Improvement Soot Particles Separation Equipments for Rubber Smoking Chamber

Soot produced from rubber tree wood combustion during the process of drying natural rubber sheets has been shown to produce negative health effects for workers, as well as poorer quality rubber sheets. This study examined ways to improve soot particle separation equipment for rubber smoking chambers. An impaction wall was able to catch all soot particles ranging in size from 3.3-4.7 microns and a modified electrostatic precipitator, installed outside of the rubber smoking chamber in order to decrease danger to workers and with the highest available voltage supply of 12 kV, captured soot particles 0.43-3.3 microns in size. This modified precipitator operated longer than the usual time 120 hours or more before requiring cleaning; equivalent to smoking about 12 lots of rubber sheets or operating for 3 months during high rubber production season. The total efficiency of soot particle separation through the use of the modified collection equipment was more than 50% for all tests. In addition, the color index of rubber sheets used in the testing improved from 6.0 to 12.0. Coloration, which is a measure of quality, was better than sheets from smoking chamber lacking soot particle separation equipment.


INTRODUCTION
Soot particles are produced from incomplete combustion of carbonaceous materials (Ndiema et al., 1998), including wood combustion for drying agriculture products, such as natural rubber (Hevea brasiliensis) sheets in Thailand.Soot emitted during the rubber drying process produces environmental and human health problems, such as pulmonary diseases and cardiovascular illness.Thailand has the largest natural rubber (NR) production and export in the world.The total amount of natural rubber production in 2006 was about 3.2 million metric tons (Thailand Rubber Research Institute, 2007).About 90% of the NR produced was exported.Generally, most natural rubber products are used in vehicle tire production; hence, the direction of NR production follows consumption in the world market.The amount of NR exported from Thailand in the last three years has increased about 0.2 million tons per year; however, the increasing quantity is insufficient for the growth of the tire industry.In the coming years, it is expected that NR production will double.To meet these expectations, while also limiting atmospheric and human health impacts, improvement of NR production for 7 million Thai farmers (particularly in southern Thailand) must be carefully planned and implemented.
Natural rubber in Thailand is produced in four types; ribbed smoked sheet (RSS), block rubber, rubber concentrated latex, and miscellaneous other forms.The proportion of each type is 35%, 35%, 28% and 2%, respectively.Production of each type depends on the world's NR consumption.In the world's rubber market, the United States, Japan, China and India are the largest consumers, while Japan and China (Tekasakul et al., 2008) are Thailand's largest exporting markets.At present, the worldwide consumption of rubber-concentrated latex is nearly constant.In contrast, both RSS and block rubber consumption is increasing due to increasing demands of the rubber tire industry.
In RSS production, Thailand can produce from July to March, while December to February is the high season.In the rubber smoking process, sheets of rubber are washed in a pool and hung on bamboo bars for drying before transferring to smoking chambers where rubber-tree wood is burned as a heat supply to reduce the moisture content in the sheets.In this process, wood burning produces wastes, such as carbon, water vapor, ash residue, and soot particles.Soot particles in the rubber smoking chamber produced from wood combustion have negative effects for the workplace environment, on worker health and also in quality of the rubber sheets themselves.Excessive deposition of soot particles on the sheet surface results in a darker color, which is considered below market standards and thus lower prices (Kalasee, 2005).In order to reduce these problems, a method for soot particle reduction is needed.However, the gas does not require a complete cleanup before entering the rubber smoking chamber, since soot particles contain some phenolic compounds (Simonelt et al., 1993) which hinder the formation of molds and bacteria that reduce the shelf life of rubber sheets.Kalasee et al. (2003) found that the size distribution of soot particles from rubber wood combustion in the smoking room has an average mass median aerodynamic diameter (MMAD) of 0.95 microns.The average value of geometric mean standard deviation (GSD) is 2.51.The mass concentration of the smoke particles depends on the moisture content of the fuel wood.Kalasee et al. (2003) also found that particle concentrations range from 47-1,358 mg/m 3 for 34.5-107.5% moisture content in fuel wood.
Several methods and devices were used for particle collection to reduce or prevent emission of particles escaping into the atmosphere; e.g., filtration, centrifuged cyclones, scrubbers, gravitational settling, etc.However, these methods need large spaces for installation.Filtration induces pressure drop when filters are contaminated with particulates during long operation times.Cyclones, scrubbers and gravitational settling can capture large particles greater than 10 microns; however, the collection efficiency is lower for smaller particles.In this work, a corona discharge device, or an electrostatic precipitator (ESP), was used to collect soot particles with variable concentration in the submicron range.This device has the advantage of high collection efficiency and less pressure drop.
In a previous work, Kalasee (2005) and Tekasakul et al. (2005) found that ESP utilizing a corona discharge was suitable for catching soot particles in wood combustion systems and improving the color of drying rubber sheets in a smoking chamber.However, ESP had some negatives for tar and volatility which decreased the life of the sheets.Installing an ESP in the smoking chamber would be dangerous for the electrical system because of the high amount of waste water produced in the drying process.During the high season it is important to change and clean electrodes before operating the system; therefore, the collection equipment was newly designed for use outside of the smoking chamber.Previously, ESPs were located inside chambers.

