Printer Friendly

An optimized electrostatic precipitator for air cleaning of particulate emissions from poultry facilities.

INTRODUCTION

Particulate matter emissions from poultry feeding operations have been associated with human and animal health concerns and wide-scale environmental issues. In addition to physical deposition to respiratory systems, PM in and from animal facilities carry chemical and biological hazards such as ammonia, infectious microorganisms, and endotoxins (Heederik et al. 2007; Just et al. 2009). PM exposures have been linked with increased respiratory and cardiovascular diseases of workers in animal facilities (Essen and Romberger 2003). Animal health and productivities are also affected by excessive PM exposure (Guarino et al. 1999). Increased mortality has been correlated with increased [PM.sub.10] and [PM.sub.2.5] levels in the atmosphere (Lippmann 2007).

Effective mitigation of PM emissions from poultry feeding operations needs to be developed to improve health and reduce their evironmental impacts. Many promising PM control technologies have already been tested such as multistage scrubbers, misting sprays, space ionization, and ESPs. It has been shown that multi-stage packed type scrubbers (Zhao et al. 2011; Melse et al. 2012) can reduce [PM.sub.10] emissions from 61% to 93%, but they require stronger or more ventilation fans to overcome the added equipment pressure drops. It is also just an exhaust treatment, which does not solve indoor air polution. Misting sprays applied inside swine facilities using water (Zhu et al. 2005), vegetable oil (Aarnink et al. 2011), and sulfuric acid (Jensen 2002) have reduced total PM by up to 87%. However, the long-term mist application can lead to enhanced microbial growth on walls and farm equipment due to availability of either moisture or oil substrates. Air ionization provides a mess-free indoor treatment to collect PM onto grounded surfaces inside animal barns (Cambra-Lopez et al. 2009). Total PM reduction efficiencies of up to 39% have been shown from tests in a swine facility (Jerez et al. 2011). However, attachment of dust into grounded surfaces made it easy for dust re-entrainment into the airstream. The low PM collection performance can also be due to the weak electrostatic attachment of particles to the collecting surfaces.

ESPs can provide better separation and handling of PM deposition and can achieve high PM removal efficiencies. St. George and Feddes (1995b) first tested a parallel-plate single-wire ESP prototype inside a swine facility, which was operated at a voltage range of-10.3, -11, and -12.1 KV, and air speeds of 0.55, 0.76, and 0.95 m/s. Although the highest PM collection efficiency of 96.4% was obtained at a setting of -12.1 KV and 0.76 m/s speed, the device was not feasible due to the very slow air speed (< 1 m/s), high power consumption of about 0.14 watts per [m.sup.3]/h of treated air, and generation of ozone concentrations of up to 210 ppb in the airstream. Negative charging systems are known to generate more ozone compared to positive ones despite providing better performance. Chai et al. (2009) tested an ESP using bi-polar charging systems and showed that an ESP operated with -30 KV produced higher collection efficiency (79%) compared to +30KV (62%). Therefore, the current ESP technology is not designed and optimized for operation in animal facilities, especially those for poultry. The high voltage and low air speed requirements and high ozone production have limited its application in animal facilities. There is a need to optimize ESP technology for mitigation of PM emissions from animal barns.

An ESP is a particulate collection device that uses very low electric current but very high voltages to ionize the airflow field and charge particles. The high electrostatic field rapidly attracts charged particles toward the collection plates. ESP collection efficiency exponentially increases with particle size, the field strength of the charger and collecting plates, and collector area, while it exponentially decreases with airflow rate as shown in Equation 1:

[eta][alpha] exp [D[E.sub.0][E.sub.p]A/Q] (1)

where

[eta] = the collection efficiency (dimensionless)

D = the diameter of the particle (m)

[E.sub.0] = the local electric field strength near the charger (V/m)

[E.sub.P] = the local electric field strength near the collector (V/m)

A = the area of the collecting electrode ([m.sup.2])

Q = the gas flow rate ([m.sup.3]/s)

An ESP can be designed using different charger and collector geometries: a single wire charger surrounded by either a (1) tubular or (2) parallel plate collectors, (3) multiple charger wires in between parallel collecting plates, and (4) two stage designs. ESP design greatly affects its performance (Manuzon and Zhao 2009). In general, the gap between the electrode and collector determines the maximum permissible electric field be fore occurrence of sparkover. The electric field varies with a gradient that is strongest near the source pole and weakest nearest the opposite pole. The strength of electric field and its distribution is very crucial in determining particle collection.

A tubular collector plate offers the best electric field distribution, which gets attenuated if parallel plates of the same area were used instead. However, particles are normally removed by mechanical agitation, which makes parallel plates preferable. The optimum length of a parallel collection plate is equal to twice the electrode and plate gap. Going longer than the optimum length further attenuates the average electric field strength of the total air space. If longer collecting plates are necessary, the electric field can be sustained by using multiple wires. However, using more wires leads to greater power consumption.

Lower power consumption can be achieved in a two stage design, which eliminates the need for maintaining a high electric field throughout the entire active space. The first stage, which consists of a small area at the ESP inlet, operates at a higher field strength in order to charge the particles. The second stage, which is located right after the first stage, occupies most of the ESP. It also operates at a lower field strength but with increased collection surface for enhanced particle collection. Unfortunately, this design is best suited for low particle loading since the second stage is usually compacted for increased surface, which makes cleaning difficult.

Treatment of PM emissions from poultry facilities requires an ESP that can handle large PM loads and provide better cleaning mechanisms, which can be most easily done with multiple wire and plate designs. The size and type of dust also needs to be considered in ESP processes. Larger particles are collected more efficiently. Material composition does not influence particle charging and migration, but it greatly influences adhesion to the collector plates. Poultry PM mostly consists of organic matter (Ellen and Takai 2000; Takai and Pedersen 2000), which tends to have low particle resistivity (less than [10.sup.8] [ohm] cm [[10.sup.7] [ohm] in.]). This property causes PM to lose charge easily which explains the re entrainment issues previously reported (Chai et al. 2009; Manuzon and Zhao 2009).

Gas flow rate or particle retention time is also an important factor in designing an ESP for poultry facilities. Most ESPs are operated from 1 to 2 m/s (200 to 400 ft/min), while airflow near poultry facility exhaust fans can be up to 7 m/s (1400 ft/min). To enable the ESP to function well, the fast moving airstream of the barn needs to be slowed down by increasing the cross sectional area of the ESP relative to the ESP air inlet. Therefore, to minimize ESP size, the ESP needs to be run at the fastest allowable air speed.

