From huge centralised conventional fuel-based generation systems to decarbonised and localised power generation units, a slow global transition is taking place. As a large number of intermittent renewable energy assets including wind and solar power plants with energy storage are being added to the grid, more efficient systems are required to monitor them. Moreover, greater coordination is required among generation, transmission and distribution. This is where digitalisation comes into the picture. It plays the role of an enabler. Digitalisation improves connectivity between computers and assets. It also enhances cloud computing, data collection and the ability to interpret the collected data using artificial intelligence (AI), predictive machine learning and blockchain.
Digitalisation can facilitate operations and maintenance (O&M) of solar power assets spread over hundreds of acres of land, and significantly reduce the scope for error. Traditional O&M models for power generation plants at a single location with a dedicated permanent team will not be able to cater to the financial and operational demands of future plants. The upcoming large solar capacity, with a high share of distributed solar assets, will require efficient O&M with remote monitoring and quick fault diagnosis and rectification by a rotating skilled team. Moreover, with increasing solar power integration into the grid and reducing solar tariffs, it is becoming imperative to reduce the dependence on manpower for carrying out O&M activities. Thus, the industry is focused on reducing operational costs and improving efficiencies through more digitalisation and automation.
Big data analytics
For proper monitoring of solar projects, it is important to track the current, voltage, power and energy generation along with the performance of the equipment, ranging from solar modules to small cables. In addition, solar resource and weather data also needs to be tracked regularly to accurately forecast energy generation. If done on a real-time basis, monitoring can generate a staggering volume of data from one solar asset. This could be utilised by developers that own similar projects or operators that carry out O&M for many solar assets.
To remove complications, automatic data acquisition can be carried out using embedded sensors integrated with SCADA and web-based monitoring systems. Monitoring systems differ across projects based on plant size, location, distance from O&M facilities, availability and performance ratio guarantees, and financial capability. However, just collecting the right data is not sufficient until it is carefully analysed to improve project output. This requires a connection between data on the digital side and a strong initiative from developers and operators towards performance improvement. Big data analytics with proper AI and machine learning tools are used to analyse this “high volume of high variety data with high velocity”.
AI is based on the simple principle of making a machine mimic the human brain. Today, software and computer applications have become so advanced that a machine can be programmed to learn, adapt and react to the various performance parameters and operational processes of a solar power plant. In fact, various technologies are being used to make these machines even smarter so that they can not only identify faults but also self-learn to predict such breakdowns before they happen. This can significantly improve plant diagnosis and enable the rectification of faults and breakdowns, leading to improved plant performance. Thus, AI applications when applied to solar can accelerate the due diligence proccess, reducing the time invested in planning and analysis.
The use of AI in energy storage is being explored to maintain grid frequency and avoid outages. Maintaining the frequency of a solar power plant at the accepted grid range is difficult due to the intermittent nature of solar. AI can help address this issue. It can be used for accurate forecasting and scheduling of power generation so as to avoid penalties for over-or under generation. AI can help in predicting weather patterns by analysing large real-time weather data sets from multiple sources like satellites, weather stations and other devices, and comparing them with historical weather data. It can also predict how this weather would impact solar production, thus allowing power producers and operators to adjust and schedule power accordingly. Moreover, a replica of an actual physical asset or a “digital twin” can be created through AI, which can be used as a benchmark for identifying faults through data anomalies. This is helpful in O&M as operators are alerted about possible breakdowns, which can be rectified beforehand to improve plant performance.
Intelligent solar operations
Even though AI application in solar plants is at a very nascent stage, there are some examples of AI being put to use in forecasting, monitoring and maintenance activities in solar plants. For instance, NEXTracker has a self-adjusting tracker control system called True Capture. In this system, machine learning software continually improves the tracking algorithms of each module row as per present weather conditions. Virtual 3D models of the plant site are created using real-time shading information, and tracking instructions for each row are updated. The system can increase energy production by 2-6 per cent.
Another example is of GreenPowerMonitor (GPM), which is a part of DNV GL. The company has developed the GPM Horizon software, which is an integrated tool for wind, solar and energy storage. It caters to asset owners looking for integrated asset portfolio management instead of siloed management of their wind or solar or energy storage assets. GPM Horizon assists in the easy and quick visualisation of asset data brought together from a diverse range of technologies and locations to a single location. It is used by engineers and analysts at DNV GL’s control centre to monitor and operate all assets remotely. The company has one such control centre at Bengaluru where GPM Horizon, renewable energy data and various software systems are used by analysts and engineers to monitor the operating assets of the company’s clients.
Krypton provides products to O&M companies for real-time remote monitoring. Krypton Collect gathers data from multiple sources, Krypton Decision Engine detects and correlates anomalies and Krypton Applications prevents or rectifies faults with advanced machine learning tools. Even software giant IBM has a self-learning technology that integrates multiple forecasting models with weather, solar generation and power grid operation data. It predicts solar generation in advance for every 15 minutes to 30 days, and has shown up to 30 per cent improvement in solar forecasting.
Increasing digitalisation and automated remote monitoring are expected to revolutionise the way solar O&M is carried out. In India, which depends on its cheap labour for much of the O&M activities, this can certainly improve project performance to a great degree. At a time when the entire global solar industry has come to a halt due to the coronavirus outbreak, digitalisation has gained even more importance. For instance, if operating solar assets were more digitalised, with drones for detecting faults, advanced sensors for collecting data, robotic systems for cleaning modules and remote automated monitoring systems, then projects could run with very limited manpower despite the lockdown. While under-construction projects would still have suffered, operational ones could have functioned without any severe impact on performance due to the unavailability of skilled manpower.
It is estimated that if the current pace of digitalisation and automation continues, the majority of the O&M activities will be carried out through AI-enabled technologies in the future. AI field assistants will be increasingly used in lieu of manpower for activities like monitoring and inspection, supply chain optimisation, installation and commissioning and identification of faults in materials through microwave and ultrasonic transmitters and receivers. Overall, with AI and advanced predictive analytics, the solar O&M landscape is poised for a digital makeover.