the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Sep 2026
Simulation-Based Assessment of Flap Scheduling and Operating Height for Ground‑Gen Airborne Wind Energy Systems
Abstract. Airborne wind energy systems (AWES) are able to access stronger and more persistent winds at higher operating altitudes, although system performance strongly depends on the choice of aerodynamic actuation, flight-path geometry, and height of operation under realistic wind conditions. In case of rigid-wing ground-generation (ground‑gen) AWES, the influence of such a combination of operating parameters on traction-phase performance is still inadequately understood at constant settings of the control algorithms and limitations of the hardware. This paper tries to fill this knowledge gap by studying the impact of several key operating parameters of the ground-gen system with the aid of a validated nonlinear kite–tether–winch simulator of the Kitemill KM1 rigid-wing Ground-Gen AWES. Using a one-factor-at-a-time simulation campaign, the influence of fixed flap deflection, lateral path geometry, wind-field formulation, minimum production height, and elevation angle of the helix axis on the reel-out mechanical power and operation margins was examined. Under the formulation of an altitude-dependent wind field with the time-varying ground-reference wind input, the baseline design produces an average reel-out power of 7.56 kW. Application of speed-scheduled flap deflection in the range of conservative values of 0°–10° increases the average reel-out power up to 11.10 kW, which corresponds to an increase of around 47 % while repeated crosswind operation was maintained in the evaluated simulations. Increasing the minimum production height further improves the average reel-out power to 12.96 kW, which is about 71 % above the baseline one. On the other hand, within the tested circle–ellipse comparison, lateral path shape has only a minor influence on average traction power, whereas circular loop radius has a clear effect. Tuning of the elevation angle does not allow increasing the reel-out power further than the minimum height optimization does in the considered operating envelope.
- Preprint
(3657 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on wes-2026-148', Anonymous Referee #1, 30 Sep 2026
-
RC2: 'Comment on wes-2026-148', Anonymous Referee #2, 10 Oct 2026
The manuscript examines how selected operating parameters affect the reel-out performance of the KM1 system. The scope and limitations are generally clear. I recommend minor revision to clarify a few methodological details and make the parameter comparisons easier to follow.
-
EC1: 'Comment on wes-2026-148', Roland Schmehl, 13 Oct 2026
Dear authors,
About the manuscript's introduction:line 35-36: The sentence "For all simulations reported here, ..." does not make any sense here, in the context of the introduction. Up to this point, you have introduced AWE only at a very high level, describing the concept of pumping-cycle operation and mentioning that the study is based on a rigid-wing kite. Your sudden reference to a VTOL subsystem and symbols F_R and M_R is completely out of context. The VTOL subsystem is irrelevant to your study, and the symbols F_R and M_R have no place in your introduction. Move the statement that the VTOL subsystem is not active in this study to line 72, where you reference Fig. 2. It is here, in Fig. 2, that a reader sees a VTOL subsystem, and it is good that you explain that this is not part of the study. Then remove the VTOL subsystem from Fig. 1, as it is irrelevant and not modelled. The schematic aircraft is sufficient. Finally, move the statement about F_R and M_R to where you describe the model.
lines 40-51: Starting from "In the present context, ..." up to "... thereby influencing both traction power and flight stability." You define several parameters and features whose influence you investigated in your study. I recommended that the physics of the problem under investigation and the mechanisms of interaction be explained more clearly. However, I find this listing a bit too simplistic; you also do not explain how these parameters and features interact. For example, it is not necessary to define the "flap deflection strategy" as "prescribed or scheduled variation
of the aerodynamic flap angle, which modifies the lift, drag, and pitching-moment characteristics of the rigid-wing kite". The average reader of the article is knowledgeable about this. On the other hand, the effects on "airspeed, tether tension, reel-out power, and control margin during traction" are useful to know. Ideally, you refer to one or two studies that have previously examined "flap deflection strategies", whether for AWES or similar applications.
