How IoT data analytics is remaking the energy sector

Advanced analytics is key to managing the dynamic distribution grid, writes Neil Barry, Senior Director EMEA, SpaceTime.
Published: Wed 27 Jul 2016

Energy companies are pioneers in working with multiple, large data sets and real-time challenges, so the current buzz about the Internet of Things (IoT) doesn’t sound as new and loud in the energy sector compared to other industries. That doesn’t mean that utilities are immune to the IoT trend. IoT enables the connection of multiple new physical devices to the power grid and to the data networks that support the power grid. Rooftop solar, electric cars, home energy batteries, smart meters, smart thermostats and smart appliances all change the local distribution grid into a dynamic, bi-directional, and multi-party marketplace for energy, rather than the old one-way system of energy delivery.

These new connected devices can cause chaos on distribution grids that were never designed to handle these new dynamics, yet customers still expect energy service that is safe, reliable, affordable and increasingly, sustainable.

Advanced analytics on the grid

All those connected devices generate volumes of data. Applying that data to decisions about the grid could help tame the chaos, but the amount and variety of data is too much for people to analyse manually to inform their decisions. This makes applying advanced analytics to that data a key capability in establishing a new order on the grid. Advanced analytics allows people to correlate, understand and act on data from multiple, disparate sources of data generated by connected devices.

Exactly what do we mean by advanced analytics? It is powerful algorithms, based on statistics and modelling, that carry out fundamental and useful actions with data such as:

  • Detecting anomalies

  • Grouping data about people, things, and events

  • Determining the root cause of behaviour or phenomena

  • Performing what-if simulations

  • Analysing large volumes of text

  • Determining when equipment or systems will fail

  • Identifying the factors leading to the best outcome

  • Forecasting future values.

These are powerful actions to have computers perform to help tame the chaos happening on the distribution grid. These actions are even more powerful when used in combination, such as finding the root cause of past equipment failures, then forecasting values to identify possible future failures.

Advanced analytics impacts on utilities

The ability of computers to perform advanced analytics on data from connected devices is starting to remake much of business as we know it. The remaking of business takes one of three forms:

  • Optimising existing operations

  • Enabling new products and services

  • Transforming business models.

What does this mean in concrete terms? How will these impacts change energy companies? Here are some examples.

Vegetation management is often the top operational expense for an electric utility. Vegetation management is crucial for preventing power outages. Vegetation interacting with powerlines and substations is the main cause of outages. IoT devices such as drones equipped with video and LiDAR, connected weather stations and soil sensors provide valuable data about vegetation. Analytics uses that data, along with the location and criticality of distribution infrastructure correlated with rainfall trends, short-term weather forecasts and vegetation types, to build what-if simulations of where and when crews trim or remove vegetation to make the most of operational budgets and lower the risk of outages.

Demand response and variable rates are new products that utilities offer to customers to shape demand for power. By giving customers signals and incentives to use less energy, utilities fit consumption to the available supply, rather than increasing supply to match unconstrained consumption. IoT devices such as smart appliances, thermostats, meters, and load control switches both supply data and accept commands to enable demand response and variable rates. For instance, British Gas now offers customers with smart meters free electricity during the day on weekends.

IoT data analytics

Applying analytics to IoT data enables utilities to forecast future supply, demand and prices; group customers according to behaviour and preferences; and identify what mix of equipment, programmes and customers yields the best reduction in consumption. Those analytics then drive signals back to the IoT devices to take action based on the price signals and demand response programme parameters.

Connected rooftop solar systems combined with smart storage such as home batteries and electric vehicles turns individual customers into power suppliers. Customers can sell surplus power back to the grid and buy extra power to store when rates are low. As customers become more active in the local energy market and more independent of their local utility, the utility must transform its business model. One possible transformation is becoming a distribution service provider that owns and operates the wires and infrastructure to support the local buying and selling of energy. Becoming a distribution service provider requires analytics for forecasting and balancing supply and demand, finding the best combination of buyers and sellers, and detecting fraud to ensure a fair marketplace.

We’re already seeing these impacts on energy companies. Managers are discovering new savings and better results from operations. Marketers are offering a slew of new products. Executives are crunching numbers and placing bets on new business models. Computers and the analytics that they run are now new partners in delivering safe and reliable power that is increasingly affordable and sustainable.