Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: http://www.tandfonline.com/loi/uawm20

Atmospheric Particulate Emissions from Dry Abrasive Blasting Using Coal Slag Bhaskar Kura , Kalpalatha Kambham , Sivaramakrishnan Sangameswaran & Sandhya Potana To cite this article: Bhaskar Kura , Kalpalatha Kambham , Sivaramakrishnan Sangameswaran & Sandhya Potana (2006) Atmospheric Particulate Emissions from Dry Abrasive Blasting Using Coal Slag, Journal of the Air & Waste Management Association, 56:8, 1205-1215, DOI: 10.1080/10473289.2006.10464533 To link to this article: http://dx.doi.org/10.1080/10473289.2006.10464533

Published online: 27 Feb 2012.

Submit your article to this journal

Article views: 349

View related articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uawm20 Download by: [37.44.207.136]

Date: 21 January 2017, At: 07:44

TECHNICAL PAPER

ISSN 1047-3289 J. Air & Waste Manage. Assoc. 56:1205–1215 Copyright 2006 Air & Waste Management Association

Atmospheric Particulate Emissions from Dry Abrasive Blasting Using Coal Slag Bhaskar Kura, Kalpalatha Kambham, and Sivaramakrishnan Sangameswaran Department of Civil and Environmental Engineering, University of New Orleans, New Orleans, LA Sandhya Potana Eco-Systems, Inc., Houston, TX

ABSTRACT Coal slag is one of the widely used abrasives in dry abrasive blasting. Atmospheric emissions from this process include particulate matter (PM) and heavy metals, such as chromium, lead, manganese, nickel. Quantities and characteristics of PM emissions depend on abrasive characteristics and process parameters. Emission factors are key inputs to estimate emissions. Experiments were conducted to study the effect of blast pressure, abrasive feed rate, and initial surface contamination on total PM (TPM) emission factors for coal slag. Rusted and painted mild steel surfaces were used as base plates. Blasting was carried out in an enclosed chamber, and PM was collected from an exhaust duct using U.S. Environment Protection Agency source sampling methods for stationary sources. Results showed that there is significant effect of blast pressure, feed rate, and surface contamination on TPM emissions. Mathematical equations were developed to estimate emission factors in terms of mass of emissions per unit mass of abrasive used, as well as mass of emissions per unit of surface area cleaned. These equations will help industries in estimating PM emissions based on blast pressure and abrasive feed rate. In addition, emissions can be reduced by choosing optimum operating conditions. INTRODUCTION Abrasive blasting is the most widely used method of surface preparation. Shipyards, refineries, automobile industries, and several other industries use abrasive blasting to remove rust, old coatings, and other surface contamination. This process creates clean surfaces with even roughness and, thus, prepares the surface for a new coating. In

IMPLICATIONS This paper presents equations to predict emission factors for uncontrolled TPM applicable for coal slag used as an abrasive in the dry abrasive blasting process. The equations presented will help in assessing the influence of blast pressure and abrasive feed rate on particulate emissions, thus helping to minimize particulate emissions. Industries, regulatory agencies, and scientific groups will be able to use these equations in particulate emissions estimation, air permitting, compliance evaluation, risk assessment, and development of best management practices.

Volume 56 August 2006

the case of dry abrasive blasting, abrasive materials are propelled against a surface with the aid of compressed air. Abrasive materials are widely categorized into metallic abrasives (steel grit, steel shot, etc.), slag abrasives (coal slag, copper slag, etc.), synthetic abrasives (aluminum oxide, silicon carbide, etc.), and natural oxides, such as silica sand.1,2 When bombarded against a metal surface at higher velocities, these abrasive materials breakdown and result in particulate emissions. The amount of particulate matter (PM) emitted depends on the breakdown rate of the abrasive material, as well as important process conditions, such as blast pressure and abrasive feed rate.3,4 Dry abrasive blasting results in both airborne emissions and spent abrasive material. Airborne emissions include particles of various sizes and particulate metals, such as arsenic, cadmium, chromium (trivalent and hexavalent), lead, manganese, nickel, titanium, and others.5– 8 A study conducted to assess the toxicity of spent abrasives showed that metal concentrations can exceed the Toxicity Characteristic Leaching Procedure criteria limits.9 Particulate emissions are of great concern because of the potential health effects, visibility impairment, ecosystem imbalance, and esthetic damage. Inhalation of PM causes respiratory problems, asthma, chronic bronchitis, and decreased lung function.4,10 –16 Thus, study of emissions from this process is important in evaluating both worker exposure and ambient air quality assessments because of these emissions. Emission factors are key inputs to estimate emissions and develop emission inventories. To calculate PM and hazardous air pollutant (HAP) emissions from dry abrasive blasting, AP-42 (U.S. Environmental Protection Agency [EPA]) has documented emission factors for various abrasive materials. This document contains candidate emission factors for total PM (TPM), coarse PM (PM10), fine PM (PM2.5), and metals obtained from different research resources based on surface type and wind speed for sand and garnet. Because the data in the AP-42 document was collected from different resources, test conditions and monitoring methods varied widely. Lack of documentation and process conditions resulted in poor data quality in some cases.2 Various state agencies reported general emission factors for TPM, PM10, and PM2.5 for garnet, coal slag, sand, metallic grits, and mineral slags but did not specify the process conditions or surface type.17–20 National Shipbuilding Research Program (NSRP)21 sponsored Journal of the Air & Waste Management Association 1205

