Accepted Manuscript

Reducing cyclone pressure drop with evases P.A. Funk

PII: S0032-5910(14)01001-8

DOI: doi: 10.1016/j.powtec.2014.12.019

Reference: PTEC 10680

To appear in:

Received date: Revised date:

Powder Technology

27 June 2014 4 December 2014

Accepted date: 11 December 2014

Please cite this article as: P.A. Funk, Reducing cyclone pressure drop with evasés, Powder Technology (2014), doi: 10.1016/j.powtec.2014.12.019

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title:

Reducing cyclone pressure drop with evases

Author:

P. A. Funk1

:USDA Agricultural Research Service, Southwestern Cotton Ginning Research Lab, 300 E. College Dr., PO Box 578, Mesilla Park, NM 88047, USA paul.funk@ars.usda.gov Corresponding author: Paul Funk (575) 526-6381; FAX: (575) 525-1076

Abstract

Cyclones are widely used to separate particles from gas flows and as air emissions control devices. Their cost of operation is proportional to the fan energy required to overcome their pressure drop. Evasés or exit diffusers potentially could reduce exit pressure losses without affecting collection efficiency. Three rectangular evasés and a radial evasé with a variable opening were tested on two cyclones. Pressure drop was recorded for inlet velocities from about 10 to 20 m s-1. The radial evasé reduced cyclone pressure drop by between 8.7 and 11.9 percent when its exit area was equal to the flow area of the cyclone vortex finder or gas exit. A simple payback based on avoided energy costs was estimated to be between 3600 and 5000 hours, not including installation cost.

Keywords

Cyclone, Energy recovery, Evasé, Pressure regain, Pressure drop, Return on investment

1. Introduction

1.1 Rationale

Cyclone separators are widely used to separate particles from gas flows and as air emission control devices in the chemical, metals, mining, petroleum, pharmaceutical and processing industries. Cyclones have low capital costs compared to other control devices. With no moving parts, cyclones intrinsically have low maintenance costs. The operating cost of a cyclone is the cost of the energy required to overcome its pressure drop. In the cotton ginning industry energy costs have risen more than other inputs in recent years [1]. Other industries likely face similar financial pressure. An evasé (â'Vâ-zâ') is a diffuser located at the exit of a duct or fan. An evasé on a cyclone exhaust was not expected to have a significant impact on particle collection, since it is located after the separation region. The goal of this research was to estimate the potential economic return of using rectangular and radial evasés at cyclone outlets by quantifying pressure drop reductions and calculating corresponding energy savings.

The benefit of an evasé may be twofold; first, it may reduce the exit loss and second, it protects the inside of the cyclone from weather. An evasé could be of a rectangular (two or three dimensional) design, or of a radial design (Fig. 1). With either option, the smooth transition in the flow area of the expansion region converts the gas

kinetic energy into pressure energy, minimizing exit losses. Theoretically, cyclone operating energy may be reduced by the amount of pressure energy regained [2].

Figure 1

1.2 Antecedents

There has been considerable research conducted on cyclone pressure drop, both empirical and computational. The focus of past research has frequently been on the dimensions of the cyclone [3-5], or occasionally on modifications to the cyclone inlet or particle outlet [6, 7], but no publications were found reporting research on the impact of evases on cyclone pressure drop. For empirical studies, this may be due to the practice of directing exhaust through a duct or filter to facilitate emissions quantification.

Attempts to predict and minimize cyclone pressure drop are numerous. Many semi-empirical pressure drop models have been proposed [8-14]. Additional work has also been done using computational fluid dynamics (CFD) [15-27], among many others, but none included modeling evases.

Early empirical research on devices that reduced cyclone pressure drop included an inlet vane that protruded into the cyclone cylinder [28]. When this inlet vane was shortened, pressure drop doubled [29]. In another study, vertical, horizontal and spiral grooves were cut into the cyclone cylinder wall. These altered the velocity profile, reducing pressure drop [30]. Another modification that reduced pressure drop was a stick inserted through the top of the cyclone, extending to the cone bottom. Empirical tests [31] and numerical simulations [32] indicated a decrease in pressure drop up to 37% with the stick. These modifications would not be appropriate when handling materials that tend to agglomerate. Inlet angle modifications have also been attempted [17], [33]. None of the aforementioned experiments included measuring pressure drop reduction through evases.

