Many of the most important questions regarding energy use and energy’s role in the economy and the environment are long-term in nature. For example, understanding how energy use will change under cap-and-trade policy is a question with answers that depend strongly on the time frame in question. In response to the important questions about energy systems that are long-term in nature, researchers have developed some of the most complex models in the world to represent the current state and predict the future state of energy systems. In general, these models represent the interactions between energy technologies, economic conditions and principles, and environmental systems. And while these models often take a similar set of inputs and predict estimates for the same type of output (e.g. level of carbon-dioxide emissions), they vary significantly in how they represent components of the energy-economy-environment nexus and thus vary significantly in projections of the future. This post will take a look at several of the most important general classes of energy-economy-environment models, and in the next post, I will take a look at some of the most central controversies among modelers that lead them to reach different conclusions, even when starting from the same original datasets.

1. Bottom-Up Models

Bottom-up models of energy-economy-environment systems begin with the fundamental components of what is being modeled and build up components towards the entire system. Bottom-up models represent technologies in rich detail, incorporating the technical details that engineering studies provide. Typically, bottom-up models take prices of inputs and outputs as given. For example, a bottom-up model could tell you how much it would cost to generate 500 kWh from a natural gas combined cycle power plant if the model user inputs the price of raw materials to construct the plant, the price of labor, the price of natural gas, and the prices of any other inputs to the system. A bottom-up modeler has great control over the representation of technologies in the economy, and while the microeconomic foundations of bottom-up models can be quite detailed, the macroeconomic representation in these models is often quite simple (exceptions do exist). Bottom-up models are able to represent the entire economy by aggregating over many individual components. Environmental impacts also can be calculated by summing across all pollution sources. However, because bottom-up models represent reality by aggregating many individual components, they often become quite data-intensive and user-dependent. For example, many bottom-up models will represent each power plant in a region individually, therefore requiring large datasets of power plant operating characteristics.

Bottom-up models are typically used when a researcher wants to study the effects of very specific changes in technology or policy. Because of rich energy technology representation and detailed accounting of energy systems, bottom-up models can predict very specific impacts to externally-forced changes in conditions. However, because the macroeconomics of bottom-up models are typically less detailed, bottom-up models are typically thought of as weaker at making long-term projections (i.e. greater than 20-30 years in the future) than the alternative approach to modeling, top-down models.

2. Top-Down Models

Top-down models of energy-economy-environment systems begin with the entire system and break it down into more fundamental components. Compared to bottom-up models, top-down models sacrifice technological detail for more accurate macroeconomic representation. Rather than representing each individual power plant, industrial facility, and component of the remaining sectors of the economy, top-down models aggregate the energy-economy-environment system into representative “black boxes.” Each black box in a top-down model shares the overall important characteristics of the aggregation it represents, but do so without specifying the mechanics and fundamental units that are actually at work. At one extreme, the most simple top-down models aggregate the energy system into one modeling unit, and at the other extreme, the most computationally-complex top-down models can approach the technological detail of simpler bottom-up models. While bottom-up models explicitly model changes in technology deployment, top-down models use rates of substitution between inputs (e.g. the substitution rate between petroleum and biofuels in running a light duty vehicle) to model changes in energy systems.

Top-down models are particularly useful for researchers interested in long-term trends and broad shifts in energy use. Because of the higher level of aggregation and black-box representation of aggregated sectors, top-down models typically can’t be used to predict changes relevant for individual energy producers and consumers. Whereas a power plant operator would typically use a bottom-up model to make business forecasts for the short to medium run, a policymaker setting energy or environmental policy for the next fifty years would typically use a top-down model.

3. Hybrid Approaches

This post gave a simplified depiction of the differences between top-down and bottom-up energy-economy-environment models. This debate is by no mean constrained to these types of models. And like other types of models, hybrid energy-economy-environment models are now being developed with both bottom-up and top-down properties. In some approaches, bottom-up style insights from engineering studies are incorporated into the characteristics of representative black boxes in top-down models. For example, an engineering study of a single particular coal power plant retrofitted with carbon capture could be used to extract parameters of cost increases for the inputs into the power plant. These parameters could then be used to update the representation of the aggregated black box for all coal plants that retrofit for carbon capture. In other hybrid approaches, bottom-up models are forced to meet economic constraints supplied by top-down models. For example, a bottom-up model may be forced into a constraint for aggregated GDP growth where the GDP growth parameter is calculated in a top-down model.

The hybrid approach represents one important area of focus that modelers have paid significant attention to. However, many disagreements over how to best represent energy-economy-environment systems still persist. In my next post, I will take a look at some of the most persistent debates and present the schools of thought on both sides of these debates.