Kura, Kambham, Sangameswaran, and Potana a study that evaluated some of the most commonly used abrasives at blast pressures of 80 and 122 pounds per square inch (psi) for PM1 (⬍1 ␮m in diameter), PM2.5, PM4 (⬍4 ␮m in diameter), and PM10. Although this document provides the emission factors for six abrasives, these emissions were estimated based on a mass-balance method, and limitations were acknowledged. Also, emission factors are not available for any intermediate pressures. A survey of U.S. shipyards showed that coal slag is a predominantly used abrasive on the East Coast and the Gulf of Mexico (JPCL 2000).22 Blasting with coal slag results in dust emissions and may result in HAP emissions based on contaminants present in the abrasive and on the metal surface.9,21,23–30 Some studies have shown that coal slag has higher emissions than other commonly used abrasives in all of the size ranges. In addition, these emissions are more toxic and can cause greater pulmonary damage and inflammation.24 –26 In view of the importance of being able to quantify abrasive blasting emissions from the use of coal slag, University of New Orleans (UNO) has initiated research to quantify TPM emissions.31 The important objective of this research was to develop TPM emission estimation equations for coal slag based on blast pressure and abrasive feed rate. This paper presents the experimental methodology and variation of TPM emission factors with blast pressure and feed rate using coal slag for two types of surface contamination, flash rust and marine paint. Emission estimation equations developed to calculate emission factors at intermediate blast pressure and feed rate are also presented in this paper. These equations will be useful in calculating the emissions more accurately under specific operating conditions. In addition, by carefully selecting optimum blast pressure and abrasive feed rates, industries can minimize PM emissions, air pollution control costs, and environmental impacts. In general, abrasive blasting emission factors are expressed in mass of emissions per unit mass of abrasive used.2 The state of Texas reported emission factors for coal slag being 0.00286 and 0.00034 (lb emissions/lb usage) for TPM and PM10, respectively.17 Thus, emission factors in this study are calculated as kilograms of TPM emitted per kilogram of coal slag used (kg/kg). In addition, to determine the emissions based on area cleaned, emission factors were calculated in terms of kilograms of TPM emitted per square meter of area cleaned (kg/m2). The feed rate is expressed in both units, “number of turns” and “kilograms per hour.” Although the mass flow rate occurring with each number of turns is not precisely reproducible, currently industries do not have any mechanism to measure kilograms per hour, because the mass flow meters are not yet in use. Most shipyards currently use Schmidt abrasive feed valve. Thus, expressing abrasive feed rate in number of turns helps in applying the research results in industries. Significance of Abrasive Feed Rate and Blast Pressure Controlling the abrasive feed rate is important to maintain uniform flow of material and blast pattern. It is also important to achieve maximum productivity (area 1206 Journal of the Air & Waste Management Association

cleaned per unit time) while consuming less abrasive material. Low feed rates may result in lower emissions but take a longer time to clean and decrease productivity. At high feed rates, the particles traveling from the nozzle interact with the rebounding particles from the base plate and breakdown resulting in higher emissions. In addition, high feed rates may result in turbulent flow of material thus creating uneven surface profile, decreasing productivity, and increasing consumption.10,32–34 Blast pressure determines the velocity of abrasive particles. Kinetic energy acquired by the particles is proportional to mV2 (“m” is mass and “V” is the velocity of particle), which is responsible for removal of rust, paint, or other surface contaminants in addition to creating a rough surface profile. The high energy action of particles on the surface results in breakdown of larger particles into smaller particles and they become airborne. In addition to abrasive particles, rust and paint chips also break down and become airborne. Friability of abrasive particles influences the breakdown rate of abrasives and, thus, influences particle emissions.10,32–34 EXPERIMENTAL WORK Material Description Coal slag is a residue produced from the combustion of coal in coal-fired utility boilers. The molten slag from the combustion of coal is quenched in water resulting in an amorphous, noncrystalline particulate. This rapid cooling also breaks the coal slag into rough angular particles, which are then separated into various particle size grades using screens. The quality of this material is often improved by crushing and screening followed by magnetic separation. Some of the other characteristics of coal slag, such as hardness, uniform density, low friability, and low free silica content, are useful in removing heavy rust and proving a high-profile finished surface. Commercially known as Black Beauty or Black Diamond, coal slag is, thus, used by many industries for dry abrasive blasting. This abrasive may contain high levels of heavy metals, such as arsenic, beryllium, chromium, nickel, and lead, as well as iron and aluminum.23–30 Coal slag gradation (abrasive particle size), contamination level, and type of coal used in producing coal slag are expected to influence atmospheric particulate emissions during blasting Emissions Test Facility Description A layout of the emissions test facility (ETF) constructed at UNO is shown in Figure 1. This facility is equipped with the following: (1) blasting equipment, (2) test plates; (3) test chamber; (4) exhaust duct and particulate collection system, and (5) stack sampling system. The blasting equipment consisted of a 273-kg (600 lb) capacity Abec blast pot with Schmidt feed valve at the bottom of the pot to control the abrasive feed rate, air compressors (Sullair Model 375 Hr and Ingersoll Rand), and fitted pressure gauges to supply compressed air. A secondary air supply unit to provide air to the blaster, a Bazooka no. 6 blast nozzle, and rubber hoses were used. Moisture traps were also used to eliminate the moisture in the air supplied. Mild steel test plates of 2.5 ⫻ 1.5 m (8 ⫻ 5 ft) were used as base plates. Two types of surfaces were used in this study: rusted panels and painted panels. The Volume 56 August 2006

