WHAT WE DO

We model complex systems for policy analysis, market forecasting, and planning using System Dynamics – a time-tested methodology invented at MIT that accounts for positive and negative feedback loops, stocks and flows, inertia from delays, and agent behavior.

WHY WE DO IT

We don't live in a linear world. Good models for policy and decision-making recognize the dynamic complexity of the systems in which we operate and incorporate the structure that truly drives system behavior.

HOW WE DO IT

We leverage the Analytica modeling platform for all of our System Dynamics models, due to its transparency, flexibility, extensibility, structured optimization, and integrated Monte Carlo analysis. This platform provides us with far more capability than is present in classical System Dynamics software platforms (Welch 2017).

Take a peek at a few examples of how staff at Lumidyne have applied System Dynamics to help our clients solve pressing needs.

BASS DIFFUSION OF DISTRIBUTED ENERGY RESOURCES

Lumidyne staff have simulated the adoption of distributed energy resources (e.g., solar PV, combined PV & storage, and advanced technology vehicles) for 15 years using well-established methodologies, including System Dynamics, Bass diffusion, consumer choice theory, optimization, and discounted cash flow analysis.

FORECASTING ADOPTION OF ENERGY EFFICIENCY TECHNOLOGIES

Lumidyne staff (Cory Welch) originally developed a model to simulate the adoption of energy efficiency technologies by integrating Bass diffusion with stock turnover dynamics while employed at the National Renewable Energy Laboratory (NREL). The model formed the basis for NREL's Stochastic Energy Deployment Systems (SEDS) model.

REGIONAL PORTFOLIO MODEL (RPM)- A STOCHASTIC, DYNAMIC, NONLINEAR OPTIMIZATION MODEL

Lumidyne staff (Cory Welch) is providing ongoing development and maintenance support for the Regional Portfolio Model (RPM) for the Northwest Power and Conservation Council. The model employs many classic System Dynamics methods, including multiple stocks and flows, feedback loops, material and information delays, behavioral expectation trending algorithms, and co-flows. It is highly transparent for stakeholder review, and can be run using an online interface not requiring any software installation.

online model