Assessment of physical human health risks and hazards of crude oil. transportation through Gitga at Territory. December 2011

Transcription

1 Assessment of physical human health risks and hazards of crude oil transportation through Gitga at Territory December 2011 Prepared by Laurie Hing Man Chan, Ph.D. Laurie Chan is a Professor and Canada Research Chair in Toxicology and Environmental Health at the University of Ottawa. He is also the Director of the Center for Advanced Research in Environmental Genomics (CAREG). He was formerly the holder of Dr. Donald Rix B.C. Leadership Chair in Aboriginal Environmental Health and Professor at the School of Health Sciences at the University of Northern British Columbia from 2006 to 2011 and held one of the 6 NSERC Northern Research Chairs between 2002 and 2006 at McGill University. Prof. Chan s research in environmental and nutritional toxicology spans from the lab developing new techniques for contaminant analysis, to participatory research in the community on the risk and benefits of traditional foods and impact of environmental change on food security. His research is supported by NSERC, CIHR, and various government departments. He is the Principal Investigator of two national projects on First Nations and Inuit food safety and environmental health. Prof. Chan has also served as an advisor for international and national governments and organizations and numerous Aboriginal communities on environmental health issues. Prof. Chan was selected as a Fellow by the Leopold Leadership Program of Stanford University in 2008 and served on an expert panel on benefits and risks of fish consumption by the World Health Organization in

2 Executive Summary The proposed Enbridge Northern Gateway Project will involve about 220 loaded crude oil and condensate tankers per year traveling from Kitimat down the Douglas Channel and through the Inside Passage of Coastal British Columbia to the open Pacific Ocean. This has created concern from the Gitga at and other coastal Aboriginal communities on potentially significant risks to their food supply by compromising access to fishing grounds and through habitat destruction and toxicity from petrochemical spills, as well as other risks to their safety from the tanker traffic. The objective of this report is to identify the health hazard faced by the Gitga at First Nations from chemical pollution in air, water and food in case of an oil spill and to assess the health risk related to exposure to toxic chemicals from consumption of traditional foods and loss of traditional food in the diet of the Gitga ata people as a result of contamination of the traditional harvest by an oil or condensate spill. Because of the lack of information on the relationship between dose and mental health effects, this report only focuses on the physical health effects of chemical exposure. Hazard identification was conducted using the oil spill statistics reported in Alaska between 1985 to Given the proposed frequency of tanker traffic, the probability of an average oil spill of 68,700 bbl (11 kilo tonnes) along Douglas Channel and the Inland Passage would be 1 spill in 11.3 years. That equates to an average input of 1 kilo tonnes per year of crude oil into the marine environment. The chemical composition of Alberta crude oil and condensate and their toxicity was reviewed. The major chemicals of concern are polycyclic aromatic hydrocarbons (PAH) that can potentially cause cancer and have adverse effects such as stunted growth, anemia, and kidney disease. The heavy shellfish consumers, women of childbearing age and children are the populations most vulnerable to the effects of PAH. The Levels of Concerns of PAH in shellfish were calculated to be ug/g. As the baseline levels of PAH in shellfish in the harvest grounds of the Gitga at are just below these levels, any increase in PAH as a result of oil spill or contamination will increase the risk of cancer and make the shellfish deemed to be unsafe for human consumption. Other 2

3 indirect health effects will likely result from loss of traditional food because of: 1) lack of availability of traditional food due to closures; 2) avoidance of an area by the local harvesters in locations that are perceived to be contaminated by an oil spill; 3) chemical effects on density and/or distribution of certain key species of traditional food that reduce its availability; and 4) increased marine traffic resulting in increased hazard of harvesting. Loss of traditional food will increase the risk of adverse effects of nutrient deficiency such as stunted growth, comprised immune functions, anemia, etc. The risk for chronic diseases such as diabetes will also increase. Any threat to the integrity of the traditional food system will have a significant adverse effect on the socio cultural and spiritual well being of the Gitga at people. Key findings of this study are summarized below: 1. Using the oil spill statistics reported in Alaska between 1985 to 1999 at 0.46 spills/bbbl handled and a mean spill size of 68,700 bbl/spill, the probability of an average oil spill along Douglas Channel and the Inland Passage was calculated to be 1 spill in 11.3 years or an average input of 1 kilo tonnes per year of crude oil into the marine environment at the proposed handling capacity of 525,000 bbl/day (p.11 12). 2. Using the oil spill statistics reported for the Trans Alaska Pipeline System between 1977 and 1998 at 0.12 spills/bbbl handled and a mean spill size of 742 bbl/spill, the probability of an oil spill from the pipeline was estimated to be 1 spill in 44 years or an average input of 2.7 tonnes per year (p.13 14). 3. As 92% of oil spill occurred at the terminals, the effects of oil pollution due to routine operation on the Gitga at will be less than that in Kitimat but the risk cannot be assessed due to lack of data (p.14). 4. Crude bitumen from the Athabasca tar sands is relatively heavy and requires blending with condensate that is a mixture of benzene, toluene and xylenes. Up to one third of this oil will evaporate in the first 24 hours after a spill but the remainder will form persistent residues and has a moderate to high risk of seafood contamination (p.16 17). 5. Alberta crude oil contains significant amount of PAH including the very toxic benzo[a]pyrene. PAH compounds are known carcinogens (p.17 19). 3

