A wind turbine may appear to be straightforward, but it includes a number of intricate components that work together to sustain its overall structure. In recent years, such machines have suffered from severe underperformance in operations. For the wind sector to grow at par with solar power and other emerging segments such as green hydrogen, understanding the reasons for such underperformance, and thereby correcting them, is essential. At a recent webinar, “Overcoming Underperformance Challenges in Wind Turbine Operations: Analysing Losses and Improving Performance”, organised by Renewable Watch, a panel of experts discussed the methods and solutions available to overcome the underperformance of wind power turbines and improve their energy generation. The discussion delved into the key reasons for energy losses in wind farms and explored how analysis of these losses can help improve wind turbine performance and reduce operations and maintenance (O&M) costs. Edited excerpts…
Kumud Niranjan Mishra, General Manager, Wind O&M, Hero Future Energies
The primary difficulties we have encountered are comparable to those of all other independent power producers (IPPs) who have installed less than or close to 1 GW of wind energy. A significant one is original equipment manufacturers (OEMs) facing restrictions in accessing data. Given its significance in assessing our assets and enhancing performance analysis, data today is comparable to gold. Additionally, calibration of tools such as sensors and wind anemometers is quite important. The market has not yet developed to the point where we can recognise even the smallest flaw in each sensor. Although we examine all performance metrics with these sensors, the calibration is subpar.
The supervisory control and data acquisition (SCADA) system offered by OEMs is actually not particularly interactive, in that it can only detect small variations in the sensors and cannot always provide pop-up notifications to aid in calibrating such equipment. As a result, we frequently use our time, energy and money inefficiently.
Moreover, with respect to a turbine’s ability to generate thrust and power, the blade is among its most crucial parts. The blade, however, is also the component that is neglected the most. OEMs and IPPs frequently do not pay enough attention to a blade’s pitch alignment or timely cleaning. I think we can increase performance by, say, 0.5 to 1 per cent if we maintain blades properly, even on a biennial basis. Furthermore, because most of these turbines were developed by European companies, they are not well suited for India, especially with the persistently high temperatures in Rajasthan and even in some areas of Madhya Pradesh. High temperatures put these devices under a lot of stress, especially when combined with strong winds.
In addition, OEMs focused on installing high quality components in the past 10 years because they made substantial profits from these machines, but the scenario has changed. OEMs now use repaired parts in machines to utilise the inventory they have. As a result, low quality components cause assets to underperform within a span of six months to a year.
In India, O&M is moving away from OEM-based maintenance and towards independent service providers (ISPs) and self-O&M. Once you start using these modes, all of your data becomes available. This will undoubtedly clean the data market. As a result, one can allocate more funds to predictive maintenance and data analytics. Our current operations rely primarily on reactive maintenance. But because the entire industry is now shifting towards predictive maintenance, the company is also doing it gradually and methodically. As of today, no business can claim to be entirely dependent on predictive maintenance, because individuals still rely on external technology and OEMs to conduct fundamental tasks. We have a central monitoring station for preventive maintenance that was created by BACS. It provides us with prompt alerts. To enable us to act quickly, the system also offers fast analysis. Additionally, we have teamed up with Onyx Insight and are conducting a proof of concept at two of our sites to see how well we can operate such proactive systems locally.
Subhakanta Nayak, Group Head Wind O&M, Tata Power Renewable Energy
The major wind operation challenges that we are facing are primarily related to the underperformance of OEMs as well as technical issues such as failure of the gearbox and the generator. The failure of components is often attributed to India’s environmental conditions, as turbines face temperature errors in regions with extreme weather conditions. Areas such as Jaisalmer offer tremendous potential as wind sites, but during summers, very high temperatures pose a threat to the functioning of wind turbines. It has been observed that even in high-wind seasons, turbines tend to run smoothly only when the temperature is low. As the temperature starts rising, the turbine’s performance starts to deplete. Such conditions cause automatic degradation in turbine performance. Thus, we recommend that the cooling factor be checked thoroughly. To this end, we have developed a dedicated team to ensure that the work is done and things are put in order before the high winds set in. Improper asset management, such as not paying attention to gear performance parameters, also causes problems.
However, the biggest challenge, I believe, is the incompetency of manpower. Manpower, especially the budding youth, has not being specifically trained for turbine O&M, and we end up hiring from the existing pool of workers in the industry. This sector demands core workmanship. Issues in operation can be sensed by engineers when a turbine is not running on a proper power curve.
Further, to keep up with the digitisation of the sector, we have installed a conditional maintenance device that can sense vibrations, helping implement a proper action plan. We have also developed a performance department to carry out power curve verification for every single turbine.
Due to commercial obligations, OEMs are not able to perform as per their own requirements. If such commercial liabilities can be addressed by IPPs – for instance, if they could adopt the self O&M and ISP routes – a hybrid model can be implemented to address various conditions more effectively. Such a model would be a conditional model, wherein maintenance is done through self O&M in some conditions and ISPs in others. This may be the way forward to ensure that maintenance protocols are followed as per the requirements of a turbine, in a timely fashion. Maintaining grid parameters and the stability of internal feeders will also be key areas of focus, going ahead.