Rubber Smoking Chamber
Most of the RSS manufacturers in Thailand are located in the south, where production is shifting from large-scale industries to community-level rubber cooperatives.About 700 cooperatives are currently operating throughout the country and each could produce about 500-1,000 metric tons per year of ribbed smoked sheets (RSS).Fig. 1 shows the rubber smoking chambers are constructed of brick and mortar, and the average chamber is 2.5 × 6 × 4 m 3 in size.The front wall is typically a steel gate for loading and unloading rubber.Ventilating windows are on the rear wall and the ceiling.The rooms are generally equipped with sensors for controlling the temperature.Heat and smoke distribution tunnels are on the floor.The rubber sheets are hung on bamboo poles placed on 2 × 2 × 2 m 3 steel crates.A forklift is used to maneuver the crates into and out of the chambers.Furnaces are located at the rear of the chambers and about 0.5 meters below the floor.Hot gas and smoke are conveyed through the distribution duct and grid into the chamber.The furnace normally has dimensions of 0.6 × 0.6 × 1.5 m 3 and is constructed from brick.Combustion rate is controlled by an adjustable steel gate in the front.

Particle Sampling Method
Soot particles are defined as all particles emitted from the combustion.Sampling was performed between the impaction wall and ESP in the tunnel which carries the combustion gas into the smoking chamber.An eight-stage Andersen Impactor (Dylec, Model   with cut sizes of 0.43, 0.65, 1.1, 2.1, 3.3, 4.7, 7.0 and 11.0 microns was used for the experiment.A backup filter was used to collect particles smaller than 0.43 microns (Tekasakul et al., 2008).Prior to sampling, the sampling plates were cleaned in an ultrasonic bath, dried overnight in a desiccator, and weighted in an analytical balance (Mettler, AB 204-S).After sampling, particles were placed in the desiccator prior to weighting in the same analytical balance in order to determine the mass of collected particles.

Particle Collection Method
Although, many methods (i.e., filtration, centrifuged cyclones, scrubbers, gravitational settling, etc.) could separate out particles, they are not suitable for use in the rubber smoking chambers.For this study an electrostatic precipitator (ESP) and an impaction wall were found to be more appropriate.

Electrostatic Precipitator
An ESP is a particulate collection device which uses electrostatic force to remove or separate the particles from a gas stream.The principle is to give an electrostatic charge to particles in a gas stream and then pass the particles through an electric field that drives them to a collection electrode.ESP requires maintenance of a high-potential difference between two electrodes, one a collecting electrode, which has a larger radius of curvature, e.g., a cylinder or a flat plate, and is electrically grounded.The discharging electrode, which has a smaller radius of curvature, e.g., thin wire or a sharp point, is held at a high voltage (Chang, 1998;Chen, 2002).Because of the high potential difference between the two electrodes, a powerful ionizing field is formed.
Every particle either has, or can be given, a positive or negative charge.If all particles in the gas stream have a negative charge and the collection plate is positively charged, the negatively charged particles will migrate to the grounded collection plate and be captured.The particles will quickly deposit on the plate, and accumulate a layer of dust that is possible to remove, either by rapping or sweeping the plate, or spraying it with a liquid.
Many researchers such as: Kim and Lee (1999) Jedrusik et al. (2001); Laskin and Cowin (2002); Elayyan et al. (2002); Kulkarni et al. (2002); Xiangrong et al. (2002); Jedrusik et al. (2003); Fujishima et al. (2004); Xiangrong et al. (2005) and Clack (2006) considered ESP for the removal of small particles (e.g., dust, fumes, smoke, diesel exhaust and soot particles); but the expensive design and voltages above 25 kV for use in RSS manufacturing are still unsuitable since it would be too complicated for workers to operate.Kalasee (2005) found that removing soot particles from the collection plate by spraying caused moisture content to increase, resulting in molding.This study modified an ESP by installing it between the furnace and the rubber smoking chamber to collect the soot particles before entering the rubber smoking chamber.This installation increased the life of the ESP and decreased the danger of electrocution.

Impaction Wall
The impaction wall (similar to a flat plate) is a simple collection method which uses mechanical force to remove particles from the gas stream.In principle, soot particles colliding with large inert object are able to pass from the streamline through the nozzle in front of impaction wall as shown in Fig. 2.

