CLIENT

CHALLENGE

The Northwest Power and Conservation Council needed to completely redevelop a planning model used for electric resource capacity planning, since their legacy tool was a loose connection of several model platforms (e.g., .Excel, Visual Basic, etc.) and was not user friendly or transparent. They sought a platform and a tool that was robust, flexible, and transparent, and that could readily deal with uncertainty and risk, while optimizing their supply and demand resource plans.

approach

Staff at Lumidyne (Cory Welch), while employed at Navigant Consulting, selected the Analytica® software platform as the ideal solution to meet the needs of the needs of the Council. The model integrated multiple distinct elements (previously in 4 separate software platforms) into a single decision analysis tool that Council staff could operate themselves. The model employs Monte Carlo analysis, linear programming, and nonlinear optimization into a System Dynamics behavioral model. Additionally, the model takes advantage of distributed computing to reduce model run time for this computationally intensive analysis. The model runs stochastically and optimizes the resource capacity plan, considering uncertainty and risk over a 20-year time horizon, including both supply and demand side resources (e.g., conservation, demand response) in its analysis.

RESULTS

The Power Council received a model that exceeded expectations for the project and was delivered on time under a very tight schedule. They used the model to develop the 7th Power Plan for the entire Pacific NW, and gained stakeholder buy-in with comparative ease, due in part to the transparency and accessibility of the model through an online interface. They continue to use the model for regional planning purposes, with model enhancements delivered as desired to meet evolving regional needs. The model is accessible for public viewing and operation through a cloud player (click link below) requiring no modeling software installation.

run online model

GRAPHICAL USER INTERFACE

The transparency of the online user interface facilitated stakeholder buy-in.

CAUSAL LOOP DIAGRAM

The model employs time-tested System Dynamics methods, including numerous stocks and feedback loops, delays, and behavioral expectation trending algorithms (Image adapted from online model).

OPTIMAL BUILD STRATEGY

The model employs nonlinear optimization using a genetic algorithm to help Council staff find optimal build strategies.

EFFICIENT RISK FRONTIER

The model accepts key parameters input as probability distributions, rather than point estimates. The model runs using Monte Carlo simulation, permitting incorporation of portfolio risk metrics (e.g., CVaR) into the optimization objective function.

INTEGRATED RESOURCE PLANNING