SEED develops coupled, multiscale models of energy and industrial systems. Below are the methods we build and the areas we apply them to. Each is shown with a short description and an example of the kind of problem it addresses.
How we work: methods
Multiscale modeling
We represent a system at several scales at once, from the unit process and the plant to the supply chain, the economy, and policy, within one framework. A change at one scale and its effects at the others are then evaluated together, rather than in separate, disconnected studies.
Example problem: linking a chemical process design to its supply chain and to regional economic response in a single optimization, so the design is chosen with its downstream consequences visible.
Process design, control, and optimization
We design and operate unit processes and full flowsheets using process simulation, dynamic modeling and control, and mathematical optimization, including mixed-integer nonlinear and superstructure formulations. This also produces physically consistent performance data that feed the larger-scale models.
Example problem: representing competing reactor and separation configurations for a biorefinery as a superstructure and optimizing it against cost and emissions.
Life-cycle and techno-economic assessment inside the design
We formulate life-cycle assessment and techno-economic analysis inside the design and optimization, using process-based, hybrid, prospective, and consequential methods. Environmental and economic performance then shape the design directly, instead of only grading it after the fact.
Example problem: computing life-cycle inventories and impacts inside a mathematical program, so that the least-cost design also satisfies an emissions constraint.
Absolute sustainability and techno-ecological synergy
We assess systems against absolute limits, such as planetary boundaries and the capacity of local ecosystems, not only against relative improvement. Techno-ecological synergy treats ecosystems and the people affected as part of the design, as variables and constraints rather than as an afterthought.
Example problem: designing a biofuel supply chain together with treatment wetlands, so that nutrient runoff to a lake stays within the watershed’s ecological capacity.
Industrial ecology and material-flow analysis
We track how materials and energy move through supply and value chains over time, using dynamic material-flow analysis and industrial-ecology methods. This supports circular-economy design and the recovery of critical materials.
Example problem: tracking the stocks and flows of wind-turbine blades over time to compare recycling, repurposing, and landfill pathways.
Capacity expansion and climate-aware decision-making
We couple energy-system capacity-expansion and integrated-assessment models, such as ReEDS, GCAM, and IMAGE, with our design models, and add input-output and general-equilibrium economics. This makes assessments climate-aware and forward-looking, reflecting how the grid, technologies, and economy evolve.
Example problem: deriving time-evolving grid emission factors from a capacity-expansion pathway, so a technology’s carbon footprint reflects the grid it will actually run on.
AI, data, and optimization under uncertainty
We use machine learning to make these models faster and more scalable: large language models and retrieval to assemble life-cycle inventories, and surrogate and physics-informed models to approximate expensive simulations. We also design under uncertainty, using stochastic and robust optimization.
Example problem: assembling life-cycle inventories automatically from documents and databases, and optimizing a design so it performs well across many future scenarios rather than a single forecast.
What we work on: application areas
Low-carbon fuels and the bioeconomy
Sustainable aviation fuel, hydrogen, and biorefinery supply chains, from feedstock to end use. We study which technologies, sites, scales, and policies make low-carbon fuels viable.
Example problem: choosing where to site, how large to build, and which conversion pathway to use for a sustainable aviation fuel supply chain under uncertain incentives.
Advanced nuclear and grid decarbonization
Small modular reactors and the decarbonization of the electricity grid, including how the generation mix evolves over time.
Example problem: comparing the life-cycle carbon of small modular reactors against alternatives on a grid whose mix is changing across the coming decades.
Circular economy and critical materials
End-of-life and recovery pathways for wind blades, photovoltaics, batteries, and plastics, and the supply of critical minerals.
Example problem: evaluating end-of-life options for wind-turbine blades, or the recovery of critical minerals from retired photovoltaics and batteries.
Food, energy, and water systems
The connections among agriculture, energy, and water, including agrivoltaics, agricultural electrification, and biomass supply.
Example problem: assessing agrivoltaics, solar generation over farmland, for its combined effects on energy output, crop yield, and water use.
Process and manufacturing industries
The footprint and redesign of industrial processes, from chemicals and pulp and paper to food processing, aluminum, and semiconductors.
Example problem: identifying where process integration or electrification most reduces the footprint of an industrial plant, and at what cost.
Work with us
SEED is recruiting PhD and MS students, and welcomes undergraduate researchers, with interests in process systems engineering, optimization, industrial ecology, and AI.