The manuscript “Performance of Multi-Band MDE-Based Virtual Sensing for Estimating Lifetime Fatigue Damage Equivalent Loads for the IEA 15 MW Reference Wind Turbine” investigate the well-known multi-band Modal Decomposition and Expansion virtual sensing approach on simulation data of the IEA 15-MW Offshore Reference Wind Turbine to estimate section moments and compare them with simulated responses. The results show errors near the tower top and close to the water level. This study provides a comprehensive overview and literature application of multi-band MDE for virtual sensing in wind turbines, but it’s novelty in methodology is unclear. Although the paper mentions the issue of modelling RNA as a lumped mass, it only demonstrates the presence of errors without thoroughly investigating their root causes. Only a detailed investigation of the cause, such as the use of the virtual sensing approach with an RNA-based lumped mass model and the complete modelling of the blades, would allow a credible statement about the error due to blade modelling.

• Clarify the novelty of your approach compared to existing studies, highlighting specific contributions and advances.
• Consider shortening the Data section, moving detailed information to an appendix to maintain focus on virtual sensing
• Provide a rationale for the multi-band approach boundaries in Table 6, indicating if they are standard or proposed by the authors.
• Consider conducting a direct comparison between models with and without accurately modelled RNA, including rotor blades, to isolate error sources.
• Offer a sample time series of strain data and analyse where discrepancies originate, enhancing the understanding of the study's context.
• Investigate errors in the frequency domain to offer deeper insights into their origins and behaviour.

This work is a well-written and understandable read with precise figures. The reviewer congratulates the authors for its high level.

 

The only major point of critics concerns the motivation and relevance of the overall concept. The authors argue MDE is applied at positions not accessible for sensor application and than later show the significant errors to be close to the tower top. This position, however, seems rather easy to access and equip with sensors (as also mentioned by the authors and displayed in Figure 11). This contradiction seems unresolved to the reviewer and I ask the authors to comment on this. 

Another aspect which might merit some additional explanation in writing: According to my understanding, DLC4.1 as per IEC61400 does not include turbulence. It's frequency in the standard also is significantly lower than its practical occurrence, mostly due to curtailment of power output as a function of energy trading. Recent presentations on the WESC 25 mentioned up to 50 turbine stops per day and also claimed this to be critical as they happen also under unfavorable operating conditions. I fully understand this cannot be covered with the present work, but I think a word or two in the conclusions to mention the possibility of DLC4.1 playing a much more significant role in real-world turbines would add value to it.

One last thing that escaped my understanding and where I merely ask for a layman's explanation is the seemingly contradiction of the tower top sensor being there and the huge error of MDE at tower top (also pointed to in a dedicated comment).

All comments including above statement are in the attached pdf.