4 6. Because of the volatile properties of the condensate, the main health concerns are the acute effects of inhalation of these chemicals from air pollution. As they have short half life in the environment, the environmental health consequence of a spill of condensate will be much less than that of a spill of crude oil in a similar scale (p.22). 7. As many of the chemicals in the oil are volatile and toxic, there is a significant health risk to the Gitga at from air pollution after an oil spill. Potential health symptoms include headaches, dizziness, nausea, vomiting, cough, respiratory distress, and chest pain and reversible lung function impairment. The risk will increase if they participate in the remediation operation (p.32 34). 8. Most of the studies reported in the literature provide evidence on the relationship between exposure to spilled oils and the appearance of acute physical, psychological, genotoxic and endocrine effects in the exposed individuals. Indigenous peoples are more susceptible to mental health illnesses such as depression and Post traumatic stress disorder (p ). 9. Level of Concerns (LOC) of PAH was calculated using the standard approach from the Food and Drug Agency in the US. The LOCs of PAH in shellfish were calculated to be ug/g that is just below the baseline levels of PAH in shellfish in the harvest grounds of the Gitga at suggesting any increase in PAH as a result of oil spill or contamination will increase the risk of cancer and make the shellfish deemed to be unsafe for human consumption (p.40 43). 10. The supply of key species of traditional food of the Gitga at such as mussel, clam, oysters, seaweed and eulachon will likely be significantly decreased after an oil spill that will worsen their food security. Intake of key nutrients such as protein, iron, zinc, vitamin D and vitamin A will become problematic. Loss of traditional food is also a significant risk factor for chronic diseases such as diabetes (p.47 50). 11. Degradation of the environment and loss of the traditional food system will have a significant adverse effect on the socio cultural and spiritual well being of the Gitga at people (p.51). 4

5 Table of Contents 1. Introduction Probability of spills Spills at Sea Spills from Pipelines Pollution due to routine operation Chemical composition of Alberta crude oil and condensate Toxicity of Crude Oil Toxicity of Condensate Toxicity of Dispersants Toxicity from Controlled Burning of Crude Oil Potential human health effects Air pollution Toxicity of Volatile Organic Compounds (VOCs) Toxicity from Controlled Burning of Crude Oil Adverse effects caused by air pollution after crude oil spill into sea water Cumulative human health effects Risk Assessment for PAH exposure from seafood consumption Indirect health risk of loss of traditional food Conclusion References

6 1. Introduction Currently, one of the most environmentally risky projects proposed for the Pacific North Coast of British Columbia (B.C.) is the Enbridge Northern Gateway Project. This project involves the construction of two 1,170 kilometer pipelines between the oil sands of Alberta and the coastal community of Kitimat, B.C. One pipeline would carry oil (bitumen) from the oil sands to Kitimat, the other would convey condensate (used to thin tar sands oil so it can be transported through pipelines) from Kitimat to Alberta. The pipelines would cross approximately a thousand significant streams and rivers, many of which are in the Skeena and Upper Fraser watersheds. These two river systems represent B.C. s most important wild salmon rivers (Living Oceans Society 2010). The oil reaching Kitimat from Alberta would be loaded onto oil tankers which would then travel down the Douglas Channel and through the Inside Passage, bound primarily for Asian markets. It is estimated that approximately 525,000 barrels (75,000 metric tonnes) per day will be loaded onto tankers (Enbridge 2010). Condensate, moving in the reverse direction, is classified as a dangerous good by the federal government and is extremely toxic (Living Oceans Society 2010). The transport of oil and condensate will require to about 220 loaded oil tankers per year (depending on tanker size). Tankers and pipelines are essentially the most efficient means of transporting oil over long distances, costing on average 2 3 cents (US) per US gallon (3.8 L) per average route. As a result, these modes of transport are the favored option for the oil industry, and thus, tankers today transport approximately 2 billion metric tonnes of oil per year worldwide (Huber 2001, United Nations Council on Trade and Development 2006). 6

7 The Gitga at like other Aboriginal communities of the Pacific North Coast of British Columbia rely heavily on the availability of traditional foods, especially those harvested from marine systems, as a source of nutrition, income and the preservation of livelihoods, national identity, culture and traditional knowledge. Current proposals for industrial development on the North Coast especially those that would increase oil tanker traffic in northern waters could introduce significant risks to this food supply by compromising access to fishing grounds and through habitat destruction and toxicity from petrochemical spills. Determining the real impact of these risks and losses is, however, a complex undertaking. Traditional food supply represents more than a commodity (i.e. a source of food and income) to Aboriginal communities. It also plays an important role in the preservation of culture and community, as well as education and health services, including traditional medicines (Columbia River Intertribal Fish Commission 2010). These factors are recognized by Health Canada (2004) as the indices of the social determinants of human health. In recent years, and especially in the wake of the 1986 Ottawa Charter for Health Promotion, much awareness has been generated around the relationship between ecological health and human health. Numerous case studies demonstrate that damage to ecosystems, and thus to the resources that those ecosystems sustain, can become the root cause or an aggravating factor for many human health issues (Corvalan et al. 2005, Gilles and Lebel 2001). This relationship is particularly significant in the context of Aboriginal communities where there exists the common belief of a deep connection with the natural world (Assembly of First Nations 2003). As Greenwood and de Leeuw (2009) argue, the health and wellbeing of Indigenous peoples arises from this connection with the land (and water). Other studies within Aboriginal communities have demonstrated links between traditional food supply, the practice of traditional knowledge, and ultimately, the resilience of the community (Ommer et al. 2007; Parrish et al. 7

8 2007; Stephens 2008). It is clear then that any factor that compromises the health, quality and abundance of marine traditional foods is likely to have complex and adverse impacts to the Aboriginal populations of the Pacific North Coast (specifically, the Coastal First Nations and more particularly, the Gitga at). This report, however, will only focus on the assessment of the physical health risk related to i) exposure to toxic chemicals from consumption of traditional foods, and; ii) loss of traditional food in the diet of the people as a result of contamination of the traditional harvest by an oil spill. 8

