Toxicology and Industrial Health 2002; 18: 28 ¡/38

www.tihjournal.com

Airborne exposure and soil levels associated with lead abatement of a steel tank John H Lange Envirosafe Training and Consultants, Pittsburgh, PA, USA

This study reports on airborne exposure levels and soil concentrations of lead in regard to abatement of a steel structure (water tank). The tank was de-leaded by abrasive sand blasting. The ball of the tank had a lead surface level that exceed ed the Environmental Protection Agency (EPA) de®nition of lead-based paint (LBP) (0.5% lead), but paint on stem and base was below this criterion. Personal and area airborne samples were collected during different activities of lead abatement of the tank. Summary results suggest during abrasive blasting of ball and stem/base personal exposure levels, as reported with arithmetic and geometric means, exceed the Occu pational Safety and Health Administration (OSH A) permissible exposure limit (50 m g/m3). Highest personal exposure (occupational exposure) was associated with blasting of ball. D istribution of airborne and soil samples suggest non-normality and is best represented by a logarithmic form. Geometric standard deviations for air and soil lead support a non-normal distribution. Outlying values were found for personal and area air samples. Exposure levels associated with blasting stem/base section of tank support OSHA’s policy requiring air monitoring of work at levels below the criterion established by EPA in identifying LBP. Area samples were statistically lower than personal samples associated with blasting ball and stem/base of tank. Exposure data suggest that workers performing abatement on steel structures have elevated lead exposure from surface lead. Respirator protection requirements are discussed. Soil lead concentration was suggested to decrease as distance increased from tank. Soil lead is suggested to be a result of deposition from LBP on tank surface. M inimal efforts were required to reduce average lead soil levels below EPA’s upper accept able criterion (1200 ppm Pb). Toxicology and Industrial Health 2002; 18: 28 ¡/38. Key words: environmental remediation; lead abatement regulations; lead regulations; lead standard; soil lead; occupational engineering controls Abbreviations: AAS, atomic absorption spectroscopy; AL, action level; AM, arithmetic mean; APR, air-purifying respirator; BDL, below detection limit; BLL, blood lead level; EPA, Environmental Protection Agency; GM, geometric mean; GSD, geometric standard deviation; HEPA, high ef®ciency particulate air ®lter; HUD, US Department of Housing and Urban Development; LBP, lead-based paint; lpm, liters per minute; ln, natural logarithm; mg Pb/kg, milligrams of lead per kilogram of soil; ND, not done; NIOSH, National Institute for Occupational Safety and Health; Pb, lead; PEL, permissible exposure limit; PF, protection factor; PPE, personal protective equipment; ppm, parts per million; SAR, supplied air respirator; SF, square feet; OSHA, US Occupational Safety and Health Administration; mg/m3 , micrograms per meter cubed; USDHHS, US Department of Health and Human Services; TCLP, toxicity characteristic leaching procedure; TWA, time-weighted average.

Introduction Address all correspondence to: John H Lange, Envirosafe Training and Consultants, PO Box 114022, Pittsburgh, PA 15239, U SA E-mail: [email protected]

# Arnold 2002

Lead (Pb) has been recognized as an environmental, occupational and public health hazard for 10.1191/0748233702th127oa

Exposure and soil levels associated with lead abatement JH Lange

29

centuries (G ersberg, et al., 1997; Lange, 1991; Rom, 1983; Steenland and Boffetta, 2000; U S D epartment of H ealth and H uman Services ¡ U SD H H S, 1997; Vernon, 1994). G overnmental agencies in the U nited States and other countries have established regulations for abatement of leadbased paint (LBP) from building and structural components (Lange et al., 1998). LBP is applied to steel structures for preventing rust and corrosion. H istorically, little control or concern was associated with removal of LBP (Lange et al., 1998a; Reynolds et al., 1999; Rom, 1983; Spee and Zwennis, 1987; U SD H H S, 1997). Elevated exposure and blood lead levels (BLL) commonly occur in workers performing these activities (Booher, 1988; Conroy et al., 1996; H unt et al., 1998; Langrigan et al., 1982; Rom, 1983; Spee and Zwennis, 1987; Waller et al., 1994; Zedd et al., 1993). It was only when `gross’ symptoms became apparent that preventative action was implemented (U S Occupational Safety and H ealth Administration ¡ OSH A, 1993; Rom, 1983; Waller et al., 1994). M ost preventative measures did not involve exposure reduction, but rather intervention to lower BLL, below which resulted in observed symptoms (Rom, 1983). Today, a greater concern exists for exposure to lead (G idikova, 1998; Reynolds et al., 1999) and it is reported that this metal can effect organ systems at low levels (K aufman et al., 1994; Levin et al., 1997; N agyma et al., 1998; U SD H H S, 1997). Studies (Booher, 1988; Jacobs, 1998; Lange and Thomulka, 2000; Zedd et al., 1993) of occupational exposure during abatement of LBP from steel structures have reported both levels above and below the current OSH A permissible exposure limit (PEL), which is 50 m g/m 3 ¡ time weighted average (TWA) (OSH A, 1993). H owever, there remains a void on information for exposures associated with different work practices (abatement) and abatement of different types of components (U S Environmental Protection Agency ¡ EPA, 1997; Lange, 2001a; Lange and Thomulka, 2000; 2000a; Lange et al., 1993; 1997; 1998a; Rabin et al., 1994). This study collected data on area and personal (occupational) airborne exposure levels during blasting of a water tank (steel structure) that was painted with LBP. Soil samples for lead were also collected before, toward the end of work and after work had been completed. /

