluation of the carcinogenic properties of cracking residues of high- and low-paraffin petroleum. Gig Tr Prof Zabol 11:53–55. [cited in IARC 1989b].

Giddings JM, Parkhurst BR, Gehrs CW, Millemann RE. 1980. Toxicity of a coal liquefaction product to aquatic organisms. Bull Environ Contam Toxicol 25:1–6. [cited in CONCAWE 1998].

Gobas F, Clark K, Shiu W, Mackay D. 1989. Bioconcentration of polybrominated benzenes and biphenyls and related superhydrophobic chemicals in fish: role of bioavailability and elimination into the feces. Environ Toxicol Chem 8:231–245.

Golder Associates Ltd. 2006. Final Draft Report on Lakewater and Aquatics Long-term Monitoring – 2005–2006 Interpretive Report, Wabamun Lake Derailment Site, Canadian National Railway Company. PIN 2401257. Wabamun, Alberta.

Gordon T. 1983. Nine-day repeated dose inhalation toxicity study in rats. Aromatic pyrolysis oil. Project No. 2035. Prepared for Gulf Oil Chemicals Co., Houston, TX. Pittsburgh (PA): Gulf Life Sciences Center. [cited in U.S. EPA 2005].

Gossiaux DC, Landrum PF, Fisher SW. 1996. Effect of temperature on the accumulation kinetics of PAHs and PCBs in the zebra mussel, Dreissena polymorpha. J Great Lakes Res 22(2):379–388.

Gustafson JB, Tell JG, Orem D. 1997. Selection of representative TPH fractions based on fate and transport considerations. Vol. 3. Total Petroleum Hydrocarbon Criteria Working Group Series, Association for Environmental Health and Sciences Foundation Publication. 109 p.

Hardy R, Mackie P, Whittle K, McIntyre A. 1974. Discrimination in the assimilation of n-alkanes in fish. Nature 252:577–578.

Health Canada. 1998. Exposure factors for assessing total daily intake of priority substances by the general population of Canada. Ottawa (ON): Health Canada, Environmental Health Directorate. Unpublished report.

Hebert GW, Kussat RH. 1972. A laboratory evaluation of the toxicity of certain oils and chemical oil dispersants to juvenile coho salmon and staghorn sculpins. Canada, Department of the Environment Fisheries Service. Technical Report 1972-6.

Hoberman AM, Christian MS, Lovre S, Roth R, Koschier F. 1995. Developmental toxicity study of clarified slurry oil (CSO) in the rat. Fundam Appl Toxicol 28:34–40.

Hollister TA, Ward GS, Parrish PR. 1980. Acute toxicity of a #6 fuel oil to marine organisms. Bull Environ Contam Toxicol 24:656–661.

Hopkinson R. 2008. Priority substances under Environment Canada’s Chemicals Management Plan for the petroleum sector. Richmond (BC): Levelton Consultants Ltd.

[HYDROWIN]. Aqueous Hydrolysis Fate Program for Windows [Exposure Assessment Tools and Models]. 2008. Version 2.00. Washington (DC): U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics; Syracuse (NY): Syracuse Research Corporation.

[IARC] IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. 1984. Mineral oils. IARC Monogr Eval Carcinog Risks Hum 33:87–168.

[IARC] IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. 1989a. Fuel oils. IARC Monogr Eval Carcinog Risks Hum 45:239–270.

[IARC] IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. 1989b. Occupational exposures in petroleum refining. IARC Monogr Eval Carcinog Risks Hum 45:39–117.

Jimenez BD, Cirmo CP, McCarthy JF. 1987. Effects of feeding and temperature on uptake, elimination and metabolism of benzo(a)pyrene in the bluegill sunfish (Lepomis macrochirus). Aquat Toxicol 10:41–57.

[JNITE]. 2010. Japanese National Institute of Technology and Evaluation. 2010. Official Bulletin of Economy, Trade and Industry [database on the Internet]. [accessed 2010 Sept]. Biodegradation and Bioconcentration of the Existing Chemical Substances under the Chemical Substances Control Law.

