U. Minn. team proposes strategy for automated selection of optimal biomass-derived fuel blends and synthesis paths
|Proposed strategy for connecting automated network generation and optimization. Credit: ACS, Marvin et al. Click to enlarge.|
Researchers at the University of Minnesota are proposing a novel strategy that simultaneously identifies (a) the most desirable biomass-derived chemical products for an application of interest, such as fuels, and (b) the corresponding synthesis routes.
In a paper published in the ACS journal Energy & Fuels, they describe the strategy, and then apply it to identify potential renewable oxygenates and hydrocarbons obtained from heterogeneous catalysis of biomass that can be blended with gasoline to satisfy ASTM specifications.
The strategy for combined product and chemistry selection consists of i) constructing an exhaustive network of reactions consistent with an input set of chemistry rules and ii) using the network information to formulate and solve an optimization problem that yields an optimal product distribution and the sequence of reactions that synthesize them.
...the network of all possible reactions that can upgrade biomass to fuels is large and complex. Consequently, there are several potential biomass-derived compounds that can be blended with gasoline, and multiple chemical routes may exist to synthesize each one of them. This leads to two related problems: (a) the identification of optimal feasible biofuel-gasoline blends in terms of techno-economic objectives and (b) the selection of efficient synthesis routes to produce renewable blend components in terms of kinetics, thermochemistry, selectivity, etc.
...In this paper, we propose and apply a scalable strategy to identify optimal gasoline blends and synthesis routes to produce the renewable additives from biomass. The strategy involves combining rule-based network generation to construct an exhaustive network of reactions that upgrade biomass, with the formulation of a Mixed Integer Linear Programming (MILP) problem to identify the mix of renewable additives (to be blended with gasoline) and their synthesis routes from biomass that is optimal on the basis of thermochemistry, kinetics or economics. We specifically consider heterogeneous catalytic routes to convert biomass into oxygenates and hydrocarbons, while ensuring that the gasoline blends identified satisfy all ASTM specifications.—Marvin et al.
The researchers first used RING, a rule-based network generator developed in their group, to construct a reaction network pertaining to upgrading biomass-derived platform oxygenates. An optimization module then accepts the generated network, thermochemistry of the species and kinetics of reactions in the network, and physical properties of possible products and formulates a MILP with ASTM fuel specifications as constraints. The output of the optimization module is a list of alternative blends and synthesis routes that can then be explored manually.
A number of objective functions such as cost, environmental impact, and efficiency can be formulated to select for desirable chemical processes. In their paper, the researchers presented results for three objective functions (energy loss, catalyst requirement, and heat duty), but others (e.g., input cost, revenue, separability) could be constructed, if the required data were available, they noted.
Minimization of energy loss (ΔE) is a better indicator of investment cost than even the facility size for a data set of fuel and chemical manufacturing plants.
Minimization of catalyst requirement is a useful objective to reduce catalyst costs and avoid slow reactions, which may require large reactors or high residence times.
Minimization of absolute heat duty reduces the costs of heat transfer by favoring energetically neutral reactions. This could lead to reductions in operating (e.g., less fuel and cooling water) and capital (e.g., smaller heat exchangers) expenses for utilities.
The paper focused on the ΔE objective for two scenarios: oxygenates with and without hydrocarbons; results for the other objectives are in the paper’s Supporting Information.
Among their findings were:
For the given objective function and product constraints, both oxygenate and hydrocarbons can be components of optimal blends.
For stricter product constraints, the optimal solutions can differ. For example, reducing the limit on water adsorption leads to solutions that involve a greater fraction of heavier oxygenates or hydrocarbons.
Hydrocarbons do not appear in the optimal blends of the other objective functions (catalyst requirement and absolute heat duty). These objectives favor having fewer reactions and as such would avoid the increased processing to remove oxygen from the inputs.
The proposed strategy lends itself to several advantages. First, it allows for large reaction networks to be generated and considered for blend optimization in a fast, automated, and scalable manner. It should be noted that because compounds from the generated reaction network (thus synthetically feasible) are considered for blend optimization, the number of compounds evaluated is a fraction of that explored in the molecular design methods discussed earlier. Second, this strategy has a significant potential for “discovering” new reaction routes.
...Third, the strategy is generic and flexible enough to be applied to other combined product and chemistry selection problems involving different sets of chemistries, objectives, and physicochemical properties. For example, optimal diesel blends consisting of alkanes and oxygenates such as fatty esters and their synthesis routes from biomass can be identified subject to physical property constraints involving properties such as cetane index, density, cold flow properties, boiling point, etc.—Marvin et al.
W. Alex Marvin, Srinivas Rangarajan, and Prodromos Daoutidis (2013) Automated Generation and Optimal Selection of Biofuel-Gasoline Blends and Their Synthesis Routes. Energy & Fuels doi: 10.1021/ef4003318