Toxicological Risks of Selected Flame-Retardant Chemicals (2000)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

EXPOSURE ASSESSMENT METHODOLOGY 35 3 Exposure Assessment Methodology DETERMINATION of whether particular uses of flame-retardant (FR) chemicals in fabrics are safe is complicated by our lack of information in multiple areas. First, all available toxicity studies are on the pure chemicals that are present in fibers before yarn and fabric formation or are applied to the fabrics during processing. The chemicals might be incorporated into backcoatings applied to the fabric, or they might be introduced into the fibers. They might have different physical and chemical forms from those used in the toxicity testing. Thus, the finished or treated fabrics possibly contain derivatives of the applied chemicals as a consequence of cross-linking, polymerization, or oxidation processes having occurred (see Appendix B for a discussion of the finishing and treatment technologies and the processes that occur therein). These derivatives might bear little chemical resemblance to the original species in terms of generic structure, types of functional groups present, solubility, volatility, and general reactivity. Toxicity tests are usually performed by gavage or oral feeding studies using the pure and unreacted chemical mixed with a carrier vehicle or animal food. The chemical is not adsorbed to fabric fibers. Second, for most of the FRs evaluated, only limited toxicity information is available for the oral route of exposure, and even less is available for the inhalation route. For the dermal route, the minimal toxicity information that is available is not adequate for developing dermal reference doses. In most cases, dermal and inhalation toxicities have to be inferred from the toxicity information available from the oral route. 

EXPOSURE ASSESSMENT METHODOLOGY 36 Third, the toxicity information that is available (for example, reference doses (RfDs) and reference concentrations (RfCs) is expressed in terms of the applied amount—not in terms of the amount that is available inside the body to cause toxicity. The difference in absorption with each route makes the route-to-route extrapolations more uncertain. Finally, potential exposures to FRs applied to furnishing fabrics within the home have not been studied. Thus, there is little basis for estimates of exposure to such materials. There are few, if any, measurements of exposures under relevant conditions of exposure, and the subcommittee located no quantitative measurements of such exposures. The volume of literature related to residential pesticide exposures was reviewed but not considered relevant because pesticides are not ordinarily applied to fabrics during the manufacturing process. The effectiveness of pesticides is not dependent on their ability to remain in the fabric during use. The subcommittee does not believe that using the small amount of pesticide data that might relate to fabrics would reduce uncertainty. The subcommittee believes that pesticides are not good surrogates for FR chemicals. To make progress, therefore, the subcommittee adopted some extremely conservative assumptions (that is, corresponding to high concentration and exposure conditions) about potential exposures. In these estimations, the subcommittee determined whether each FR chemical would pose an acceptable risk even with such assumptions. If the risk was acceptable, then that chemical could be dropped from further consideration. Subsequent iterations of the procedure would then depend on finding more defensible information about the exposure conditions. The subcommittee was unable to find any such information and recommends collection of such information—a process that should be relatively straightforward but is outside the subcommittee's charge. People can come in contact with furnishing fabrics through direct contact (for example, by sitting on them). However, most of the time the contact is likely to be very small because of the presence of clothing and because the FR is incorporated into the fabric fiber structure or is present in a backcoating formulation added on the reverse face of the fabric. Young children, in particular, might suck on furnishing fabrics; therefore, possible dissolution in saliva and ingestion of the FRs present must be considered unless they can be demonstrated to remain “locked” within the fibers or in the backcoating resin formulation. Finally, as fabrics wear they can shed small fibers, the majority of which are likely to be too large to be inhaled to any substantial degree. However, some inhalation exposure of FR chemical species could result from the generation of particles of respirable size (≤10 µm). Actual exposures to FR chemicals applied to fabrics are likely to be limited by multiple factors. FR chemicals are of little use unless they stay in the 

