ABSTRACT
ABSTRACT
The lack of an available particulate matter (PM) PM2.5 emission factor for almond harvesting operations has become a challenge for particulate matter regulations and emissions inventory in California. Low-dust harvesters are viewed as one of the strategies to reduce PM emissions and help achieve the state’s PM2.5 attainment targets. This paper evaluates the potential emission reduction from using low-dust harvesters compared to the conventional. Orchard boundary measurements of PM concentrations were collected to back-calculate emission rates using inverse dispersion modeling. Emissions from four low-dust harvesters (Flory 850, Exact E3800, Weiss-McNair 9800 and Jack Rabbit) were compared to those from a conventional harvester (Flory 480) in two orchards, located in the Fresno County. Emissions of PM2.5, PM10 and total suspended particles were observed to be lower for all new harvesters compared to the conventional harvester. The range of reductions varies from about 40% to 77% in PM2.5 emissions based on emission factors generated. The average ratio of PM10 to PM2.5 emissions is about 12.5%. The results of these tests imply that these new low-dust harvesters are capable of reducing PM emissions without affecting product quality. Therefore, the San Joaquin Valley Air Pollution Control District should consider including the use of these new harvesters in the conservation management practices (CMP) for the reduction of PM emissions in the valley.
Implications: The results of this research indicate that almond low-dust harvesters could potentially reduce PM emissions over traditional harvesters without any negative effect on product quality. Therefore, the use of these new harvesters should be considered as part of almond best management practices and updating of emissions inventory in the San Joaquin Valley.
Introduction
California almond industry produces 80% of the world’s almonds (ABC 2017). In 2016, the Almond Board of California (ABC) reported a total harvest of 936 Gg of almonds from approximately 364,200 ha of farmlands. Over 70% of these farmlands are located within the San Joaquin Valley (SJV) Air Pollution Control District. An estimated of 12 Gg of PM10 emissions was accounted to almond harvesting operations. The SJV was once considered as a serious non-attainment area for PM10 based on the U.S. National Ambient Air Quality Standards (NAAQS) until a compulsory Conservation Management Practice (CMP) program was implemented. The whole SJV air shed was declared in compliance with PM10 NAAQS in 2008.
The ABC has been aggressively involved in developing abatement strategies toward PM emission reductions from almond harvesting (Downey, Giles, and Thompson 2008; Faulkner and Capareda 2012; Faulkner et al. 2011; Goodrich et al. 2009). These studies have indicated potential significant reductions in PM emissions, resulting into the development of the current California Air Resource Board (CARB) PM10 emission factor (EF) of 3,500 kg/km2 (31.2 lbs/acre) from the original value of 4,570 kg PM10/km2 (40.8 lbs/acre) for all almond harvesting operations. Out of the total PM10 EF, 90% are attributed to nut pickup, with the remaining 10% being attributed to shaking and sweeping operations (Faulkner 2013). Aside from changes in sweeping and harvesting operations, there were also a number of published works involving machine design modifications for reducing PM emissions (Southard et al. 1997; Whitelock et al. 2007).
In 2010 and 2011, a series of orchard tests were conducted in Kern County, CA to investigate emissions of “low-dust” almond harvesters (Faulkner 2013). The low-dust harvesters were provided by the three major almond machine manufacturers in California: Flory Industries (Salida, CA), Weiss-McNair (Chico, CA) and Exact Harvesting Systems (Modesto, CA). The concentrations of TSP, PM10 and PM2.5 were measured at the edge of an orchard and emission rates were back-calculated using inverse dispersion modeling. It was reported that emissions of TSP and PM10 trended lower for all low-dust harvesters. The study has also shown a promise in significantly reducing PM2.5 from two new harvesters, which is vital in developing a PM2.5 emission factor for almond operations. The use of old harvesting equipment coupled with large acreage of almond orchards could possibly contribute significantly to PM2.5 emissions inventory (ABC 2017). The use of low-dust harvesters is viewed as a potential CMP in the SJV.