EXPERIMENTAL
The experiment was divided into two parts.First, a burner (Fig. 3) was built for field-testing and evaluating soot particle separation by impaction wall and a modified ESP.Then the equipment was designed for soot particle collection (Fig. 3) in an actual rubber smoking chamber.

Field Testing
In this experiment, collection efficiency and soot particle size range distribution were measured.A burner was built for this test.

Collection Efficiency of a Simple Designed Electrostatic Precipitator
For field-testing and use with a rubber smoking chamber, a low-cost and simplified ESP was designed to use for collecting soot particles.The analytical expression for the efficiency of the ESP is given by the Deutsch-Anderson equation (Hinds, 1999): Where A is the collector surface area, V e the particle migration velocity, Q the volume flow rate of gas, and η is the ESP fractional collection efficiency.
The collection efficiency of ESP increases with the migration velocity.The relationship for migration velocity is given by: Where q is the electrical charge on particles, E the electric field strength, C the slip correction factor, μ the air dynamic viscosity, and D p is the particle diameter.
Fig. 4 shows the designed ESP with a single copper wire electrode (0.5-mm-diameter, length 150 cm) placed in the middle of the ESP between two grounded, stainless steel plate electrodes.The plate electrode lengths were 150 cm each; while the plate-toplate electrode spacing was 10 cm (total spacing in the tunnel remained 30 cm width, 150 cm length, and 60 cm high).The ESP collects soot particles before the hot gas enters the smoking chamber through 12 8-inch-diameter tubes at floor level.The collecting device is connected to a simple electrical circuit that supplies a high potential difference between the electrodes.The inlet voltage of 220 VAC is adjustable with a slide regulator between 0-220 V (Chuan Hsin, SRV-10).This voltage is then transformed into high voltage (0-15 kV) by a neon transformer (LECIP, EX230A15N).A Wheatstone bridge circuit consisting of four high-voltage diodes (16 kV) is used to transform the highvoltage AC current to a rippled DC current.In this work, the input operating voltage of the ESP was kept constant at 220 V which corresponded to 12 kV (Kalasee et al., 2006;Srisang et al., 2006) at peak output.The corona discharge wire was located at the negative electrode and the collecting electrode carried a positive charge.This simple circuit was used because it is easy to construct and affordable for the manufacturers.

Collection Efficiency of a Simply Designed Impaction Wall
For this work, the impaction system was also designed to be simple.It was installed across the flow of hot gas and soot particles between the furnace and modified ESP (primary collection).It was made from brick, concrete and mortar.
According to conventional impaction theory (for flat, rigid substrates), the dimensionless stokes number (Stk) is the principle factor for impaction and is given by: Where μ is the air dynamic viscosity; W is throat width; ρ P is particle density; D p is particle diameter; U is air velocity in the throat; and C C is Cunningham slip correction factor.The slip correction factor is specified by (Hinds, 1999): Where λ is the mean air molecule free path.Fig. 5 shows the impaction system which consisted of a hot gas flow in the throat tunnel (stopping distance) as a rectangular acceleration nozzle and an impaction wall.In this design, the dimensions of the impaction wall were 150 cm (length) and 60 cm (high).The throat width (W) was 60 cm and its length (L) was 30 cm.The stopping distance (S) was 30 cm.The ratio of S/W was 0.5.Hence, the value of Stokes number in theory for the stopping distance was 0.5 (Hinds, 1999).The velocity range of hot gas flow in the throat tunnel was 200-500 cm/s.Its value was estimated and considered from simulation data with an air velocity inlet into the furnace by computational fluid dynamics (CFD) modified from Kalasee (2005), Promtong (2006) and Promtong and Tekasakul (2007).The collection efficiency of the impaction-wall separation of soot particles was determined by the before and after weight of the zinc plate in each test (for this design, the zinc plate is built on the impaction wall due to dust load) and this equation: Where C exit and C inlet are the particle concentration at the exit and inlet of the collecting device, respectively.

Soot Particle Distribution Range
In this experiment, an 8-stage Andersen impactor (Dylec, Model AN-200) replaced the filter holder as shown in Fig. 3.And the aerosol was pulled through the flue by a vacuum pump used in the preceding section.The flow rate was 28.3 L/min.These vacuum pumps were then operated only within 30 minutes for all tests.The mass of the plates was measured by a 4-digit analytical balance (Mettler, AB204-S) before and after the experiment.

