Improving the Second law efficiency of a cryogenic air separation unit

Abstract

One-quarter of the worldwide greenhouse gas emissions is emitted by fossil fuel based power plants. In order to limit future climate changes caused by these emissions, several types of CO2-capturing power plants are currently being developed. An integrated gasification combined cycle (IGCC) is one of the most promising alternatives. It is the mission of a European collaboration project called DECARBit to enable the commercial use of this type of power plant. One of themain process units of an IGCC is the air separation unit (ASU). It provides both oxygen and nitrogen to the gasifier, and nitrogen to the gas turbine. The main objective of this thesis is to improve the Second law efficiency of a cryogenic ASU, with a focus on the use of novel distillation concepts.

Improving the Second law efficiency of a process is equivalent to improving its exergy efficiency. In this thesis, the exergy efficiency is defined as the desired change in exergy contents of the ASU products divided by the net amount of added work. Using exergy analysis, it is shown that the exergy efficiency of a state-of-the-art two-column ASU with a pumped liquid cycle is approximately 35%. Most of the exergy destruction is located in the compressor after-coolers, the distillation section, and the main heat exchanger.

The irreversibilities in the compressor after-coolers are caused by the use of cooling water. They can almost completely be eliminated by transferring the heat of compression to the ASU products instead of to the cooling water. The achieved reduction in exergy destruction corresponds to almost 1% of the net electric efficiency of the IGCC and it increases the exergy efficiency of the ASU to approximately 70%.

Two alternatives are presented that can improve the distillation section efficiency. The first one is the addition of a third distillation column; it can reduce the exergy destruction in the distillation section by approximately 30%. The second option is to improve the heat integration of the two distillation columns, by using heat-integrated distillation stages (HI stages). These HI stages are the basis of a relatively novel distillation column configuration called a heat-integrated distillation column (HIDiC). In an ASU, the use of HI stages enables a lower operating pressure in the high-pressure column, which reduces the required work input to ASU. Depending on the amount of heat-transfer capacity per HI stage, the exergy destruction in the distillation section can be reduced by 20 to 30% due to the use of HI stages.

HI stages and HIDiCs are not yet in industrial use. So far, only two complete HIDiCs have been built, both using concentrically-integrated columns equipped with structured packing. They have proven the feasibility of the HIDiC concept, but detailed knowledge on the performance of the columns is still very scarce. As a result, simulations of packed concentric HIDiCs still involve several uncertainties. They are related to the achievable overall heat-transfer coefficient, to the performance of a ring-shaped distillation column, and to the effects that a radial heat flux has on the column performance. In order to obtain more insight into these phenomena, two research directions have been pursued: a theoretical one and an experimental one.

The theoretical work concerns the further development of a model for the simultaneous transfer of mass and thermal energy, based on the theory of irreversible thermodynamics. The model describes the vapour–liquid interface region of a mixture as a series of connected control volumes that together represent a vapour film, the interface, and a liquid film. This interface region is located in between the bulk vapour and bulk liquid phases; the conditions at its boundaries are equal to the adjacent bulk phase conditions. A routine has been developed that calculates the thermal and molar fluxes through the interface region, based on input values for the boundary conditions, or driving forces. The film thicknesses ratio is found by requiring consistency between the entropy productions calculated using the entropy balance and using the product-sum of conjugate fluxes and driving forces.

By applying this model to a nitrogen–oxygen mixture, it has been shown that the direct coupling between heat and diffusion fluxes has a considerable influence on the calculated values of the fluxes. The measurable heat flux is most sensitive to the coupling effect, which makes a correct description of the effect especially important when simulating a HIDiC. Another important model parameter is the number of control volumes that is used to represent the films. The effect of the interface resistances on the calculation results was relatively small.

The experimental work concerns the development of a new experimental HIDiC. The designed set-up consists of a cylindrical inner column with a diameter of 14 cm that is surrounded by a ring-shaped outer column with a diameter of 22 cm. A difference in operating pressures causes thermal energy to be transferred from the high-pressure inner column to the low-pressure outer column. Both columns will be equipped with 1.6 m of structured packing and will operate at total reflux conditions. The set-up is designed to operate at cryogenic temperatures, elevated pressures, and high oxygen concentrations. At the top of the set-up, two copper-brazed plate heat exchangers will be used as condensers, using evaporating nitrogen as coolant. Electrical heaters with a maximum duty of 25 kW will be used as reboilers.

Radial and angular temperature and composition gradients inside the columns will be measured directly at several height levels, in both the vapour and liquid phases. These measurements can also be used to determine the separation efficiency of the columns. The total amount of thermal energy transfer will be obtained based on two independent measurements of the condenser and reboiler duties of both columns. The set-up can also be used to assess the coupling between thermal and molar fluxes.