The biggest roadblock to the energy transition is neither technology nor finance, nor is it a lack of ambition; it is uncertainty. We are operating on limited information and trying to make an educated guess for a better, cleaner future. The transition requires answers to essential questions, such as when electric vehicles and heat pumps will be scaled up? How much energy will data centers require? When will we reach net-zero emissions?

These are not technical or financial questions; instead, they are questions about the future, and the most crucial tool we use for answering them is forecasting and scenario modeling.

The era of flat demand growth in the U.S. is coming to an end. We are transitioning to a period of unforeseen dynamism in the energy sector. Given our history of flat demand, incremental technology changes and predictable costs, traditional forecasting tools were a sufficient indicator of the future.

The energy system is no longer the slow, predictable, centralized machine that it used to be. It is now an intricate, data-driven web where decisions adapt in real time to shifting weather, markets and human behavior. Naturally, this dynamism breeds considerable volatility that we cannot accurately predict but can effectively explore.

For far too long, we have relied on moving averages and exponential smoothing to forecast the perfect number. There is no problem with that, except that it is not ideal. As Joseph DeCarolis, former U.S. Energy Information Administration administrator, said at MIT Energy Initiative Fall Colloquium, “Whatever you do, don’t start believing the numbers.” No one knows, and no one can predict the future exactly, but we can explore it in a structured way. The energy transition demands a different kind of modeling. One that moves us from forecasts to foresight.

It doesn’t predict the future; it reveals what might happen and how we can respond. Scenario models are frameworks that allow us to test today’s decisions and their future outcomes, uncovering blind spots and biases, forcing us to consider alternative possibilities. They are upgrades to traditional models, helping us ask: what if the world doesn’t evolve as planned? As research by scholars Leonard Göke, Jens Weibezahn and Christian von Hirschhausen suggests, scenarios are collective blueprints and not crystal balls. They have the power to align governments, organizations and citizens on emerging futures.

When I build various power market scenarios, I have seen how tweaking a few key assumptions, such as electrification growth rates, capital costs, fuel prices, etc., can significantly transform strategic decisions. Scenario modeling defines acceptable ranges and confidence bands. They help us identify dealbreakers in investments. For example, imagine a wind power plant operating at 100%, 50% and 40% of its generation capacity. Running these cases through the model reveals how revenue changes under different operating conditions. Suppose the plant is profitable at 100% and 50%, but not at 40%. We can then define 40% capacity as the critical threshold for profitability. That’s the subtle yet powerful role of foresight. It transforms uncertainty into informed agency.

With billion-dollar investments that affect all lives, understanding their gravity and impact is essential. Simple stress tests in different conditions can help identify what builds or breaks the investment. Governments use long-term scenarios to quantify some of the most critical macroeconomic trends and policy challenges facing the global economy. Utilities rely on scenario planning for rate cases, testing reliability and anticipating flexibility. Similarly, investors have described scenario-planning as vital in investment management, to help explore options and inform decision-making.

Scenario models help plan energy systems more effectively. They help planners simulate futures varying in demand, weather, policies and technologies, to identify optimal load zones for the interconnection of renewable projects, such as solar, wind and storage, thereby avoiding curtailment and transmission constraints. Using these models enables operators, traders and investors to test various operational strategies. For instance, what happens if a heatwave drives up demand and a major transmission outage occurs, or if wind generation is 10% less than expected? Is the investment still profitable?

By using these “what-if” scenarios through advanced power market and network models like production cost models, operators can effectively dispatch or curtail supply and demand, reduce costs and maintain reliability. Similarly, capacity expansion models become practical decision-making tools for policymakers, who can test their proposed policies against different technological mixes, investments and operational scenarios, and assess their impact on household energy costs. Thus, these models hold the power to bridge the gap between system design and household impact.

As expected, scenario modeling has its limitations too. Not all the predictions the model makes will be perfect. I have seen the models give counter-intuitive results that aren’t neat numbers. But that does not mean they aren’t helpful. Instead, they force us to plan for alternative actions to take when projections diverge from expectations.

If assumptions are sound and transparent, results make sense. That visibility is critical and holds the power to align stakeholders. The inputs and assumptions, if discussed and debated by the stakeholders first and then finalized, have a higher chance of bringing together multiple voices and shaping the future. Every scenario is an opportunity to develop a shared framework. When all facets of a community use the same framework, they build clarity and a better understanding of uncertainty. This foresight enables the community to take ownership of its future. It empowers them to collectively shape a mutually beneficial, just and equitable future.

The clean energy transition is unfolding as a complex and messy challenge. Each step forward reveals new challenges. However, the way forward is not chasing correct answers but asking better questions. Scenario modeling and forecasting provide us with a structure and framework to ask those questions. If done well, it gives us the ability to answer multiple “what-ifs” and explore many futures before committing to one. Hence, it is a crucial tool in this journey. Just like life, to get it right, we must get out of autopilot mode and focus on collective, proactive action towards shaping a clean, fair and equitable transition.