This study is a continuation of the previous efforts of ABC and CARB to develop a PM2.5 emission factor upon considering CMPs such as the adaptation of low-dust almond harvesters. Several machine manufacturers voluntarily provided their new harvesters during the PM emissions testing. These are the Flory 860 and 8550 models (Salida, CA), E3800 Exact harvester (Modesto, CA), 9800 California Special by Wiess-McNair (Chico, CA) and the newcomer Jack Rabbit harvester (Ripon, CA). The Flory 480 harvester was used as the control machine similar to that of the previous study (Faulkner 2013).
An updated and comparative PM emission factor is necessary in order to implement an appropriate incentive program for reducing PM emissions using low-dust harvesters. The results of this study could be utilized in generating a PM2.5 emission factor based on the improved and non-conventional harvesting machinery systems.
The objectives of this study were to:
Compare any measurable reductions in PM emissions between a conventional harvester (Flory 480) and a new low-dust emission harvester.
Compare the collection efficiencies between the conventional and new harvester.
Generate a ratio of the particle size range of PM10 and PM2.5 based on Federal Reference Method (FRM) samplers.
Methodology
Sampling location
PM emission measurements were conducted in two orchards in the Fresno County near Riverdale, CA in October 2017. Both orchards were planted in a subsurface drip-irrigated sandy loam soil, with general composition of sand (63%), silt (20%) and clay (17%), as determined using a hydrometer (Soil, Water and Forage Testing Laboratory: Texas A&M Agrilife Extension 2018). Almond trees were estimated to be about 15 years old and were planted in approximately 400-m (0.25 mi) rows oriented in a north-south direction. The first orchard has a total area of 28 ha (70 ac) while the second orchard has a total area of 15 ha (37 ac). In both orchards, there is a 7.3 m (24 ft) spacing between rows and 6 m (20 ft) between trees in the same row.
An orchard is typically planted with a combination of almond varieties to achieve cross-pollination. A nonpareil cultivar is usually mixed with a “pollinator” cultivar or with two “pollinator” cultivars. For both orchards where sampling was conducted, the nonpareil cultivars are planted every other row with two cultivars, Monterey and Fritz, on an alternating basis.
Experimental design
During a weeklong almond harvesting, emissions from the low-dust harvesters and those from the traditional harvester were measured. Four low dust-emission harvesting machines from different manufacturers (Flory 850, Exact E3800, Weiss-McNair 9800 and Jack Rabbit) were tested for their PM emissions. A control run using the Flory 480 conventional harvester was used in between low dust machinery runs. The tractors used to pull the harvesters were operated at a ground speed close to 4.8 km/h (3 mi/h).
Manufacturers approved the use of Flory 480 as the control machinery prior to emissions sampling. A code was assigned for each harvester in order to generate a randomized order of runs per replicate. There were four treatments (harvester) represented by letters A, B, C and D. The control runs are those with small letter “c” attached before the treatment letters. There were three replicates for a total of 21 tests (Table 1).
Table 1. Order of runs.
Each test plot consisted of 9 tree rows. To facilitate ease of harvester turn-around, only the trees on the first, fourth and ninth rows were considered. The windrows on both sides of the tree row in consideration were used for pick-up. Therefore, emissions were evaluated during the nut pickup of six windrows for each test plot.
Emissions calculations
The data collection and analyses for this study followed a similar approach based on the previous almond harvest emission studies (Faulkner 2013; Faulkner and Capareda 2012; Faulkner et al. 2011). In summary, there were three sets of samplers (S1, S2 and S3) deployed on the downwind and one set of samplers (S4) on the upwind side of the sampling plot (Figure 1). Each sampler set consisted of collocated low-volume TSP and federal reference method (FRM) PM10 and PM2.5 samplers. These samplers were used to measure the ambient PM concentrations during the nut pick up operations. Each downwind sampler was positioned directly perpendicular to the tree rows used for pick-up, with a distance of approximately 8.5m (28 ft) from the edge of the plot.
Published online:
09 September 2019Figure 1. Sampler configuration (not to scale) 28 ft. away from edge of sampling plot; S1-S3 represent downwind samplers; S4 represents upwind samplers.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/uawm20/2019/uawm20.v069.i11/10962247.2019.1655500/20210524/images/medium/uawm_a_1655500_f0001_b.gif"})
Figure 1. Sampler configuration (not to scale) 28 ft. away from edge of sampling plot; S1-S3 represent downwind samplers; S4 represents upwind samplers.