/

/

/

Materials and methods Worksite description/practices LBP abatement was conducted in late summer and early fall of 2000. This project involved removal of LBP from outside a steel structure (water tank) that was approximately 140 feet high and was base/stem/ ball shaped (golf ball structure). The site location was the mid-Atlantic region of the U SA and there were no houses or building structures within 150 m of the tank. The tank was empty at the time of work. Lead concentration in paint was 0.2% on the stem/base and 6.9% on the ball. Analysis of lead concentration for air and soil samples was performed by atomic absorption spectroscopy (AAS). Paint chip samples were analyzed using the EPA SW-846. The tank was built and painted about 30 years ago and consisted of a red lead primer and a finish coat. The ball part of the tank had a primer coat, at least on most of the surface, while there were only spot locations with primer on the stem and little if any on the base area of the tank. The paint was `degraded’ and faded from weather and age. The paint was not peeling or chipped, except at the base where about 25% of the surface had cracked paint. M ost of the time on-site involved set-up and preparation for blast activities. The tank had a surface area of about 11 000 square feet (SF ). The thickness of steel for ball and stem/base was estimated to be quarter-inch thick. Abatement was performed by sandblasting (abrasive blasting) (U S D epartment of H ousing and U rban D evelopment ¡ H U D, 1995; Vincent, 1995). A polyethylene plastic containment (about five feet high) was built for blasting the `flat’ part of the top of tank and a containment system attached to a crane was used for the ball part, including the bottom-curved base of the ball. The stem and base were blasted using a spider lift and the lowest part of the base was blasted directly from the ground. N o emissions occurred from the containment on the top of the tank. The contaminant system attached to the crane was constructed with a steel frame and had plexiglass windows with a foam seal on the face. This allowed the containment to be partially `sealed’ to the tank. Sand was collected using a funnel system to hoppers on the ground. This also allowed a draw of air across the face of /

Exposure and soil levels associated with lead abatement JH Lange

30

the containment. When the containment worked properly, only small amounts of dust were released. H owever, wear and damage occurred to the containment as work progressed in days, resulting in increased emissions. Open air blasting was performed when using spiders and working from the ground. The total number of days that abatement was performed was approximately 16. Approximately 20 days were required for setup and preparation for abatement. Abatement was not in continuous days. For example, abatement may occur for a day and the next day only setup would be conducted. Abatement and setup were conducted only on days that had good weather (no rain and temperatures above 458F ). R iver sand and black beauty were used as the blast media. At various times one or the other type of sand was used, while periodically they were mixed. To prevent dust, a dust suppressant agent (transmission fluid) was added to black beauty. Blasting pressure varied from day to day, but was highly effective in removal of paint from surfaces. Rates of work varied from a few hundred SF per day to over 1000 SF per day. Personal protective equipment (PPE) employed included clothing (disposal and non-disposable) and Bullard air supplied equipment (respirator). PPE was used during all blasting, including when performed in spiders and from the ground.

factor (PF ) for a respiratory category or OSH A PEL (Leidel et al., 1977). Standards used in calculating the probability employed PF s of: 10fold for half-mask air-purifying respirator (APR ) (500 m g/m 3), 500-fold (50 000 m g/m 3) for half-mask supplied air respirators (SAR ) and 1000-fold for Bullard helmets (CE respirator or full face SAR ) (100 000 m g/m 3) (OSH A, 1993). PF s represent exposure measurements outside the respirator (Lange and Thomulka, 2000). Samples were collected as area and personal measurements (Lange, 1999). Personal measurements were collected from the breathing zone of the worker (lapel of collar) (Lange, 1999; 2001a; Lange and Thomulka, 2000b; Lange et al., 2000). Area samples were collected during blasting from various locations inside the tank and around the outside of the tank. These samples were collected for determination of non-occupational exposure and/or environmental exposure (Lange, 1999). Two samples were collected inside the tank during blasting and were included in calculations for area samples. M ost days when abatement occurred, one to two personal samples and two to three area samples were collected. F ield blank samples were collected and these consisted of the opening sample cassette for 15 seconds and then resealing (closing).

Air monitoring

Soil samples were collected at approximately 5-, 10-, 15-, 20-, 40-, and 50-foot intervals from the tank. Four sets of samples were collected around the tank, north, east, south and west, for each interval at 5, 10, 15, and 20 feet. Samples were collected before, during and after abatement. After abatement only one set of samples at a five-foot interval was collected. Approximately 50¡ 100 g of soil were collected for each sample and soil was obtained from about the top 2 cm (H U D, 1995). Analysis was performed using AAS employing the EPA method SW-846. Soil sample results were in mg Pb/kg of soil. Statistical comparisons were conducted for samples at 5, 10, 15, and 20 feet distances. The detection limit was set at less than 1 mg/kg. Toxicity characteristic leaching procedure (TCLP) was performed on spent sand for lead (H U D, 1995). This sample was taken from waste material generated while blasting the ball part of the tank.

Air sampling was conducted using a calibrated sample pump with a nominal flow rate of 2 lpm (Lange et al., 1997; 1998a). F low was determined using a calibrated rotometer. Samples were collected closed face using a 37 mm mixed cellulose ester filter cassette (0.8 m m nominal pore size) (Lange et al., 1997). Analysis of samples was performed using the N IOSH 7082 method and a detection limit of 1.0 m g Pb/m 3 was arbitrarily established (Lange and Thomulka, 2000a) ¡ below detection limit (BD L). Exposure results, area and personal, were reported as a TWA in m g Pb/m 3 (Lange and Thomulka, 2000a). Sampling methods are identical to that previously described by Lange and Thomulka (2000a). A graphic method was used to determine the confidence coefficient (probability) that personal (occupational) exposure of at least 5% of samples exceeds the OSH A protection /

