A building is often only as intelligent as the electrical distribution network it connects with. That's why smart buildings are often seen as an extension of the smart grid.  

Meters, building controls, intelligent lighting and HVAC systems, distributed energy systems and the software layered on top are indeed valuable for controlling localized energy use within a building. But in many cases, the building relies on the utility or regional electricity grid to value those services.

Some analysts define these technologies as the "enterprise smart grid" because of their interaction with the electricity network. 

But what if there were no supporting centralized grid? How would buildings be designed then?

That's what Bruce Nordman, a researcher at the Lawrence Berkeley National Lab, often thinks about. 

In today's grid-centric framework, the building works for the benefit of the larger grid. But Nordman thinks there's a future where the opposite is true. And that future is the nanogrid.

"In dealing with the grid, it's been about incremental adjustments rather than shifting to something fundamentally better," said Nordman in an interview. "The grid should be a slave to the building, not the master."

In May, Nordman published a paper (PDF) with Ken Christensen, a computer science professor at the University of South Florida, outlining why they think nanogrids are the best way to manage energy consumption and production.

A nanogrid is different from a microgrid, according to the authors. Although some microgrids can be developed for single buildings, they mostly interface with the utility. Some aren't even fully islandable. A nanogrid, however, would be "indifferent to whether a utility grid is present." Rather, it would be a mostly autonomous DC-based system that would digitally connect individual devices to one other, as well as to power generation and storage within the building.

Nordman and Christensen describe it this way: "A nanogrid is a single domain of power -- for voltage, capacity, reliability, administration, and price. Nanogrids include storage internally; local generation operates as a special type of nanogrid. A building-scale microgrid can be as simple as a network of nanogrids, without any central entity."

The nanogrid is conceptually similar to an automobile or aircraft, which both house their own isolated grid networks powered by batteries that can support electronics, lighting and internet communications. Uninterruptible power supplies also perform a similar function in buildings during grid disturbances.

Here's a simple diagram illustrating how a nanogrid -- based on the concept of "local power distribution" -- would function. Essentially, it would allow most devices to plug into power sockets and connect to the nanogrid, which could balance supply with demand from those individual loads.

"It presumes digital communication among entities, embraces DC power, and is only intended for use within (or between) buildings. Building-scale microgrids are built on a foundation of nanogrids and pervasive communication," wrote the authors. The system capacity could be anywhere from a few kilowatts to hundreds of kilowatts.

There are lots of potential benefits to structuring local DC power distribution in this way, they argue. Conversion losses would be cut, investments in inverters and breakers would be reduced, and device-level controls would enable a much more nimble way to match generation or storage capabilities with demand. The building would also theoretically be immune to problems more likely to be encountered with a local microgrid or the broader centralized grid.

However, obstacles currently outweigh those benefits by a significant degree.

Nordman thinks nanogrids should be universal, meaning they would operate on the exact same communications and voltage standards. But because nanogrids are still mostly conceptual, no organization has attempted to create those standards. Nanogrids can't really scale without them.

Nordman is attempting to find funding for simulation work, but without money, he admitted that the concept is "dead in the water."

There's also another structural issue to deal with on the grid. Theoretically, these localized, autonomous systems could be scaled without much interference with the utility. "I'm trying to drive utilities out of the building," said Nordman.

But he also envisions nanogrids being tied to larger microgrids, which are ultimately tied to the central grid. That inevitably brings utilities into the picture, and they likely wouldn't have much incentive to support a system designed to drastically cut their electricity sales. That's why growth in nanogrids will likely occur in developing countries with weak grid services where it makes sense to power buildings in isolation.  

The concept is compelling. It's also so completely different from the status quo that it currently has little chance of scaling in this grid-centric world.

When asked whether he thought nanogrids would gain traction anytime soon, Nordman answered bluntly: "I would say 'no' because what we’re all being told is that we have to organize around the needs of the grid. This is the case against the smart grid as we know it."


To learn how to effectively manage, analyze and take action on the data generated by grid networks, join Greentech Media at The Soft Grid: Data, Analytics and the Software-Defined Utility conference on September 10-11, 2014 in Menlo Park, CA. Look forward to seeing you there!