Kura, Kambham, Sangameswaran, and Potana

Figure 1. ETF at UNO.

steel plates were sprayed with water and allowed to develop flash rust. For the painted panels, a commercially available Rust Oleum Safety Yellow paint and thinner were used. The average thickness of the coating was estimated to be 0.73 mil (1 mil ⫽ 0.001 in. ⫽ 25.4 ␮m), assuming average transfer efficiency of 50%. Abrasive blasting experiments were carried out in an enclosed chamber of 3.7 ⫻ 3 ⫻ 2.5 m (12 ⫻ 10 ⫻ 8 ft). The emissions from the process were vented through a horizontal duct using a fan with 60-rpm capacity. The duct was 0.31 m (1 ft) in diameter, and average volumetric flow rate produced by the fan was 85 m3/min (3000 cfm). Thus, an average linear velocity of 9.5 m/min was maintained within the test chamber. To collect emissions effectively and control their discharge into the atmosphere, a two-stage particle collection system was installed downstream of the fan. In the first stage, the coarse particles were collected as a result of changing the direction of flow, and in the second stage, fine particles were collected using fabric filters. Velocity measurement and emissions sampling was performed as specified by 40 CFR Part 60 Appendix A Test Methods 1–5.35–39 The sampling train consisted of a heated probe (with a nozzle, Type S Pitot tube, and differential pressure gauge), temperature sensor, filter holder, four glass impingers, ice bath container with crushed ice, and dry gas meter.35–39 The method of sampling is briefly described in the experimental procedure. A detailed description of these methods can be obtained from EPA website at http://www.epa.gov/ttn/emc. Experimental Parameters Atmospheric emissions of PM from dry abrasive blasting are influenced by a number of parameters; the most important ones include: (1) abrasive type; (2) abrasive gradation; (3) type of base plate (mild steel, aluminum, concrete, etc.); (4) type and level of initial surface contamination; (5) final surface finish desired; (6) blast Volume 56 August 2006

pressure; (7) abrasive feed rate; (8) blast nozzle size; (9) angle of blast nozzle; (10) distance between the blast nozzle tip and the surface; (11) dwell time (average time spent by abrasive jet on the surface); (12) wind velocity in case of outdoor blasting operations; (13) blast room size, orientation of the surface being cleaned, and the ventilation rate in case of indoor blasting operations; and (14) the number of reuses of abrasive.4,21,32,40 In view of the significance of blast pressure and abrasive feed rate, the research was focused on studying the effects of these parameters on particulate emissions. From the literature, it was noted that industries commonly use blast pressures in the range of 80 to 120 psi. Thus, to estimate emission factors as a function of blast pressure and feed rate, three blast pressures (80, 100, and 120 psi) and three feed rate conditions (three, four, and five turns of opening of Schmidt valve) were selected. To study the effect of surface contamination, rusted mild steel panels (with flash rust) and painted mild steel panels were used as base plates. For each operating condition, three test runs were performed to assess the repeatability of the test runs. Therefore, nine combinations of operating conditions (three pressures ⫻ three feed rates) were studied for each surface type in this research. Each of these nine testing conditions was repeated three times, which resulted in 27 total runs for rusted panels and 27 total runs for painted panels. A medium-grade coal slag was used in this study. Experimental Procedures Abrasive blasting was carried out in an enclosed test chamber using medium-grade coal slag to remove surface contaminants from mild steel test plates. As mentioned earlier, atmospheric PM emissions were sampled as per EPA standard methods, Methods 1–5 for stack sampling.35–39 Because the exhaust duct was circular and 0.31 m (12 in.) in diameter, eight traverse points were selected and marked on the sampling probe. A sampling Journal of the Air & Waste Management Association 1207