9 2. Probability of spills Even when every effort is made to prevent oil spills, accidents will happen. The following section makes an attempt to calculate the probability of an oil spill using statistics published in the literature. The purpose to identify the magnitude of the hazard posed by the proposed plan for oil transportation. 2.1 Spills at Sea According to Anderson and LaBelle (2000), the rate of spills from oil tankers at sea has dropped dramatically over the past 35 years. Prior to this period (i.e. pre 1974), the spill rate worldwide, in the U.S. and for tankers carrying Alaskan North Slope (ANS) crude oil, had remained constant at approximately 0.83, 0.51 and 0.77 significant spills per billion barrels (Bbbl) handled, respectively; where any spill over 1000 bbl is considered significant and 1 Bbbl = 10 9 bbl. These values dropped to 0.67, 0.36 and 0.56 spills per Bbbl handled ( data). And finally, to 0.46, 0.29 and 0.46 spills per Bbbl handled ( data). In total, 11 tanker spills 1,000 bbl associated with the transportation of ANS crude oil occurred between 1977 and These 11 spills are made up of the 240,000 bbl Exxon Valdez spill (1989) and 10 other spills which were less than or equal to 15,000 bbl in size. The earliest three spills (1978, 1980, and 1981) occurred outside of U.S. waters, and all subsequent spills occurred in U.S. waters. Given the climatic and geographic similarities between Alaskan waterways and those of the narrow passageways of the Pacific North Coast, and in particular those leading down Douglas Channel, it is assumed that the risk of experiencing a major spill from an oil tanker in the waters of the Pacific North Coast will be equivalent to those of Alaska; equating to 0.46 spills per Bbbl handled (see 9

10 Table 1). The mean quantities of oil released per spill are summarized in Table 2. Table 1. Worldwide tanker spill rates ( data) Worldwide U.S. coastal & offshore waters Alaska North Slope crude oil tanker spill rates Spill Source Vol. oil handled (Bbbl) No. of Spills Spill Rate Vol. oil handled (Bbbl) No. of Spills Spill Rate Vol. oil handled (Bbbl) No. of Spills Spill Rate Spills greater than 1,000 bbl All Spills In Port At Sea Spills greater than 10,000 bbl All Spills In Port At Sea Spills greater than 100,000 bbl All Spills In Port At Sea Adapted from Anderson and Labelle (2000) Table 2. Average and median spill sizes by spill source ( data) Spill greater than 1,000 bbl Spill greater than 10,000 bbl Number of Mean Spill Median Spill Number of Mean Spill Median Spill Spill Source Spills Size (bbl) Size (bbl) Spills Size (bbl) Size (bbl) Worldwide Tankers Total ,100 7, ,800 56,900 In Port 50 56,300 3, ,300 43,000 At Sea 63 81,000 14, ,000 60,000 Tankers U.S. Waters Total 20 22,800 5, ,000 15,400 In Port 12 5,600 4, ,900 12,900 At Sea 8 48,600 14, ,100 17,800 Tankers Alaska North Slope Crude Total 8 37,100 7, ,900 15,000 In Port 4 5,600 5,

11 At Sea 4 68,700 14, ,900 15,000 Average (Worldwide, US, Alaska; In Port) 22 22,500 4, ,400 18,600 Average (Worldwide, US, Alaska; At Sea) 25 66,100 14, ,300 30,900 Adapted from Anderson and Labelle (2000) It is estimated that the proposed Enbridge pipeline would deliver approximately 525,000 barrels of crude oil per day to tankers. Given that Albertan tar sand oil (called bitumen) is an extremely heavy and thick form of petroleum, with a density of 1015 Kg/m 3 (CAPP 2008, Environment Canada 2008a), a million barrels would weigh approximately 173,000 metric tonnes. That equates to 31 million tonnes of crude oil handled each year (191.6 million barrels). At a mean spill rate of 0.46 spills per billion barrels (Bbbl) handled (or alternatively 1 spill per 2.2 Bbbl handled), the probability of an average oil spill of 68,700 bbl (11 KT) would be 1 spill in 11.3 years. That equates to an average input of 1 KT per year of crude oil into the marine environment (Table 3). Table 3. Calculations for spill rate One barrel = 159 L = m 3 Oil Density = 1,015 kg/m 3 Pipeline delivers: 525,000 bbl/day = 191,625,000 bbl/yr = 30,468 million L/yr = million tons/yr One barrel = 161 Kg = tons Mean spill Rate: 0.46 spills/bbbl handled Mean spill size: 68,700 bbl/spill Spill frequency 1 spill of size 68,700 bbl per 11.3 years Spills Spill frequency 1 spill of size 11 KT per 11.3 years or 1 spill/2.2 Bbbl handled Mean (KT/yr) Mean (bbl/yr) 1 6,000 The above calculations are based on the statistics reported between and may be somewhat outdated. With the improvement of technology and tighter regulations, the number and amounts of oil spills have decreased in the last decade ( ). According to the statistics reported by the International Tanker Owners Pollution Federation Limited (ITOPF), numbers of large spills (over

12 tonnes) have decreased from 7.9 spills per year on average between to 3.3 spills per year on average between ( services/data andstatistics/statistics/). The total quantity of oil spilt also decreased from 1,137,000 tonnes in the 1990s to 212,000 tonnes in the 2000s. In 2010, there were 4 large spills recorded with a total oil spilled of 10,000 tonnes. It is not possible to predict whether this downward trend will continue. Assuming the improvement of technology has plateaued, the estimate reported above may have over estimated the risk by a factor of 2 (from the number of spills) to 5 (from the quantity spilt). On the other hand, the probability of spill is calculated based on the estimated handling capacity of 525,000 bbl/day stated in the Enbridge proposal and has not taken into the account that the proposed pipeline is designed to support a capacity increase from 525,000 bbl/day to 850,000 bbl/day in future expansion phases. The estimated probability of spill will, therefore, increase by 60% if the pipeline is used at its full capacity. 12