Jonsson G, Bechmann RK, Bamber SD, Baussant T. 2004. Bioconcentration, biotransformation, and elimination of polycyclic aromatic hydrocarbons in Sheepshead minnows (Cyprinodon variegates) exposed to contaminated seawater. Environ Toxicol Chem 23: 1538–1548. [Cited in Golder Associates Ltd. 2006]

Khan SH, Goode JW. 1984. Micronucleus test: aromatic pyrolysis oil orally for 2 days. Project No. 2082. Prepared for Gulf Oil Chemicals Co., Houston, TX. Pittsburgh (PA): Gulf Life Sciences Center. [cited in U.S. EPA 2005].

Kizu R, Kato S, Usui O, Hayakawa K. 1999. Estrogenic activity of heavy oil and its assay method. Bunseki Kagaku 48:617–622.

Kizu R, Ishii K, Kobayashi J, Hashimoto T, Koh E, Namiki M, Hayakawa K. 2000. Antiandrogenic effect of crude extract of C-heavy oil. Materials Science and Engineering: C 12(1-2):97–102.

Koyama J, Uno Seiichi, Kohno K. 2004. Polycyclic aromatic hydrocarbon contamination and recovery characteristics in some organisms after the Nakhodka oil spill. Mar Poll Bull 49:1054–1061.

Lampi M, Parkerton T. 2009. Bioaccumulation assessment of PAHs. Prepared for CONCAWE. Documents Library.

Lampi M, Paumen ML, Parkerton T. 2010. An evaluation of the persistence, bioaccumulation and toxicity of petroleum hydrocarbons. Report prepared for CONCAWE, Brussels, Belgium. Annandale (NJ): ExxonMobil Biomedical Sciences, Inc.; Machelen (BE): ExxonMobil Petroleum and Chemical.

Lewis SC. 1983. Crude petroleum and selected fractions. Skin cancer bioassays. Prog Exp Tumor Res 26:68–84. [cited in IARC 1989b].

Lyman W, Reehl W, Rosenblatt D, editors. 1990. Handbook of chemical property estimation methods. Environmental behaviour of organic compounds. Washington (DC): American Chemical Society.

MacLean MM, Doe KG. 1989. The comparative toxicity of crude and refined oils to Daphnia magna and Artemia. Manuscript Report EE-111. Ottawa (ON): Environment Canada. 72 p. [cited in Environment Canada 2010a].

Martin-Skilton R, Saborido-Rey F, Porte C. 2008. Endocrine alteration and other biochemical responses in juvenile turbot exposed to the Prestige fuel oil. Sci Tot Environ 404:68–76

McCain BB, Hodgins HO, Gronlund WD, Hawkes JW, Brown DW, Myers MS. 1978. Bioavailability of crude oil from experimentally oiled sediment to English sole Parophrys vetulus. J Fish Res Board Can 35(5):657–664.

McCarthy JF, Jimenez BD, Barbee T. 1985. Effect of dissolved humic material on accumulation of polycyclic aromatic hydrocarbons: Structure-activity relationships. Aquat Toxicol 7:145–156.

McKee RH, Nicolich MJ, Scala RA, Lewis SC. 1990. Estimation of epidermal carcinogenic potency. Fundam Appl Toxicol 15:320–328.

McLeese DW, Burridge LE. 1987. Comparative accumulation of PAHs in four marine invertebrates. In Capuzzo IM, Kester DR, editors. Oceanic processes in marine pollution. Malabar (FL): Kirger RE. p. 109–118 [cited in European Commission 2008b].

Michigan. 2010. Michigan Department of Natural Resources and Environment. Oil intoxication. [accessed 2010 Feb 17].

