OTEC condenser heat exchanger analysis

Abstract

The increasing demand in renewable energy motivates the search for new renewable energy sources. The ocean is a vast energy source of both thermal and kinetic energy, however methods to harness this energy haven’t reached maturity yet. Ocean Thermal Energy Conversion (OTEC), a method to harness the thermal energy of the ocean, could change this. In order to get to this stage of maturity, experimental and numerical research on the OTEC power cycle was done by Goudriaan [12] and Kuikhoven [23]. However, the numerical model developed showed inaccuracy for the condenser and lacked physical insight into the processes of heat transfer. The condenser and consequently, the cold water pipe, are a significant part of the investment costs of an OTEC plant. Therefore, the goal in this research was to increase the accuracy of the performance calculations of the condenser and to gain more knowledge on the physical phenomena in the condenser heat exchanger.

Two different methods were tested in order to improve the condenser model. In the first part a heat transfer correlation is fitted to the experimental data. A literature study was done on two phase heat transfer correlations to investigate what phenomena occur in two phase condensation flow. Relevant dimensionless quantities were chosen accordingly. The quantities used are the conventional Reynolds and Prandtl number. The mass transfer that occurs through absorption was represented by the Schmidt number.

The resulting correlation fit was tested on accuracy using the experimental data obtained by Goudriaan [12] and Kuikhoven [23]. To be able to compare the results of this research with their research, the in-depth validation was done for the same working fluid mass flow of 0.010 kg/s that was used by Goudriaan [12] and Kuikhoven [23]. The Schmidt number showed a positive effect on the accuracy.

The correlation was extrapolated for other working fluid mass flows as well. It was found that the correlation fit is a useful tool for the working fluid mass flows that have a large amount of experimental data available (the cases 0.007, 0.010 and 0.013 kg/s) but isn’t as accurate for working fluid mass flow with a small amount of experimental data available (0.005 kg/s).

In the second part of the report a detailed condenser model was developed using the method of Fernández- Seara et al. [10]. This model was used to provide more physical insight into the two phase mixture behaviour in a condensation process. The model is also capable of predicting the required pressure on the working fluid side to provide full condensation.

Literature research was done on different aspects of the condensation process in a plate heat exchanger. Film thickness correlations were tested. Different heat transfer correlations developed for plate heat exchangers were investigated. Flow pattern theory and flow pattern maps were used to predict the flow behaviour. The method of Fernández-Seara et al. [10] was used to predict mass transfer through the vapor-liquid interface in the working fluid. A pressure iteration loop was added to predict the condenser pressure on the working fluid side.

The initial results were slightly off compared to the experimental measurements. The reason for the inaccuracy is expected to be the neglection of surface tension effects on the wettability of the working fluid. Dry surface voids are expected to occur in the liquid film with decreasing concentration. A correction factor based on surface tension was added to include this effect. The 0.010 kg/s case was validated in detail and showed decent accuracy. At 0.007 kg/s working fluid mass flow deviations were slightly more significant but the trends of the characteristic variables were similar to the experimental data. The mass transfer iteration loop showed inaccuracies and instabilities at the working fluid mass flows of 0.005 and 0.013 kg/s.

The model showed that for the range tested the condenser performs better with increasing concentration of ammonia. The non-linear behaviour of the mass diffusivity decreases the mass transfer coefficients at decreasing concentrations. The surface tension correction factor also adds to the decrease in performance, however, if the mass flux of the working fluid would be higher, this correction factor might not be valid because the dry surface voids might not occur any more. The overall heat transfer coefficient increases towards the outlet of the condenser.