Ocean Thermal Energy Conversion (OTEC) uses the temperature difference between the surface and deep layers to generate electricity. The temperatures in the ocean hardly vary between day and night or between the seasons, which ensures a very constant energy production with OTEC. In order for OTEC to compete with the current forms of energy production, optimization is necessary such that costs are reduced to make OTEC power generation economically feasible. In cooperation with the Delft University of Technology, Bluerise B.V. constructed a small scale OTEC-plant to test the performance of the OTEC-cycle and optimize its outputs. The 100 W OTEC demo consists of an organic Rankine cycle using pure ammonia as the working fluid that is downward condensed in a gasketed plate heat exchanger (GPHE).

Plate heat exchangers (PHEs) are known to have a wide range of applications due to their superior performance in relation to favourable heat transfer coefficients (HTCs), compactness, design flexibility and thermal effectiveness. However, the two-phase behaviour inside the PHEs is yet not fully understood, resulting in over- or underestimating by several published heat transfer and pressure drop correlations, which are limited by the range of conditions they cover. Tao et al. (2018) conclude that better predictions of flow patterns and dominant physical phenomena in PHEs will improve the calculation of heat transfer and pressure drop. Therefore, the goal in this research is to identify the flow patterns that occur in the GPHE and gain more knowledge on the dominant physical phenomena that are apparent in the downward condensing ammonia. In the future, this can be used to increase the accuracy of performance calculations of the condenser. Flow visualization experiments are performed on the GPHE to identify the flow patterns and how they relate to the performance of the GPHE. A visualization section is designed in the OTEC demo that allows for flow visualization. The visualization section involves the GPHE, a transparent visualization plate, illumination by a LED-strip and a high speed camera.

The first part of this thesis involves material research on glass and several polymers to produce the visualization plate. The setup requires a plate that is chemically compatible with ammonia and has a mechanical performance that is fit for the operational pressures inside the GPHE. Chemical tests prove that Polystyrene is chemically well resistant to ammonia for four days. The results of a finite element analysis predict that the mechanical performance of PMMA is promising but glass is critical for the GPHE. To ensure a safe setup, a material combination of PS/PMMA is selected. A visualization plate is manufactured from a corrugated PS layer on top of a thick PMMA base plate. The PS layer acts as a chemical resistant barrier between ammonia and the PMMA base, while PMMA ensures the mechanical performance. However, after 7 days of performing the visualization experiments, the plate showed limited durability with respect to ammonia, as cracks in the material emerged and the plate lost its transparency.

The second part of the research in this thesis involves two types of experiments. The first type involves flow visualization experiments on a single-channel configuration. For all conditions, only film flow and partial film flow are observed. A low mass flux shows partial film flow with a smooth film, and a high mass flux corresponds to film flow with rough film characteristics. Partial film flow includes dry-out areas on the plate, and it is concluded that these surface area voids increase with vapor quality and volumetric void fraction. Mass flux contributes to a pressure drop increase and therefore film flow corresponds to a higher pressure drop in the PHE with respect to partial film flow. The HTCs increase with mass flux but mostly with vapor quality. However, there seems to be an optimum for the HTCs with increasing vapor quality when partial film flow is observed. For partial film flow, the surface area voids increase with vapor quality and for large voids the slope of the HTCs decay. It is concluded that surface area voids have a negative influence on the HTCs. Proposed flow patterns maps by previous studies are not in accordance with the observed flow patterns inside the GPHE for the current experimental conditions. For this reason, a flow pattern map for downward condensing ammonia is proposed for this experimental configuration and conditions. The second type of experiments involves multi-channel experiments without visualization to predict the flow patterns for lower mass fluxes. The slope of the HTCs followed the same trend with increasing vapor quality for the observed partial film flow in the visualization experiments. This indicates that for this multi-channel configuration partial film flow occurs in the GPHE.