Three studies looked at cyclone pressure drop with gas outlet modifications. A numerical simulation and empirical test examined vortex finder length and diameter [34]. An empirical study conducted with tangential inlet cyclones indicated a reduction in pressure drop with vortex finders having a cone -shape [7]. A more recent empirical study of axial inlet cyclones also indicated a reduction in pressure drop with vortex finders having a cone-shape [35]. Though these latter two studies did not examine pressure drop in relation to evases specifically, the tested modifications to the shape of the vortex finder may have had a similar function in that flow area gradually increased, possibly resulting in partial static pressure regain. Unfortunately, this modification cannot be made to existing cyclones inexpensively.

Patents describing cyclones with evases and diffusers have been awarded throughout the last century [3638], but evases are not widely used at present. The common practice is for the gas outlet (exhaust, or vortex finder) to end just above the cyclone top (Fig. 2a), or to cover it with a shallow cone ("rain hat") to keep rain out (Fig. 2b). Though less common, some installations have covers (called spin caps) that direct exhaust away from prevailing winds or an adjacent structure (Fig. 2c). There may be regulatory as well as practical reasons for these various designs.

Figure 2

1.3 Cautions

In some jurisdictions, such as North and South Carolina, rain hats (and thus evases) would not be in compliance with existing regulations, at least for cyclones controlling cotton gin emissions, as permit authorities prohibit rain hats or require cyclone exhaust to be directed upwards [39, 40]. Currently these jurisdictions only allow exhaust flappers to keep rain and snow out of the cyclone. Although installing evases would be in violation of the existing rules, a petition for alternative controls may be filed, provided adequate justification is given. Though precedent is lacking, permission might be granted since reducing electrical energy consumed reduces the power plant contribution to pollution in the regional airshed [41]. Note that directing particulate upwards results in a greater effective source height. Increasing effective height increases the distance from the source where maximum concentration occurs and it reduces maximum ground level concentration [42]. Although local concentration is less, the time particles remain suspended and the distance they travel is more. In the U.S., industrial sources are to control emissions, not merely disperse them [43].

In South Carolina, air permit conditions stipulate monitoring pressure drop to verify proper cyclone operation. The rule is based on published values for pressure drop that result from operating 2D2D and 1D3D cyclones in a range bracketing their design inlet velocities [44]. The values, from 748 to 1495 Pa, were claimed to be a compromise between collection efficiency and power requirement (though empirical evidence has not been found in the literature). Since the purpose of the evase is to reduce pressure drop, adding evases would again require petitioning for a variance.

2. Materials and Methods

2.1 Apparatus

Two 30.48 cm diameter cyclones, A and B, were built according to the fully enhanced 1D3D design [45]. The cyclones' gas outlet, or vortex finder, was 15.24 cm in diameter. A sealed container (dust bin, Fig. 3) was attached to the particle outlet at the bottom.

Figure 3

A test apparatus was constructed that held one of the two cyclones. Air was supplied to the cyclone by a 5.6 kW centrifugal fan. The fan was controlled by a variable frequency drive (VLT 8000 AQUA, Danfoss, Nordborg, Denmark). A 10 cm diameter duct connected the fan to the cyclone. The duct included sections that provided flow straightening, temperature sensing, and velocity and pressure measurement functions (Fig. 3). Pitot-static velocity pressure, Venturi differential pressure, and cyclone pressure were indicated on inclined manometers and also were converted to 4-20 mA signals (Model 614, Dwyer Instruments, Inc., Michigan City, Indiana, USA). Signals were recorded continuously at 1-s intervals using a data logger (Model 34970A, with 34908A switch unit, Agilent Technologies, Santa Clara, California, USA). 2.2 Rectangular evasés

Three rectangular evasés with matching spin caps were tested. The length of an evasé is normally selected to balance pressure regain with friction losses. When designing evasés to fit on a cyclone, stresses caused by wind load and weight also require consideration. This favors a shorter evasé. Charts presenting plane (two dimensional) diffuser performance, Figs 2.29 to 2.31in Fan Engineering [2], guided the design of the three dimensional rectangular evasés used in this study. Evasés inlet dimensions were selected to match either the cyclone inlet or vortex finder flow area.