Soil monitoring

/

Exposure and soil levels associated with lead abatement JH Lange

31

Data analysis D ata were reported using summary statistics (arithmetic mean ¡ AM , geometric mean ¡ G M , standard deviation, geometric standard deviation ¡ G SD and median) (Lange and Thomulka, 1999; 2000). G M and G SD were determined using natural logarithms (ln) (Lange and Thomulka, 2000; Lange, 2001b). D istribution was determined using the Shapiro ¡ Wilk W-test (Lange and Thomulka, 1999; Shapiro and Wilk, 1965). Outliers were evaluated non-transformed and transformed (ln) using the G rubbs test (G rubbs, 1950; Lange and Thomulka, 2000). Statistical differences (nontransformed data) were evaluated using Wilcoxon Rank Sum test/M ann ¡ Whitney U-test and K ruskal¡ Wallis test (D aniel, 1991). Significance was established as less than 5%. Any value reported BD L were included in calculations at the established detection limit (Lange et al., 1998a). /

/

/

/

/

/

Results Exposure data and soil lead concentrations are shown in Tables 1 and 2. The largest exposure values for area (284 m g/m 3) and personal samples (28 989 m g/m 3) were outliers. Both area and personal exposure samples were non-normally distributed when non-transformed and normally distributed when transformed. N on-normality of area and personal samples is supported by G SD. A statistical difference (Wilcoxon Rank Sum test) exists between area and personal blasting samples (blasting and stem/base ¡ ground work) with and without outliers. D ue to the small sample size for /

Table 1.

a

(blastinga) (blastinga) (blasting from ground b) (in crane) (working on ground c)

N umber of samples

Arithmetic mean

G eometric mean

Standard deviation

G eometric standard deviation

Range

12 11c 3 3 3 30 29

4854.2 2172.6 98.0 2.6 7.5 13.8 4.5

483.0 305.6 56.4 1.9 5.9 3.2 2.7

9470.0 4471.2 97.6 2.5 5.6 51.2 5.7

16.0 12.3 4.4 2.5 2.4 3.5 2.6

3 ¡/28 989 3 ¡/13 851 11 ¡/203 BD L ¡/5.5 3 ¡/13 BD L ¡/284 BD L ¡/22

Blasting for ball part of tank. Blasting stem and base of tank. N ot included in other blasting categories. c While blasting was being performed on ball of tank. d Without outlier. b

/

Summary statistics for air samples, in m g Pb/m3.

Type of sample

Personal Personal Personal Personal Personal Area d Area

other exposure comparisons, statistical analysis was not performed, but a difference exists when evaluated descriptively. The lowest exposure category was for crane operator. Two samples were collected inside the tank (area with mechanical system for pumping water, which is located inside base of tank) and values were approximately 4 m g/m 3 and 6 m g/m 3, respectively. Three field blanks (sampling cassettes) were collected and AM for these were below established detection limit ( B 1 m g/m 3). Exposure results indicate that those blasting on the ball and stem/base on average exceeded the OSH A PEL for airborne lead. Other personal measures and area concentrations on average did not exceed the OSH A action level (AL; 30 m g/m 3). M ost exposure concentratio ns for blasting activities on the ball exceed the OSH A PEL and some samples for stem/base exceed this standard as well. One area sample, highest and outlier (284 m g/m 3), exceeded the OSH A PEL, however, the second highest concentration (22 m g/m 3), which was not an outlier and was below the OSH A AL. When outliers were removed for area and personal sample data, average exposure concentrations are dramatically reduced. After transformation, no sample for area or personal measurements was an outlier. Two samples for blasting exhibited low exposure (3 m g/m 3 and 11 m g/m 3) and neither was an outlier for non-transformed data. Both samples were included in the blasting category since this is the activity that occurred when they were collected. H owever, during most of the time when these samples were collected repair to the blasting machinery and/or containment occurred.

Exposure and soil levels associated with lead abatement JH Lange

32 Table 2.

Summary statistics for soil concentration, in m g Pb/kg, of lead at various distances from base of tank.

Sample collection period and location

Arithmetic mean

G eometric mean

Standard deviation

G eometric standard deviation

Range

Initial, 5 feet Initial, 10 feet Initial, 15 feet Initial, 20 feet Initial, 40 feet a Second, 5 feet Second, 10 feet Second, 15 feet Second, 20 feet b Second, 50 feet c F inal, 5 feet

2085 321 202 191 34 762 451 277 315 150 397

1835 180 190 118 ND 705 239 158 53 [197] 119 286

1026 405 134 238 ND 562 500 293 324 105 379

1.9 3.4 2.0 2.8 ND 2.9 4.0 3.7 16.9 [3.4] 2.5 2.5

751¡/2890 50¡/922 74¡/373 53-547 ND 260¡/1461 75¡/1129 50¡/671 BD L ¡/711 43¡/253 116¡/914

a

Is average of two samples. One sample was below detection limits and not included in average. c Is average of three samples. N D not done. [ ] with three values, not including value which was BD L. b

Containment used for blasting the ball initially had some negative pressure draw, which passed through a H EPA filter. The containment was attached to a crane and pulled spent sand out the bottom and into barrels. The area of the containment was approximately eight-foot wide, seven-foot high, and five-foot deep. Foam rubber wrapped in rubber sheeting was used to `seal’ the containment against the tank. Plexiglass windows were on the top and upper sides of containment and allowed light to enter. Blasting caused some damage, particularly to the plexiglass windows on the sides and leaks began to develop in the containment after a short period of time. Periodically one of these plexiglass windows would break during operations. Visible dust was observed inside the containment when blasting. Probability that at least 5% of employees’ exposure above the PF for CE respirators can not be determined as a result of the large G SD. H owever, the highest concentration did not exceed the OSH A limit for a half-mask SAR . For blasting from ground there is greater than an 80% probability of at least 5% of employees’ exposure exceeding a PF of 10. Exposure to those in the crane and working on the ground while blasting had a probability of exceeding the OSH A PEL by about 25% and 30%, respectively. Soil samples concentrations for lead generally exhibited a higher level near the tank with a concentration decreasing as distance increases from tank (Table 2). A statistical difference exists among initial samples at 5, 10, 15, and 20 feet using the K ruskal¡ Wallis test, but there was no difference /

for the second set of samples (P ¾ 0.42). There were statistical differences (Wilcoxon Rank Sum test) for samples at five feet, initial and second, and initial and final. N o statistical difference exists between samples at 10-, 15-, and 20-foot for initial and second. All initial and second soil samples were non-normally distributed and were normal in distribution when transformed. The highest value in the second set of samples (1461 mg/kg) was an outlier when non-transformed and not an outlier when transformed. There were no outliers for the initial sample set (non-transformed and transformed). TCLP results was 3.6 mg Pb/L. EPA regulations identify a material to be hazardous waste for lead at 5 mg Pb/L (H U D, 1995). /

