Professional Documents
Culture Documents
293-310 (1999)
Donna J. Vorhees
Cheri L. Butler
ABSTRACT
Hazardous waste site managers and regulators can decide easily whether site
contamination requires remediation to protect human health, but defining the extent oj‘
cleanup is often an enormous challenge. When is the site clean enough for human access
and use? Risk assessment tools can be used to help answer this question. One can
calculate chemical-specific risk-based screening levels (RBSLs) in environmental media
(e.g. soil, water, and food) that correspond to “acceptable” levels of risk. RBSLs can be
calculated for individual chemicals and chemical mixtures, specific to various exposure
pathways. In calculating RBSLs, one must account for the fact that people may be
exposed to more than one chemical by more than one exposure pathway. Fate and
transport models may be needed to predict how people might become exposed to site
contamination in the future. Example RBSLs for non-cancer effects are calculated to
illustrate the utility and limitations of RBSLs for making risk-bused decisions ut
petroleum release sites.
293
INTRODUCTION
Risk-based screening levels (RBSLs) represent maximum chemical concentrations in soil,
sediment, water, air or food that correspond to acceptable levels of human health risk.
These values are often calculated within a risk-based decision framework, such as the
American Society for Testing and Materials (ASTM) Standard Guide for Risk-Based
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
site collected in the preliminary assessment. Tier 1 RBSLs are conservative, employing
exposure pathway characteristics and modeling assumptions that result in RBSLs that are
lower than those that would be calculated if site-specific limitations on exposure or
contaminant transport were taken into account. The investigator compares maximum
contaminant concentrations at the site to the RBSLs. If the site concentrations exceed the
RBSLs, the investigator can choose from two options: (1) remediate site contamination to
meet the RBSLs;or (2) proceed to Tier 2 with collection of site-specific information and
calculation of site-specific target levels (SSTLs).
In Tier 2, the investigator calculates SSTLs by conducting simple fate and transport
modeling using site-specific information to predict the attenuation of contaminants away
from the source area. If maximum site contamination levels exceed the SSTLs, remedial
action may be implemented or analysis proceeds to Tier 3. Tier 3 is a thorough, site-
specific risk assessment. As each tier is completed, remedial action goals become more
site-specific and presumably, less conservative and less costly to achieve than those
defined by the previous tier. The reduction in conservatism and cost is due to
consideration of site-specific information, not a reduction in goals for human health
protection.
RBSLS FOR PETROLEUM 295
the source to a person, and a person who may be exposed to the contaminant3. There are
numerous potential exposure pathways, including inhalation of contaminated indoor air
and outdoor air, dermal contact with and ingestion of contaminated soil and groundwater,
and many other pathways.
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
RBSLs can be calculated in excess of C,,, for individual compounds or for fractions.
Therefore, C,,, values must be compared to RBSLs for all individual compounds and
fractions, and the lower value selected as the screening level. At CSat,the flux of
contaminants from soil to air and water reaches a plateau, no matter how much more
chemical is added to the soil. Therefore, exposure via these cross-media pathways cannot
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
increase beyond CSat. Exposure routes with soil RBSLs in excess of C,,, are not of
concern for human health.
A toluene RBSL is calculated for the “incidental ingestion of surface soil” exposure
pathway as follows:
HQ* AT * BW * R f o
(2) c =
EF * ED * IR *
$1
where: C = concentration of toluene in surface soil (mgkg)
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
HQ = hazard quotient = 1
Toluene also may leach from site soils into groundwater. A toluene RBSL is calculated
for the “leaching to groundwater pathway as follows:
H Q * AT * BW * R f o
(2) c =
EF*ED*IR
HQ = hazard quotient = I
BW = bodyweight = 70 kg (adult)
r 1
(3) RBSL,, [ X I=
kg soil LFnV
Where:
RBSL, = 2 17 m a g .
Where:
Ow, = volumetric water content in vadose zone soils (cm3 H20/cm3 soil)
O,, = volumetric air content in vadose zone soils (cm3 air/ cm3 soil)
This leaching factor equation results in conservative RBSLs because it assumes that no
dilution or attenuation of the leachate occurs as it migrates from the vadose zone to the
saturated zone.
