Deepak Pandey still remembers the chaos of the world’s worst electrical outage. At 1:02 p.m. on Tuesday, July 31, 2012, millions of people across India suddenly found themselves waiting for hours on paralyzed trains or stuck in elevators when power went out for half of the country’s 1.3 billion inhabitants across 22 states. Traffic snarled across the capital New Delhi, backup generators ran out of diesel and the water supply became tainted because pumping stations had no backup systems to counteract the unprecedented outage. In the soaring heat — with temperatures touching 100 degrees Fahrenheit and humidity as high as 88 percent — it was “misery,” says Pandey, who lived in a gated community in Delhi and was lucky enough to have access to electricity from a generator. The outage brought the country to a standstill and became known as Blackout Tuesday.
Crippling outages have plagued India for years, but the world’s fourth-fastest-growing economy is now taking a step toward making blackouts a thing of the past. Power Grid Corporation of India, a state-owned grid operator that transmits half the country’s electricity, has begun the first stage of installing the world’s largest wide-area monitoring system (WAMS) from GE Power across its network. The combined suite of hardware and software is designed to detect dangerous power fluctuations and should help put an end to crippling outages. The technology is on display this week in Paris at CIGRE, the world’s largest power transmission conference.
Under the grid’s former system, control-center operators were getting electrical data from the grid every 5 seconds. When it comes to grid protection, 5 seconds is a lifetime. The system made it difficult to respond to problems quickly, or predict where glitches might occur. It is a delayed and “reactive” system, says Pandey, the software services leader for GE Power’s Grid Solutions business in India.
India’s huge, 360-gigawatt national electricity grid is already one of the largest and most complex in the world. While China’s grid is built from separate isolated parts fitting together like a jigsaw puzzle, India’s system is synchronized, meaning that if one region goes down, it can take others with it. It will come under more strain as India likely doubles its energy consumption by 2040. Some of that electricity will come from intermittent renewables like wind and solar, threatening to further destabilize the system.
With GE’s newer WAMS, operators will see 25 samples of data per second flash across visual displays, giving them time and insights to isolate a problem in the massive grid and respond quickly to fluctuations before they spread. “It’s like moving from an X-ray to an MRI,” says Pandey, who’s helping implement the new monitoring system.
The project is rolling out initially in northern India — the most complex section of the grid. It’s where the 2012 outage originated before spreading across 2,000 miles — an area the size of the United Kingdom and Denmark put together. The region includes the sprawling metropolises of Delhi, Punjab and Haryana. Technicians are installing around 1,180 phasor measurement units, each about the size of a microwave oven, across the network to measure minute changes in voltage phase, a nuanced way of detecting how electricity behaves across a grid.
The units will work together to collect voltage data from the grid and send it over a dedicated fiber-optic network to 32 different control centers. In aggregate, the new system will act like a heart monitor to keep track of the grid’s health in real time. “This scale and size are unparalleled in the history of power transmission in India,” says Sunil Wadhwa, who runs GE Power’s Grid Solutions business in South Asia.
But stabilizing India’s electricity network involves more than collecting a mountain of data every day — it needs intelligent software too. Three years ago, the algorithms to parse this level of real-time data didn’t exist. That’s why in 2015 GE assembled an international team of around 50 software developers from Redmond, Washington; Edinburgh, Scotland; and Noida, India, to design software that could analyze India’s enormous electricity demands in real time.
The result was e-terraphasorpoint, an AI-powered system that India’s grid operator is installing at 34 of its control centers. It will track all new data collected by the monitoring units and protect the grid.
The sprawling new system will give grid operators the ability to see problem areas pop up and grow on a map of the network, Pandey says, and teach the monitoring technology to improve over the coming years. GE’s system will have the capacity to store 500 terabytes of data — roughly equivalent to 100,000 4.7-gigabyte DVDs — to analyze why problems might have occurred, and train the network to become even more intelligent over time.
Besides having one of the largest grids, India also has the most residents in the world without reliable access to electricity — some 240 million people, according to the World Bank. But government officials have promised to tackle the challenge head-on: Prime Minister Narendra Modi has pledged to connect all India’s citizens to stable power by 2019, and after the 2012 blackouts, the government invited power-generation experts from around the world to figure out how to prevent more failures. They recommended a wide-area monitoring system like the one now being installed, Pandey says.
The new installation could set a template for other nations to follow. “There are lots of repeat opportunities which are available across national grids in other countries,” says Pandey. “It’s a good reference to apply elsewhere.”