PM10 and PM2.5 concentration measurements were conducted using FRM samplers. Each set was equipped with a FRM PM10 sampling inlet (model PQ100 inlet; BGI Inc.; Waltham MA) and a Very Sharp Cut Cyclone (VSCCA, BGI, Inc.; Waltham, MA) for PM2.5 collection. TSP concentrations were also measured alongside the FRM samplers using a TAMU sampler designed by Wanjura et al. (2005). The programmable FRM samplers operate at a controlled flowrate (accuracy ± 0.50%) of 16.7 L/min. The preparation of Whatman filters, PM recovery and post-weighing were all based on the Standard Operating Procedure for PM Gravimetric Analysis (U.S. EPA 2008).
Dispersion modeling often involve pollutants assumed to be aerodynamically spherical particles (U.S. EPA 2005). The particles collected were predominantly slightly aspherical in shape (Figure 2), as determined by a Vega 3 TESCAN electron microscope (TAMU MIC: College Station, TX) with a particle density of 2.3 g/cm3 determined by a pycnometer (AccuPyc 1330, Micrometrics, Norcross, GA).
Published online:
09 September 2019![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/uawm20/2019/uawm20.v069.i11/10962247.2019.1655500/20210524/images/medium/uawm_a_1655500_f0002_b.gif"})
Figure 2. Scanning electron microscope image of collected particles.
In order to estimate the harvester PM emissions, in terms of kg/km2, a back-modeling approach (Faulkner 2013; Faulkner and Capareda 2012; Goodrich et al. 2009) was used using the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) by Lakes Environmental (Ontario, Canada). AERMOD is a steady-state plume model that incorporates flow and dispersion to estimate downwind pollutant concentrations from a given pollutant source even in complex terrains. The horizontal dispersion of pollutants throughout the planetary boundary layer (PBL) is assumed to follow Gaussian distribution. The vertical dispersion in the convective boundary layer (CBL) is described using a bi-Gaussian probability density function (U.S. EPA 2004). The measured downwind net PM concentrations were used to estimate back the ground-level area emissions coming from each test plot. Each sampler downwind concentration provided independent estimates of PM2.5, PM10 and TSP emission factors such that for each plot, three sets of emission factors can be calculated based from S1, S2 and S3 sampler sets (three PM2.5 EF, three PM10 EF and three TSP EF). These measurements were considered repeated emission factor estimates of a given plot.
Meteorological instruments were also deployed in an open area, 10-m away upwind of the orchard. A two-dimensional WM-350 anemometer (Weatherhawk Instruments, Logan, UT) was used to measure wind speed (accuracy ± 3%) and direction (accuracy ± 3%) 4m above the ground surface. A barometric pressure sensor (MetOne Instruments; accuracy ± 0.30%) was used, recorded every 5 min. A temperature (accuracy ± 0.16°C) and relative humidity (accuracy ± 2%) probe mounted in a solar radiation box at 2 m (BGI, Mesalabs NJ) was used, recorded every 5 min. Additional meteorological parameters were derived from available meteorological data provided by the National Oceanic and Atmospheric Administration (NOAA) – National Climatic Data Center (NCDC) and Earth System Research Laboratory (ESRL). The surface hourly data were based from Leemore Naval Air Station, 30 kilometers away from the sampling site. To compensate for missing data, 5-minute Automatic Surface Observing System (ASOS) weather data were processed in the AERMINUTE (a program within the AERMOD). The upper air data were based from the Oakland weather station as recommended by the California Air Resources Board (CARB 2006). Average moisture was assumed during the sampling period.
Additional meteorological and surface parameters were estimated according to the U.S. EPA guidance for AERMOD (U.S. EPA 2018). Surface characteristics require parameters such as albedo, Bowen ratio and surface roughness. Albedo is the fraction of radiation reflected by a surface; Bowen ratio describes the relationship between the sensible heat flux to the evaporation heat flux; and surface roughness is described as the height above the ground at which zero horizontal wind speed is achieved (U.S. EPA 2004). Surface characteristics and meteorological observations are input to the AERMET (a meteorological data preprocessor for AERMOD) to construct similarity profiles of the PBL parameters on a local scale.
