For more than two years, wind 7 studied the market closely to purchase a management system based on the requirements of an operator of wind farms (see the report Turbine Management: Operator Requirements). All products on sale were tested and evaluated. A number of providers had programmed comprehensive databases, but few systems were designed convincingly with respect to the aforementioned requirements. In particular, all of them failed to offer the fully automated capture of "unfiltered" data from different types of wind turbines. Although some 10 different companies have been working towards this goal for years, often in close cooperation with major project developers, no one was able to present a working solution.
Therefore, in the fall of 2003 wind 7 decided to develop its own solution. Managing Director Stefan Groos, who was in charge of the project, found the optimal partner in DrehPunkt GbR of Rostock, whose combination of experienced operators and creative programming experts was very promising. Indeed, the first test version was installed after only a few months. Since April of 2004 -- after only six months of development -- wind 7 has been managing all of its wind farms with "ROTORsoft."
ROTORsoft is an integrated IT system for all relevant events and processes in the management of wind energy projects. Various interfaces are provided for the wind farms to read out all of the raw data automatically and store it permanently in an SQL database (or another one if the customer so requests). Process control ensures that important information is immediately sent to any designated recipient by fax, e-mail, or text message. Operation is made as easy as possible by using a Web-based technology that everyone is familiar with today. The software can practically be used around the clock from anywhere in the world via Intranet (company network) or the Internet. Users do not need to install any special software.
The program's start page provides an overview of all of the important information at a glance -- either as an overview map (Figure 1) or as a list. The various wind farms have round symbols in different colors assigned to them. The vertical segments tell us what the technological status of the systems is, while the horizontal ones symbolize the status of communication. Flags are used to indicate the direction and speed of wind as is done on classic weather maps.
Figure 1.
Basically, the traffic light principle is used: if everything is running well, the symbols are green, but if there is not enough wind, the vertical part is yellow, while red indicates a malfunction. Blue is used to indicate maintenance. If any alarms have been issued, the wind farm's symbol has a red ring blinking around it so that the manager can immediately see where work needs to be done. The same applies for communication: if there are any problems during remote communication and data cannot be read out smoothly, the horizontal circle segments of the wind farm symbols are red.
As common with communication controllers in other industries, subordinate levels always report the most negative features to the next level. The symbol on the overview for the wind farm is only green if everything is running smoothly. Even if only one component of a wind turbine is malfunctioning, the red warning signal will appear on the overview. The manager can get a quick overview just by using a mouse. On the next level, the various turbines in the wind farm appear. One level down from there, additional details about the components, etc. are provided.
To the left of the overview, all of the wind farms are displayed in a commonly used tree structure known from other programs. Each wind farm has its own main directory, with each wind turbine constituting a subdirectory, etc. There is no limit to the amount of details that can be stored here. If it makes sense, even the smallest parts -- such as screws, fuses, etc. -- can be recorded. As wind 7 already handles the procurement of spare parts and related logistics itself for individual wind farms, such options are anything but niceties; rather, they are an elementary management tool.
There is a task field at the top right of the start page where important events are displayed. Users only see the tasks assigned to them because everyone has to register with their personal password. The management can see exactly what has been finished thanks to automated event tracking.
Point the mouse cursor to one of the wind farm symbols in the overview to display some initial details about the wind turbines at the bottom right. For example, the time at which the performance data of individual machines, wind speeds, etc. was last checked is indicated. Click on this information to have additional details displayed.
The wind farm overview (Figure 2) provides a table of all of the important information about individual machines, such as energy yield, operating hours, temperatures, etc. The management can define which data are to be displayed here. Just like the start page, the farm overview can also be displayed as a map. Here, each individual machine then appears as a round symbol with colors based on the traffic light scheme.
Figure 2.
Another important management task is determining availabilities, especially when it comes to warranty issues. Disputes are bound to crop up if, as happens so often, availability is as poorly defined as the data on which the determination is based. Many managers blindly trust the figures that the manufacturer's software provides. While no one wishes to suggest that such data is manipulated, technically such fraud is possible. Therefore, management is well advised to determine the exact figures and document results here.