Discussion This study provides information on airborne levels of lead, personal, area, and soil samples associated with LBP abatement of a steel structure tank (Conroy et al., 1996; Spee and Zwennis, 1987; Zedd et al., 1993). Personal samples were collected for different categories of workers on the same project. G enerally, individual workers did not change tasks, except who blasted on one occasion and then had the remainder of their time either as the crane man or non-blasting ground personnel. LBP abatement of steel structures is becoming common in the U nited States (Forst et al., 1997; H immelstein et al., 1993). Previous studies have suggested elevated exposure occurs to workers

Exposure and soil levels associated with lead abatement JH Lange

33

during abrasive blasting of steel structures with LBP (Conroy et al., 1996; EPA, 1997; Frumkin et al., 1993) and during lead abatement in general (Chambers et al., 1995; Conroy et al., 1996; Forst et al., 1997; G oldberg et al., 1997; Jacobs, 1998; Lange, 2001a,b; Lange and Thomulka, 2000a; Levin et al., 1997; Reynolds et al., 1997; Spee and Zwennis, 1987; Vork et al., 2001). Both air and soil samples were non-normally distributed and appear to better fit a logarithmic distribution. G SD for these data support nonnormality for distribution. Previous studies have also suggested that occupational exposure (Lange and Thomulka, 1999; 2000b; Lange et al., 1997; 1998a; M axim et al., 1997; 2000a) and environmental contaminants (Brown, 2001; Esmen and H ammad, 1977; H erut, 2001; H ill et al., 2001; Lange, 2001a,b; N ayebzadeh et al., 1999) are nonnormal in distribution. Form of distribution is important for understanding exposure and evaluation of summary information (Lange et al., 2000; M axim et al., 1997). Blasting of the ball part of the tank resulted in exposure levels greatly exceeding the OSH A PEL for airborne lead. A single outlier for this category reduced the AM and G M by approximately one half. H owever, there was tremendous variation for these data and this variation is enhanced by two low exposure results (3 m g/m 3 and 11 m g/m 3). Since these values were not outliers, all calculations included these samples. If both values are excluded, AM and G M of air samples for blasting are 6256 m g/m 3 and 1422 m g/m 3, respectively. The highest exposure observed in this investigation is similar to a previous report (Reynolds et al., 1999) of `airborne lead concentrations as high as 29 000 m g/m 3, and suggest that such periodic elevated exposure levels may not be uncommon. Exposure results for blasting of stem and base were above the OSH A PEL, but much lower than for the ball of the tank. These data suggest that exposure levels above the OSH A PEL can result when paint concentrations are below the EPA definition of LBP (0.5% lead) (EPA, 2001) and is supported by previous investigations (EPA, 1997; Zedd et al., 1993). This, in part, supports OSH A’s interpretation of their Lead Standard that air monitoring is required regardless of the lead concentration on surfaces (OSH A, 1999). F urther

support is provided in a previous study where lead exposure was observed to occur even after a steel structure (bridge) had been deleaded (Reynolds et al., 1997). Exposure results in this study exceed lead airborne concentrations reported or inferred in previous abrasive blasting investigations (Conroy et al., 1996; Frumkin et al., 1993). H owever, one `blaster’ in a previous study (Conroy et al., 1996) did have a median airborne lead concentration of 2241 m g/m 3 and a range of 81¡ 4401 m g/m 3, which is not dissimilar to exposures to blasters on top of tank in this study. OSH A, through its lead standard, provided a requirement that abrasive blasting conducted before collection and analysis of air samples be considered to have an airborne exposure concentration in excess of 2500 m g/m 3 (OSH A, 1993). Results for blasting top of tank support the OSH A requirement (29 CF R 1926.62, d, iv) that those conducting abrasive blasting prior to information on exposure levels be protected at a level exceeding 2500 m g/m 3. Respirator protection for abrasive blasting according to OSH A is at a minimum a `half-mask SAR operated in pressure demand or other positive¡ pressure mode’. Other studies of airborne lead levels in the construction and demolitio n industry have reported exposures above the PEL (EPA, 1997; Froines et al., 1990; G oldberg et al., 1997; Lange and Thomulka, 2000a; Lange et al., 1993; Spee and Zwennis, 1987; Waller et al., 1994; Zedd et al., 1993). H owever, excluding two studies of lead exposure during tank (steel structure) demolition (Spee and Zwennis, 1987; Waller et al., 1994), airborne concentrations were descriptively lower than reported in this investigation for blasters on top of the tank. M ost of these studies (Froines et al., 1990; Lange and Thomulka, 2000a) have found an AM airborne lead concentration in the range of around 100¡ 600 m g/m 3. H owever, higher values have been reported (G oldberg et al., 1997; Spee and Zwennis, 1987) and exposure appears to be depend on task, lead concentration in paint, engineering controls, and work practices (Lange and Thomulka, 2000a; Vork et al., 2001). This information supports the requirements of employment of PPE during abatement. G round and crane personnel did not have exposure values that exceed the OSH A AL for /

