Jump to content

Julia Austin

Editor
  • Content Count

    135
  • Joined

  • Last visited

  • Days Won

    60

Everything posted by Julia Austin

  1. Q: Is there any relationship between the vibratory mode and wind direction and/or boom orientation? A: We have seen no significant correlation between the vibratory mode and boom orientation or wind direction. However, in order to see the signature of the vibratory mode, neither sensor can be in the tower shadow.
  2. Q: What is the root cause of the vibratory mode or mechanism that causes it to occur? Will the fix entail replacement of defective material or change to the manufacturing process? A: The root cause of the vibratory mode, the mechanism of how it occurs, and possible changes to the manufacturing process are facts we will readily share after successful field validation of the modified sensor.
  3. Q: Could you discuss the onset of the vibratory mode and time it takes to present itself? Could any sensor be impacted at any time in its life? A: Based largely on customer field data, we?ve found that pairs of sensors exhibit scatter as soon as 2 weeks and as late as 26 weeks. While sensors may enter the vibratory mode anywhere in this 2-26 week period, it is more likely to occur in the 2-6 week range.
  4. Q: Until you have new sensors to sell, what will you sell? Will NRG supply a different anemometer with tower orders until the problem is corrected? A: NRG continues to offer the #40C sensor for sale either as part of our complete NOW systems or stand alone. Many customers have continued to choose this sensor because it continues to provide usable data. As always, the data should be quality-checked, and depending upon the tower configuration, any potential slow-down can be mitigated through detailed analysis. The Symphonie data logger can support other types of anemometers. However we do not provide recommendations on specific manufacturers and leave that decision to customers.
  5. Q: Will NRG issue refunds or eventually, replacements? A: Once the validation of the revised sensors is complete, NRG will provide details on a warranty program to replace sensors that exhibit scatter in their data.
  6. Q: Even though it may not be of interest to everyone, will the cause and modifications be documented? A: We will be writing a white paper that documents our investigation.
  7. Q: Confirm that slides 8 and 10 show one, then both, then no bad sensors from top to bottom. A: On slides 8 and 10 the top plots show both sensors dragging, the middle plots show one sensor dragging, and the bottom plots shows both sensors as normal.
  8. Q: How far back does the historical data go and has there ever been a period when no sensor dragging existed? Is the target series of anemometers still the same? January 2006 or later? A: The first report of sensor dragging was received in 2007. Review of data from the past five years indicates that this problem began in 2006.
  9. No, damage to equipment due to lightning, vandalism, and birds is not covered by NRG's warranty policy.
  10. Q: What anemometers will work with the Symphonie logger besides the #40? What anemometer does NRG recommend in lieu of the #40? A: There are a variety of anemometers that will work with the Symphonie logger. Consult NRG Systems Tech Support or your sales rep for details. We do not recommend one anemometer manufacturer over another. Every cup anemometer has its own strengths and weaknesses which need to be evaluated in light of the project.
  11. Q: When will I get another update? A: The webinar of 12/02/08 was part of ongoing communication. We will continue to provide other updates and opportunities to speak with our team as more information becomes available. As always, please feel free to contact Sales or Technical Support directly via phone or through the Technical Forum on the web.
  12. Q: When will NRG be shipping good sensors again? How do I get some test sensors? A: Once we have fully validated and confirmed the performance of the modified sensor we will communicate those results. The timing for manufacturing and shipping new sensors is based on our ability to collect field data.
  13. Q: What are your success criteria for the validation of the revised design? A: We will determine success based on these three criteria: Data from field tests of 50 pairs of revised sensors do not exhibit excessive scatter or mean bias. Initial results will be assessed after one month with on-going testing. Customer review and acceptance of field results Validation of post-cal data
  14. Q: What role do booms play in the occurrence of #40 slow down? If you?re not ready to share the root cause, then how can you be so sure it?s NOT the boom? A: The boom could contribute to the vibratory mode through coupling, but it is not part of the root cause. Sensor dragging has been observed on many different boom types. This is one factor that has led us to rule out the boom as the root cause.
  15. Q: Can a ball bearing anemometer get into a vibratory mode? A: Based on our understanding of this issue, yes, it is theoretically possible, though we did not study this in our investigation.
  16. Q: How do we explain the 2-26 week onset of vibratory mode? A: The answer to this question involves a more detailed description of the root cause than we are able to share at this time. That information will be forthcoming upon field test verification.
  17. Q: Why did NRG not inform us in July that TSB008 was no longer valid? A: We did not yet know if there was a problem with the sensor itself. We believed there was, but did not think it was responsible of us to go public with any information before full testing and validation of our hypothesis.
  18. Q: Why has NRG continued to ship sensors that were possibly defective? A: There are five reasons: Prior to September we were not certain there was a problem with the sensor itself. While some believed the root cause to be boom vibration, our initial investigation told us this was likely not the case. That is why we continued to investigate the sensor itself. According to many of our customers, the data was still usable and was within the typical margin of testing error. Not all sensors exhibited dragging. Some customers reported no incidents, while others reported higher frequencies of incidents. Many variables contribute to uncertainty in wind measurement data (e.g. reference data, MCP methods, modeling, tower configuration, long term variability). Because the reported slow-down was a small piece of the whole, a piece that couldn?t be isolated, it did not make sense to pull the #40 from the market. After we identified the vibratory mode in September, we continued to ship sensors with all these factors in mind. Additionally, if we had pulled the #40, we would have taken huge capacity out of the industry.
  19. Q: If NRG can make sensors go bad in its new test lab, why not test all sensors before they leave NRG? Why continue to ship the current model if there is a test to screen for bad sensors? A: The lab test can reproduce the vibratory mode of sensors that have exhibited high scatter in the field, or by intentionally substituting problem parts in a new sensor. However, this test is unable to screen for future problems of normal production sensors.
  20. Q: If I have an effected sensor, how do I interpret my data? Do I scrap my existing data? Should I manipulate my data? What does NRG recommend doing to ensure reliable and accurate wind measurement data right now? A: The data should still be analyzed and scrutinized as one would normally do. Although additional uncertainty may exist, it can be accounted for and should be considered in the context of other sources of uncertainty (e.g. reference data, MCP methods, modeling, tower configuration, long term variability). We do not recommend any wholesale changes to equipment deployment and we continue to recommend installing sensor pairs at each level. It?s important to remember that, in most cases, sensor dragging leads to an under-reporting of wind speed, not an over-reporting.
  21. Q: How do I know if I have an affected sensor? A: It?s important to remember that not all sensors are affected by this phenomenon. The first step is to look at your data and analyze the scatter plots (similar to ones shown in the webinar presentation) to determine if two sensors at the same height diverge.
  22. Hello, Sebastian - We've spoken with some customers who report no incidences and some who report very high incidences of affected anemometers. Based on all of the data that we?ve looked at, and all of the sensors we?ve looked at, it?s somewhere between 20% and 30%. We?ve had it reported as high as 100% by some customers. So we can only base that on looking at the entire population of sensors, and our best estimate is it is somewhere in that range.
  23. Please click on the link above to view the #40 Progress Report Webinar transcript. #40 Webinar Transcription 120208.pdf
  24. Thank you for attending the Webinar. There were many good questions asked, and our Team is preparing answers ? answers should post by end of day Monday, December 8.
×
×
  • Create New...