New ESP design improvements are needed to handle problems such as low PM adhesion, high power consumption, and ozone generation. A multiple wire and plate design with longer plate lengths and sustained electric field strength was developed to address low PM adhesion issues and allow greater collection surface for prolonged operations. Use of parallel plates allowed for mechanized automatic cleaning. An optimized ESP geometry was developed using CFD simulations (Manuzon 2012). This study aimed to empirically optimize the operation of the new ESP for PM air cleaning at poultry facilities where the greatest PM concerns exist.

The specific objectives of the study were to accomplish the following:

1. Determine optimized settings of voltage and air velocity for the new ESP.

2. Quantify effects of dust concentration, dust type, and operation time on the PM collection efficiency of the new ESP in the laboratory.

3. Evaluate the performance of the new ESP in a commercial poultry facility.

MATERIALS AND METHODS

The ESP Module

The optimized ESP design was developed (Figure 1b) based on previous studies and CFD simulations (Manuzon 2012). The CFD study showed that two wires were sufficient to develop sustained field strength to charge and collect poultry PM. The ESP design shown in Figure 1a was fabricated to fit a 929 [cm.sup.2] (1 [ft.sup.2]) area. The collector plate assembly consisted of seven parallel plates arranged 5 cm (2 in.) apart with 12-1.27 mm (0.04 in.) diameter charger wires. Each plate contained an area of 619 [cm.sup.2] (0.66 [ft.sup.2]), which made a total of 7433 [cm.sup.2] (8 [ft.sup.2]) per ESP section. The complete details of the ESP are summarized in Table 1.

The Laboratory ESP Test Apparatus

The ESP unit was mounted at the center of a wind tunnel for laboratory testing shown in Figure 2(a). The unit was powered using a high voltage power supply that can provide a maximum of 15,000 V and 4 mA. To simulate various air speeds, flow rate was controlled using a variable speed centrifugal blower located at the inlet of the wind tunnel. A turntable dust feeder was used to meter the dust, which was introduced into the wind tunnel using a venturi nozzle. During the ESP operation, PM was measured about 0.3 m (1 ft) or 1 diameter away from the ESP at the duct centers using isokinetic nozzles. PM sensors were switched before and after the ESP unit to measure PM concentration.

The Field ESP Unit and the Commercial Poultry Facility

The ESP unit was tested in a mechanically ventilated poultry layer facility housing 200,000 layer hens (Hyline W36,48 weeks old) from June 26 to July 7, 2012. The facility had 48-137 cm (54 in.) diameter exhaust fans located on two end walls and nine baffle air inlets located along the center of the roof section. The five stage ventilation system was thermally controlled with a setpoint of 28[degrees]C (82[degrees]F) during testing, while part of the indoor air is recycled to continuously dry the manure belts.

The field ESP unit was designed similar to the lab unit, except for an additional automated cleaning system as shown in Figure 2(b). The inlet (a) and outlet (b) transition was reduced from 2.4 m (8 ft) to 0.6 m (2 ft) just enough to ensure proper PM measurements. A fan (f) was used to ensure constant air supply thru the system. The unit was automatically cleaned using a programmable logic controller (g), which drove a set of pneumatic nozzles to blow high pressure air sequentially into each of the six sections inside the ESP unit every four hours. Simultaneously, a vacuum cleaner (g) was activated to draw knocked off PM during the cleaning cycle which lasted for six minutes.

Laboratory Tests Experimental Design

The following experiments were conducted in the laboratory to optimize the ESP operating parameters and quantify the ESP PM collection performance:

* effects of operating voltage and velocity

* effects of PM type and loading

* effects of ESP operation time

The first study determined the optimum setting for volt age and velocity. The second study examined how dust type and concentration can affect ESP performance using previously optimized ESP settings. The last test verified if the findings from the previous tests were sustained for longer ESP operation and if poor PM adhesion issue was resolved.

The Effect of Operating Voltage and Velocity on ESP Performance. The response surface curve of the ESP at various voltages and velocities (E-V Curve) was used in determining the best operating voltage and velocity of the ESP. However, commercially available flour was initially used as a preliminary screening tool instead of poultry dust due to the limited amount of samples. The size distribution and density of flour was similar to that of poultry dust, so it was expected that their collection performance would be equivalent. The effect of dust type was tested in another section to eliminate any uncertainties from this test.

A central composite design (CCD) with axial and center points was used to obtain the response curve (Table 2) of ESP performance. Three replicate runs were performed for each point. The voltage was varied from 9.4 KV to 13.6 KV, while the velocity was varied from 0.8 m/s to 2.2 m/s (157 to 433 ft/min).

The best fit response surface model was fitted based on the following general formula:

Y = [A.sub.0] + [A.sub.1][X.sub.1] + [A.sub.2][X.sub.2] + [A.sub.3][X.sub.1][X.sub.2] + [A.sub.4][X.sup.2.sub.1] + [A.sub.5][X.sup.2.sub.2] + [epsilon] (2)

where

Y = the response variable (collection efficiency, %)

[X.sub.i] = the predictor variables

i = 1 (voltage, KV), 2 (velocity, m/s or ft/min)

[A.sub.i] = regression coefficients

[epsilon] = the random error

The Effect of Dust Type and Concentration on ESP Performance. The next study was performed to determine the effect of dust type and concentration on ESP performance as shown in Table 3. Three types of dust were tested, namely poultry, flour, and Arizona dust (ISO 12103 1,A2, Powder Technology Inc., Burnsville, MN). Poultry dust was tested at 2,4, and 9 mg/[m.sup.3] (12.4,24.9, and 26.2 x [10.sup.-8] lb/[ft.sup.3]); flour was tested at 1, 3, and 7 mg/[m.sup.3] (6.2, 18.7, and 43.7 x [10.sup.-8] lb/[ft.sup.3]); while Arizona dust was tested at 7 and 14 mg/[m.sup.3] (43.7 and 87.4 x [10.sup.-8] lb/[ft.sup.3]). Each test was performed in triplicate runs.

The resulting dust concentrations were not similar for each dust type, which was limited by the possible adjustments made on the dust feeder. Similarly, only two levels were attained in testing the Arizona dust, since at 7 mg/[m.sup.3] (43.7 x [10.sup.-8] lb/[ft.sup.3]) were the minimum possible dust concentration, while above 14 mg/[m.sup.3] (87.4 x [10.sup.-8] lb/[ft.sup.3]), would possibly cause damage to the PM sensors.