The first evasé had a 7.6 x 15.2 cm inlet, for an area equal to the cyclone inlet, 116 cm2 (Fig. 4a). The second and third evasés both had inlet areas equal to the cyclone outlet, 183 cm2. The second evasé was only slightly taller than it was wide, with a 12.1 x 15.2 cm inlet (Fig. 4b). The third evasé was much taller, with a 7.6 x 24.1 cm inlet (Fig. 4c). All three of the rectangular evasés had a flat top, and sides and a bottom that diverged 7.5° from parallel. All three extended from the spin cap until their outlet areas were approximately twice their inlet areas. This resulted in a length of 21.3, 29.2 and 23.8 cm for the first, second and third rectangular evasé, respectively.

Figure 4

Rectangular evases can only be attached to a cyclone by a spin cap. Analyses of spin cap flow dynamics were not found in the literature. Each spin cap had a cylindrical portion that was the diameter of the gas outlet, equal in height to its matching rectangular evase, with a collar extending 5 cm below to attach to the gas outlet. Each spin cap had one vertical wall that intersected the cylindrical portion radially. The first and third spin caps had an outside vertical wall that intersected the cylindrical portion tangentially, parallel to the radial wall. The second spin cap had a volute transition to the outside vertical wall, to accommodate its wider evase. To quantify the contributions of the spin caps, each rectangular evase spin cap was tested alone as well as with its evase. Each spin cap discharged air from the side complimenting the direction of rotation established by the cyclones.

2.3 Radial evase

Numerous publications reported radial diffuser research, primarily on turbo machinery (typically compressors or jet engines). These articles were considered somewhat applicable since turbo machinery and cyclone exhausts both present the radial diffuser or evase with a swirling flow. Since flow from cyclones used as air emissions control devices contain particulate, a vaneless diffuser would be preferred. The best vaneless designs appear to be ones with near constant flow area [46], or similarly, ones with a slightly restricted flow area [47]. The tested design had a flow area that was restricted at the throat, and that then expanded.

The radial evase, Fig. 5, was equal in diameter to the cyclone, 30.5 cm. In theory, an evase would need to shed water, to protect the cyclone from rain and snow. Therefore, the radial evase was designed so that the top had a negative slope of 7.5° and the bottom a negative slope of 5.0°. This slight linear convergence still resulted in an increasing flow area with increasing radius. The radial evase top was supported by threaded rod so that the space between the top and bottom could be adjusted. The exit opening (hi, at the terminus radius of 15.24 cm) was set to 1.11, 1.27, 1.67, 1.91, 2.22, 2.54, and 2.86 cm. The throat (h3, at a radius of 9.17 cm) was 0.27 cm greater than the exit opening (h4). The ratio of exit area to inlet area ranged from 1.34 to 1.52 for the smallest to the largest exit openings, respectively. The ratio of the inlet to the exit radii was a constant 1.66. Due to swirling flow, the radial evase flow length was about 6.28 cm (based on photos of tell-tails made through a clear plastic evase), for an approximate height to length ratio that varied from 0.20 to 0.48 for the smallest to the largest exit openings, respectively.

Figure 5

2.4 Procedures

Barometric pressure, air temperature and relative humidity were recorded and entered into a laptop computer to calculate local air density. All gages and instruments were zeroed. The data logger was initiated and the fan was operated at six power levels; first, for two minutes at each power level in descending order, then for two minutes at each power level in ascending order. The power levels were 5% increments, between 25% and 50% of full fan power, to obtain data over a range of cyclone inlet velocities. Inlet velocities were approximately 10 to 20 m s-1, depending on local air density.

In the case of the radial evase, the exit opening was set at the next level and the test series repeated. In the case of the rectangular evases, the evase was removed and the test series was repeated for the spin cap alone, then the spin cap was removed and the test series was repeated for the cyclone alone. The rectangular evases were tested at each fan setting on each cyclone; the radial evase was tested at each opening and at each fan setting twice on cyclone A, and once on cyclone B. 2.5 Analysis

At least 100 observations from the data logger record were averaged to obtain a velocity and cyclone pressure value for each combination of cyclone and evase at each fan setting. Velocity pressure and local air density at the time of the test were used to compute the value for cyclone true inlet velocity. Cyclone pressure drop (Pa) and inlet velocity (m s-1) values were plotted using a spreadsheet (Microsoft Excel 2010) and a second order polynomial curve with zero intercept was fit for each cyclone and evase combination. The coefficient of determination was used to evaluate how well each curve fit the data.