Kura, Kambham, Sangameswaran, and Potana

Figure 2. Variation of TPM emission factors with feed rate (number of turns).

port was located upstream of the exhaust fan to minimize flow disturbances and maintain isokinetic flow conditions. A nozzle with i.d. of 4.57 mm (0.018 in.) was used to maintain isokinetic flow conditions during sampling. A measured quantity of coal slag was added to the blast pot, 1208 Journal of the Air & Waste Management Association

and blast pressure and feed valve were then set to desired operating conditions (e.g., 80 psi and four turns). Leak checks were performed once the sampling train was assembled. As the blasting was carried out in the chamber, samples were drawn isokinetically through the sampling Volume 56 August 2006

Kura, Kambham, Sangameswaran, and Potana

Figure 3. Variation of TPM emission factors with pressure (psi).

probe. The probe was sufficiently heated to prevent water condensation. The particles were collected on a preweighed Whatman No. 10 filter paper that was already desiccated to eliminate moisture. The filter assembly box was maintained at a temperature of 120 ⫾ 15 °C to collect PM from the sample gas stream while preventing moisture condensation. Velocity head and temperature at the Volume 56 August 2006

eight traverse points were measured with a Type S Pitot tube. Static pressure in the stack, dry molecular weight of stack gas, sample volume, atmospheric pressure, and temperature were recorded as per EPA guidelines. A series of four impingers was used to collect the moisture from the sampled gas. The first two impingers were filled with 100 mL of water, the third impinger was left empty, and the Journal of the Air & Waste Management Association 1209

Kura, Kambham, Sangameswaran, and Potana

Figure 4. TPM emission factor trends with feed rate (kg/hr).

fourth impinger was filled with 200 –300 g of silica gel. After each test run, the impingers were weighed, and increase in the weight was recorded to calculate moisture content of the stack gas. At each traverse point, the exhaust gas sample was collected for 2 min. Thus, total sampling time was 16 min for each of the test runs. 1210 Journal of the Air & Waste Management Association

Once the sample was collected, the filter paper was removed from the filter box and placed in a desiccator. PM from the probe was collected in a beaker by rinsing the probe with acetone. The final weights of PM in filter paper and beaker were recorded after desiccating them for 24 hr. The sum of the weights of both containers was used Volume 56 August 2006

Kura, Kambham, Sangameswaran, and Potana Table 1. Coefficients for EF estimation.

EF Units Rusted Panelsa kg/m2 kg/kg Painted Panelsa kg/m2 kg/kg Rusted Panelsa kg/m2 kg/kg Painted Panelsa kg/m2 kg/kg

Feed Rate Units

a

b

c

d

e

f

R2

No. of turns No. of turns

1.78E ⫹ 00 ⫺3.85E-01

⫺2.91E-02 3.88E-03

9.07E-01 1.61E-01

1.85E-04 ⫺2.72E-05

⫺1.74E-01 ⫺2.28E-02

7.19E-03 3.78E-04

0.57 0.50

No. of turns No. of turns

9.61E ⫹ 00 2.38E-01

7.18E-02 ⫺2.20E-03

⫺6.58E ⫹ 00 ⫺5.37E-02

⫺1.75E-04 1.46E-05

8.90E-01 7.32E-03

⫺2.67E-03 ⫺2.47E-05

0.97 0.94

kg/hr kg/hr

⫺6.56E ⫹ 00 ⫺3.14E-01

2.37E-01 9.77E-03

⫺1.44E-02 3.33E-05

⫺1.06E-03 ⫺4.21E-05

1.87E-05 6.30E-07

1.09E-05 ⫺6.07E-06

0.51 0.62

kg/hr kg/hr

1.57E ⫹ 00 1.17E-01

2.55E-02 ⫺1.12E-03

⫺8.66E-03 ⫺1.45E-04

⫺5.05E-04 ⫺4.41E-07

⫺1.14E-05 ⫺3.16E-07

2.14E-04 4.26E-06

0.58 0.85

Notes: aCoefficients for blasting of this panel type.