The experiment was constrained by the inability of the dust feeder to generate similar levels of dust concentrations for each type. Thus, several statistical tests were performed to carefully evaluate the effects of dust type and concentration on ESP collection efficiency. Initially, multi-variate linear regression with analysis of covariance (ANACOVA) was performed to test the effects of dust type on ESP performance with dust concentration as the covariate factor. ANACOVA enabled simultaneous comparisons of the regression lines for poultry, flour, and Arizona dust, despite having an unbalanced experimental design for levels of dust concentration. The following statistical model was used:

Y = [B.sub.0] + [B.sub.1] x + [B.sub.2][z.sub.A] + [B.sub.3][z.sub.B] + [B.sub.4]x x [z.sub.A] + [B.sub.5]x x [z.sub.B] + [epsilon] (3)

where

Y = the response variable (collection efficiency, %)

x = the predictor variable (dust concentration, mg/[m.sup.3])

[z.sub.A], [z.sub.B] = indicator variables, representing different dust types with [z.sub.A] = 1, [z.sub.B] = 0 for Arizona dust; [z.sub.A] = 0, [z.sub.B] = 1 for flour; and [z.sub.A] = 0, [z.sub.B] = 0 for poultry dust

[B.sub.i] = the regression coefficients, with i = 1 ... 5

[epsilon] = the random error

The regression lines correlating dust types (poultry, flour, and Arizona dust) against collection efficiency were individually examined to test for the effect of concentration. The model used for the test was

Y = [B.sub.0] + [B.sub.1]x + [epsilon] (4)

Equation 4 used the same definitions as Equation 3 and is similar except it was missing third to fifth terms, since only one type of dust was evaluated at a time. Additionally, since only two levels (7 and 14 mg/[m.sup.3] [43.7 and 87.4 x [10.sup.-8] lb/[ft.sup.3]]) were obtained for the Arizona dust correlation, analysis of variance (ANOVA) was performed to test the means for significant difference. The null hypothesiss (Ho) used for the test was that the mean efficiency obtained at 7 mg/[m.sup.3] (43.7 x [10.sup.-8] lb/[ft.sup.3]) was not statistically significant (alpha = 0.05) with that obtained at 14 mg/[m.sup.3] (87.4 x [10.sup.-8] lb/[ft.sup.3]).

The Effect of Time on ESP Performance. The last study aimed to evaluate the performance of the ESP for extended periods using poultry dust. A mean inlet concentration of 2 mg/[m.sup.3] (12.4 x [10.sup.-8] lb/[ft.sup.3]) was used in the test based on established literature values (Guarino M. A Caroli and Navarotto 1999; Cambra-Lopez et al. 2009; Aarnink et al. 2011). The purpose of the test was to examine the behavior of the PM deposits within six hours after being trapped into the collecting plates and determine the appropriate time required before plate cleaning. Only a single lab test was performed due to the limited availability of poultry dust. Selection of the appropriate cleaning time was further verified by comparing the ESP's lab performance with its performance inside the actual poultry house for three four-hour cycles after the automated cleaning system was activated. Only descriptive statistics were performed to compare the lab run and the three field runs due to the dissimilar conditions between each type of test.

Measurement of PM Concentrations and Environmental Conditions

Two types of real-time aerosol sensors were used that measure mass concentration (Sensor 1) and particle size distribution and concentration (Sensor 2). Sensor 1 (last calibrated in June 2011) provides mass concentration data with a measurement range of 0.001 to 150 mg/[m.sup.3] (1.6 x [10.sup.-12] to 2.4 x [10.sup.-7] lb/[ft.sup.3]) and presents the measurement in size-segregated mass fraction concentrations corresponding to [PM.sub.1], [PM.sub.2.5], respirable, [PM.sub.10], and total PM size fractions. Sensor 2 (last calibrated December 2010) provides high-resolution aerodynamic particle size distribution ranging 0.5 to 20 [micro]m (2 to 79 in. x [10.sup.-5]) classified into 52 bin sizes. It also provided total particle concentration data up to a range of 1000 particles/[cm.sup.3] (16,000 particles/[in.sup.3]). There were major differences between aerosol sensors. Sensor 1 was more applicable for field and long-term measurements, due to its versatility of providing direct mass concentrations of [PM.sub.1], [PM.sub.2.5], respirable and total PM mass concentrations. It was also more portable and required less maintenance compared with Sensor 2. Sensor 2 had higher sensitivity and provided high-resolution particle count data in 52 bin sizes. Sensor 2 was very useful for differentiating types of dust based on their particle size distribution (PSD) and ESP grade efficiencies, but it would be very difficult to use them for long-term and repeated tests. Grade efficiency data will still have to be converted to average efficiencies, which was necessary to completely describe ESP performance. Aside from being cumbersome, converting count data into averaged mass concentration data can lead to huge errors, especially if large particles were miscounted. Both PM sensors recorded dust concentration averages and counts every minute at a 1 s sampling frequency for the total test duration.

A data logger was used for recording process temperature and relative humidity every minute, while an anemometer was used to check air velocity before, after, and during checks in between the test. Ozone concentration was measured using gas detector tubes.

PM Collection Efficiency Calculation

The performance of the ESP was determined with the following particle collection efficiency formula:

[[eta].sub.i] = 1 - [C.sub.o]/[C.sub.i] = 1 - [P.sub.i] (5)

where:

[[eta].sub.i] = the particle collection efficiency of the ith measurement

[C.sub.o] = the outlet or during test particle concentration (particles/[cm.sup.3] [particles/[in.sup.3]] or mg/[m.sup.3] [lb/[ft.sup.3]])

[C.sub.i] = the inlet or before-test particle concentration (particles/[cm.sup.3] [particles/[in.sup.3]] ormg/[m.sup.3][lb/[ft.sup.3]])

[P.sub.i] = the particle penetration of the it h measurement

If efficiency was obtained using particle count data, the formula given in Equation 5 calculates the grade efficiency or penetration, which is applicable only within the specified size range. Particle collection greatly depends on size, while this varies according to the PSD. The average collection efficiency for a PSD can be calculated using the formula:

[[eta].sub.i,AVE] = [summation over i][[eta].sub.i][C.sub.i]/[summation over i][C.sub.i] = 1 - [P.sub.i,AVE] (6)

where

[[eta].sub.i,AVE] = the average collection efficiency of the PSD

[P.sub.i,AVE] = the average penetration of the PSD

Equation 6 can be used to convert grade efficiency data calculated from Sensor 2 measurements to average efficiency data. Hence, average efficiency data directly calculated from Sensor 1 measurements can be compared to average efficiency from Sensor 2 provided that the count concentrations were first converted to volume or mass concentrations in Equation 6.

Statistical Analyses

All analysis of variance, covariance, and multi-variate regressions were performed using commercial statistical soft ware and evaluated based on an alpha of 0.05 (P < 0.05) significance level. In particular: (1) multivariate regression analysis was performed to obtain the best fit response surface model of efficiency as a function of voltage and velocity; (2) regression analysis, ANOVA, and ANACOVA were performed to deter mine the effects of dust type and uneven levels of concentration on collection efficiency. The complete details of each model equation evaluated were described in the previous sections (3). Lastly, general descriptive statistics were calculated for the plots of efficiencies during the runs for the lab testing and the actual field testing.

