Page 153 - Advanced Mine Ventilation

Generation of Respirable Coal Dust 133

9.5 Statistical Analysis of Data

Because the data presented and discussed earlier show great scatter, a statistical
analysis was done to determine the correlation coefficients and regression coefficients
of RDI with respect to the various properties of coal.
Standard library programs were used for this purpose. It was shown earlier [5] that a
dimensional analysis of the problem suggests a relationship between the RDI and the
various properties of coal, which may be expressed as:

x y
a b
RDI ¼ constant $ z
c
where a, b, and c are properties of coal, and x, y, and z are their exponents.
To obtain such a relationship in this case, the logarithm of the variables were used
as input to the statistical analysis program. RDI was the dependent variable, and the
moisture content, ash content, volatile content, fixed carbon content, ratio of fixed
carbon/volatile and fusain content were used as the independent variables.

9.5.1 Correlation Between the Logarithm of Variables
This was obtained using library program “PPMCR” (Pearson product moment corre-
lation coefficient). The results are given in Table 9.4.
Moisture appears to be well correlated with fixed carbon with a correlation coeffi-
cient of 0.4224. It was, therefore, not used in the subsequent regression analysis.

9.5.2 Regression Analysis
The data from the PPMCR analysis were used as input into another library program
called “UPREG,” which carried out a regression analysis. The fixed carbon content,

Table 9.4 Correlation of Respirable Dust Index (RDI) With Coal Properties

Coefficient of Correlation with
Dependent Variable, RDI
Variable Number Independent Variable (Variable No. 1)
2 Moisture 0.1680
3 Ash (dry) 0.1998
4 Volatile content (DAF) 0.3043
5 Fixed carbon (DAF) 0.4011
6 Ratio Fixed Carbon 0.3226
Volatile Content
7 Fusain 0.3664