5.

0 Tracer Gas Studies and Ventilation Rate Measurements

Clayton Group Services (2004) performed air movement studies by releasing smoke into the trailer. Very little
air movement was observed. The smoke dispersed slowly in all directions from the center of the room. Sulfur
hexafluoride tracer gas was used to identify leaks from the trailer to the outside. Air exchange rates and
ventilation rates were determined by measuring decay characteristics of carbon dioxide released into the space.
The ventilation rate in the large room was 17.49 cubic feet per minute (cfm) with an air exchange rate of 0.659
air exchanges per hour, whereas the small room had a ventilation rate of 24.92 cfm with an air exchange rate of
1.67 air exchanges per hour. These results were used in the air modeling presented in Section
6.1. They reflect the conditions that existed at the time the measurements were made and, since the trailer is not
airtight, are likely to change depending on environmental conditions such as wind speed and direction.
6.0 Empirical Model for Indoor Air Mercury Emission
Several models were developed and evaluated to empirically describe indoor air mercury vapor concentrations
resulting from evaporation of an elemental mercury source. The initial evaluation was based on a simple box
model presented in Riley et al (2001), which provided an order of magnitude estimate of potential mercury vapor
exposure in a room resulting from cultural and religious practices.
The box model has the form:
C()t =
Q
S
``
`
1 e
Q t
V
``
`
(1)s>
where,
C(t) = concentration at time t C(t) = 0 at t=0
t = time (hours)
S = rate of evaporation (micro gram per hour)
Q = air flow rate from the room (cubic meters per hour)
V = room volume (cubic meters)
The box model predicts an exponential rise in mercury vapor concentration to a final equilibrium concentration
of S/Q. The rate of exponential increase is governed by the V/Q time constant which is the number of hours per
air exchange; Riley, et al. (2001) suggest a typical value of two hours for V/Q. The authors acknowledge that
their simple model only provides an order of magnitude estimate of potential exposure because the fate and
transport
15
of mercury vapor inside a house is complex and case-specific, and requires data for a variety of variables,
including adsorption and desorption characteristics.
Examination of the voluminous data obtained using Lumex and Tracker real-time mercury vapor analyzers
indicates that the simple box model does not adequately predict final equilibrium mercury concentrations.
Typically, mercury concentration rises to a maximum in the first few hours and then decreases (decays) with
time until the final equilibrium concentration is reached. The decay mechanism appears to be exponential in
nature. Several potential decay models were evaluated.
The decay model best suited for modeling mercury emission data was:
C
d
()t = C()t -
OE

e
Dt
-``
s
`
1
S
E
Q
``
>
`
+
S
E
Q
oc

.
(2)
Dt Dt
= C()t -

e +(1e )-
S
E
Q
.
oc

OE
where,
Cd(t) = decay model concentration
C(t) = box model concentration
D = exponential decay factor
E = final equilibrium concentration
This model provides a smooth transition to the final equilibrium concentration and predicts concentrations that
are always less than or equal to the conservative box model concentration (upper limit). The decay component of
the model is consistent with the observed mercury emission (concentration) decrease with time, possibly due to
oxidation of elemental mercury.
Figure 27 presents Lumex monitoring data for a 45-hour time period. The data were fit to Equation 2 using the
Sigma Stat (v2.03) statistical analysis software package to perform weighted non-linear regression. The final
equation, with an r
2
= 0.998, is as follows:
C
d
()t = 7121 -(1e
0.732 (t + 0.345)
)
-
OE