RESULTS AND DISCUSSIONS

The E-V Curve and Optimum Voltage and Air Velocity

Figure 3a shows the voltage and velocity contour response surface (E V Curve) for the efficiency of the ESP. The plot showed the expected trend that efficiency increased from approximately 30% to 85% as voltage was increased from 9.5 KV to 13.5 KV and velocity was decreased from 3 m/ s to 0.5 m/s (590 to 98 ft/min). The best fit response surface described in Equation 7 had a regression coefficient ([R.sup.2]) of 0.8125. The good correlation between actual and predicted data can be verified by examining the residuals plot. Figure 3b shows the randomly distributed residuals ranging from -6% to 10% for 27 data samples.

[eta] = [A.sub.0] + [A.sub.1](E - 11.5/1.5) + [A.sub.2](V - 1.5/0.5), if V in m/s [eta] = [A.sub.0] + [A.sub.1](E - 11.5/1.5) + [A.sub.2](V - 295/98), if V in f/min (7)

where:

[eta] = the particle collection efficiency (dimensionless)

E = the voltage (KV)

V = the air velocity (m/s [ft/min])

[A.sub.0] = 0.67282286470821

[A.sub.1] = 0.11039285845043

[A.sub.2] = -0.0448572318091

The second term inside the parenthesis of Equation 6 transformed the operating voltage into a dimensionless volt age by deducting 11.5 KV (the center point of the experimental design) and dividing by 1.5 KV (half of the voltage operating range). A similar procedure was performed on the third term using 1.5 m/s (295 ft/min) as centerpoint and 0.5 m/s (98 ft/min) range instead. Therefore, the entire equation was expressed in a dimensionless form. Since the operating voltage and velocity range also correspond to the only feasible working range of the equipment, it can be deduced that the magnitudes of [A.sub.1] and [A.sub.2] coefficients in Equation 7 reveal the relative influence among both variables. Thus, having a voltage coefficient that is 2.5 times larger than the coefficient for air velocity suggests that ESP performance was more sensitive to voltage change than air velocity.

The optimized operating point, which is where collection efficiency achieved maximum, was identified at 13.6 KV and 1.7 m/s (334 ft/min). Operating at higher voltages is close to the spark limit observed at 14 KV, while having an air velocity of 1.7 m/s (334 ft/min) was the fastest allowable speed. Previous observations (Manuzon and Zhao 2009) have shown that operating above 2 m/s (394 ft/min) can lead to re entrainment in less than one hour after poultry PM were collected into plates, especially the particles whose sizes were greater than 10 um. The choices of both optimum voltage and velocity were therefore selected by deciding to operate slightly below the equipment limitations as a safety factor and by the goal to achieve at least 80% absolute collection efficiency.

One limitation of this study is that flour was used instead of actual poultry dust to test the clean plate performance of the ESP due to the limited amount of poultry dust samples. It is reasonable to use flour to simulate poultry dust in the tests, since its PSD was close to that of the poultry dust. In this specific study the count median diameter (CMD) and geometric standard deviation (GSD) of the flour was 2.1 and 2.0 um (8.2 and 7.9 x [10.sup.-5] in.), while that for actual poultry dust was 1.3 and 1.7 [micro]m (5.1 and 6.7 x [10.sup.-5] in.), respectively. Flour also has the same particle density (1.5 g/[cm.sup.3] [90 lb/[ft.sup.3]]) compared to the poultry dust.

The Effect of Dust Type and Concentration

Figure 4 shows the ESP PM collection efficiencies for different types of dusts and loads. Poultry dust collection efficiencies varied from 80% to 89% as PM loads were increased from 2 to 9 mg/[m.sup.3] (12.4 and 56.2 x [10.sup.-8] lb/[ft.sup.3]). Flour dust collection efficiencies varied from 79% to 90% as PM loads were increased from 1 to 7 mg/[m.sup.3] (6.2 to 43.7 x [10.sup.-8] lb/[ft.sup.3]). Lastly, Arizona dust collection efficiencies were 35% and 52% for 7 and 14 mg/[m.sup.3] (43.7 and 87.4 x [10.sup.-8] lb/[ft.sup.3]) loads, respectively.

The effect of dust type was analyzed by means of regression analysis with ANACOVA. The resulting model was obtained based on Equation 3 and is shown in Equation 8:

Y = 65.13 + 1.06x - 32.54[z.sub.A] + 18.01 [z.sub.B] (8)

Equation 8 had a regression coefficient ([R.sup.2]) of 0.93724. Although the ANOVA showed that both the effects of concentration (P > F = 0.0093) and type (P > F = < 0.0001) were significant; power calculations showed that the least significant number (LSN) necessary to verify the F-tests for concentration was 14 and for dust type was 6. Since the actual sample sizes were only 6 for Arizona dust and 9 each for poultry and flour, the power analysis suggests that there was not enough power to conclude that concentration had a significant effect on ESP performance. However, the analysis verified that there was enough power to conclude that dust type significantly affected ESP performance.

ANACOVA enabled simultaneous comparisons of the three dust types using their least square means instead of their arithmetic means to account for other effects in the model caused by having an unbalanced design. The least squares mean of efficiency for poultry, flour, and Arizona dust were 86.2%, 89.7%, and 39.1% respectively. Pairwise mean comparison of those means using Tukey HSD showed that the mean efficiencies for poultry and flour were not significantly different, while the mean efficiency for Arizona dust was significantly different compared to the other two based on an alpha = 0.05 significance level. The result also suggested that using flour instead of poultry dust can yield the same ESP performance, which validated the material substitutions performed in the earlier tests.

The effect of dust concentration was more carefully examined next by considering each group of lines shown in Figure 4, individually. Separate linear regression analyses were performed for each dust type to determine the effect of concentration on ESP collection efficiency. For both poultry dust ([R.sup.2] = 0.1538) and flour ([R.sup.2] = 0.0756), the low correlation coefficient computed clearly suggests that dust concentration did not affect collection efficiency within the range of concentrations examined. The higher correlation results obtained for Arizona dust ([R.sup.2] = 0.6854) can be due to having only two levels, which makes regression analysis inappropriate for this case. ANOVA was used instead, which showed that there was no significant difference (a = 0.05) between the mean efficiencies obtained at 7 and 14 mg/[m.sup.3] (43.7 and 87.4 x [10.sup.-8] lb/[ft.sup.3]) for Arizona dust. The observation that dust concentration had no significant effect on ESP performance was the expected result since ESP performance should not be affected by PM concentration unless the plates were overloaded.

