Energy conversion systems today, whether they are mining operations, power plants, refineries, or gasification reactors are generally characterized by large installations, long lifetimes, slow capital stock turnover, and substantial capital commitments. Smaller units that are mass produced, modular, and controlled in aggregate by cheap automation and control systems represent a new approach to scale in energy production and conversion infrastructure.

This research seeks to test the scaling laws for the energy industry. Do the conditions that placed a premium on large-scale energy infrastructure persist, or are there economies of scale to be reaped from an aggregation of smaller units? This question will be applied to the Fischer-Tropsch synthesis, a process by which an array of liquid hydrocarbons is produced from carbonaceous synthesis gas, through a case study that is in its own right a deliverable result. This case study will produce a model testing the potential advantages of liquid hydrocarbon production under the small-scale paradigm.

The Fischer-Tropsch process entails hydrogenation of adsorbed CO to form CH₂ "monomers" for stepwise oligomerization. At each stage, adsorbed hydrocarbon can desorb, hydrogenate, or continue chain growth with another monomer. The result is a suite of hydrocarbons of varying chain length and industrial applicability. Historically, research and development of the process assumed singular large-scale structures, embracing and discarding reactor designs under that assumption. A number of challenges to Fischer-Tropsch reactor design, however, such as pressure drops with bed length and the need to rapidly shed the heat generated by these exothermic reactions, are absent from smaller units. Maintaining process conditions is of critical importance, as increased temperature favors elective methane formation, deposition of catalyst-damaging carbon, and reduced chain length of products.

The Fischer-Tropsch synthesis that is the focus of this research, then, takes these advantages of smaller units one step further, as this process is highly sensitive to temperature, pressure, input gas stoichiometric ratios of carbon and hydrogen, catalyst type, and promoters. Existing reactors separate and recycle the output streams back into their own input stream to maximize conversion, but a network of smaller-scale reactors allows output streams to be refined in terms of these parameters and redirected to different units whose conditions are optimized for the products of choice. Of particular importance is the study and management of the secondary reactions that occur in a Fischer-Tropsch reactor, as recycled olefins have been demonstrated to be catalytically reabsorbed for further transformation and synthesis. Understanding the conditions under which this occurs and the effect of various operating conditions on selectivity of products informs a networking control strategy through which the advantages of an aggregate network might be realized; enhanced control of reactants allows more selective control of products.

The model to be developed and demonstrated has three primary objectives:

• Demonstrate its own viability through simulations that can be correlated with existing Fischer-Tropsch synthesis data in order to fine-tune the model's algorithms and assumptions to be maximally relevant to deployable technology.
• Identify and quantify cost drivers to existing large-scale optimization strategies and cost savings of smaller scale aggregated networks that take advantage of mass production, reactivity to technological change, and industrial learning curves.
• Perform analyses of a simulated aggregate network of small scale modules from a control standpoint, identifying and optimizing the extent to which process environments can be controlled and varied to maximize selectivity and efficiency of liquid hydrocarbon production.

Researchers

Tom Socci, Ph.D. Candidate, Earth and Environmental Engineering, tas2013@columbia.edu
Klaus Lackner, Ewing-Worzel Professor of Geophysics, klaus.lackner@columbia.edu