EXPOSURE ASSESSMENT METHODOLOGY 37 backcoating or the fabric throughout its life, so durability is a requirement for their selection. An FR treatment that wears off easily, washes off, or evaporates from the fabric is less suitable for its intended purpose, and might actually fail any prescribed durability test-performance requirements. The subcommittee chose to evaluate three exposure scenarios that are intended to represent the three routes of exposure (dermal, ingestion, and inhalation). For each of these exposure scenarios, exposures estimates were obtained for the general population1 with deliberately conservative assumptions, as detailed below. For dermal exposure, the exposure scenario is that of an adult sitting on a couch for a substantial fraction of his or her time, with potential exposure of the skin directly or indirectly (through clothing) in contact with the couch. For the ingestion exposure scenario, the subcommittee examined the scenario of a child repeatedly sucking on the treated fabric (for example, chair or couch fabric). For inhalation exposure, the scenario is that of a person spending a considerable fraction of his or her time in a room containing FR-treated furniture from which the FR is shed as small (respirable) particles or from which the FR chemical evaporates. In all these scenarios, it is plausible that some quantities of FR can be transferred from furniture to humans. The subcommittee aims to overestimate the quantities by using exaggerated estimates of values for such controlling factors as the time of exposure or the rate of movement of the FR. DERMAL EXPOSURE SCENARIO The subcommittee chose to address the problem in two iterations based on ease of determining the information and impact on the final answer. Release from the fabric might result in direct contact of the chemical with the skin surface. Transfer of this chemical through the skin surface would provide an internal dose. In the first iteration, the release rate was the only limiting factor and any chemical on the skin was assumed to completely transfer within the body and chemicals that would not cause health concerns under this scenario were eliminated from further calculations. For the remaining chemicals, an estimate of the transfer across the skin was used to estimate the internal dose and the hazard index. 1Exposures to workers involved in the manufacture of FR chemicals, manufacture of FR-treated fabric or furniture using FR-treated fabric were not evaluated. 

EXPOSURE ASSESSMENT METHODOLOGY 38 First Iteration As a first estimate of exposure, it was assumed that the skin and clothing of the person sitting on fabric would present no barrier to movement of a nonionic FR. Further, it was assumed that there would be sufficient water present (for example, from sweat) to allow dissolution of the nonionic FR in that water, with subsequent transfer to the skin and then into the body of the sitting person. With those assumptions, the only limiting factor on the transfer rate of nonionic FRs is assumed to be the limited dissolution rate of the FR on the fabric. All the FR that dissolves is assumed to be immediately absorbed by the sitting body. It is generally believed that the skin is an extremely good barrier to ionic chemicals (Grasso and Lansdown 1972). Swarbrick et al. (1984) showed that the ionized form of carboxylic acids penetrates the skin four orders of magnitude more slowly than the nonionized form. For ionic FRs, the subcommittee assumed that the permeability of skin to water (10−3 cm/hr) would provide a conservative estimate (EPA 1992). Under the exposure conditions described above, the dose rate for the dermal route for FRs is calculated using Equation 1: (1) where the meanings and values assigned to the symbols are as follows: D = The dose rate of chemical (mass of chemical per unit body weight per unit time). This rate is the desired value calculated using the above formula. Sa = The area density of the FR (the application rate to the fabric or back-coating—mass per unit surface area). This value is chemical specific and was chosen at the highest value likely to be used. It ranged from 2 to 7.5 mg/cm2 depending on the treatment type. (The range was chosen from the experience of the UK's textile market in meeting the UK's furniture-fire regulations). Ab = The area of body in contact with the couch was chosen to be 2,200 cm2. This value is based on 8,880 cm2 for the total body surface of the upper extremities (trunk, arms, and neck) of an adult (EPA Exposure Factors Handbook, Table 4–4). A worst-case estimate of body surface repeatedly in contact with furniture for long periods would be about 1/4 of the bare upper torso, or 2,200 cm2. 

