General comments

The paper proposes a novel approach to wake redirection that includes derating yawed turbines to operate within a ‘safe envelope’ for turbine loading. The effect of yawed and derated operation on turbine loading was simulated using BEM, while the effect on wake characteristics was evaluated with the steady-state wake model FLORIS. Based on the acquired simulation results, the proposed method shows great promise and hence the results are suitable for publication. The writing in the paper is mostly clear, with some minor errors that are detailed below. Furthermore, a small change in the structure of the paper is suggested. Additional comments can also be found in the list below.

Specific comments

• The paper contains many instances where one-sentence paragraphs could easily be combined with the preceding or following paragraph.
• Ln 13: Power maximization and cost minimization are interconnected and should not be regarded as two separate aspects.
• Ln 22: Please add a reference supporting this statement.
• Ln 32: Please specify whether this concerns the yawed turbine, downstream turbines, or both.
• Ln 45: Please specify how derating is achieved in this paper.
• Ln 182-185: Similarly, wake redirection is only employed for a limited range of wind directions. This could also be considered during the load-constrained optimization, as the safe region corresponds to average fatigue loads, correct?
• Ln 220: Can you be more specific on how the optimization is limited to a smaller number of turbines? For example, by making the yaw misalignments of downstream turbines a function of the upstream turbines.
• Ln 279-281: Not only will derating the turbine result in lower power output, but the decrease in thrust will also make wake steering less effective. Please also comment on this aspect of the combined control approach.
• Section 3.1 does not present any results and would fit better in the methodology section. This would also prevent some questions that arise after reading the current section 2 on how the combined control strategy is implemented.
• Ln 323: Which wake deflection model is employed within FLORIS?
• Figure 3: While the 3d bar plots clearly show the results of some of the edge cases, it is not possible to distinguish the effects of yaw misalignment and derating in the center of the plots. Please switch to a 2D contour plot, or 2D bar plot, or make the current bars uniform in color (no gradient). The same applies to Fig. 5.
• Figure 6: Is this constraint function independent of wind speed? Please specify if this is not the case.
• Ln 390: Rated speed of the 10MW turbine is set at V=11.4 m/s and not 14 m/s.
• 16: When considering the test cases, we see that most wind speeds are above-rated. What is the reasoning behind this parameter choice, given that wake redirection is most effective at below-rated conditions (if at all effective above-rated)?
• Figure 7: Would it not make more sense to refer to spacing s as the vertical between the turbines, and not the absolute distance between turbines as is used at the moment?
• Figure 11: Please increase font size. Furthermore, I suggest changing the axes to express distance as a function of rotor diameter D.
• Ln 455: The second case should refer to 45 deg instead of 90 deg.
• While the results with the 9 turbine farm demonstrate the effectiveness of the proposed control methods, the symmetry of the layout results in some redundant results (as also indicated in the table captions). Therefore, it would be interesting to consider a more complex/less uniform layout.
• The constrained optimization currently only considers blade root moments and blade tip deflection. Are these also the most critical components to consider, or will including other structural components further decrease the performance gain of wake redirection with derating?
• While the choice of a steady-state wake model to demonstrate the proposed combined wind farm controller is understandable, these models are prone to overestimating the effects of derating on the wake statistics. Please comment on this in the paper in the discussion of the simulation results with 3D distance.

Technical corrections

• Title: “load-constrained wind farm power maximization”
• Ln 10: “… while maintaining an unaltered design load envelope …”
• Ln 20: “Among all …, wake redirection (WR) has proved to be highly effective for increasing wind farm energy harvesting.”
• Ln 21: “… deflect its wake away from downstream rotors.”
• Ln 23-24: Single sentence paragraph, combine with subsequent paragraph.
• Ln 27: incomplete reference “Services (2004)”.
• Ln 36: “misaligned” instead of “misalignment”.
• Ln 45: “… integrated in load-constrained wind farm control.”
• Ln 60: “increased design loads”.
• Ln 74: Two consecutive sentences starting with “moreover”.
• Ln 94: Rewriting the sentence is suggested, for example: “Clearly, to keep every turbine operating below or at its design loads, a wind farm controller should maximize power output within the safe envelope of each turbine in a farm, thereby combining derating and wake redirection.”
• Ln 182: “… seldom considered problematic.”
• Ln 186: “That being said, …”
• Ln 186-189: For clarity, please rewrite into multiple smaller sentences.
• Ln 193: “ … a wind farm controller …”
• Ln 200: “steady-state wind farm controller”
• Ln 201: “… consists of …”
• Ln 202: “… yaw angle \phi …”
• Ln 205: Break in two sentences. “Similarly, the ambient characteristics are collected into an array p …”
• Ln 224: “…, an optimal load-constrained controller is proposed …”
• Ln 251: “… in Eq. (5).”
• Ln 276: “On that plot, …”. I assume this is referring to Fig. 2, but please specify which plot.
• Ln 287: “… define the turbines’ operative setpoints through a simpler sub-optimal optimization algorithm without a …”
• Ln 304: “The LQR, being model-based, …” Please rewrite the sentence for clarity.
• Ln 339: Missing reference for “Standards”.
• Ln 342: Capital letter “in order …”
• Ln 354: I believe there is a typo here, blade tip deflection is highest for positive misalignment angles (Fig. 3b).
• Ln 351: “As in the previous analyses, the positive impact of derating on loads can be clearly noticed.”
• Ln 391: “…, with D being the rotor diameter, …”
• Ln 407: “However, the gain increment in the sub-optimal case (i.e. 3.8%), appears to be only marginally lower than that of the combined control approach equal to 3.9%.”
• Ln 410: “…, especially in farms containing a large number of turbines.”
• Ln 423: “To this end, Fig. 9 and Tab. 5 report …”
• Ln 439: “… for 60% of the cases, … for 85% of the cases, … for 95% of the cases.”
• Ln 514: typo “analyze”