The application of friction/heat-transfer correlations to cooling-tower design

The thesis of this paper is that friction and heat-transfer properties of cooling-tower packings are correlated, i.e. that an improvement in the heat-transfer properties can be obtained at the expense of an increase in friction. For any given shell design and packing surface a definite friction factor will then produce the best cooling-tower performance. The precise nature of the friction/heat-transfer relation depends on the method of changing the shape of the packing, e.g. a progressive change in the alignment of corrugated sheets to produce progressively narrower constrictions in the air-flow path. The limited experimental evidence now available, however, suggests that differences between various methods are not pronounced, i.e. that the results for all film-type packings free from the more obvious design faults are reasonably well represented by a general friction/heat-transter correlation. For this correlation a chart is prepared from which the optimum friction factor, the economic packing surface per unit ground area and the corresponding cooling-tower performance coefficient can be read off for any values of two design constants, namely the air-flow resistance of the tower shell, and the comparative cost of extending the tower ground area and the packing surface. The values of these design constants are discussed for various applications, and it is shown that the optimum friction factor can vary over a range of 0.1?0.6, being lowest for small mechanical-draught towers situated on expensive ground, and highest for large natural-draught towers. Methods of determining the economic fan power or chimney height are described and illustrated by worked examples. With the aid of the design chart, economic mechanical- and natural-draught tower designs can be prepared very quickly. For those film-type packings which do not quite satisfy the general correlation, the design chart may be used in conjunction with two correction factors. For larger departures from the genera- l correlation, such as those to be expected from splash-bar packings, individual design charts can be constructed by the methods described.