to determine the actual mass of emissions in the stack gas. The base plate was cleaned to obtain an SP10 or nearwhite finish. The area cleaned was measured using a measuring tape with appropriate approximations for nonquadrilateral geometries, and blasting time was recorded using a stopwatch. For each combination of pressure and feed rate, three test runs were carried out to minimize experimental errors and test the repeatability of the results. Because the samples were collected upstream of the exhaust fan and two-stage particulate collection system, results presented in this paper correspond with uncontrolled emission factors. RESULTS Variation of Emission Factors with Feed Rate (number of turns) To quantify the effect of feed rate on TPM emissions at a given pressure, emission factors were plotted against feed rate, in terms of number of turns. Figure 2 shows the variation of emission factors (in both kg/m2 and kg/kg) with feed rate for the rusted panels, as well as painted panels. In general, emission factors were higher for rusted panels when compared with painted panels. In the case of rusted panels, higher emission factors were mostly observed at four turns as compared with three and five turns of opening of the Schmidt valve. However, at 120 psi, emissions per square foot were lower at four turns than three and five turns. In the case of painted panels, minimum emission factors were observed at four turns for all of the blast pressures. This may be because of more uniform flow of material at four turns as compared with three and five turns. In addition, the effect of rebounding particles at five turns resulted in higher emission factors (kg/m2), which can be observed in Figure 2 for painted panels. Higher correlation coefficients (R2) were obtained for painted panels than for rusted panels. Except for emission factors in kg/kg at 80 and 100 psi, significantly higher R2 values (ⱖ0.92) were observed for painted panels. In the case of rusted panels, R2 value varied between 0.30 and 0.74. Polynomial equations were also developed to estimate emission factors for intermediate operating conditions. Volume 56 August 2006

Variation of Emission Factors with Pressure (psi) Figure 3 demonstrates variation of TPM emission factors for coal slag with blast pressure. Emission factors increased with pressure in most cases for both rusted and painted panels. Because at high velocity the abrasive particles suffered more damage, higher pressure resulted in increased emissions. This phenomenon was observed in Figure 3 for the rusted and painted panels at four and five turns (both in kg/m2 and kg/kg). However, in the case of rusted panels, at three turns, emission factors (in kg/kg) at 120 psi were lower than those at 100 psi. Also, lower emission factors (in kg/m2) were observed at 100 psi than at 80 and 120 psi. To estimate emissions at intermediate blast pressures, polynomial equations were developed, and these are shown in Figure 3 for each feed rate condition. Weakest correlation was observed at five turns for emission factors (in kg/kg) for rusted panels (0.08). Emission factors in terms mass of TPM per unit area cleaned depicted for painted panels were strongly correlated with pressure at all of the feed rate conditions (R2 ⱖ0.98). Although emission factors (in kg/kg) increased with pressure at all of the feed rates for painted panels, correlation coefficient ranged between 0.63 and 0.95. Variation of Emission Factors with Feed Rate (kg/hr) Mass flow rate of the abrasive depends on blast pressure and feed valve opening. Because different feed valves result in different feed rates, an attempt was made to determine emissions at varying abrasive feed rates. Because there was no means of directly measuring or monitoring the flow rate, abrasive feed rate (in terms of kg/hr) was calculated from the mass of abrasive used and actual blast time. Worker weariness and differences in worker to worker could have affected the blasting time, and because three persons carried out the blasting operations, calculated feed rate (kg/hr) varied significantly. Thus, poor correlations were obtained for emission factors in some cases for both rusted and painted panels. Figure 4 depicts the variation of emission factors (kg/m2 and kg/kg) with feed rate in terms of kg/hr for both rusted and painted panels. Although variation of emission factors (kg/m2 and kg/kg) with feed rate was not significant at 100 psi, lower Journal of the Air & Waste Management Association 1211

Kura, Kambham, Sangameswaran, and Potana Table 2. Calculated vs. measured EFs. Surface Type Rusted

EF Units kg/m2

kg/kg

Painted

kg/m2

kg/kg

Pressure (psi)

Feed Rate No. of Turns

Calculated EFs Using Eq 1

Measured EFs from Experimental Results

80 80 80 100 100 100 120 120 120 80 80 80 100 100 100 120 120 120 80 80 80 100 100 100 120 120 120 80 80 80 100 100 100 120 120 120

3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5

3.52 3.78 3.69 4.03 4.44 4.49 4.69 5.24 5.44 0.120 0.151 0.137 0.122 0.161 0.155 0.103 0.150 0.151 1.86 1.29 2.51 2.50 1.89 3.05 3.01 2.34 3.45 0.054 0.050 0.060 0.061 0.056 0.066 0.080 0.074 0.083

2.44–4.31 3.64–3.87 3.46–3.86 3.72–5.61 4.38–5.19 4.22–4.53 4.53–5.35 3.86–5.76 4.98–6.31 0.093–0.137 0.138–0.155 0.142–0.153 0.105–0.158 0.150–0.175 0.133–0.151 0.060–0.128 0.146–0.177 0.129–0.179 1.85–1.92 1.31–1.37 2.41–2.49 2.32–2.40 1.89–2.02 3.01–3.19 2.98–3.22 2.08–2.13 3.46–3.50 0.051–0.059 0.044–0.053 0.059–0.060 0.059–0.063 0.043–0.061 0.062–0.065 0.079–0.080 0.074–0.075 0.081–0.086

