Reliability Quest Pushes Utilities Beyond Storm Response
By Rod Walton, Senior Editor
Who could blame operators who feel overwhelmed in dealing with outage management. Storm damage conjures up the same old dread, whether from wind, rain or wildfire.
And, on top of that, a fourth horseman rides in the form of bi-directional distributed energy which complicates things for workers trying to figure which side of the circuit is safe.
As if that wasn't enough, hackers both near and far never seem to sleep, always probing for a weakness in the software and hardware keeping the grid alive.
Meanwhile let us count the ways that customers want to know about the when, why and how long of their disrupted power. Communication is key, whether by computer, cell phone or even door to door if it's local enough. And they don't want to just passively wait for the outage alert; they want a two-way route to ask questions and get specifics.
Weather, cybersecurity, customer service. Yada yada yada. It's a wonder that utility pros don't blow a fuse with all of this.
The customer is always first, so let us begin there.
"The flow of information is just as important to the customer as the restoration is; in some instances it is more important," said Scott Thomsen, senior strategic advisor for communications at Seattle City Light. "Research by Chartwell shows that the cause of an outage is becoming more important to a customer. It appears to be a case where people want the context, which might help them decide how good an estimate for restoration might be."
Responding promptly and predicting the future are now part of the equation that utilities must solve during every emergency. And money is part of math, with Navigant Research predicting that the global OMS market will grow from about $929 million in 2014 to $1.3 billion in 2023 (Figure I). Utilities everywhere are moving fast to install smart meters, sensors, switches and fault indicators to give them the eyes and ears they need from the grid's edge to the substation.
Seattle City Light-which serves about 415,000 customers- is among a handful of utilities in the U.S. which share their outage data with other utilities, vendors and emergency departments at what used to called hyper-drive speeds. Zac Canders, co-founder and CEO of software-as-a-service (SaaS) provider DataCapable, which is working with Seattle City Light on many of these outage communication challenges, is wagering that mega-speed sharing will revolutionize operations and maintenance savings at utilities.
"In the future, if there was a hack or a very large storm, every utility will understand the impact and be able to support (with crews, messaging, supplies, etc.)," Canders said. "We have a hunch you can bypass network-based management systems via a natural language processing of text to identify when/where/why an event has occurred."
New-age, de-centralized outage reporting. Sounds pretty far out, right? But Seattle City Light is hardly in the minority of utilities investing in not only hardening their systems but also embedding those lines with all kinds of systems which can communicate across the interconnections and maybe even predict where the fault will happen.
Switching Schemes and Fine Machines
Southern California Edison (SCE) is one of the biggest, providing power to about 15 million people in the Los Angeles and nearby regions. SCE installed about 5 million smart meters from 2009 to 2012 and now is seeking regulatory approval for a massive grid modernization plan that would install remote fault indicators, intelligence switches and other sensor technology all along the utility's massive distribution system.
Jeff Gooding, who is heading up that grid modernization effort for SCE, pointed out that the smart meter rollout covered one end of the outage management needs, while substation automation is helping smooth things on the other side. The next phase is the in-between stuff of about 95,000 miles of distribution lines and almost 5,000 circuits.
Southern California deals with a lot of wind and wildfire threats which damage the grid components. But another huge complication for the OMS, Gooding noted, is all of the rooftop solar installed at homes and businesses, as well as the energy storage scaling up due to regulatory edicts.
"The distribution network was designed for one-way power flow, using a bulk generator connected to the transmission grid," he said. "Because the landscape is changing with solar panels, electric vehicles and battery storage...at times on the grid we're seeing reverse power flow, making reliability more difficult."
|FIGURE 1: Outage Management System Spending by Region, World Markets: 2014-2023|
Grid operations teams have switching schemes in mind for particular areas based on what they know about the bulk generation and transmission as well as where circuits are located to control electricity coming from one way. The struggle gets real fast when it also arrives from behind meters and who knows how many locations.
"Sometimes we work hot (with power on) but most of the time we take it out of service if we can," Gooding said. "With reverse power flow, that changes the switching scheme and operators don't see what they're expecting to see."