In order to estimate the surface parameters, the sampling site was divided into six sectors, one-kilometer in radius around the orchard. The site land use map was based from the National Land Cover Database (NLCD), with 90m resolution, and generated using the AERSURFACE (surface parameter estimator of AERMOD). The surface parameters were defined per sector based on EPA guidelines (U.S. EPA 2018) shown in Table 2.
Table 2. Surface parameters used for AERMOD.
Size fractionation
Ten-liter resealable plastic containers were used to collect samples of almonds and foreign materials from each harvested windrow for each run. Sampling was done before and after nuts were picked up. The sample of almonds and foreign materials collected from the load-out stream as it entered the hopper bottom trailer was used for size fractionation. Every sample was sealed and transported to the BETA laboratory, Texas A&M University for mechanical fractionation.
The samples were placed in a sieve series in accordance to the American Society for Testing and Materials (ASTM) Standard C 136–06 (ASTM 2006) after the nuts were manually separated. The materials retained on each sieve series were collected and weighed to establish the mass fraction. The size ranges used as fractionation categories are shown in Table 3.
Table 3. Fractionation categories.
An analysis of variance (ANOVA) test was conducted using JMP Statistical Software (JMP v. 13; SAS Inc.; Cary, NC) to determine whether significant differences existed in the composition of products delivered to the huller between the new and conventional harvesters. The null hypothesis tested was that the mean of each mass percentage for every size range between the two harvesters were equal.
Results and discussion
Collection efficiency
The results of the mechanical size separation of the harvested samples are shown in Tables 4–7. Soil content in the harvested sample coming from conventional harvester was relatively higher compared to that of the new harvesters. However, no pairwise differences were found significant between traditional and new harvesters as revealed by ANOVA results at α = 0.05 (Machine A, p = .09; Machine B, p = .18; Machine C, p = .58; Machine D, p = .09). In actual practice, the differences in soil content are considered functionally insignificant. Most California almond processors would not penalize growers due to soil content present in the harvested almond samples, whether from using a conventional or a low-emission harvester.
Table 4. Size separation results (Machine A).
Table 5. Size separation results (Machine B).
Table 6. Size separation results (Machine C).
Table 7. Size separation results (Machine D).
Emission calculations
The summary of the processed meteorological conditions are shown in Tables 8–11. Emission reductions were calculated using ambient PM concentration measurements and inverse dispersion modeling. The emissions calculated were screened based on prevailing wind direction and location of the plume relative to the downwind samplers. Emissions were further streamlined by benchmarking the results with that of the previous low-emission harvester study conducted in 2010–2011 (Faulkner 2013).
Table 8. Meteorological parameters used for emissions calculation (Machine A).
Table 9. Meteorological parameters used for emissions calculation (Machine B).
Table 10. Meteorological parameters used for emissions calculation (Machine C).
Table 11. Meteorological parameters used for emissions calculation (Machine D).
Tests 1 and 2 results were omitted since the prevailing average winds as soon after the tests were started, were mainly blowing from the southwest such that the samplers were no longer downwind of the test plots. Emissions calculated based on S2 and S3 of Test 3 was also not included. During this period, the wind direction shifted such that the samplers were no longer downwind and away from the plume center. Samplers positioned at the edge of the plume might produce some uncertainties during back-calculation of emission rates (Faulkner 2013).
Tests 13 and 14 were omitted from analysis due to extremely high calculated emission rates. During this period, extremely stable ambient conditions were recorded which would typically result in an unusual estimate of emissions (Faulkner 2013). The prevailing wind direction during Tests 18 and 19 was observed to be coming from southeast such that the samplers were no longer downwind of the orchard. Tests 15 was also omitted due to extremely high calculated emission rates which was way beyond three standard deviations from the mean emission rate for the particular test. Test 21 was omitted since the samplers were no longer downwind of the orchard (prevailing winds coming from southwest).
PM emission factors
After these samples were omitted, the data were tested for goodness of fit (Shapiro-Wilk Test) to ensure that ANOVA assumptions are still valid.
A square-root transformation was applied to TSP emission rates and log transformations for PM10 and PM2.5 emission rates in order to fit the calculated emission rates into a normal distribution. The calculated average emission rates from inverse dispersion modeling were reported in the form of emission factors (kg/km2) as shown in Table 12.