emission factors (kg/m2) were observed at intermediate feed rates for 80 and 120 psi for painted panels. Emission Estimation Equations and Their Application The polynomial equations presented in Figures 2– 4 depict either effect of pressure on emission factors at a given feed rate or effect of feed rate at a specific pressure. To study simultaneous effects of both, pressure and feed rate, on emission factors, three-dimensional equations were developed. Datafit curve fitting application was used for regression analysis and for plotting three-dimensional charts. Equations were chosen based on the physical phenomenon and the best R2 values. To evaluate the statistical significance of these regression models, two-way analysis of variance was used. F static was calculated for H0: a ⫽ 0, b ⫽ 0, c ⫽ 0, d ⫽ 0, e ⫽ 0, and f ⫽ 0, where a, b, c, d, e, and f are coefficients in the regression model. Null hypothesis (H0) was rejected if Prob(F) ⬍ 0.025. In addition, t static was calculated for H0: a ⫽ 0, b ⫽ 0, c ⫽ 0, d ⫽ 0, e ⫽ 0, or f ⫽ 0 to test the significance of each coefficient. Null hypothesis was rejected if Prob(t) ⬍ 0.05. Equation 1 1212 Journal of the Air & Waste Management Association

represents emission factors (EF) for uncontrolled TPM as a function of pressure and feed rate. Coefficients for this equation are given in Table 1. EF ⫽ a ⫹ 共b ⴱ P兲 ⫹ 共c ⴱ F兲 ⫹ 共d ⴱ P 2 兲 ⫹ 共e ⴱ F 2 兲 ⫹ 共f ⴱ P ⴱ F兲

(1)

where EF is emission factor in kg/m2 or kg/kg (given in Table 1); P is blast pressure (psi; applicable range: 80 –120 psi); F is feed rate (applicable range: three to five turns or 100 –700 kg/hr); and a, b, c, d, e, and f are coefficients to be read from Table 1. EFs within the tested ranges of pressure and feed rate (number of turns and kg/hr) can be estimated easily using eq 1. For example, in the case of a painted panel with a blast pressure of 85 psi and abrasive feed rate of 3.5 turns on Schmidt feed valve, eq 1 gives uncontrolled TPM emissions of 1.52 kg/m2. Similarly, for rusted panel involving a blast pressure of 90 psi and abrasive feed rate of 4.5 turns on Schmidt feed valve, use of eq 1 gives uncontrolled TPM emissions of 4.12 kg/m2. Table 2 shows calculated EFs Volume 56 August 2006

Kura, Kambham, Sangameswaran, and Potana

Figure 5. Variation of TPM emission factors with pressure and feed rate.

using eq 1 and those measured in the field through stack testing procedures for all blast pressures and abrasive feed rates (number of turns). Volume 56 August 2006

Variation of EFs with Pressure and Feed Rate Figure 5 illustrates EFs as a function of both pressure and feed rate (number of turns as well as kg/hr) for both rusted Journal of the Air & Waste Management Association 1213

Kura, Kambham, Sangameswaran, and Potana and painted panels. EFs for painted panels showed significantly stronger correlations with pressure and feed rate (number of turns) as compared with rusted panels. Although reading EFs from these plots is a little cumbersome, the trends and simultaneous effect of varying pressure and feed rate can be clearly observed. This figure will also help in identifying blast pressure and feed rate for obtaining minimum EFs. For example, Figure 5c shows the variation of emissions factors (kg/m2) with blast pressure and feed rate (number of turns) for painted panels. It can be observed that EFs increased with increase in blast pressure, and at a given blast pressure, a “U” shape variation can be observed with increase in feed rate. Minimum EF can be observed at 80 psi and four turns of feed rate.

3.

4. 5. 6. 7. 8. 9.

CONCLUSIONS The TPM emission estimation equations developed in this research are applicable to dry abrasive blasting using medium-grade coal slag on mild steel with flash rust and paint. Uncontrolled TPM emissions were measured using EPA stack testing procedures, and equations were developed as a function of blast pressure and abrasive feed rate. Uncontrolled TPM emitted can be calculated using the blast pressure and abrasive feed rate either in terms of area cleaned (kg of PM/m2 area cleaned) or mass of abrasive used (kg of PM/kg of coal slag consumed). Because the emissions were measured upstream of the control equipment, the EFs given by the equations correspond with uncontrolled atmospheric emissions. The equations assist in determining the best process conditions (blast pressure and feed rate) to lower the atmospheric PM emissions from blasting operations using coal slag. In addition, these equations can be used to calculate accurate emissions under intermediate blast pressure and feed rate conditions and determine compliance with applicable state and federal emission standards. Individual industries, shipyards performing dry abrasive blasting, in particular, and the NSRP that assists shipyards comply with environmental regulations will benefit from the developed equations because they will help them identify the optimum process conditions to achieve lowest TPM emissions. The equations presented in this paper are applicable within the tested ranges of blast pressure and abrasive feed rate. EFs were found to be higher for rusted panels than painted panels for the surfaces used in this study. Further research is recommended to evaluate size distribution and metal analysis of PM emitted from dry abrasive blasting using coal slag.

10. 11. 12. 13. 14. 15.

16. 17. 18. 19.

20.

21. 22. 23. 24.

