Despite the hype around what new technology can do, we remain in a transitional period as new systems are incrementally added to legacy systems, with effects that ripple across operations. Thinking through the implications of each technology addition or change can pay dividends, operationally and strategically.
Hype is out, return-on-investment is in.
Though "electric power has been dominated by customer solutions," today "integration of the pieces has become the dominant challenge." Add the presence of intelligent electronic devices (IEDs) and the resulting reams of data and processing that data into "practical, efficient and actionable instructions" is another critical challenge.
So writes Ron Willoughby, formerly with KEMA Inc., now an independent consultant, in an article last month in Distributed Energy titled, "Power System Automation Drives the Need for Smart Grid." 
The article is succinct, graphically illustrated, and provides thought leadership on the growing complexity of grids as they are modernized in myriad ways.
I caught up with Willoughby by phone recently and we hashed over a few key points.
Despite the hype around what new technology can do—providing situational awareness of power flows, power quality, faults, the behavior of systems—we remain in a transitional period as new systems are incrementally added to legacy systems, Willoughby told me.
That hits on his point about integration in a pre- plug-and-play world built entirely on global standards. (In my own, personal view, the new holy grail of the power industry.)
"There's a big challenge for manufacturers here," Willoughby said. "Depending on the manufacturer, they've left their proprietary technology in place and provided interfaces to talk to the outside world. Having been in manufacturing I can say the situation is difficult. On-going revisions to control software is not easy. Because as manufacturers advance their product offerings by developing new features, they still have to support legacy systems.
"So, there needs to be a transition period for manufacturers from legacy systems to new," Willoughby continued. "And for the end-users, there also needs to be a transition, because they've got such an investment in legacy equipment and associated training, no one can afford wholesale replacements. So, you've got to find a way to make new technologies work with old."
"In many cases, this boils down to data management," he said. "How critical is a specific data stream? Does it need a path that allows near real-time flows? Or will periodically collected (interval) data suffice? Can a system or application use distributed intelligence or does that data need to come all the way back to a centralized operations center with commands sent back the same way, which increases communication delays?"
"So, the transition period in which new systems are integrated with old is not simply a linear addition," Willoughby pointed out. "As IEDs or end-use meters are added in grid modernization projects, they need to be accompanied by an evaluation of their impact on the overall system. What is the nature of the data generated? Is it needed in near real-time? Or will historic data serve a purpose? Will distributed intelligence keep the network from bogging down?"
These thoughts tied back to my first conversation with Willoughby two years ago, which we captured in "The `Next Big Thing.' " By that phrase he meant data management and analytics. Every device or system added to an existing grid carries myriad implications, and how an entity handles decisions about distributed intelligence, about data flows, what is critical to real-time operations, and which data sources can be addressed by distributed intelligence in the field can shape investment decisions and business cases for decades.
"Here's the starting point," Willoughby said. "What I tell people is that at the very beginning, you've got to first decide what data do I really need, and how am I going to use it? That will define where you need to place devices to collect that data, and how frequently you need to collect it. Then, the communications system can be designed to fit the need.
"With that information, I have a plan for collecting and processing the data," Willoughby added. "Where people have failed is they've collected information because they think they're going to need it. And then they get overwhelmed by it. So be careful to collect only what you need and decide how far upstream it needs to go. That will define storage requirements, back office equipment needs, and whether that storage and processing might be better done at the substation level, for example."
Take dynamic voltage control, load management and distributed resources, for instance.
"How do these three go together?" Willoughby asked, rhetorically. "Let's assume the value story is `how can I save the end-user money?'"
The traditional approach of sending more energy than needed to guarantee acceptable voltage at the end-user's home or business wastes energy, he pointed out. But with a smart meter at that endpoint, you can know with certainty how much voltage is reaching the end-user. You now have an option to act on that information, assuming equipment exists on the distribution system and at the end-user to dynamically regulate the voltage.
On the distribution line, you can have single-phase voltage regulators with controls. You can also have pole-mounted capacitors with controls, reclosers with controls and you've got relays, regulators, tap changers with controls. In other words, you can add equipment with controls all along the distribution line. The efficiencies achieved through the resulting dynamic voltage control saves fuel at the generation end, a savings passed through to the customer.
The same data allows the utility to levelize the voltage along the distribution line, cutting line losses. That's where distributed resources come in, because they can provide energy closer to the load. They also raise the voltage at the connection point, further complicating the control systems.
If you also have load control for end-user systems such as air conditioning or pool pumps, you have additional tools to achieve efficiencies.
The upshot is that devices, systems and functionalities can be added to a grid and at the end-user to achieve synergies of efficiency, if you will. But the foregoing points about data management, flows and distributed intelligence apply.
Another important factor in this process, according to Willoughby, is that when utilities are transparent and communicative about the efficiencies and potential cost savings with customers—as they are with regulators—acceptance of these new technologies will come much faster and with much less resistance.