Development of Correlation Between GCV and Proximate Analysis of Indigenous Coals

type

below bituminous coal. The heating value of sub location of the coal mines. The heating value is bituminous coal is between 8,300 and 13,000 Btu's expressed in two different ways on account of per pound. Lignite contains 25 to 35 percent carbon moisture present in the coal. Heating value usually content. Lignite is also used to generate electricity. expressed as higher heating value (HHV) or gross Sometimes lignite is known as brown coal because calorific value (GCV) and lower heating value of its brown color. The heating value of lignite (LHV) or net calorific value (NCV). Coal contains ranges between 4,000 and 8,300 Btu's per pound moisture as an essential component so; difference between both these heating values is the latent heat Pakistan is moving towards the large scale use of of condensation of water vapors produced during coal today because huge coal deposits have been combustion process. When coal burns the moisture found in Pakistan. The Thar coal deposit in Sindh is in coal evaporates taking away some heat of one of the largest in the world. It contains of 175 combustion which is not available for our use. The billion tons of coal in four sections. Punjab has about higher calorific value presumes that all the vapors 600 million tons of coal deposits. The Salt Range produced during combustion process are fully alone has about 500 million tons of coal that can be condensed and the lower heating value presumes exploited and used as an energy resource that the water is removed with the combustion economically. Beside Thar, there are seven other products without being fully condensed. When we coal fields in Sindh; two of them are developed while say Higher Heating Value or Gross Calorific Value the others are un-developed including Thar. KPK it is the total heat released when burning the coal. (Hangu and Cherat) and Azad Kashmir (Kotli) are When we say Lower Heating Value or Net Calorific also among the developed coal fields. Balochistan Value it is the heat energy available after reducing has more than ten developed coal fields the loss due to moisture. Coal with greater contributing a major part of coal production in percentage of volatile matter and fix carbon Pakistan. Germany has been developed a process produces more heating value on combustion as they for the up gradation of the Kalabagh iron ore, using are the combustible constituents of coal and greater indigenous coal of Makarwal. Coal is the best future percentage of non-combustibles (moisture and energy resource for Pakistan. The most important mineral matter) contents lowers the heating value. uses of coal are in steel production, cement In bomb calorimeter, the heating value of coal is manufacturing, electricity generation and either determined by an adiabatic process or by production of different chemicals and manufacture static method (isothermal) with the correction made of gaseous and liquid fuels. Coking coal also known if net heating value is of interest for analysis of coal. as metallurgical coal is mainly used in steel The unit is calories per gram, which may be

Production:
converted to the alternate units. Heating value is 2.1 Heating Value the direct indication of heat content (energy value) of coal. The heating value represents the combined Heating value of coal is the heat produced by heats of combustion of carbon, hydrogen, nitrogen combustion of a unit quantity of coal in a bomb and Sulphur in organic matter and of Sulphur in calorimeter with oxygen under a specified set of pyrite and the higher heating value with correction conditions prescribed by standard test like (ASTM applied if the lower heating value is of interest The D-121; ASTM D-2015; ASTM D-3286; ISO 1928) [1-significance of the correlation of heating value with 7]. The heating value of coal is neither the part of composition in ordinary fuel usage is shown by the proximate analysis nor part of ultimate analysis it development, as early in 1940's 9 different formulas is one of many physical properties of coal. It is often for calculating heating value of coal from the found in the various sections that deal with the ultimate analysis and 11 formulas for calculating it physical properties. The heating Value varies on the from the proximate analysis. Formulas have been coalification, geographical age, ranking and prop osed with in the last thre e year s. The heating value of coal. The fixed carbon of different correlation is perhaps of even greater importance coals is assumed of a fixed composition and hence of for the rationalization and modeling of conversion fixed heating value. The composition and heating processes now being developed much work has been value of the volatile matter differ from coal to coal done on measur ements of heatin g value of and are assumed to depend upon the nature of coal indigenous coal samples, where the calorific value as indicated by the volatile matter on dry as free was found to vary with percentages of fixed carbon, basis. These assumptions limit the utility of the volatile matter, moisture and ash contents. These Goutal formula. parameters can be used to estimate the calorific The following model have been developed by value coal. Some of the models proposed originally Central Fuel Research Institute, Dhanbad (CFRI), for correlation of heating value of coal with its for the calculation of heating value of Indian coal proximate analysis.
form their proximate analysis. unaltered minerals, oxides and sulfites. Chemical [12] changes during the "ashing process" that occurs in Empirical formulae are also available in the the mineral matter produces ash and Fixed Carbon literature for the calculation of the heating value of constituents in coal that left behind after the loss of coal based on ultimate and proximate analyses. ash, volatile matter and moisture, is referred to as HHV=82F+a.V [13] fixed carbon content. The fixed carbon value is F= percentage fixed carbon V= percentage volatile matter basically the value that is used for measuring efficiency of coal on burning [15][16][17][18]. a= a constant depending upon the value of volatile matter expressed as dry ash free basis

Regression Analysis:
This model assumes the coal consisting of volatile It is a which is a multivariate matter and fixed carbon, each contributing to function for examining the statistical technique linear correlations