ACKNOWLEDGMENTS The authors wish to acknowledge the support and funding from the Gulf Coast Region Maritime Technology Center, the Office of Naval Research, U.S. Environment Protection Agency (Region VI), and the Maritime Environmental Research and Information Center. REFERENCES 1. U.S. Army Corps of Engineers. Painting: New Construction and Maintenance, Chapter 7. Surface Preparation; EM 1110-2-3400; U.S. Army Corps of Engineers: Washington, DC, 1995, pp 4-15. 2. U.S. Environmental Protection Agency. Emission Factor Documentation for AP-42 Section 13.2.6 Abrasive Blasting Final Report; EPA-Contract 1214 Journal of the Air & Waste Management Association

25.

26. 27.

28.

68-D2-0159; Work Assignment No. 4-02; MRI Project No. 4604-0; U.S. Environmental Protection Agency: Washington, DC, 1997. The National Shipbuilding Research Program. Ultra-High Pressure Water Blasting Project; Final Project Results Summary; Maritech-ASE; Agreement No: 2000932; The National Shipbuilding Research (NSRP)Program: Charleston, SC, 2002. U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards. Health and Environmental Impacts of PM. Available at: http://www.epa.gov/air/urbanair/pm/hlth1.html (accessed 2005). National Institute for Occupational Safety and Health. Evaluation of Substitute Materials for Silica Sand in Abrasive Blasting; National Institute for Occupational Safety and Health: Atlanta, GA, 1998. MacKay, G.R.; Stettler, L.E.; Kommineni, C.; Donaldson, H.J. Fibrogenic Potential of Slags Used as Substitutes for Sand in Abrasive Blasting Operations; Am. Ind. Hyg. Assoc. J. 1980, 41, 836-842. Vallyathan, V.; Blake, T.; Leonard, S.; Greskevitch, M.; Jones, W.; Pack, D. In Vitro Toxicity of Silica Substitutes Used for Abrasive Blasting; Am. J. Ind. Med. Suppl. 1999, 1, 158-160. Conroy, L.M.; Menezes-Lindsay, R.M.; Sullivan, P.M.; Cali, S.; Forst, L. Lead, Chromium, and Cadmium Exposure during Abrasive Blasting; Arch. Environ. Health 1996, 51, 95-99. Obery, A.M.; Landis, W.G. Assessment of Environmental Toxicity from Spent Recycled Glass Abrasives Part A; Data Review; Institute of Environmental Toxicology and Chemistry, Western Washington University: Bellingham, WA, 1999. Hansink, J.D. Minimizing Dust during Abrasive Blast Cleaning; J. Protect. Coat. Lining 1995, 12, 18-19. U.S. Environmental Protection Agency. Particulate Matter Research Program Five Years of Progress; EPA 600/R-04/058; U.S. Environmental Protection Agency: Washington, DC, 2004. U.S. Environmental Protection Agency. Air Quality Criteria for Particulate Matter; National Center for Environmental Assessment: Research Triangle Park, NC, 2004. Daigle, C.C; Chalupa, D.C.; Gibb, F.R.; Morrow, P.E.; Oberdorster, G.; Utell, M.J.; Frampton, M.W. Ultrafine Particle Deposition in Humans during Rest and Exercise; Inhal. Toxicol. 2003, 15, 539-552. Wilson, F.J., Jr.; Hiller, F.C.; Wilson, J.D.; Bone, R.C. Quantitative Deposition of Ultrafine Stable Particles in the Human Respiratory Tract; J. Appl. Physiology 1985, 58, 223-229. Voutilainen, A.; Kaipio, J.P.; Pekkanen, J.; Timonen, K.L.; Ruuskanen, J. Theoretical Analysis of the Influence of Aerosol Size Distribution and Physical Activity on Particle Deposition Pattern in Human Lungs; Scandinavian J. Work Environ. Health 2004, 30(Suppl. 2), 73-79. Oberdorster, G. Lung Particle Overload: Implications for Occupational Exposures to Particles; Reg. Toxicol. Pharmacol. 1995, 21, 123-135. Texas Natural Resource Conservation Commission. Abrasive Blast Cleaning; Texas Natural Resource Conservation Commission Air Permits Division: Austin, TX, 2001. Bay Area Air Quality Management District. Sandblasting, Miscellaneous Standards for Performance; Regulation 12, Rule 2; Bay Area Air Quality Management District: San Francisco, CA, 1990. South Coast Air Quality Management District. Unconfined Abrasive Blasting, South Coast Air Quality Management District Permit Processing Handbook; Section 2; South Coast Air Quality Management District: Diamond Bar, CA, 1989. County of San Diego Air Pollution Control District. Abrasive Blasting Procedures, Emission Factors; The County of San Diego Air Pollution Control District: San Diego, CA. Available at: http://www.sdapcd.org/ toxics/emissions/ablast/ablast.html (accessed 2005). National Shipbuilding Research Program. Particulate Emission Factors for Blasting Operations and Other Potential Sources; NSRP 0552; N1-97-4; National Shipbuilding Research Program: Charleston, SC, 1999. NSRP Studies Minimizing Waste from Mineral Slags. J. Protect. Coat. Lining 2000, 8, 73-77. National Surface Treatment Center. General Discussion of Abrasive Blasting. Available at: http://www.nstcenter.com/frames.aspx?menu_file⫽. js&main⫽/universal.aspx?id⫽2 (accessed 2005). Alberta Human Resources and Employment. Work Place Health and Safety Bulletin, Crystalline Silica in the Workplace, CH059 —Chemical Hazards; Alberta Human Resources and Employment: Alberta, Canada, 2004. Hubbs, A.F.; Minhas, N.S.; Jones, W.; Greskevitch, M.; Battelli, L.A.; Porter, D.W; Goldsmith, W.T.; Frazer, D.; Landsittel, D.P.; Ma, J.Y.C.; Barger, M.; Hill, K.; Schwegler-Berry, D.; Robinson, V.A.; Castranova, V. Comparative Pulmonary Toxicity of 6 Abrasive Blasting Agents; Toxicol. Sci. 2001, 61, 135-143. Buchan, B.; Cao, C. Coal-Fired Generation: Proven and Developing Technologies; Office of Market Monitoring and Strategic Analysis, Florida Public Service Commission: Tallahassee, FL, 2004. Naval Facilities Engineering Service Center. Recycling and Reuse Options for Spent Abrasive Blasting Media and Similar Wastes; Technical Memorandum by Battlle, TM-2178-ENV; Naval Facilities Engineering Service Center, Port Hueneme, CA, 1996. Chesapeake Specialty Products, Inc. Available at: http://www.chesprod. com/html/ met_abras.html (accessed 2005). Volume 56 August 2006