EF * ED * IR * C
HQ =
AT * BW * R p
H Q * AT * BW * R p
C=
EF * ED * IR
HQ = hazard quotient = 1
BW = bodyweight = 70 kg (adult)
RBSL,, [XI=
kg soil VF,,
Where:
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
Where:
Dseff= effective diffusion coefficient between groundwater and soil surface (cm'/s)
C,;'[ mg
kg - soil
]=? *
P,~
+8,,+k,vp,J* 10' L- g
cm' - kg
302 VORHEES AND BUTLER
Where:
Any of the pathway-specific RBSLs for toluene could exceed CSarfor toluene. Therefore,
these values must be compared to the CSarfor toluene before selecting the RBSL that
either represents a toluene cleanup goal or triggers Tier 2 analysis. The “leaching to
groundwater” pathway has the lowest RBSL of 217 mgkg, which is lower than Csat for
toluene. Therefore, at this hypothetical site, 217 mgkg should be selected as the
screening level for toluene.
Petroleum release sites are difficult to evaluate and remediate because of the changing
composition and distribution of complex petroleum products following release to the
environment. Their composition changes over time and space due to individual
constituents partitioning differently among environmental media and due to
environmental degradation processes such as photolysis and microbial action. These
phenomena are collectively referred to as product “weathering.” For this reason, relying
on whole product (e.g, gasoline, #2 fuel oil, jet fuel) analytical data may be appropriate
for fresh spills, but not for older spills that have had time to weather. Recognizing this
dilemma, the MA DEP and the TPHCWG developed fraction-based approaches for
evaluating human health risk at petroleum release sites. The TPHCWG defined 13
RBSLS FOR PETROLEUM 303
RBSLs for TPH fractions are calculated using the method described for individual
compounds, substituting fraction-specific transport parameter values’ and toxicity
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
criteria5. C,,, for each fraction must be calculated and compared to the fraction RBSLs.
RBSLs for each TPHCWG fraction are listed in Table I. In Table I, “Csal”refers to those
fractions where C,,, is lower than the RBSL. Leaching and volatilization pathways have
RBSLs lower than C,,, only for the most mobile (i.e. soluble or volatile) and toxic
fractions. Cleanup decisions could be based on the TPH fraction with the lowest RBSL.
In calculating RBSLs, one assumes that the toxicity of a given fraction does not change
with different petroleum product sources. For example, the toxicity of the C ~ to
O CIZ
For personal use only.
aliphatic fraction measured at a gasoline spill site is the same as the toxicity of the Clo to
C12 fraction measured at a weathered #2 fuel oil spill site. Consequently, if exposure
assumptions and environmental transport model assumptions are held constant, the same
RBSLs will be calculated for each fraction, regardless of petroleum type.
RBSLs for TPH
A measured TPH concentration cannot be plugged into an RBSL equation to get a total
TPH RBSL because toxicity criteria do not exist for TPH. In addition, the toxicity of TPH
measured at one site is not necessarily the same as TPH measured at another site.
Therefore, TPH RBSLs can be calculated by weighting individual fractions by their
composition in the total petroleum mixture (f, = C, /CTPH)and assuming their toxic
effects are additive. TPH RBSLs should be calculated using petroleum fraction data that
are representative of TPH contamination at a site, taking into account all possible source
areas and differential degradation across the site.
Example TPH RBSLs in this section are calculated based on the 13 TPHCWG fractions.
Using petroleum fraction data, we can derive an RBSL for TPH that accounts for
additivity of toxic effects across fractions:
HI = C HQ, = C MIN [fi*CTpH/ RBSL, , Cisat/RBSL] S I
304 VORHEES AND BUTLER
Aliphatics
>10-12 86 4,300
>12-16 38 4,300
>16-21 16 87,000
Aromatics
Note: “Csal(’- substituted for pathway RBSLs that exceed Csalfor a given fraction.
RBSLS FOR PETROLEUM 305
For direct routes of exposure such as soil ingestion, dermal exposure and inhalation of
particulates, the degree of saturation does not limit exposure. However, as the degree of
saturation increases, particulate generation decreases. TPH RBSLs for the ingestion of
soil, dermal contact with soil, and soil particulate inhalation pathways are calculated as
follows:
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
The values for CTPHcalculated using the equations above are the pathway-specific RBSLs
for TPH.
When compounds are present in a mixture, equilibrium is achieved for each component
between the aqueous and organic phases. By assuming the organic phase is an ideal
solution, one can express the solubility of the component as its mole fraction multiplied
by its pure component solubility (Raoult’s Law)’. This calculated solubility is sometimes
For personal use only.
referred to as “effective solubility,” and its use results in calculation of lower C,,, values.
Table II lists example TPH RBSLs for four common petroleum products. These RBSLs
were calculated based on fraction composition data for fresh product, rather than
petroleum that has weathered in the environment. All of the TPH RBSLs reported in
Table II were calculated by assuming non-detected fractions were present in the mixture
at one half the detection limit. Table III lists TPH RBSLs for an example petroleum
mixture calculated by varying this assumption. The differences are due largely to the
detection limit for the benzene fraction, which was not detected in the mixture.