It was interesting to observe that Arizona dust had the lowest collection efficiency among the PM tested. As mentioned earlier, PM collection efficiency was greatly affected by the particle size distribution and adhesion characteristics. Therefore, the size distribution and other inertial characteristics of each type of PM tested were compared in Table 4. The values presented in the table were the result of measurements and literature search. Poultry dust, flour, and Arizona dust were arranged in size from the smallest to the largest with CMD of 1.3, 2.1, and 2.9 [micro]m (5.1, 8.3, and 11.4 x [10.sup.-5] in.), respectively. It must also be noted that the GSD or the size spread of the three types of particles were very close ranging from 1.7 to 2 [micro]m (6.7 to 7.9 x [10.sup.-5] in.). Among the three types of dust, flour particles were the lightest with a particle density at 1.49 g/[cm.sup.3] or [93 lb/[ft.sup.3]], followed by poultry (1.53 g/[cm.sup.3] [96 lb/[ft.sup.3]]), and Arizona (2.65 g/[cm.sup.3] [165 lb/[ft.sup.3]]).

Larger and heavier particles do not easily follow changes in airflow direction compared to smaller and lighter ones. The relaxation time of the particle is a measure of how fast a particle can change its direction if an external force is applied. Table 4 showed the calculated relaxation time for the three particles. Arizona dust had one order of magnitude slower relaxation time compared to poultry and flour dust, which had almost the same. This confirms why Arizona dust had low-collection efficiency compared to poultry and flour dust, which had the same collection performance. In impaction and filtration processes, larger particles tend to get collected more easily. However, in electrostatic precipitation, wherein horizontally moving particles must be diverted toward the vertical direction of the collecting plates, heavier particles will be harder to collect than lighter ones if their size and charge characteristics were equal.

Different types of dusts inherently come with different size distributions due to their difference in composition. Because of this, it is very difficult for two types of dust to have equivalent collection efficiencies. For example, although the flour and poultry dusts had close total collection efficiencies, their grade efficiencies were completely different as shown in Figure 5. Sensor 2 was used as PM sensor for calculating the efficiency results, which provided greater resolution on grade efficiencies. The PSDs of both types of PM were similar but still completely different, as shown by the different peak values. The peak for the PSD of poultry dust occurred at 1.4 um (5.5 x [10.sup.-5] in.), while that of flour dust was at 3 um (11.8 x [10.sup.-5] in.). The grade efficiency plots of each type of PM were also completely different, having the curve for flour lie above that of poultry dust. However, similar average collection efficiencies can still be obtained even if their grade efficiencies were completely different as long as their PSDs were close enough. The average efficiency of flour and poultry dust was calculated using the grade efficiency data from Figure 5 and Equation 6. The average total collection efficiency computed was 88.7% for poultry dust and 92.5% for flour, which was about 4.2% difference only.

The Effect of Run Time on ESP Performance

Figure 6 shows the effect of run time on the performance of the ESP in the laboratory using poultry dust and with efficiencies obtained by measurements from Sensor 1. The plot showed how total collection efficiency varied from 69% to 96% during the entire six hour run for an inlet concentration of 1.813 [+ or -] 0.206 mg/[m.sup.3] (11.2 [+ or -] 1.28 lb/[ft.sup.3]) of total suspended particles, 1.21 [+ or -] 0.148 mg/[m.sup.3] (7.6 [+ or -] 0.9 lb/[ft.sup.3]) of [PM.sub.10], 0.719 [+ or -] 0.08 mg/[m.sup.3] (4.5 [+ or -] .5 lb/[ft.sup.3]) of respirable PM, 0.612 [+ or -] 0.076 mg/[m.sup.3] (3.8 [+ or -] 0.5 lb/[ft.sup.3]) of [PM.sub.2.5], and 0.562 [+ or -] 0.075 mg/[m.sup.3] (3.5 [+ or -] 0.5) of [PM.sub.1]. After four hours, there was a steep drop in collection performance which immediately went back up within 30 minutes. This was due to malfunctions in the feeding rate of the dust feeder and was not a fault of the ESP system. During the test, the dust supply was diminished, so a new batch of dust had to be loaded to the dust feeder. The introduction generated the observed feed fluctuations. The mean performance of the ESP for the entire run were 89.4% [+ or -] 7.3% for total PM, 87.1% [+ or -] 8.4% for [PM.sub.10], 86.5% [+ or -] 8% for respirable PM, 87.9% [+ or -] 7.5% for [PM.sub.25], and 89.5% [+ or -] 7.4% for [PM.sub.1]. As the plates of the ESP became loaded with the PM layer, the electric field changed due to the insulating nature of the PM components. This was accompanied by a decrease in current consumption and ESP performance due to the voltage drop development across the plates. Although a slight drop in current was observed from 2.3 to 2.1 mA, the ESP efficiency decrease was not observed for the entire six hour run, except during the time when feed rate became unstable. This would suggest that plates do not need to be cleaned within six hours, as long as the PM load was not higher than 2 mg/[m.sup.3] (12.4 x [10.sup.-8] lb/[ft.sup.3]).

Figure 6 also showed that even the poultry particles that were smaller than 1 um (4 x [10.sup.-5] in.) had high collection efficiency averaging at 89.5%. Similar results were shown by the time plot of collection efficiencies measured with the Sensor 2 (Figure 7). Count measurements provided a better resolution of the grade efficiencies of the ESP performance. The results showed that larger particles were easier to collect compared to smaller ones as shown by the decreasing trend in collection performance of 62.8%, 72.5%, 77.9%, 84%, 83.4%, and 92% for 0.5, 0.8, 1,2.3,4.1, and 10.4 [micro]m (2,3,4, 9, 16,41 in. x [10.sup.-5]) particles, respectively.

The general statistics of the grade efficiencies for important particles sizes were obtained from Sensor 2 and summarized in Table 5. The table showed that there was a certain degree of variation ranging from 7.7% to 14.1% relative standard deviations. In general, the wide variations were a consequence of dust concentration variations, mainly due to feeding rate fluctuations of the dust feeder. The most important value to look at is therefore the average values, unless a consistent drop in collection efficiency is observed, which would mean PM re-entrainment from the plates that would require cleaning.

The grade efficiency values and particle size distribution obtained from Sensor 2 measurements for poultry PM (shown in Table 5) were used to calculate the average [PM.sub.1], [PM.sub.2.5], respirable, and [PM.sub.10] collection efficiency shown in Table 6 and was compared with the calculated efficiencies obtained from Sensor 1. Table 6 shows very close agreement between sensor readings except for the [PM.sub.1] values, which had about 16% difference in readings between the two instruments. It was expected to have poorer agreement between PM sensor readings at the smaller sized ranges due to their difference in sensitivity for measuring small sized particles.

