IBM Deep Thunder Takes Precision Energy Analytics by Storm

Subscribe To Download This Insight

By Ryan Martin | 3Q 2016 | IN-4229

Back in 2012, IBM and the Vermont Electric Power Company (VELCO) entered into an agreement to build an advanced fiber network to connect transmission substations to the state’s distribution utilities. Following this came a two-year joint development project using coupled data models and related software to increase grid reliability, lower weather event-related OPEX costs, and optimize the utilization of renewable energy generation resource. Today, this system provides both the communications infrastructure to relay usage as well as equipment status information back to utilities and the analytical backend to improve grid resilience and storm response. It also has the support of The Weather Company’s B2B, mobile, and cloud-based assets (proprietary data) acquired by IBM (proprietary technology) late last year.

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.

 

The Smart Grid Resource Mix is Changing

NEWS


Back in 2012, IBM and the Vermont Electric Power Company (VELCO) entered into an agreement to build an advanced fiber network to connect transmission substations to the state’s distribution utilities. Following this came a two-year joint development project using coupled data models and related software to increase grid reliability, lower weather event-related OPEX costs, and optimize the utilization of renewable energy generation resource. Today, this system provides both the communications infrastructure to relay usage as well as equipment status information back to utilities and the analytical backend to improve grid resilience and storm response. It also has the support of The Weather Company’s B2B, mobile, and cloud-based assets (proprietary data) acquired by IBM (proprietary technology) late last year.

Finding Parity in the Mix of Produced vs. Purchased Power

IMPACT


Vermont Electric Power Company (VELCO) is the transmission entity that monitors, manages, and moves large amounts of power around and through the state of Vermont. Organizations like VELCO supply this power to distribution utilities, which then pipe it to their residential and business customers.

ISO-NE is the larger body that overseas all of the transmission companies in New England – Convex (CT), Maine Power, (ME), NGRID (MA, NH, RI), PSNH (NH), REMVEC (MA, RI, NH), VELCO (VT) – to ensure enough power is generated to serve the load (the amount of electricity used by the population of New England) at any given point in time. When power generation doesn’t match load, very bad things can happen (system blackouts, outages) – and it’s a constant battle to get this right.

Providers like ISO-NE are charged with predicting future load requirements – generally for the next day – to determine the amount of energy that needs to be produced or purchased to meet demand (e.g., VELCO produces a lot from Natural Gas and Nuclear; it also purchases power from Hydro Quebec). The challenge with renewable energy sources such as wind and solar is that it can be extremely difficult to predict how much power they’ll produce on a given day, and this makes it harder to accurately match align generation and load.

IBM’s goal is to alleviate service issues that spawn from an imbalance in generation/load by allowing these kinds of companies to more accurately forecast the amount of wind and solar they’ll be able to produce during each hour of the following day. VELCO views the use of such tools as critical to the continued build-out of its renewable energy portfolio; the subsequent risk is that transmission utilities, in general, could be disintermediated by the same last-mile providers they currently serve.

Here's What We Think

COMMENTARY


There can be a strong case to push parts of the analytics workflow from the cloud to the edge in some of the more distributed and mobile implementations. Other times, such as in the fixed industrial space, the more tangible near-term opportunity starts with historical or systems log data, which companies like Glassbeam, Seeq, and Sight Machine are tapping into to extract insights – and therefore value – from the information that’s already available. Good things can happen when these worlds collide.

IBM Deep Thunder is a prime example; it combines the short-term forecasts developed by IBM Research (used by VELCO) with The Weather Company’s global forecast model, using historical weather data to train machine learning models to predict the impact of future events on business operations.

The Weather Company uses 3 billion weather forecast reference points from weather sensors and aircraft, as well as millions of smartphones, buildings, and moving vehicles to collect real-time weather data to produce an average of 15+ billion forecasts each day. Flexing this data gives IBM and its partners a competitive advantage commensurate with the proprietary nature of the data they collect. The benefit to energy customers is that IBM Deep Thunder caters to higher degrees of grid automation and therefore the ability to meter and manage consumption at the micro-level.

Services