Geothermal heat-energy based power-generation has been an area of intense focus and development in New Zealand for more than five decades now. There are immense benefits of geothermal power generation in terms of environmental conservation. Other benefits include constant base load capability, lower foot print etc. The geothermal heat energy resource can be classified as high-enthalpy and low-enthalpy heat source based on the temperature and whether geothermal brine comes out in liquid or vapor phase. New Zealand geothermal resources are largely under the low-enthalpy category which makes it necessary to use Organic Rankine Cycle (ORC) as the plant thermodynamic cycle.

ORC-based power plants use fluids with boiling point lower than water. ORC power plants are complicated systems with a significant influence of components on each other and performance of the entire system. Heat exchangers form a crucial part of geothermal power plants at all stages of the plant – starting with claiming a major portion of the capital cost for commissioning the plant right up to governing the working (and hence profitability) of the plant in terms of deciding the plant downtime (for maintenance) and plant performance (preheaters/vaporizers/superheaters should be able to provide a minimum degree of superheating to working fluid and this minimum value decides the maximum value of working fluid mass flow rate through the plant which translates into plant’s net electric power output). Although heat exchangers have such a critical influence on the life-time profitability of power plants, their designs are still based on empirical/semi-empirical models, developed over the years through extensive experimentation, which form a constraint towards predicting their performance if the unit is operating in off-design conditions or is different in some geometrical aspects from the units that were used in experiments generating the empirical/semi-empirical models.

The vaporizer is a critical component in a power generation system. It is where the motive/working fluid undergoes change of phase to power the prime rotor (expander) and generate electrical or shaft power. Figure 1 shows a schematic of an organic Rankine cycle power plant of the type of interest for the present work. It has been estimated that more than half of the heat exchangers employed in process industries involve two-phase flow on the shell-side [50], and yet two-phase flow patterns in cross-flow have received much less attention than in-pipe two-phase flow patterns. There have been a number of experimental studies on various single and multi-tube geometries [44, 50, 51, 73, 95, 106, 116, 153] with main focus on areas such as void-fraction prediction correlations and frictional pressure drop prediction correlations. The void fraction and pressure drop measurements obtained by these investigators produced bundle average or pitch average values that were used in the formulation of various correlations.

In this thesis, the aim is to develop a Computational Fluid Dynamics (CFD) approach towards simulation of a full-scale pentane vaporizer with the aim of building capability for troubleshooting installed vaporizer units. CFD analysis of single-phase heat exchangers (e.g. preheaters, superheaters) was also carried out to present a comprehensive CFD approach towards troubleshooting the heating side of a power-plant. Heat exchangers are also used on the condensing side but that was excluded from the CFD scope of the project although MATLAB™ codes were written for analysing condensers as well.

In order to achieve the target the project was divided into different stages – a) CFD simulation and validation of preheaters, b) MATLAB™ code development for preheaters and vaporizers, c) CFD simulation and validation for single- and multi- tube geometries, d) selection of a substitute fluid to ascertain CFD set-up tuning factors in view of the absence of pentane boiling experimental data. This is done by – i) comparing thermophysical properties of three fluids – water, R113 and Pentane, ii) analysis of dimensionless numbers characterizing bubble dynamics (Bubble Reynolds number, Eötvös number, Morton number) and boiling process (Weber number, Jakob number, Boiling number) , e) parametric CFD analysis of a Shell-and-Tube Heat Exchanger (STHE) slice geometry to obtain relationship between liquid level and heat transfer performance of the unit, f) CAD modeling of a full scale vaporizer unit, g) CFD simulation on a scaled down representative geometry of repeating unit within the full scale vaporizer to verify the CFD set-up, h) droplet carryover analysis with the help of MATLAB™ code and CFD simulation. The project was originally aimed at getting the simulation for the full-scale vaporizer’s repeating unit done at its end, but the time requirement estimation for multi-phase simulations proved to be an under-prediction. In the tasks completed, validation is carried out at each step and constraints/assumptions of the set-up clearly outlined and explained in the respective sections.

The validation for single-phase CFD analysis has been done against well-established theoretical models along with a mesh-dependence study. CFD analysis is also conducted on novel heat exchanger geometries to demonstrate capabilities of thermal and flow field visualization. A MATLAB™ code has been developed for PHEs with validation against experimental data. Phase-change (boiling) CFD analysis of single- and multi- tube geometries has been validated against experimental data. A substitute fluid has been selected for tuning of pentane vaporization CFD studies on the basis of dimensionless numbers that characterize bubble dynamics and boiling. CFD studies have also been conducted on a STHE slice geometry with five different inflow rates to obtain information about the impact of inflow rate on heat transfer performance and vapor volume fraction escaping the vaporizer unit. A CAD model for a generic industrial scale kettle-type vaporizer has also been developed. Based on the representative structure within the full scale geometry a scaled down representative geometry is used for CFD analysis. The CFD-Post™ data of the vapor velocity field of this representative geometry is used as an input to a MATLAB™ code to plot droplet trajectories. The MATLAB™ code is unable to factor in droplet vaporization. Droplet vaporization is simulated by CFD analysis imagining a worst case scenario where the droplet is traveling straight up a vapor outflow channel.

The main tasks for future work on this project would be – a) do CFD analysis of the full scale vaporizer and benchmark computational requirements, investigate any potential optimizations with regards to computational time and hardware resources, and b) integrate droplet vaporization code into the droplet trajectory plotting code. A preliminary design for a lab has also been presented that can be used to conduct experiments providing detailed information about parameters required to tune CFD model set-up.

In conclusion, it can be said that the project has demonstrated a validated CFD analysis approach towards understanding and troubleshooting heat exchangers used in geothermal power plants and in general this approach can be extended to any process industry that uses heat exchangers.