3. Results

3.1 Rectangular evase results

Second order polynomial curves fit the data well (in each case, R2 > 0.99). Results for each evase at representative velocities are shown in Table 1. Rectangular evase results were disappointing. With cyclone A there was a small reduction in cyclone pressure drop with rectangular evases (Fig. 6). At 1D3D cyclone design inlet velocity (16.26 m s-1) evases 1, 2, and 3 on cyclone A resulted in a decrease in pressure drop of 13, 40, and 0 Pa, a potential energy savings of 1.5, 4.9, and 0%, respectively. With cyclone B, pressure drop increased slightly with the addition of rectangular evases (Fig. 7). At design inlet velocity evases 1, 2, and 3 on cyclone B resulted in an increase in pressure drop of 18, 16, and 23 Pa, a potential energy loss of 2.2, 2.0, and 2.8%, respectively. It is not

known why cyclone B had a different response than cyclone A. Dimensions were similar, but cyclone B had more internal roughness. Thus cyclone B may have had greater turbulence, and this may have caused flow in the rectangular evases to separate from the wall, preventing velocity recovery.

Table 1 Figure 6 Figure 7

Results also differed by cyclone for the spin caps alone. With cyclone A, spin caps 1, 2, and 3 conferred a pressure drop penalty of 5.4, 0, and 6.6% respectively. With cyclone B, spin caps 1, 2, and 3 conferred a pressure drop penalty of 9.4, 5.0, and 4.5% respectively, at design inlet velocity.

3.2 Radial evase results

Radial evases decreased pressure drop for both cyclones. For cyclone A, the greatest reduction in cyclone pressure drop was with a separation between the top and bottom of 1.67 cm at the radial evase exit (h^). This resulted in a pressure drop reduction of 97 Pa (11.9%) at the design inlet velocity of 16.26 m s-1 (Fig. 8). For cyclone B the greatest reduction in cyclone pressure drop was with an evase exit separation height of 1.91 cm. This resulted in a pressure drop reduction of 70 Pa (8.7%) at 16.26 m s-1 (Fig. 9). Interestingly, the flow area in the vortex finder was 182 cm2, and at an evase exit separation height of 1.91 cm the exit flow area was 182 cm2.

Figure 8 Figure 9

3.3 Return on investment calculation

What would be the potential return on investment if such an apparatus were to be installed? Potential savings in air power, Pair, is the product of volumetric air flow, V, and pressure reduction, Ap, Eq. 1, where the volumetric air flow is determined by the size of the cyclone and its inlet velocity:

APair (W) = V (m3 s-1) * Ap (Pa) (1)

In this example, a 1.52 m diameter 1D3D cyclone with a D/2 high by D/4 wide inlet operating at a design inlet velocity of 16.26 m s-1 has been assumed, resulting in a volumetric airflow, V, of 4.72 m3 s-1. The potential electrical power savings, APelectricity, Eq. 2, is a function of fan efficiency and motor efficiency:

APelectricity (W) = APair (W)/(f * ^motor) (2)

where the divisor f is the fan efficiency and nmotor is the motor efficiency, here assumed to be 55% (centrifugal fan) and 95% (high efficiency motor), respectively. A simple payback period in hours can be calculated with Eq. 3 if the cost of an evase and the cost of electricity are known:

Payback (h) = Costevase ($) / [ APelectncity (kW) * Costdectriclty ($ kWh-1)] (3)

For this example published values from the cotton ginning industry were used; the average cost of electricity per bale, $3.79, reported for 2010 [48] was divided by the average kWh per bale, 34.5, [49] to arrive at an average cost of $0.11 kWh-1. The cost of obtaining a radial evase for a 1.52 m diameter cyclone was estimated to be $350 (W & R Industrial Services, Brownfield, TX), not including installation. This example calculation resulted in a simple payback of 3,614 to 5,000 hours depending on pressure drop reduction (97 or 70 Pa for a radial evase on cyclone A or B, respectively). Assuming 75 days operation at 24 hours per day, the modification would pay for itself in between two and three seasons (Table 2). Assuming continuous operation, the simple payback would be 5 to 7 months. Installation cost was not included because it was difficult to estimate accurately; the cost could be significant if the work were performed by an outside contractor on an existing cyclone, or negligible if the evase and cyclone were installed at the same time.