Journal of the Pakistan Institute of Chemical Engineers
between a single dependent variable(DV) and two determination, the adiabatic bomb calorimeter or more independent (IV). This type of analysis is method was used in which a weighed sample is used for forecasting and prediction, and also used to burnt completely in oxygen under controlled determine the relationships between the dependent conditions The calorific value is computed from var iab le and ind epe nde nt var iab les . Man y temperature observations made before, during and techniques have been developed in Regression after combustions by Heating value= m.CpDT analysis of which linear regression analysis and 5.1 Determination of Pearson correlation (r) nonlinear regression analysis are vital for the Pearson's correlation was calculated by dividing the current analysis. Multiple linear regression sum of the xy values (Óxy) (dependent variables and analysis was conducted in order to get predicted independent variables) by the square root of the gross calorific value of coal by applying function on 22 product of the sum of the x values (Óx ) and the sum combustibles (fix carbon and volatile matter) and 22 of the y values (Óy ) The resulting formula is: non-combustibles (moisture and ash contents) components of coal against calculated gross calorific r = value of coal respectively In this process dependent variable is illustrated as a function of different However, the correlation between these parameters independent variables with corresponding was determined by using the software IBM SPSS coefficients, along with the constant term. In general, any increasing function of the absolute air drying oven maintained at 40 C for one hour.
distance would serve to measure the goodness of a The air-dried samples were cooled in desiccators, predictor weighted and again placed in the air-drying oven for MSE = one hour. The experiment was repeated until the loss in weight of total samples was not more than Where, 0.1% per hour. Each sample was then thoroughly n = numbers of total experiments performed. mixed and gradually reduced in size to -60, +80 K = number of predictors used in the model. matter), M (Moisture), F.c (fixed carbon) and ash As the proximate contents of coal (moisture, ash, contents on air dried basis were used as fixed carbon and volatile matter) are directly independent variables while HHVs MJ/kg (higher related by their percentages as follows; heating values in Mega Joules per kilo gram) were Moisture% + Ash% + Volatile matter% + used to target the output dependent variable. The Fixed carbon% = 100 studies included two models; Model 1 contained all So, according to above relation for the proximate the proximate analysis components as predictors of components of coal, the Model 2 ultimately has HHV ash% as predictor [19][20][21][22][23][24]. Descriptive statistics of (Y = a+b X+b2X+bX+bX) 11 223 34 4 the data set considered in the model development are presented in Table 1. While, the predictors of model 2 included fixed carbon, moisture and volatile matter. Ash contents were excluded.

Figure 2. Effect of Moisture and Ash contents on HHV of coal
It means that it is necessary to use a linear model to the prediction of HHVs. Statistically; it was make a better estimation models.
observed that there was a strong negative correlation between fixed carbon and ash. To see (Y = a+b X+b2X+bX+bX) 11 223 34 4 correlation between fixed carbon, ash and all other Therefore, on the basis of the considered model predictors for higher heating value of coal, Pearson structures, multiple linear regression method correlation was employed. The results are based modeling was applied to estimate the higher presented in Table 2. heating values of the coals as the best fit models for

Sig. (2-tailed)
.000 Results show significant negative correlation to be strongly negative, thus showing that no more between fixed carbon and ash contents of coal. different information was obtained due to fixed Ho we ve r, th e co rr el at io n be tw ee n th es e carbon and ash for prediction of heating value of parameters, fixed carbon and ash, which appeared coal. It refers to a situation in which two or more explanatory variables in a multiple regression may change erratically in model are linearly related meaning that one can be response to small changes in the model or the data. linearly predicted from the others. So, model 2 does Numbers of solutions are present in statistics to not contain percentage value of ash contents overcome multicolinearity problem for regression actually, but indirectly it contains percentage value analysis. Two of these are proposed here as follows; of ash contents of coal because all proximate Ø One is to change the original values by taking components of coal are directly related with each logarithm of one of the collinear predictors other by their percentage values as follows;

Pearson Correlation
shown in Model 1. Moisture% + Ash% + Volatile matter% + Fixed Ø Second is to exclude one of collinear predictors carbon% = 100 to evaluate the outcome shown in Model 2. In this situation the of the multiple regression coefficient estimates Model 1         The value of determination coefficient (R ) for model Indian, Indonesian, South African and Afghan coals I and mode II of the present study have been found from the published literature. The results for to be 0.93 and 0.92 respectively, which are Ind one sia n a nd Sou th Afr ica n c oal s s how reasonably close to its maximum value 1.00. The reasonable agreement with their experimental models developed in the present study have also values and those computed by model II of the been tested by taking proximate analysis results of present study shown in figure 5 And 6 Respectively.
However, values computed by model I of the those computed by model II of the present study and present study and those by models by Majumdar, by CRFI (Central Fuel Research Institute) models Gaut al's and CRFI (Cen tral Fuel Rese arch are reasonably close while those computed by model Institute) are significant by different than the I the present study and other models differ experimental values. The results for experimental significantly as shown in figure 7. HHV (Higher Heating Value) of Afghan coals and