In its pursuit of a permanently lower levelized cost of energy (LCOE), the wind energy industry is developing an exciting new generation of larger, more advanced turbines and pushing these highly complex machines to operate more effectively in harsher conditions around the world. This is evidenced by the scale of U.S. wind power, which has increased significantly in recent years.
For instance, the largest utility-scale onshore turbine in 2012 was around 3 MW, whereas now, the smallest available turbine is generally over 2 MW. On top of this, as scrutiny on performance grows, almost all turbines since 2012 have come stocked with condition monitoring sensors (CMS) as standard, compared to less than half before 2012.
It would be a mistake, however, to think that this progress means operations and maintenance (O&M) costs will reduce naturally as the industry matures and the size of turbines increases. A data-driven, digitalized approach is becoming the clear strategy for driving down costs of assets, but there are technical and cultural obstacles that remain. Surmounting challenges in data collection, data handling and data access will leave the industry well-placed to optimize turbine performance and reduce the LCOE, ultimately maintaining the competitiveness of the U.S. wind market.
Changing mindsets in data collection
Until now, U.S. wind operators had been using a number of traditional data collection and management processes to stay on top of project performance. Currently, many smaller operators still rely on the original equipment manufacturers’ (OEM) supervisory control and data acquisition data (SCADA) system, as well as vibration and lubrication data, to track performance curves, identify outliers and quantify the overall health of turbine components.
As part of the shift toward a data-driven approach, computerized maintenance management systems are increasingly being used by maintenance teams and increasingly providing good ways to gather maintenance data for parts consumption and time spent on maintenance work. However, despite these advances in technology, we need a cultural shift to reap the full benefits of digitizing turbine performance data collection.
It’s not easy to change the everyday working styles of technicians who have been experts in turbine inspection for a long while. In particular, moving from pen and paper recording practices to inputting data on a digital device is harder than it might seem. This may ultimately be a simpler process of data collection, especially if done using sensors built into a mobile device, but shifting mindsets is never an overnight task.
Once this barrier is overcome, detailed tracking of these data in digital format makes them much easier to understand and analyze than notes from a paper report. Digitizing O&M procedures also allows data collected during inspection to be integrated with condition monitoring data sources to deliver a more accurate picture of a wind turbine component’s health.
In order to realize these benefits, and before they can expect to accurately organize and understand their data, owners and operators should make sure to assess the technology they are buying to collect it very carefully. It is important to understand the technology well, which can be done using in-house engineering and third-party experts to complement the OEM’s support.
Organizing turbine data and using digital technologies
Once you’ve taken steps to collect your data in the right format, how do you make effective use of this to improve asset reliability?
The primary barrier faced in data handling for turbine performance monitoring is that owners often have multi-brand fleets where technologies vary, and data is tagged differently across machines. For example, some turbine models feature hydraulic pitch systems whilst others feature electric. Organizing data into a centralized database when it is collected in a variety of formats is done more efficiently when the data is digitalized.
It is particularly critical for U.S. wind asset owners and operators to stay on top of asset performance data and to have the tools to understand what it is telling them. It’s not just a case of having different types of machinery, as older turbines fall out of the production tax credits and owners and operators face the challenge of maintaining a healthy profit and loss sheet. In low-cost markets such as Texas, this can be especially challenging if owners are unable to monitor the performance of their fleet in one place and compare like for like performance data across all their different turbines.
Equally, using technology that makes reading the data simpler enables better isolation of an issue that might be affecting a single turbine’s performance or reliability. For owners, operators, and their asset management and operations management teams, using a comprehensive software platform to track the data of all their turbines means the scope of any given issue can be narrowed down easily, without wading through dozens of spreadsheets, to ensure an accurate call to maintenance action can be made.
Digital technologies and the application of machine learning (ML) to turbine performance data can speed up reading and analysis processes. However, if used improperly, ML approaches can deliver false positives, meaning an alarm could indicate technicians are to conduct maintenance that might not be necessarily urgent, leading to a potential waste of time and money. Supported by thorough engineering expertise and principles, ML can be connected to and informed by turbine mechanics, aligning with the constant evolution of turbine technologies. With the input of human engineering expertise to evaluate the urgency of an outlier in data sets and the necessary solution, digital predictive maintenance technologies can result in reliable, actionable insight.
Getting the most out of your investment in digitalization
By using digitalized predictive analysis to optimize O&M procedures with the aim of extending the life of assets, operations managers can accurately track the reliability of individual components and make small improvements in performance: for example, different control scenarios or better lubrication.
Asset life extension would not only improve the value of the industry, but U.S. asset operations managers are motivated to do so due to the renewal of production tax credits for upgrades rather than for scrap and rebuild. Initial return on investments and state-by-state legal obligations for a percentage of power to come from renewables are also motivators for turbine life extension in U.S. wind.
Good condition monitoring analysis develops objective criteria to define the severity of a feature or fault of interest, and condition monitoring suppliers typically do this with yellow (warning) and red (critical) alarm thresholding. Such alarms can be very useful indicators of emerging changes in turbine drive train health, but what’s most important is understanding the nature of the fault and whether it could lead to consequential further damage.
With digital data collection, organization and analysis rapidly developing, U.S. wind farm owners should not be settling for ineffective condition monitoring programs. In this day and age, nearly every major failure associated with rotating parts can be detected at least three to six months in advance of needing repair by using simple technologies to record and digitize the technology, present it and analyze it.
Data restriction poses a final obstacle
However much digital technology may make it easier to collect and analyze data and allow the U.S. wind market to realize significant O&M savings, there remains a final obstacle that puts these gains at risk – restriction of access to turbine performance data. Without full access to unprocessed raw data, owners cannot fully understand the health of their asset and manage it in the most optimal way possible. Ultimately, predictive maintenance analysis is only as good as the data available.
There are several barriers to free data access, but there should be little doubt that owners are entitled to access the data of the turbines they own. Barriers fall on a range from a failure of design, which might mean it isn’t collected in the first place, to a contractual limit that means owners and operators aren’t able to freely access all the data collected about their turbine performance. A number of vibration equipment manufacturers provide products and services with data secured by encryption or under a contract charging additional fee for access.
Unfortunately, this means there is a heavy reliance on OEM software, and owner/operators are tied to the specific turbine OEM for analysis and diagnosis of the equipment. In these instances, rather than being able to benefit from the focused expertise of an ISP, asset owners may need to go to great lengths with the condition monitoring equipment manufacturer to gain access to their data.
Moving forward, it will be important for the industry to recognize the importance of turbine owners’ access to raw, original data in order to implement the most accurate predictive maintenance analysis procedures. Data access should be designed in such a way that any entity can access raw data for analysis purposes.
Data access should not be considered a threat by turbine OEMs, after all. It isn’t about access to the kinds of design data that should rightly be protected as intellectual property. It is about unlocking the potential of the turbines they have supplied.
The wind industry is increasingly using a smarter, data-driven approach to turbine performance monitoring, but there are still cultural and technological barriers to overcome: most prominently, incentivizing turbine inspection and maintenance teams to fully digitize data collection, applying the engineering expertise required to enhance the effectiveness of data handling and analysis using digital technologies, and bringing about the industry mindset shift necessary to overcome turbine performance data restriction.
Ultimately, however, as the commercial and operational benefits of a digitalized approach are demonstrated by an increasing number of U.S. wind energy firms, this growing industry momentum will make these barriers seem less and less significant.