Kura, Kambham, Sangameswaran, and Potana 29. FairMount Minerals and Subsidiaries. Available at: http://www. Fairmountminerals.com/interior.asp?page⫽Black%20Magnum&category⫽ Markets&level1⫽Abrasive (accessed 2005). 30. Hansink, J.D. An Introduction to Abrasives for Protective Coating Removal Operations. J. Protect. Coat. Lining 2000, 4, 66-73. 31. Datar, S.R. Environmental Performance of Coal Slag and Garnet as Abrasives. Thesis. University of New Orleans: New Orleans, LA, 2003. 32. Paddison, R.D. What Type of Abrasive to Use? An Overview on Selection and Use of Common Blast-Cleaning Materials; Protect. Coat. Eur. 2000, 5, 10-16, 63. 33. Clemco Industries. Achieving Productivity from Abrasive Blast Cleaning Systems; J. Protect. Coat. Lining 1989, 9, 31-36. 34. Holt W.S; Austin, D.M. How Nozzle Pressure and Feed Rate Affect the Productivity of Dry Abrasive Blasting; J. Protect. Coat. Lining 2001, 10, 82-104. 35. U.S. Environmental Protection Agency. 40 CFR Part 60; Appendix A; Method 1-Sample and Velocity Traverses for Stationary Sources; U.S. Environmental Protection Agency: Washington, DC, 1997. 36. U.S. Environmental Protection Agency. 40 CFR Part 60; Appendix A; Method 2-Determination of Stack Gas Velocity and Volumetric Flow Rate (Type S Pitot Tube); U.S. Environmental Protection Agency: Washington, DC, 1997. 37. U.S. Environmental Protection Agency. 40 CFR Part 60; Appendix A; Method 3-Gas Analysis for the Determination of Dry Molecular Weight; U.S. Environmental Protection Agency: Washington, DC, 1997. 38. U.S. Environmental Protection Agency. 40 CFR Part 60; Appendix A; Method 4-Determination of Moisture Content in Stack Gases; U.S. Environmental Protection Agency: Washington, DC, 1997.

Volume 56 August 2006

39. U.S. Environmental Protection Agency. 40 CFR Part 60; Appendix A; Method 5-Determination of Particulate Matter Emissions from Stationary Sources; U.S. Environmental Protection Agency: Washington, DC, 1997. 40. Seavey, M. Abrasive Blasting above 100 psi. J. Protect. Coat. Lining 1985, 7, 26-37.

About the Authors Bhaskar Kura is a professor and Kalpalatha Kambham and Sivaramakrishnan Sangameswaran are graduate students in the Department of Civil and Environmental Engineering at the University of New Orleans. Sandhya Potana is an environmental engineer at Eco-Systems, Inc. Address correspondence to: Bhaskar Kura, Department of Civil and Environmental Engineering, Engineering Building, Room 828, New Orleans, LA 70148; phone: ⫹1-504-280-6572; fax: ⫹1-505-280-5586; e-mail: [email protected].

Journal of the Air & Waste Management Association 1215