A desire to transition away from fossil fuels to more sustainable alternatives has driven the development of energy conversion systems at pressures beyond the vapor-liquid critical point. Transcritical heat pumps, for instance, can decarbonize industrial processes that require heat beyond 100°C, such as drying processes in the food and chemical industry, when powered by renewable electricity. Additionally, supercritical power cycles enable the conversion of heat from previously unsuitable sustainable heat sources into electricity. Therewith, these power cycles can meet the rising demand for non-fossil power, partly driven by the electrification of heat.
However, these systems operate in a highly non-ideal thermodynamic region, where a fluid no longer undergoes a discrete phase transition from liquid to gas when it is heated. Instead, at supercritical pressures, fluids undergo a boiling-like process in which they remain in a continuous phase. During this pseudo-boiling process, fluids exhibit considerable non-linear variations of thermodynamic properties. These variations are most pronounced in the pseudo-critical region, where thermophysical gradients strongly influence flow behavior and heat transfer. The consequences of the variations are most pronounced in the heat exchangers of these energy conversion systems, which operate in the near-pseudo-critical region. In these heat exchangers, buoyancy effects are significantly more dominant than in similar heat exchangers operating with subcritical pressure fluids, leading to highly configuration-dependent heat transfer behavior. Currently, the understanding of these non-ideal effects remains limited, especially in an experimental context, hindering the design of efficient and safe equipment for supercritical energy conversion systems. To bridge this knowledge gap and support the successful implementation of supercritical pressure energy conversion systems, this dissertation investigates the influence of buoyancy on heat transfer in supercritical carbon dioxide (sCO2) flows, with a specific emphasis on thermal stratification in near-pseudo-critical heat exchangers.
To explore these phenomena, a novel experimental facility was developed, featuring a naturally circulated flow loop operating at supercritical pressures. The four-meter-tall structure exploits the substantial density gradients of sCO2 to induce a buoyancy force over its vertical legs, driving the flow. By integrating buoyancy-driven natural circulation as the flow-driving mechanism, the facility eliminates the need for mechanical pumps and ensures stable flow conditions for controlled experimentation. The current Natural Circulation Loop (NCL) provides a steady and stable circulation across a broad range of operating conditions. In this work, a steady-state flow rate equation for the natural circulation of supercritical pressure fluids is proposed and validated against experimental data, demonstrating close agreement between predictions and measured flow rates.
Under specific conditions, when the system’s mass flow rate is sufficiently reduced at a constant heating rate, the natural circulation loop becomes unstable. This instability manifests as system-wide oscillations in temperature and pressure, posing a thermal fatigue risk for high-pressure loops. These oscillations are identified as dynamically induced, driven by traveling density waves resulting from periodic deterioration in heat transfer within the NCL heaters. However, these oscillations occur only under a narrow set of conditions and can be mitigated by diffusing the density waves through the system. When appropriate mitigation measures are implemented, natural circulation loops can provide a reliable and stable passive circulation mechanism, making them well-suited for critical applications such as nuclear reactor cooling or for sensitive applications such as the current heat transfer experiments.
A test section integrated within the circulation loop enables optical investigations of flow behavior in a horizontal plate heat exchanger channel, providing insight into buoyancy-induced thermal stratification. The test section employs optical techniques that visualize refractive index variations to capture CO2 flow motion. These optical methods, used alongside heat transfer measurements, reveal highly transient flow phenomena within the heat exchanger channel for the first time in experiment. Shadowgraphy, in particular, proves effective in visualizing fluid motion in turbulent supercritical CO2 flows with imhomogeneous temperatures.
In the test section, a hydrodynamically developed flow is examined, with heating applied either from the top or bottom to impose a one-sided density gradient in the CO2. The results reveal strong stratifications in both heating configurations, occurring much earlier than expected compared to subcritical fluids.
In the two heat transfer configurations, buoyancy effects act in opposite ways. When CO2 is heated from the bottom, creating an unstable stratification, the channel flow is characterized by secondary motion away from the heated wall. Here, buoyancy enhances heat transfer, enhancing the effect of an increasing specific heat near the wall. As buoyancy effects intensify at higher heating rates, heat transfer progressively improves. Conversely, under stable stratification, when CO2 is heated from the top downward, the imposed density gradient suppresses turbulence-induced vertical motion, hindering heat removal from the heated wall. As a result, buoyancy opposes the heat transfer enhancement by variable specific heat. As buoyancy effects become more significant, the heat transfer deteriorates beyond an optimum. The heat transfer rates between the two configurations differ by up to an order of magnitude in the current experiments.
The current findings confirm that supercritical CO2 flows are highly susceptible to buoyancy effects and demonstrate how buoyancy can significantly alter heat transfer relative to a neutrally buoyant setting, often overshadowing other property variation effects. As such, this research contributes to the validation of numerical models and the resolution of previously divergent experimental heat transfer results, thereby aiding in the development of improved predictive models and optimized heat exchanger designs. By enhancing the reliability of supercritical energy conversion system designs, this work supports the broader goal of defossilizing industrial heating and power generation.