Umesh Raichandani, Senior Manager, Operations, EDF Renewables India
At EDF Renewables India, we currently have a capacity of 605 MW in operation, and we plan to expand in a span of one and a half years. I believe that OEMs in India are not open to accepting any external changes. For instance, no OEM in India is willing to make changes in their turbines as per the data obtained from monitoring systems. They rely more on their own designs and do not accept too many inputs from any other source. Another challenge is that the conditions of OEMs in India are not as one would hope. There is a scarcity of many essential parts, and assets are not maintained efficiently. For instance, the maintenance of gearboxes is often neglected for long periods of time. The majority of this negligence is either due to financial challenges, or scarcity of manpower and other resources.
An IPP’s main focus is on optimising a plant as much as possible. To overcome hurdles, it is important to have the capacity to transition to energy-based availability (EBA). Through EBA, we can get more insights into losses incurred. While availability of data is an important factor, the data must be as accurate as possible. We have seen, in a lot of cases, that wind speed is manipulated by a variety of factors. If the wind measurement is not correct, then a turbine may over- or underperform. A key aspect of performance analysis is the power curve. In power curve analysis, wind plays a very important role, as seen in most cases. Another major aspect that is neglected by OEMs is alerts and warnings. If these are noticed and addressed in time, we can save some major components from failing.
With respect to digital operations in India, we have an operational control centre, wherein the first step is to monitor assets effectively for wind speed, active power, alarms, warnings and curtailment. The second step involves taking actions based on the findings of the monitoring and operational control centres. A primary level of analysis is based on SCADA data, such as power curve analysis, and alarm and warning analysis, followed by other forms of analysis, which provide us with a deeper understanding of wind turbines. We also have a performance analysis and enhancement arm, which is part of asset operations. This arm derives results from generated data, which it shares with OEMs for better functioning and operation of turbines. We are using predictive maintenance tools for our day-to-day operations. Digital operations in these domains are really important for overall improvement in the performance of assets. I believe that the market is evolving more now, and the competition is increasing with respect to tariff and turbine models. In my view, from the IPP perspective, the challenges are still the same. Going forward, greater involvement and collaboration with OEMs can make our assets perform better.
Niraj Shah, Head, Operations & Maintenance (Wind), Apraava Energy
Speaking of the wind sector in India, there are two major issues with respect to turbine underperformance that need to be highlighted. One of them is the intangible part, which is when we make a business assumption and realise that underperformance is happening against our business estimates. Generally, wind estimates have always been on the higher side, and the actual wind has been declining year on year.
The other major problem we see is the kind of manpower that exists at present. Service providers do not possess the required materials when they move into outage, so you either have to take multiple outages, or the work is only half done, resulting in inefficiencies in turbine performance once it is put into service again. One of the major reasons for this is that most turbines in India are designed to incur the lowest capital cost. As these turbines are built in the cheapest way possible, they possess minimal sensors. Unless you have sensors that can receive and interpret data regarding the functioning of a turbine, you cannot analyse its performance effectively.
For instance, any gas or steam turbine, where most of the critical sensors are placed, will always have at least two sensors – a main sensor and a standby sensor. A comparison between the two allows the system to assess whether a machine needs to stop or continue working. This is fairly straightforward. However, as many people do not have so much information on their SCADA systems before moving to outage, they cannot plan the utilisation of materials well. Hence, preventive maintenance is suboptimal at a certain level. As a turbine continues to work, there will be a cascading effect caused by ignored maintenance issues, subsequently leading to a critical failure of the system. Another key challenge is that most OEMs and investors do not dive too deeply into the root causes of such failures. An attitude of “replace it and move on” has crept into the industry, which in my view is a key challenge.
We dived into the predictive maintenance space with an artificial intelligence (AI)-machine learning (ML)-based platform, which we have been using for the last two years. The interesting part is that it is a common belief that AI/ML platforms can help save a business millions of dollars, yet the ground reality is slightly different. The first practical problem is the amount of data coming from the SCADA system and OEM machines is very limited. The second problem is that, if I want to install additional sensors, then the system has to be able to accept internet of things data. As a result, I am forced to derive these data points from outside the OEM/SCADA system, since the OEM and SCADA do not allow me to put this additional data into their systems.
Going forward, electronic obsolescence will also be a key concern, as it will occur for most products developed over the past 10-15 years. Moreover, suboptimal manpower continues to be a limiting factor, as everybody is consuming manpower from the OEMs, and the OEMs have stopped developing their manpower further. The number of trained workers who have kept up with the growing demands of the industry is limited. The industry, as a whole, is surviving on the pool of workers generated by the OEMs.
In terms of outlook, I believe that the Indian wind market is moving towards higher PLCs on wind turbines. Consequently, wind turbines will be equipped with more instruments. This will result in more data being available to SCADA systems, leading to better-informed, data-driven decisions. Moreover, many players and investors in the market are considering self O&M as an option, unlike the current scenario of signing contracts with ISPs and OEMs. Such a transition is expected to change the whole approach of investors. They will seek greater digitalisation of wind turbines, so that they can make informed decisions.