13 2.2 Spills from Pipelines Nine crude oil spills 100 bbl associated with the Trans Alaska Pipeline System (TAPS) occurred between 1977 and Five of those spills were 1,000 bbl; all five occurred prior to 1982, with an average spill size of 4,200 bbl. Only one larger spill occurred after 1982 (i.e. a spill of 811 bbl in 1996). Thus, the spill rate for spills 500 bbl was calculated at 0.12 spills per Bbbl transported over the years No spills 1,000 bbl occurred after Thus, a rate for spills 1,000 bbl was not computed for (Table 4). Alaskan North Slope crude oil and condensate spills 500 bbl from onshore pipelines over the years averaged approximately 0.12 spills /Bbbl transported, with an average spill size of 672 bbl (Table 5) (Anderson and LaBelle 2000). In summary, the spill rates for the more recent data ( ) are consistent between the Trans Alaska Pipeline System and the onshore Alaska North Slope crude oil pipeline. Spill sizes range from a mean of 811 bbl for the former and 672 bbl for the latter. Thus, a mean of 742 bbl was used in the spill rate calculations below (Table 6). Table 4. Spills associated with Trans Alaska Pipeline System, spill size 500 bbl and 1,000 bbl. From Anderson and LaBelle (2000) Years Spill Source Number of Spills Throughput (Bbbl) Spill/Bbbl moved Spills 500 bbl 6* Spills 1,000 bbl 5^ Spills 500 bbl Spills 1,000 bbl 0 ^^ * Average spill size 3,635 bbl; median spill size 3,250; largest spill 7,000 bbl, ^ Average spill size 4,200 bbl; median spill size 4,000 bbl Only one spill, 811 bbl, ^^ No spills 1,000 bbl occurred between 1985 and 1998; therefore, no rate was calculated. 13

14 Table 5. Crude oil and condensate spills associated with onshore Alaska North Slope crude oil production, spill size 500 bbl, From Anderson and LaBelle (2000) Years Spill Source Number of Spills Production (Bbbl) Spill/Bbbl moved All Spills 5* Facilities Pipelines * Average spill size 672 bbl; median spill size 650 bbl; largest spill 925 bbl. Table 6. Spill rate calculations for pipelines Pipeline delivers: 525,000 bbl/day = 191,625,000 bbl/yr Mean spill size: 742 bbl/spill * Mean spill Rate: 0.12 spills/bbbl handled or 1 spill/8.3 Bbbl handled Spill frequency 1 spill of size 742 bbl per 44 years * mean of 811 and 672 bbl Spills Spill frequency 1 spill of size 0.12 KT per 44 years Mean (T/yr) Mean (bbl/yr) Pollution due to routine operation Accidental discharges of fuel, oil, condensate, or bilge will likely occur during the life of the project despite all regulations in place. In case of an accidental spill of oil or fuel at an anchorage or on route, there could be adverse local effects to marine seaweed/plants and invertebrates depending on the location. It was estimated that 92% oil spill occurred at the terminals during loading and unloading (Environment Canada 2011). Therefore the impact may be more significant in the harbor area of Kitimat and less on the harvest grounds of the Gitga at. It is, however, not possible to estimate the extent that these effects might be. 14

15 3. Chemical composition of Alberta crude oil and condensate Crude oils are essentially a mix of thousands of organic compounds (hydrocarbons) and a few inorganic compounds (sulfur, oxygen, nitrogen and a variety of metals) (Geraci and Aubin 1990). The major groups of hydrocarbons found in crude oil are listed in Table 7. However, depending on the origin of the oil, it can vary widely in its appearance and chemical composition (Chevron Phillips 2010). Table 7. The broad categories of hydrocarbons found in crude oil and their relative concentrations Hydrocarbon Description Composition by weight (%) * Alkanes Aromatics Also known as: paraffins or saturates The basis of petroleum fuels. Aliphatic compounds. Range in size from C1 (methane) to C78 Include the polycyclic aromatic hydrocarbon (PAHs). There are over 100 different PAHs Resins High molecular weight hydrocarbons containing S, O and N. 4.2 Asphaltenes High molecular weight complex compounds. Chemically altered fragments of kerogen (organic compounds found in the organic matter of sedimentary rocks). Soluble kerogen is called bitumen. 1.7 * Data for unweathered Athabasca Sweet Mixed Blend #5 (Environment Canada 2008a) Foster (2010) broadly classified crude oils into three groups as follows: A. Paraffin based crude oils: contain higher molecular weight paraffins which are solid at room temperature, but contain little or no asphaltic (bituminous) matter. They can produce high grade lubricating oils. 15

16 B. Asphaltic based crude oils: contain large proportions of asphaltic (bituminous) matter and little or no paraffin. Many are predominantly naphthenes. This group is further divided into two classes: i. Those that contain over 1% sulfur (high sulfur crude oils) by weight and consist of aromatics and asphaltenes; mostly found in North America (Alberta, Saskatchewan, California and Mexico), South America (Venezuela, Colombia and Ecuador) and the Middle East (Kuwait and Saudi Arabia) ii. Those that contain less than 1% sulfur (low sulfur crude oils) and consist of aromatics, naphthenes and resins; mostly found in Western Africa (Chad), Central Africa (Angola) and East Africa: (Madagascar). C. Mixed based crude oils: are between the two types above, containing paraffins and naphthenes, as well as aromatic hydrocarbons. Most crude oils fit into this category. Alberta s Athabasca oil sands are a group B (class i) crude oil. They are a mix of sand, silt, clay and water with 10 12% crude bitumen and % sulfur by weight, and have a viscosity of 5,000 cp (5 Pa s) 1 at 25 0 C (about that of molasses), increasing to as high as 300,000 cp for the most viscous deposits (Dusseault 2001, Environment Canada 2008a). Bitumen is viscous (thick) and heavy, due to its high proportion of aromatics and naphthenes (i.e. high molecular weight compounds with more than 60 carbon atoms) versus paraffins. It is also heavy due to its high content of nitrogen, sulfur, oxygen (NSO s) and heavy metals. Thus, crude bitumen is often blended with a light hydrocarbon liquid (condensate) before it is shipped by pipeline to markets where it undergoes further processing (McColl et al. 2008). Condensate is a mixture of benzene, toluene and xylenes (Environment Canada 2008b). 1 A measure of dynamic viscosity where 1 centipoise (cp) = 10 3 Pascal seconds (Pa s) 16

