The Alberta Oil Sands Community Exposure and Health Effects Assessment Program : report

42 Pages·1997·7.9 MB·English

by  Gabos, S. (Stephen)

y^i
m
m m -
wm^
2a A ^ '44|
“f
^ ^ m s s m
^ ,3, ,<{ ■ ■ 1 "^, T ? , ^ 3 \*^ F V
ilB^
For more information contact:
Health Surveillance
Alberta Health
P.O. Box 1360
10025 Jasper Avenue
Edmonton, Alberta
T5J 2P4
Phone: [phone]
Fax: [phone]
ISBN 0-7785-0029-2
TABLE OF CONTENTS
SStteupdhyan TGeabaoms,: M .DC.a MADIANA
INTRODUCTION 4
Vince Gagner, >998
Dennis Prince, M.Sc., REng.
BACKGROUND AND RATIONALE 5 Don Schopflocher, Ph.D.
Susan Shaw, MA
DEVELOPMENT OF THE MEASUREMENT INSTRUMENTS 6 Shaole Wu, Ph.D.
Background: Active vs Passive Sampling 6
Data Quality Objectives 7 Siu Chan, Ph.D.
Phase One: Developing and Testing the Measurement Instruments ... .8
Sulphur Dioxide (SO2) 8
Nitrogen Dioxide (NO2) 9
Field Team:
Exposure Chamber Studies 11
Alexander MacKenzie
Ozone (O3) 12
Volatile Organic Compounds (VOC s) 13
Yvonne WalshS
Particulates (PM-iq and PM2 5) 13
DEVELOPMENT OF ANCILLARY INDIVIDUAL MEASURES 17
Interview Teams:
The Time Activity Diary 17
Demographic and Exposure Questionnaire 17
Sharon Anderson
The Health And Nutrition Survey 17
Jill Henderson- Grainger
Theresa Espejo
FIELD STUDY 18
Goals 18
Sonia Gosse
Design 18
Sampling 19
Cheryl Chimco
Field Operations 19
Interview Teams 19 Theresa Bilou
Housing 19
Michelle Anne Snook
Data Entry 19
Lisa Rose
ANALYSIS OF EXPOSURE DATA 20 Stella Stewart
Analysis 20
Rose Boucher
Results 20
NO2 20
SO2 23
O3 25
Volatile Organic Compounds 27
Particulates 29
Conclusions and Recommendations 34
ANALYSIS OF ANCILLARY INDIVIDUAL MEASURES 34
REFERENCES 35
APPENDIX A 37
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
EXECUTIVE SUMMARY
The Alberta Oil Sands Community Specifically, the objectives of the pilot
Exposure and Health Effects Assessment study were to develop a d ata collection
Program was developed to ensure that a method for personal/population
long-term, systematic approach to data exposure assessment of exposure to
gathering was implemented that would sulphur dioxide (SO2), nitrous oxides
improve our knowledge about the link
(NOx), ozone (O3), volatile organic
between the environment and human compounds (VOCs), particulates, and
health. The purpose of the pilot study heavy metals (HM) using the Total
was to develop the methodology to be Exposure Assessment method. Including
used in the main investigation and to development of the protocols for field
address the technical, laboratory, and data collection, laboratory analysis, and
logistical aspects of the Program. This quality assurance; development of
included developing the appropriate methods for the collection of individual
exposure assessment techniques, field ancillary data on exposure conditions,
activities, analytical laboratory testing health status, and nutrition; conducting a
procedures, and data analysis feasibility study on a convenience sample
capabilities. of the population; and development of a
statistical analysis methodology for the
environmental data.
Phase One: Developing and Testing the Measurement Instruments
The Personal Exposure Monitoring Atmosphere Generating System (TAGS),
Devices (PEMs) were developed and owned by Health Canada and operated
tested in two phases. During the first by Bovar Environmental in Toronto; these
phase, the Initial design for the samplers tests would determine an accurate
was tested In several controlled
sampling rate for the badges. There were
environments to ensure that the data
three separate 24-hour exposures for
quality objectives were attainable. The each type of sampler - S O2 and NO2 - at
second phase focused on developing and
high (50 g/m^), medium (10 g/m^), and
testing the PEMs in the field, to evaluate low (5 g/m^) concentrations. Eighty
the applicability of the technology for samples for each contaminant type
testing exposure In a random population. were collected.