[Mobil] Mobil Oil Corporation. 1985. A modified Ames pre-incubation mutagenesis assay for determination of specific mutagenicity of the DMSO extract of heavy vacuum gas oil. Study No. 52261. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1987a. Micronucleus assay of bone marrow red blood cells from rats treated via dermal administration of heavy vacuum gas oil. Study No. 61591. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1987b. Metaphase analysis of Chinese hamster ovary (CHO) cells treated in vitro with a DMSO extract of heavy vacuum gas oil. Study No. 52262. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1987c. Clarified slurry oil: developmental toxicity study in rats. Study No. 50541. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in ECB 2000b].

[Mobil] Mobil Oil Corporation. 1987d. A static 96-hour acute toxicity study of process oil to bluegill sunfish. Pennington (NJ): Mobil Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1987e. A static 48-hour acute toxicity study of process oil to Daphnia magna. Pennington (NJ): Mobil Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1987f. A static 96-hour acute toxicity study of process oil to Selenastrum capricornutum. Pennington (NJ): Mobil Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1988. Thirteen-week dermal administration of syntower bottoms to rats. Study No. 62710. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in ECB 2000b].

[Mobil] Mobil Oil Corporation. 1990. Developmental toxicity study in rats exposed dermally to Ferndale syntower bottoms. Study No. 62934. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in ECB 2000b].

[Mobil] Mobil Oil Corporation. 1991. Developmental toxicity study in rats exposed dermally to heavy atmospheric gas oil. Study No. 64146. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1992. Thirteen-week dermal administration of heavy atmospheric gas oil to rats. Study No. 63456. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in API 2004].

[Mobil] Mobil Oil Corporation. 1994a. Developmental toxicity study in rats exposed dermally to heavy coker gas oil (HCGO). Study No. 64168. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in ECB 2000b].

[Mobil] Mobil Oil Corporation. 1994b. Thirteen-week dermal administration of Joliet heavy coker gas oil to rats. Study No. 64165. Princeton (NJ): Mobil Oil Corporation, Environmental and Health Science Laboratory. [cited in ECB 2000b].

[MSDS] Material Safety Data Sheet. 2007. HAGO [heavy atmospheric gas oil]. Calgary (AB): NOVA Chemicals.

Muijs B, Jonker M. 2010. A closer look at bioaccumulation of petroleum hydrocarbon mixtures in aquatic worms. Environmental Toxicology and Chemistry, V29, 9, pp 1943 – 1949.

[NCI] National Chemical Inventories [National Chemical Inventory (NCI) System Requirements]. 2006. Columbus (OH): American Chemical Society, Chemical Abstracts Service. Available from:

[NEB] National Energy Board. 2008. Focus on safety and environment: a comparative analysis of pipeline performance 2000–2006.

Neff JM, Anderson JW. 1981. Response of marine animals to petroleum and specific petroleum hydrocarbons. London (UK): Applied Science Publishers Ltd. [cited in CONCAWE 1998].

Neff JM, Cox BA, Dixit D, Anderson JW. 1976. Accumulation and release of petroleum-derived aromatic hydrocarbons by four species of marine animals. Mar Biol 38:279–289.

Newsted JL, Giesy JP. 1987. Predictive models for photoinduced acute toxicity of polycyclic aromatic hydrocarbons to Daphnia magna, Strauss (Cladocera, crustacean). Environ Toxicol Chem 6:445–461.

Nfon E, Cousins IT, Broman D. 2008. Biomagnification of organic pollutants in benthic and pelagic marine food chains from the Baltic Sea. Sci Total Environ 397:190–204. [cited in Lampi et al. 2010].

Niimi A, Palazzo V. 1986. Biological half-lives of eight polycyclic aromatic hydrocarbons (PAHs) in rainbow trout (Salmo gairdneri). Water Res 20(4):503–507.

Nishimoto S, Yamawaki M, Akiyama K, Kakinuma Y, Kitamura S, Sugahara T. 2009. Severe abnormalities in the reproductive organs of mice caused by chemical substances contained in heavy oil. J Toxicol Sci 34:239–244.

[OECD] Organisation for Economic Co-operation and Development. 2009. Emission scenario document on transport and storage of chemicals. OECD series on emission scenario documents, No. 24.