EXPOSURE ASSESSMENT METHODOLOGY 39 µw = The fractional rate (per unit time) of FR extraction by water (e.g., sweat) assumed to be present under the given conditions. This rate is chemical specific; it was generally estimated from extraction measurements or laundering tests and ranged from 0.0004 to 0.038 per day. fc = The fraction (dimensionless) of time spent on the couch by the adult was assumed to be 1/4, or 6 hr/d (every day). The subcommittee believes, based on measurements of how people spend their time, that 6 hr/d may be considered a reasonable upper bound. Wa = The adult body weight (mass) was assumed to be 70 kg. The computed internal dose rate was then divided by the dermal RfD (external dose) to determine a hazard index. In practice, the oral RfD was substituted for the dermal RfD, because the latter was not available. For those FRs considered to be possibly carcinogenic, an overestimate of lifetime risk was obtained by multiplying the dose rate by the carcinogenic potency slope (q1*) for the FR. If the hazard index for a particular chemical was less than one with the assumptions just described, the subcommittee considered the exposures via the dermal route to be sufficiently small to merit no further examination. If the hazard index exceeded one in the first iteration, an alternative iteration of the exposure assessment was performed in which some consideration was given to the skin as a barrier to penetration. Alternative Iteration For the alternative iteration of the dermal assessment, the exposure assumptions were the same as those in the first iteration, except that the assumption of 100% immediate absorption of all the FR that dissolved was modified. Instead, an estimate of the rate at which the FR could penetrate the skin was made, assuming that the FR dissolved up to its solubility limit in water. That rate of penetration was then factored into the exposure assessment. The rate of penetration of a chemical through skin may be estimated using the skin permeability coefficient (Kp, with dimensions of velocity)—the total mass penetration rate is the product of water concentration, permeability coefficient, and skin area. Such coefficients have not been measured for the FR chemicals, but they may be estimated from the octanol-water partition coefficient (Kow, dimensionless) and molecular weight (m, mass/unit amount of substance) by using a correlation (Potts and Guy 1992) that may be written in dimensionless form (Equation 2): 

EXPOSURE ASSESSMENT METHODOLOGY 40 (2) where Kp0 = 1 cm/sec is a reference permeability coefficient, and m0 = 1 g/mol is a reference molecular weight. Using the permeability coefficient, the dose rate for the dermal route (alternative iteration) is obtained using Equation 3: (3) where the meanings and values assigned to the symbols are the following: D = The dose rate of a chemical (mass per unit body weight per unit time). This rate is the desired value calculated using the above formula. Cw = The water solubility of FR (mass/unit volume), which is different for each chemical. Kp = The permeability coefficient (length/time), which is different for each FR and calculated from the correlation given above. Ab = The area of the body in contact with the couch (equal to 2,200 cm2) as described for the first iteration. fc = The fraction (dimensionless) of time spent on the couch by the adult (1/4, or 6 hr/d), as described for the first iteration Wa = The adult body weight (mass) (70 kg), as described for the first iteration. Because the exposures of interest were 6 hr per day over a lifetime, this equation uses the historical steady- state-flux relationship instead of an alternative which adjusts for the concentration of chemical in the skin (EPA 1992). Once again, this dose rate was divided by the oral RfD (as the best estimate of the internal dose for comparison with the calculated internal dose from dermal exposures) to obtain a hazard index. If the hazard index was less than one, the subcommittee considered the dermal exposure route to be sufficiently small to merit no further attention. Also, as in the first iteration, for those FRs considered to be possibly carcinogenic, an overestimate of lifetime risk was obtained 