The SO2 and NO2 badges were Second, several samples were collected
developed for the study, because there from locations in Fort McMurray where
was no commercially available the values were expected to be high, to
instruments designed to measure compare to the levels observed from
relatively low levels of these blank samplers. The level of contaminant
contaminants over a short period of measured on the blank samplers was
time (24 hours). consistently lower than the level of
contaminant measured by the exposed
Initially, badges for both contaminant samplers, and the precision of the
type were exposed under controlled exposed samplers was deemed
laboratory conditions, using the Test acceptable. There was a consistent trend
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
of measurable contaminant in relation to A n umber of survey instruments were
concentration and length of exposure. developed to collect individual
demographic, health, and time
Standardized, commercially available activity information.
samplers were used for VOCs, ozone and
particulates, so preliminary testing of the
sampling devices was not required.
Phase Two: The Field Study
The primary goals of the Field Study Although the SO2 levels of exposed
included generating preliminary samples generally exceeds the levels
exposure information to support the measured on blanks, the
effective design of the Main Study, differences are slight. In fact, without the
establishing and testing protocols for presence of extreme samples in the
deploying, collecting, exposed samplers, we would not be able
packaging/shipping, and analyzing to conclude that measurable quantities
exposure monitoring samples, and field of SO2 exist In the test area. As with the
testing of ancillary individual
NO2 samplers, only a very small
questionnaire instruments. In general, proportion of samples show high levels
the pilot study sought to test the data of exposure. In addition, median levels
collection Instruments and methods on a for outdoor and ambient levels tend to
small sample. Consequently many exceed those for indoor and personal
important aspects of the data were not levels. Separate analysis of the results of
examined. Including analysis of the ambient monitoring does not provide
responses to the survey instruments. strong evidence that the sites differ in
SO2 concentration. The analysis does
I Each of the sampler types was deployed show, however, that the third day of the
I to collect information on ambient test had lower ambient levels than did
conditions, using the outdoor monitoring the first two days (Main Effect for Day,
sites located within the test area (Fort
F=3.8, df =2, 22, p<0.05). This suggests
McMurray, Fort MacKay, and Suncor Plant that day to day variability could be of
j Site), as well as to collect Information
considerable Importance in the
I about individual exposure conditions, monitoring of SO2.
I using outdoor, indoor, and personal
I monitoring of specific individuals within
The distribution of exposure to O3 was
I the test area. only mildly skewed. Only the Outdoor
sample appears to differ from any other
! The NO2 levels of all exposed samples type including blank samples. Statistical
t exceeded the levels measured on blanks,
analysis lends support to this last
indicating that measurable quantities of observation. Further analysis lends no
NO2 do exist in the test area. In addition, support to for any variation in levels
neither the ambient nor outdoor
across ambient sites or days, or for any
locations show a markedly skewed characteristic level of individual
distribution; the Indoor and personal exposure. There is slight support for a
measures account for a l arge majority of relationship between personal and
the most extreme measures. Separate indoor exposure levels across Individuals,
I analysis of the results of ambient but no support for a relationship
i monitoring suggests that the Fort
between outdoor levels and either
i McMurray town site had higher levels of
personal or indoor levels.
I NO2 concentration, and that the first day
I of testing showed generally lower levels The distribution of VOC levels Is
of NO2 concentration than did the two extremely skewed, indicating that a small
I succeeding days. Neither of these effects
number of participants were exposed to
I was significant in conventional tests of relatively high levels of particular VOCs.
significance, however. None of the groups showed that the
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
ambient or outdoor levels different from higher than indoor levels) for all
blank levels (except for a very small chemical groups.
number of extreme levels in outdoor
samples). Indoor and personal levels Very few particulate samples were
exceeded blank levels and did not differ obtained, so analysis was confined
(with the possible exception that to description.
personal levels of the hexane group were
INTRODUCTION
The Alberta Oil Sands Community The Alberta Oil Sands Community
Exposure and Health Effects Assessment Exposure and Health Effects Assessment
Program was established following public Program was developed to ensure that a
hearings conducted by the Energy long-term, systematic approach to data
Resources Conservation Board in relation gathering was implemented that would
improve our knowledge about the link
to Syncrude's Mildred Lake Development
Project (1994). Human health concerns between the environment and human
related to air quality were raised by health. The Program combines two
various participants including aboriginal broad concepts in an integrated
groups, environmental associations, and population-based environmental health
Alberta Health. The Energy Resources framework: (1) the direct measurement
Conservation Board views and of personal and population exposure to
recommendations of the human health environmental factors, and (2) the
issue were: epidemiologic surveillance of health
outcomes in the population. The current
document reports on activities relevant to
"The Board acknowledges the
CM>ncerns of many of the interveners the first objective; consideration of the
that atmospheric emissions from the second objective is deferred to the Main
oil sands plants are impacting on the Study Proposal.
