From the past few decades a gamut of sectors like IT, Telecommunication, Finance, Publishing and Media began leveraging the information extracted from various devices, machines, websites, and social media platforms hence opening a door to digital transformation.
The pace of digital advancement was thwarted in the Energy industry owing to government regulations and control. After deregulation and arrival of private players in the market, various pilot projects have been launched in this direction.
The electric power grid is a nexus of generating stations, transmission lines and distribution lines which is responsible for the transfer of electric power from producer to the consumer while keeping the power quality in check. The electric power grid is the largest machine ever created by mankind thus it has enormous potential of digitalization and automation.
Drivers of Data Analytics in the Energy Sector
Power utilities generate large amounts of data as the grid is extensively connected with sensors. In India, most of the data go unanalyzed due to the lack of infrastructure and expertise in the domain of analysis. According to a report published by PWC for the fiscal year 2017-18, 24% of the energy injected in the grid is dissipated as AT&C (Aggregate Technical and Commercial) losses squandering billions of dollars from the GDP.
The ever increasing hunger of the world for electricity and the drifting energy trends towards clean and renewable energy demands a more flexible and efficient system that can only be achieved by making our grid smart through digitalization. The Digital revolution has begun, thus the velocity, variety and volume of the data is rising and will rise up to many folds in the near future. Storage and extraction of valuable information from such data will be an ordeal but if analyzed it can help pacify major technical and non-technical challenges currently faced by power utilities.
IoT and the Data Ocean
Internet of Things refers to the network of devices which enables them to connect to other devices and communicate in order to work more efficiently by optimizing operations and minimizing human errors. IoT devices are usually installed in remote locations and are used to convert physical phenomena to processed electrical signals in order to retrieve information. Advanced analytics is deployed to make sense of the large amounts of data(Big Data) garnered via interaction of sensors on IoT devices.
The IoT technology is so penetrative and pervasive that in the near future each of us will be communicating to thousands of “things” on a daily basis. A forecast from IDC from 2019 estimated that there will be 41.6 billion IoT devices which will generate 79.4 zettabytes (ZB) of data in 2025. IoT will create a huge amount of data and will become the biggest driver of Big Data Analytics in the upcoming years. This data will help organizations to dive deeper into the data ocean of unprecedented sources and extract information to build actionable tactics.
How Machine Learning Fits in the Picture?
Machine Learning is a study of computer algorithms that focuses on programs that learn from experience related to a task with performance as a metric and its performance increases with the experience. In short, we can make machines do repetitive and routine based work more efficiently without putting too much effort into writing a program explicitly for that job.
Along with gaining insights from data using descriptive-analytic techniques we can also parse the data through machine learning algorithms for predictive and prescriptive analytics. Predictive analytics is the field of analytics in which we predict future outcomes using the historic and current data with the help of statistical and machine learning tools. Whereas, prescriptive analytics provides us with recommendations for making decisions at which the machine learning algorithm arrived after multiple simulations. It helps businesses in making better data-driven decisions.
Some Use Cases of Data Analytics in the Energy Sector
- Generation forecasting coupled with optimized procurement of resources for generating stations using plant data analytics. Plant variables like plant capacity, load demand, fuel reserves, fuel GCV (Gross Calorific Value), optimum PLF (Plant Load Factor), CUF (Capacity Utilization Factor), station heat rate, scheduled maintenance, weather conditions can be used for data modeling.
- Resource assessment of a region before setting up solar or wind farms using generation analysis and forecasting with the help of solar and wind data respectively and other linear and nonlinear variables. It helps in finding optimum locations that promise good ROI.
- Automatic scheduling for preventive maintenance or replacement of machinery and equipment on the basis of the number of operating cycles (lifetime) or monitoring the condition of equipment by NDT (Non-Destructive Testing) methods with the help of IoT devices.
- Implementation of smart grid projects for improving grid stability and operability by employing data analytics and machine learning algorithms for Volt-VAr control and optimization, renewables integration, early fault detection, isolation and restoration. Smart grid infrastructure allows more customer participation by providing a graphical representation of real-time usage and real-time pricing (Time of Use metering) of electricity with the help of smart metering systems which also requires data analytics and visualization tools.
- Theft analytics by discoms using machine learning algorithms like Random Forest with Decision Trees on consumer data. Possible suspects are grouped together using clustering techniques, further steps can be taken to pinpoint the defaulters.
- Tracking of man and machine using GIS (Geographic Information System) technology for workforce management and asset monitoring. GIS technology is also finding its way in supply chain monitoring.
- An accurate day-ahead / week-ahead demand and generation forecasting using machine learning algorithms like SVR (Support Vector Regression) on big data for power trading.
Producing electricity has always been a resource exhaustive process taking a toll on the environment. With the integration of new-age technologies, the energy sector has an important role in mitigating the climate crisis. These technologies help in making our energy infrastructure sustainable, environment friendly and empower the customers with information allowing them in making conscious decisions.
In this competitive age, analytics provides a cutting edge in making organizations future-proof.
Data is the gold mine and analytics is the extraction tool.
“The goal is to turn data into information, and information to insight”.
About the author
Vivek Vats is an AI evangelist and a technology for the social good advocate. He deeply believes that the vision of sustainability comes in tandem to Industrial Revolution 4.0. He has previously worked with Power generation and transmission companies and is interested in R&D to transform the face of the Energy sector of our country.