EXPOSURE ASSESSMENT METHODOLOGY 41 by multiplying the dose rate by the carcinogenic potency slope (q1*) for the FR. In cases where both iterations were performed for an FR, the lower of the two calculated dose rates was taken as the subcommittee's best available estimate of a conservative dose rate. Each iteration examines the effects of just one mechanism that limits dose rates. The first examines just the dissolution rate of the FR in water, and the second just the barrier presented by the skin to permeation by the FR. In reality, both of these mechanisms, and more besides, act to limit dose rates. It would be possible to include both of these mechanisms in a single model, but to do so would require information beyond that required when modeling each mechanism in assumed isolation. Each of the mechanisms considered introduces a resistance to movement of the FR into the body, and their combination introduces more resistance than either alone. However, the combined effect of both mechanisms cannot reduce the dose rate by more than a factor of two below the lower of the two dose rates estimated in the two iterations-that is, for each mechanism acting in isolation. There are many uncertainties in the dermal-exposure parameters and the calculations of the hazard indices. The subcommittee believes that the actual exposures are at least 100-fold lower than calculated. Parameters and calculations were deliberately chosen to provide a worst-case estimate. The subcommittee wanted to be sure that chemicals for which no further research was recommended would be safe, so conservative choices were made at every practical juncture. The most conservative assumptions in the dermal-exposure parameters relate to the body-surface area exposed, fraction of the day on the couch, and the daily exposure for a lifetime. The assumption of 2,200 cm2 is based on 8,880 cm2 for the total body surface of the upper extremities (trunk, arms, and neck) of an adult (EPA Exposure Factors Handbook, Table 4–4). One-quarter of the body-surface area not covered by clothing and being in contact with a couch is a very high estimate of exposed surface area. According to Table 14–2 of the Exposure Factors Handbook, the greatest time spent on any of the activities that might be done on the couch (such as watching TV, reading, or conversing) was 4.4 hr on a weekend for a male 12–17 yr old. For males and females of other ages, exposure time was significantly less. On weekdays, exposure for all groups was less. The assumption that the exposure occurs daily for a lifetime is also very conservative, and it assumes no deviation from the exposure for any reason. The most conservative assumptions in calculating hazard indices are that the constant release rate of FRs is for the life of the fabric, permeability coefficients are probably overestimated for the high-molecular-weight chemicals, the applied chemicals do not react in the fabric, and there is no attempt to limit the exposures by the actual amount of chemical in the fabric. An assumption that is not conservative is the use of the oral RfD as the internal dose for compari 

EXPOSURE ASSESSMENT METHODOLOGY 42 son with the calculated internal dose. The dermal iterative calculations assumed that the lesser of the release rate or the dermal absorption rate was rate limiting, when in actuality both would be rate limiting. Toxicity based on oral studies for chemicals that might have been incompletely absorbed would underestimate risk by the dermal route when the hazard index was calculated. INHALATION EXPOSURE SCENARIO Particles In the inhalation exposure scenario, a person spends some fraction of his or her time in a room containing FR-treated upholstered furniture. Some of the FR might be worn away during everyday use of the upholstery, and some of the particles so eroded might be small enough to be entrained into the air of the room and be inhaled. The concentration of such small particles in the room air will depend on the amount of upholstery in the room, the volume of the room, and how fast air is drawn through the room (the air exchange rate). The average concentration of FR present on the upholstery fabric or as small airborne (respirable) particles is estimated using Equation 4: (4) where the meanings and values assigned to the symbols are Cp = The average concentration (mass/unit volume) of FR attached to respirable particles in the room, calculated using the above formula. Sa = The area density of the FR (application rate to the fabric or back-coating—mass per unit surface area). This value is chemical specific, and the highest value likely to be used was chosen. It ranged from 2 to 7.5 mg/cm2, depending on the treatment type. Ac = The area of FR-treated fabric within the room. A suite of furniture using 30m2 of fabric was chosen as reasonably large compared with the small room size (see Vr below). µr = The release rate (per unit time) for the FR as respirable particles that are entrained into room air; see the discussion below. Vr = The room volume, chosen as 30 m3 (about 12 ft×11 ft×8 ft) to represent a fairly small room to contain such a suite of furniture. The ratio of fabric area to room volume drives the FR concentration. 

