Digital Operations

Wind industry incorporates AI and ML into its O&M practices

Digitalisation is transforming every industry, from manufacturing to transport, banking and energy. The wind power segment is no exception to the digital revolution. Wind power assets generate large volumes of data pertaining to power generation and other parameters. In most cases, only a small percentage of this data is actually used as the majority of the wind asset operators depend on a few essential indicators for carrying out daily operations. The remaining data may be taken into account in case of a failure if time and costs permit, and if the operators have the required expertise to analyse and understand the data.

At present, most operators and asset owners view digitalisation tools not as  “must have” but as “good to have”. These advanced tools are often sold as add-ons instead of critical components. Moreover, when new wind power projects are set up, the emphasis is on saving capital costs. Incorporating digital tools is considered an added expense in most cases, especially as most new turbine owners expect more advanced and high performing machines. Thus, digitalisation may not be an important consideration for them owing to marginal savings in O&M costs and limited energy yield improvement. That said, digital tools offer various benefits such as efficient O&M, reduced downtime, less repair and replacement, improved asset life, advanced forecasting and energy prediction, making them an attractive proposition.

Digital technologies can create new opportunities for wind asset operators and owners, helping them plan their development strategies. Digitalisation in the wind power space is at a nascent stage, and hence there is a huge untapped potential for the deployment of digital tools for better O&M of wind power plants. Digital transformation in the wind power segment is largely driven by the increasing economic pressures on wind power owners owing to technology advancements and intense competition. Digitalisation can help bring improvements in cost efficiencies.

Enabling predictive analytics

The application of digital tools like advanced software, sensors, artificial intelligence (AI) and machine learning (ML) assists in efficient remote monitoring and analysis for identifying any likely faults before they occur. Thus, O&M becomes more “predictive” rather than “corrective”. Defects can be recognised and rectified before any large damage occurs, thus preventing any significant cost overruns. Predictive maintenance ultimately helps in reducing the long-term operating costs of a project, preventing system faults and prolonging the equipment life.

Predictive analytics becomes even more important when the world is grappling with the impact of a global pandemic, which shows no signs of abating. The second and third waves of the Covid-19 infection have already begun, and many countries are going into lockdown again. In this scenario, predictive O&M can help acquire data, analyse it and take action accordingly, especially since many O&M engineers will now be confined to homes and will not be able to visit project sites. Moreover, wind power projects require huge tracts of land and are often located in remote regions. Thus, predictive O&M is more suitable for these plants as compared to traditional O&M, which requires  dedicated manpower at project sites.

Creating digital twins

Digitalisation can help create a dynamic visual representation of a wind power asset. This virtual representation, called the digital twin, can be used to predict the performance of a physical asset in certain conditions. The real data parameters are used by the digital twin to help carry out the day-to-day O&M of its physical twin by predicting energy generation, identifying faults, improving reliability and defining the maintenance strategy. This virtual asset receives data through sensors, and uses AI- and ML-based advanced digital tools to carry out predictive and prescriptive analytics. Through the creation of a digital twin, O&M engineers can identify faults, and predict when and where they are going to occur with great accuracy. This can greatly help in saving time and money, as operators can decide for how long the turbines can work without maintenance and when will the scheduled maintenance activities yield maximum benefit.

Digital twins can also help bridge the gap between pre-sales generation claims and after-sales operational outcomes as they provide insights into the post-installation performance of turbines even before they are put to use. Further, a digital twin can help predict the remaining life of a wind power turbine and its performance over its lifetime.

Operators and engineers can vary data parameters like wind speeds and create various simulations to assess the turbine’s performance in each of these virtual conditions. Moreover, if planned well in advance, digital twins can help generate higher revenues in the case of time-of-day tariffs. Further, in disaster-prone areas or places with damaging wind speeds, digital twins can help predict any health and safety calamity on the costly equipment and to human life.

Facilitating virtual power plants

Wind power is infirm by nature and it is important to supplement it with balancing systems such as battery energy storage, pumped hydro, gas power plants and other stable power sources. Various such small- and medium-scale power generation units and balancing systems can come together to create a large virtual power plant (VPP). This VPP can form a single centralised system through high quality data exchanges between wind assets and the surrounding power systems, thus maintaining grid stability and reliability. Another function of a VPP is that it can link smaller power producers with larger assets. They can then supply power together in energy markets or even bid as a single entity, especially with the current thrust on hybridisation and round-the-clock power. It can create a win-win situation for all and is possible through the integration of advanced digital tools in power systems.

The way forward

In most cases, wind equipment manufacturers are also wind power asset operators, hired by IPPs to carry out O&M of wind turbines. Thus, these manufacturers are responsible for ensuring that their equipment delivers as promised. Many leading global turbine manufacturers have integrated highly advanced digital techniques in their day-to-day O&M practices. With their smart remote sensing and AI-enabled analytics, these manufacturers can assess the health of their assets from centralised data and analysis hubs.

For instance, GE’s cloud-based Predix platform allows wind farm operators to connect, monitor, predict and optimise unit and site performance. It creates a digital doppelganger to identify faults through data anomalies and the asset performance management software then assists in condition-based maintenance, using AI to execute repairs before an asset breaks down. Likewise, Siemens Gamesa’s Pythia platform enables early prediction of potential damages, digital twins and risk-based planning. Vestas acquired Utopus Insights, an energy analytics and digital solutions company, to assist in portfolio-wide asset visualisation, predictive maintenance and wind power forecasting.

Like manufacturers, a few large asset owners have started integrating some level of digital solutions to assess the performance and expected revenues of their projects. However, the uptake is very small when compared to the immense potential of digitalisation in the wind power space. Going forward, digitalisation can indeed become an enabler for improving wind power project performance and generating the required returns.


Enter your email address