Telecommunications Needs for Distributed Generation
By Jim McClanahan
The best policy in distribution planning is to anticipate telecommunications requirements and plan comprehensively. The costs of not doing so can include installation of incompatible or redundant systems, an inability to scale systems up and excessively difficult or expensive maintenance. Failing to plan properly might as well be planning to fail.
It is no stretch to say that the significant economies of scale associated with generation were among the underpinnings of electric power in its first century. Building centralized generation stations and then distributing the power to customers was simply the way things were done.
Over the last two decades we have seen the emergence of generation technologies that challenge this model and, in ways, turn the electric utility on its head. The amount of distributed generation being added to the modern grid is significant and continues to grow. It has become so integral to the industry, the U.S. Department of Energy considers the ability to support a variety of generation and storage options one of the seven cardinal traits of the smart grid.
Considerable planning goes into the reliable and efficient integration of distributed resources with the electric system. There are typically a number of applications necessary: Metering information for hourly energy accounting and monthly billing must be collected; real-time data and controls need to be integrated into the utility's SCADA system; and high-speed protection coordination for schemes such as transfer trip can be required. Traditionally, these different types of applications have, due to different requirements, used different types of telecom connectivity. The problem for utilities is that with the rapidly expanding number of distributed sites, if each site has two or three independent and different telecom systems there is an increase in the time, complexity and cost of deployment.
When evaluating telecom options three key considerations concern latency, bandwidth and reliability. Things such as the coordination of transfer trip operation between equipment at the substation and equipment at the distributed generation (DG) site require a reliable communications system with low latency. SCADA requires a nearly continuous flow of information, but with limited bandwidth requirements. Energy accounting data and billing data have even lower requirements.
There are several solutions that meet these needs and a few that can provide significantly more bandwidth than is currently required, which could enable future applications such as video surveillance of the site for enhanced safety and security.
When looking at wireless options, typically the types of latency required for an application such as transfer trip—signaling to remote circuit breakers to assure complete clearing of a fault—will be point-to-point in nature. This implies use of a dedicated pair of radios that only communicate with each other (possibly with a dedicated repeater site between them).
One common question is whether radios available for license-free operation are suitable for these types of critical applications. Experience at many utilities has shown that solutions using the 900 MHz ISM (Industrial, Scientific and Medical) band and other license-free frequencies can be reliable. The key with license-free solutions is having a solid design, a thorough site survey of the path being considered and use of directional antennas. It is also important to remember that even licensed solutions occasionally incur interference. The difference is that with licensed solutions you will have some recourse to address the issue, but there is certainly no guarantee you will never see interference or that the process for handling interference issues will lead to a quick resolution. As with many things, the question of licensed versus license-free solutions does not have a simple answer. Instead it requires identifying and evaluating a variety of considerations.
Fiber optic connectivity is another option used by some utilities at certain distributed generation sites. While this type of link can provide low latency and plenty of bandwidth it can be expensive to implement. A typical radio link might cost between five and ten thousand dollars, while it typically costs twenty-five to fifty thousand dollars per mile just install fiber (excluding the end equipment that lights it). In general utilities seem to be finding the point-to-point wireless solution more cost effective and quicker to deploy than fiber.
For sites larger than roughly 2 MW, mechanisms like transfer trip are critical for the safety and stability of the grid. For mid-sized distributed generation sites (under 2 MW but larger than 500 kW), the requirement may not be as critical. IEEE 1547 requires the detection and elimination of unintentional islanding within two seconds. While larger distributed generation sites may need transfer trip to ensure they meet this condition, the minimum feeder load is typically much higher than the capacity of mid-sized generation sites, so local protection is sufficient to disconnect them should the feeder breaker trip.
While these sites may still require metering and SCADA data, they do not have the need for the extremely low latency required by some protection systems. For these sites some type of carrier option such as cellular data or a mesh radio system that supports traditional distribution automation may be sufficient.
The important thing is to stay ahead of the curve. A hodge-podge of telecom solutions can easily become entrenched as the number of distributed sites grows. So it is much better to start with a clearly defined approach, rather than find in the middle of multiple deployments that the approach adopted simply does not scale to the number of sites you are dealing with. Limits to scaling can appear in many ugly forms—excessive effort to install, saturation of the available bandwidth (especially in mesh networks), or unduly cumbersome maintenance, among others.
Starting with some analysis and a well-defined approach ensures consistency, makes it easier to explain to site developers what is expected of them, helps the utility reach efficiency of scale and ensures that the concerns and needs of all stake-holders and applications are identified and considered.
It is often said that failing to plan is the same as planning to fail. This definitely proves true for technologies such as distributed generation and storage that are rapidly evolving and growing, tending to outpace complete integration with the grid.
Jim McClanahan, an IEEE member, is a senior principal in the energy and utilities practice of West Monroe Partners. Previously he worked for S&C Electric as Director of Smart Grid Services leading the teams that planned, designed, installed, commissioned and supported complex smart grid solutions. In addition to being involved in dimensions of the smart grid such as distribution automation, distributed generation and AMI, he also worked on aspects of telecommunications like radio frequency and data network design. He earned a B.S.in electrical engineering at the University of Tulsa in 1985 and an M.B.A. at Oklahoma State University in 1992.