Thermofluid Network Simulation of a Natural Draft Direct Dry Cooling System for a 50 MWe sCO2 Power Cycle for CSP Application

The promise of high thermal efficiencies and low initial capital cost make supercritical carbon dioxide (sCO2) power cycles an attractive technology for concentrated solar power (CSP) applications. As CSP plants are usually located in arid or semi-arid regions, cooling options that minimize water consumption are preferred. Natural draft direct dry cooling systems (NDDDCSs) offer an alternative to traditional dry cooling technologies such as air-cooled condensers (ACCs) or indirect natural draft dry cooling systems. NDDDCSs combine the benefits of low auxiliary power consumption with direct heat rejection, while minimising operational costs and noise. This study investigates the performance of a NDDDCS when utilised as cooling technology for a precooler and intercooler (combined in one tower) of a partial cooling with reheat (PCRH) sCO2 cycle for a conceptual 50 MWe CSP plant. Initial work involves the development of a zero-dimensional (0D) lumped parameter sizing model to assess overall system performance and aid in selecting the best-performing geometry between the evaluated cases. Thereafter, a higher fidelity simulation model is developed in Flownex SE software, which employs a thermofluid network approach to solve the governing mass-, momentum-, and energy balance equations for the air- and sCO2 streams. Empirical heat transfer and pressure drop correlations are also integrated for both air- and sCO2-sides to capture component characteristics. After verification with the 0D model, the thermofluid network model (TNM) is extended to include discretised finned-tube heat exchangers and a one-dimensional (1D) discretised air-side tower to simulate the natural draft. This model is valuable for providing detailed insights into potential maldistribution effects, particularly under off-design conditions, and is used to evaluate the performance of the cooling system under varying ambient temperature and part-load conditions. The best-performing system was selected based on minimising both lost work (sCO2-side pressure drop) and estimated total heat exchanger and tower material costs. The selected configuration features an inlet diameter of 55.79 m, inlet height of 11.14 m, tower height of 55.79 m and outlet diameter of 27.89 m, designed to meet combined heat rejection requirements for the precooler (PC) and intercooler (IC) at design point. At design conditions, the system’s air mass flow rate and total heat rejection rate are 3702 kg/s and 85.03 MW, respectively. Simulation of the cooling system using the discretised network model shows reductions of 4.7% and 0.8% in overall heat rejection and air mass flow rate at the design point, compared to the lumped model. While significant variations in fluid properties, pressure drops and heat transfer coefficients are observed on the discretised sCO2-side, the resulting air-side effects dominate differences from the lumped model results. Further simulations investigate the effects of off-design ambient temperature and crosswind variations. Significant differences in sCO2-side temperatures and heat transfer coefficients arise from ambient temperature changes. Flow and pressure drop characteristics during partial bypass operation (to avoid overcooling impact on the cycle) of heat exchangers are also illustrated. Simulations highlight the differing impacts of crosswinds on heat rejection, sCO2 outlet temperature and pressure drop for the PC and IC.