health of the region’s population. The
Board believes that there is an The approach to measurement of
obligation on industry to address this personal and population exposure was
modeled after the United States
issue as effectively and rapidly as
Environmental Protection Agency
possible. The Board also
approach for particle total exposure
acknowledges Syncrude’s commitment
assessment method (TEAM). The purpose
to support and participate in a regional
health study that is broadly based and of the pilot study was to develop the
involves all stakeholders. The Board methodology to be used in the main
notes, however, that concerns about investigation and to address the
the health effects from atmospheric technical, laboratory, and logistical
emissions have, despite a number of aspects of the Program. This included
efforts, continued to exist in the oil developing the appropriate exposure
sands region as well as other areas of assessment techniques, field activities,
the province. The Board does not analytical laboratory testing procedures,
support a h ealth study carried out and data analysis capabilities.
simply for its own sake and expecrts
any health study undertaken in the
region to be meaningful, with terms of
reference sufficiently broad to
demonstrate both short and long
term effects.”
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
Specifically, the objectives of the pilot II. Develop methods for the collection
study were to: of individual ancillary data on
exposure conditions, health status,
and nutrition;
for personal/population exposure
assessment of exposure to sulphur III. Conduc:! a f easibility study on a
dioxide (S02), nitrous oxides convenienco sample of the
I. Dc(eNovO mexlp)oo,p u n odaz s od n ae t (a V ( O0cC3os)l),l, e cv topilaoarntt iil mem ie ltoahrtgoeasd,n ic population; and
and heavy metals (HM) using the IV. Develop a statisticol analysis
TEAM methodology including: methodology for the
environmental data.
• d evelopment of the field data
collection protocol,
• d evelopment of protocols for
laboratory analysis, and
• d evelopment of a q uality
assuranc:e protocol;
BACKGROUND AND RATIONALE
In general, exposure can be defined as interfere with the normal biological
any contact between a substance, functions, causing effects ranging from
biological agent or radiation and an subtle biochemical changes to clinical
individual or community. We are all disease and even death. The concept of
exposed to low levels of contamination a continuum from source of
in the air we breathe, in the food we eat, contamination to the final health effect
the water we drink, and the consumer is a b asic feature of all contemporary
products we use. Contaminants can risk models.
Determining the risk posed by The output of each component in the
environmental contaminants to chain of events serves as Input to the
populations requires knowledge about next. The lack of information on any one
the following fundamental components: component thus impairs our ability to
make accurate assessments of the
• s ource(s) of contaminants; associated population health risks. Our
• t ransport of agents in the knowledge about the source and
environment; transport of chemicals and other agents
In the ambient environment is increasing
• exposure of individuals and
as the result of environmental
communities to cdiemicols;
• d ose recoived by those monitoring programs, however, there is a
exposed (biologicol markers need to Integrate these data with
of exposure); information on population exposure,
biological markers and health effects.
• early biologicol effecrts
resulting from the dose This is very Important In achieving new
(biologicol markers of effecrt); health based protection levels.
and
• o vert health effecrts (cHinioal
disease, death).
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
In dealing with population health these issues, more than ever, there is a
outcomes which may be attributable to need to look beyond one-time
long-standing, exposures to low-levels of epidemiologic studies.
contaminants, we are confronted with
Environmental health surveillance is a
the difficult and complex problem of
chronic health effects. A n umber of tool which can be used to gather data
conditions such as cancers, disorders of and Information on the health of people
the cardiovascular system, neurological for the purpose of tracking and detecting
disease, chronic respiratory ailments, and trends and associations among a b road
many other diseases have Important range of environmental and health
environmental, behavioral, social, and related variables. The process consists of
genetic links. The causes of these an on-going, systematic collection,
conditions are multifactorial in nature. analysis, and Interpretation of selected
Other characteristics such as multistage data on health outcomes, environmental
development, long Induction time, and quality parameters, and population
the absence of information on individual exposure. In addition, data on
and population exposure make progress behavioral, lifestyle, social, economic,
In chronic disease prevention slow and and other confounding variables are
tenuous. In order to be able to address also considered.
