Particle Size Distribution and Cyclone Efficiency Distribution Essay

In this experiment collection efficiency of a cyclone has been determined for two types of particles – Fly Ash and MgO in an air stream. Such a dust is commonly found in industries using coal and refractories. Particle size distribution of ambient air as well as cyclone exhaust has also been measured using Electrical Low Pressure Impactor (ELPI). The collection efficiency of the cyclones was in 85 -95% range for Fly Ash dust and approximately 96% for MgO dust under experimental conditions. The particle size distribution in the ambient air as well as in the cyclone exhaust was showed a log normal distribution and each of these distributions was composed of more than one size distributions. A. Introduction: In modern industrial era we have to live with dust and powders on continuous basis. There are many industrial processes that use raw materials in powder form like powder metallurgy, sintering plants in integrated steel plants, cement industry, polymer engineering etc. to name a few. It is not unreasonable to expect that the industries that use powders as raw material, throws lot of particulate matter into the atmosphere around it. Even in the cases, when the raw material in not a powder, the emissions contains lot of particulate matter. Some examples are emissions from blast furnaces, coal fired plants etc. to name a few. Vehicular pollution is one of the major sources of suspended particulate matter (spm) in the atmosphere in the urban areas. Therefore, the knowledge of particle size distribution is required and very useful in many cases. Some examples are the following: †¢ Estimation of dust hazard to the personal handling powders in industries †¢ Designing an equipment for removing dust from a gas stream like exhaust gas of blast furnace and other furnaces †¢ Selection of a suitable dust cleaning system for a given environment †¢ Estimation of the efficiency of filters and other dust collection systems †¢ Identifying the source of the dust particles †¢ Estimation of properties of an aerosol etc. Therefore, it is useful to understand the method and practices of measuring and describing particle size distribution and also the different methods and instruments that are used to clean dust from a gas stream. This experiment is concerned with sizing distribution of atmospheric dust and the efficiency of a dust collection system will be determined. The following section described different distributions of particle size in a dust sample. A. 1 Description of particle size distributions Dust particle or airborne particles are not of a given size rather there are particles of different sizes in a dust sample. This size range can be very large in the range of tens of nanometers to hundreds of micron. The exact size distribution depends upon the source of the particulate matter. For example size distribution in a blast furnace exhaust will be different from that in a motorcycle emission and so on. A particle size distribution can be described by the following mathematical expression: Here, ‘d’ is the diameter of the dust particle and dN is number of the dust particle in the diameter range ‘d’ to ‘d+dd’. ‘a’, ‘b’, ‘? ’ and ‘? ’ are the constants. Depending on the value of these constants there are two kinds of particle size distributions. One is â€Å"Nukiyama – Tanasawa† distribution and the other is â€Å"Rosin-Rammler† distribution. For â€Å"Nukiyam – Tanasawa† distribution, ? = 2 and ? = 1 and the expression is â€Å"Rosin – Rammler† distribution is described by the following expression: The â€Å"Rosin – Rammler†distribution was developed to represent size distribution of coal particles, that was received by sieving of coal particles. Here di is a particular sieve size or the minimum size of a particle retained by that sieve; R is the weight of the coal particles retained by all the sieves with size d > di and was expressed as percentage of the total coal weight; ‘b’ and ‘n’ are constants. To evaluate these constants in these size distributions, one needs to do curve fitting. Therefore, it is easier to go for simpler statistical distributions based on ‘normal’ distribution. For any distribution, there is a mean and a standard deviation. For a sample these can be calculated by using the following formulae: Sample Mean Sample Standard Deviation From these sample statistics one can calculate population parameters like true mean or population mean, ? and standard distribution ? with certain degree of accuracy. However, if the sample can be considered to be true representative of the true population then one can take sample statistics (mean and standard deviation) as population parameters ? and ?. While the mean is a measure of the central tendency, standard deviation gives distribution of particle size around the mean. If standard deviation is large then the distribution is wide and vice versa. If two more parameters ‘skew’ and ‘kurtosis’ that measure symmetry and peakedness respectively are also used in conjugation with mean and standard deviation, then can completely describe a size distribution. For a normal distribution, ‘skew’ and ‘kurtosis’ are zero and the distribution is mono-modal with peak at the mean and is symmetric about the mean. Such a distribution is applicable for simple distributions with m/s > 2. 