Simulation and Optimization

Economic relationships are characterized using many different types of mathematical functions. Input/output functions govern the local economies, optimization characterizes power systems, behavioral models represent the actions of households and individuals. Veritas uses, builds and maintains an extensive library of computer models that represent economic relationships. For new projects we are able to draw upon and integrate these allowing cost-effective simulation modeling.

Power System Models

Our suite of power system models (EPSM) is supply-focused and includes minimum cost based hourly dispatch optimization and long run market clearing simulations models for dozens of regions. EPSM has been applied in regulatory impact analyses and dozens of peer reviewed regulatory submissions. EPSM is populated with data from two main data sources: the Continuous Emission Monitoring Systems (CEMS) and the Emissions & Generation Resource Integrated Database (eGRID).  CEMS provides hourly readings of emissions and generation, eGRID provides annual data on power plant generation and emissions.

EPSM uses the most recent unit and generator year data. The Unit dataset provides unit descriptors, the unit’s operational status, the primary fuel type, annual readings of heat input in MMBtus, annual NOx emissions in tons, annual SO2 emissions in tons, and CO2 emissions in tons. The Generator dataset provides the same descriptor variables, as well as the generator nameplate capacity in megawatts, generator capacity factor, and generator annual net generation in megawatt hours. For EPSM’s purposes, the Unit and Generator data sets are merged to provide one set of data for each unit that describes the units fuel type, heat input, nameplate capacity, capacity factor, and annual net generation.  These unit data are combined to represent regions and hourly load from each region is used to dispatch units.

Energy Penalty Simulation Models

Engineering and economic issues intersect at the regulation of power plant cooling water intake structures. A particularly important effect is the “energy penalty” that accompanies closed cycle cooling conversions. Understanding energy penalty effects at a detailed level is important because energy penalty effects vary hourly and tend to be at their highest when atmospheric conditions are already contributing to high air-conditioning loads, generation costs, and wholesale electricity prices. Veritas has produced dozens of simulation models that take in hourly measurements of ambient conditions to produce hourly estimates of generating unit efficiency. These are connected to hourly dispatch models to simulate effects in the context of power markets.

Recreation Site Choice Simulation Models

Recreation site choice models consider recreator locations, along with site characteristics such trails, bathrooms, fish catch rates, launches and other amenities to understand how people choose among recreation opportunities. Veritas conducts survey research and econometric modeling to identify specific recreation preference functions. We also conduct sophisticated functional transfers that link preference functions to different regions. In these transfer applications site choice simulation models are created by fusing a transferred function to the geography, population densities and site characteristics. Veritas has developed scores of such models in regions throughout the United States.

Electric Car Dynamic Adoption Model

Veritas developed the personal transportation/electric vehicle penetration module of the Electric Power Research Institute’s (EPRI) REGEN macroeconomic model. The Transportation Module consists of an integrated set of equations and data that connect to the REGEN macroeconomic model.  These equations and data are brought together to characterize vehicle purchases, driving behaviors, fuel consumption and emissions.  The structural economic model can be calibrated to generate a realistic and internally consistent representation of baseline personal transportation conditions into the far future (decades).  The module is partial equilibrium; counterfactual experiments can be conducted by specifying conditions that are different from baseline and running the model.  Electric vehicle adoption is modeled using econometric equations in a system dynamic framework.

Dynamic Pharmaceutical Choice Model with Bayesian Updating

A dynamic model of adoption rates for new pharmaceutical products. The model incorporates preferences for health outcomes into initial choices. Subsequent to realizing outcomes patients use Bayes Rule to update expectations and choice preference. The model is made dynamic using a Markov approach. (Bingham, Miller, and Johnson 2001).