Rubber Smoking Test
An impaction wall and a single modified ESP (a wire-plate type) were built for soot particle collection and replaced the earlier versions of ESP used for the rubber smoking process, which utilized 12 identical wire-cylinders.The new system was installed between the furnace and rubber smoking chamber.The impaction wall also increased the life of the modified ESP.Kalasee, (2005) and Tekasakul et al. (2006) operated an ESP only for 30 minutes after the introduction of fresh wood to the burner, since the smoke is at its greatest density during this period.In this study, the total operating time for testing the drying of one lot of rubber sheets using the ESP was 10 hours.During four days of processing, fresh wood combustion took place 20 times.A total of three months, or 12 lots of the rubber sheets of drying tests were performed and the modified ESP was in operation 120 hours.

RESULTS AND DISCUSSION
The results for this research are divided into four parts.In the first part, dust load and size particle collection by an impaction wall and ESP are evaluated.The second part, the life of a modified ESP is studied.The third part presents a look at soot particle distribution for the equipment.And finally, the color of the dried rubber sheets for all experiments is observed.

Dust Load and Size Particle Distribution
Table 1 shows the total particle mass concentration and the efficiency of the impaction wall.Particles from upstream and downstream of the ESP were sampled and measured.This study assumed that particles downstream of an impaction wall equalized the particles upstream of the ESP.In Table 1, runs #3, #5, and #8, as well as the total particle concentration are very high, due to the increased moisture content of the rubber tree wood, which was stored outdoors and absorbed rain water.The size-fractionated collection efficiency of soot particles for the impaction wall and ESP are shown in Figs. 6 and 7, respectively.The results show that the impaction wall has significant effect for all size of soot particles range.The quantity of large particles range from 3.3 to 4.7 microns which collected by the impaction system was higher than small particles.However, Fig. 6 shows that for a part of another curve, the small particle collection was near the large particles.It indicates that the hot gas flow in the tunnel was turbulent.So, an impaction wall also has significant influence on collecting small particles according to the Brownian motion and diffusion method.Fig. 7 shows the influence of size particles on the collection efficiency of the designed ESP during on experiments.The small particle amount was higher than large particles due to the influence of the Brownian motion and diffusion charging mechanism of small particles.
The collection efficiency of soot particles under dust loading of an impaction wall and ESP are shown in Figs. 8 and 9.The collection efficiency is plotted against the dust-loading parameter (c v t); where c is the particle mass concentration, v is the aerosol velocity, and t is the collection time.In Fig. 8, the collection efficiency of the impaction wall is rather constant as the particle concentration and operation time increase.This result indicates that the quantity of particles and time operation did not have significant influence.But the results for the ESP found that as particle concentration and operation time increased, the collection efficiency of the ESP decreased.This is in agreement with Chang's experiments (1998).The loading of dust on the electrode reduced the electrical interactions between the collecting surface and charged particles.
The main problems of an impaction wall are the pressure drop, and the velocity and temperature variation in the rubber smoking chamber.So, in this experiment it was necessary to control the combustion rate by using an adjustable steel gate on the front of furnace to adjust the temperature in the smoking chamber to between 50-65°C.For the rubber smoking process, the temperature was set about 49-52°C on the first day, about 52-57°C in second day, about 57-60°C the third day, about 60-63°C the fourth.Over a period of days, the temperature was constant at about 50°C, and the hot gas flow velocity constant at about 20 L/m (modified from Prasertsan, 1993;Kalasee, 2005;Tekasakul et al., 2006).

Life Span of a Modified ESP
Table 1 shows the collection efficiency of the modified ESP (wire-plate) for the clean ESP electrode is decreased from 78% to 40% after operating at 7200 minutes (120 hours) or the rubber sheets 12 lots.This result is better than compared with the results from Kalasee (2005) and Tekasakul et al. (2006).After 120 hours of operation, the collection efficiency of this ESP design decreases and the ESP electrodes must be changed or cleaned.

Escaping Soot Particle Distribution in the Rubber Smoking Chamber
In this study, gas is not required to complete the cleanup system before entering the rubber smoking chamber, because soot particle distribution decreases about 50% of total particle production from the rubber wood combustion.Soot particles contain phenolic compounds (Simonelt et al., 1993) which hinder the formation of molds and bacteria, which affect rubber sheet

Cut size diameter [micron] Size fractionated collection efficiency [-]
Run  shelf life.Small particles were distributed in the rubber smoking chamber and perched on the rubber sheet surface, resulting in a color a little darker than normal.Parts of small particles were attached to the rubber-sheet surface, as well as the walls of the smoking chamber.And finally, the escaping soot particles will probably flow and spread to the environment.

Fig. 4 .
Fig. 4. Diagram of the corona discharge device and electric circuit.

PhysicalFig. 5
Fig. 5 Schematic diagram and the physical characteristics of impactor.

Table 1 .
Upstream, downstream and total collection efficiency of soot particle distribution using an impaction plate and ESP.