17 The physical characteristics of this medium crude oil determine its pathway and fate after a spill in the marine environment, including its potential for and likely extent of seafood contamination (Yender et al. 2002). It usually floats on the surface, although it can mix with sand when stranding on beaches or in the surf zone occurs, and is deposited in the nearshore. Up to one third of this oil will evaporate in the first 24 hours but the remainder will form persistent residues. It has moderate to high viscosity, can be dispersed by natural processes only very early in the spill, but it readily emulsifies. There is a moderate to high risk of seafood contamination from this type of oil because of its relatively high content of low molecular weight, water soluble aromatic hydrocarbons, which are semi volatile and therefore evaporate slowly; the coating of fishing gear and of intertidal species can be significant. Data on the chemical composition of crude bitumen from the Athabasca tar sands is sparse. A recent study by Yang et al. (2011) collected six Alberta oil samples from various sources three raw Alberta oil sands (AOS), a diluted crude oil sands bitumen (DOB, diluted with diluents), Albian Heavy Synthetic (AHS), and Alberta Sweet Mixed Blend (ASMB) and measured their chemical fingerprints. Quantitation results of five target petroleum characteristic alkylated polycyclic aromatic hydrocarbon (PAH) homologues and other US EPA priority PAHs are summarized in the following Table. 17

18 Table 8. PAH concentrations in six Alberta oil samples (extracted from Yang et al. 2011) Environment Canada ( 2008a) also reported the PAH distribution in Alberta Sweet crude oil, and is listed in Table 9 below. The total amount of spillage of each chemical in the estimated spillage scenario is also presented. 18

19 Table 9. PAH distribution of Alberta Sweet crude oil including concentrations^ for each component PAH and the predicted quantity of each chemical released by spills i) annually* and ii) once every 11 years** (see Table 3 above for calculations). Chemical Name Concen. (ug/g oil) Spillage (KT spilt every 11 yrs) Spillage (KT/year) Chemical Name Concen. (ug/g oil) Spillage (KT spilt every 11 yrs) Spillage (KT/year) Acenaphthene E E 05 Phenanthrene E E 04 Acenaphthylene E E 06 C1 Phenanthrenes E E 04 Anthracene E E 06 C2 Phenanthrenes E E 04 Benz[a]anthracene E E 06 C3 Phenanthrenes E E 04 Benzo[a]pyrene E E 07 C4 Phenanthrenes E E 04 Benzo[b]fluoranthene E E 06 Chrysene E E 05 Benzo[k]fluoranthene E E 07 C1 Chrysene E E 05 Benzo[e]pyrene E E 06 C2 Chrysene E E 05 Benzo[ghi]perylene E E 06 Dibenz[a,h]anthracene E E 06 Dibenzothiophene E E 04 Fluoranthene E E 06 C1 Dibenzothiophenes E E 04 Fluorene E E 05 C2 Dibenzothiophenes E E 04 C1 Fluorene E E 04 C3 Dibenzothiophenes E E 04 C2 Fluorene E E 04 Naphthalenes E E 04 Indeno[1,2,3 cd]pyrene E E 07 C1 Naphthalenes E E 04 Biphenyl E E 05 C2 Naphthalenes E E 03 Perylene E E 06 C3 Naphthalenes E E 03 Pyrene E E 05 ^ Concentration data (ug/l) for Alberta Sweet Mixed Blend #5from Environment Canada online database (2008a) * based on the mean 1 KT crude oil spilt per year (i.e. 11 KT / 11 years = 1 KT/year) ** based on one average sized oil spill event of 11 KT 19

20 4. Toxicity of Crude Oil Petroleum hydrocarbons in crude oil are categorized into four fractions: saturates (aliphatics), aromatics, resins and asphaltenes (Singh, 2006). These fractions and their derivatives are considered pollutants when released into the environment. They are a serious concern worldwide because of the hazards they pose to the health of humans and the biota (Burgherr 2007). The aliphatics, and some aromatic fractions, are the most biodegradable components while the resins and asphaltenes are believed to be resistant to biodegradation (ATSDR 1999). Amongst the aromatics, the PAHs attract considerable concern because of their known toxic and bioaccumulative effects in animals and foodwebs, as well as their health risks to humans, including heightened risks of cancer and DNA damage (Pérez et al. 2008). The specific health effects experienced will depend mainly on the extent of the exposure (both in terms of duration of exposure and intensity of exposure), the innate toxicity of the PAH and the exposure pathway (i.e. inhalation, ingestion or skin contact). Other factors include pre existing health status and age (Scientific Services 2009). There are 17 PAHs (including some variations of certain ones as illustrated in Table 9) that have been identified as being of great concern because of their associated human health effects. These include: acenaphthene, acenaphthylene, anthracene, benz[a]anthracene, Benzo[a]pyrene, benzo[e]pyrene, benzo[b]fluoranthene, benzo[g,h,i]perylene, benzo[j]fluoranthene, benzo[k]fluoranthene, chrysene, dibenz[a,h]anthracene, fluoranthene, fluorine, indeno[1,2,3 c,d]pyrene, phenanthrene and pyrene (Scientific Services 2009). The toxicity of the PAHs of interest (i.e. those from Table 9) is presented in Table 10 and Table 11, including the principal supporting studies used by the U.S. EPA to determine the nature and extent of 20

21 the toxicity. The tables are separated by those PAHs that may have carcinogenic effects (including any U.S. EPA published slope factors) and those that have no known carcinogenicity (along with their U.S. EPA published oral reference doses). 21

