Intelligent Azure Data Management Platform - enables Renewable Energy forecasting of Power generation

EnergyCo is India’s leading renewable energy companies and one of the top five Wind Turbine Generator (WTG) manufacturers in the world. Its services span across the entire life of wind energy projects and has provided solutions of installed capacity of more than 17,000 MW worldwide. It operates from 18 countries across 6 continents with a support network of around 8500 employees.

EnergyCo belongs to the Global Energy Industry. The industry generated a massive amount of data every day. This data was meant to speed up decision making related to power generation and asset behavior. But since it came in a variety of sizes and forms, it provided less value. It did not help in joining all dots and form a clear picture.

There were opportunities for better service levels to investors and customers. But a lack of a unified system to collect, store and manage data hampered the vision of working toward a sustainable social, economic and ecological development.

Here’s how this affected EnergyCo-

Time & Effort of Decision Making

EnergyCo engineers faced difficulties while locating the problem source in its assets i.e. Wind Turbine Generators (WTGs). This was mainly due to the disparate data sources. There was no unified view of all data sources which offered complete visibility over the asset condition. It increased the time and effort of decision making while combating asset maintenance challenges.

The WTGs were equipped with several systems, sensors and alarms. They monitored the asset condition. There was the SCADA System for wind farm overview & control, turbine overview, log view and report generation. Numerical Weather Prediction Data by NWP and IRIE provided weather data for the next 7 days. Met Mast Data generated regional weather data based on temperature, humidity, wind direction etc. Internally, EnergyCo used SAP ERP System for resource management and File systems for administrative purposes. But there was no unified view of the data and data sources.

This posed a threat while tackling asset maintenance issues. When an alarm went off, the site engineer and repair crew spent time figuring out the source of the problem. It could be any of the WTG components such as the gearbox, brake pad, blade speed. The problem source could also be the impact of weather changes. This cast a shadow on WTG behavior, especially its downtime. It would have meant certainly mean big losses when multiple WTGs on multiple wind farms experienced downtime simultaneously.

Predictive Analytics & Risk Maintenance

Risk maintenance meant early WTG failure detection and notifications. It would give the site engineer a levy to arrange resources and resolve the impending issues. This would ensure less expenditure on asset maintenance and in turn, prolong its efficiency.

With accurate asset behavior predictions, the maintenance cycles could be planned for every WTG. These WTGs were installed with sensors to monitor blade revolutions, resonance, vibrations, surrounding wind direction and velocity. The site engineer could start, stop, pace up, slow down and shut down a WTG, if required. If the wind conditions were to get extreme, it would be ideal to shut down the WTG and limit damages.

EnergyCo had opportunities for practicing ‘Predictive Analytics’ opportunities out there. But it had to focus on threats of ‘Non-Compliance’ reports which suggested high latency in resolving issues.

Forecasting of Power Generation

Forecasting power generation required historical and real-time WTG data in terms of asset lifecycles and weather conditions. But the lack of a unified view of operational data caused and created operational obstacles for the engineers.

EnergyCo used the Met Mast Data and Numerical Weather Prediction Data to predict the weather conditions for the next 7 days. However, this data was not enough to forecast WTG output for a long period of time. It got stored and sourced in an unstructured manner. It resulted into a lack of predictability in the WTG Data. Even though the WTGs were equipped with sensors and alarms, they lacked Machine Learning Algorithms to predict outcomes based on historical and real-time data.

Customer Confidence

EnergyCo faced challenges with forecasting power generation. This meant they could not transmit the required and promised amount of power to the downstream utilities. This meant paying the Utilities huge sums of money in the form of penalties. The amount of penalties depended proportionally to the lack of power transmitted. These penalties would affect the revenue of the organization, customer confidence and imply operational inefficiency.

Generating more power was not the solution to the problem of paying penalties. Surplus power generation meant over use of assets and loss of power.

The Solution: Big Data Management & Business Intelligence