Date of Award


Document Type


Degree Name

Master of Science (MS)


Chemical Engineering



The convective heat transfer coefficient of coal water slurry flowing turbulently in a horizontal heated pipe was measured. The values of heat transfer coefficients from different slurry concentrations and flow rates were correlated using two approaches: the Dittus-Boelter type equation, and an analogy between heat and momentum transfer using Metzner and Friend equation.

The heat transfer data were taken in a 100 inches long, 1/2 inch inside diameter horizontal stainless steel pipe. The pipe was electrically heated with a DC power supply unit. Temperature measurements were taken using Copper/Constantan thermocouples embedded into the pipe wall while pressure drop was measured by a U-tube mercury manometer. The rheological properties of the slurries were measured with a Haake RV 100 rotational viscometer. The slurries were prepared using Beulah coal (North Dakota) with a mean particle diameter of 43 microns and Rosebud coal (Wyoming) with a mean particle diameter of 52 microns. The coal water slurry concentration ranged between 10.8 percent to 52.8 percent (by weight) while the flow rate ranged between 20 to 53 lb per minute.

The rheogram of the slurries indicated the existence of non-Newtonian behavior. The rheological behavior of the slurries were analyzed using the Power Law and the Yield Power Law models. The resulting fluid behavior and fluid consistency indices were used to calculate modified Reynolds and Prandtl numbers. In addition, an empirical equation was also used to estimate slurry viscosity from the viscosity of water and solid content. These dimensionless numbers were then used to correlate the heat transfer data represented as dimensionless Nusselt number in Dittus-Boelter equation, and Stanton number in Metzner and Friend equation.

The dimensional approach using the Power law model for calculating Prandtl and Reynold numbers resulted in the following correlation for heat transfer during slurry flow: 0.4 0.90 (Nu/Pr' ) = 0.023 ( Re') A similar correlation was developed when the Yield Power law model was used. The above correlation is recommended, because of its simplicity and because both the correlations were found to be of similar accuracy. However, the Metzner and Friend correlation was found to be much less accurate. Our data indicated wide scatter in the B-factor in the Metzner and Friend correlation and did not show the expected exponential relationship with the Prandtl number.