The most important objective for conducting the effect of time in the laboratory was to evaluate if the selected four hour cleaning cycle was sufficient and does not affect ESP performance in the field. Figure 8 illustrated the field reliability of the ESP unit. Lab1 was performed during the lab testing for six hours, while field1 to field3 were the performance of the ESP inside the actual poultry farm during a four hour cycle when the plates were still considered clean. The plots showed that the performance of the ESP was still satisfactory even in actual field conditions. The average efficiencies of the field trials were 85.7% [+ or -] 1.9% (field1), 80.8% [+ or -] 5.3% (feld2), and 82.9% [+ or -] 3.1% (field3), compared to the 89.4% [+ or -] 7.3% average in lab1. The initial concentrations for each test were different. The lab test was done at an inlet concentration of 2 mg/[m.sup.3] (12 x [10.sup.-8] lb/[ft.sup.3]), while the field test was done at an average of 0.1 mg/[m.sup.3] (6.2 x [10.sup.-9] lb/[ft.sup.3]). This was the limitation of the test, since it was not possible to change the concentration of PM inside the farm during the field test, which also varied with time of day due to differences in temperature, humidity, dust source, and type. The lab tests were run at a temperature of 17[degrees]C [+ or -] 2 [degrees]C (63[degrees]F [+ or -] 4 [degrees]F) and relative humidity of 33% [+ or -] 5%.

Verification of the ESP PM Collection Performance at a Commercial Poultry House

Figure 9 showed the performance of the ESP during the last five days of the 10-day field test. The ESP performed very well with a consistent total PM collection efficiency of 81.8% [+ or -] 6.5%. The efficiency was fairly close to its performance during the lab testing, which was at 89.4% [+ or -] 7.3%. Furthermore, the inlet concentration during the run was 0.1 [+ or -] 0.0 mg/[m.sup.3] (6.2 x [10.sup.-9] lb/[ft.sup.3]), which ranged from 0.098 to 0.205 mg/[m.sup.3] (6 to 12 x [10.sup.-9] lb/[ft.sup.3]). The range of concentrations where the ESP was exposed to was lower than the 2 mg/[m.sup.3] (12 x 10-8 lb/[ft.sup.3]) value used during the lab testing so it is just reasonable to expect that the field test should showed performance that is similar to the lab test. The outlet concentration during the run was 0.02163 [+ or -] 0.00826 mg/[m.sup.3] (12 [+ or -] 0.5 x [10.sup.-9] lb/[ft.sup.3]), which ranged from 0.005 to 0.092 mg/[m.sup.3] (0.3 to 5.7 x [10.sup.-9] lb/[ft.sup.3]). The average collection efficiencies for each size range were: 86.1% [+ or -] 5.5% for [PM.sub.1] particles, 85.8% [+ or -] 5.4% for [PM.sub.2.5] particles, 85.3% [+ or -] 5.4% for respirable particles, and 84.1% [+ or -] 5.4% for [PM.sub.10] particles. Thus, even the small-sized particles were being collected well by the ESP. However, the trend showing that collection efficiency was increasing as the particle size decreases was an inconsistent result. This observation was not seen in the lab test and is also inconsistent with theory. Smaller particles should be more difficult to collect. However, there were also several factors that can cause the mentioned observations. It can be due to inefficient sampling, since sampling tubes were difficult to align during the test and they were very sensitive to the air stream. Although the tube samplers were designed for isokinetic conditions, the movements of the tubing during the test could have affected the sampling efficiency. Another explanation is the generation of larger particles, which might occur if the plates were getting saturated.

Figure 10 shows the temperature and humidity profile during the entire 10-day run. The mean values for the measurement were 28.3 [+ or -] 3.70 C (83 [+ or -]70 F) for temperature and 56.4% [+ or -] 9.5% for relative humidity. The temperature ranged from 23[degrees]C to 36[degrees]C (73[degrees]F to 97[degrees]F), while the relative humidity ranged from 32% to 75% during the testing. Exposure of the ESP to greater variations in environmental condition were ideal since these were the conditions that the lab test cannot simulate. Exposing the ESP to these conditions is the ultimate test that warrants the applicability of the designed ESP to the farm environment.

The ozone levels of the ESP outlet were also measured during the field run. The levels were below the lowest range of the detector tube which was 50 ppb.

Comparison of this ESP versus other Electrostatic Devices for PM Control of Animal Facilities

Table 7 shows the comparison of the optimized ESP design with other electrostatic PM collection devices used for poultry facilities. The optimized ESP had the highest collection performance among all types tested, even with the faster flowing air of 1.7 m/s (334 ft/min). Re-entrainment is the primary reason why PM collection efficiency is lower in air ionization systems. It was shown in an air ionization system that particle collection dropped from about 75% to about 30% between five to 19 days of its operation (Cambra-Lopez et al., 2009). Similar results have been shown in a previous study where the particle collection efficiency dropped within less than an hour of operation starting from the larger particles (Manuzon and Zhao 2009). Another design (Chai et al. 2009), which was meant for exhaust treatment, had almost 80% collection efficiency of total particles but with only 55% for [PM.sub.2.5]. The design was also not tested using actual poultry dust but with cornstarch instead. The effect of long-term operation was also not tested.

In order to collect fine particles and make them stay in the collecting walls, the electric field must be strong and sustained enough to attract and hold them together. The design tested here is the optimum design, which means that spacing plates further apart would lead to lower current consumption but would collect the particles less efficiently (Manuzon and Zhao 2012). Placing the collecting plates farther apart limits the strength of the electric field in some areas such that increasing the operating voltage further does not result in a favorable electric field gradient. This was shown in a study (Chai et al. 2009) where increasing the voltage from 30 KV to 60KV no longer resulted in improved collection efficiency. In order to collect above 80% of the [PM.sub.10] and [PM.sub.2.5], enough energy must be spent so they remain within the collector plates before they are cleaned and collected.

CONCLUSION

An ESP for collecting poultry PM was developed and evaluated in the laboratory with an improved collecting efficiency of 89.4% [+ or -] 7.3% for total PM, 87.1% [+ or -] 8.4% for [PM.sub.10], 86.5% [+ or -] 8% for respirable PM, 87.9% [+ or -] 7.5% for [PM.sub.2.5], and 89.5% [+ or -] 7.4% for [PM.sub.1].

The current consumption of the ESP was 0.047 watts per [m.sup.3]/h (1 e-4 hp/cfm) at the optimized voltage setting of 13.6 KV and 1.7 m/s (334 ft/min).