EXPOSURE ASSESSMENT METHODOLOGY 43 Rν = The air exchange rate (air changes per unit time) within the room, chosen as 0.25 air changes per hour, corresponding to the low end of the range of measured values in houses (EPA 1997, Murray and Burmaster 1995). The subcommittee is unaware of any measurements of µ r, the release rate of the FR as respirable particles. It was therefore estimated by considering the possible loss of FR over an average lifetime (about 15 yr) of furniture might amount to 50% in a 25%-fraction of upholstery fabric surfaces (the upper cushion surfaces and the upper backrest of a couch, for example) that might receive heavy wear. Of the worn material, perhaps 1% might be in the form of particles small enough to be considered respirable. Thus, the Release Rate was estimated using Equation 5: (5) where the meanings and values assigned to the symbols are the following: µr = The release rate (per unit time) for the FR as respirable particles that are entrained into room air, calculated using this equation. fl = The fraction (dimensionless) of FR remaining in worn areas after the lifetime of the upholstery (0.50). fw = The fraction (dimensionless) of the upholstery that is relatively heavily worn (0.25). fr = The fraction (dimensionless) of particles released by wear that are respirable (0.01). T = A typical lifetime for upholstery (15 yr). The upholstery is assumed to be replaced after this period, so that exposure f continues for a lifetime. With these assumptions, the release rate is approximately 2.3×10−7/d. With the other parameter values discussed above, the resulting estimated FR particle concentration ranges from 0.8 to 3 µg/m3. Such concentrations may be compared with typical total indoor air concentrations of respirable particles on the order of 100 µg/m3 (EPA 1996). From the average indoor concentration, a time-averaged exposure concentration for a person using the room was estimated using Equation 6: (6) 

EXPOSURE ASSESSMENT METHODOLOGY 44 where the meanings and values assigned to the symbols are the following: Cp,avg = The time-average exposure concentration (mass/unit volume), calculated using the above equation. Cp = The average concentration (mass/unit volume) of FR attached to respirable particles in the room, estimated as already described. fi = The fraction (dimensionless) of time spent in the room containing the upholstery (0.25). For each FR, the time-average exposure concentration was divided by the provisional inhalation RfC derived from the chemical's oral RfD in order to calculate a hazard index. When necessary, the estimated RfC was calculated using Equation 7: (7) where the meanings and values assigned to the symbols are the following: RfC = Inhalation reference concentration (mass/unit volume). RfD = Oral reference dose rate (mass per unit body weight per unit time). Wa = The adult body weight (mass), assumed to be 70 kg. νb = The nominal adult breathing rate (volume/unit time), assumed to be 20 m3/d. If the hazard index for a particular chemical was less than one with the assumptions just described, the subcommittee considered that inhalation exposures to FR particles would be sufficiently small to merit no further examination. For those FRs considered to be possibly carcinogenic, an overestimate of lifetime risk was obtained by multiplying the average dose rate by the carcinogenic potency slope (q1*) for the FR. Vapors In addition to the possibility of release of FR chemicals as particles worn from upholstery fabric, the subcommittee considered the possibility of their release by evaporation. For the ionic chemicals with vapor pressures that are 

EXPOSURE ASSESSMENT METHODOLOGY 45 extremely small, such releases are negligible. The vapor pressures of some of the nonionic FR chemicals, however, are not negligible in their as-applied state, but the chemicals might be in considerably altered form (for example, cross-linked or oxidized, chemically bound to the fabric, or physically adsorbed to the fabric). Therefore, the calculations discussed below, using the vapor pressure of the pure and unmodified chemical, must be carefully interpreted in each individual case and might have to be rejected outright as being so far from the truth as to be useless. The rate of flow of vapor from the room is given by expression 8: (8) and the rate of emission from the upholstery may be estimated by expression 9: (9) where the meanings and values assigned to the symbols are the following: C = The equilibrium vapor concentration (mass/unit volume) within the room (calculated as shown below). Cv = The saturated vapor concentration (mass/unit volume) at room temperature. This varies for each chemical. λ = The saturation fraction (dimensionless) of the chemical in air entering the room (less than or equal to one). For all the FR chemicals, this value is assumed to be zero, because the air entering the room is not expected to be contaminated with the FR vapor. For water vapor (see below for context), a value of 0.7 is used (i.e., 70% relative humidity). Ac = The area of FR-treated fabric within the room. A suite of furniture using 30 m2 of fabric was chosen as reasonably large compared with the small room size discussed below. Vr = The room volume, chosen as 30 m3 (about 12 ft×11 ft×8 ft) to represent a fairly small room to contain such a suite of furniture. The concern is the ratio of fabric area to room volume. Rv = The air exchange rate (number of air changes per unit time) within the room, chosen as 0.25 air changes per hour, corresponding to the low end of the range of measured values in houses. 