/

/

Exposure and soil levels associated with lead abatement JH Lange

34

lead. This finding suggests that OSH A’s requirement for employment of PPE based on these exposure data are not applicable to these categories of workers (OSH A, 1999). H owever, these results do not account for all potential routes of exposure to personnel (hand-to-mouth). Area monitoring suggests that average airborne concentrations were below the OSH A AL, and when the single outlier is removed these values approach a suggested exposure standard for outdoor environments (3 m g/m 3) associated with lead abatement (Lange, 1991). M eans, AM and G M , did exceed the N ational Ambient Air Standard (1.5 m g/m 3) established by EPA (EPA, 1984). The outlier sample was collected at the base of tank while blasting was conducted around the base from ground level. This area sample was within 10 feet of blasting during at least part of the sampling time period. Area samples were statistically lower in concentration compared to blasting and stem/base (blasting from ground) personal samples. Studies that compared area and personal samples have reported a similar finding of a statistical difference between sampling methods (Lange, 1999; Lange et al., 2000), with personal samples usually exhibitin g higher exposure levels when compared to area samples (Corn, 1983; 1985; Linch et al., 1970; Scheeper et al., 1995). These data support previous findings that area samples can not be used as a measurement for occupational exposure (Lange, 1999; Lange et al., 2000). These study results suggest that airborne concentration levels for abatement of LBP from tanks can result in exposure well above the OSH A PEL. Exposure results indicate that respirator protection is required for personnel conducting blasting, but not for those assisting in operations. Even when LBP on the tank surface was below the EPA standard of 0.5% lead respiratory protection was required. If averages are examined in regard to respiratory protection levels AM for blasting requires a PF of about 98 and for the G M about 10. Protection factors are determined by the ratio of the exposure level to the PEL (OSH A, 1996). OSH A does not clearly identify which mean should be used, but if calculations are followed in OSH A standards the AM would appear to be inferred (OSH A, 1996). H owever, for averaging sample

results, a court ruling for lead exposure found that the G M is the appropriate summary statistic in evaluating compliance (U nited States Court of Appeals, 1991). Based on the court ruling, G M is legally applicable for summation of exposure concentrations (Letters to the Editor, 1998), however, investigations evaluating the relationship of exposure and disease events, particularly related to occupational chronic diseases, have suggested that AM is the best measure (Armstrong, 1992; Sexias et al., 1988). F undamentally, the most important question in regard to overexposure and risk is whether lead is to be considered a chronic toxicant or an acute toxicant. Such distinction is important when viewing exceedance probabilities of PELs (M axim et al., 2000b). If G M s were considered for personnel conducting blasting, for both with and without outliers, a half-mask APR with H EPA filters would be the minimum OSH A requirement. Even when using this summary statistic day-to-day variation must be considered in selecting appropriate respiratory protection (Lange and Thomulka, 2000a; Spee and Zwennis, 1987). Since the highest exposure value for blasting did not exceed 50 000 m g/m 3 and the AM is above 2500 m g/m 3 a half-mask SAR would be suggested as appropriate based on summary statistics, exposure variation and OSH A’s respiratory selection criteria in the lead standard. Since a `full-face’ Bullard blasting respirator (CE abrasive blasting respirator) was used and is the common form of respiratory protection used in this industry (Conroy et al., 1996), adequate protection would exist (OSH A, 1995). CE respirators, according to OSH A, provide a protection up to 100 000 m g/m 3. These exposure data suggest that use of the common form of respiratory protection in the abrasive blasting industry will provide adequate protection to workers (Conroy et al., 1996). For blasting LBP below the EPA criteria of 0.5% lead exposure results suggest that a half-mask APR respirator will provide adequate respiratory protection when compared to AM , G M and highest exposure concentration. This suggests, as has others (Lange and Thomulka, 2000a; M axim et al., 1997), respiratory requirements will likely vary with functional job category, activities performed, methods employed, and concentration of `contaminate’ encountered.

Exposure and soil levels associated with lead abatement JH Lange

35

Personnel working on the ground and operating the crane while blasting did not exceed the AL. These exposure levels would not require implementation of respiratory protection and use of respirators at these values may be considered inappropriate by some (Lange, 2000; Lange and Thomulka, 2000a). In this case, both probability of overexposure and the maximum exposure level reported support this conclusion. Exposure concentrations for personnel work on ground and in crane appear to be descriptively the same as area samples. Probabilities of exceeding set exposure limits (PF for half-mask APR and PEL) for blasting from ground, those working on ground during blasting and in the crane, were all equal to or exceeded 25%. When the AM , G M and range for these values are examined, descriptive exceedance of set exposure limits (OSH A PEL or AL) did not occur in this study. Reason for probabilities of at least 5% of employees having an exposure over the OSH A PEL is a result of the large variation (G SD ) for these data (Lange and Thomulka, 2000). The lowest variation in this study was a G SD of 2.4, which represents about a 75¡ 80% variation in sample concentration (Leidel et al., 1977). Variation of non-blasting samples observed in this study are similar to that in previous airborne investigations during lead abatement (Conroy et al., 1996; Lange, 2001a,b; Lange and Thomulka, 2000a; Lange et al., 1998a) and are also within the range reported for other occupational exposure studies (Esmen, 1998). Lower variation is usually associated with results that have a large number of values at or near the BD L (Lange and Thomulka, 2000a) or employ highly effective engineering controls (Lange and Thomulka, 2000a). Besides variation associated with occupational and environmental sampling (Esmen and H ammad, 1977), environmental conditio ns for this work, variations in work practices and small number of samples collected would also influence observed sample variation (Lange and Thomulka, 2000a; Lange et al., 1996; 1997; 1998). Extreme variation for personal samples collected during blasting of the top of the tank is a result of the varied conditions, inadequacies in engineering controls, variation in work practice performed, and different materials employed (Lange and Thomulka, 2000a; Lange et al., 1997; 1998a). H owever, /