e
0.117 (t + 0.345)
+(1e
0.117 (t + 0.345)
)
-
140
oc
. 7121
The final equilibrium concentration predicted by this equation was 140 ng/m
3
(0.14 Fg/m
3
); this value is
reasonable based on the data in Figure 27. The t + 0.345 term (t + t0) accounts for time offset between time zero
and the start of monitoring measurements.
Table 3 presents decay model (Equation 2) non-linear regression results for several sets of mercury concentration
vs. time data (r
2
range = 0.910 to 0.998). Lumex and Tracker monitoring data, box model results and decay model
calculation results are presented in Figures 27-34. The room volume was fixed at 25.37 m
3
for all nonlinear
regression analyses.
16
The data in Table 3 show a wide range of air exchange rate (Q/V) values (0.099 to 1.54, average = 0.68) for
the mercury monitoring data sets evaluated. The data in Table 3 are generally in agreement with the range of
mean residential air exchanges per hour (0.53 to 1.1) noted in a National Research Council report on the risk
associated with radon in drinking water (NRC, 1999), and with those (0.25-1.57) reported in a study of
residential air exchange rates in the United States (Murray et al., 1995). Fit values for the E term indicate
that the decay model final equilibrium concentration is generally 2-4 percent of the box model equilibrium
value. The fit parameters for the August 19, 2002 Lumex monitoring data set (see Figures 31 and 32) may be
unreliable because the time offset parameter reached the defined upper limit (0.5 hours) within the first three
iterations of the regression. The August 5, 2002 Lumex monitoring data (Figure 27) and August 7, 2002
Tracker monitoring data (Figure 28) are from the same 45-hour time frame. Regression results for Q, D, and E
terms are in good agreement for the two monitoring data sets. There are a number of individual Tracker or
Lumex readings in Figures 27-34 that are lower than the adjacent readings on the figures. These readings are
normal and occur during automatic monitoring instrument zero adjustments, and do not reflect actual
measured concentrations.
Overall, this decay model (Equation 2) is adequate for describing elemental mercury emissions provided all
environmental factors are stable (constant). The factors include temperature, ambient pressure, air exchange
rate, and electrostatic effects. In addition, the elemental mercury source must be undisturbed. It is highly
unlikely that all these conditions are met during ritualistic uses of mercury. This is evident from the observed
bumps in the mercury concentration vs. time data sets (Figures 27-34).
The empirical decay model cannot predict the final equilibrium concentration due to the lack of data for
elemental mercury oxidation as a function of time, temperature, etc. Mercury monitoring results indicate that
the final equilibrium concentration is typically less than 5 percent of the simple box model predicted
concentration. The final concentration appears to be reached after 50-60 hours of stable, undisturbed elemental
mercury vaporization.
Figure 35 presents mercury concentration vs. time data when the mercury container was shaken for the first 16
hours. The box model appears to accurately predict mercury concentration for the first nine hours (Figure 36)
before mercury emission rate decay begins. Figure 37 shows the final model with a rate decay time offset of
9.04 hours. The final model, with an r
2
= 0.957, is:
C(t)= Box Model = BM
` `
`
23.49
-(t + 0.137)`
`
`= 7.322 -
`
`
1 e
s
25.37 >
`
`

s>(0.926 -(t + 0.137))
= 7.322 -(1 e ) t C 9.04 hours
17




0.0378
0.0378
( ))
0.124 (t 9.041
C()t
=BM
e

1
+
7.322
7.322

(0.124 (t 9.041))
=BM (e (1 0.005163)+0.005163) t c9.04 hours
where, S/Q = S/23.49 = 7.322; therefore, S = 172 g/hour and the final equilibrium concentration is
0.038 g/m
3
.
6.1 Model for Predicting Average Indoor Air Mercury Concentration
Additional studies were carried out to develop a simple model to predict average mercury vapor
concentrations in indoor air based on average emission over various
Table 4 presents mercury emission rates based on weight loss from mercury beads of different
diameter. Figures 38 and 39 present Tracker mercury concentration (twohour average) vs. time data for
nominal 0.5 cm beads. Figure 40 presents the nonlinear regression analysis for the nominal 0.5 cm bead
average mercury emission rate in micro gram per hour per square centimeter (g/hr/cm
2
) vs. time data
(22-864 hours). Figure 41 includes emission rate data for nominal 0.5 cm beads and other bead sizes.
Total bead surface areas were based on the effective bead diameter, which was calculated assuming a
spherical bead with weight equal to the starting weight divided by the number of beads and density of
13.6 g/cm
3
. The beads tend to flatten and spread out on the surface upon which they rest, therefore, the
bead active emitting surface area is less than 100 percent. The fraction of bead surface area available
for
time
inte
r
v
a
l
s.

