Sudip Sharma
About
I am a PhD candidate in the Department of Energy and Process Engineering at NTNU. I work in Professor Thomas A. Adams II's Process and Power group. My research develops next-generation CO2 capture. I focus on Magnetic Induction Swing Adsorption, or MISA. MISA regenerates a magnetic adsorbent using an alternating magnetic field. This replaces the usual approach of heating the bed with hot gas or steam. My work sits at the boundary of first-principles process modelling and machine learning. I build open-source IDAES/Pyomo models of adsorption cycles. I then pair these models with Gaussian-process surrogates and active learning. I use multi-objective optimisation to balance system performance.
Before my doctorate, I completed an Erasmus Mundus MSc in Biological and Chemical Engineering. I studied at four European universities: Reims, Tallinn, Liège, and Graz. My thesis covered the techno-economic and life-cycle assessment of polyaniline catalysts for PEM fuel cells. I hold a bachelor's degree in chemical engineering from Kathmandu University, Nepal.
Research
Almost 50% of global CO2 emissions come from large stationary point sources like power plants, cement, steel, etc. where capturing carbon directly from flue gas is one of the few realistic near-term routes to deep decarbonisation. My research develops a new way of doing this: Magnetic Induction Swing Adsorption (MISA), in which a magnetic adsorbent is heated directly and rapidly by an alternating magnetic field, instead of bby slowly forcing hot gas or steam through the bed, which is also usually produced using fossil fuels.
I a building a detailed first-principles models of these cycles in the open-source python framework and use them to push past "does it work"? to the harder question: under what conditions, and at what energy and cost, does MISA actually beat the established alternatives? Because every simulation is expensive, I coupled the model with machine learning surrogates (Gaussian process or Artificial Neural Networks) and multi-objective optimisation. This helps is understanding the limitations of MISA system after carefully examining the trade-offs.
The goal is an honest, reproducible, open-source assessment of weather induction-driven adsorption can cut the energy penalty of post-combustion carbon capture by enough to matter.
Publications
2026
-
Sharma, Sudip;
Adams, Thomas A..
(2026)
An Open-Source IDAES Framework for Simulating Inductively Heated Adsorption Processes.
Systems and Control Transactions (SCT)
Academic article
Journal publications
-
Sharma, Sudip;
Adams, Thomas A..
(2026)
An Open-Source IDAES Framework for Simulating Inductively Heated Adsorption Processes.
Systems and Control Transactions (SCT)
Academic article