Table 12. Almond harvesting emission factors (kg/km2) at two different periods (2010/2011 and 2017) using low-emission and conventional harvesters.
Differences within new harvesters were found to be not significant (α = 0.10) for any of the particle sizes (TSP, p = .79; PM10, p = .46; PM2.5 = 0.53), indicating that the calculated emissions based from the three sets of downwind samplers did not vary significantly. However, differences in emissions between the new and conventional harvesters were significant for only some of the particle sizes analyzed.
For Machine A, upon comparison with the Flory 480 (conventional), calculated emission rates were significant at a given 90% confidence level for TSP (p = .039) and PM10 (p = .018); however, not significant for PM2.5 (p = .23).
For Machine B, upon comparison with the Flory 480 (conventional), calculated emission rates were significant at a given 90% confidence level for PM10 only (p = .026); however, not significant for PM2.5 (p = .21) and TSP (p = .1079).
For Machine C, upon comparison with the Flory 480 (conventional), calculated emission rates were significant at a given 90% confidence level for all particle size range analyzed, TSP (p = .013), PM10 (p = .059) and PM2.5 (p = .058).
For Machine D, upon comparison with the Flory 480 (conventional), calculated emission rates were significant at a given 90% confidence level for all particle size range analyzed, TSP (p = .029), PM10 (p = .013) and PM2.5 (p = .045).
For this study, the calculated emission factors have shown positive reductions for all particle size range analyzed upon comparison with that of the conventional harvester (Table 12). Machine A produced 77% fewer TSP emissions, 43% few PM10 emissions, and 41% fewer PM2.5 emissions compared to Flory 480. Machine B was able to reduce TSP emissions by 44%, PM10 emissions by 56% and PM2.5 emissions by 51%. Relatively higher emission reductions were calculated for Machine C (TSP = 68%, PM10 = 74% and PM2.5 = 62%) and Machine D (TSP = 54%, PM10 = 63% and PM2.5 = 57%).
The resulting emission factors for PM2.5 were fairly consistent with the previous study by Faulkner (2013) in an orchard in Kern County, CA, with some slight improvements in the values of PM2.5 emission reductions (Table 12). Similar trends were also observed for PM10 reductions, except for a lower % reduction for Machine A compared to the previous study. For TSP, the % reductions were fairly consistent for both study periods, with a higher % reduction for Machine D from the present study.
Ratio of PM2.5 and PM10
The ratio of PM2.5/PM10 can provide vital information on the origin of particles, its formation, and possible health effects (Sugimoto et al. 2016; Xu et al. 2017). For regulatory purposes, identifying and quantifying PM2.5/PM10 can serve as a useful tool for retrospective prediction of PM2.5 without direct PM2.5 measurements, given the existing PM10 emission factor by the CARB.
Table 13 presents the ratios of PM2.5 to PM10 for each machinery as well as for all machineries tested. The calculated ratios were based from collocated measured concentrations of PM2.5 and PM10 from the FRM samplers. These individual ratios generated from the collocated samplers were then averaged. By arithmetic average based from the results of the valid test runs, the ratio between PM2.5 and PM10 was estimated to be around 12.5%.
Table 13. Calculated ratios of PM2.5 to PM10 based on FRM samplers.
Conclusion
The four “low dust” harvester emissions were compared to those of a conventional Flory 460 harvester. For both study periods, improvements in PM emissions for TSP, PM10 and PM2.5 were observed from 40% to 70% across four low dust harvesters relative to the conventional harvester. The average ratio between PM10 and PM2.5 emissions is about 12.5%. Fractionation analyses showed that there were no significant differences observed between the production composition collected from the harvester, and to be delivered to the huller between the new and the conventional harvesters.
The presented results were conducted in the Fresno County while the previous one was from a single orchard in Kern County. Emission reductions could vary under different production areas and different types of management practices. However, the results of this present study validates the previous one and further implies that these new low-dust harvesters are capable of reducing PM emissions without affecting product quality.