22 5. Toxicity of Condensate As indicated above, condensate is a mixture of benzene, toluene and xylenes (Environment Canada 2008b). The toxicity of these chemicals of potential concern is presented in Table 10 and Table 11, along with the principal supporting studies used by the U.S. EPA to determine the nature and extent of their toxicity and any listed reference doses and slope factors. Because of the volatile properties of the condensate, the main health concerns are the acute effects of inhalation of these chemicals from air pollution. As they have short half life in the environment, the environmental health consequence of a spill of condensate will be much less than that of a spill of crude oil in a similar scale. Table 10. Toxicity of threshold PAHs (non carcinogens) in crude oil and condensate including health effects, U.S. EPA published reference doses (RfD) and other related information Substance Effects Study type RfD / RfC 3 Confidence 4 Supporting Studies Acenaphthene 1, 2 No treatment related effects on survival Study Low Mice oral sub RfD = Liver weight changes CASRN chronic study 6 X 10 2 Database Low Increased cholesterol levels RfD Low (USEPA 1989a) No treatment related effects on lifespan (USEPA 1989a) No treatment related effects on gross/histological appearance of tissues (Schmahl 1955) Anthracene 1 CASRN Naphthalene CASRN Fluoranthene 1 CASRN Fluorene 1 CASRN ,1 Biphenyl 1 CASRN Mice oral sub Study Low chronic study RfD = 3 X 10 1 Database Low RfD Low No tumors observed (USEPA 1989a) (Schmahl 1955) No tumors observed Mice lung implant (Stanton et al. 1972) Some tumors observed Mice dermal (Scribner 1973) Diarrhea, lethargy, hunched posture, rough coats, decreases in terminal body weight, Rats gavage (BCL 1980) incidences of lesions in some organs General toxicology and immunotoxicology. Sig. decrease absolute weights of brain, liver, and spleen and relative weight of spleen in high dose females. Nasal effects from inhalation: Hyperplasia in respiratory epithelium Metaplasia in olfactory epithelium Nephropathy, increased liver weights, hematological alterations and clinical effects. Decreased RBC, packed cell volume and hemoglobin Kidney damage, reduced hemoglobin levels, decreased food intake, and decreased longevity RfD = 2 Study High 2 X 10 Database Low RfD Low Albino mice gavage (Shopp et al. 1984) Mice chronic inhalation study Mice subchronic gavage Mice subchronic gavage Albino rat chronic oral study Study Med RfC = 3x10 3 Database L M RfC L M Study Med RfD = 4 X 10 2 Database Low RfD Low Study Med RfD = 4 X 10 2 Database Low RfD Low Study High RfD = 5 X 10 2 Database Low RfD Med (NTP 1992) (USEPA 1988) (USEPA 1989b) (Ambrose et al. 1960) Benzene Decreased lymphocyte count Oral RfD based on RfD = Study Med (Rothman et al. 1996) 22

23 CASRN Ethylbenzene CASRN Toluene CASRN Liver and kidney toxicity Developmental toxicity Increased kidney weight Neurological effects Decreased body weight, increased mortality Xylenes CASRN Impaired motor coordination (decreased rotarod performance) 1 No data for inhalation reference concentration (RfC) extrapolation workers exposed to benzene inhalation Rat subchronic to chronic oral bioassay Rat and rabbit developmental inhalation studies Rats gavage occupationallyexposed workers 4 X 10 3 Database Med RfD Med RfC = 3x10 2 RfD = 1 X 10 1 Study Low Database Low RfC = 1 RfD Low Study Med RfD = 8 X 10 2 Database Med RfD Med RfC = 5 Study High Database High RfC High (NTP 1986b) (Wolf et al. 1956) (Andrew et al. 1981) (Hardin et al. 1981) (NTS 1990) (Cavalleri et al. 2000) (Eller et al. 1999) (Nakatsuka et al. 1992) (Neubert et al. 2001) (Schaper et al. 2003) (Zavalic et al. 1998) Chronic rat study RfD = 0.2 Study Med (NTP 1986a) oral gavage Database Med Subchronic rat RfC = 0.1 RfD Med (Korsak et al. 1994) inhalation study 2 No data for carcinogenicity 3 Based on the assumption that thresholds exist for certain toxic effects such as cellular necrosis. Oral RfD (mg/kg day) and Inhalation RfC (mg/m 3 ) are estimates (with uncertainty included) of the daily exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. 4 Confidence in principal study = quality of design; did it identify LOAEL & NOAEL for several sensitive endpoints using adequate number of animals? Confidence in the database = presence of toxicological data from other species showing effects of exposure on development, reproduction, etc. Confidence in RfD reflects confidence in principal study and in the database. 23

24 Table 11. Toxicity of potentially carcinogenic PAHs in crude oil and condensate including expected health effects, U.S. EPA published slope factors and other related information Substance Acenaphthylene CASRN Benz[a]anthracene CASRN Benzo[a]pyrene CASRN Benzo[b]fluoranthene CASRN Benzo[b]fluoranthene CASRN Benzo[g,h,i]perylene CASRN Benzo[k]fluoranthene CASRN Naphthalene CASRN Effects Concentration 3 Principal and Supporting Study type Oral Slope Factor Studies Not classifiable as to human carcinogenicity 1 (USEPA 2011f) No tumors lifetime study, but (Cook 1932) Mice dermal exposure 65% mort at 1 year Positive mutation assay Salmonella typhimurium (Kaden et al. 1979) Probable human carcinogen 1, 2 (USEPA 2011f) Mice gavage (Klein 1963) Mice dermal (IARC 1973) Exposure caused increase in (Steiner and Edgecomb 1952) tumors in many studies Mice subcutaneous injections (Steiner and Falk 1951) Mice intraperitoneal injections (Wislocki et al. 1986) Probable human carcinogen 1, 2 (USEPA 2011f) Stomach tumors (Neal and Rigdon 1967) Rat dietary and gavage, (papillomas & carcinomas). (Brune et al. 1981) Increased respiratory tract & upper digestive tract tumors Proportion of animals with tumors increased steadily with increased dose Increased incidence of tumors at injection sites 5 x 10 3 ug/l SF = 7.3 per (mg/kg)/d ] Syrian golden hamsters Inhalation and intratracheal instillation Mice, rats, guinea pigs, rabbits, hamsters and some primates intraperitoneal BAP injections (Thyssen et al. 1981) (USEPA 1991) (Knauf and Rice 1992) (USEPA 1991) (IARC 1983) Probable human carcinogen 1, 2 (USEPA 2011f) Increased incidence epidermoid Osborne Mendel rats lung carcinomas and pleomorphic implants sarcomas in lung and thorax (Deutsch Wenzel et al. 1983) Increased incidence of liver adenomas and hepatomas & Mice intraperitoneal injections (LaVoie et al. 1987) lung adenomas Injection site sarcomas XVIInc/Z mice injections (Lacassagne et al. 1963) Complete carcinogenic activity (Wynder and Hoffmann 1959) Mice skin painting assays and initiating activity (Amin et al. 1982) Not classifiable as to human carcinogenicity 1 (USEPA 2011f) Increased incidence epidermoid carcinomas in lung & thorax (not sig.) Benign and malignant tumors at high doses, but located at other sites Papilloma observed at highdoses after tenth month Osborne Mendel rats lung implants Mice skin painting assays Mice injected subcutaneously (Deutsch Wenzel et al. 1983) (Muller 1968) (Wynder and Hoffmann 1959) Mice dermal application (Hoffmann and Wynder 1966) Probable human carcinogen 1, 2 (USEPA 2011f) Increased incidence epidermoid Osborne Mendel rats lung (Deutsch Wenzel et al. 1983) carcinomas in lungs and thorax implants Increased incidence hepatic adenomas and hepatomas (not Mice intraperitoneal injections (LaVoie et al. 1987) sig.) Increase in lung adenomas Dose related increase in % of with tumors & number of tumors per animal Probable human carcinogen (via inhalation) Carcinogenicity via oral is uncertain Decreases in survival. Increases in non neoplastic lesions in lungs and nose. Increase in alveolar/bronchiolar adenomas. Mice skin painting assay (Van Duuren et al. 1966) Mice whole body exposure to naphthalene vapors (USEPA 2011f) (NTP 1992) Tumorigenic response in lungs (Adkins et al. 1986) 24