DEVELOPMENT OF
MEASUREMENT INSTRUMENTS
The Personal Exposure Monitoring quality objectives were attainable. The
Devices (PEMs) were developed and second phase focused on developing and
tested In two phases. During the first testing the PEMs in the field, to evaluate
phase, the initial design for the samplers the applicability of the technology for
was tested in several controlled testing exposure In a random population.
environments to ensure that the data
Background; Active vs Passive Sampling
Five contaminant classes were designated explain the principles of passive sampling.
for the study, including volatile organic The following Is taken from Shields and
compounds, sulfur dioxide, nitrogen Weschler (1987):
dioxide, ozone, and inhalable
particulates. All of the compounds,
(Eq.l)
except for particulates, may be measured
m/(t Ca) = D ( A /I) where
through the use of a p assive sampling
"sampler". A p assive sampling sampler is m / ( t Ca) = badge sampling rate
a small plastic container that holds a p ad
(mL of air per minute)
saturated with a solution designed to
react with specific chemicals in the air, m = mass of substance that diffuses
such as ozone or sulphur dioxide. After onto the sampler sorbent (ug)
the sampler is exposed to the air for a t s= sample duration (minutes)
specific period of time, the pad Is Ca= time weighted average
removed and treated to determine the
concentration of substance (ug/mL)
amount of chemical that reacted with
D = difhision coefficient (cm^/minute)
the solution, and the amount of chemical A = cross-sectional area of die sorbent
contaminant originally In the air can then
be determined using a simple formula.
I = diffusion path length of the
spaadm p l(ecrm ^ )( cm)
One chemical characteristic of gases in air
is the diffusion coefficient. This
parameter reflects the rate that gases The sampling rate can be determined by
disperse In air at a g iven temperature exposing it to known concentrations of a
and pressure. Pick's First Law of Diffusion contaminant (Ca) for known time intervals
considers this "diffusion coefficient" and (t) and measuring the mass of
may be rearranged as an equation to contaminant (m) that it collects. Once
gB THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
the sampling rate is established, the Active sampling must still be used for
sampler can then be exposed to measuring inhalable particulates because
unknown concentrations of contaminant this contaminant is not a g as. That is,
for a standard duration and the mass of there is no diffusion coefficient to govern
contaminant can be measured in the lab, any sample rate into a sorbent material.
thereby solving for (Ca). The preferred sampling method (as used
The primary benefit of passive methods is by EPA, Research Triangle Institute) is to
the ergonomically friendly nature of a pull a k nown volume of air through a
small clip-on badge compared with a filter, determine the mass of particles
conventional battery-powered pumping collected over the sampling period and
systems typically used previously to express the measure as a mass of particle
identify exposure levels. The principles per volume of air sampled. Once again,
of passive sampling are simple; however, the fundamental principles are simple,
the challenge lies in the treatment, but care must be taken In the
handling, and analysis of the sorbent pad preparation of the collection medium (a
used to collect the target compound preweighed filter) and the handling of
from the air. the filter in the field.
Data Quality Objectives
Uncertainties In data obtained from a Precision
sampling program may be attributed to Precision is a measure of the mutual
the way that a sample was taken or the agreement among individual
way it was analyzed. It Is very important measurements. For example, when two
to explicitly recognize that there are samples are placed side-by-side for the
inherent limitations to ail types of same amount of time there should be
sampling and analysis methods. little difference in the measurement of air
Confidence in the results may be contaminant concentration. A w orkable
increased but are often dictated by precision data quality objective would be
factors such as cost, time, and availability +/- 25% between duplicate samples.
of trained personnel. Lawrence Keith Again, this margin is expected to increase
(1988) argues that all sampling programs during low level monitoring.
must first define "data quality
objectives" prior to deciding on the
Representativeness
methodology required, and he lists those
Representativeness refers to the degree to
objectives as: accuracy, precision,
which the measures obtained from the
representativeness, completeness,
samples describes the environment being
and comparability.
sampled. For example, an environmental
sampler that Is p laced in a volunteer's
Accuracy living room can not be Interpreted to be
Accuracy is a measure of how close the representative of the entire home.
