5. CHAPTER QUANTITATIVE RISK ASSESSMENT (from LaGrega et
al., p837-885)
My teaching goals for this chapter are for you:
1) to learn what information is required to perform a risk assessment
2) to learn how to calculate the risk associated with a given chemical
concentration
3) to appreciate the uncertainty associated with risk assessment
RISK: probability of suffering harm or loss
If the level of risk can be quantified we can calculate it as follows:
risk = (probability) * (severity of consequence)
CEE 440
(5.1)
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
1
Example:
If Bob smashes his finger while loading boxes he loses 5 days of work
Severity of Consequence = 5 days
The likelihood of Bob smashing his finger is 10%
Probability = 0.1
Hence,
Risk = 0.1 * 5 days = .5 person days lost
Risk must be divided into background and incremental risk:
Background risk: occurs in the absence of a particular source of risk
Incremental risk: caused by the source of risk
Total risk: Background + Incremental
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
2
Example:
U.S. Background Cancer risk = 0.25
Excess lifetime cancer risk deemed acceptable by the US EPA at Superfund
sites = 1x10-6
Total risk = 0.250001
Death/Cancer Risks
Motor Vehicle Accident
Home Accidents
Annual risk
0.00024
0.00011
Uncertainty
10%
10%
Cigarette Smoking (pack/day)
Peanut butter (4 tsp./day)
0.0036
8x10-6
Lifetime risk
1x10-6
Factor of 3
Factor of 3
Uncertainty
Factor of 10 to 100
Drinking Water with EPA limit of TCE
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
3
Risk Assessments can affect:
1) waste treatment/disposal options
2) remediation of contaminated site
How clean is clean?
What remediation technology do we use?
How do we protect the people doing the cleaning?
3) minimization of waste generation
4) placement of new treatment facilities
5) developing new products
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
4
Four Steps to Risk Assessment
1) Hazard Identification: which chemicals important
2) Exposure Assessment: where do chemicals go, who might be exposed
and how
3) Toxicology Assessment: determine numerical indices of toxicology for
computing risk
4) Risk Characterization: estimate the magnitude of risk, and the
uncertainty of the estimate
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
5
5.1 HAZARD IDENTIFICATION
The potential hazard a chemical represents can be evaluated based on
concentration and "danger"
One approach is outlined below:
1) Sort the contaminants by medium (i.e. soil, groundwater, etc.)
2) Tabulate mean and range of concentration values of each chemical at
site
3) Identify reference doses for non-carcinogens and slope factors for
carcinogens for each potential exposure route
4) Determine the toxicity scores for each chemical in each media:
For non-carcinogens:
TS = Cmax/RfD
(5.2)
where TS = toxicity score
Cmax = maximum concentration
RfD = chronic reference dose
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
6
For carcinogens:
TS = SF * Cmax
where SF = slope factor
(5.3)
5) For each exposure route (i.e. air, water, soil) rank the compounds by
toxicity score
6) For each exposure route, select those chemicals comprising 99% of the
total score
Other factors which must be taken into consideration to evaluate toxicity are:
- mean concentration
- frequency of detection
- mobility
- persistence in the environment
- treatability
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
7
It is not rare to detect as many as 100 different chemicals at a
contaminated site
Hence, a subset of chemicals (surrogates) is often chosen to be
representative of all the chemicals detected.
This limits the number of chemicals analyzed and focuses efforts on most
significant hazards
Surrogates are chosen based on which chemicals best represent the
following risks:
1) Most toxic, persistent, and mobile contaminants
2) Most prevalent (w.r.t. spatial distribution and concentration)
3) Most exposure
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
8
5.2 EXPOSURE ASSESSMENT
Assessing exposure requires identifying and/or quantifying
a) contaminant transport and fate
source characterization
advection, dispersion, sorption, transformation characterization
exposure points (i.e. where does contamination reach?)
b) general and sensitive populations who will be contaminant receptors
who is coming into contact with contaminants and
how?
c) doses to population from short and long term exposure routes
based on exposure, how much is actually taken into receptor?
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
9
Up to now, CEE440 has been focused on contaminant transport and fate.
Example of two simple but common transport and fate scenarios are shown
below. One involves groundwater, the other atmospheric transport (Figure
14-1 from LaGrega et al., p 845).