Oliver B, Niimi A. 1988. Trophodynamic analysis of polychlorinated biphenyl congeners and other chlorinated hydrocarbons in the Lake Ontario ecosystem. Environ Sci Technol 22:388–397

Ortiz-Zarragoitia M, Garmendia L, Barbero MC, Serrano T, Marigómez I, Cajaraville MP. 2011. Effects of the fuel oil spilled by the Prestige tanker on reproduction parameters of wild mussel populations. J Environ Monit 13:84–94.

Pancirov R, Brown R. 1975. Analytical methods for polynuclear aromatic hydrocarbons in crude oils, heating oils, and marine tissues. Proceedings of a Conference on Prevention and Control of Oil Pollution, San Francisco, California. Washington (DC): American Petroleum Institute. p. 103–113. [cited in Potter and Simmons 1998].

Papciak MS, Goode JW. 1984. CHO/HGPRT test using aromatic pyrolysis oil. Project No. 2081. Prepared for Gulf Oil Chemicals Co., Houston, TX. Pittsburgh (PA): Gulf Life Sciences Center. [cited in U.S. EPA 2005].

[PETROTOX] Spreadsheet model to calculate toxicity of petroleum hydrocarbons to aquatic organisms. 2009. Version 3.04. Brussels (BE): CONservation of Clean Air and Water in Europe (CONCAWE).

Potter T, Simmons K. 1998. Composition of petroleum mixtures. Vol 2. Total Petroleum Hydrocarbon Criteria Working Group Series, Association for Environmental Health and Sciences Foundation Publications. 109 p.

Rhodes SW, Risher JF. 1995. Toxicology profile for fuel oils. Atlanta (GA): U.S. Department of Health and Human Services, Agency for Toxic Substances and Disease Registry, Division of Toxicology/Toxicology Information Branch. 231 p.

[RMRI] Risk Management Research Institute. 2007. Quantitative assessment of oil spill risk for the south coast of Newfoundland and Labrador. St. John’s (NL): RMRI (Canada). Draft report for Transport Canada.

Rossi SS, Anderson JW, Ward GS. 1976. Toxicity of water-soluble fractions of four test oils for the polychaetous annelids, Neanthes arenaceodentata and Capitella capitata. Environ Pollut 10:9–18.

Saffiotti U, Shubik P. 1963. Studies on promoting action in skin carcinogenesis. Natl Cancer Inst Monogr 10:489–507. [cited in IARC 1984].

Salazar-Coria L, Amezxua-Allieri MA, Tenorio-Torres M, Gonzales-Marcias C. 2007. Polyaromatic hydrocarbons (PAHs) and metal evaluation after a diesel spill in Oaxaca, Mexico. Bull Environ Contam Toxicol 79:462–467.

Salem H, Katz SA, editors. 2006. Inhalation toxicology. 2nd ed. Boca Raton (FL): CRC Press, Taylor & Francis Group.

[SCREEN3] Screening Tool Program for Windows [Screening Model]. 1996. Version 4.10. Research Triangle Park (NC): U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Emissions, Monitoring, and Analysis Division. Technology Transfer Network - Support Center for Regulatory Atmospheric Modeling.

[SENES] SENES Consultants Limited. 2009. Review of current and proposed regulatory and non-regulatory management tools pertaining to selected petroleum substances under the Chemicals Management Plan. Report to Health Canada. Richmond Hill (ON): SENES.

Shapiro DD, Getmanets IY. 1962. Blastomogenic properties of petroleum of different sources. Gig Sanit 27:38–42. [cited in IARC 1989b].

Shubik P, Saffiotti U. 1955. The carcinogenic and promoting action of low boiling catalytically cracked oils. Acta Unio Int Contra Cancrum 11:707–711. [cited in IARC 1984].

Simpson BJ. 2005. Analysis of petroleum hydrocarbon streams on the Health Canada CEPA/DSL Draft Maximal List. Report to the Canadian Petroleum Products Institute.

Smith WE, Sunderland DA, Sugiura K. 1951. Experimental analysis of the carcinogenic activity of certain petroleum products. Arch Ind Hyg Occup Med 4:299–314. [cited in IARC 1984].