resulting estimate is to the true value. In Furthermore, it must be recognized that it
this case, the sample methods chosen for is merely an assumption that personal
the Oilsands Program must be within a exposure monitors measure the same air
tolerable range of variance accepted by that the volunteer is actually Inhaling.
other methods that are formally Accordingly, the objective here is to
recognized as "true" measures. A properly describe and define all data to
practical accuracy objective for the avoid underestimating the Inherent
samplers is to remain within +/- 25% of limitations of the samplers.
the recognized reference sampling
method. It Is to be expected that this
Completeness
margin will increase significantly at very
low contaminant concentrations. This is Completeness measures the amount of
valid data obtained from the research
not a p rimary concern because the results
compared to the amount expected to be
from very low readings may be
obtained under absolutely correct
interpreted as being below a safe level
of exposure. conditions. Temperature, pressure, wind
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM
velocity, and humidity are constantly Comparability
changing in both indoor and outdoor Comparability concerns the confidence
environments. It is theoretically with which the results of the research
conceivable that these changes in can be correlated to data from the same
environmental conditions may alter the study or other studies of similar design.
sampling rates of both the passive, and The reason for including the objective of
to a l esser extent, the active samplers. comparability in the Oilsands Program is
Sources of bias are also introduced to establish consistency with previous
during transport, storage, and handling TEAM studies (where technology permits)
of the samplers. The objective is and to maintain a standard methodology
therefore to recognize uncontrollable and protocol for sampling and analysis so
sources of bias and account for as many that results obtained from the different
as possible through a sound sampling components of the Oilsands Program are
protocol and blank sample analysis. comparable to each other.
Phase One: Developing and Testing the Measurement Instruments
Sulphur Dioxide (SO2}
Like most gases, the measurement of The sampling rate of the badge was
determined to be 41.1 ml_/mlnute and
sulfur dioxide (SO2) in the air has
traditionally been performed through the sensitivity of the sampling method
active sampling. For example, NIOSH was 200 ppb over a 4 h our sample
Method 6004 is the typical method used duration. Assuming a constant sampling
in the workplace for measuring SO2 rate, this would translate Into sensitivities
exposure and it involves using a pump to of about 35 ppb over a 2 4 hour sampling
pull a k nown volume of air through a duration. Unfortunately, this poses some
filter that has been treated with sodium problems for the Oilsands Program
carbonate. Researchers have known for because the 1993 Air Quality Monitoring
years that sodium or potassium Report For Alberta (Myrick, 1995)
carbonate will collect SO2 from the air. indicates that typical concentrations of
After the sample Is exposed, the treated S02 in Fort McMurray and Fort McKay
filters are processed in the lab with are between 5 a nd 10 ppb (13 to
hydrogen peroxide and the lab can then
26 ug/m^).
determine the mass of SO2
collected per sampler. The Science Team agreed that the
Leaderer design was acceptable, but the
The filters used for the NIOSH method sorbent pad area needed to be larger to
6004 (filters treated with Na2C03) were facilitate the collection of more
also used in passive samplers to measure contaminant. In addition, the diffusion
high levels (such as 500 ppb) of SO2 in path between the diffusion membrane
workplace environments. The OGAWA and the treated sorbent pad needed to
passive sampler was originally Identified be decreased to Increase the badge
as the most appropriate sampler for sampling rate (See Equation I). The final
measuring SO2 in the Pilot Study. sampler badge design for the Oilsands
However, field tests carried out by Program Is a clear light-weight plastic
Alberta Environmental Protection and holder with a d iameter of 55 mm. The
the Clean Air Strategy for Alberta found cost of the badge is approximately 1/4
that the blank levels of the samplers the price of the originally planned
were very irregular and that the small OGAWA samplers and is similar In
size of the surface area of the sampler
appearance to the OVM-3500.
would render it virtually useless for 24-
hour ambient air monitoring. The badge is constructed from a
modified 55-Plus Milllpore Filter Holder
Leaderer, et al. (1994) used a passive with a removable TEFLON diffusion
sampler badge to measure typical urban barrier designed to protect the sorbent
air concentrations over 24 hour periods. pad from wind and rain. The sorbent
THE ALBERTA OIL SANDS COMMUNITY EXPOSURE AND HEALTH EFFECTS ASSESSMENT PROGRAM