25 Table 11 cont. Substance Phenanthrene CASRN Chrysene CASRN Dibenz[a,h]anthracene CASRN Fluoranthene CASRN Fluorene CASRN Indeno[1,2,3 cd]pyrene CASRN ,1 Biphenyl CASRN Benzene CASRN Ethylbenzene CASRN Toluene CASRN Effects Concentration 3 Principal and Supporting Study type Oral Slope Factor Studies Not classifiable as to human carcinogenicity 1 (USEPA 2011f) No mammary tumors Sprague Dawley rats oral dose (Huggins and Yang 1962) No tumors Mice skin painting assays (Kennaway 1924) Skin papilloma Mice dermal application (Scribner 1973) Probable human carcinogen 1, 2 Increased incidence of liver and (Wislocki et al. 1986) lung adenomas or carcinomas Mice intraperitoneal injections Increased incidence of hepatic (Chang et al. 1983) tumors Complete carcinogen Mice skin painting assays (Wynder and Hoffmann 1959) Probable human carcinogen 1, 2 (USEPA 2011f) Weight loss, emaciation and dehydration. Pulmonary (Snell and Stewart 1962) adenomas and carcinomas. Mice oral route (Snell and Stewart 1963) Mammary carcinoma. Hemangioendothelioma Mammary carcinomas Mice gavage (Biancifiori and Caschera 1962) Complete carcinogen Mice skin painting assays (Wynder and Hoffmann 1959) Fibrosarcoma development Mice subcutaneous injections (Lubet et al. 1983) Not classifiable as to human carcinogenicity 1 (USEPA 2011f) (Hoffmann and Wynder 1966) No tumors observed Mice skin painting assays (Wynder and Hoffmann 1959) (Suntzeff et al. 1957) Not classifiable as to human carcinogenicity 1 (USEPA 2011f) Tumors observed (not sig.) Buffalo rats diet (Morris et al. 1960) Probable human carcinogen 1 (USEPA 2011f) Increased incidence epidermoid Osborne Mendel rats lung carcinomas in lungs and thorax implants (Deutsch Wenzel et al. 1983) Increased incidences of Swiss albino mice topical papillomas and carcinomas applications (Hoffmann and Wynder 1966) 100% tumor incidence Mice skin painting assays (Rice et al. 1986) Not classifiable as to human carcinogenicity (USEPA 2011f) No increase in the incidence of Mice gavage (BRL 1968) tumors Several malignant and benign Albino rat chronic oral study (Ambrose et al. 1960) tumors (not sig.) Classified as a "known" human carcinogen (USEPA 2011f) Leukemia, preleukemia, other malignancies 1 to 10 µg/l SF =1.5 x 10 2 to 5.5 x 10 2 per (mg/kg)/d ] Workers exposed in industry (Aksoy et al. 1972) (Infante et al. 1977) (Rinsky et al. 1981) (Rinsky et al. 1987) Not classifiable as to human carcinogenicity (USEPA 2011f) Pos with E. coli + hydroperoxide Bacterial mutation assays (Dean et al. 1985) Inadequate data to assess human carcinogenicity (USEPA 2011f) No effects Rats and mice inhalation carcinogenicity study (Huff 2003) (NTP 1990) (CIIT 1980) (Gibson and Hardisty 1983) Suggestive of potential carcinogenic effects Sprague Dawley rats oral route (Maltoni et al. 1997) Inadequate data to assess human carcinogenicity (USEPA 2011f) Xylenes CASRN No sig. incidences of tumors Rat gavage (NTP 1986a) (Maltoni et al. 1983) (Maltoni et al. 1985) 1 No data for inhalation reference concentration (RfC) 2 Cancer has been shown to be induced in humans by various PAH mixtures (i.e in cigarette smoke, roofing tar and coke oven emissions, etc). It is not possible, however, to conclude from this information that this PAH is the responsible agent. 3 Predicted concentrations for lifetime exposure at which 1 in 10 6 individuals will develop one or more cancerous tumors 25