even with this extremely large variation, there is a good level of respiratory protection afforded by a helmet CE respirator. Average (AM and G M ) soil lead results suggest that levels were below the EPA criterion for nonbare soil (1200 ppm Pb) (EPA 2001; Lange and Thomulka, 1994). Concentration of lead in soil generally decreased through distance from tank (H U D, 1995). These results suggest that lead in soil is a result of rain washing lead off the tank over time creating a gradient. This gradient was statistically different at sampled distances for the initial set of samples. The second sampling set was not statistically different and this may be a result of soil disturbance during work. Initial and second set of soil samples collected were non-normally distributed, but normally distributed when transformed. D uring abatement considerable disturbance resulted to soil surrounding the tank from use of `Bobcats’, boom crane for containment and other light equipment. This caused the sod and soil to be dislodged and redistributed. Sand was also deposited on the surface near the tank during activities. Both disturbance and deposition of sand, spent and non-spent, changed the gradient and diluted soil lead levels. There were statistical differences among initial, second and final soil concentrations at 5 feet. This change in soil lead concentration is a result of disturbance of the soil around the tank. Sand was spilled around the base of tank and some of the soil lead was removed when this sand was `scraped up’ using a Bobcat. Soil disturbance was minimal further away from the tank’s base. EPA established a `maximum’ soil lead concentration of 1200 ppm for non-play areas (EPA, 2001) and 5000 ppm as the definition of this metal constituting a hazardous waste condition (H U D, 1995). Although these values are for residential dwellings and `determination’ of hazardous waste conditions, it does provide a standard for reference purposes. After abatement, second and final soil samples were below both criteria. Since a heavy grass cover exists over the soil, it is unlikely that exposure from this source is of great importance. When soil lead is considered for an abatement project, these results demonstrate the requirement of pre- and post-soil testing. Soil lead concentrations also appear to vary greatly from similar

Exposure and soil levels associated with lead abatement JH Lange

36

locations. Based on this study, minimal efforts were necessary to reduce soil lead around the tank. TCLP results suggest that waste (sand) generated during blasting does not meet the criterion of hazardous waste. Sand used in this abatement would therefore be identified as construction debris.

Conclusion These data support previous investigations that high airborne exposure concentrations can result from abrasive blasting LBP from steel structures (Conroy et al., 1996). Variation for air samples was extremely large for blasting top of tank, but similar to that previously reported for other lead abatement projects (Lange, 2001a,b) and occupational sampling (Lange and Thomulka, 2000; 2000a). Summary occupational airborne lead exposure to personnel blasting stem/base exceed the OSH A PEL even though the LBP concentration was below EPA’s standard of 0.5%. These results support OSH A’s policy of exposure monitoring without regard to the surface lead concentration. Area samples were generally below the OSH A AL, but do exceed on average a proposed airborne lead limit and the N ational Ambient Air Standard for lead. Personal samples (blasting and stem/base) were statistically higher in concentration as compared to area samples. Soil lead samples are suggested to be non-normally distributed and lead concentration decreased with distance from tank. These data suggest that soil lead is primarily a result of LBP being `washed’ off the tank over time. Pre-and postsoil lead samples are warranted for any work on steel structures. M inimal remediation of soil can lower lead levels in soil after abatement. This study provides information on occupational and soil lead levels that can be expected during abatement of steel structures. F urther studies on exposure and soil levels associated with lead abatement of steel structures are warranted.

References Armstrong, B.G. 1992: Con®dence intervals for arithmetic means of lognormality distribution exposures. American Industrial Hygiene Association Journal 53, 481¡/85.

Booher, L.E. 1988: Exposure in a ship overhaul facility during paint removal. American Industrial Hygiene Association Journal 49, 121¡/27. Brown, S.K . 2001: Air toxics in a N ew Australian dwelling over an 8-month period. Indoor-Built Environment 10, 160¡/66. Conroy, L.M ., M enezes-Lindsay, R.M ., Sullivan, P.M ., Cali, S. and Forst, L. 1996. Lead, chromium, and cadmium exposure during abrasive blasting. Archives of Environmental Health 51, 95 ¡/99. Corn, M . 1983: Assessment and control of environmental exposure. Journal of Allergy and Clinical Immunology 72, 231¡/41. Corn, M . 1985: Strategies of air sampling. Scandanavian Journal of Work, Environment and Health 11, 173¡/80. Chambers, C., M orris R ., Salzman, D., Willis, T., Schnitzer, P., Brender, J. and Perrotta, D.M . 1995: Epidemiological notes and reports lead poisoning among sandblasting workers ¡/ G alveston, Texas, M arch 1994. Morbidity and Mortality Weekly Report 44, 44 ¡/45. D aniel, W.W. 1991: Biostatistics: a foundation for analysis in the health sciences. N ew York: Wiley and Sons. U S Environmental Protection Agency. 1984: Health effects assessment for lead . Cincinnati, OH : Of®ce of Health and Environmental Assessment, EPA/540/1-86/055. U S Environmental Protection Agency. 1997: Lead exposure associated with renovation and remodeling activities: environmental ®eld sampling study, Volume I: technical report. EPA 747-R-96-007. Washington, DC. U S Environmental Protection Agency. 2001: Lead; identi®cation of dangerous levels of lead: ®nal rule (40 CFR 745). Federal Register 66: 1205 ¡/ 12401. Washington, D C. Esmen, N.A. and H ammad, Y. 1977: Log-normality of environmental sampling data. Journal of Environmental Science and Health A12, 29 ¡/41. Esmen, N.A. 1998: Exposure estimation in four major epidemiological studies in the acrylonitrile industry. Scandanavian Journal of Work, Environment and Health 24, 63 ¡/70. Forst, L., Persky, V., Freels, S., Williams, R . and Conroy, L. 1997: Lead exposure in ironworkers. American Journal of Industrial Medicine 32, 540¡/43. Froines, J.R., Baron, S., Wegman, D.H . and O’Rourke, S. 1990: Characterization of the airborne concentrations of lead in US Industry. American Journal of Industrial Medicine 18, 1¡/17. Frumkin, H., G err, F., Castaaed a, F., Leal, A. and Santiago, L.R. 1993: Lead poisoning in bridge demolition workers ¡/ Georgia, 1993. Morbidity and Mortality Weekly Report 42, 388¡/90. G ersberg, R .M ., Gaynor, K ., Tenczar, D., Bartzen, M ., Ginsberg, M ., G resham, L.S. and M olgaard, C. 1997: Quantitative modeling of lead exposure from glazed ceramic pottery in childhood lead poisoning cases. International Journal of Environmental Health Research 7, 193¡/ 202. G idikova, P. 1998: Air lead pollution and lead exposure of experimental animals and children in Stara Zagora