EXPOSURE ASSESSMENT METHODOLOGY 46 ξ = The fraction (dimensionless) of surface area of the fabric that is not occluded and so might release vapor (the areas beneath cushions would be occluded, for example) (assumed to be 0.75). Da = The diffusivity (area/unit time) of the chemical vapor in air. For water (see below), the vapor diffusivity in air is approximately 2.2×10−5 m2/sec. Most organic chemicals have vapor diffusivities in air within a factor of two of 5×10 −6 m2/sec, and this value is used in the estimates. d = A boundary layer thickness (length) corresponding to a relatively undisturbed layer of air around the upholstery (estimated as 0.01 m (1 cm), as explained below). γ = A dimensionless factor (less than or equal to one) to account for adsorption of the chemical in, or binding of the chemical to, the fabric or other materials incorporated as part of the FR treatment. For these conservative estimates, this factor is assumed to be one. Equating the rate of flow of vapor from the room and the rate of emission of vapor from the upholstery allows estimation of the equilibrium vapor concentration in room air. It is calculated using Equation 10: (10) where the meanings and values of the symbols are as previously assigned. The time for which this equilibrium vapor concentration could be maintained within the room, before all the FR originally applied to the fabric evaporates may be calculated using Equation 11: (11) where the meanings and values assigned to the symbols not defined immediately above are the following: tv = The time for which the equilibrium vapor concentration could be maintained (computed from the above equation). Sa = The area density of the FR (application rate to the fabric or back-coating—mass per unit surface area). This value is chemical spe 

EXPOSURE ASSESSMENT METHODOLOGY 47 cific, and the highest value likely to be used was chosen. It ranged from 2 to 7.5 mg/cm2 depending on the treatment type. The boundary layer thickness (d) was initially estimated as 1 cm. That value was confirmed approximately by applying the theoretical approach just described to the drying of wet upholstery. With influent air at 70% relative humidity (lamda=0.7) and with the fabric assumed to be wetted at 1.5 kg/m2 (150 mg/cm2), the resultant estimated drying time is about 35 hr if drying is not limited by the air-change rate (as obtained in the equations above by setting η → ). That drying time corresponds within a small factor with common experience. With the other parameter values used, estimates of FR concentration are relatively insensitive to the exact value chosen for the boundary-layer thickness or to the diffusivities of the individual FRs—evaporation is principally limited by the room air-exchange rate. From the equilibrium vapor concentration in room air, the short-term time-average vapor exposure concentration was estimated using Equation 12: (12) where the meanings and values assigned to the symbols are the following: Cs,avg = The short-term time-average vapor exposure concentration (mass/ unit volume) (calculated from the above equation). C = The equilibrium vapor concentration in room air (mass/unit volume) (calculated as described earlier). fi = The fraction (dimensionless) of time spent in the room containing the upholstery (0.25). To calculate a hazard index for each FR, the short-term time-average vapor exposure concentration was divided by the RfC or provisional inhalation RfC derived from the oral RfD as calculated in Equation 7. If the hazard index for a particular chemical was less than one with the assumptions just described, the subcommittee considered that exposures via the vapor inhalation route would be sufficiently small to merit no further examination. For all the FRs examined in this report, the fraction γ used to account for adsorption of the FR to the fabric has been taken to be one, since no better information was available for any of the FRs. In practice, for several of the FRs this assumption leads to results that are plainly implausible, in that the time for which the FR would remain on the fabric would be very limited. Any useful FR obviously has to remain on the fabric for a period of years, so that in such cases 

