Project Summary: Analysis of Exposure, Dose and Risk Associated with Tap Water Contaminated with Chloroform


PROJECT SUMMARY:
ANALYSIS OF EXPOSURE, DOSE AND RISK ASSOCIATED WITH TAP WATER
CONTAMINATED WITH CHLOROFORM
FOR:
Mahnomen Public Water System
Fond du Lac Band of Lake Superior Chippewa
BY:
U.S. EPA
Office of Research and Development
Center for Environmental Solutions and Emergency Response
Revision 8
Dated July 2020
EPA/600/S-20/254

-------
ANALYSIS OF EXPOSURE, DOSE AND RISK ASSOCIATED WITH TAP WATER
CONTAMINATED WITH CHLOROFORM
INTRODUCTION:
Many municipal water supply systems use chlorine for primary disinfection or residual protection
for drinking water. Because it is difficult to completely remove organics from the drinking water, organic
carbon reacts with residual chlorine to create disinfection by-products. As organic matter and chlorine react,
trihalomethanes (THMs) are formed, one of which being chloroform. This is of concern due to the high
volatility of chloroform and the carcinogenic risk that it potentially presents to individuals from inhaling
chloroform. The most significant exposure pathway is believed to be showering (Davis et al. 2016). Here,
we analyze water concentrations and estimate air concentrations. We also analyze potential exposures,
doses, and risk from the inhalation of chloroform during showering. We perform these analyses for a range
of chloroform tap water concentrations that were observed in the Belich Addition (Mahnomen) Public
Water System in November 2019. The Mahnomen Public Water System is part of the Fond du Lac Band
of Lake Superior Chippewa in Minnesota. This analysis was conducted at the request of EPA Region 5's
Water Enforcement and Compliance Assurance Branch.
APPROACH:
The goal of these analyses is to evaluate the range of observed chloroform tap water concentrations
in the Mahnomen Public Water System with respect to several criteria. First, we compare the range of
observed chloroform tap water concentrations to a Minnesota water criterion level for chloroform in
drinking water (Minnesota Department of Health, 2016) to provide a reference point of comparison. We
use California Environmental Protection Agency's acute reference exposure level (mg/m3) for chloroform
exposure to evaluate estimated shower enclosure air concentrations that were determined using the range
in observed chloroform water concentrations for the Mahnomen community. Also, we use the United States
Environmental Protection Agency's (EPA) previously published carcinogenic slope factor for inhalation
dose of chloroform to estimate a lifetime cancer risk from the inhalation of chloroform during showering
reported by Nazaroff and Alvarez-Cohen in 2001. Finally, we use EPA's unit risk factor (URF) for
exposure to chloroform, (ug/m3)"1, to estimate an individual's lifetime cancer risk (U.S. Environmental
Protection Agency 1976).
The analyses provided here are based on a range of chloroform concentrations observed in the tap
water of the Mahnomen Public Water System obtained on November 6, 2019. These chloroform
concentrations ranged from 100 (micrograms (10~6 grams) per liter) ug/L to 250 ug/L. The approach
described here is also described in the associated Microsoft Excel workbook
(Chloroform_exposure_calculations_REV6_7-14-20.xlsx), "OVERVIEW_ASSESSMENT," tab. All the
calculations outlined here are performed in the Excel workbook. The approach utilizes Henry's Law
Constant to estimate the concentration of chloroform in the shower enclosure air during a showering event
using two models: "SIMPLE MODEL" (Sanders, 2002) and "COMPLEX MODEL" (Davis, et al. 2016).
The parameter values used in each model were obtained from the Davis et al. (2016) paper because they
represent a typical shower as well as behavior of individuals taking showers. A typical shower occurs in
an enclosure of 2 m3, has a water flow rate of 9 L/min, and has a duration of 7.7 minutes. We assume
individuals take one shower per day. Each model is based on equilibrium conditions.
Sanders (2002) developed a simple screening model to determine the concentration of contaminant
in air. Assumptions in this model consist of water temperature, the total amount of water used during the
showering event (Vwater), and the volume of the shower stall (Vair). This model is derived from the
dimensionless expression of Henry's Law Constant and can be found in the Microsoft Excel spreadsheet,
tab "SIMPLE MODEL."
2 I