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
10
Below are schematics of
potential soil and
groundwater transport
and fate pathways: (Fig
14-2a,2b,2c from
LaGrega, p846-848)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
11
Once we figure out where the contaminants are going we must determine:
Potentially Exposed Populations
present population in vicinity of site
future population in vicinity of site
sub-populations of special concern (i.e. children, fishermen)
Identifying these populations requires identifying different exposure scenarios:
worker scenario
trespasser scenario
residential use scenario
recreational use scenario
construction scenario
The last step is to identify:
Receptor Doses
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
12
Three exposure routes are commonly considered:
ingestion
inhalation
dermal contact
A general formula that represents the relevant parameters to consider for dose
is:
I = (C * CR * EF * ED) / (BW * AT)
(5.4)
I = intake (mg/kg of body weight)
C = concentration at exposure point (mg/L in water, mg/m3 in air, mg/kg in
soil)
CR = contact rate (e.g., L/day, m3/day)
EF = exposure frequency (day/year)
ED = exposure duration (yr)
BW = body weight (kg)
AT = average time (days)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
13
Modified forms of this equation have been developed to account for specific
exposure pathways. For example, the intake dose from the inhalation of
fugitive dust may be calculated as follows:
I = (C * CR * EF * ED * RR * ABS) / (BW * AT)
(5.5)
RR = retention rate (decimal fraction)
ABS = absorption into bloodstream (decimal fraction)
C = Cs * Pc
Cs = concentration of chemical in fugitive dust (mg/mg)
Pc = concentration of fugitive dust in air (mg/m3)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
14
We can also represent the intake via dermal exposure as follows:
I = (C * A * (DA/Exposure Event) * ABS * SM * EF * ED) / (BW * AT) (5.6)
C = mg chemical/kg of solid
A = surface area of skin exposed (say 20% of total skin area in cm2)
DA = dust adherence (0.51 mg/cm2 per exposure event)
ABS = skin absorption rate (~6%)
SM = effect of soil matrix (~15% of contaminant on soil available for
dermal contact)
EF = frequency of contact (day/year)
ED = exposure duration (yr)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
15
Typical Exposure Parameter Table (if living by landfill)
Parameter
Average body wt(kg)
Skin surface area (cm2)
Water ingested (L/day)
Air breathed (m3/hr)
Retention rate (inhaled air)
Absorption rate (inhaled air)
Soil ingested (mg/day)
Bathing duration (min)
Exposure frequency (days)
Exposure duration (yrs)
CEE 440
Adults
70
18,150
2
0.83
100%
100%
100
30
365
30
Child age 6-12
29
10,470
2
0.46
100%
100%
100
30
365
6
Child age 2-6
16
6,980
1
0.25
100%
100%
200
30
365
4
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
16
Example:
Calculate the average daily intake of chloroform for on-site workers over one
year from dermal contact of a soil with a mean concentration of 1.12 mg/kg
of soil, where 2 exposure events per day occur over a period of 156 work
days/year.
Solution:
I = (C * A * (DA/ExposureEvent) * ABS * SM * EF * ED) / (BW * AT)
A = 18,150 cm2 * (0.2) = 3630 cm2
DA = 0.51 mg/cm2 per exposure event
ABS = 0.06
SM = 0.15
EF = 2 exp. events/day * 156 days/year
ED = 1 year
BW = 70 kg
AT = 1 year = 365 days
I=
mg
Exp. Ev.
day
- kg
0.51 mg
1.12 kg * 3630 cm 2 * Exp. Ev. 2 * 0.06 * 0.15 * 2 day * 156 yr * 1yr * 10 6 mg
cm
[]
I = 2.27x10-7 mg/kg-day
CEE 440
70[kg]* 365days
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
17
5.3 TOXICOLOGY ASSESSMENT
When assessing contaminant transport and fate we evaluated contaminant
intake via three different exposure pathways (inhalation, ingestion, dermal
contact). We now want to understand the relationship between dose (or
intake) and response (or adverse health effect).
This involves determining the dose-response relationships for each chemical
of interest (surrogate chemicals) and determining the confidence in these
relationships.
Note: as an engineer you do not have an adequate background to conduct
toxicological evaluations. However, in order to understand how toxicological
decisions are made you should understand the general methods that
toxicologists use.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
18
Toxicology is, to a large extent, supported by information obtained from animals.
Extrapolating from animals at high doses to humans at low doses (i.e. doses
usually encountered) carries with it a high degree of uncertainty. Hence,
estimates of toxicological risk can be orders of magnitude off. Nevertheless, site
cleanup is being more and more driven by a risk based approach and we must
understand the basics of toxicology to interpret risk based decisions with any
intelligence.
Chemicals are classified as non-carcinogens and carcinogens. In hazardous
waste management we evaluate the toxic nature of non-carcinogens via the
reference dose (RfD) and of carcinogens via the slope (SF).