emission depends upon several factors including bead diameter, resting surface roughness, and surface
tension. The bead active surface area for emission was assumed to be 50 percent for this study. The
final model (Equation 3) can be used to predict average emission rate, S, for 22-864 hours exposure
time (r
2
= 0.943).
S' =avg g / hr / cm
2
=96.947 (e
(0.0188 hours )
+(0.0000033 hours)+0.0968) (3)
The nominal 0.5 cm data in the first two sections of Table 4 (first 11 data sets) were used to determine model
parameters in Equation 3; the data in the last set was not


shellfish (e.g., tribal and other subsistence fishers and their families who rely heavily on locally caught fish for
the majority of their diet).
Although people are exposed to methylmercury via the dietary route, there are also some non-dietary sources
of mercury exposure. Many consumers are not aware that mercury has been used for years in common
household products such as thermostats. Releases from the manufacture of mercury-containing products and
inappropriate disposal of these products have contributed to mercury entering the environment and ultimately
the food chain. Misuse of or accidental breakage of some products can create indoor air health risks and expose
consumers to dangerous levels of mercury. In addition, certain cultural or religious uses of mercury may also
result in harmful mercury exposure. The number of individuals exposed in the U.S. in this way is very small.
The Agency will make it a priority to provide consumers with reliable risk information about mercury
exposure so that they can make informed choices about the fish they eat and the products they use.
Progress to date. EPA has directed most of its mercury risk communication activities toward raising
awareness about dietary practices. The FDA-EPA national advisory,
What You Need to Know About Mercury in Fish and Shellfish, provides advice for women who might become
pregnant; women who are pregnant; nursing mothers; and young children.
2
This advisory represents the first
time FDA and EPA have combined their advice into a single uniform advisory. During the summer and fall of
2004, the two agencies distrib
IV. Communicating to the Public about Mercury Exposure Risks - 49
uted brochures about the advisory to approximately 200,000 medical providers in the U.S.
In September 2005, EPA sponsored the Eighth Annual National Forum on Contaminants in Fish
(Fish Forum). The forum provided an opportunity for people who have an interest in the subject
of advisories, from both the public and private sectors, to discuss scientific and policy issues, risks
and benefits, and communication strategies associated with exposure to chemical contaminants in
sport- and subsistence-caught fish and shellfish. In September 2005, the 13th straight year, EPA
released its National Listing of Fish Advisories, a summary of information on locally-issued fish
advisories and safe-eating guidelines.
3
This information is provided to EPA annually by states,
territories, and tribes.
States and tribes issue fish consumption advisories if elevated concentrations of chemicals such as
mercury are found in local fish. States monitor their waters by sampling fish tissue for persistent
pollutants that bioaccumulate. States issue their advisories and guidelines voluntarily and have
flexibility in what criteria they use and how the data are collected. As a result, there are significant
variations in the number of waters tested, the pollutants tested for, and the threshold for issuing
advisories. Based on self-reporting, the national trend is for states to monitor different waters each
year, generally without retesting waters monitored in previous years.
4
As new waters are tested and
results are added to previous years findings, the number of fish advisories continues to rise. EPA
makes information on the fish advisories, as well as Fish Forum proceedings, easily accessible to the
public on its website.
Although most of EPAs risk communication efforts have been directed to increasing awareness of
mercury in the food chain, the Agency has also investigated non-dietary sources of mercury
exposure about which the public should be aware. Risk communication has been conducted in
conjunction with mercury reduction activities, such as school clean-outs or thermometer collection
programs. In many cases, critical mercury outreach to schools and communities would not
otherwise occur without EPA assistance. For example, EPAs Region 6 has identified a particular
need for such support in communities on the U.S./Mexico border.
EPAs national efforts on mercury risk communication have been aimed at making information
widely available to the public and at co-sponsoring national conferences that bring together people
from across the country to share information on mercury risk communication. A unique exposure
concern is raised by ritualistic use of mercury in certain cultural communities. For this reason, in
January 1999,