5. If this ratio is smaller then the distribution, generally shows large positive skew. To tackle such a problem one goes for log-normal distribution, which is a normal distribution of the logarithm of the particle size. Most of the natural size distributions are best described by log-normal distribution. As mass distribution is more appropriate and used frequently, therefore, one can deduce mass distribution from size distribution. To do this one needs to calculate mean and standard distribution of the mass of the particles and this is done by dividing individual size measurements by while calculating the mean and the standard deviation. For most of the pollution control applications log-normal distribution is used. Probability distribution function (pdf) for such a distribution with mean ? and standard deviation ? is given by the following expression Such a distribution is shown in figure 1, below. In this case, the peak shifts in left direction with increasing standard deviation, ?. Fig. 1: Shift of the probability distribution peak in left direction with increasing standard deviation of the sample There is another very important aspect of particle size distribution. Generally a dust sample collected from certain location does not consist of a single distribution; instead it consists of many size distributions. It is easier to identify and separate these distributions when the peaks are well separated. However, many times the peaks are so close that these distributions mingle up as a single composite size distribution and one needs to extract individual distributions out of this composite distribution by carefully deconvoluting the composite size distribution. Before describing a size distribution, one needs to first measure the size distribution. There are many instruments that help in measurement of particle size and the size distribution. Some important techniques are described in the following section. A. 2 Measurement of Particle Size and Size-Distribution Sieving: This is the most conventional and easiest method for particle size-distribution measurement. In this case a representative sample of the particulate matter is taken by suitable sampling method like divided cone method. In divided cone method, the particulate matter is made as a cone and one quarter of the cone is taken. This process is repeated hill the final sample size is taken. This sample is then sieved by using a series of sieves of different sizes in a consecutive order. The particle that remains above the sieve of a particle number (size) is given that size and in this manner the size-distribution of the entire sample is measured. Sieve size is given by a number. That number represents number of aperture in a linear inch. Thus a sieve of size 75 means, there are 75 apertures in one inch of that sieve and so on. Therefore, a larger sieve number corresponds to a smaller sieve size. This method is very easy and suitable for coarse particles of size greater than 50 ? m. However, for finer sized particles, this technique becomes very unreliable. Optical Microscopy Optical microscope is also very useful method to measure size distribution of particulate matter. This method can be used to measure size distribution of particulate matter from any source. Different sampling methods can be used to collect the sample for size distribution measurement. Some of these are: (i) Filtration: Membrane filters are generally used to collect samples that have different color or refractive index than the filter. (ii) Sedimentation on a glass slide is another useful technique, especially for large particles. However, for getting a representative sample one should be careful. (iii) Thermal precipitation is another useful technique; however, care should be taken to avoid segregation of sizes. (iv) Electrostatic precipitation on a glass slide or electron microscope grid is another technique that is commonly employed. However, optical microscopy is limited to ~ 1 ? m sized particles as maximum magnification is 1000 only. This is because, light is the probing signal and its wavelength is of that order. For measuring the size distribution if still finer particles one needs to use scanning electron microscope (SEM) and transmission electron microscope (TEM). In case of SEM a focused electron beam is scanned in a raster and the image is formed by collecting the different type of electron signals like secondary electron or back scattered electron. In case of TEM, the focused electron beam is transmitted through the particle and an image of the particle is formed at higher magnification ~ 100,000. Cyclones: Cyclones are used mainly as dust removing system; however, these are also used for separating particles into different size groups. One example is use of cyclone in personal dust sampler to separate the powder into two fractions – one which is respirable and another which is not. Cascade impactors: In this system, particles are collected in different size groups in different stages according to the aerodynamic impaction onto a substrate. Each stage can be analyzed chemically, measured using a microscope or can be even weighed electronically. These are used for sampling of particulate matter in atmosphere or in chimney or furnace exhaust gas. These are capable of sizing the particulate matter in 0. 05 to 10 mm range. These systems have evolved considerably and modern systems are equipped with quartz crystal microbalance for detecting mass number of the dust particle and even electrical detection of the particles, which has been charged before classifying into different sizes. One such system is Electrical Low Pressure Impactor (ELPI). This equipment will be described in somewhat detail in the next subsection. Diffusion Battery: In this system particles are classified based on their ability to diffuse through a series of mesh screens. Smaller particles diffuse faster and vice versa. Finer particles are thus collected easily than the coarser particles. This system can be used for particles smaller than 1 ? m. Electrical Mobility Analyzers: In this system, the particles are charged prior to separation. The charged particles are separated by applying electric field. Smaller particles have higher mobility due to smaller mass than the larger particles. Therefore, electric field removes the particles in selective manner at different stages according to their sizes. This system can do sizing of particles in the size range 1 ? m to 1 mm. Light Scattering Devices: These are based on scattering of light by the dust particle in the suspension. It can use dry as well as wet suspension. The angle of scattering of light is related to its size. Normally a highly collimated laser light is used. Though the minimum size is limited by the wavelength of the light and is ~ 0. 3 mm; newer systems have been designed that can measure particles in nm size range as well. However, these are costly equipments.