Abstract:
The application of plate heat exchangers has increased significantly because of their higher heat transfer and lower fouling tendency compared with tubular heat exchangers. Several studies have been carried out to investigate fouling in plate heat exchangers. However, while crystallisation fouling is probably the most serious fouling mechanism, it has never been studied in detail. The current study involves the investigation of calcium sulphate fouling in three different plate heat exchangers, namely, Alfa-Laval POl, APV TRl, and a flat plate heat exchanger. The main objective of this research project was to investigate the effects of various operating parameters such as flow velocity, solution concentration and temperature (wall and bulk) on fouling. Fouling increased with increasing solution concentration and temperature and decreasing flow velocity. The solution concentration was found to be an important parameter in the initial ' stages of the process only. Its importance decreased significantly once the deposits were formed. Pressure drop across the plate heat exchanger was measured along with the overall fouling resistance. The increase in pressure drop with deposit formation directly reflects the partial blockage of the passages. The crystalline samples were analysed using Energy Dispersive X-Ray Analysis (EDX) and X-Ray Diffraction Analysis (XRD). A scanning electron microscope and a X-ray diffractometer have been used for this purpose. The deposits were found to be pure gypsum crystals. Fouling in parallel and series flow plate heat exchangers has been investigated and it was found that both arrangements fouled to the same extent when the mass flow rate was the same. However, the series flow arrangement fouled comparatively more for the same flow velocity. The design of corrugations near the fluid inlet and the outlet, which are responsible for the fluid distribution over the whole plate, is very important. Under clean conditions, better design of the APV TRl plates resulted in higher heat transfer and lower pressure drop compared with the Alfa-Laval POl plates. Once deposition started, the fluid distribution deteriorated resulting in considerably more fouling on the APV TRl plates. Some experiments have been carried out in a flat plate heat exchanger to investigate fouling under simplified flow conditions, for better understanding of the fouling mechanism. The presence of calcium sulphate particles (undissolved matter and crystals removed by the fluid) was investigated by installing an in-line filter before the plate heat exchanger and by recycling, without any filtration, part of the solution coming out of the plate heat exchanger. This increased fouling by up to 60-70%. Particulate fouling did not contribute much. It was the generation of extra nucleation sites which enhanced the crystallisation rate significantly. Fouling in the plate heat exchanger (Alfa-Laval POl) has been compared with that in a double pipe heat exchanger. The plate heat exchanger fouled significantly less at the same flow velocity. It was concluded that fouling would still be less in the plate heat exchanger even for identical wall shear stress in both heat exchangers. Several mitigation techniques such as teflon coating, addition of alumina particles to the process fluid, and pulsating flow have been tested successfully to reduce calcium sulphate fouling. 3-4% acetic acid was used to clean the fouled heat exchanger. It usually took around 2 hours to remove all the deposits. Attempts have been made to model the fouling behaviour. In the model, the deposition rate has been expressed as a function of the deposit mass. Removal of the gypsum deposits by turbulent bursts was found to be negligible. Some efforts have also been made to simulate the flow in a clean and a fouled flat plate heat exchanger and in a channel with a single corrugation on the bottom surface to understand local fouling mechanisms.