The largest source of emissions to the atmosphere is currently
coal fi red power stations, however, the widespread use of
mercury in dental amalgam means that in the future, crematoria
could become the most signifi cant source. Hospitals with
medical waste incinerators are also a major contributor to the
mercury problem, and although emissions are decreasing as
the number of medical waste incinerators is reduced, there is
concern within the health community that the use of mercury
in healthcare products is exposing patients and other vulnerable
groups. The chlor-alkali industry, the biggest single user
of mercury in Europe, has released many tonnes of mercury
to the aquatic environment over the years, and contributed to
fi sh contamination52.
Approximately 70% of environmental mercury now comes
from human activities including a variety of industrial processes;
coal burning, incineration or disposal of mercury-containing
products, the use of mercury for chlorine production
in the chlor-alkali industry, production of zinc, steel and other
metals; cement production, mining and product recycling.
Mercury is used in a variety of industrial, consumer and medical
products.
It is also released into the environment through natural phenomena
(volcanoes, degradation of minerals or evaporation
from soils) and manmade processes.
PRODUCT EXAMPLES AT A GLANCEg:
fl uorescent light bulbs and batteries,
medical devices: thermometers, blood pressure
instruments (sphygmomanometers),
laboratory chemicals, preservatives in some vaccines
and pharmaceuticals, and in dental amalgams53.
various temperature and moisture measurement and
sensing devices (barometers, hydrometers, fl ame
sensors).
g A very detailed list of mercury use in products is available, see: The
European Commission, DG Enterprise. Risks to Health and the Environment
Related to the Use of Mercury Products. Prepared by Risk & Policy Analysts
Limited, Norfolk. J372/Merkury. August 2002. Annex 3.
Hospital waste can contribute to
mercury emissions
21
Stay Healthy, Stop Mercury campaign
Table 2. USE, EXPOSURE ROUTES AND TOXICITY OF MERCURY AND ITS COMPOUNDS54 55
Mercury
Form
Elemental (Hgo) Inorganic
(mercury salts)
(Hg1+)
Organic- methyl
mercury (CH3Hg-)
Organic
ethyl mercury
(C2H5Hg-)
Organic
phenyl mercury
(C6H5Hg-)
Main Use Dental fi llings (amalgam is
a mix of mercury and other
metals like Ag, Sn, Cu, In, Zn)
Medicines,
Cosmetics
(used as a
preservative)
No intentional uses,
when deposited
into water, mercury
is transformed into
methylmercury by
micro organisms
and bacteria.
Vaccines (the
preservative
thimerosal
is 49% ethyl
mercury
Fungicide,
bactericide
Other uses Goldmining
Chloralkali plants
Products (batteries, switches,
fl uorescent bulbs, measuring
and control devices eg.
thermostats)
Medical devices (thermometers,
gastrointestinal tubes,
sphygmomanometers)
Santoria and other ethnic and
religious rituals
Disinfectants and
anti-microbials
Electrical
equipment
Photography
Source of
exposure
Hospital spills eg. broken
thermometers
Dental amalgam
Home spills
Children playing with
quicksilver used in ethnic/
religious rituals
Fish consumption
(the fi sh have
ingested methyl
mercury and it is in
their muscle tissue)
Route of
exposure
and
absorption
rate
Inhalation: 80% absorbed
Ingestion: 0.01% absorption
Dermal: minimal absorption
Ingestion: ~10%
absorbed
Dermal: lethal
doses can be
absorbed
Inhalational: well
absorbed
Ingestion: 90-
100% absorbed
Injection: 100%
absorbed
Ingestion: 80-
100% absorbed
Dermal: see
salts
Toxicity -
primary
Lungs, Skin, Eyes, Gingiva Kidneys, Gastrointestinal
tract
Central nervous
system
under study Kidneys
Toxicity -
secondary
Central nervous system,
Kidneys
Central Nervous
system
Cardio-vascular
under study
Central nervous
system
Transport in
body
Crosses blood- brain barrier
Crosses the placenta
Found in breastmilk
Does not easily
enter the brain
or cross the
placenta
Crosses blood-brain
barrier
Crosses the
placenta
Found in breastmilk
Crosses bloodbrain
barrier
Crosses the
placenta
Found in
breastmilk
NOTE: Dose and Timing of Exposure are not refl ected in this table