4. Conclusion

This research has shown that the pressure drop in a cyclone can be reduced by between 8.7 and 11.9% with the addition of a radial evase. Because the evase is outside of and after the cyclone, it was assumed that particulate separation would not be affected. Since the evase alone is inexpensive, the estimated simple payback period, not including installation cost, is between 2 and 3 seasons for a facility operating 24 hours a day for 75 days each season. For a year-round operation, the return on investment would be within 5 to 7 months. It is important to check the air permit rules governing a facility when considering a radial evase. If the facility is located in a jurisdiction where cyclone exhausts must be directed vertically, or where pressure drop is used as a measure of correct cyclone operation, the modification would not be allowed unless appropriate variances could be obtained.

Acknowledgments

The author appreciates W & R Industrial Services, Brownfield, TX, for assistance in estimating evase costs.

References

[1] P. Funk and R. Hardin IV, "Cotton gin electrical energy use trends and 2009 audit results," Applied Eng. in Agric., vol. 28, no. 4, pp. 503-510, 2012.

[2] R. Jorgensen, Ed., Fan Engineering, Eighth Edition, Buffalo, New York: Buffalo Forge Company, 1983.

[3] K. Elsayed and C. Lacor, "Modeling and Pareto optimization of gas cyclone separator performance using RBF type artificial neural networks and genetic algorithms," Powder Technology, vol. 217, pp. 84-99, 2012.

[4] K. Elsayed and C. Lacor, "CFD modeling and multi-objective optimization of cyclone geometry using desirability function, artificial neural networks and genetic algorithms," Applied Mathematical Modeling, vol. 37, pp. 5680-5704, 2013.

[5] P. Funk, G. Holt and D. Whitelock, "Novel cyclone empirical pressure drop and emissions with heterogeneous particulate," Journal of Aerosol Science, vol. 74, pp. 26-35, 2014.

[6] R. Baker, M. Gillum and S. Hughs, "Preseparators and cyclones for the collections of stripper cotton trash," Trans. ASABE, vol. 38, no. 5, pp. 1335-1342, 1995.

[7] G. Holt, R. Baker and S. Hughs, "Evaluation of static pressure drops and PM10 and TSP emissions for modified 1D-3D cyclones," Trans. ASABE, vol. 42, no. 6, pp. 1541-1547, 1999.

[8] M. Lissman, "An analysis of mechanical methods of dust collection," Chemical and Metallurgical Engineering, vol. 37, no. 10, pp. 630-634, 1930.

[9] C. Stairmand, "Pressure drop in cyclone separators," Engineering, p. 409, 21 Oct. 1949.

[10] W. Barth, "Design and layout of the cyclone separator on the basis of new investigations.," Brennst. Warme Kraft, vol. 8, pp. 1-9, 1956.

[11] J. Dirgo and D. Leith, "Cyclone collection efficiency: comparison of experimental results with theoretical predictions," Aerosol Sci. and Tech., vol. 4, pp. 401-415, 1985.

[12] H. Mothes and F. Loffler, "Prediction of particle removal in cyclone separators," International Chemical Engineering, vol. 28, pp. 231-240, 1988.

[13] E. Muschelknautz and M. Trefz, "Pressure drop and separation efficiency in cyclones," VDI-HeatAtlas, pp. 18, 1992.

[14] B. Zhao, "A theoretical approach to pressure drop across cyclone separators," Chemical Engineering and Technology, vol. 27, no. 10, pp. 1105-1108, 2004.

[15] W. Griffiths and F. Boysan, "Computational fluid dynamics (CFD) and empirical modelling of the performance of a number of cyclone samplers," Journal of Aerosol Science, vol. 27, pp. 281-304, 1996.

[16] P. Funk and S. Hughs, "Dust cyclone modeling," in Beltwide Cotton Conferences, Cordova, Tenn., 2000.

[17] P. Funk, S. Hughs and G. Holt, "Dust cyclone design," Applied Engineering in Agriculture, vol. 17, no. 4, pp. 441-444, 2001.

[18] W. Peng, A. Hoffman, P. Boot, A. Udding, H. Dries, A. Ekker and J. Kater, "Flow pattern in reverse-flow

centrifugal separators," Powder Technology, vol. 127, pp. 212-222, 2002.