EXPOSURE ASSESSMENT METHODOLOGY 48 it is clear that the assumption γ=1 is incorrect—the FR is likely to be strongly bound to the fabric, which would have the effect of making γ much smaller. The subcommittee, nevertheless, lists the results of the calculations and points out the situations where it considers the results implausible. Where appropriate, the results are used to indicate that further information on the evaporation potential of the particular FR would be desirable. For each FR, the long-term time-average vapor exposure concentration was estimated from the equilibrium vapor concentration in room air using Equation 13: (13) where the meanings and values assigned to the symbols are the following: Cl, avg = The long-term time-average vapor exposure concentration (mass/ unit volume) (calculated using the above equation). C = The equilibrium vapor concentration in room air (mass/unit volume) (calculated as described earlier). fi = The fraction (dimensionless) of time spent in the room containing the upholstery (0.25). tv = The time for which the equilibrium vapor concentration could be maintained (computed as described earlier). Tf = A typical lifetime for upholstery (15 yr). The upholstery is assumed to be replaced after this period, so exposure continues for a lifetime. For those FRs considered to be possibly carcinogenic, an overestimate of lifetime risk was obtained by multiplying the long-term time-average vapor exposure concentration by the inhalation unit risk for the chemical. If an inhalation unit risk was not available, the unit risk was estimated from the oral carcinogenic potency by using Equation 14: (14) where the meanings and values assigned to the symbols are the following: 

EXPOSURE ASSESSMENT METHODOLOGY 49 U = The estimated unit risk (volume/mass) for the specific FR (calculated using the above equation). q = The oral carcinogenic potency (time times body weight per unit mass) for the specific FR. Wa = The adult body weight (mass) (assumed to be 70 kg). νb = The nominal adult breathing rate (volume/unit time) (assumed to be 20 m3/d). It was pointed out earlier that the estimates for short-term exposure concentrations for the FRs might be substantially erroneous because of lack of information about the binding of the FR to the fabric. For the estimates of lifetime risk presented here, the uncertainty is somewhat smaller, because of the correction introduced to take into account the time for which vapor concentrations could be maintained. The error in short- term estimates comes from an overestimation of the emission rate, but the lifetime risk estimate is independent of the emission rate if the FR completely evaporates within the typical lifetime of the upholstery. Uncertainty in the Inhalation Exposure Estimates There are various uncertainties in inhalation exposure estimates. Given the conservative nature of the estimation procedure, the subcommittee believes that actual exposures are likely to be at least 100-fold lower than calculated. The most conservative assumptions relating to the inhalation exposure scenario are that vapor and particles will be released uniformly for the duration of the exposure and the ratio of room volume to fabric surface area. The most conservative assumptions in the calculation of hazard indices are constant release rate for the life of the fabric, complete absorption from the breath, the assumption that applied chemicals did not react in the fabric, no attempt to limit the exposures by the actual amount of chemical in the fabric, and no attempt to limit the vapor concentration by the actual achievable air concentration. ORAL EXPOSURE SCENARIO The exposure scenario for oral exposure is a child repeatedly sucking on upholstery fabric. It is assumed that a young child might repeatedly suck on the fabric for some fraction of time, wetting a different area with saliva each time it occurs. Such behavior might continue for a couple of years. The limiting 

