With wind power installations crossing the 500 GW mark globally, wind power has emerged as the cheapest source of power in many parts of the world. In India itself, wind tariffs are hovering around Rs 2.76 per kWh, which was the winning tariff in the last Solar Energy Corporation of India wind power auction. In the face of declining wind power tariffs, which are an outcome of competitive bidding, the wind power industry is scrambling for positive financial gains. This has increased the demand for products and technologies that can improve project performance and lower life cycle costs, thus highlighting the role of efficient operations and maintenance (O&M).
As per industry estimates, O&M contributes to 20-30 per cent of the levellised cost of energy of a wind power project. However, until recently, the O&M of wind power plants did not receive the same importance as the initial stages of project development and execution. The majority of wind farms, spread across acres of land, still depend completely on manpower for O&M-related activities. In developed countries, some degree of automation has been incorporated for carrying out these routine tasks. However, with the increasing size and scale of wind assets, automation has become inevitable as traditional O&M models require a permanent and dedicated O&M team for each wind asset. As such, this existing model will not be able to meet the future O&M demands as manpower costs continue to rise and O&M data gets more complex.
The constant monitoring of wind turbines and related equipment to measure the current, voltage, power and energy generation is essential for ensuring that the plant is performing as required. Automatic monitoring systems not only monitor the plant yield, but also assist in timely diagnosis and rectification of faults, so as to prevent plant downtime. In addition, the analysis of data sets helps in providing energy generation forecasts. As the number and scale of wind assets has increased for each developer, the need for integrated asset portfolio management has surfaced. This requires greater connectivity and much higher levels of automatic monitoring.
Automatic monitoring systems vary across projects depending on project specifications and can be remotely controlled to not just “read” but both “read and write”. While automatic data acquisition is important for O&M, it is impossible to manually track and analyse these huge and complex data sets. This is where big data analytics is needed. Automated monitoring systems integrated with big data analytics can handle huge volumes of a large variety data with high velocity to assess faults and save operating time and expenses.
Drones and robotics
Drones and robots are increasingly being used in wind power plants to reduce the manpower requirement. Flying drones and unmanned aerial vehicles provide greater detail than ground crews, and are the most useful for site assessment and O&M. Studies suggest that using a fully automatic drone without a human pilot can significantly reduce monitoring time and costs. Autonomous drones are also being developed with real-time artificial intelligence (AI) and complex analytics capabilities, which can automatically detect faults in the wind plant and rectify them, resulting in significant manpower and cost savings.
Apart from these, crawling robots and driving robots have been developed. Crawling robots can come close to a structure’s surface and detect faults in materials by using radiation. Driving robots can manage the entire supply chain in a wind power plant right from transporting equipment to unloading and installing it as per the directions.
Role of digitalisation
Digitalisation is not a new concept for the wind power segment, which has moved from computer-based systems, databases and communication networks and servers to internet of things (IoT), cloud-based platforms, advanced analytics, predictive data analytics, asset performance management software, smart sensors and intelligent forecasting solutions. Such digital systems can perform multiple functions in the wind O&M segment, improvement of power plant efficiency, reduction of operational expenses, and decrease in unplanned outages. Digitalisation can help in improving plant efficiencies on a real-time basis, leading to enhanced equipment life.
In AI, a machine mimics the human brain and responds to issues in the same manner as a human. In the O&M of wind power plants, AI can be applied to machines to manage the day-to-day operations as well as the maintenance activities of the plant, with minimal or no human intervention. These smart machines significantly reduce the time required for such activities, saving significant operating costs in the long run.
The greatest advantage is that these machines can be programmed to self-learn and integrate more advanced softwares, thereby helping in achieving higher performance efficiencies. A lot of research is being done in the machine learning domain to develop systems that can self-learn and evolve to such an extent that they are capable of not only addressing present issues but any future faults in wind power plants. Machine learning applications are also being developed for frequency regulation and weather forecasting.
Innovation and research worldwide is centred on developing intelligent systems that can bring digitalisation in the O&M of renewable energy assets. These systems and softwares can enable predictive analytics and energy forecasting, and increase the efficiency of assets. For instance, a DNV GL-owned renewable energy software company, GreenPowerMonitor (GPM), developed the GPM Horizon. This software assists in easy and quick visualisation of all data from multiple locations and diverse technologies to a single monitoring and analytics centre. The data is then used by experts at the control centre for remote O&M, thus saving precious time and manpower costs.
Technology giant IBM is also working on weather forecasting models by assimilating multiple data sources so as to predict weather conditions and the expected renewable energy generation based on the weather forecast. Another example is that of Predix, a software solution developed by General Electric. Predix analyses sensor data to predict failures and maximises the efficiency of an asset. Such intelligent digital applications aim to create a “digital twin” or replica of the actual asset, which can be used for forecasting and predictive analytics. Different simulations can be carried out on this digital twin, and any data anomalies can then be analysed to predict future faults.
Highly advanced remote monitoring stations and analytics centres are being developed in various parts of the world that incorporate all these technologies. These centres remotely collect data from all the assets in an operator’s portfolio, and undertake O&M. These centres are equipped with such advanced digital tools that they can not only predict weather and accordingly forecast energy generation, but also estimate faults and respond to price signals like peak pricing based on the grid requirements.
Traditionally, original equipment manufacturers (OEMs) of wind turbines have been responsible for the O&M of wind power plants. However, with shrinking margins due to high competition, developers and owners of large project portfolios are now resorting to self O&M. Going forward, OEMs will have to reduce the dependence on manpower and automate their O&M procedures to stay in the game or risk losing significant O&M business. Similarly, developers that own multiple assets also need a high degree of automation and digitalisation to manage all their assets with limited physical interference.
This is where the transformative application of AI and machine learning is needed. Digital technologies in the form of AI field assistants and forecasting technologies have already been introduced in the industry to save O&M costs. Even though the initial costs of deploying these technologies are high, the long-term project life cycle gains compensate for the high capital cost. With the current pace of innovation, it is expected that automation and digitalisation will change the face of wind power O&M in a few years.
By Khushboo Goyal