-------
Davis etal. (2016) designed a slightly more detailed (complex) model to estimate the concentration
of contaminant in the air. More assumptions were taken into account in this model, including air exchange
rate (/.„«). generation rate (G2), and transfer efficiency shower duration (7). This model is described in the
tab "COMPLEX MODEL."
For the estimation of an individual's dose, Chowdhury (2013) developed a dose model to estimate
the daily intake of the contaminant through inhalation. This model took into account the assumptions of
adsorption efficiency of the contaminant (Er) into a person's lungs, breathing rate (R), shower duration (t),
shower frequency (/¦)• exposure frequency (EF), exposure duration (ED), averaging time (AT), and body
weight (BW). This model can be found in the tab "DOSEMODEL."
EPA previously published a carcinogenic slope factor for chronic exposure to chloroform through
inhalation (http://www.dartmouth.edu/~cush.man/courses/engs37/RiskAssessment.pdf) Nazaroff and
Alvarez-Cohen (2001). EPA currently does not seem to provide an inhalation slope factor for chloroform
("https://www.epa.gov/iris'). This previously published slope factor, (8.2 x 10"2 mg/kg-day)"1, for chloroform
intake via inhalation was used to help assess an individual's risk. To be conservative, we used the Sanders
(2002) "SIMPLEMODEL," results for chloroform air concentrations to estimate individual doses and then
multiplied each dose by this slope factor to estimate risk. Additionally, male and female doses, due to their
different average body masses, were assessed for risk.
We also evaluated risk using EPA's Unit Risk Factor (URF) of 2.3 x 10~5 for the inhalation exposure
of chloroform ("https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfin7substance n.mbr=25 ) (U.S.
Environmental Protection Agency 1976). Here risk is for exposure to the chloroform during showering.
For these estimates of risk, using the observed range of chloroform tap water concentrations, we used the
chloroform air concentrations estimated from the "SIMPLE MODEL."
RESULTS:
In the SUMMARY table below (taken from the "OVERVIEW_ASSESSMENT" tab), we first
compare the Minnesota water quality value of 20 ug/L (ppb) with the chloroform concentrations observed
in the tap water of Mahnomen community. This comparison is a ratio of observed tap water concentration
divided by the Minnesota water quality criterion to show how many times the criterion is exceeded. The
Minnesota water quality criterion for chloroform does not directly apply to showering but instead is a
general criterion for chloroform in tap water. The observed chloroform concentrations exceed the
criterion by about 5 to 13 times. We compare the observed range of chloroform concentrations to the
Minnesota criterion as a reference point for comparison.
Using the "SIMPLE MODEL" and "COMPLEX MODEL" approaches described in the
associated Excel spreadsheet, chloroform air concentrations are calculated using the Sanders and Davis et.
al. models, respectively. Next, we compare the calculated air concentrations from each model with the
California EPA chronic air exposure level for chloroform as a ratio of calculated air concentration
(mg/m3) divided by the California criterion. The ratios show how many times the calculated shower air
concentration for the low and high observed chloroform water concentrations exceed the criterion
(approx. 10 to 50 times).
In the "DOSE MODEL" tab of the associated Microsoft Excel spreadsheet, we use a dose
equation to calculate a person's (male and female) absorbed dose from inhaling chloroform during
showering. For these dose calculations (Chowdhury, 2013) we use the "SIMPLE MODEL" calculated
air concentrations. After doses are calculated, risks are estimated using an earlier EPA carcinogenic slope
3

