Heat transfer and pressure drop for flowing wood pulp fibre suspensions

Abstract

Heat transfer operations are an integral part of many chemical processes. The efficient utilisation of thermal energy in these systems has in recent years become increasingly significant because of the rising cost of energy. The primary aim of this work was to investigate aspects of heat transfer to wood pulp fibre suspensions. The flow of wood pulp fibres is unique. The ability of the fibres to flocculate results in a suspension which flows with a pressure drop greater than that of water at low flow rates, and less than that of water . . at high flow rates. These effects have been investigated in detail. It might be expected that the fibres would have a similar effect on the heat transfer characteristics (this concept being the basis of the Reynolds analogy between heat and momentum transfer), however until now there has been no experimental evidence to confirm this. Heat transfer to suspensions of bleached Kraft pulp fibres was measured first in an annular test section. A limited range of pulp concentrations (0% < C < 1 %) and bulk velocities (0.2 mis < u < 4 mis) were tested because of limitations of the test rig. The addition of a small amount (0.08 %) of fibres caused an increase in the heat transfer coefficient above the water value. Further addition of fibres decreased the magnitude of the heat transfer coefficient towards the water value. A· new test rig was designed and built, and heat transfer and pressure drop were measured simultaneously for a much wider range of suspension concentrations (0% < C < 4%) and bulk velocities (0.5 mis< u < 10 mis). The suspension heat transfer coefficients were higher than for water at low flow rates and lower at high flow rates, and showed similar trends to the pressure drop data. However the cross-over points of the curves with their respective water curves differed, indicating some differences between the two transfer mechanisms. An equation derived using a curve-fitting technique correlated well with the data. The results were also modelled in terms of the modified eddy diffusivity. The model accurately predicted the observed trends in the pressure drop data. However the inability of the model to account for the influence of the plug core on the temperature profile caused errors in the heat transfer predictions. The effects of fibre type and fibre aspect ratio were determined by comparing the results for the bleached Kraft fibres with data measured for unbleached Kraft pulp, mechanical pulp, and model Nylon fibres. While both types of Kraft pulp gave similar results, the mechanical fibres gave different trends because of their greater size distribution and the effects of the smaller fibrillar material. It was observed for the nylon fibres that the flow patterns were strongly dependent on the aspect ratio. The fibres with low aspect ratios (r < 60) did not flocculate and their influence was to damp turbulence within the fluid, reducing both the pressure drop and the heat transfer. Conversely, at high aspect ratios (r > 60) the fibres behaved more like wood pulp fibres, causing regions of heat transfer enhancement and/or reduction depending on the flow conditions. The effect of the addition of wood pulp fibres to a fouling solution was also investigated. The aim of these experiments was to see if the fibres could be used as an unconventional fouling mitigation technique. The addition of a small amount (0.05%) of fibres caused the rate of fouling to decrease. This decrease in the rate of growth of the deposit was the result of fibres abrading the heated surface, and damping diffusion of the fouling species through the fluid. A decrease in the rate of deposit build-up would allow for a longer time between cleaning cycles in heat exchangers and pipelines. There may therefore be applications where small amounts of pulp fibres could be added to a process fluid to reduce the rate of fouling without negatively influencing the heat transfer rate.