Grid, Heal Thyself

What’s so “smart” about a smart grid? In short, it’s the ability to communicate, predict certain failures before they occur, fix itself when something goes wrong, automatically report failures when they occur, and isolate faults more quickly.


In the past I’ve written about the smart grid and the need for distributed generation, both of which are necessary for a variety of reasons, including the ability to allow renewable energy to be a significant player in the electricity production business. Today I want to tell you about several aspects of a smart grid: the ability to predict certain failures before they occur, fix itself when something goes wrong, automatically report failures when they occur, and isolate faults more quickly. It also allows automatic meter reading and communication with customers’ smart appliances.

The existing grid was built a little at a time, in effect combining a bunch of separate microgrids into a unified whole. In much the same way, the smart grid is being implemented piece by piece. Earlier this year San Diego Gas and Electric added a variety of smart devices – including wireless communication devices, smart meters, and sophisticated sensors – to its arsenal of “self-healing smart grid” tools. I spoke with engineers from On-Ramp Wireless, the company that makes the communication equipment. They gave me a rundown of the technology and how it’s being used.

Communication

What’s so “smart” about a smart grid? In short, it’s the ability to communicate. While many devices on the grid act autonomously, they also need to share information with each other and with a central grid controller. The On-Ramp Total Reach Network offers one communication platform suitable for the smart grid. It’s a low-bandwidth, high-capacity wide-area network that enables communication between sensors, distribution automation devices, utility operators, smart meters, and smart appliances. Operating on a 2.4GHz ISM band, which doesn’t require a license, the network provides 128-bit encryption for data security. It’s capable of reaching underground devices for grids with buried cables. On the consumer side, On-Ramp devices communicate with smart appliances and home automation equipment via the ZigBee wireless specification.

Hardware

In addition to the communication hardware, the self-healing grid also includes the following gear. (Much of this equipment is already on the “standard” grid, except that these devices include communication and additional remote-operation capabilities.)

  • Fault Circuit Indicators – These monitor the power lines, constantly sensing voltage, current, and fault conditions. If things are working properly, they report their information every 24 hours. If they detect an existing or impending fault, they relay that information immediately. In the case of a power failure, they’ll identify the location so technicians can isolate the fault, reroute power around it, and fix it quickly.

  • Transformer Monitors – About one for every five houses, these sensors monitor transformer performance, looking for signs of potential failure. Like fault circuit indicators, they report once a day unless there’s something wrong.

  • Voltage regulators with power-factor measurement – Voltage regulators make automatic corrections for low-voltage conditions, but they also report the condition to the utility in case it’s a sign of a larger problem. If the power factor becomes too large, they’ll signal an operator who can remotely make corrections using integrated capacitor banks. (Automating that process should be in the next generation of self-healing grid devices.)

  • Reclosers – On the standard grid, these circuit breakers will open when they detect a short circuit (say, if a tree branch knocks a wire to the ground.) They’ll also try to reconnect automatically by waiting a certain amount of time, reconnecting, and checking to see if the short is still there. The smart grid version of a recloser will also report the failure over the wireless network and await further instructions. In certain situations, such as fire season in the southwest, operators don’t want the reclosers to reconnect automatically, so they’ll send a signal telling the recloser exactly when it’s safe to reconnect.

  • Grid-level Storage – Batteries are strategically placed throughout the community and at various substations, providing short bursts of power when needed. This makes brown-outs much less likely.

  • Distribution Automation – This equipment automatically reroutes power based on varying loads, generation, and fault conditions. Again, these exist on the standard grid, but with the ability to communicate with other devices, they can make more intelligent decisions. This is also one key to incorporating renewables onto the grid. (More on that later.)

Software Analytics

Fault-detection is all well and good, but why wait until it fails when you can predict a potential failure and fix it before it actually fails? Analytic software looks for patterns that might indicate impending faults or inefficiencies, and reports that information to a central operator. (In the future they expect the ability to make real-time corrections automatically.)

In addition to that, the software can keep track of how much renewable power is going onto the grid, make predictions about available renewable power based on past history and weather forecasts, and determine if a net-metering customer is putting power onto the grid during a power failure (AKA “islanding”).

For the Consumer

With the current grid, the utility often doesn’t know about a power failure until a customer reports it. The smart grid obviously improves that situation by automatically reporting the fault and its location, allowing the utility to make repairs much more quickly. But on the day-to-day side, the smart grid also benefits the consumer by working with the smart meter and smart appliances. In places where rates vary based on demand, the grid can coordinate the powering of appliances so that the refrigerator, dishwasher, pool pump, and air-conditioner don’t all come on at the same time and don’t run during peak-demand (i.e. high-cost) hours. It can also work with ZigBee-based home automation equipment to turn lights off when nobody is in the room, etc. Hotels are taking advantage of this ability – if the occupants leave the AC on but are gone all day, the system will turn it off until they return.

About Those Microgrids…

Earlier I alluded to the grid being built out of separate microgrids. Well, sometimes an idea comes full circle. Microgrids – stand-alone grids that may or may not be connected to the main grid – are popping up all over the place. Hospitals, military bases, college campuses, and other organizations are building their own microgrids for a variety of reasons. I’ll discuss that in a future article. Stay tuned…