Why Building the Smart Grid Will be a Long-Term Project
- Written by George W. Arnold and Wanda K. Reder
Medium-term prospects for the smart grid will be among the key technology topics addressed next month at the IEEE's Technology Time Machine conference in Hong Kong. The purpose of the small and, frankly, elite meeting is to assemble people who are betting their corporate and national futures on when critical technologies will mature and take off. To judge from preliminary assessments laid out in a white paper prepared for the conference, some technologies, such as cloud computing, already are at a hockey-stick inflection point, while others, such as the so-called "Internet of Things," will reach that point in perhaps ten years' time. With the smart grid, due to immense technical challenges and acute engineering shortages, the inflection point may be closer to two decades away.
The smart grid will mean using information and communications to make electric power delivery systems more efficient, flexible and dynamic, to save energy and to accommodate intermittent renewable sources of energy as well as electric vehicles (EVs) and plug-in hybrids (PHEVs).
There is a stark contrast between the grid as it exists today and the grids of other sectors, such as telecommunications. In their essential features, power delivery systems are not much different than they were a hundred years ago. What is more, the information and communications technology that is starting to be applied in the grid is, to a great extent, similar to what was done 30 years ago to telecom infrastructure.
In a nutshell, the smart grid represents the convergence of technologies that control the flow of electrons on the one hand and bits on the other. Electrons, in comparison with pieces of information, are much less tolerant of impairments and manipulation—delays, latency, compression, coding and so on. To remain stable, the grid must be balanced in real-time at all times.
Smart grid technologies are being introduced at a time when power systems are in urgent need of renewal and expansion, anyway. In the advanced industrial countries, transformers and substations are, on average, about 40-years-old. In some of the rapidly industrializing countries, such as India, electric infrastructure is inadequate and unreliable. And in some of the world’s poorest countries, there is no infrastructure at all. According to a United Nations agency, about 1.5 billion people have no access to electricity.
In the next 20 years, as systems are incrementally improved or built out, techniques will be borrowed from control system theory and IT to improve performance in transmission and distribution, which will tend to converge somewhat. Intelligence will be embedded in the grid, supported by a great many sensors, so that the system can self-correct and negotiate its way around problems. This is not trivial. Using such intelligence as the basis for modifying the behavior of the system is like changing the operation of a Boeing 747 in midair. It is going to take time to demonstrate that we actually know how to reliably control a stochastic system.
Engaging the consumer usefully in the workings of the smart grid is another major challenge. Smart meters are being installed by the millions around the world (in the United States, the 2009 Stimulus Bill provided billions of dollars for smart grid programs). But for the customer—and society as a whole—to benefit, consumers must have visibility into their energy use and the tools to manage it. So far, many utilities and energy providers do not seem to have kept this priority front and center in their planning.
Though a lot of innovation will take place at the granular level as smart grid technologies are introduced, the general features of the grid will not change radically in the next two decades; we will not see the kind of architectural sea change witnessed in telecom during the last 30 years.
When there is a sea change in electric power, if there is one, it will be from centrally generated power to ubiquitous distributed power. Every home will have a solar array, personal windmill or fuel cell, and each such system will, at times, feed energy into the grid as well as draw power from it. The grid, as a whole, will be an enormously dynamic federation of micro grids, in which supply and demand are constantly being rebalanced in response to price signals and physical constraints. This vision is more likely to be realized 30 or 40 years from now than in the next two decades.
That said, there are several potentially disruptive technologies that could accelerate the evolution of the smart grid. One is cost-effective storage. With EV and PHEV development programs driving R&D in batteries, some manufacturers are starting to offer grid-scale arrays, drawing on new technology developed for cars.
The theoretical potential is huge. On average, about half the power grid’s generating capacity, built to handle peak loads, is unused, and automobiles are parked 90 percent of the time. So if car batteries could store energy for the grid and feed it back as needed, ubiquitously, it would be a revolution in power. What may be needed most is not so much technical innovation in the narrow sense but innovation in business models and regulation; think of Google and Facebook, where business ingenuity has been at least as important as engineering.
Batteries are not the only possible disrupter in energy storage. Fuel cells, tapping the thermal energy stored in buildings, pumped hydro and compressed air are other candidates.
Whenever you graft existing technology onto another technology where reliability is absolutely paramount, it takes time, money, effort and talent to innovate. This is the main constraint on the smart grid. At present, the U.S. utility industry spends about 0.3 percent of its revenues on R&D; literally less than the dog food industry spends for research. The Center for Energy Workforce Development has found that, in the next five years, about half the U.S. workforce in electric power will be lost to attrition. The situation appears to be similar in other advanced industrial countries, from Australia to Europe.
To begin remedying the situation in the United States, a 2007 National Science Foundation workshop was launched: the IEEE/U.S. Power and Energy Engineering Workforce Collaborative. The group’s April 2009 report recommended doubling the rate at which degrees are conferred in power engineering, hiring 80 new university faculty members in the next five years, and increasing federal research funding. The current acute shortage of seasoned experts affects not just implementation of the smart grid but regulation of it as well. Supervisory authorities at all levels lack the talent they need. And although standards setting has gone rather well, there is a reluctance to give up proprietary approaches and adopt procedures that will make different systems fully interoperable.
Cultural and subcultural barriers will also have to be overcome. Power engineering, communications, computing and IT all have their separate vocabularies, their special ways of talking. So there is a social aspect to getting people with diverse skills and backgrounds to work together meaningfully.
Traditionally, utilities take a very conservative and cautious view of new technology. Equipment depreciates over 20 to 30 years and amortization rates are key to rates of return. So utilities do not make rash decisions, and robustness and reliability are still paramount.
Conversely, in communications, there has been a trade-off of reliability for functionality in recent decades. Because there is so much redundancy and so many enticing new features, we are willing to put up with impairments that would have been intolerable when a five-nines philosophy ruled (99.999 percent reliability). But now telecom engineers are entering the realm of power system control, where five-nines still prevails.