In order to prepare and implement a new PM2.5 SIP in the valley, the SJVAPCD could consider phasing out older harvester machinery, while providing incentives for owning newer harvesters. PM2.5 reduction targets for the SIP could also be realized by aggressively adopting best management practices such as investing on paved surfaces in between orchards, shifting to one-harvest pass new varieties and proper machinery adjustments.
| Rep 1 | Rep 2 | Rep 3 |
|---|---|---|
| cA | D | B |
| A | cD | cB |
| cC | A | C |
| C | cA | cA |
| B | C | A |
| cB/cD | cC/cB | cA/cD |
| D | B | D |
| Sector | Albedo | Bowen | Surface Roughness | Category |
|---|---|---|---|---|
| 1 | 0.18 | 0.70 | 0.10 | Orchard |
| 2 | 0.18 | 0.70 | 0.10 | Orchard |
| 3 | 0.18 | 0.70 | 0.10 | Orchard |
| 4 | 0.18 | 0.70 | 0.10 | Orchard |
| 5 | 0.18 | 1.00 | 0.30 | Low Intensity Residential |
| 6 | 0.18 | 1.00 | 0.30 | Low Intensity Residential |
| Particle Size | ||
|---|---|---|
| Minimum | Maximum | Category |
| 18.5 mm | – | Nuts |
| 10 mm | 18.5 mm | Leaves, Small Twigs, Small Nuts |
| 2 mm | 10 mm | Leaves and Grasses |
| – | 2 mm | Soil |
| Size Separation Category | Mass fraction (%) | |
|---|---|---|
| Machine A | Conventional | |
| Nuts | 63.2 (2.0) x | 58.76(2.4) x |
| Leaves, Small Twigs, Small Nuts | 32.1 (2.0) x | 31.2 (0.7) x |
| Leaves and Grasses | 0.4 (0.1) x | 0.2 (0.1) x |
| Soil | 4.3 (3.9) x | 9.8 (1.9) x |
Notes. No significant differences were detected in means in the same row followed by the same letter (α = 0.05). Standard deviations are shown parenthesis.
| Size Separation Category | Mass fraction (%) | |
|---|---|---|
| Machine B | Conventional | |
| Nuts | 62.4 (3.1) x | 62.6 (2.4) x |
| Leaves, Small Twigs, Small Nuts | 29.9 (2.6) x | 25.9 (5.5) x |
| Leaves and Grasses | 1.4 (0.7) x | 0.2 (0.1) x |
| Soil | 6.3 (2.0) x | 11.3 (7.3) x |
Notes. No significant differences were detected in means in the same row followed by the same letter (α = 0.05). Standard deviations are shown parenthesis.
| Size Separation Category | Mass fraction (%) | |
|---|---|---|
| Machine C | Conventional | |
| Nuts | 64.1 (6.6) x | 62.0 (5.9) x |
| Leaves, Small Twigs, Small Nuts | 25.5 (3.9) x | 23.1 (3.9) x |
| Leaves and Grasses | 1.2 (0.9) x | 0.8 (0.7) x |
| Soil | 9.2 (10.8) x | 14.0 (9.2) x |
Notes. No significant differences were detected in means in the same row followed by the same letter (α = 0.05). Standard deviations are shown parenthesis.
| Size Separation Category | Mass fraction (%) | |
|---|---|---|
| Machine D | Conventional | |
| Nuts | 62.5 (2.0) x | 58.56 (2.8) x |
| Leaves, Small Twigs, Small Nuts | 31.6 (2.1) x | 31.2 (0.7) x |
| Leaves and Grasses | 0.6 (0.3) x | 0.2 (0.1) x |
| Soil | 5.3 (2.9) x | 9.98 (2.3) x |
Notes. No significant differences were detected in means in the same row followed by the same letter (α = 0.05). Standard deviations are shown parenthesis.
