Join on Zoom Stream on YouTube
Heather Liddell
Assistant Professor of Mechanical Engineering, and Sustainability Engineering & Environmental Engineering, Harrington Fellow
Speaker Biography:
Heather Liddell is an Assistant Professor at Purdue University with joint appointments in Mechanical Engineering and Sustainability Engineering & Environmental Engineering. She studies energy and environmental impacts in manufacturing, integrating economy-level and technology-level modeling approaches. Before joining Purdue in 2023, Dr. Liddell was a group leader at the energy consulting firm Energetics, where she worked closely with the U.S. Department of Energy to develop strategies for resource-efficient and decarbonized manufacturing in the United States. She holds BS and MS degrees in mechanical engineering and a PhD in materials science from the University of Rochester and completed postdoctoral training at the U.S. Naval Research Laboratory. As a UT Harrington Fellow for the 2026-2027 academic year, she is collaborating with colleagues at the Energy Institute to develop new energy systems modeling approaches for a rapidly evolving industrial economy.
Abstract:
In Leonard Read’s famous 1958 essay “I, Pencil,” the eponymous narrator insists, rather gleefully, that no single person on Earth knows how to make a wooden pencil. Manufacturing know-how is scattered across complex global supply chains, with each entity contributing only a tiny piece. The story captures a marvel of economics. But for industrial system modelers, this marvel is simultaneously a conundrum. Individual firms may know what they make and in what quantities—but because U.S. statistical agencies do not comprehensively compile that information, this knowledge is similarly scattered. In this talk, I will show how sparse and decentralized data on physical production can limit our ability to build accurate mass-balanced models of the national production economy. Energy intensity (energy per unit of output) underpins how we track industrial efficiency and evaluate policy, but without physical output data these metrics must be assessed with a monetary-unit denominator. A monetary basis blurs energy efficiency gains with market price volatility and inflation. I will describe our group’s effort to build physically and environmentally extended input-output (IO) models of U.S. goods-producing industries, including approaches we have developed for operating under severe data sparsity and uncertainty. Lastly, I will explore how industrial energy use has shifted over time, and what that means for model obsolescence.