RfD factor = the acceptable daily intake as established by the EPA [mg/kg-day]
SF =
the slope of the dose-response curve at very low exposures, often
referred to as the carcinogen potency factor (CPF) [kg-day/mg)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
19
Calculating RfD
1. Select the most sensitive species for which adequate studies are
available (unless other species more appropriate reference for humans). If
human data are available these are given priority
2. Select the principal or critical studies using the appropriate route of
exposure. RfDs are route specific (i.e. inhalation, ingestion, dermal).
3. Select supporting studies. These studies may relate effects in animal
species of interest to humans, etc.
4. Identify the "No Observed Adverse Effect Level" (NOAEL) or, if NOAELs
are not available, the "Lowest Observed Adverse Effect Level" (LOAEL).
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
20
5. Adjust the NOAEL downward by orders of magnitude to reflect
uncertainty:
Reduce the NOAEL found in humans by an uncertainty factor of 10 to
account for variations in the general public
Reduce the NOAEL by an additional uncertainty factor of 10 when
extrapolating from animals to humans
Reduce the NOAEL by an additional uncertainty factor of 10 if the data are
derived from a sub-chronic instead of a chronic study
If no NOAEL were available (i.e. adverse health effects were observed at all
health levels), reduce the LOAEL by a an uncertainty factor of 10.
The EPA also applies a modifying factor, which ranges from 1 to 10, to
reflect a qualitative professional judgment of uncertainties which are
not accounted for by the preceding uncertainty factors.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
21
Example:
In a subchronic oral toxicity study in mice, a LOAEL of 5 mg/kg-day was
determined for a specific agent. The quality of the data is given a high rating
by the expert evaluating the data. What is the RfD?
Area of uncertainty
Uncertainty
factor
Variation within population
10
Extrapolation from animals to humans
10
Extrapolation from subchronic to chronic
10
Extrapolation from LOAEL to NOAEL
10
Modifying factor
1
RfD = 5 mg/kd-day / (10*10*10*10*1) = 0.5 g/kg-day
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
22
Calculating SF
1) The carcinogenic response of an animal to a specific exposure route can be
obtained at different doses and plotted as shown below (Figure 5-23, LaGrega,
1994)
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
23
2) In order to obtain any carcinogenic response high doses are typically used.
Hence, these must be extrapolated down to low doses to obtain our SF.
Note: Extrapolation of data from the 10-90% carcinogenesis range of test
animals to 0.0001% carcinogenesis may yield results which are off by several
orders of magnitude. The range of error in experimental dose-response
extrapolation for low carcinogenic frequencies in potential exposed human
populations is indicated below (Figure 5-24 LaGrega, 1994):
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
24
EPA classification system for carcinogenicity
A
Human carcinogen
B1
Probable human carcinogen, limited human data available
B2
Probable human carcinogen, no human data available (animal data)
C
Possible human carcinogen
D
Not classifiable as to human carcinogenicity
E
Evidence of noncarcinogenicity for humans
Note: approximately 50 human carcinogens are known, 13 of these are found
in the environment, and the rest are primarily pharmaceutical products
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
25
Human environmental carcinogens
Carcinogen
Arsenic
Asbestos
Benzene
Benzidine
Bis-chloromethyl-ether
Chromium, hexavalent
Diethylstilbestrol
2-Naphthylamine
Nickel
Plutonium-239
Radium-226
Radon-222
Vinyl chloride
CEE 440
Cancer
Lung, skin
Lung, mesothelioma
Leukemia
Bladder
Lung
Lung
Female
Bladder
Lung
Lung
Lung, osteosarcoma
Lung
Liver, angiosarcoma
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
26
5.4 RISK CHARACTERIZATION
The final step in risk assessment is to estimate the risks.
This consists in part on calculating estimates of both carcinogenic and noncarcinogenic risks to susceptible receptors for all exposure scenarios
considered important at both maximum and average exposure
concentrations.
It is important to note that the calculated values are only as good as the
data used to calculate them. Each piece of data has a confidence level
associated with it. And this all factors into the confidence in the final
esimate.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
27
Carcinogenic Risks:
Risk = I * SF
(5.7)
I = chronic daily intake calculated in the exposure assessment (mg/kg-day)
SF = carcinogen slope factor (kg-day/mg)
The EPA has designated a risk < 1x10-6
acceptable.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
28
Non-Carcinogenic Risks
Normally characterized in terms of a hazard index (HI).