| Machine A | Conventional | |||||
|---|---|---|---|---|---|---|
| Min | Max | Ave | Min | Max | Ave | |
| Sensible Heat Flux (W/m2) | 152.1 | 196.4 | 167.5 | 86.9 | 174.77 | 127.8 |
| Albedo | 0.19 | 0.19 | 0.2 | 0.19 | 0.22 | 0.2 |
| Wind Speed (m/s) | 0.50 | 5.09 | 2.8 | 0.68 | 6.11 | 2.9 |
| Temperature, K | 295.4 | 300.9 | 298.7 | 293.8 | 300.9 | 297.2 |
| Relative Humidity, % | 20 | 26 | 22.3 | 22 | 25 | 23.3 |
| Machine B | Conventional | |||||
|---|---|---|---|---|---|---|
| Min | Max | Ave | Min | Max | Ave | |
| Sensible Heat Flux (W/m2) | 139.9 | 164 | 151.6 | 88.1 | 175.1 | 121.7 |
| Albedo | 0.19 | 0.2 | 0.2 | 0.19 | 0.24 | 0.2 |
| Wind Speed (m/s) | 0.50 | 2.1 | 2.0 | 1.77 | 2.33 | 2.1 |
| Temperature, K | 297 | 304.2 | 300.9 | 294.9 | 307.5 | 302.1 |
| Relative Humidity, % | 11 | 15 | 13.7 | 8 | 24 | 15.0 |
| Machine C | Conventional | |||||
|---|---|---|---|---|---|---|
| Min | Max | Ave | Min | Max | Ave | |
| Sensible Heat Flux (W/m2) | 112.3 | 164 | 138.1 | 101.8 | 165.8 | 123.1 |
| Albedo | 0.19 | 0.2 | 0.2 | 0.18 | 0.22 | 0.2 |
| Wind Speed (m/s) | 1.35 | 5.8 | 2.9 | 0.50 | 2.11 | 1.9 |
| Temperature, K | 302.5 | 304.2 | 303.6 | 294.9 | 300.4 | 296.7 |
| Relative Humidity, % | 7 | 18 | 13.7 | 19 | 24 | 22.3 |
| Machine D | Conventional | |||||
|---|---|---|---|---|---|---|
| Min | Max | Ave | Min | Max | Ave | |
| Sensible Heat Flux (W/m2) | −22 | 151.7 | 83.3 | 88.1 | 198.9 | 159.4 |
| Albedo | 0.19 | 0.74 | 0.4 | 0.19 | 0.24 | 0.2 |
| Wind Speed (m/s) | 0.79 | 5.39 | 3.3 | 0.66 | 4.34 | 2.7 |
| Temperature, K | 295.9 | 304.2 | 300.0 | 298.8 | 307.5 | 303.1 |
| Relative Humidity, % | 8 | 36 | 23.0 | 11 | 32 | 18.7 |
| 2017 Sampling | 2010–2011 Sampling | |||||
|---|---|---|---|---|---|---|
| Machine | TSP | PM10 | PM2.5 | TSP | PM10 | PM2.5 |
| A Control (cA) | 2,153 9,360 | 492 864 | 326 551 | 1,149 4,835 | 561 1,981 | 264 401 |
| % Reduction | 77%* | 43%* | 41% | 76% | 72% | 34% |
| B Control (cB) | 1,590 2,820 | 747 1,706 | 225 456 | 2,281 3,891 | 1,034 1,599 | 359 280 |
| % Reduction | 44% | 56%* | 51% | 41% | 35% | −28% |
| C Control (cC) | 2,911 9080 | 1,360 5,200 | 108 281 | -- | -- | -- |
| % Reduction | 68%* | 74%* | 62%* | |||
| D Control (cD) | 2,643 5,748 | 1,530 4,100 | 371 855 | 5,095 6,865 | 1,453 2,628 | 121 313 |
| % Reduction | 54%* | 63%* | 57%* | 26% | 45% | 61% |
Notes. Machine C was not able to participate during the 2010–2011 emission tests.
*Reductions were significant at α = 0.10
| Machine | Average from all replicates (%) |
|---|---|
| A | 14.4 |
| cA | 15.7 |
| B | 17.9 |
| cB | 8.3 |
| C | 8.4 |
| cC | 10.7 |
| D | 15.9 |
| cD | 11.5 |
| Overall Average (%) | 12.5 |
Acknowledgment
The authors would like to express their gratitude to the Almond Board of California as the major funding entity, the San Joaquin Valley Air Pollution Control District for the additional funding and modeling guidance, Matthew Efird and Roger Isom for providing the orchard and harvester operators and to the BETA Lab Crew of Texas A&M University for providing the manpower during sampling.