The ORFEUS team is made up of four Principal Investigators (PIs), two research software engineers (RSEs), and two postdocs across three research groups. The lead group at Princeton consists of Prof. Rene Carmona (PI), Prof. Ronnie Sircar (PI), Michal Grzadkowski (RSE), Alice Fang (RSE), and Xinshuo Yang (postdoc). The rest of the team members include Prof. Mike Ludkovski (PI) and Guillermo Terrén-Serrano (postdoc) from the University of California at Santa Barbara and Dr. Glen Swindle (PI) from Scoville Risk Partners.
ORFEUS Data Products
An overview of the ORFEUS project can be found at our website (orfeus-dev.princeton.edu). As part of this project, we have developed two simulation platforms — PGscen and CLNSim — that support the joint simulation of renewable asset production and loads conditioned on short-term forecasts.
PGscen is open source. It can be found at github.com/PrincetonUniversity/Pgscen for producing joint Monte Carlo scenarios composed of power load demand, as well as solar and wind generation for a given grid. The PGscen model is trained on historical forecasted and realized demands and outputs. CLNSim, is also calibrated to historical forecasts and realizations of wind, solar and hydro assets, as well as loads. The platform supports live listener-based simulation requests, automatically returning ad hoc user requests within a few minutes.
ORFEUS Software Products
We have implemented a parallelized computational pipeline to simulate over these scenarios using Vatic, our grid simulation engine (github.com/PrincetonUniversity/Vatic). Vatic was adapted from Prescient, a platform developed by a team at Lawrence Livermore National Laboratory and includes several new features as well as technical improvements that facilitate running hundreds of thousands of grid simulations in parallel on a compute cluster.
Beyond the data products described above, the team maintains a website https://orfeus-dev.princeton.edu/ containing a presentation of the members of the team and of the original research objectives, together with scenario samples which can be produced on demand. Also features on the website are plots of some of the risk indices produced by the team’s original algorithms as well as the LMP resulting from the ensuing risk adjustments.