[19] L. Hu, L. Zhou, J. Zhang and M. Shi, "Studies on strongly swirling flows in the full space of a volute cyclone separator," AIChE Journal, vol. 51, no. 3, pp. 740-749, 2005.

[20] J. Gimbun, T. Chuah, A. Fakhru'l-Razi and T. Choong, "The Influence of temperature and inlet velocity on cyclone pressure drop: a CFD Study," Chemical Engineering and Processing, vol. 44, pp. 7-12, 2005.

[21] T. Chuah, J. Gimbun and T. S. Choong, "A CFD study of the effect of cone dimensions on sampling aerocyclones performance and hydrodynamics," Powder Technology, vol. 162, no. 2, pp. 126-132, 2006.

[22] C. Cortes and A. Gil, "Modeling the gas and particle flow inside cyclone separators," Progress in Energy and Combustion Science, vol. 33, pp. 409-452, 2007.

[23] J. Derksen, H. Van Der Akker and S. Sundaresan, "Two-way coupled large-eddy simulations of the gas-solid flow in cyclone separators," AIChE Journal, vol. 54, no. 4, pp. 872-885, 2008.

[24] K. Elsayed and C. Lacor, "Optimization of the cyclone separator geometry for minimum pressure drop using mathematical models and CFD simulations," Chemical Engineering Science, vol. 65, pp. 6048-6058, 2010.

[25] K. Elsayed and C. Lacor, "Modeling, analysis and optimization of aircyclones using artificial neural network, response surface methodology and CFD simulation approaches," Powder Technology, vol. 212, pp. 115-133, 2011.

[26] K. Chu, B. Wang, D. Xu, Y. Chen and A. Yu, "CFD-DEM simulation of the gas-solid flow in a cyclone separator," Chemical Engineering Science, vol. 66, pp. 834-847, 2011.

[27] J. Chen, K. Chu, R. Zou, A. Yu, A. Vince, G. Barnett and P. Barnett, "How to optimize design and operation of dense medium cyclones in coal preparation," Minerals Engineering, p. in press, 2013.

[28] C. Shepherd and C. Lapple, "Flow pattern and pressure drop in cyclone dust collectors," Industrial and Engineering Chemistry, vol. 31, no. 8, pp. 972-984, 1939.

[29] C. Shepherd and C. Lapple, "Flow pattern and pressure drop in cyclone dust collectors," Industrial and Engineering Chemistry, vol. 32, no. 9, pp. 1246-1248, 1940.

[30] H. Kim, K. Lee and M. Kuhlman, "Exploratory design modifications for enhancing cyclone performance," Aerosol Science, vol. 32, pp. 1135-1146, 2001.

[31] L. Wang, Q. Yan and L. Liu, "Effect of a stick on the flow field in a cyclone and the pressure drop reduction mechanism," Aerosol Sci. Tech., vol. 35, no. 5, pp. 909-913, 2001.

[32] A. Gong and L.-Z. Wang, "Numerical study of gas phase flow in cyclones with the Repds," Aerosol Science and Technology, vol. 38, pp. 506-512, 2004.

[33] S. Bernardo, M. Mori, A. Peres and R. Dionisio, "3-D computational fluid dynamics for gas and gas-particle flows in a cyclone with different inlet section angles," Powder Technology, vol. 162, pp. 190-200, 2006.

[34] F. Kaya and I. Karagoz, "Experimental and numerical investigation of pressure drop coefficient and static pressure difference in a tangential inlet cyclone separator," Chemical Papers, vol. 66, no. 11, pp. 1019-1025,

[35] Z. Xiong, Z. Ji and X. Wu, "Development of a cyclone separator with high efficiency and low pressure drop in axial inlet cyclones," Powder Technology, vol. 253, pp. 644-649, 2014.

[36] J. H. Keenan, "Cyclone separator". USA Patent 2,033,470, 1936.

[37] H. C. Reeves, "Cyclone dust separator". USA Patent 2,193,883, 1940.

[38] Kilian, "Process and apparatus for scrubbing exhaust gas from cyclone collectors". USA Patent 3,989,485, 1976.

