NEMS remains the integrated modeling system the U.S. Energy Information Administration uses to develop the Annual Energy Outlook. Current NEMS-based cases examine how U.S. energy markets could evolve through 2050 under different assumptions—not what will inevitably happen.
The National Energy Modeling System (NEMS) is the U.S. Energy Information Administration’s integrated modeling system for analyzing and projecting U.S. energy markets. The analysis originally published on this page covered 1998 through 2020, but NEMS did not end with that forecast period.
Current NEMS-based Annual Energy Outlook cases extend through 2050 and reflect major changes in energy production, consumption, technology, infrastructure, and policy. The model continues to project energy production, imports, exports, conversion, consumption, and prices using economic, technological, resource, behavioral, and market assumptions.
NEMS evaluates interactions across the energy system. Changes in one sector can affect others through prices, technology costs, infrastructure constraints, consumer behavior, and resource availability. Its results are projections based on defined assumptions—not predictions.
The original analysis used a version of NEMS with a forecast horizon ending in 2020. It focused heavily on cellulosic ethanol and the effect that federal incentives and technological improvements might have on commercial production.
Since then, the U.S. energy landscape has changed substantially. Shale development expanded domestic oil and natural gas production, renewable generation grew, natural gas displaced much coal-fired generation, and electric vehicles and battery storage became more important. NEMS evolved as well.
EIA substantially modernized NEMS for the Annual Energy Outlook 2025, replacing much of its connecting Fortran and shell-script code with open-source Python packages. EIA also expanded the system’s treatment of hydrogen, carbon capture, oil and gas resources, and other emerging technologies.
NEMS represents fuel supply, energy conversion, and end-use consumption through interconnected modules. Its projections incorporate assumptions involving:
Changing these assumptions can produce substantially different outcomes. EIA therefore publishes multiple cases to examine uncertainty rather than identifying one case as a guaranteed or most likely future.
Liquid-fuels modeling remains an important part of NEMS, although it is broader than the petroleum-market representation described in the original article. Modern cases consider conventional gasoline, diesel, jet fuel, ethanol, biodiesel, renewable diesel, sustainable aviation fuel, and other alternative pathways.
Crude production, refinery capacity, product demand, international trade, renewable feedstocks, policy, and alternative-fuel supply remain interconnected. The modern model must represent a wider range of technologies and market constraints than the earlier 1998–2020 framework anticipated.
The original study examined ethanol made from agricultural residues, forest products, wood waste, crop residues, and dedicated energy crops. Its central relationships—feedstock availability, collection and transportation costs, conversion technology, capital requirements, and market demand—remain relevant.
Commercial liquid cellulosic ethanol, however, did not develop along the growth path envisioned in the original scenarios. Research and investment continued, but deployment faced challenges involving logistics, capital, conversion costs, reliable operation, and competition from established fuels and other renewable-fuel pathways.
This distinction matters: the Renewable Fuel Standard’s broader “cellulosic biofuel” category includes pathways other than liquid cellulosic ethanol, including qualifying fuels produced from biogas. Regulatory cellulosic-biofuel volumes should therefore not be treated as equivalent to commercial cellulosic-ethanol production.
The earlier analysis assumed that technological improvements would steadily reduce cellulosic-ethanol conversion costs. It examined multiple cases with different rates of cost reduction, deployment, and federal support. Those scenarios reflected the information available at the time.
Modern long-term analysis must consider a much wider technology set, including electric vehicles, batteries, renewable generation, hydrogen, carbon capture, advanced biofuels, refinery changes, and industrial processes. Small differences in capital costs, efficiency, infrastructure, fuel prices, and expected returns can produce significantly different results over a horizon extending to 2050.
New technologies rarely move directly from demonstration to widespread use. Commercial expansion depends on technical performance, financing, operating costs, infrastructure, regulation, competing technologies, demand, and expected investment returns.
Under favorable policy and technology assumptions, the original reference case projected U.S. cellulosic ethanol production reaching approximately 850 million gallons by 2020. Other cases examined how removing the assumed federal support would affect competitiveness.
These were scenario results, not predictions. Commercial-scale liquid cellulosic ethanol subsequently encountered limitations that kept production well below the anticipated trajectory. Meanwhile, shale production, renewable-electricity costs, batteries, and electric vehicles developed in ways that were difficult to anticipate at the beginning of the century.
Comparing those historical projections with actual market development illustrates both the value and the limitations of long-term energy modeling.
Current Annual Energy Outlook cases use NEMS to evaluate U.S. energy markets through 2050, including:
A longer horizon does not make the results more certain. It increases the importance of assumptions about technology, economics, behavior, resources, and policy. The value of the model is its consistent framework for exploring how different assumptions interact across the energy system.
NEMS was extended through 2050. To view the original article, click the button below.
NEMS continues to provide an integrated framework for examining energy supply, demand, technology, prices, infrastructure, and policy. The original cellulosic-ethanol analysis is best understood as a historical example—one that offers a useful comparison between long-range model assumptions and actual market development.