HI = I / RfD
(5.8)
HI = hazard index (dimensionless)
I = chronic daily intake (mg/kg-day)
RfD = reference dose (mg/kg-day)
Note: reference doses are dependent on the route of exposure and
may only be used with exposure data for the same route
A HI < 1 is acceptable. If greater than one chemical is present, then the sum
of the HI for all chemicals should be <1 for the risk to be acceptable. If
different chemicals affect different organs, then the HI for all chemicals that
affect a particular organ can be summed individually to determine risk.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
29
Example:
Chloroform has reached a drinking water well for a small town and is being
consumed by the town occupants. The average concentration of chloroform in
the well is 4.30x10-4 mg/L. For an adult, calculate the carcinogenic and
noncarcinogenic risk of drinking the groundwater contaminated with chloroform
assuming water is drunken daily over a year, the parameters given in the Typical
Exposure Parameter Table and I = (C * CR * EF * ED) / (BW * AT). SF = 6.1x10-3
kg-day/mg; RfD = 1.0x10-2 mg/kg-day
C = 4.30x10-4 mg/L
CR = 2 L/day
EF = 365 day/yr
ED = 1 yr
BW = 70 kg
AT = 365 days
I = 4.30x10-4 [mg/L] * 2 [L/day] * 365 [day/yr] * 1 [yr] / (70 [kg] * 365
[days])
I = 1.22x10-5 mg/kg-day
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
30
Carcinogenic Risk = I * SF = 1.22x10-5 * 6.1x10-3 = 7.44x10-8
This is below the EPA risk threshold of 1x10-6 so the townsfolk are "safe"
Non-carcinogenic Risk = HI = I/RfD = 1.22x10-5 / 1.0x10-2 = 0.00122
This is below 1 so the townsfolk are "safe"
Up to now we have only been examining human health effects. There is an
increasing emphasis to look at ecological health effects.
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
31
5.5 ECOLOGICAL RISK ASSESSMENT
Estimate of the adverse effect to an ecosystem from a hazardous waste site
Such assessments at most Hazardous Waste sites are minimal at best.
The basic steps in an ecological risk assessment are the same as those
previously covered when examining the risks to humans. However, ecological
risks assessments are more complicated because:
- multiple biological endpoints: multiple species
- more complex exposure pathways: determined by the biological endpoints
- indirect effects: habitat impairment, disruption of intertrophic relationships
- evaluating impacts on ecosystems: very complex environment
On the other hand, there are advantages to an ecological risk assessment:
exposures and hazards can often be estimated directly on the species of
concern
CEE 440
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
32
Tiered Approach to Corrective action Objectives (TACO)
Initial Site Assessment
-Conduct site investigation and identify principal chemical(s) of concern, extent of
affected environmental media, and potential migration pathways and receptors.
Compare
measured values
to look-up table
values
Involves
changing input
parameters on
standard risk
equations based
on site specific
data
Site Classification and Initial Response Action
-Classify site per specified scenarios and implement appropriate initial response action.
-Reclassify site as appropriate following initial response action, interim remedial action, or
additional data collection.
Interim Remedial Action
-Conduct partial source removal or
other action to reduce risk(s) and
site classification
Tier 1 Evaluation
-Identify reasonable potential sources, transport pathways, and exposure pathways (use
flowchart in Figure ??).
-Select appropriate Tier 1 risk-based screening levels (RBSLs) from Tier 1 Look-Up
Table, or other relevant criteria (taste, odor thresholds, etc.) Compare these values with
site conditions
no
Chemical(s) of concern
concentrations exceed
RBSLs?
yes
no
Remediation to Tier 1
RBSLs practicable?
Interim remedial
action appropriate?
yes
no
Tier 2 Evaluation
-Collect additional site data as needed.
-Conduct Tier 2 assessment per specified procedures. Compare Tier 2 site-specific
target levels (SSTLs) with site conditions.
no
Chemical(s) of concern
concentrations exceed
SSTLs?
yes
no
Remediation to Tier 2
SSTLs practicable?
Interim remedial
action appropriate?
yes
Involves
transport and
fate modeling
based on site
specific data
yes
yes
no
Tier 3 Evaluation
-Collect additional site data as needed.
-Conduct Tier 3 assessment per specified procedures. Compare Tier 3 site-specific target
levels (SSTLs) with site conditions.
no
Chemical(s) of concern
concentrations exceed
SSTLs?
yes
Interim remedial
action appropriate?
yes
no
Remedial Action Program
-Identify cost-effective means of achieving final corrective action goals, including combinations of
remediation, natural attenuation, and institutional controls. Implement the preferred alternatives.
Continued monitoring
required?
no
yes
Compliance Monitoring
-Conduct monitoring program as needed to confirm that corrective action
goals are satisfied.
CEE 440
No Further Action
2011 Charles J. Werth, University of Illinois at Urbana-Champaign. All rights reserved.
33