D11.2 Experimental investigation of CO2 liquefaction for CO2 capture from cement plants

This report gives a detailed description of the 10 ton per day low temperature CO2 liquefaction and phase separation laboratory rig installed in SINTEF Energy's laboratories in Trondheim, including specifications for its main components. This is followed by experimental results from the first experimental test campaigns with gas chromatography (GC) instruments connected.
The phase separation part of the experimental set-up consists of two separator stages. A binary gas mixture first enters the main separator, where the bulk separation is performed. The liquid CO2 is then heated and expanded to a lower pressure before it enters the second purification separator. The rig is well instrumented with pressure, temperature, mass flow and level sensors, as well as a gas chromatograph.
On 05.09.2018, experiments with a feed composition of approximately 63 mol% CO2 and 37 mol% N2 to the first separator were conducted. Tests were performed with liquid levels between 16 and 32 cm, and pressure levels between 30 and 35 bar in the first separator and 12 and 8 bar in the second separator. The final liquid CO2 product purities were measured. Close to equilibrium conditions were achieved. The results show that the product purity can be controlled by the pressure level of the second separator stage.
On 10.09.2018 an experiment with 32 cm liquid level in both separators, 30 bar pressure in the first separator and 8 bar pressure in the second separator were performed for a feed gas composition of 76 mol% CO2 and 24 mol% N2. Measurements of the composition were performed at the inlet to the first separator and at the vapour and liquid lines from both separators. Close to equilibrium conditions are achieved in the two separators. The experimental results for purity and CO2 capture ratio (CCR) are in good agreement with results from a HYSYS simulation of the process. A CCR of 83 % is estimated based on the experiments compared to 82 % in the simulations. It should be mentioned that the experimental rig and the simulation model thereof lacks an internal recycle stream which can increase the obtainable CCR considerably.