[SPARC]. 2009. Sparc performs automated reasoning in chemistry [Internet]. U.S. Environmental Protection Agency. Ecosystems Research Division.

Sprague JB, Carson WG. 1970. Toxicity tests with oil dispersants in connection with oil spill at Chedabucto Bay, Nova Scotia. Fisheries Research Board of Canada. Technical Report No. 201.

Stegeman JJ, Teal JM. 1973. Accumulation, release and rention of petroleum hydrocarbons by the oyster Crassostrea virginica. Mar Biol 22:37–44.

Strobel CJ, Brenowitz AH. 1981. Effects of bunker C oil on juvenile horseshoe crabs (Limulus polyphemus). Estuarine Res Fed 4(2):157–159.

Sun Petroleum Products Co. 1979. Letter from Sun Petroleum Products Company submitting additional information requested in USEPA letter dated February 6, 1979, with attachments. EPA/OTS No. 8EHQ-0379-0140. NTIS/OTS-0200438. [Abstract]. [cited in TOXLINE 2009].

Szaro RC. 1979. Bunker C fuel oil reduces mallard egg hatchability. Bull Environm Contam Toxicol. 22: 731-732.

Szaro RC, Albers PH, Coon N. 1978. Petroleum: effects on mallard egg hatchability. J Wildl Manag 42(2):404–406.

Tagatz ME. 1961. Reduced oxygen tolerance and toxicity of petroleum products to juvenile American shad. Chesapeake Sci 2:65–71.

Takeuchi I, Miyoshi N, Mizukawa K, Takada H, Ikemoto T, Omori K, Tsuchiya K. 2009. Biomagnification profiles of polycyclic aromatic hydrocarbons, alkylphenols and polychlorinated biphenyls in Tokyo Bay elucidated by δ¹³C and δ¹⁵N isotope ratios as guides to trophic web structure. Mar Pollut Bull 58:663–671.

Tatem HE, Cox BA, Anderson JW.1978. The toxicity of oils and petroleum hydrocarbons to estuarinecrustaceans. Estuarine Coastal Mar Sci 6:365–373.

Tolls J, van Dijk J. 2002. Bioconcentration of n-dodecane and its highly branched isomer 2,2,4,6,6-pentamethylheptane in fathead minnows. Chemosphere 47:1049–1057.

[TOPKAT] Toxicity Prediction by Komputer Assisted Technology [Internet]. 2004. Version 6.1. San Diego (CA): Accelrys Software Inc.

[TOXLINE] Toxicology Literature Online [database on the Internet]. 1974– . Bethesda (MD): National Library of Medicine (US). [revised 2009 Apr 18; cited 2009 Apr 21].

Transport Canada. 2010. Spill Prevention: National Aerial Surveillance Program.

[UBTL] Utah Biomedical Testing Laboratory Inc. 1990. 28 day dermal toxicity study in rats. Salt Lake City (UT): UBTL. Report No.: ATX-90-0066. [cited in API 2004].

[UBTL] Utah Biomedical Testing Laboratory Inc. 1994. A developmental toxicity screen in female rats administered F-228 dermally during gestation days 0 to 20. Study No. 66479. Salt Lake City (UT): UBTL. Report No.: ATX-91-0267. [cited in API 2004].

[U.S. EPA] United States Environmental Protection Agency. 1992. Screening procedures for estimating the air quality impact of stationary sources. Revised. Research Triangle Park (NC): U.S. EPA, Office of Air and Radiation, Office of Air Quality Planning and Standards. Report No.: EPA-454/R-92-019. Available from: EPA National Library Network - OLS : Record.

[U.S. EPA] United States Environmental Protection Agency. 1995. Great Lakes Water Quality Initiative Technical Support Document for the Procedure to Determine Bioaccumulation Factors, Report No. EPA-820-B-95-005.

[U.S. EPA] United States Environmental Protection Agency. 2005. Appendix II. Robust summaries of studies used to characterize the fuel oils category. Washington (DC): U.S. Environmental Protection Agency, High Production Volume Chemical Program.