[39] SCDHEC, "Bureau of Air Quality, Registration Permit, Cotton Ginning Operations," 2011. [Online]. Available: http://www.scdhec.gov/environment/baq/docs/permitting/Registration_Permit-Cotton_Ginning_Operations.pdf.

[40] NCDENR, "15A NCAC 02D .0542 Control of particulate emissions from cotton ginning operations," 2008. [Online]. Available: http://daq.state.nc.us/rules/rules/D0542.pdf.

[41] P. Funk, "The environmental cost of reducing agricultural PM2.5 emissions," J. Air and Waste Management Assoc., vol. 60, no. 6, pp. 681-687, 2010.

[42] C. Cooper and F. Alley, Air Pollution Control - A Design Approach (2nd ed.), Prospect Heights, IL: Waveland Press, 1994.

[43] R. Schulze, Fundamentals of Air Dispersion Modeling, seminar notes, 21-22 May 2007, Minneapolis, MN, Dallas, TX: Trinity Consultants, 2007.

[44] W. Anthony and W. Mayfield, Cotton Ginners Handbook; Agricultural Handbook No. 503, Washington, DC: U.S. Government Printing Office, 1994.

[45] P. Funk and K. Baker, "Dust cyclone technology for gins - a literature review," J. Cotton Sci., vol. 17, no. 1, pp. 40-51, 2013.

[46] Z. Yingkang and S. A. Sjolander, "Effect of geometry on the performance of radial vaneless diffusers," Transactions of the ASME, vol. 109, pp. 550-554, 1987.

[47] A. Jaatinen, A. Gronman, T. Turunnen-Saaresti and P. Roytta, "Effect of vaneless diffuser width on the overall performance of a centrifugal compressor," Proc. IMechE Part A: J. Power and Energy, vol. 225, pp. 665-673, 2011.

[48] T. Valco, H. Ashley, J. Green, D. Findley, T. Price, J. Fannin and R. Isom, "The cost of ginning - 2010 survey results," in Beltwide Cotton Conferences, Cordova, Tenn., 2012.

[49] P. Funk and R. G. Hardin IV, "Changes in cotton gin energy consumption apportioned by ten functions," J. Cotton Sci., vol. 17, pp. 174-183, 2013.

Figure Captions

Figure 1. Section showing proposed rectangular and radial evases on cyclones.

Figure 2. Three types of cyclone gas outlet (exhaust, or vortex finder) terminations found in industry: a) direct; b) with rain hat; and c) with spin cap.

Figure 3. Schematic of test apparatus showing supply duct with flow straightening, temperature sensing, velocity, and pressure measurement sections, and a 30.5 cm cyclone with dust bin and radial evase.

Figure 4. Rectangular evases 1 (a), 2 (b) and 3 (c), with inlet areas equal to either the cyclone inlet (a) or the cyclone outlet (b and c). The top line (photos) shows the evases with their dedicated spin caps.

Figure 5. Radial evase shown in section. Exit separation heights, h4, of 1.11, 1.27, 1.67, 1.91, 2.22, 2.54, and 2.86 cm were tested. The throat opening, h3, was 0.27 cm greater than h4, the exit separation height.

Figure 6. Cyclone A pressure drop v. cyclone inlet velocity for each rectangular evase. Poly indicates the second order polynomial curve with zero intercept that was fit to the data for cyclone A and each evase.

Figure 7. Cyclone B pressure drop v. cyclone inlet velocity for each rectangular evase. Poly indicates the second order polynomial curve with zero intercept that was fit to the data for cyclone B and each evase.

Figure 8. Cyclone A pressure drop v. inlet velocity for the radial evase at various openings. Poly indicates the

second order polynomial curve with zero intercept that was fit to the data for cyclone A and each radial evase exit separation height.

Figure 9. Cyclone B pressure drop v. inlet velocity for the radial evase at various openings. Poly indicates the

second order polynomial curve with zero intercept that was fit to the data for cyclone B and each radial evase exit separation height.

Table Captions

Table 1. Cyclone pressure drop over a range of inlet velocities for each cyclone and each evase.

Table 2. Example return on investment calculation assuming 1.52m diameter 1D3D cyclone, straight-blade fan

with 55% efficiency, large high efficiency electric motor with 95% efficiency, total cost for electrical energy $0.11 kWh-1, and radial evase cost of $350 (calculation does not include installation cost).