The operating conditions of the ESP was optimized by obtaining the voltage and velocity curve (E-V curve) for performance using response surface methods at levels of 9.6 to 13.6 KV and 0.8 to 2.2 m/s (157 to 433 ft/min) air velocity. An equation with good fit ([R.sup.2] = 0.8125) had prediction with absolute residuals not greater than 10% difference from the actual values. The prediction equation also showed that ESP performance increased with increasing voltage and decreased with air speed and that air speed was less sensitive to voltage in terms of their effect on collection efficiency.

The optimized ESP had the best performance compared to other types of ESP designs tested in animal facilities. The improved ESP design collected greater than 80% of both [PM.sub.10] and [PM.sub.2.5] components at a speed of 1.7 m/s (334 ft/min), which was higher compared to other designs taking the air speed into consideration. The ESP, however, may have about 60 times more power consumption compared to air ionization systems for improved performance in collection of finer particulates.

The effects of dust types and concentrations were also tested using poultry, flour, and Arizona dust. Poultry dust collection efficiencies varied from 80% for 2 mg/[m.sup.3] (12.4 x 10-8 lb/[ft.sup.3]) and 89% 9 mg/[m.sup.3] (56.2 x [10.sup.-8] lb/[ft.sup.3]) dust loads. Flour dust collection efficiencies varied from 79% for 1 mg/[m.sup.3] (6.2 x [10.sup.-8] lb/[ft.sup.3]) and 90% for 7 mg/[m.sup.3] (43.7 x [10.sup.-8] lb/[ft.sup.3]). Lastly, Arizona dust collection efficiencies were 35% for 7 mg/[m.sup.3] (43.7 x [10.sup.-8] lb/[ft.sup.3]) and 52% for 14 mg/[m.sup.3] (87.4 x 10-8 lb/[ft.sup.3]) loads, respectively.

The effect of operating time on ESP performance was examined. It was shown that the ESP can be operated for as long as six hours without needing to clean the plates.

The ESP was further tested inside an actual poultry facility for 10 days of continuous operation after being equipped with an automated cleaning system operating at a four hour cycle. The performance of the ESP for collecting actual poultry dust was 86.1% [+ or -] 5.5% for [PM.sub.1], 85.8% [+ or -] 5.4% for [PM.sub.2.5], 85.3% [+ or -] 5.4% for [PM.sub.4], 84.1% [+ or -] 5.4% for [PM.sub.10], and 81.8% [+ or -] 6.5% for total PM.

ACKNOWLEDGMENTS

The authors would like to thank Mark Meyer, Dong Jugang, Lara Hadlocon, and Larry Heckendorn for their asistance with this research.

The authors would also like to thank the ASHRAE Grant-in-Aid program for providing funding for this research.

REFERENCES

Aarnink, A.J.A., J.V. Harn, T.G.V. Hattum, Y. Zhao, N.W.M. Ogink. 2011. Dust Reduction in Broiler Houses by Spraying Rapeseed Oil. Transactions of theASABE 54:1479-89.

Cambra-Lopez, M., A. Winkel, J.V. Harn, N.W.M. Ogink, A.J.A. Aarnink. 2009. Ionization for Reducing Particulate Matter Emissions from Poultry Houses. Transactions of the ASABE 52: 1757-71.

Chai, M., M. Lu, T. Keener, S.J. Khang, C. Chaiwatpongsakorn, J. Tisch. 2009. Using an improved electrostatic precipitator for poultry dust removal. Journal of Electrostatics 67:870.

Ellen H.H., R.W. Bottcher, E. von Wachenfelt, and H. Takai. 2000. Dust levels and control methods in poultry houses. Journal of Agricultural Safety and Health 6:275-82.

Essen, S.V., Romberger, D., 2003. The respiratory inflammatory response to swine confinement building environment: adaptations to respiratory exposures in the chronically exposed worker. Journal of Agricultural Safety and Health 9:185-196.

George, S.D.S. and J.J.R. Feddes. 1995a. Electrical Properties of Organic and Respirable Swine Dust. Canadian Agricultural Engineering 37:97-101.

George, S.D.S. and J.J.R. Feddes. 1995b. Removal of airborne swine dust by electrostatic precipitation. Canadian Agricultural Engineering 37:103-7.

Guarino M., A Caroli, P. Navarotto. 1999. Dust concentration and mortality distribution in an enclosed laying house. Transactions of the ASAE 42, 1127-33.

Heederik, D., T. Sigsgaard, P.S. Thorne, J.N. Kline, R. Avery, J.H. Benlekke, E.A. Chrischilles, J.A. Dosman, C. Duchaine, S.R. Kirkhorn, K. Kulhankova, J.A. Merchant. 2007. Health effects of airborne exposures from concentrated animal feeding operations. Environmental Health Perspectives 115, 298-302.

Jensen, A., 2002. Changing the Environment in Swine Buildings Using Sulfuric Acid. Transactions of the ASAE45:223-7.

Jerez, S., B., S. Mukhtar, W. Faulkner, Casey, K., D., M.S. Borhan, R.A. Smith. 2011. Evaluation of Electrostatic Particle Ionization and BioCurtain[TM] Technologies to Reduce Air Pollutants from Broiler Houses. Proceedings of the 2011 ASABE Annual International Meeting. Louisville, Kentucky, August 7-10, 2011. Just, N., C. Duchaine, B. Singh. 2009. An aerobiological perspective of dust in cage-housed and floor-housed poultry operations. Journal of Occupational Medicine and Toxicology. London, England. 4:13.

Lim, T.T., C. Wang, A.J. Heber, J. Ni, L. Zhao, S.H. Hanni. 2008. Effects of Electrostatic Space Charge System on Particulate Matter Emission from High-Rise Layer Barn. Proceedings of the 2008 ASABE Annual International Meeting. Providence, Rhode Island, June 29-July 2, 2008.

Lippmann, M., 2007. Health effects of airborne particulate matter. The New England Journal Of Medicine 357, 2395-7.

Manuzon, R.B. 2012. Electrostatic Precipitation Technologies for the Mitigation of Particulate Matter Emissions from Poultry Facilities. Doctoral dissertation. The Ohio State University Library, Columbus, OH, 243.

Manuzon, R.B., Zhao, L., 2009. Laboratory Evaluation and Modeling of Electrostatic Precipitation of PM Emissions from Poultry Buildings. ASHRAE Transactions 115, 831-49.

Melse, R.W., P. Hofschreuder, and N.W.M. Ogink. 2012. Removal of Particulate Matter (PM10) by Air Scrubbers at Livestock Facilities: Results of an On-Farm Monitoring Program. 55, 689-98.

Mitchell, B.W., L.J. Richardson, J.L. Wilson, Hofacre, C.L., 2004. Application of an electrostatic space charge system for dust, ammonia, and pathogen reduction in a broiler breeder house. Applied Engineering in Agriculture 20, 87-93.