-------
factor, (mg/kg-day)1, for inhalation of chloroform reported by Nazaroff and Alvarez-Cohen (2001). This
slope factor ("https://www.epa.gov/iris') no longer seems to be available via U.S. EPA's Integrated Risk
Information System (IRIS), but is used here for comparison. The "DOSEMODEL" tab estimates a
person's dose (in units' mg/kg-day) from the inhalation of chloroform during showering and uses the
approach described by (Chowdhury 2013). Multiplying the slope factor by the dose results in a unitless
risk value. The added risk ranges from 2.8E"4 to 8.3E"4 or about 3 in 10,000 to 8 in 10,000 added risk.
The lifetime exposure risk is calculated by multiplying the URF by the estimated chronic, lifetime
air concentration to which an individual is exposed. This results in an estimated lifetime exposure risk to
chloroform during showering. The added risk ranges from 7.0E"5 to 1.8E4 or about 7 in 100,000 to 2 in
10,000 added risk. These results are provided in the SUMMARY table below.
SUMMARY TABLE: EXPOSURE, DOSE, AND RISK ASSESSMENT
(Results are a measure of how much greater chloroform water and air concentrations are compared to criteria. Also, an assessment of inhalation dose and risk
during showering considering various established criteria.)
WATER (Chloroform concentrations observed in the Belich
Road community tap water ranged from lOOug/L (ppb) to 250
ug/L)
Evaluation criteria values
Ratio of Obs/Criterion (Low cone)
Ratio of Obs/Criterion (High
cone)
Minnesota water quality guidance value for
Chloroform (ug/L) in drinking water.
(https://www.health.state.mn.us/communities/enviro
nment/risk/docs/guidance/gw/chlo roforminfo.pdf).
Ratio values (observed concentration divided by
criterion) exceeding 1 are deemed unacceptable.
20
5.0

12.5

| |
AIR
Evaluation criteria values
Simple model ratio (air
conc/criterion)
Complex model ratio (air
conc/criterion)
The chloroform concentrations observed in the
Mahnomen community tap water ranged from 100
ug/L (ppb) to 250 ug/L.

{Based on low
coric chloroform
observed)
{Based on high coric
chloroform observed)
{Based on low
coric
chloroform
observed)
{Based ori high coric
chloroform observed)
California EPA acute reference exposure level (mg/m3)
is deemed an acceptable acute air exposure
concentration for chloroform. Ratio values exceeding
1 are deemed unacceptable.
0.15
20
51
10
24
DOSE
Evaluation criteria values
Risk fro
¦	
m Dose (male)
......
Risk from
¦		
Dose (female)
Risks calculated using air cone based on
SIMPLE MODEL

(Low cone obs)
(High concobs)
(Low concobs)
(High concobs)
EPA IRIS work published an earlier carcinogenic slope
factor (mg/kg-day)1 for chloroform via inhalation,
which assumes chronic, long-term exposure. Note
that the SIMPLE_MODEL estimated chloroform air
concentrations were used for dose calculation. Risks
greater than 1E-6 are deemed unacceptable.
8.10E-02
2.8E-04
7.1E-04
3.3E-04
8.3E-04
RISK
Evaluation criteria values
.............
Risk from Air
........................
Exposure (URF*Air
Cone)
Risk from Air Exposure
(URF*Air Cone)
Risks calculated using air cone based on
SIMPLE MODEL

{Low obs cone)
{High obs coric)
EPA's unit risk factor (URF) for lifetime exposure (per
ug/m3). Estimate of lifetime risk from inhalation
exposure during showering. Risks greater than 1E-6
are deemed unacceptable.
https://cfpub.epa.gov/ncea/iris2/chemicalLanding.cf
m?substance_nmbr=25
2.30E-05
7.0E-05
1.8E-04
* http://www.dartmouth.edu/~cushman/courses/engs37/RiskAssessment.pdf

