By Khushboo Goyal
Ensuring optimum solar plant performance and quality of assets is perhaps the most critical consideration for a developer today. At a time when solar tariffs have hit rock bottom at sub- Rs 2.50 per unit, affecting project revenues, developers cannot afford to compromise on energy generation and run the risk of becoming unprofitable. Investors must be kept happy and debt must be repaid, while expanding their project portfolio and increasing the asset base. Thus, in the face of the cut-throat competition that prevails in the Indian solar market, it is important for developers to have skilled operations and maintenance (O&M) crews as well as the most advanced equipment, while keeping costs in check. However, this is easier said than done. The solar power landscape in India has changed radically over the past five years – project sizes have increased four to five times (if not more), the geographical spread has expanded and developers have amassed multi-GW project portfolios. Thus, a single developer now owns double-digit projects that are spread across multiple states and can be of various types, including rooftop, utility scale, and floating solar.
Consequently, traditional O&M models, with a large, dedicated O&M team for each asset, will not work or will become too expensive for developers. Moreover, such a model would not be able to meet future O&M demands, as monitoring data gets more complex. Thus, the focus of both asset owners and O&M teams (in-house and third-party service providers) is now turning towards incorporating more automation and digitalisation in their O&M practices, using drones, robotics, artificial intelligence (AI)-enabled monitoring and analysis, advanced fault diagnostics techniques and intelligent remote monitoring software. Such O&M tools and services not only help in getting the job done quickly but also ensure higher quality and greater reliability. Moreover, they can significantly reduce O&M costs owing to lower manpower requirements, even though a small initial capital investment may be required to deploy such advanced tools.
Apart from the direct cost saving from reduced manpower requirements, such automated and digital technologies have an indirect benefit. The high precision and quality of delivered services – module cleaning, inspection, data collection, analysis or even fault diagnostics – mean overall better O&M with more energy generation, improved equipment life and less project downtime. All these ultimately lead to cost benefits for asset owners through increased revenue, operational savings and reduced maintenance costs.
In this article, we explore some of the trends that are likely to dominate the solar O&M market in the near future and also assess the role of Covid-19 in accelerating the transition towards these advanced technologies…
Drones for inspection
With the increasing size and scale of projects, manual inspection may not be possible and is also time-consuming. For manpower to inspect each piece of the equipment spread across hundreds of acres of land, it would take days for fault detection. Thus, the use of drones for inspection activities has emerged as a popular alternative to manual inspection. Drones today can be equipped with thermal cameras to capture infrared signatures and detect defects, dirt and soiling on panels. These defects are geotagged and sent to O&M teams. Drones can provide more granular details than ground crews and can detect malfunctioning modules, specifically hotspots, which reduce electricity generation. They can also point out faulty strings with greater accuracy and reliability.
Robots in module cleaning
One of the most critical as well as expensive steps in the O&M process is the periodic removal of dust and soiling from solar panels. Traditional methods in India involve manual cleaning, using buckets of water in smaller projects and hosepipes attached to water tankers in bigger projects. These techniques not only entail large manpower requirements but also huge volumes of water. According to industry estimates, the cost involved in such a process can be 10-35 per cent of the total O&M expense, depending on the cleaning technique employed. It is the highest when tractors are used to spray water due to high manpower costs. The costs could go down if there is free access to water. Affixing a wet brush on a vehicle to clean modules greatly reduces both labour and water costs.
Costs can be further reduced with the help of robotic technologies, which limit human involvement as well as water usage in cleaning activities. These robots, in some cases, also use AI to self-assess the cleaning requirement by analysing dust deposition patterns and electricity generation data, subsequently carrying out O&M activities on their own. Many wet cleaning robots have wipers, scrubs, brushes, water and detergent for cleaning solar panels. In regions where water availability is a challenge, dry-cleaning robots, with large microfibre brushes on wheels, can be used, which rotate at high speeds to generate air flow and remove dust from solar panels.
