New service package optimizes operation of wind turbines

Keeping the technical availability of wind turbines at a continuously high level is not an easy task. Many operational problems are still not recognized in time. With the new WINDcenter, Steag Energy Services wants to change this. The ew editorial team spoke with Fabio Wagner, head of the WINDcenter, about the objectives and also the special features of an OEM independent service package for wind turbines.

Mr. Wagner, how much experience does Steag Energy Services have in the field of wind energy?

Steag Energy Services supports its customers as an international service provider in the planning and implementation of wind farms: starting with the design, the approvals, through negotiations with the turbine manufacturers to the management of construction supervision. In addition, the company offers support in the operation and maintenance of wind farms with the experience that we have accumulated from many different wind energy projects worldwide. We are a leader in the management of wind farms in Brazil and have a total capacity of more than 1 GW under contract there. For example, we operate the wind farms of Assuruá (234 MW), Itarema (207 MW) and Gestamp (234 MW). We have also gained extensive experience in India in the field of wind farm operation and maintenance.

With the WINDcenter you are providing an OEM independent solution package for the monitoring and optimization of wind farms. What are the components of the package?

The package is based on our tried and tested IT systems SI®/PAM, SR::x and SR::SPC. With the operation and maintenance system SI®/PAM we cover the entire area of technical operations and maintenance management for wind farms. With the data management system SR::x we collect, store and visualize the operating data of the wind turbines. SR::SPC analyses the data from SR::x using high-quality key performance indicators (KPIs) for early detection of changes in the operating behavior of the plant. If anomalies are found, the experts from the WINDcenter carry out a detailed root cause analysis and provide specific recommendations for action to solve the problem.

At a wind farm which went into operation in 2014, our IT system generated a signal in November 2014 which initially indicated that the temperature of the generator winding was higher than expected, but still below the SCADA alarm threshold defined by the manufacturer.

Fabio Wagner demonstrates how the system works: In case of relevant deviations from the reference value, the system automatically triggers an alarm signal. It is not based on fixed limit values or alarm thresholds in the process control system, CMS or SCADA system, but on the actual behavior of a wind turbine taking account of the current operating conditions.

Can you explain how the system works more precisely?

The key performance indicators I mentioned are configured by our experts for each plant. It is important to select suitable wind turbine parameters with regard to the process to be monitored. From the available data, our system learns how these parameters correlate with each other. It generates a reference value for the operating condition defined as “good”, which is then compared with the current measured value. In the case of relevant deviations from the reference value, the system automatically triggers an alarm signal. Our system is therefore not based on fixed limit values or alarm thresholds from the process control system, from the condition monitoring system (CMS) or from the SCADA system, but rather on the actual behavior of a wind turbine taking account of the current operating conditions.

Can you give us some practical examples?

A whole lot of them, but here is one case in particular. At a wind farm which went into operation in 2014, our IT system generated a signal in November 2014 which initially indicated that the temperature of the generator winding was higher than expected, but still below the SCADA alarm threshold defined by the manufacturer (figure 1). The signal came directly after the wind turbine went back into operation after a shutdown. Whenever we receive such signals, we graphically display all the related influencing factors in the SR::x visualization tool in order to identify dependencies and establish correlations. The further analysis showed that there was no increase in the cooling water temperature. Rather, it became clear that the cooling water pressure had risen (figure 2). In this context, we also realized that the absolute level of the cooling water pressure had also risen in comparison with the other turbines in the wind farm. We finally identified a clogged filter in the cooling water system as the cause, and recommended replacing or cleaning it.

The further analysis showed that there was no increase in the cooling water temperature but that the cooling water pressure had risen. Finally, a clogged filter in the cooling water system was identified as the cause and shortly afterwards it was recommended to replace or clean it.

Wouldn’t the manufacturer’s software have recognized the cause of the damage?

No, a conventional condition monitoring system would not even have detected the temperature rise of the generator winding at that point. The wind turbine’s SCADA system did of course issue a fault signal when the winding temperature reached the system’s SCADA alarm limit. However, that only happened ten days after we had already received the alarm signal from our system. In response to the increased winding temperature, the SCADA system shut down the wind turbine until the temperature had fallen to a non-critical level again. This process was repeated several times (figure 3). There were therefore production losses totaling around 75 MWh. In addition, the generator winding was most probably damaged as a result of the late SCADA alarms and repeated start-up processes. That could shorten the service life of the generator.

How would you summarize the advantages of your service package for wind farm operators?

Our IT solutions enable us to monitor a large amount of process data. We can react quickly to changes in the operating behavior of a plant. Our analysis of SCADA and CMS data issues warning or alarm signals only when they are relevant.

Since our experts evaluate the data while considering all influencing variables for each specific alarm signal, we can identify the cause of a problem or a change in operating behavior very accurately. In addition, our solutions enable us to recognize very early on when a system is no longer able to perform as it should.

In response to the increased winding temperature, the SCADA system shut down the wind turbine until the temperature had fallen to a non-critical level again. This process was repeated several times, bringing the total production losses to around 75 MWh.

Do these advantages also have an impact on the maintenance of wind turbines?

Of course. With the WINDcenter, a predictive and therefore condition-based strategy can be implemented. This not only reduces plant downtimes, but also avoids additional costs for unplanned maintenance measures or for the repair of consequential damage. With this background, the WINDcenter is particularly interesting for offshore wind farms, since early detection of possible technical problems enables maintenance measures to be planned in good time and, above all, much better and more cost-effectively. This applies from the availability of spare parts to the complete logistics.

How is the industry reacting to your new service package?

The reactions are extremely promising, and that is also because we offer an OEM independent service. An experienced wind farm operator can use our IT systems directly. Here, we can deal with many different arrangements. For this reason, our solutions are used worldwide in wind farms with a total capacity of 1.3 GW. This shows us that we are on the right track.