26 6. Toxicity of Dispersants The use of oil dispersants is a potential option in response to off shore oil spills when sensitive shoreline habitats are present, and the objective is to mitigate the quantity of oil reaching those shorelines. Dispersants essentially disperse the film of surface oil into small droplets which then allow the oil to better intermix with seawater. If applied at depth, oil that might otherwise rise to the surface of the water is broken up by the dispersant into small droplets that can then remain suspended in the water. Dispersants themselves also have toxicological effects. Thus, the use of dispersants in oil spills can involve tradeoffs between impacts to shorelines and impacts to pelagic and deep sea environments. There are 14 dispersants listed on the U.S. Environmental Protection Agency s (EPA) National Contingency Plan (NCP) Product Schedule (USEPA 2011a). The U.S. EPA recently conducted toxicity testing on eight of those dispersants (Table 12). The eight were selected based on three criteria: 1) Lower toxicity of the dispersant or of the dispersant when mixed with oil 2) Availability of sufficient quantities (specifically during the 2010 Gulf of Mexico spill) 3) Immediate availability of samples for testing Table 12. The eight crude oil dispersants selected from the National Contingency Plan Product Schedule and tested for toxicity by the U.S. EPA, including some chemical data Dispersant Manufacturer Chemical composition (the composition of dispersants is largely proprietary) Corexit 9500 JD 2000 DISPERSIT SPC 1000 Sea Brat #4 Nokomis 3 AA Nokomis 3 F4 ZI 400 SAF RON GOLD * (Nalco Company 2010b, a) Nalco Inc., Sugarland TX GlobeMark Resources Ltd., Atlanta, GA U.S. Polychemical, Chestnut Ridge, NY Alabaster Corp., Pasadena, TX Mar Len Supply, Inc., Hayward, CA Mar Len Supply, Inc., Hayward, CA Z.I. Chemicals, Los Angeles, CA Sustainable Environ. Technol., Mesa, AZ Nalco Inc. recently released the main ingredients of Corexit 9500 * CAS Main chemical composition of Corexit Sorbitan, mono (9Z) 9 octadecenoate Sorbitan, mono (9Z) 9 octadecenoate, poly(oxy 1,2 ethanediyl) derivs Sorbitan, tri (9Z) 9 octadecenoate, poly(oxy 1,2 ethanediyl) derivs Butanedioic acid, 2 sulfo, 1,4 bis(2 ethylhexyl) ester, sodium salt (1:1) Propanol, 1 (2 butoxy 1 methylethoxy) 26

27 The U.S. EPA conducted acute toxicity tests with multiple concentrations of Louisiana Sweet Crude Oil alone, and with mixtures of Louisiana Sweet Crude Oil (LSC) and each of the eight dispersants. It tested the effects of each treatment on: i) endocrine activity (i.e. some of the dispersants contain nonylphenol ethoxylates which break down in the environment to nonylphenol a potential endocrine disruptor), ii) iii) toxicity to living cells (i.e. cytotoxicity), and acute toxicity to mysid shrimp (Americamysis bahia) and small fish (Menidia beryllina) None of the eight dispersants displayed biologically significant endocrine activity via the androgen or estrogen signaling pathways. All displayed cytotoxicity in at least one cell type; with JD 2000 and SAF RON GOLD being less cytotoxic than the others, and DISPERSIT SPC 1000 more cytotoxic. The JD 2000 was shown to be the least toxic in the whole animal assay. In general, the results indicated that the eight dispersants were similar in toxicity: classified as ranging from highly toxic to moderately toxic. The ZI 400/LSC mixture was the exception, having only slightly toxic effects on one of the test species (Menidia beryllina, a small fish). The dispersant, Corexit 9500A, which was applied to offshore surface waters and in the deep ocean during the Gulf of Mexico spill, fell into the moderately toxic category for both species tested, namely mysid shrimp (Americamysis bahia) and Menidia beryllina (USEPA 2010). The tests also demonstrated that, in most cases, adding the dispersant rendered the mixture less toxic than the crude oil alone (Hemmer et al. 2010). To date, available information indicates that the dispersants being used to combat oil spills do not appear to accumulate in seafood and therefore, there is likely little public health concern from them due to seafood consumption. However, new evidence has shown that inhalation of the dispersants can cause neurotoxicity (Sriram et al. 2011). Effects of air pollution will be discussed in Section

28 7. Toxicity from Controlled Burning of Crude Oil Deliberate burning of crude oil on the sea surface, a practice termed, in situ burning, is one approach used to reduce the spread of spilt oil. The burning oil in this manner (i.e. the incomplete combustion of organic matter in the presence of chlorine) has the potential to form polychlorinated dibenzo p dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). The release of these chemicals into the environment can, in turn, result in human exposure. There is very little information available on the generation of dioxins from in situ open burning of oil in water bodies. One study found that PCDD s and PCDF s in particulates downwind of in situ burning were no higher than background levels; thus concluding that dioxins were not being produced by the burning of crude oil or diesel oil (Fingas et al ). 28

29 8. Potential human health effects 8.1 Air pollution Air pollution resulting from oil spills into the environment could be of concern to the public, as exposure to the constituent chemicals has been shown to cause a variety of health hazards including neurological symptoms and cancer. There are two main sources for such pollutants into the air: those that escape directly from the oil spill (namely volatile organic compounds present in relatively high concentrations) and those that are produced when at sea crude oil spills are burnt (i.e. in situ burning). The following section examines the primary chemicals involved and their respective potential health effects Toxicity of Volatile Organic Compounds (VOCs) Benzene, toluene, ethylbenzene and xylene (BTEX) are considered the major VOCs present in crude oil. Thus, these chemicals also tend to be the major compounds tested for in air samples following oil spills. For example, the U.S. EPA has been testing air samples for BTEX since the 2010 Gulf of Mexico BP oil spill. These pollutants are selected for monitoring because they are present in relatively large quantities in oil and because, at elevated concentrations, they may cause human health problems. Other volatile chemicals found in crude oil include certain polycyclic aromatic hydrocarbons (PAHs). Thus, certain PAHs also tend to be monitored following oil spills due to their toxicity and relative quantities present in weathered oil. For example, the U.S.EPA is focusing on benzo(a)pyrene, benzo(a)anthracene, benzo(b)fluoranthene, benzo(k) fluoranthene, chrysene, dibenz(a,h)anthracene, indeno(1,2,3 cd)pyrene, and naphthalene (USEPA 2011). To date the levels of PAHs and VOCs detected in air samples in the Gulf of Mexico, following the BP oils spill and averaged over a period of approximately 5 months, have not exceeded the screening level thresholds set by the U.S.EPA, except for one incident with the VOC benzene. 29