Wood Pellet Markets as Feedstock Infrastructure for Renewable Diesel: Supply Dynamics, Price Formation, and Clean-Fuel Policy Implications
Graduate Student Name
Chukwuemeka Valentine Okolo
Email Address
Faculty mentor/Supervisor
Dr. Andres Susaeta
Email Address (Faculty mentor/Supervisor)
Department Affiliation
Forest Engineering Resources & Management
Job Location
Peavey Science Center
Description of project or research opportunity
Renewable diesel (RD) is rapidly becoming the compliance fuel of choice under California's Low Carbon Fuel Standard (LCFS) and Oregon's Clean Fuels Program, yet the economics of producing RD from woody biomass feedstocks remain poorly understood, particularly with respect to feedstock cost risk and supply volatility. This project fills that gap by conducting the first systematic empirical analysis of the U.S. densified biomass (wood pellet) sector as a feedstock platform for RD production.
The core question driving this work is how national pellet supply conditions, export competition from European biomass energy markets, and inventory dynamics translate into feedstock costs and supply risk for RD facilities in the Pacific states. Despite its direct relevance to project finance decisions and state-level decarbonization targets in Oregon and California, this connection has received almost no systematic empirical attention.
The project is built on the U.S. Energy Information Administration's Monthly Densified Biomass Fuel Report, which is a mandatory census survey of approximately 90 U.S. pellet manufacturing facilities that has been collected since January 2016 but remains largely unexploited by academic researchers. The dataset includes monthly figures on production volumes, feedstock costs by input type, end-of-month inventories, and domestic and export sales prices, broken out across three U.S. census regions covering the period 2016 through 2025. As of late 2025, these facilities together account for more than 13 million tons of annual production capacity, a scale that makes this market directly relevant to Pacific Northwest forest and energy policy.
Methodological Approach
The analysis uses three methods that build on each other:
1. Non-parametric cost-supply curve construction. Historical price-quantity observations from EIA-63C are sorted to map out how much biomass the market has supplied at each price level and how that envelope has shifted over time as export demand expanded.
2. Panel fixed-effects econometric modeling. Region-month panel regressions estimate the price elasticity of pellet supply and quantify the role of inventory conditions in short-run price formation, the core parameters for assessing feedstock cost risk in a prospective RD project.
3. Techno-economic pass-through analysis. The feedstock price distribution from the data is then linked to the minimum fuel selling prices for RD, using process parameters from Aspen Plus simulations.
This work draws on agricultural and resource economics, energy economics, environmental policy, and chemical engineering. The findings speak to three active policy debates: the design of biomass additionality conditions under LCFS, the role of long-term supply contracts and contracts-for-difference in aligning commodity markets with clean-fuel objectives, and the feedstock market conditions under which Pacific Northwest wood fiber can credibly anchor an RD supply chain.
The core question driving this work is how national pellet supply conditions, export competition from European biomass energy markets, and inventory dynamics translate into feedstock costs and supply risk for RD facilities in the Pacific states. Despite its direct relevance to project finance decisions and state-level decarbonization targets in Oregon and California, this connection has received almost no systematic empirical attention.
The project is built on the U.S. Energy Information Administration's Monthly Densified Biomass Fuel Report, which is a mandatory census survey of approximately 90 U.S. pellet manufacturing facilities that has been collected since January 2016 but remains largely unexploited by academic researchers. The dataset includes monthly figures on production volumes, feedstock costs by input type, end-of-month inventories, and domestic and export sales prices, broken out across three U.S. census regions covering the period 2016 through 2025. As of late 2025, these facilities together account for more than 13 million tons of annual production capacity, a scale that makes this market directly relevant to Pacific Northwest forest and energy policy.
Methodological Approach
The analysis uses three methods that build on each other:
1. Non-parametric cost-supply curve construction. Historical price-quantity observations from EIA-63C are sorted to map out how much biomass the market has supplied at each price level and how that envelope has shifted over time as export demand expanded.
2. Panel fixed-effects econometric modeling. Region-month panel regressions estimate the price elasticity of pellet supply and quantify the role of inventory conditions in short-run price formation, the core parameters for assessing feedstock cost risk in a prospective RD project.
3. Techno-economic pass-through analysis. The feedstock price distribution from the data is then linked to the minimum fuel selling prices for RD, using process parameters from Aspen Plus simulations.
This work draws on agricultural and resource economics, energy economics, environmental policy, and chemical engineering. The findings speak to three active policy debates: the design of biomass additionality conditions under LCFS, the role of long-term supply contracts and contracts-for-difference in aligning commodity markets with clean-fuel objectives, and the feedstock market conditions under which Pacific Northwest wood fiber can credibly anchor an RD supply chain.
Tasks student will perform
Task 1, Data Acquisition and Panel Construction (~25 hours)
The student will download, clean, and integrate the EIA-63C monthly survey files spanning 2016 through 2025. This involves reconciling inconsistencies in variable naming, unit definitions, and regional coding that arise across ten years of survey administration, a necessary step that is analytically critical but labor-intensive. The student will supplement the EIA-63C panel with two additional public sources. The final product will be a single, analysis-ready panel dataset in Stata or R, fully documented with a cleaning log and data dictionary. This task builds practical skills in working with government survey data, assembling multi-source longitudinal datasets, and writing clean, well-documented code, all of which transfer directly to graduate school, federal agency work, or private-sector analyst roles.