Rectangular évasé

Radial évasé

Spin Cap

Figure 1

Figure 3

K 7.6 4"--21.3

(Not to Scale)

116 cm2 183 cm2

183 cm2

Figure 4

N- K4 = 15.24 (Exit)

7.5° |

^ R3 = 9.17 —4 . h4

(Throat) + 5.0° +

R2 = 7.62 —

(Inlet)

Figure 5

_ 1200

a 1000

400 200 0

Rectangular Evasés - Cyclone A

♦ Cyclone A ■ Evasé 1 Evasé 2 Evasé 3

— Poly. (Cyclone A)

— Poly. (Evasé 1) - Poly. (Evasé 2)

Poly. (Evasé 3)

In let Velocity (m s-1)

Figure 6

_ 1200 CL

a 1000

Rectangular Evasés - Cyclone B

♦ Cyclone B ■ Evasé 1 Evasé 2 Evasé 3 -Poly. (Cyclone B) - Poly. (Evasé 1 ) Poly. (Evasé 2) Poly. (Evasé 3)

In let Velocity (m s-1)

Figure 7

_ 1200 CL

a 1000

Radial Evasé - Cyclone A

■ 1.11 A 1.27 1.67 x 1.91

• 2.22 + 2.54

2.86 -Poly. -Poly. Poly. - -Poly. Poly. Poly. Poly. Poly.

(1.11)

(1.27)

(1.67)

(1.91)

(2.22)

(2.54)

(2.86)

I n let Velocity (m s-1)

Figure 8

1600 1400

_ 1200

a 1000

■ 1.11

A 1.27

X 1.67

X 1.91

• 2.22

- 2.86

---- •Poly. (B)

Poly. (1.11)

Poly. (1.27)

Poly. (1.67)

--- -Poly. (1.91)

Poly. (2.54)

Poly. (2.86)

Radial Evasé - Cyclone B

In let Velocity (m s-1)

Figure 9

Cyclone Inlet Velocity (m s-) 12.26 14.26 16.26 18.26 20.26

Cyclone A alone 442 618 822 1055 1317

A With Spin Cap 1 469 653 866 1110 1383

A With Evase 1 439 610 809 1036 1290

A With Spin Cap 2 438 612 814 1045 1304

A With Evase 2 421 587 781 1003 *1252

A With Spin Cap 3 475 660 876 1122 1398

A With Evase 3 446 620 822 1053 1312

Radial Evase with Separation Height of:

1.11 cm 421 581 768 981 1220

1.67 cm 391 545 724 928 1158

2.22 cm 406 567 753 966 1205

Cyclone B alone 435 608 811 1042 1302

B With Spin Cap 1 480 668 887 1137 1417

B With Evase 1 447 623 828 1062 1325

B With Spin Cap 2 457 639 851 1093 1366

B With Evase 2 443 620 827 1063 1328

B With Spin Cap 3 458 638 847 1086 1355

B With Evase 3 447 625 834 1071 1339

Radial Evase with Separation Height of:

1.11 cm 430 597 791 1013 1262

1.91 cm 400 557 740 949 1183

2.22 cm 407 569 758 975 1218

Cyclone A B

Reduction in Pressure Drop (Pa) 97 70

1.52m Cyclone Airflow (m3 s"1) 4.72 4.72

Air Power Reduced (W) 460 333

Centrifugal Fan Efficiency 0.55 0.55

Motor Efficiency 0.95 0.95

Electrical Power Reduced (kW) 0.880 0.636

Electricity Costs ($ kWh-1) 0.11 0.11

Evase Cost ($)z 350 350

Simple Payback (h) 3,614 5,000

ROI(y) operating 24 h d-1 * 75 d 2.0 2.8

installation cost not included.

-.1200 fiai 000 o

« 600 v

¿ 400

11.9% Pressure Reduction with Radial Evasé

y = 3.6149x2 - 8.2454X R2 = 0.9972

♦ Cyclone Only

• Radial Evase

y = 3.1521X2 - 6.7144X R2= 0.9986

10 15 20

In let Velocity (m s-1)

Graphical abstract

Highlights

• This investigation is the first to test cyclone performance with evases

• Evases may reduce cyclone pressure drop without affecting collection efficiency

• Investment in an evase can be recovered through reduced operating energy costs

• Potential energy savings for the US cotton ginning could industry exceed $1.5 million