The software was designed not to reproduce the availability data calculated by manufacturer's software, but to read out the raw data of control systems (generally values calculated every 10 minutes) and come up with its own results even as it leaves a lot of leeway for a definition of availability. For example, manufacturers often count maintenance and repair time as availability although the machine is not working. ROTORsoft filters out such events. In other words, there is not merely a sole distinction between "available" and "unavailable," but rather additional modes (such as "maintenance") that can be statistically analyzed.
Another important item is the analysis of the frequency of errors. Control systems for wind turbines almost always have hundreds of different messages to output. The software subsumes all of them as "ROTORsoft messages" to standardize this data so that analyses can be conducted for more than one manufacturer, which is especially important when new wind turbines are to be purchased. It used to be that only insurance companies had such knowledge, which they unfortunately kept confidential for competitive reasons. But operators need to create their own statistics to improve quality in the midterm.
Germany's Wind Energy Association (BWE) has therefore already begun setting up an anonymous database. ROTOR will allow information from your own management to be added to the new BWE database automatically so that no extra manual work is required. Likewise, automated messages can be sent to IWET and IWR, for instance, to create wind indices. Such events used to cost operators a lot of time and effort. And a lot of them thus did without such messages, which only damaged the industry as a whole. Everyone would benefit from great detail in such a database.
Management is often obligated to report to customers / investors regularly and often even to the bank. Generally, such reports have to list the monthly production in detail and provide a comparison to benchmarks. This comparison shows what the production would have been in a normal wind year / month, with the usual proviso of fluctuations. Fund initiators generally publish these assessments as a sort of "balance on current accounts."
The amazing thing is that it sometimes takes up to a year before such analyses are published. Indeed, such analyses used to take a lot of work. Nonetheless, any fund initiator should be able to perform this task within a few days. ROTORsoft can handle this job in just a few seconds. Once the index is published, it is automatically imported at the data center. The system already has all of the energy yield data for each wind turbine anyway. And the index region is entered into the system when a new wind farm is set up. All you need to do to compare the current production data to the index is press a button.
Comparisons of current and target figures are also important, especially for fund models. Forecasts (such as from a fund prospect) are entered into ROTORsoft when a wind farm is created. For the set intervals, analyses can be presented in just a few seconds as tables or charts (Figure 3).
Figure 3
All kinds of other analyses -- such as for production, wind speed, wind direction, availability, frequency of errors, etc. -- can be created easily as tables or charts. All of the information in the database can be processed in any combination or easily exported to other programs, such as Excel.
The log book is the central controlling tool. It automatically records all of the events that occurred during the operation of the turbine, any malfunctions, status messages, etc. with an indication of the date, time, etc. The most important part of these records is the ability to track all (!) acknowledgments and user actions: a red warning blinker on the overview site is only deactivated, for instance, if a staff member with the proper authority has acknowledged the error. The log book records exactly which staff member acknowledged which error at which time. In addition, comments can be added about the causes of the malfunctions and the actions taken. The owner of the wind farm thus has absolute transparency at all times and can hold the manager liable for failures.
For technical management, the machines must truly be controlled holistically. Preventive maintenance should always be performed. A Condition Monitoring System can be smoothly integrated.
A fully automated management report function is also about to be completed. Here, users will first define the wind farm, time periods, and other details and then determine which components the report will contain: overview of energy yield, log book, event report, analysis of causes, availability, performance curves, etc. Users then need only click on "Create report" to produce a PDF document containing all of the desired information, which can then be sent by e-mail.
The software is already being used for all of the wind turbine types commonly installed in Germany, such as Enercon, GE (formerly Enron / Tacke), REpower, NEG Micon, Nordex, Vestas, AN-Bonus, Frisia, etc. Less commonly used turbines, such as those of Mitsubishi, Pfleiderer and WindMaster, have not yet been integrated. Gamesa is the only major manufacturer that has not yet been implemented. If customers are interested, an interface can be provided for these customers in the short term.
In the past few months, a number of talks have been held with potential customers. Interest is tremendous. WPD of Bremen, the market leader among fund initiators with more than 500 wind turbines in its portfolio, has already chosen to use ROTORsoft, and other operators and operating companies are soon to join. By the end of the year, a total of almost 1,000 turbines in Germany and neighboring countries will be managed using ROTORsoft. It thus also pays to have quality in the wind industry, too.