Microgrids: Smart or Dumb?

Peter Asmus | Nov 11, 2010

Microgrids may be a hot topic among those forecasting key future trends shaping the world's energy infrastructure, but few significant state-of-the-art commercial microgrids are actually up and running in North America, the world's leading market for microgrids. One leading domestic developer claims that not a single microgrid is providing energy services today in the U.S., but that firm uses a very narrow definition of what a microgrid is, excluding remote, off-grid microgrids within its qualifications, for example.

At present, regulations governing energy have not kept pace with emerging microgrid islanding technology, frustrating immediate progress. Most of the public and private investment dollars pouring into modernization of the globe's electric grid have been soaked up by utility smart grid deployments, with very little funding filtering down to the microgrid level of design and deployment.

Academics from the University of Wisconsin-Madison -- an institution often credited with the birthing of the microgrid concept (at least in engineering terms) -- predict it could take 30 years for the microgrid to become ubiquitous. Yet current trends appear to make microgrids an inevitable augmentation of today's centralized grid infrastructure. Aggregation platforms similar to microgrids will be absolutely necessary if our energy infrastructure follows in the footsteps of telecomm and the evolution of today's Internet. No doubt the existing radial transmission grid will still provide the majority of power supplies to the industrialized world. But renewable distributed energy generation (RDEG) will also play a larger role in providing energy supply, reliability, security and emergency care services.

Given consumer pushback on smart meters -- the very underpinning of the utility-dominated "smart grid" -- in California, Texas, Colorado and elsewhere, the microgrid represents an alternative business model for boosting the quality of grid services. It is becoming self-evident that the hype behind the Obama Administration's stimulus spending on smart grid upgrades raised expectations to unrealistic heights. Furthermore, utilities focused too much on the benefits meter data might bring to their own operations -- and forgot to connect the dots with consumers, many of which only saw higher bills, and no coordinated programs and tools to respond to real-time price signals with more efficient consumption patterns and protocols. And then there were the concerns about data security.

The goals of both the smart grid and the microgrid are the same: to maximize generation assets through embedded intelligence while dramatically boosting efficiencies, thereby minimizing costs. However, they appear to offer two potentially different paths forward.

Both "supergrid" and "microgrid" will need to get smarter, though it is the distribution system that is currently the prime source of outages and unreliability. Today's distribution grid network is clearly inadequate to support the type of innovation now occurring with distributed resources, including devices such as plug-in hybrid electric vehicles (PHEV) serving as distributed storage batteries. The question is: Do we need bottom-up or top-down innovation?

Microgrids installed in developing nations or rural regions of the United States may be quite simple, even "dumb" if compared to the hyperbole often attached to descriptions of the smart grid. The Consortium for Electric Reliability Solutions' (CERTS) demonstration projects show that microgrids do not necessarily need to rely on all of the sensors and fast, real-time communication protocols that are hallmarks of the smart grid.

Among the current microgrid control options are centralized management systems requiring high-bandwidth links between the inverters and central controller. Other prototype microgrids rely upon distributed on-board control that reduces the bandwidth needed -- but at the cost of synchronization difficulties. More recent work has investigated a hybrid control scheme where proximate inverters operate in a master-slave arrangement. Still others are focused on remote or smaller microgrids are sticking with common frequency droop method, commercialized through the CERTS work, which greatly reduces the need for any high-bandwidth communications over large distances.

Control systems fall into two major camps. The purists -- epitomized by the CERTS software -- believe that microgrids should operate without any central command and control system, with generators and loads harmonizing autonomously based on local information. This is the view espoused by leading academics and localization advocates and the rationale is compelling. This system will work for the majority of smaller microgrids with a single owner and whose top priority is reliability and sustainability during emergencies. These are the "dumb" microgrids, if you will.

In the other camp are what you might call the pragmatists. They lean toward systems that can be described as "master/slave," (whereas the CERTS approach has been described as being "like a commune.") These operating systems are much more focused on optimization of services outside the microgrid. The benefits of reliability may come second to generating new revenue streams from excess generation (or even demand reductions.)

There are also those systems that can straddle these two views. There are few clear cut direct competitors in the space since no standards exist and microgrids are so modular, diverse and optimize such a broad array of energy-related services. It is these control systems -- still literally being defined -- where the fiercest competition may reign within the microgrid space. This is the guts of the microgrid, if you will, and the focus of current software innovation.

Related Topics


So what's the future in your opinion? Gated communities with standalone-capable micro-grids, surrounded by "everyone else" with third-world-level service?

Thank you for reminding your readers that the distribution system is the prime source of outages and unbreliability. That is too often overlooked in discussions of the need for baseload power and new transmission. If the smart grid community continues to ignore the progress being made on small grids to incorporate high penetrations of renewable power it may indeed take 30 years for them to benefit from today's technology. Meanwhile, the cost of oil-fired generation will have motivated the small grids to leadfrog over them and lead the way to clean, smart micro-grids. http://homerusersgroup.ning.com/profiles/blogs/clean-smart-microgrids

Another aspect of the SmartGrid/MicroGrid controversy is that there's a strong leaning to use standardize protocols for it. My fear is that ubiquitous use of standardized protocols will put at risk the entire grid. I feel that regional, or even utility level variations should be encouraged to provide a defense in depth to cyber crime/warfare.

The Smart Grid and a microgrid are things different. The Smart Grid is a social paradigm. A microgrid is an electrothecnical tool for aggregation renewable energy sources (mainly distributed generation) and other kind of "smart" devices to a distribution grid. What still has not been solved (or envisaged) is which will be the proper "aggregation approach" in the future. A microgrid is one of the proposed "aggregation approaches", but "vrtual power plants" is another one. Some "aggregation approaches" are market oriented and others are "grid oriented". The underlying control schemes to each "aggregation approach" is then different. Which one will prevail? That is one the most improtant "to be or not to be" questions regarding distribution networks (MV&LV) nowadays.

J Martin proposes only two possibilities exist, but he is clearly wrong about that. Among others, IMEUC is a distinct option, though more favourable to customers than to large utilities, so obviously off the table.

It's actually a bit fun to watch how much the utilities hate the concept of "competing for customers". Sorta like watching dowagers crossing a ditch.