Using Simulations to Guide Material Development of Solid Sorbents for CO2 Capture

Friday, November 12, 2010: 10:15 AM
250 A Room (Salt Palace Convention Center)
Adam Berger and Abhoyjit Bhown, EPRI, Palo Alto, CA

One technique for separating carbon dioxide from flue gas involves the use of a solid sorbent in a packed that will preferentially adsorb the CO2. Once the sorbent becomes saturated with CO2, it can be regenerated by increasing the temperature, causing the CO2 to desorb and form a high purity stream that can be compressed for geologic sequestration.

Work has been completed on a finite element model that uses the physical properties of the flue gas and a given sorbent to thermodynamically model mass-transfer-dependent multi-component adsorption, temperature evolution, velocity variations, pressure effects, and energy usage. The model employs a four step cycle that includes an adsorption step in which flue gas is flowed through the bed, a heating step that desorbs most of the CO2, a purge step that further regenerate the bed, and a cooling step that readies the bed for adsorption. The results from this cyclic analysis can then be used to calculate metrics of performance such as CO2 recovery, product purity, and specific energy penalty for cyclic operation. Parametric analysis can be used for a specified sorbent to study the effects of cycle timing, adsorption conditions, regeneration temperature, column sizing, and particle size to maximize the performance metrics. Moreover, we can perform parametric analysis on the sorbent properties to determine which increases in material properties will most enhance the sorbents capture potential. This analysis can help guide sorbent development by comparing energy or recovery tradeoffs between individual sorbents or among different classes of materials, such as between chemisorption and physisorption sorbents.

Sorbent development is an active field with many new materials being developed and tested for working capacity and theoretical regeneration energies. Each of the sorbents being researched employs one of the two possible adsorption mechanisms: physical adsorption or chemical adsorption. In physisorption, the target molecules are attracted to the pore walls within a high surface area sorbents by Van der Waals forces and have a very low heat of adsorption roughly equal to the heat of sublimation. In chemisorption, the target gas undergoes a chemical reaction to bind to certain sites on the sorbent with a much greater heat of adsorption. A larger heat of adsorption intensifies the temperature dependence of the equilibrium adsorption isotherms, which means that under the right conditions a small temperature rise can cause a more thorough sorbent regeneration than in a similar sorbent with lower heat of regeneration. However, the greater heat of adsorption also requires a greater cooling load during adsorption and heating load during desorption to maintain the target temperatures, which ultimately results in a higher parasitic load on the power plant. Further, chemisorption sorbents tend to break down thermally at high temperatures and the binding sites can often be poisoned by molecules such as SO2 that cannot be readily regenerated or purged at low temperatures. On the other hand, though physisorption sites attract CO2 preferentially, they will also competitively adsorb other flue gas components, including nitrogen and water, which can dilute the product stream when they desorb during sorbent regeneration. This paper will examine and quantify the metrics of performance, specifically energy usage at a specified purity and recovery, for representative physisorption and chemisorption materials with otherwise identical properties. The results will be used not only to quantify the difference in specific energy between the modeled sorbents, but also to calculate the capacity increase necessary to compensate for the changes in other material characteristics.

Beyond this general comparison of adsorption mechanisms, we can also identify material properties that will most influence the metrics of performance within specific classes of sorbents. We can use this information to identify threshold and benchmark material property levels for each adsorption mechanism, which provides a mechanism for rapid screening within a class of synthesized sorbents. It also can be used to determine which measurements are necessary to determine the capture potential of a sorbent. For example, if the regeneration temperature has a large effect on working capacity, then measuring the sorbent stability at higher temperatures become important. Similarly, looking at or synthesizing materials that are more thermally stable becomes more of a priority. By identifying these areas of need both for screening and development through the use of simulation, we will help guide sorbent development towards advances that will have the greatest impact on capture effectiveness.

Extended Abstract: File Not Uploaded
See more of this Session: CO2 Capture by Adsorption II
See more of this Group/Topical: Separations Division