This piece was originally published in the April 2017 issue of electroindustry.
Kerrick Johnson, Vice President of Strategy and Communication, Vermont Electric Power Company (VELCO)
Picture a tool that sees precisely into the near future and enables utilities to better protect communities, meet customer needs, and realize renewable energy’s full value. That was the vision of the Vermont Electric Power Company (VELCO) in creating the Vermont Weather Analytics Center (VWAC).
The ways in which we generate, move, and use power—and how we pay for it—are changing faster than at any time since electric power came of age. Throughout the United States, steadily increasing amounts of solar and wind generation—intermittent power—are plugging into a grid that was designed to rely on stable and predictable baseload resources.
At the same time, public policy in many states is propelling the development of a grid that is more reliant on renewable sources. Customers, too, are driving the transformation with their own desires for greater control and the ability to generate some of their own power but still rely on the grid when they need it.
Severe weather events pose additional threats to the power grid. For example, Vermont has experienced eight federally declared weather disasters since 2011. Extreme weather has increased storm response budgets by 35 percent in three years for Vermont’s largest distribution utility. Similar weather trends create safety and economic threats to communities, as well as utility and public infrastructure.
Adapting the legacy grid to these radically changing demands and conditions will require new tools and new ways of managing the system. In 2014, VELCO began work with IBM, Vermont’s distribution utilities, and other partners to develop VWAC, which integrates IBM’s precise weather forecasting with Vermont’s customer load data and output data from the state’s renewable generators to turn this mass of data into actionable information using leading-edge analytics.
VWAC’s analytics innovation lies in the powerful ability to feed multiple data streams into four integrated forecasting models.
- Weather: The Vermont-specific version of IBM’s weather model produces high-resolution, accurate, and specialized forecasts. Traditional 72-hour forecasts provide information to a 12 km2 resolution, updated every one to three hours; VWAC’s tools provide 72-hour advance forecasts to the 1 km2 level, including wind speed and direction by elevation, solar irradiance, and precipitation for every ten-minute increment. It is, in effect, a four-dimensional forecast.
- Electricity demand + net metering: This model uses smart meters, weather forecasting, and other data sources to analyze customer-owned generation to predict demand and better plan for future system reliability needs.
- Renewable power: This model forecasts solar and wind production and separately correlates hydro to improve power supply planning efficiency.
- Renewable integration stochastic engine (RISE): This coupled model with a probabilistic framework synthesizes the other models’ output for best use of Vermont’s renewable generation, efficiency, demand response, and transmission resources.
Transforming Data into Intelligence
Which data, and in what form, matters. VELCO and IBM explored terabytes of data from telemetry, revenue metering, smart meters, geographical information systems, weather archives, and numerical weather prediction models. Through painstaking, collaborative analysis, we now can secure the right data at the right time in the right format necessary to generate actionable information.
In order to represent relations between input data, the electrical grid, contextual information, and predictive models, VELCO and IBM developed a generic data model, which establishes the foundation for future business application development. This data model is the key for replicating the VWAC breakthrough with other energy companies.
More than a hundred users from VELCO, 17 distribution utilities, five state agencies, and one regional transmission organization (ISO New England, an independent system operator) access the information and insights through a weather and grid portal that includes forecast videos, statistical or interactive mapping, graphs, data tables, or archived data. The portal turns millions of bits of data into a model that gives users unified, structured, and meaningful access to information that enables smarter power supply decisions, disaster preparedness, storm restoration, grid planning, and more.
The development of a big data architecture for the ingestion, cleansing, and exploration of terabytes of smart meter data enables an unprecedented level of analysis and understanding of solar energy in Vermont. We now know the installed capacity, location, size, and output of grid-connected renewable generation. VELCO and IBM’s work also revealed precisely, for the first time, the exponential growth rates of photovoltaics: by the end of 2015, up to 15 percent of Vermont’s energy was supplied by solar systems on sunny days—a percentage that continues to rise.
VWAC has already produced significant benefits. It strengthens emergency response coordination calls and utility crew augmentation decisions and enhances field crew protection by providing targeted wind chill and lightning potential indices. VWAC enables more accurate, site-specific customer updates and has improved traffic safety through an effective partnership with the state transportation agency.
The system has substantially improved outage scheduling and contingency analyses with reliable 72-hour forecasts of expected system conditions. It gives utilities the ability to determine grid capacity for additional solar generation from the transmission system down to the substation level. It also enables demand analysis to the substation level—another first.
Furthermore, VWAC increases planning assessment reliability due to smart meter data integration, significantly improving the analysis of non-transmission alternatives to system upgrades. It refines our ability to assess transmission constraints and storage options and reduces power supply market risk by more accurately assessing supply needs. VWAC improves developer-customer collaboration on solar installations and enables refined comparative generation assessments of prospective solar and wind sites.
Finally, VWAC provides greater visibility to potential demand response events based on demand forecasts, from the substation to the distribution service territory to the statewide level. VWAC enhances the utilities’ ability to manage statewide peak demand, and helps analyze the cost-benefit of energy efficiency programs more precisely.
Value Continues to Grow
Next steps are underway to further strengthen core services and generate even greater value for more partners. Linking VWAC output to VELCO’s energy management system will improve core grid reliability and more accurately integrate weather-dependent generation in it.
Connecting VWAC to Vermont emergency management, environmental, and agriculture agencies helps emergency operations keep communities safer and provides a detailed weather record for federal cost reimbursements. The Agency of Natural Resources can better monitor stream health, manage habitats, and protect field crews, and the Agency of Agriculture can provide farmers with information that will improve their operational efficiency and protect against damaging farm run-off discharges.
A pilot study initiated by ISO New England will share data streams and lessons learned to help this regional grid operator best meet its need to accurately account for rapidly growing renewable generation, especially solar, in the six New England states. Using VWAC to unlock additional value from renewables through documented performance results may support project and policy changes and could save customers millions in future system upgrades tied to load growth.
Reduced uncertainty of production from behind-the-meter (BTM) solar will result in better commitment and dispatch decisions by ISO New England, reducing production costs and avoiding unnecessary supplemental commitments and fuel consumption. Conservatively, savings associated with better commitment estimates (based on current BTM penetration levels) are on the order of $1 million per summer season, growing as BTM penetration increases.
Expanded use of VWAC tools will improve accuracy and reduce uncertainty associated with grid-connected solar and wind resources in the day-ahead market, leading to more efficient commitment and dispatch decisions by the system operator. Applied nationally, the savings would be in the billions.