Israel-based Ecoppia, for instance, offers various robotic module cleaning solutions that are completely automated and can be monitored remotely. Similarly, Arcelor Mittal-owned Exosun supplies battery-operated robotic cleaning solutions for solar panels. A domestic company, Skilancer Solar, has built a self-powered automatic robot that requires no manual intervention to operate. Although for some developers initial costs might be a challenge, the overall benefits are worth the investment. Moreover, with increasing uptake and technology advancements, these robotic technologies are likely to become more economically viable in the near future.
Automation and digitalisation can be effectively put to use in fault detection and rectification. Remote monitoring systems with real-time updates can help in predicting faults, diagnosing their causes and taking corrective action to prevent equipment or project downtime. Renewable energy developers are also focusing on setting up advanced monitoring systems to avoid paying penalties for over- or under-generation, according to the respective scheduling, forecasting and deviation settlement regulations. Thus, advanced asset management software with AI-enabled monitoring platforms is being increasingly used for predictive O&M by helping developers and O&M service providers analyse data from a number of projects and take corrective action accordingly.
Predictive maintenance technologies have evolved and advanced greatly to ensure higher reliability and efficiency. Cloud-based remote monitoring systems are being developed, which collect data based on specified critical parameters and make them accessible to O&M teams from anywhere in the world via cloud computing. Data loggers transmit this data to cloud-based internet of things (IoT) platforms, which can be accessed in raw form or aggregated forms, or as visual representations, so as to make monitoring of projects and even stand-alone equipment less complicated. For instance, TrackSo, developed by Free Spirits Green Labs, is an IoT-based energy management platform that helps track solar power plant performance and predict failures by providing proactive maintenance of assets.
Role of AI and machine learning
A technology with immense scope in O&M activities is AI-based digital twins. This involves the creation of a replica of the actual physical asset to be used as a benchmark for identifying faults through data anomalies. O&M teams can, thus, predict possible faults or breakdowns in a solar power project based on the behaviour of this digital twin in certain simulated conditions. Moreover, through simulation and the digital twin’s performance in these conditions, energy outputs can be enhanced if possible. Thus, a digital twin can be a powerful tool to improve plant performance, reduce breakdown-related expenses and improve revenues.
AI-driven robots are now being used in the module cleaning space. These highly efficient robots can take into account various parameters such as inclination of solar panels, geographic location, wind direction, and speed and dust content to clean panels effectively, without any manual intervention. They can also decide the frequency of cleaning and clean the panels as and when required.
Further, accurate weather forecasting services are being integrated with monitoring systems to predict energy generation and possible deviations from the expected energy output. In such cases, large real-time weather data sets sourced from satellites, weather stations and other devices are analysed and compared with historical weather data. Further, with the help of AI-enabled analysis of historical, present and future weather data, the impact of weather on solar production can also be determined. Moreover, machine learning is being applied in advanced solar O&M. Equipment is programmed to learn and react to the various operational processes and issues of a solar power plant so that through self-learning, machines can identify possible faults and diagnose them. At present, efforts are focused on developing advanced self-monitoring and self-operating machines that will, in time, evolve to such an extent that they will be able to address most (if not all) of the minor O&M issues on their own.
The way forward
During the first Covid wave and the resulting nationwide lockdown, developers faced issues in carrying out proper O&M and solar plant O&M activities suffered in many regions. Since O&M in India is highly dependent on manpower, the unavailability of skilled technicians became a major challenge in many remote locations, even for basic services such as security. Regular availability of water for module cleaning also became an issue. The summer months compounded the problem, with soiling losses being quite high and modules requiring regular cleaning. Many projects faced operational challenges due to the unavailability of spares. Moreover, in the case of rooftop projects, O&M teams could not get access to project sites.
Developers and O&M service providers that already had remote monitoring systems or some level of automation at their project sites were able to ride out the storm with better results than many of their counterparts who were dependent solely on manpower. Now, most developers and their O&M teams have transitioned to some level of automation and use digitally enabled tools for their daily O&M needs, which somewhat helped them during the disastrous second wave of Covid when manpower was extremely scarce.
While this digital transformation was a logical step and would have taken place over time, the Covid outbreak accelerated this process. Many developers have in the past few months invested in e-security, robotic cleaning, drones for inspection and advanced remote management systems. Going forward, this trend is likely to continue till automation and digitalisation move from a marginal role to a mainstream one in the O&M of solar projects.