[U.S. EPA] United States Environmental Protection Agency. 2008a. Transportation and marketing of petroleum liquids. In: AP 42, Compilation of air pollutant emission factors, vol 1. 5th ed. Washington (DC): U.S. EPA.

[U.S. EPA] United States Environmental Protection Agency. 2008b. Biota-Sediment Accumulation Factor Data Set. [cited in Lampi et al. 2010].

Vandermeulen HH, Lee RW. 1986. Lack of mutagenic activity of crude and refined oils in the unicellular alga Chlamydomonas reinhardtii. Bull Environ Contam Toxicol 36:250–253. [cited in IARC 1989a].

Vandermeulen HH, Foda A, Stuttard C. 1985. Toxicity vs mutagenicity of some crude oils, distillates, and their water soluble fractions. Water Res 19:1283–1289. [cited in IARC 1989a].

Varanasi U, Stein J, Nishimoto M. 1989. Biotransformation and disposition of PAH in fish. In Varanasi U, ed, Metabolism of Polycyclic Aromatic Hydrocarbons in the Aquatic Environment, Boca Raton (FL): CRC Press. pp 93–150. [cited in Burkhard and Lukasewycz 2000].

Vrabie CM, Candido A, van Duursen MBM, Jonker MTO. 2010. Specific in vitro toxicity of crude and refined petroleum products: II. Estrogen (α and β) and androgen receptor-mediated responses in yeast assays. Environ Toxicol Chem 29(7):1529–1536.

Vrabie CM, Candido A, van den Berg H, Murk AJ, van Duursen MBM, Jonker MTO. 2011. Specific in vitro toxicity of crude and refined petroleum products: 3. Estrogenic responses in mammalian assays. Environ Toxicol Chem 30:973–980.

Wan Y, Jin X, Hu J, Jin F. 2007. Trophic dilution of polycyclic aromatic hydrocarbons (PAHs) in a marine food web from Bohai Bay, North China. Environ Sci Technol 41:3109–3114.

Weil CS, Condra NI. 1977. Experimental carcinogenesis of pyrolysis fuel oil. Am Ind Hyg Assoc J 38:730–733. [cited in IARC 1989b].

Weinstein JE, Oris JT. 1999. Humic acids reduce the bioaccumulation and photoinduced toxicity of fluoranthene to fish. Environ Toxicol Chem 18(9):2087–2094.

White DH, Kirke KA, Coon NC. 1979. Effect of No. 2 fuel oil on hatchability of marine and estuarine bird eggs. Bull Environ Contam Toxicol 21:7–10.

Yakata N, Sudo Y, Tadokoro H. 2006. Influence of dispersants on bioconcentration factors of seven organic compounds with different lipophilicities and structures Chemosphere 64:1885–1891. [cited in Lampi et al. 2010].

Yvette C, Van Tongeren M, Urbanus J, Cherrie JW. 2011. An assessment of dermal exposure to heavy fuel oil (HFO) in occupational settings. Ann Occup Hyg 55(3):319–328.

Zhou S, Heras H, Ackman RG. 1997. Role of adipocytes in the muscle tissue of Atlantic salmon (Salmo salar) in the uptake, release and retention of water-soluble fraction of crude oil hydrocarbons. Mar Biol 127:545–553.

Appendix

Appendix of the Draft Screening Assessment

• Appendix 1: Petroleum Substance Groupings
• Appendix 2: Physical and Chemical Data Tables for Industry-restricted HFOs
• Appendix 3: Measures Designed to Prevent, Minimize or Manage Unintentional Releases
• Appendix 4: Release Estimation of Industry-restricted HFOs During Transportation
• Appendix 5: Modelling Results for Environmental Properties of Industry-restricted HFOs
• Appendix 6: Modelling Results for Human Exposure to Industry-restricted HFOs
• Appendix 7: Summary of Health Effects Information from Pooled Toxicological Data for HFO Substances

---

Footnotes

---