Placing scores of remote fault indicators-softball-size devices which can go on the end of hot sticks or right on the wire to confirm the breaks-and remote intelligence switches can help open switches and isolate outages automatically-will help answer of those unknowns involving distributed energy resources. SCE installed about 1,000 remote fault indicators in its service area last year (2016).
The plan is to install more all along the 90,400 miles of SCE distribution line, replacing more than 14,000 mechanical fault indicators. The sensitivity could get so good that hyper-intuitive sensors perhaps will detect falling power lines and cut power even before the cable hits the ground.
"It's aspirational," Gooding admitted. "But we believe it's possible once we get all the grid modernization out there."
Bradley Williams, vice president of industry strategy at Oracle Utilities, said the industry to doing better protecting against the two-way danger of distributed energy with the IEEE 1547 standard. IEEE 1547 requires that the interconnection with a distributed generation source trips out when the outlying grid is down, thus protecting workers from unexpected current. However, utilities still struggle with grasping the level of generation out there behind the meter. Their customer-facing side may know where all the rooftop solar is, but the operations side doesn't always know what the level of bi-directional electricity is.
Another component of outage management that is only slightly less high-tech but certainly just as important is customer communication. Customers want to know that an outage is an outage and not just their breaker tripped. They want to know how it happened and how long it will endure. Finally, they want to be alerted in multiple ways on various devices.
Sometimes they may act like they want to kill the messenger.
"A few utilities that have been on the forefront of doing more frequent and proactive outage communications have had some challenges and gotten a bit of bad press for not getting things right when they rolled out their new programs," Terry Nielsen, executive vice president at outage management specialist GridBright, pointed out. "But, in the end, even with the temporary setbacks I haven't seen anyone pull the plug and retreat to doing outage restoration behind closed doors."
Determining outage extent and estimated restoration times dictates how early and detailed that utilities can be in their communications with customers. Jumping out too early can result in inaccurate estimates of an outage's scope and also notify people who are not even affected by it, Nielsen noted. Waiting until it is fully confirmed by an onsite crew can be too late in a customer's mind.
Damage assessment teams which are not responsible for outage repair but patrol and record downed wires and broken poles can certainly help pinpoint the often bedeviling details.
"This can dramatically improve the ability estimate the amount of work required to do restoration and therefore improve outage restoration estimates long before a crew is able to make it to the actual repair," Nielsen said.
OMS Shouldn't Take Back Seat to Renewables
Williams, of Oracle Utilities, has argued that investment in the customer engagement for outage response could be much stronger. He also stressed that utilities should be prepared more on the field side for the big storms as well as the typical, more limited outages from downed lines and blown transformers.
"As utilities move to combined OMS and DMS storm-proven system performance becomes even more important," Williams said. "I am not hearing the same emphasis on systems performance like we used to. We have not had a huge storm like Hurricane Sandy in years, yet that is what these systems must support."
Certain utilities, such as those on the east coast blasted by Hurricane Sandy five years ago, have invested heavily in system hardening. However, elsewhere the focus has been on renewable energy resources and advanced distribution management systems, he contended.
"That is very important, and we at Oracle certainly support that," Williams added. "But because of that, the need to invest in scalable outage management restoration has taken a back seat."
Several years ago, Williams wrote an article about outage management that included a prospective report card in which utilities could judge themselves (Figure 2). He still contends that utilities need to be their own harshest critics, ever observant about the threats to their pieces of the grid.
He also would like to see utility planners incorporate IEEE 1457 and distributed generation sources to aid the grid during outages. Why not put schemes in place which could incorporate the still producing solar PV and help keep the lights on in sections otherwise shut down because of an outage? Plus, manufacturers of things like air conditioners-which when trying to restart stress the load while grids are trying to get back up and running-could incorporate delays that spread the load out over a longer time and avoid new faults.
All in all, several sources agreed that system average interruption duration index (SAIDI) and frequency index statistics have improved. Smart meters are finding their second level of value in helping read the grid better and sensoring technologies are locating the faults quicker.
"I think we're headed the right direction," SCE's Gooding said. "Intelligent switches and sensors are the key."