Exposure and soil levels associated with lead abatement JH Lange

37 (Bulgaria) International Journal of Environmental Health Research 8, 303¡/13. G oldberg, M ., Levin, S.M., D oucette, J.T. and G rif®n, G. 1997: A task-based approach to assessing lead exposure among ironworkers engaged in bridge rehabilitation. American Journal of Industrial Medicine 31, 310¡/18. G rubbs, F.E. 1950: Sample criteria for testing outlying observations. Annals of Maths and Statistics 21, 27 ¡/58. H erut, B. 2001: Levels of aerosol trace metals along the M editerranean Coast of Israel 1996¡/87. Fresenius Environmental Bulletin 10, 89 ¡/93. H ill, M .T.R ., D imitroulopoulou, C., Ashmore, M .R . and Batty, K . 2001: M easurement and modeling of shortterm variations in particle concentrations in UK homes. Indoor-Built Environment 10, 132¡/37. H immelstein, J., Wolkson, M ., Pansky, G., M orse, D., Ross, H.A. and Gill, J. 1993: Lead poisoning in bridge demolition workers ¡/ Massachusetts. Morbidity and Mortality Weekly Report 38, 687¡/88, 693¡/94. H unt, A., Hawkins, J., G illigan, E. and Bhatia, S. 1998: A comparison of the lead particle content of indoor dust before and after a lead paint abatement: a new source of lead recontamination. Indoor-Built Environment 7, 32 ¡/46. Jacobs, D.E. 1998: Occupational exposures to lead-based paint in structural steel demolition and residential renovations. International Journal of Environmental Pollution 9, 1¡/14. K aufman, J.D., Burt, J. and Silverstein, B. 1994: Occupational lead poisoning: can it be prevented. American Journal of Industrial Medicine 26, 701¡/12. Lange, J.H. 1991: A suggested air standard for lead to protect the public during lead abatement. Medical Hypotheses 36, 211¡/21. Lange, J.H . 1999: A statistical evaluation of asbestos air concentrations. Indoor-Built Environment 8, 293¡/303. Lange, J.H . 2000: H ealth effects of respirator use at low airborne concentrations. Medical Hypotheses 54, 1005¡/ 1007. Lange, J.H . 2001a: Letter to the Editor, regarding Levels of aerosol trace metals along the M editerranean Coast of Israel 1996¡/87. Fresenius Environmental Bulletin 10, 741¡/ 42. Lange, J.H . 2001b: Occu pational exposure during removal of windows with lead-based paint and asbestos caulking. Bulletin of Environmental Contamination and Toxicology 66, 146¡/49. Lange, J.H. and Thomulka, K .W. 1994: Sampling and evaluation of soil for lead contamination: criteria for determining permissible concentrations in soil. Fresenius Environmental Bulletin 3, 751¡/56. Lange, J.H . and Thomulka, K.W. 1999: An evaluation of personal airborne asbestos exposure measurements during abatement of drywall and ¯oor tile/mastic. International Journal of Environmental Health Research 10, 5 ¡/19. Lange, J.H . and Thomulka, K .W. 2000: Occupational exposure during lead abatement of steel surfaces by needle gun methodology. Bulletin of Environmental Contamination and Toxicology 64, 463¡/66.

Lange, J.H . and Thomulka, K.W. 2000a: Occu pational exposure to airborne asbestos during abatement of asbestoscontaining pipe and boiler insulation. Fresenius Environmental Bulletin 9, 477¡/82. Lange, J.H . and Thomulka, K.W. 2000b: Effectiveness of engineering controls for airborne lead exposure during renovation/demolition of a commercial building. IndoorBuilt Environment 9, 207¡/15. Lange, J.H ., G retz, J.B., Wade, J.W. and Thomulka, K.W. 1993: A pilot study of a small-scale lead-based paint abatement project that evaluated a blast machine which utilizes a soft recyclable plastic media: a case report. Fresenius Environmental Bulletin 2, 671¡/76. Lange, J.H ., Lange, P.R., Reinhard, T.K. and Thomulka, K .W. 1996: A study of personal and area airborne asbestos concentrations during asbestos abatement: a statistical evaluation of ®bre concentration data. The Annals of Occupational Hygiene 40, 449¡/66. Lange, J.H., Sites, S.L.M. and Thomulka, K .W. 1997: Personal sample measurements of airborne lead during abatement procedures. Bulletin of Environmental Contamination and Toxicology 58, 659¡/66. Lange, J.H ., Bruce, K.M., Johnson, R .L., Phillips, M .J., Smith, D.M . and Weidenboerner, K.M . 1998: Survey of leadbased paint abatement projects in public building of Erie and Crawford Counties, Pennsylvania, U SA. Regulatory Toxicology and Pharmacology: RTP 28, 73¡/78. Lange, J.H ., Sites, S.L.M. and Thomulka, K .W. 1998a: Airborne lead exposure to workers during activity as part of a building demolition project. Indoor-Built Environment 7, 210¡/16. Lange, J.H., Kuhn, B.D., Thomulka, K .W. and Sites, S.L.M. 2000: A study of matched area and personal airborne asbestos samples: evaluation for relationship and distribution. Indoor-Built Environment 9, 192¡/200. Leidel, N.A., Busch, K.A. and Lynch, J.R. 1977: Occupational Exposure Sampling Strategy Manual. DEH W (NIOSH ) Publication Number 77 ¡/173, N ational Tech nical Information Service N umber PB-274¡/792. Cincinnati, OH : National Institute for Occu pational Safety and H ealth, 1977. Letters to the Editor. 1998: A log-normal distribution-based exposure assessment method for unbalanced bias. The Annals of Occupational Hygiene 42, 413¡/22. Levin, S.M., G oldberg, M . and D oucette, J.T. 1997: The effect of the OSHA lead exposure in construction standard on blood lead levels among ironworkers employed in bridge rehabilitation. American Journal of Industrial Medicine 31, 303¡/309. Linch, A.L., Weist, E.G. and Carter, M .D. 1970: Evaluation of tetraethyl lead exposure by personal monitoring surveys. American Industrial Hygiene Association Journal 31: 170¡/ 79. M axim, L.D., Allshouse, J.N., Kelly, W.P., Walters, T., Waugh, R. 1997: A multiyear workplace-monitoring program for refractory ceramic ®bers: ®ndings and conclusions. Regulatory Toxicological Pharmacology 26, 156¡/71.

