“As utilities shift toward integrating increasing amounts of DERs into their systems, they will be relying upon these resources to complement energy procurements from the wholesale market. The nature of these DERs and associated properties with respect to intermittency and various levels of reliability, however, need to be integrated into the planning process. Therefore, the Guidance Proposal recommends that the utilities identify a process to move from deterministic to a probabilistic modeling approach for distribution system planning.”
 State of New York Public Service Commission Case 14-M-0101 -Proceeding on Motion of the Commission in Regard to Reforming the Energy Vision. Order Adopting Distributed System Implementation Plan Guidance. Issued and Effective: April 20, 2016
Why Model Grid Infrastructure Probabilistically?
Given the small spatial scale of DER, the temporal variability, and incorporation of human preferences, optimization modeling of DERs is problematic. To cope with the number of variables and new types of uncertainty that are presented by a distributed future, it has been well documented that a probabilistic approach to modeling more distributed energy futures is desirable. An agent based model (ABM) of a probabilistic grid, one that is realistic but not real, can help integrate feedback between DER adoption and use futures with physical investments in the grid.
Procedural and Pattern oriented modeling (POM) can help improve to move beyond fixed asset or market optimization towards probabilistic infrastructure investments. Cities are characterized by self-similarity, or fractal scaling. This means that a subsection of a city will demonstrate the same overall properties as larger sections and other sections. The use of the self-similar properties of cities can help, by enabling methods to “grow” realistic , but not real, energy grids based on GIS inputs. For a complete description of the model and documentation see: https://www.comses.net/codebases/6006/releases/1.0.0/
Findings: Focus on Differences in Demand not Supply
More on the model results will are forethcoming