IEEE Power & Energy Magazine - May/June 2018 - 26
By Hamed Mohsenian-Rad,
Emma Stewart, and Ed Cortez
In the evolutIon of advanced sensIng technologies, transmission systems have led distribution. the
visibility and diagnostics of the transmission grid have
been transformed over the past decade with the systematic
deployment of phasor measurement units (PMus). similar and even more advanced new information sources are
now becoming available at the distribution grid, using distribution-level PMus, also called micro-PMUs (µPMUs).
µPMus provide voltage and current measurements at higher
resolution and precision to facilitate a level of visibility
into the distribution grid that is currently not achievable.
however, mere data availability in itself will not lead to
enhanced situational awareness and operational intelligence. data must be paired with useful analytics to translate these data to actionable information. In this article, we
explore some of the opportunities to leverage µPMu data,
combined with data-driven analytics, to help electrical distribution system planners and operators to get out in front
of problems as they evolve.
the data generated by µPMus are a prominent example of big data in power systems. each µPMu generates 124,416,600 readings per day. therefore, µPMus
installed on a handful of utility distribution feeders can
generate terabytes of data on daily basis. Because µPMus
Digital Object Identifier 10.1109/MPE.2018.2790818
Date of publication: 18 April 2018
ieee power & energy magazine
stream their measurements continuously, the data must be
collected, cleansed, and processed, all in real time.
the collected µPMu data must then be dissected into
descriptive, predictive, and prescriptive analytics. While
descriptive analytics focuses on what happened in the past,
predictive analytics aims at what may happen in the future.
Both are stepping stones toward prescriptive analytics-
optimizing the future with informed decisions. here, we
consider case studies in both descriptive and predictive
analytics and provide a sampling of the benefits derived
from µPMu data.