Exposure and soil levels associated with lead abatement JH Lange

38 M axim, L.D., Allshouse, J.N. and Venturin, D.E. 2000a: The random-effects model applied to refractory ceramic ®ber data. Regulatory Toxicology and Pharmacology: RTP 32, 190¡/99. M axim, L.D., Allshouse, J.N., Kelly, M .P., Walters, T. and Waugh, R. 2000b: A multiyear workplace-monitoring program for refractory ceramic ®bers: ®ndings and conclusions. Regulatory Toxicology and Pharmacology: RTP 26, 156¡/71. N ayebzadeh, A, Cragg-Elkouh, S, Rancy, R . and Dufresne, A 1999: Source of indoor air contamination on the ground ¯oor of a high-rise commercial building. Indoor-Built Environment 8, 237¡/345 N agyma, L., D esi, I., Schultz, H. and Papp, A. 1998: Consequences of lead exposure of rats during pregnancy, lactation, and poisoning: a combined behavioural and neurotoxicological study. International Journal of Environmental Health Research 8, 121¡/35. U S Occupational Safety and H ealth Administration. 1993: Lead exposure in construction: interim ®nal rule (29 CFR 1926.62). Federal Register 58: 26590-26649. Washington, DC. U S Occupational Safety and H ealth Administration. 1995: Interim interpretation concerning Type-CE respirators used in abrasive blasting that are manufactured by the E.D. Bullard Company, models 77 and 88, OSHA Standards Interpretation and Compliance Letters (8/30/95). Washington, D C. U S Occupational Safety and H ealth Administration. 1996: General Industry requirements, 29 CFR 1910. Washington, DC. U S Occupational Safety and H ealth Administration. 1999: Using X-ray ¯uorescence for analysis of lead in paint and applicability of other agencies lead levels. OSHA Standards Interpretation and Compliance Letters (3/01/99). Washington, D C. Rabin, R ., Brooke, D.R . and Davis, L.K . 1994: Elevated blood lead levels among construction workers in the M assachusetts occupational lead registry. American Journal of Public Health 84, 1483¡/85. Reynolds, S.J., Seem, R ., Fourtes, L.J., Sprince, N.L., Johnson, J., Walkner, L., Clarke, W. and Whitten, P. 1999: Prevalence of elevated blood leads and exposure to lead in construction trades in Iowa and Illinois. American Journal of Industrial Medicine 36, 307¡/16. Reynolds, S.J., F uortes, L.J., G arrels, R .L., Whitten, P. and Sprince, N.L. 1997: Lead poisoning among construction

workers renovating a previously deleaded bridge. American Journal of Industrial Medicine 31, 319¡/23. Rom, W. 1983: Environmental and occupational medicine. Boston, M A: Little, Brown and Company. Scheeper, B., K romhout, H. and Boleij, S.M. 1995: Wood-dust exposure during wood-working processes. The Annals of Occupational Hygiene 39, 141¡/54. Seixas, N.S., Robins, T.G. and M oulton, L.H . 1988: U se of geometric and arithmetic mean exposures in occupational epidemiology. American Journal of Industrial Medicine 14, 465¡/77. Shapiro, S.S. and Wilk, M .B. 1965: An analysis of variance test for normality. Biometrika 52, 591¡/611. Spee, T. and Zwennis, W. 1987: Lead exposure during demolition of steel structure coated with lead-based paints. Scandanavian Journal of Work, Environment and Health 13, 52 ¡/61. Steenland, K . and Boffetta, P. 2000: Lead and cancer in humans: where are we now? American Journal of Industrial Medicine 38, 295¡/299. U nited States Court of Appeals. 1991: American iron and steel Institute and Bethlehem Steel Corporation vs. Occupational Safety and Health Administration. Federal Reporter Second Series (D C Cir.): 939, 975¡/1010. U S Department of H ealth and Human Services (HU D ). 1997: Update: blood lead levels ¡/ U nited States, 1991¡/1994. Morbidity and Mortality Weekly Report 46, 141¡/146. U nited States D epartment of Housing and U rban D evelopment 1995: Guidelines for the evaluation and control of lead-based paint hazards in housing. Washington, DC: United States D epartment of H ousing and U rban Development. Vincent, L.D. 1995: Abrasive blast media additives: a users perspective. Materials Performance 34, 31 ¡/33. Vork, K .L., H ammond, S.K ., Sparer, J. and Cullen, M .R . 2001: Prevention of lead poisoning in construction workers: a new public health approach. American Journal of Industrial Medicine 39, 243¡/53. Waller, K ., Osorio, A.M . and Jones, J. 1994: Lead exposure in a tank demolition crew: implications for the new OSH A construction lead standard. American Journal of Industrial Medicine 26, 693¡/702. Zedd, H .C., Walker, Y.P., H ernandez, J.E. and Thomas, R .J. 1993: Lead exposure during shipboard chipping and grinding paint-removal operations. American Industrial Hygiene Association Journal 54, 392¡/96.