Petroleum Products
indoor air
Cancer risk estimates at some sites will drive the extent of remedial activities. At these
sites, calculation of RBSLs for noncarcinogenic effects may not be worthwhile when it is
clear that such analysis would not lead one to recommend more stringent remediation
goals than those resulting from the analysis of cancer risk. In some cases, RBSLs may not
be protective of ecological or aesthetic impacts. The potential for site contamination to
impact the ecosystem or aesthetics must be evaluated separately.
Assumption for
NDs = DL 70 630
Legend: ND = Non-detect
DL = Detection Limit
Does the Target Hazard Quotient Ensure the Desired Level of Human Health
Protection ?
To ensure that RBSLs attain the desired level of human health protection, one must
consider the number of potential exposure pathways and chemicals to which people might
308 VORHEES AND BUTLER
1.E+06
1.E+05 I----
Weathered Diesel
n
I-
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
w
E I.E+03
1.E+01
1.E+OO
For personal use only.
Figure 1. Comparison of TPH RBSLs for fresh and weathered product. Error bars for
weathered product indicate the range of RBSLs calculated for 4 soil samples. Product
composition data were taken from Reference 13.
be exposed in setting target hazard indices and target cancer risk levels. At a site where
more than one pathway exists, RBSLs can be calculated from combined pathways or the
lowest pathway RBSL can be chosen as the screening value. For example, TPH RBSLs
calculated by combining all complete soil pathways ensures that the desired level of
protection is attained.
lU3SLs for weathered petroleum might differ from RBSLs for fresh product due to
changes in petroleum mixtures following release to the environment. With time, the more
mobile components of the mixture migrate away from the site, leaving a less mobile
mixture behind. Concentrations decrease due to these transport mechanisms as well as
RBSLS FOR PETROLEUM 309
biodegradation. There is some evidence that organic compounds bound to soil become
less available for uptake with humans the longer they remain in the soil“’. The toxicity of
the mixture could increase or decrease, depending on whether the pattern of
contamination shifts toward more or less toxic TPH fractions.
Drug and Chemical Toxicology Downloaded from informahealthcare.com by Universitat de Girona on 11/21/14
Figure I shows RBSLs for fresh and weathered samples and diesel. Some weathered
product RBSLs are higher and some are lower than fresh product RBSLs. However, this
plot includes few data (n=5); therefore, no strong conclusions can be made for these two
products.
CONCLUSIONS
Calculation of RBSLs provides a fast and simple way to estimate “acceptable” levels of
risk for human health exposure at a petroleum release site. Many variables influence the
For personal use only.
REFERENCES
1. American Society for Testing and Materials. Standard Guide for Risk-Based
Corrective Action Applied at Petroleum Release Sites. E 1739-95, 1995.
2. United Stated Environmental Protection Agency (US EPA). Risk Assessment
Guidance for Superfund: Volume I - Human Health Evaluation Manual, Part B:
Development of Risk-based Preliminary Remediation Goals, Interim Final. (EPA540-
1-89-002). 1991.
3. United Stated Environmental Protection Agency (US EPA). Soil Screening
Guidance: Users Guide (EPA 540-R-96-018), April 1996)
4. U.S. Environmental Protection Agency (USEPA). Integrated Risk Information
Database.
5. Total Petroleum Hydrocarbon Criteria Working Group (TPHCWG), Development of
Fraction-Specific Reference Doses ( R m s )and Reference Concentrations (RfCs)for
Total Petroleum Hydrocarbons (TPH), 1997.
3 10 VORHEES AND BUTLER
9. Shiu, W.Y., Maijanen, A., Ng, A.L.Y., MacKay, D. Preparation of Aqueous Solutions
of Sparingly Soluble Organic Substances: II. Multicomponent Systems - Hydrocarbon
Mixtures and Petroleum Products. Environmental Toxicology and Chemistry, Vol.
7~125-137,1988.
10. Alexander, M. 1995. Environmental Science & Technology, 29:27 13-2717.
11. Friedman & Bwya, Inc., Results From The Analysis of Soil Samples For Aliphatic
Volatiles By GC/FID, Weathered JP-4 Characterization, 1995.
12. Mayfield, H. T., JP-8 Composition and Variability, Armstrong Laboratory,
Environics Directorate, Environmental Research Division, 1996.
13. Nakles,D.V., Lynch, J.W.,Edwards, D.,Tell, J.G., Potter,T.L., Andes, R.P., and
Barkan, C.P.L., Risk-Based Management of Diesel-Contaminated Soil, Prepared for:
Association of American Railroads, Washington, D.C., Prepared by: Remediation
For personal use only.