What's Complexity Theory Got to Do with Power Engineering?

April 18, 2012
Seen together as a dynamic system, the increasingly complex utility-customer-regulator network may have some big surprises in store for us.

In the 1990s I had the privilege of being part of a utility research organization that was leading out in developing the concept of 'distributed utility' – a major forerunner of 'smart grid'. The premise was that central generation had reached the limits of economy of scale, and smaller, distributed generation resources, particularly aero-derivative gas turbines, were more economical and easier to site closer to load. Of course, there was lots of discussion about system stability, interaction with large remote generation, and how all this would work with looming deregulation of the electric utility industry. In California, deregulation meant that utilities would soon have to delaminate. The traditional vertical integration of generation, transmission and distribution would be broken up to encourage competitive market forces to do their thing, similar to the breaking up of the telecommunications industry.

We had frequent brainstorm meetings with engineers, economists and operations research experts taking turns at the whiteboards. Arrows and symbols connected the various elements of the physical electrical system with marketers, regulators and consumers. Here was the generation resource, there was the consumer, over here is the market place where time and location based pricing is determined. Black, blue, red and green dry-erasable lines went everywhere. Trying to add a little humor I even wrote a little story titled: "The Totally Disconnected Utility" and passed it around the office. I got a few yucks but most of my colleagues just grabbed another bagel and went back to their computers and whiteboards. This was serious, industry changing stuff!

Fast-forward 20 years. The push for deregulation has sort of fizzled. The promise of cheaper, market driven rates hasn’t uniformly come true in the 24 states that have deregulated their electric supply. Other states are resisting, and congress seems baffled on how to proceed. Still, all the distributed utility conceptualizing and the anticipation of deregulation began to crack open the door to what we now refer to as smart grid.

That early creative work also opened a potential Pandora's Box and some major challenges flew out – some of them being issues that might be better understood under the framework of complexity theory. Complexity theory attempts to describe the behavior of complex systems in which large networks and sub-networks of adaptable components, without central control, can interact in ways that are unpredictable under conventional systems analysis. The tighter, the more connected these components are, the more the unpredictability.

Complex systems aren't necessarily chaotic systems, although they can have chaotic behavior (in the mathematical sense). Complex systems phenomena occur in economics, biology, network theory – wherever the behavior of the whole can't be understood from studying the behavior of individual parts.

Take the example of army ants. An individual ant’s behavior isn't very complicated. Even when you put a few dozen of them in a room they’ll follow each other's chemical pheromone trail, wandering aimlessly until they die. But, put half million of these simple creatures together and you have a horde that can destroy everything in its path. Then, interweaving their bodies, the ants will methodically build a shelter for the night, protecting larvae and queen. Next morning at the crack of dawn the ants untangle themselves and the march resumes.

All without a central command!

Nearer to home, that five pound mass sitting between your ears has been studied and modeled for decades and we still can’t explain most brain behaviors in terms of the individual characteristics of its 100 billion or so neurons.

Getting back to power systems, we didn't have to consider the implications of complexity theory when we didn’t have a complex system. For 100 years our generation and delivery systems were complicated but not complex in the formal mathematical sense. Oh sure, some academic studies have shown the potential for chaotic electrical behavior in generator-grid interaction, but nothing much has come of it. Now, however, we're going flat out to make the utility-customer systems as distributed, interactive and independent as possible using the latest telecommunications, automation and customer side control technologies. In a few years we plan to have 60 million smart meters deployed, connected to an advanced metering infrastructure and sharing data with advanced distribution automation systems and back office meter data management.

But the physical infrastructure alone won't provide all the interconnections. Customers are interacting with each other and industry regulators, more than ever, in diffuse ways - through the media, through intervener groups, through political and environmental group efforts.

We wanted to get the customer's attention and now we have it. In the long run greater customer involvement is going to produce a better utility-customer-regulator paradigm. However, along the way we can expect some surprises, both in customer behavior, regulator actions and in demands on the physical infrastructure. We're already seeing some unexpected consumer behavior in the various anti smart meter movements that have sprung up around the nation.

Nominally, when you don't know exactly what’s coming, you really want a 'robust' system – one that can adapt to surprises. But sometimes the pursuit of robustness can lead to overbuilding the physical assets and, on the regulatory side, excess bureaucracy.

For a little different slant on what's needed in the physical infrastructure, read "Smart Enough to Play Dumb" by Rick Bush, Editorial Director of T&D World. In the article, Rick quotes Terry Boston, CEO of PJM, a regional transmission organization (RTO) that coordinates the movement of wholesale electricity in all or parts of 13 states and the District of Columbia. They were discussing the value of the smart grid.

"Rick," Terry said, "I'd rather operate a robust grid than a smart grid any day, but the truth is that the smart grid will cost less to build out than the robust grid."

But will the smart grid be adaptable enough to meet all the demands that the complexity of smart grid is creating? Do we have enough infrastructure and organizational robustness to carry us through? Time will tell.

About the Editor

Paul earned his B.S. and an M.S. in electrical engineering from the University of California-Berkeley and is a registered professional engineer. He has worked in the energy industry for more than 25 years, developing and implementing advanced energy technologies. As research director for Pacific Gas and Electric Co. he pioneered methodologies used in the design, maintenance and control of energy delivery systems. As a consultant he has provided guidance to utilities and the vendor community, nationally and internationally. Email him with comments: [email protected]

Voice your opinion!

To join the conversation, and become an exclusive member of T&D World, create an account today!