km-tech: Forecasting tool for energy efficiency

V. S. K. Murthy Balijepalli, Research Scholar, IIT Bombay, has developed a tool that will lead to more efficient management of the energy sector in India. A Bachelor of Technology graduate in electrical engineering, he was concerned by the heavy power distribution losses that afflict our country and started toying with the ways to better the situation. And in the process, he decided to do his PhD in the area of Smart Grid. During that time, there was no research being done on Smart Grids and when he presented his idea before the PhD committee, he was met with skepticism, as they questioned its functionality. This did not deter him and he went ahead to publish his paper titled “Towards Indian Smart Grids”. Along the way, he interacted with many industry people and gained some insights, all of which helped him when he developed the km-tech – a load, price and grid frequency forecasting tool.     
 
Balijepalli believes that an efficient power forecasting tool is extremely essential in today’s scenario when our country is facing a  power crisis of the worst kind. And with the gap between demand and supply of power continuing to widen, the situation looks bleak. What it requires are drastic steps. Till the time that happens, an innovation like km-tech can be relied upon to ease the situation, as it will help the players in the power sector better manage their power needs and also save them some money in the process. Explaining the issues with the existing forecasting tools, he says, “The existing forecasting technologies face the drawback of using a large historian data before being able to predict to standard practical accuracy levels. Moreover, most of the existing technologies fail to capture the dynamics of even moderately irregular variation patterns of the variables, which are to be estimated. In order to attain these attributes some of these technologies tend to lose the golden property of ‘Simplicity’. Extendibility to the future electricity grid environment and applicability to a wide variety of forecasting problems are the other major drawbacks.” 

Balijepalli’s efforts have been recognised globally and have earned him several prestigious awards

What km-tech essentially does is that it sits on the existing forecasting softwares and improves their performance, leading to efficiency be – it price forecast, load forecast or even grid frequency.

More from News & Analysis

What is the US HIRE Bill and why is India’s $250-billion IT sector worried? Is the internet dead? What's this theory that OpenAI's Sam Altman says might be true?

Speaking about its functionality, he says, “It has a sophisticated propriety mathematical function that reads the behavior of the existing forecasting tool, in addition to a Fibonacci module that has been developed at IIT Bombay. Using the data characteristics matrix defined for any existing forecasting system it tries to segregate within km-tech and the different mathematical modules based on the locales, to collate data that will help improve the performance of the existing forecasting system. So basically km-tech is a core engine, you can use it for price forecasting, load forecasting, wind forecasting and even grid frequency. It also takes into account the practical factors that may affect the price or the load frequency and these can be defined by the operator.” In simpler terms, km-tech brings out the new insights and better forecasts using the existing data.

Balijepalli’s efforts have been recognised globally and have earned him several prestigious awards including the MIT Young Innovator Award, the Department of Science and Technology (DST)- Lockheed Martin Gold medal and even Gandhian Technology Edge award from SRISTI, National Innovation Foundation. The tech has been successfully tested, “The patent on this technology got commercialized with Kalki technologies and it was also used in IEX (Indian Energy Exchange). With IEX we tested the load generation, it studied the Indian regional load volume and the volume of how much to bid. We also provided additional parameters like maximum hourly load, etc. and analyzed it. The results showed that the potential electricity saving ranged between 14.45 and 35.64 per cent daily. A second study conducted at an energy distribution company in Mumbai, revealed that they can save around Rs. 1.046 billion a year for 1,000 MW capacity with the use of the price forecasting tool,” explains Balijepalli.

He even started DesiSmartGrid portal to spread awareness about smart grids and their activities

 
This forecasting tool can play a major role in the Smart Grid projects. In fact, it can play a crucial role in avoiding situations like the Northern Power Grid collapse that we recently witnessed. He says, “If you study the cause of the recent power grid failure, then you will realize that it was due to lack of proper forecasting tools and flaws in the demand response program (termed as Availability based tariff- Unscheduled interchanges). Smart grids will address these issues effectively through end consumer empowerment. Currently, the National load dispatch center does the scheduling, but doesn’t have any forecasting software in place. Mostly forecasting is done at the state level and they submit these forecasted values to the Regional Load Dispatch Centers, but they never stick to those values and they cannot be relied on. If km-tech is employed to study the behavior at the Distribution Transformer (DT) level and at the Transmission Bus level, then this will lead in efficient grid management that can be achieved by coordinating the dynamic grid load changes. Further, the decision modules can be embedded to the developed technology to take the corrective actions in real-time.” He has also started a web portal DesiSmartGrid to focus primarily on spreading the awareness about smart grid and its activities in India.