EXPOSURE ASSESSMENT METHODOLOGY 50 factor is the rate at which the FR dissolves into the saliva. It is assumed that all the dissolved material is ingested by the child. Under those conditions, the dose rate to the child is calculated using Equation 15: (15) where the meanings and values assigned to the symbols are the following: D = The dose rate of chemical (mass per unit body weight per unit time). This rate is the desired value calculated using the formula given. Sa = The area density of the FR (application rate to the fabric or back-coating—mass per unit surface area). This value is chemical specific, and the highest value likely to be used was chosen. It ranged from 2 to 7.5 mg/cm2, depending on the treatment type. Af = The area of fabric sucked on each occasion. The subcommittee estimated that 50 cm2 (about 7.75 square inches) would be a suitable value. µa = The fractional rate (per unit time) of FR extraction by saliva under the given conditions. This rate is chemical specific. It was generally estimated from extraction measurements or laundering tests and ranged from 0.0004 to 0.038 per day. fcc = The fraction (dimensionless) of the time a child sucks FR-treated fabric. The subcommittee considered a suitable estimate to be 1/24 or 1 hr/d. Wc = The body weight of the child, assumed to be 10 kg, which is close to the average weight of 1-yr-old children. The calculated dose rate for each FR was then divided by the oral RfD of that FR to determine a hazard index. If the hazard index was less than one, the subcommittee considered the oral exposure route to be sufficiently small for that FR to merit no further attention. For those FRs considered to be possibly carcinogenic, the lifetime average dose rate was calculated by taking into account the period during which a child might continue sucking behavior. The lifetime average dose rate was calculated using Equation 16: (16) 

EXPOSURE ASSESSMENT METHODOLOGY 51 where the meanings and values assigned to the symbols are the following: Dave = Lifetime average dose rate (mass per unit body weight per unit time) (calculated using the above equation). D = Average dose rate (mass per unit weight per unit time) during the period of exposure (calculated using the previous formula). Ts = Length of time of exposure during childhood (assumed to be 2 yr). TL = Standard length of lifetime used in the definition of the carcinogenic potency calculations (70 yr). An overestimate of lifetime risk was obtained by multiplying the lifetime average dose rate by the carcinogenic potency slope (q1*) for the FR. Uncertainties in the Oral Exposure Estimate There are various uncertainties in oral-exposure estimates. The subcommittee believes that actual exposures are likely to be at least 100-fold lower than calculated, given the conservative nature of the estimation procedure. The most conservative assumptions relating to the oral exposure scenario are the surface area sucked, and that this would occur daily for two years. It is hard to imagine a child actually doing this. The most conservative assumptions in the calculation of hazard indices are constant release rate for life of fabric, complete oral absorption, assumption that applied chemicals did not react in the fabric, and no attempt to limit the exposures by the actual amount of chemical in the fabric. REFERENCES EPA (U.S. Environmental Protection Agency). 1992. Dermal Exposure Assessment: Principles and Applications. EPA/600/8–91–011B. Office of Health and Environmental Assessment, U.S. Environmental Protection Agency, Washington, DC. EPA (U.S. Environmental Protection Agency). 1996. Air Quality Criteria for Particulate Matter, Volume 1 of 3. EPA 600/P-95/001aF. National Center for Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC. EPA (U.S. Environmental Protection Agency). 1997. Exposure Factors Handbook, Volume I—General Factors, Update to Exposure Factors Handbook EPA/600/8–89–043. EPA/600/P-95/002Fa. Office of Research and Development, National Center for Environmental Assessment, U.S. Environmental Protection Agency, Washington, DC. 

EXPOSURE ASSESSMENT METHODOLOGY 52 Grasso, P., and A.B.G.Lansdown. 1972. Methods of measuring, and factors affecting percutaneous absorption. J. Soc. Cosmet. Chem. 23:481–521. Murray, D.M., and D.E.Burmaster. 1995. Residential air exchange rates in the United States: Empirical and estimated parametric distributions by season and climactic region. Risk Anal. 15(4):459–466. Potts, R.O., and R.H.Guy. 1992. Predicting skin permeability. Pharm. Res. 9(5):663–669. Swarbrick, J., G.Lee, J.Brom, and N.P.Gensmantel. 1984. Drug permeation through human skin II: Permeability of ionizable compounds. J. Pharm. Sci. 73(10): 1352–1355.