The investigation of the transient start-up behavior of natural draft dry cooling towers in dispatchable power plants

Abstract

To provide electricity with clean and renewable resources, solar energy is considered as one of the best. In hybrid Photovoltaic/Concentrated Solar Power (PV/CSP) power plants, continuous electricity could be achieved when CSP component works in a dispatchable mode. As a significant part of CSP system, natural draft dry cooling towers (NDDCTs) are preferred for heat rejection because of no water and parasitic power consumption. When CSP system is called upon to start up frequently in hybrid PV/CSP power plants, NDDCTs have to start-up/shut-down ceaselessly. To understand the transient start-up process of NDDCTs under both windy and windless conditions, this thesis establishes 1-D theoretical model and 2-D, 3-D numerical models in light of a 20 m short NDDCT built up at Gatton campus of the University of Queensland.(1) A one-dimensional (1-D) theoretical model is presented to estimate the start-up process of NDDCTs. The start-up process is analyzed in two successive stages. In the first stage, the dominant mechanism is natural convection operating via the generation and propagation of hot plumes rising from the heat exchanger surface. In the second stage, the air flow is driven by the draft powered by the difference in the inside and outside densities.(2) To verify the previous theoretical analysis and more details during the start-up of NDDCTs, a two-dimensional (2-D) numerical model is established under windless condition. The study finds that, starting from cold, the air flow inside NDDCTs experiences three stages indicated by Richard number (Ri 10) before the steady-state is reached. In the first stage, natural convection is dominant, followed by mixed convection. In the third stage, the draft finally becomes the dominant mechanism and is sufficient to explain subsequent transition to steady-state and onwards. The Richardson number is introduced to identify the boundaries between these three.(3) To investigate the crosswind effects from 1 to 15 m/s on the start-up of NDDCTs, a simplified theoretical model and a three-dimensional (3-D) numerical model are built up. The simplified theoretical model identifies two distinct flow patterns that correspond to two ends of the wind speed range. The intersection of asymptotes technique is adopted to find when the flow switches from one to the other. The switching wind speed corresponds to the longest start-up period as well. Results of 3-D numerical analysis agree with these observations and provide more detailed insights. It is found that natural draft dominates the air flow at wind speeds below the cross-over and crosswind is dominant at higher wind speeds. The flow field is uniform only if a single mechanism becomes dominant: either natural draft or forced convection caused by crosswind. The start-up duration is the shortest/longest when the air flow is the most/least uniform.(4) To accelerate the start-up period of NDDCTs under crosswind, windbreak walls of different angles of attack (0°, 30° and 60°) are introduced under the heat exchanger inspired by theoretical analysis, which suggests a quicker start-up with the increasing proportion of crosswind (1~15 m/s) redirected into the tower via the heat exchanger by windbreak walls. The start-up time with crosswind always follows a similar trend, i.e., first increasing to the peak and decreasing monotonously until a critical speed, beyond which the start-up time keeps almost constant. The air flows through each individual heat exchanger bundle due to the interaction between the natural draft (due to heating) and the crosswind effects.(5) To predict the time-taken for the start-up process of NDDCTs with different geometric parameters and initial conditions under windless condition, dimensional analysis is made to first determine the significant variables governing the transient start-up process. Buckingham Pi theorem is subsequently adopted to identify the relationship between the air mass flow rate through the heat exchanger and the time-taken during the start-up process of NDDCTs. Then, numerical tests are carried out to investigate the start-up process of NDDCTs with different geometric parameters including the tower height and base diameter, over a range of initial condition, i.e., the temperature difference between the heat exchanger and the ambient air. The numerical results demonstrate that the base diameter will not affect the time-taken during the start-up process of NDDCTs in the absence of cold air inflow. However, the time-taken for the start-up process of an NDDCT rises with the increase of the tower height, but decreases with the increase of ambient air-heat exchanger temperature difference.The outcomes in this thesis give solid support to improve the dispatchability of CSP systems employing NDDCTs for heat rejection.