Skip to main content

Analytics experts converge in Orlando

By November 1, 2018No Comments

There’s little doubt that today’s utilities have entered the age of Big Data — and Big Analytics.

From customer experience ratings to SAIDI and CAIDI numbers, from equipment operations data to weather metrics, most if not all utilities are swimming in numbers and charting a course for the future based on how they can best predict future business cases with those numbers and that data.

That was the overarching theme at Utility Analytics Week in Orlando last week, where over 500 utility analytics staff attendees from 38 different countries heard from a wide range of speakers on multiple applied analytics topics.

The conference opened, fittingly enough, with Fred Hazelton, a founder of expounding on how data and analytics drive the customer experience at nearby Disney World and other Orlando theme parks. In his presentation, Hazelton offered that by recording, calculating and analyzing wait times at various attractions, his company can and does present “touring plans” — essentially annotated maps — that can save between 2.5 and 3.0 hours worth of waiting in line at the park by visiting attractions in the order proscribed by the plan. looks at numerous variables to develop its recommendations, Hazelton added, with holiday schedules, day of week, seasonal time of year and hour of the day being among the most important factors affecting Disney World line durations. Somewhat surprisingly, though, other factors that come into play include such things as opening and closing hours at neighboring parks, media mentions or “buzz” created by new attractions and even the price of gas six months prior to the date (the reasoning being that some plan their trips six months out, and lower gas prices mean lower fuel prices overall, leading to lower airfares).

Utilities, Hazelton implied, can use the same type of data to develop their own “touring plans” (perhaps better said — business plans) for their service territories. The key is to collect the data you can (e.g. outage or service call locations, expected hours/duration for work to be performed, etc.) and then map it out for efficiency. Perhaps even better, track and analyze storm restoration, for instance, to inform future storm restoration plans.

EVs on the Rise

Peter King, Electric Transportation Project Manager for Duke Energy Florida, talked about new data and analytics surrounding electric vehicle (EV) growth nationwide.

Among the numbers coming in, King says, are that more than 1 million EVs now are on the road, showing an annual growth rate of 30 percent. The Tesla Model 3 is now the fifth best-selling passenger car in the US, he said. EV battery life is changing too; more EVs now hold larger batteries, many in the 200+ mile range, at sizes of 60 kWh to 100 kWh. Over 50 different models of EV batteries are expected to be on the market within the next year or two; all of this, King said, informs what he calls the coming “Power Gap,” where utilities will need to address all the new electric vehicles in the market and the need for expanded charging stations nationwide.

Heavy duty EVs are also coming to market, King added, including forays by at least Tesla, Daimler and Volvo into delivery trucks, buses and even some off-road vehicles. In the same presentation, Kent Mathis of the Jacksonville Electric Authority talked about the potential for EV bucket trucks for utilities.

Mathis and King both referred to the need for utilities to create and execute plans for EV charging infrastructure, and King talked about Duke Energy Florida’s new EV Infrastructure Pilot Agreement with the Florida Public Service Commission, an $8 million study of the state’s electric transportation network that is intended to go through 2022. The pilot program will study current market penetration of EVs as well as existing charging networks along with projecting for future super-charger locations and track both privately-owned EVs as well as fleet vehicles as they are developed.

King and Mathis also talked about how the rollout of the EV charging network of the future might need to adapt to how, where and when EV drivers are recharging — be it at public charging stations, at home or at work. Presently, a large percentage of EV drivers recharge their vehicles at home overnight, but as Mathis pointed out, that may not be the case in the future, particularly as EV growth develops and workplace charging stations become more prevalent.

Analytics for System Optimization

Another key use of utility analytics — real-time system optimization — was laid out by Phu Pham of Southern California Edison (SCE) and Carol Bogacz of Burns & McDonnell.

Pham and Bogacz described a centralized control algorithm SCE has been using to record and track real-time telemeter data from capacitors in the field. SCE’s Distribution Volt-Var Control (DVCC) Analytics program has been in use for about a year now, deployed on 40 percent of SCE’s distribution system, with some 2500 feeders covered. System-wide deployment is anticipated by 2024.

The DVCC program investigates all SCADA discrepancies between systems, Phu pointed out, and maintains settings as circuit changes occur, triggering a maintenance process that reevaluates settings and updating systems where circuit changes occur. Program benefits include energy reduction and the optimized use of substation and field capacitors, including maintaining voltage within 5 percent of nominal, per the California Public Utilities Commission Rule 2 voltage compliance.

The system includes a dashboard that interprets and displays results for use by system engineers as a troubleshooting tool but also for prioritizing maintenance. Phu and Bogacz also noted that SCE and Burns & McDonnell have developed DVCC algorithms to identify key trends, correlations and abnormal behavior on the system that can be used to inform, plan and execute future network roll-outs. Over time, the goal is to expand the DVVC program to cover some 26,000 devices on SCE’s network.

In a subsequent presentation on the use of weather data to both accelerate rapid response to storm outages and prepare for inclement weather prior to a storm strike, Gaetan Benoit, Manager of T&D Control Room Operations of NB Power (New Brunswick, Canada), noted how careful analysis of historical weather data coupled with near-term forecast data has been used at NB Power to line up emergency restoration crews days in advance of a storm strike, a particularly valuable tool around and near holiday periods, when contract restoration crews can be harder to locate and schedule.

Benoit offered examples of data and analytics helping NB Power quickly mobilize for oncoming storms as well as being able to call off additional crews when weather data shows that a storm is failing to materialize as originally forecast. Either or both uses of the data, he said, proved to be money-savers for the utility.

Summing up the opportunity to use both weather data and other analytics in a way that benefits utilities both large and small, Rob Berglund, head of the Energy & Utilities Business Group at The Weather Company/IBM, referred to a slogan The Weather Company likes to use as to the value of data.

Citing the company’s thousands of always-on weather reporting stations, Berglund said, “we use data and analytics to monetize through accuracy.”