Communications technology is a hot topic at Autovation 2011 and Utilimetrics is pleased to have Josh Gerber of San Diego Gas & Electric and Joaquin Silva, president and CEO of On-Ramp Wireless (San Diego, Calif.) present their paper Communication Requirements for Improving Utility Efficiency in Smart Grid in the first education session Monday, Sept. 26 at 10:30 a.m.
“Our strategy for the presentation is to discuss the segmentation of the applications for the smart grid,” states Silva. For example: From the substation to the meter on the customer premise, what are the data requirements and information requirements? And how do they vary across each of the applications, whether it is transformer automation? whether it is a meter, etc.? “We want to demystify some of this from a requirements perspective, specifically how it relates to the wireless requirements,” he adds.
After discussing the applications and how the wireless requirements vary, they will move on to some of the specific wireless approaches and technologies that can be used to solve the problems. “We plan to emphasize that it takes multiple types of wireless, depending on the application,” continues Silva.
“We will also talk about how to solve some of the very challenging problems of the smart grid,” he states. One example is that a very high percentage of the distribution grid assets are either pad-mounted or below ground, such as transformers and fault detection sensors. “For example, at San Diego Gas & Electric, 60% of distribution grid feeders are below ground,” states Silva.
This makes it challenging for wireless. You have to use multiple radio technologies to solve the problem. In one case, you may have an architecture where you have a low data rate and very high receiver sensitivity, needing a distributed wireless system or repeater-based solution. You need a higher data rate set of wireless solutions for more latency-sensitive applications.
“A good example is that a lot of utilities want to move toward dynamic volt/VAR control,” notes Silva. If you take the volt/VAR control problem from the substation down to the home, you have to collect the data from the volt/VAR information on various grid points, such as distribution feeder transformer and line sensors.
This is a challenging problem from a wireless perspective, because, again, 50% to 60% of the transformers are in pad-mount steel bunkers. “As a result, you have to be able to collect that data,” states Silva. “It’s not too latency-sensitive in terms of data collection.” However, you have to take into account the control loop for the other assets that exist on the distribution grid that are designed to dynamically manage it.
An example: If there is an electric vehicle plugged into a transformer, it impacts the whole VAR level sensor in the transformer connected to a wireless link that is able to penetrate into the pad-mount. It would sense that change and send an upstream alarm to a controller in the substation, or centrally-managed, which would say: “This has exceeded its threshold.” Now a secondary command in the energy management system would send a control message to a cap bank controller or switch that has the capability to dynamically change the volt/VAR configuration, or perhaps even shed load in the electric vehicle charging station if it is going to exceed a power level threshold on the transformer, which would cause it to explode.
“This is an example of where you need the pervasive wide-area sensing layer, and then a control system,” continues Silva. “However, there is a mismatch in the radio requirements for both. You can’t provide a control loop across the entire smart grid. It would bankrupt even the most well heeled utility.” As a result, according to Silva, you have to use a combination of two radio platforms – a more focused high-speed control layer that is managing the volt/VAR dynamic control layer, and a data collection sensing engine that gives you the pervasive wide area, including below-ground coverage.
Overall, what are the benefits of wireless? “Wired is extremely expensive in an already built-out distribution grid, because you have to lay cable or fibers,” replies Silva. “As a result, it’s not practical in many application segments. The substation typically does have fiber. However, between the substation and the consumer’s home, you need wireless.” Digging trenches and laying fiber typically does not work from a cost model, except in a greenfield situation. As a result, wireless is really the only way to have a low-cost sensing and control layer for the distribution feeder network.
There is still a place for wired, though, according to Silva. However, what the smart grid is all about is pushing out the connectivity and control that exists at the substation down the low-voltage feeder network. “As a result, in terms of growth, wireless applications and technologies are going to be far faster than cable,” he concludes.
