Interview with Stefano Galli
In this interview, Dr. Stefano Galli discusses the need for a holistic understanding of Smart Grid. He explains that a system-wide view of Smart Grid is fundamental to effective research and development and to network design.
Question: Smart Grid means different things to different people. How do you view the Smart Grid and how do you define it?
Definitions do vary, and it's because people tend to look at Smart Grid from the perspective of their own areas of expertise or interest. They tailor their definitions to reflect their own perspectives and sometimes forget to look at the Smart Grid as a whole. Thus, arriving at a holistic definition of the Smart Grid is challenging because any attempt to define it usually privileges one particular aspect of the grid.
Fundamentally, when seeking to define Smart Grid, it's important to recognize that the power grid is a commodity delivery system where the commodity, which is energy, has a production-to-consumption cycle time of zero. This is the only commodity in the world I know of that has this characteristic. Generation, delivery, and consumption all have to happenat the same time. The power grid is a perfect just-in-time system where generation and demand must be balanced at every instant.
Because electrical power moves just as fast as communications signals do, and because the power industry has a production-to-consumption cycle time of zero, the power grid naturally presents fundamental challenges in sensing, communications, and control. These challenges will escalate once certain new applications and technologies are introduced, particularly new technologies that are aimed at sustainably addressing energy independence.
Question: What types of new technologies will be part of the Smart Grid?
Some of the new technologies needed to address these needs will include utility-scale renewable sources that feed energy into the transmission system; distributed and renewable energy resources that feed into the distribution system; plug-in electric vehicles, which will potentially create large load increases in sections of the grid; new demand-side management techniques that will give consumers interactive ways to participate in Smart Grid; and storage technologies that will allow us to introduce some latency between the generation and consumption cyclesto help compensatefor the time-varying nature of renewable resources such as wind or solar energy.
Question: How challenging will it be for utilities to integrate these new technologies?
The "smartness" in the Smart Grid will have to ensure the balance of generation and demand in the presence of time-varying and stochastic (random) generation and demand influences, which is a substantial departure from the deterministic paradigm of today’s grid.Integrating thesenew applications and technologies while still ensuring the stability of a perfect just-in-time system will require the addition of new control, monitoring, and protection mechanisms.
These additional requirements exceed the capabilities of the existing grid, however, and can't be achieved by simply modifying the current supervisory control and data acquisition (SCADA) network. In particular, the Smart Grid communications network mustincorporate new design features that address Smart Grid's two major requirements: (1) integrating time-dependent renewable resources; and (2) controlling the load, which will be influenced by the introduction of new, dynamic end-use behaviors.
I think these two design features, more than any others, will require introducing the most intelligence to the existing power grid. They will require a sophisticated design that combines sensing, communications and control.
Designing this system will be extremely challenging. Most of the other technologies, such as smart meters, are important and will make the aging grid run better, but they are not really necessary nor are they rocket science.
Question: What are the most expeditious approaches for determining how to integrate time-dependent renewable resources and controlling a very dynamic load?
We need a lot more research and development and a more holistic approach to R&D in Smart Grid.
Fundamentally, the issue is how to design a reliable and robust system to control and manage communications on such a wide geographic scale for a system that is becoming increasingly stochastic and that pretty much often operates near a state of instability. It's a problem that has not been solved. There are many open challenges related to this and solving them cannot be accomplished in piecemeal fashion. This requires funding fundamental R&D.
In particular, we need a lot of interdisciplinary R&D to learn how to implement distributed control under communications constraints. Furthermore, the fundamental interplay between communications and power flow dynamics is still not completely understood. We need a systems engineering approach to tackle this issue.
The problem is compounded by the fact that often we work in isolation from experts in other fields. It's long been a joke that power and communications engineers do not often talk to each other and control engineers only talk to mathematicians. These behaviors compound the problem even more.
Increasing specialization in the profession is also not necessarily a strong point when it comes to Smart Grid. I once heard a very bright colleague say, "We're very good at designing subsystems, but the Smart Grid should not become a dumb grid composed of many smart subsystems." The statement aptly describes what we're up against.
Question: Are companies allocating enough money for R&D?
The utility industry does not spend a lot on research and the research it does perform focuses a lot on subsystems and on the "D" side of R&D. This practice underscores the argument I made beforeabout the importance of looking at the whole problem in a holistic way. What is missing is basic research and correctly formulated problem statements.
Basic research is traditionally done in academia, but because not many grants have been available for Smart Grid until recently, academic R&D is definitely behind where it should be. Things are getting better, however, and research groups in universities are expanding their utility research programs.
The R&D issue is a worldwide problem, although it is perhaps more emblematic of the U.S., where the need to focus on short-term revenues greatly influences allocations for research. Wall Street wants companies to meet quarterly estimates. Nobody has really the courage to invest in a 10-year R&D program.
Question: How important are standards to the Smart Grid and to getting it off the ground?
The power grid often crosses national or jurisdictional boundaries. This is the case in Europe, for example, and even in the United States, where there are different regulations in different states. In such an environment, it is important to make sure that devices and applications communicate across boundaries regardless of who installs or operates them. A standardization process is the only means to ensure this.
Standards are also important because the implementation of the Smart Gridwill require decades to accomplish. Even with conventional technologies, equipment installed onto the grid usually stays there for decades. Power companies cannot put themselves in a position where they could be hostage to a few companies that push proprietary solutions. You never know, in 20 years the vendor of a proprietary solution could go out of business, and then the utility is without support for the proprietary solution. The industry needs to have an ecosystem that is always able to provide it with a solution that ensures interoperability, and multi-source agreements are the only way to guarantee this.
The downside of the standardization process is that it takes time to develop excellent technical solutions and the process often can't keep pace with the market's needs for new technologies and services. All standardization venues and all standardization processes are thus tempted to accept technical work that is done first in trade associations or other groups where fewer stakeholders might contribute to the development of a specification. And sometimes after a product hits the market its proponents seek to have the specification rubber-stamped into a standard.
This trend is common to all standards development organizations but it should be resisted. This is particularly important for Smart Grid. The technical work for Smart Grid should be pursued in standards development organizations where a wide variety of stakeholders from various industry segments can contribute to standards decisions.
Question: How would you describe the IEEE Standards Association's role in the Smart Grid?
IEEE standards span a wide set of disciplines, from power engineering to communications to the integration of renewable resources. The breadth of IEEE standards has no rivals. This is due to the fact that IEEE ensures the implementation of a fair and open global process for integrating and leveraging technology through the IEEE's 40-plus technical societies and councils. This process enables industry to leverage IEEE's resources, which include the best electrical engineers in the world, the best scholarly publications, the best conferences, and so on. This technical excellence is embodied in IEEE standards. An example is IEEE's Smart Grid standardization program, which integrates multiple technology platforms and cross-industry requirements.