-------
DISCUSSION:
The observed chloroform concentrations in the Mahnomen Public Water System exceed
Minnesota's Department of Health criterion for chloroform in drinking water by a factor of about 5 to 13
times (Minnesota Department of Health, 2016). While the Minnesota water quality criterion for chloroform
does not apply to showering, this comparison provides a reference point and illustrates that the observed
chloroform concentrations in the tap water of the Mahnomen Public Water System are elevated.
Results show that the estimated shower enclosure air exposure concentrations, using both the
simple and complex models, exceed the California acute air concentration criterion for chloroform by a
factor of approximately 10 to 50 times. The complex model estimates were lower because the model
incorporates an air exchange rate, or loss term, to provide less conservative results than the simple model.
The Davis et. al., model was modified to represent equilibrium conditions to be consistent with the
"SIMPLE MODEL".
Comparing the estimated doses to an earlier EPA published carcinogenic risk factor (slope factor),
considering both males and females, showed that individual added lifetime risk associated with showering
in these potential air concentrations of chloroform exceeds the EPA's acceptable risk of 1 x 10~6 acceptable
(one in a million) risk. Estimated risks from an absorbed dose of chloroform acquired during showering
ranged from approx. 3E~4 to 8E~4 or about 3 in 10,000 to 8 in 10,000 added risk. Added risks greater than
IE 6 (one in a million) are deemed unacceptable.
Finally, we also calculated risk from inhalation exposure (using EPA's URF for chloroform) during
showering. For exposure risk we calculated chloroform air concentrations in the shower using the
"SIMPLE MODEL." These risks, determined for the low and high observed chloroform concentrations,
also exceeded the EPA's acceptable lifetime risk, with risks ranging from approximately 7E~5 to 2E~4 or
about 7 in 100,000 to 2 in 10,000 added risk. Again, added risks greater than one in a million are deemed
unacceptable.
Although risks (both dose-based and exposure-based) were not specifically provided using the
more detailed, less conservative "COMPLEX_MODEL," these risks also exceed EPA's one in a million
acceptable lifetime risk.
Given the exceedance of the various criteria, mitigated measures should probably be considered.
REFERENCES
1. Sanders, P. F., (2002) A screening model for predicting concentrations of volatile organic chemicals
in shower stall air, Division of Science, Research and Technology, New Jersey Department of
Environmental Protection, P.O. Box 409, Trenton, NJ 08625.
https://www.ni.gov/dep/dsr/research/A%20Screening%20Model%20for%20Predicting%20Concentra
tions%20of%20VOCs%20in%20Shower%20Stall%20Air-RPS.pdf. accessed 6/17/2020.
2. Davis M.J., Janke R., Taxon T.N. (2016) Assessing Inhalation Exposures Associated with
Contamination Events in Water Distribution Systems. PLoS ONE 11(12): eO 168051.
https://doi.oo mrnal .pone .0168051.
3. Chowdhury S. (2016) Exposure assessment for trihalomethanes in municipal drinking water and risk
reduction strategy. Sci Total Environ. 2013;463-464:922-930. doi:10.1016/j.scitotenv.2013.06.104.
5

-------
4.	Minnesota Department of Health. (2016). Chloroform and Drinking Water.
https://www.health.state.mn.us/communities/environment/risk/docs/guidance/gw/chlorofomiinfo.pdf
accessed on 6/18/2020.
5.	Nazaroff and Alvarez-Cohen (2001). Risk Assessment (carcinogenic slope factor for inhalation of
chloroform). http://www.dartmouth.edu/~cushman/courses/engs37/RiskAssessment.pdf.
6.	U.S. Environmental Protection Agency (1976). Unit Risk Factor for Chloroform.
https://cfbub.epa.gov/ncea/iris2/chemicalLanding.cfm7substance n.mbr=25. accessed 6/17/2020.
DISCLAIMER:
This document and the associated Excel spreadsheet include contributions from EPA and individuals
outside the United States Government. It has been subjected to review by the U.S. EPA's Office of Research
and Development (ORD) and approved for sharing to individuals and parties that have an interest in these
results. The sampling and analysis results for the chloroform observed in the tap water of the Mahnomen
Public Water System were not subjected to EPA's quality system requirements. The views expressed in
this summary are those of the author(s) and do not necessarily represent the views or the policies of the
U.S. Environmental Protection Agency. The results and discussion in this project summary have not been
formally disseminated by the Agency and should not be construed to represent any Agency determination
or policy.
6

-------