Task 2, Descriptive Analysis and Visualization (~25 hours)
Using the cleaned panel, the student will produce a set of time-series and cross-sectional visualizations in ggplot2 (R) or Stata graphics, including regional production and inventory trends, domestic versus export price divergence over time, export volume dynamics and the 2021–2022 surge episode, and feedstock cost composition by input category. The student will calculate summary statistics, identify and document structural breaks or market anomalies, and draft a descriptive results section with publication-quality figures suitable for inclusion in the dissertation chapter and any resulting journal submission. This task develops skills in exploratory data analysis, economic data visualization, and scientific writing, including the discipline of presenting quantitative findings clearly for a non-specialist policy audience.
Task 3, Supply Curve Construction and Regression Support (~30 hours)
The student will sort price-quantity observations to construct empirical cost-supply envelopes by region and nationally, the non-parametric supply characterization that is the project's first analytical deliverable. They will then assist in coding and running the baseline panel fixed-effects regression models, organize regression output tables, and conduct pre-specified robustness checks, including alternative lag structures and sample period restrictions. Robustness checking is a critical but often resource-constrained activity in dissertation research; having dedicated support here directly improves the credibility and rigor of the final results. Throughout the project, the student will join weekly research meetings, contribute to a shared literature review on biomass markets and renewable fuels policy, and work within a GitHub-based reproducible research setup if possible.
The student will download, clean, and integrate the EIA-63C monthly survey files spanning 2016 through 2025. This involves reconciling inconsistencies in variable naming, unit definitions, and regional coding that arise across ten years of survey administration, a necessary step that is analytically critical but labor-intensive. The student will supplement the EIA-63C panel with two additional public sources. The final product will be a single, analysis-ready panel dataset in Stata or R, fully documented with a cleaning log and data dictionary. This task builds practical skills in working with government survey data, assembling multi-source longitudinal datasets, and writing clean, well-documented code, all of which transfer directly to graduate school, federal agency work, or private-sector analyst roles.
Task 2, Descriptive Analysis and Visualization (~25 hours)
Using the cleaned panel, the student will produce a set of time-series and cross-sectional visualizations in ggplot2 (R) or Stata graphics, including regional production and inventory trends, domestic versus export price divergence over time, export volume dynamics and the 2021–2022 surge episode, and feedstock cost composition by input category. The student will calculate summary statistics, identify and document structural breaks or market anomalies, and draft a descriptive results section with publication-quality figures suitable for inclusion in the dissertation chapter and any resulting journal submission. This task develops skills in exploratory data analysis, economic data visualization, and scientific writing, including the discipline of presenting quantitative findings clearly for a non-specialist policy audience.
Task 3, Supply Curve Construction and Regression Support (~30 hours)
The student will sort price-quantity observations to construct empirical cost-supply envelopes by region and nationally, the non-parametric supply characterization that is the project's first analytical deliverable. They will then assist in coding and running the baseline panel fixed-effects regression models, organize regression output tables, and conduct pre-specified robustness checks, including alternative lag structures and sample period restrictions. Robustness checking is a critical but often resource-constrained activity in dissertation research; having dedicated support here directly improves the credibility and rigor of the final results. Throughout the project, the student will join weekly research meetings, contribute to a shared literature review on biomass markets and renewable fuels policy, and work within a GitHub-based reproducible research setup if possible.
Special skills required
Required
• Statistical software proficiency: Ability to import, clean, and summarize a real-world dataset independently in Stata, R, or Python. The student does not need to know all three; depth in one is sufficient, and training will be provided in project-specific workflows.
• Comfort with tabular data formats: Ability to work with CSV and Excel files, understand variable types, and recognize common data quality issues such as missing values, duplicate records, and inconsistent coding.
• Attention to detail: The data assembly task involves merging multi-year government survey files where small merging or recoding errors propagate through all downstream analyses. Careful, documented work is essential.
• Reliability and communication: Ability to meet weekly deadlines, flag problems proactively, and ask questions when data issues are unclear. Research quality depends on honest, timely communication.
Preferred (not required)
• Prior coursework or interest in energy policy, environmental economics, forestry, natural resource economics, or bioenergy. Sufficient context will be provided during onboarding.
• Familiarity with ggplot2 or Stata graphics for producing publication-quality figures.
• Any prior exposure to panel data or time-series analysis, even at an introductory level.
• Statistical software proficiency: Ability to import, clean, and summarize a real-world dataset independently in Stata, R, or Python. The student does not need to know all three; depth in one is sufficient, and training will be provided in project-specific workflows.
• Comfort with tabular data formats: Ability to work with CSV and Excel files, understand variable types, and recognize common data quality issues such as missing values, duplicate records, and inconsistent coding.
• Attention to detail: The data assembly task involves merging multi-year government survey files where small merging or recoding errors propagate through all downstream analyses. Careful, documented work is essential.
• Reliability and communication: Ability to meet weekly deadlines, flag problems proactively, and ask questions when data issues are unclear. Research quality depends on honest, timely communication.
Preferred (not required)
• Prior coursework or interest in energy policy, environmental economics, forestry, natural resource economics, or bioenergy. Sufficient context will be provided during onboarding.
• Familiarity with ggplot2 or Stata graphics for producing publication-quality figures.
• Any prior exposure to panel data or time-series analysis, even at an introductory level.
Proposed dates of employment
-