PowderTechnologyInc., 2011. MSDS Arizona Test Dust. 5. Takai, H., and S. Pedersen. 2000. A comparison study of different dust control methods in pig buildings. 16, 269-77.

Zhao, Y., A.J.A. Aarnink, M.c.m.d Jong, N.W.M. Ogink, P.W.G.G Koerkamp. 2011. Effectiveness of Multi-Stage Scrubbers in Reducing Emissions of Air Pollutants from Pig Houses. Transactions of the ASABE54, 285-93.

Zhu, Z., Dong, H., Tao, X., Xin, H., 2005. Evaluation of airborne dust concentration and effectiveness of cooling fan with spraying mist systems in swine gestation houses. Proceedings of the 7th International Symposium on Livestozck and Environment, Beijing, China., 224-9.

Roderick Manuzon, PhD

Lingying Zhao, PhD

Member ASHRAE

Christopher Gecik

Roderick Manuzon is a graduate student, Lingying Zhao is an associate professor, and Christopher Gecik is a design engineer in the Department of Food, Agricultural, and Biological Engineering at Ohio State University, Columbus, OH.

Table 1. Complete Specification of the New Lab ESP Module

Parameter                             Value

Plate length (cm/in.)                 20/8
Plate spacing (cm/in.)                 5/2
Plate area                           619/0.66
  ([cm.sup.2]/[in.sup.2])
Total available collection area      7433/8
  ([cm.sup.2]/[in.sup.2])

Number of Plates                        7
Wire diameter (mm/in.)              1.27/0.04
Number of wires                        12
Inter electrode spacing (cm/in.)       5/2
Maximum operating voltage (V)        14,000
Maximum operating current               4
  (mA)

Table 2. CCD Design for Optimizing the Operating Voltage and Velocity
of the ESP (N = 3)

Codes                    Voltage (KV)   Velocity, m/s
(where 0 is center                        (ft/min)
point; a,A-are axial
points, +,- are levels
above and below the
center point)

a0                           9.4          1.5 (295)
--                            10           1 (196)
-+                            10           2 (393)
0a                           11.5         0.8 (157)
0                            11.5         1.5 (295)
0A                           11.5         2.2 (433)
+-                            13           1 (196)
++                            13           2 (393)
A0                           13.6         1.5 (295)

Table 3. Test Plan for Effects of Different Types and Levels of Dust
on ESP Performance (N= 3)

Dust Type   Dust Concentrations, mg/[m.sup.3]
            (x [10.sup.-8] lb/[ft.sup.3)

Poultry     2 (12.4)   4 (24.9)    9 (56.2)
Flour       1 (6.2)    3 (18.7)    7 (43.7)
Arizona     7 (43.7)   14 (87.4)   --

Table 4. Properties of Three Types of PM Tested with the ESP

PM Type        CMD, [micro]m   GSD, [micro]m      Density,
                  (in. x          (in. x        g/[cm.sup.3]
                [10.sup.-5]    [10.sup.-5])    (lb/[ft.sup.3])

Flour            2.1 (8.3)        2 (7.9)         1.49 (93)
Poultry dust     1.3 (5.1)       1.7 (6.7)        1.53 (96)
Arizona dust    2.9 (11.4)       1.8 (7.1)       2.65 (165)

PM Type        Relaxation         Source
                Time (s)

Flour           7.71E-4     Actual measurement
Poultry dust    7.65E-4     Actual measurement
Arizona dust    6.96E-3     (PowderTechnology-
                               Inc., 2011)

Table 5. Summary of Grade Efficiencies for the Six Hour ESP
Long-Term Test at 2 mg/[m.sup.3] PM Load at an Optimized Operating
Voltage of 13.6 KV and an Air Velocity of 1.7 m/s (334 ft/min)
Using Poultry Dust

Particle               Average     Standard     Maximum      Minimum
Size, [micro]m        Efficiency   Deviation   Efficiency   Efficiency
(x [10.sup.-5] in.)       %            %           %            %

0.5 (2)                  62.8        11.9         79.7         38.4
0.8 (3)                  72.5         8.7         91.9         40.3
1(4)                     77.9         7.7         99.4         39.1
2.3 (9)                   84          6.7         99.9         50.6
4.1 (16)                 83.4        14.1         100          28.3
10.4 (41)                 92          8.5         100          53.5

Table 6. Comparison between Mass and Count PM Sensor
Efficiency Measurements for Poultry Dust

Size Range     Sensor 1, %   Sensor 2, %   Difference, %

[PM.sub.1]        73.1          89.5            16
[PM.sub.2.5]      83.7          87.9             5
RESP              83.6          86.5             3
[PM.sub.10]       88.7          87.1             2

Table 7. Comparison of the Optimized ESP Performance with Other
Electrostatic PM Collection Devices for Animal Facilities

Study Author        Intended     ESP Type      Collection
                    PM Type                    Efficiency, %

                                               TSP    [PM.sub.10]

Optimized ESP        layer      wire-plate     89.4      88.7
George and Feddes    swine      wire-plate     91.1       --
  1995a
Mitchell and        broiler    space ionizer    61        --
  Hofacre 2004
Lim et al. 2008      layer     space ionizer    --        47
Cambra-Lopez        broiler        space                  36
  et al. 2009                     ionizer
Chai et al. 2009    poultry     wire-plate      78        --
Jerez et al. 2011   broiler    space ionizer    39        --

Study Author        Collection      Voltage    Air Speed,
                    Efficiency, %   Applied   m/s (ft/min)
                                     (KV)

                    [PM.sub.2.5]

Optimized ESP           87.9         13.6         1.7
George and Feddes        --          -12.1     0.95 (187)
  1995a
Mitchell and             --          -30.0       Still
  Hofacre 2004
Lim et al. 2008          --          -25.0       Still
Cambra-Lopez             --          -30.0       Still
  et al. 2009
Chai et al. 2009         55           -30      1.7 (334)
Jerez et al. 2011        --           -30        Still

Study Author          Ozone      Power, Watt per
                    Generation     [m.sup.3]/h
                      (ppb)         (hp/cfm)

Optimized ESP          <50        0.047 (1e-4)
George and Feddes      210         0.14 (4e-4)
  1995a
Mitchell and            --             --
  Hofacre 2004
Lim et al. 2008         --             --
Cambra-Lopez            --             --
  et al. 2009
Chai et al. 2009        --        0.002 (6e-6)
Jerez et al. 2011       --             --
COPYRIGHT 2014 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. (ASHRAE)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Manuzon, Roderick; Zhao, Lingying; Gecik, Christopher
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2014
Words:9451
Previous Article:Heat and moisture production of modern swine.
Next Article:Field & laboratory evaluation of advanced load monitoring boiler controllers.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |