Experimental Investigation of the Supercritical CO2 Condensing Cycle

by Thomas Conboy & Steve Wright
Sandia National Laboratories

Click Here To Access The Complete Paper
Click Here To Access The Presentation

The S-CO2 recompression Brayton cycle has shown promise as an efficient power conversion system for advanced high temperature nuclear reactors and other heat sources. In a typical proposed cycle, the compressor inlet operates at a pressure comfortably above the critical point to avoid two-phase conditions at any point within the cycle. However, it has also been proposed to operate the main compressor at cooler, single-phase liquid conditions below the critical pressure. For an operating CO2 Brayton cycle, this would result in gas-phase CO2 at the gas cooler inlet, and condensation within the primary waste heat rejection system. Exit conditions of the cooler would be a saturated or subcooled liquid. The net effect of expanding the operating envelope as described, creating a 'condensing' Brayton or Rankine cycle, is greater efficiency for a given turbine inlet temperature.

The downside of this approach is that it requires lower heat rejection temperatures that may not always be available. Consequently, it may not be desirable to design a plant specifically for 'condensing' operation, but instead to design a 'standard' supercritical CO2 cycle which has the flexibility to take advantage of lower environmental temperatures when possible. This is currently standard practice in steam-Rankine plants, which are known to operate at higher efficiencies in cooler weather. At many locations within the US, environmental temperatures in winter or at night are sufficiently low that CO2 condensation below 88°F (31°C) is feasible.

Sandia performed a series of experiments to demonstrate the feasibility of operating the condensing cycle using hardware designed for operation above the critical point. Initial testing used the Sandia S-CO2 Compression Loop. This is a high-pressure S-CO2 loop (up to 2500 psi) with a compressor, heater, and a downstream expansion valve, but no turbine. Tests were performed to show that the radial compressor could effectively compress liquid CO2. This also proved that the gas foil bearings, shaft seals, and other turbomachinery internals can operate with liquid CO2 at the compressor inlet. Other tests expanded on this to show that the compressor could also be operated with gas-phase CO2, and even two-phase mixtures (above 1000 psi). In addition this research loop was modified by adding a 50kW heater to warm the CO2 following compression. With this upgrade, it was then possible in a single experiment to compress liquid CO2, heat it in the heater, and then expand it in a valve, providing single phase gas to the gas chiller or a two-phase mixture to the gas chiller to demonstrate its performance as a condenser. The gas cooler used in initial tests was a spiral heat exchanger.

Following these results, similar tests were run on Sandia's larger S-CO2 full Brayton cycle. The main compressor was run with subcooled liquid and with saturated liquid CO2 inlet conditions while generating electricity. This loop is configured with Heatric printer-circuit heat exchangers (PCHEs), widely considered a critical component of the compact supercritical CO2 power conversion cycle. Therefore these tests also confirmed the ability of these advanced geometry heat exchangers to operate as high pressure CO2 condensers (again without modification).

The fundamental outcome of these tests is the flexibility implied for operating characteristics of a future CO2 power plant: a full-scale multi-megawatt plant may be designed for 'standard' S-CO2 Brayton operation, but used in the condensing cycle mode as cooler temperatures are available, without significant losses or damage to hardware. Further, gains in efficiency allowed by use of the condensing cycle increase the competitiveness of the CO2 cycle for application to lower temperature heat sources in cool climates, such as fossil, geothermal, solar thermal and light-water reactor (LWR) systems.