Wrapping up the latest update to the Pike Research Microgrid Deployment Tracker, which was published earlier this month, I had a very simple insight: What is and what is not a microgrid is really in the eye of the beholder.
The U.S. Department of Energy (DOE), guided by the perspective of academics from the University of Wisconsin and big thinkers at DOE’s Lawrence Berkeley National Laboratory, came up with a long-worded definition: “A group of interconnected loads and distributed energy resources (DER) with clearly defined electrical boundaries that acts as a single controllable entity with respect to the grid [and can] connect and disconnect from the grid to enable it to operate in both grid-connected or island mode.”
To large extent, Pike Research adheres to this definition, but with one major exception. One segment of microgrids included in our Tracker update is “remote microgrids,” networks of distributed resources that are not interconnected with a larger utility grid, primarily located in the developing world.
All told, Pike Research identified 87 new microgrids either planned, proposed or in current operation which now total over 2,574 MW in planned or operating capacity. This compares to 1,626 MW of planned and operating capacity identified in the Pike Research 4Q11 update, a 63% capacity increase.
I had lunch with Gary Seifert, a business development executive at OSIsoft, at the recent OSIsoft Users Conference in San Francisco. He opened my eyes to the perspective of a company whose data management systems are vital for the University of California-San Diego’s microgrid, whose 42 megawatts (MW) of total capacity generate over 84,000 data streams through the company’s “PI” system, a volume of data that keeps growing and can reach 100 bits per second.
From a company such as OSIsoft, the islanding requirement to define a microgrid is largely a false one proliferated by academics. In the real world (I am paraphrasing here) it’s the organization, optimization and visualization of data for customers that really constitutes a microgrid. “Don’t tell some of these military bases that they don’t have a microgrid if they cannot fully island yet,” he warned.
Data Architecture for UC-San Diego Microgrid
Seifert made a convincing case that the islanding threshold for a microgrid may be too stringent. Nevertheless, Raj Chudgar, vice president of smart grid/microgrids for Power Analytics – whose modeling software is layered on top of OSIsoft’s PI system at UC-San Diego – claimed that only 5% to 10% of the projects listed in Pike Research’s MGDT published in the 4th quarter of 2011 met his vision of what constituted a bona fide microgrid. Since his firm’s software is among the most sophisticated available, this assessment did not surprise me. (Chudgar was a panelist on our May 22 webinar, “Renewable Energy Integration.”)
Outside the Lines
Indeed, not all of the projects profiled in the Tracker database meet the Pike Research and/or DOE definitions of a microgrid. Some projects were included due to their noteworthy features and/or key contributions to the development of technologies critical to the success of the overall microgrid market. Yet Pike Research still uses the ability to safely island as the key distinguishing feature for microgrids for very practical reasons. If we followed the OSIsoft view, it would be impossible to track all projects labeled a “microgrid” due to the sheer numbers.
This is an even larger concern with remote systems. Therefore, Pike Research screens these microgrid projects according to the following criteria: (1) inclusion of a renewable energy generation resource; (2) some network controls that allow for optimization of generation, loads and (in most cases) some form of energy storage.
At present, Pike Research does not include remote Direct Current (DC) telecommunications towers in this database. These systems number in the hundreds of thousands, and so would be virtually impossible to track on an individualized basis. Furthermore, Pike Research generally looks for a microgrid to feature at least two generation sources, two different buildings (and usually some human occupants of these building structures) as basic criteria for a microgrid. As this market matures, these rather artificial screening functions may be revised. The lines between microgrids, virtual power plants and smart grid renewables integration will continue to blur, making market segmentation of the microgrid market increasingly difficult.