Heat Transfer, Mass Transfer and Scaling Characteristics in a Long Tube, Climbing Film, Pilot Plant Evaporator

ing characteristics could be investigated. Finally the pilot plant A single tube plant Kestner evaporator'was set up at was run under 1st effect conditions in order that a comparison Illovo to investigate heat transfer characteristics when could be made with industrial 1st effect evaporators (both simulating different evaporator Tests carried out have Roberts and Kestner). This particular trial would also confirm shown that the Kestner evaporator, which is used almost exOr reject the validity of previous work canried out i.e., whether. Or elusively as a 1st effect vessel in the south African sugarcane innot the pilot plant was a reasonable model of factory conditions. dustry, operates most favourably at the tail of the evaporator The aims of the project were as follows:system when compared to the conventional Roberts-type (i) The optimisation of heat transfer coefficient, juice flow rate evaporator currently in use. A seven week scaling test was and brix change across the evaporator and a study of the incarried out with the pilot plant operating in parallel with the 3rd terdependence of these factors. effect vessel at lllovo. If appears that the rate of scaling (ii) A comparison of heat transfer coefficients for Kestners and in the Kestner is substantially reduced due to the increased liconventional evaporators and, in the case of 1st effect, a quid and vapour velocities through the tube and the resulting three way comparison between pilot plant, industrial decreased juice residence time. Kestner and industrial Roberts vessels. Introduction (iii) An investigation of scaling characteristics of the pilot plant The Climbing-film evaporator was patented in 1899 by Paul compared to conventional evaporators and the effect of Kestner. Long tube vertical climbing film evaporators have been scale on heat transfer coefficient. employed in various industries (notably in fruit juice concentra(iv) An investigation of how effectively the pilot plant can be tion and the European sugar beet industry) for many years. In controlled and the problems encountered by sudden fluctuathe South African sugarcane industry this type of evaporator tions in feed, pressure and heating vapour supply for has been used on 1st effect only, subsequent effects usually convarious effects. sisting of the conventional Roberts-type evaporator which, although having several shortcomings in design, has remained Experimental Details popular due to its simplicity of construction and cheapness. A schematic diagram of the pilot plant installed at IUovo is Since the iuice residence time is minimal, Kestners reduce the shown in Fig. 1. problems .of inversion and colour formation at high temperatures. The 7m long tubes make it possible to achieve high evaporation loads on single pass operation and the high tube velocities result in high heat transfer coeficients. Easy accessibility provides simple maintenance and cleaning and if the evaporator is operating satisfactorily scale formation is slower than in conventional downflow evaporators. If the Kestner is to be provided with an independent separator more space will be taken up in the horizontal plane. As a result of the short residence time, the Kestner cannot be employed to store juice, as is often done with conventional evaporators to smooth out fluctuations in juice flow between the clarification and pan stages of the sugar manufacturing process. The Kestner requires boiling feed, i.e. boiling at the operating pressure or a supplementary juice heater and the instrumentation required for accurate control is more complex than for conventional evaporators. Allan1 emphasizes the advantages of using Kestners at the evaporator tail although some doubts have been expressed as to the validity of using Kestners on any effect except the first as it is believed that the increasing viscosity of the juice towards the evaporator tail will hamper the climbing ability of the film and hence prevent efficient operation. It is worth noting that there is a limit to the maximum viscosity and manufacturers2 emphasize this when quoting Kestner specifications. This 'maximum', about 100cP, is unlikely to be reached in a sugar mill operating normally where the final syrup viscosity is typically 75cP. The initial purpose of the pilot plarit was to determine whether the Kestner can be successfully operated in the 3rd, 4th and 5th effects of an evaporator 'quin'. A simulation test was performed in parallel with the 3rd effect (of a 'quad') at Illovo so that scalSyrup and hot water were mixed to the required feed brix in the steam heated, stirred, mixing tank. The feed was maintained at about 10°C below the required 'boiling' temperature in the thermostatically heated feed tank and was pumped using a variable speed mono-pump via juice heater to the column. Flashing was induced by boosting the juice temperature to about 5OC above the required boiling temperature in the juice heater and also by including a small orifice in the tube feed inlet pipe. The tube was 7m long,with a 50,8 mm OD, and was constructed from stainless steel and surrounded by a 101,6 mm OD, insulated, heating chamber. The vapour-liquid mixture leaving the column was separated in an empty, centrifugal-inlet chamber, the vapour being taken off under vacuum, condensed and collected in a tank. The product juice and condensed shellside vapour were also collected in tanks. (see Fig. 1). Vacuum fluctuations were minimised using two independent product tanks, which helped to ensure steady-state conditions during the tests. Feed and product were sampled for brix determination. A factory vacuum pump was used to provide the tubeside vacuum with a direct link to the top of column 'shellside' to draw off incondensible gases. The column shell-side vacuum was provided by the conditions of the heating vapour. Temperatures were measured using a 12 point recorderlthermocouple system. This proved useful in maintaining steadystate conditions as a continuous record of pilot plant temperatures was kept. (Generally a sharp temperature change indicated a deviation from steady-state operation). The tubeside vacuum was controlled using a pneumatic control valve linked to the cooling water outlet of the tubeside vapour condenser. The feed juice temperature was controlled using a thermostat/solenoid Proceedings of The South African Sugar Technologists' Association June 1978