Grams - a Dynamic Intensification and Optimization Platform for Modular Chemical Process Systems

A significant portion of future demands for chemical feedstocks is predicted to be met by unconventional and distributed resources such as associated shale gas, stranded natural gas, coalbed methane, landfill gas, biogas, and fuel gas. Current roadblocks to widespread utilization of these resources often include a lack of pipeline infrastructure, sparsity, seasonal variability and uncertainty in feedstock volumes and compositions, and occasional high levels of contamination. Small-scale, modular chemical process systems (MCPS) can be a promising solution to tap into these unconventional resources. They offer increased flexibility, agility and ability to accommodate varying market demands. However, poor economies-of-scale and high capital intensity prohibit most small-scale MCPS to be economically viable over their large-scale counterparts. We argue that this limitation can be overcome through dynamic process intensification that combines multiple phenomena, such as separation, conversion and intermittent storage, within a single equipment in a periodic manner. One example of such a multifunctional MCPS technology is the sorption-enhanced reaction process (SERP). In SERP systems, in situ capture of undesired reaction byproducts by adsorbents favors the desired reactions in catalysts, thereby increasing the overall yield of desired products. To this end, advanced modeling, simulation and optimization of such dynamically intensified and complex systems can provide critical breakthroughs needed for small-scale MCPS towards the economic production of chemicals utilizing unconventional feedstocks.

We have developed GRAMS (Generalized Reaction-Adsorption modeling, optimization and Simulation), which is a first-of-its-kind computational framework for the optimal design and operation of dynamically intensified packed-bed systems. For given feeds and products specifications, GRAMS is used to optimally design the cycle configurations, column design specifications and process operating conditions of a periodically operated MCPS [1-3]. The GRAMS platform combines an in-house dynamic process simulator with a cost estimator and a data-driven constrained grey-box optimizer [4]. The high-fidelity process simulator is based on a first-principles-based model of a generalized reaction-adsorption system to predict the performances of processes incorporating periodic gas adsorption-desorption (e.g., pressure/temperature/vacuum swing adsorption: PSA, TSA, VSA, SMB), reaction (e.g., PFR, SMBR), or a combination of both (e.g., SERP). The model predictions are extensively validated with experimental data for industrially relevant pressure swing adsorption (PSA), steam methane reforming (SMR), methanol synthesis, sorption-enhanced SMR (SE-SMR), and sorption-enhanced water gas shift reaction (SE-WGSR) processes. The framework has been used for the optimal synthesis of three multi-mode, multi-step and periodic SERP systems, namely SE-SMR, SE-WGSR and sorption-enhanced methanol (SE-MeOH). The optimized SE-SMR produces hydrogen from natural gas with 35% higher productivity and more than 10% lower cost in comparison to existing small-scale systems [2]. Furthermore, the novel SE-MeOH process, designed using GRAMS, could lead to more than 7% improvement in methanol yield with only 2% decrease in production capacity [3].

References

[1] A. Arora, S. S. Iyer, and M. M. F. Hasan, “GRAMS: A General Framework Describing Adsorption, Reaction and Sorption-Enhanced Reaction Processes,” Chem. Eng. Sci., vol. 192, pp. 335–358, 2018.

[2] A. Arora, I. Bajaj, S. S. Iyer, and M. M. F. Hasan, “Optimal Synthesis of Periodic Sorption Enhanced Reaction Processes with Application to Hydrogen Production,” Comput. Chem. Eng., vol. 115, pp. 89–111, 2018.

[3] A. Arora, S. S. Iyer, I. Bajaj, and M. M. F. Hasan, “Optimal Methanol Production via Sorption Enhanced Reaction Process,” Ind. Eng. Chem. Res., 2018.

[4] I. Bajaj, S. S. Iyer, and M. M. F. Hasan, “A Trust Region-based Two Phase Algorithm for Constrained Black-box and Grey-box Optimization with Infeasible Initial Point,” Comput. Chem. Eng., 2017.