Best regards,
Your managing editor -
RC3: 'Comment on wes-2026-148', Anonymous Referee #3, 28 Oct 2026
The manuscript investigates the sensitivity of traction-phase mechanical power to flap deflection, flight-path geometry, prescribed wind conditions, minimum production height, and helix-axis elevation angle for the Kitemill KM1 ground-generation airborne wind energy system.
The work addresses a relevant operational problem, where its contribution is best presented as a controller-dependent simulation sensitivity study rather than an optimization of the full pumping cycle.
The authors recognize several important limitations, like the one-factor-at-a-time design, idealized atmospheric inputs, and lack of direct experimental verification of the reported gains.
The following comments should be addressed:
1) Although the manuscript presents the work as an assessment rather than a comprehensive optimization study, the sensitivity analysis could be more rigorous than the adopted one-factor-at-a-time approach. More systematic methods, such as variance-based global sensitivity analysis (e.g., Saltelli, 2008), could provide additional insight into parameter interactions and their relative contributions to output variability. The authors are encouraged to consider such methods or, at least, include a brief review of the relevant state of the art and discuss the limitations of their chosen approach.
2) Equation (28) averages only over intervals with positive winch speed. This is a valid definition of "conditional mean positive reel-out power", but it is not the mean mechanical power over the entire nominal traction phase, because the negative-power intervals are omitted. The negative intervals are significant in several power traces. If full pumping-cycle outputs are available, net cycle energy or cycle-averaged power would provide a useful secondary metric, even though this latter part is not analyzed/optimized.
3) Figure 3 appears to indicate an average reel-out power of approximately 40 kW for a fixed flap deflection of 5°, increasing to more than 45 kW at 10°. However, Table A1 specifies a maximum winch mechanical power of 35 kW. The authors should clarify whether this value represents a hard constraint imposed by the simulation model, a nominal equipment rating, or an operational limit that is not enforced in the reported simulations. If the 35 kW limit is intended to be enforced, the apparent discrepancy should be resolved, either by correcting the reported results or by explaining how these values were obtained.
4) As last, the proposed flap schedule is a major part to the claimed improvement, but its implementation is not reproducible from the manuscript. However, the manuscript does not provide sufficient information to reproduce its implementation. The authors should include an explicit mathematical formulation, pseudocode, or an equivalent algorithmic description.Citation: https://doi.org/10.5194/wes-2026-148-RC3
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 386 | 101 | 52 | 539 | 54 | 47 |
- HTML: 386
- PDF: 101
- XML: 52
- Total: 539
- BibTeX: 54
- EndNote: 47
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This paper is a sensitivity study of four pumping cycle parameters in groundgen circular flight AWES: radius, minimum height, elevation angle, and flap deflections. The authors have identified that manipulating those four parameters can improve cycle power. Their decisions to choose a baseline case as the starting point, then sequentially adjusting one of the four parameters does highlight the four paremeters’ impact on power production.
The paper has some promising results that can be investigated further to produce a more comprehensive study. Unfortunately, the current version does not provide enough depth or novelty to warrant a journal publication. There are also a few technical issues related to the way power is calculated and the performance of the chosen model.
The authors can consider the comments below if they decide to improve the study.
1. The abstract only mentions what was done, not what novel results have been found.
2. The current method of calculating the average reel-out power (eq. 28) is problematic. By disregarding the negative power portion (vt < 0), the authors have not included the power consumed by the winch when it temporarily pulls the kite back in during circular flight. This approach will inflate the average power figure. I have not seen any past study that only includes the positive portion of the power vs time graph in the average power calculation, nor have the authors justified their unusual choice.
3. The kite’s inability to remain airborne while using 15 deg flaps should be investigated. A prior parameter sweep of the same KM1 model (DOI: 10.1002/we.70109) found that the same airframe have no trouble remaining airborne at 30 m/s airspeed. If I extrapolate the blue line in fig. 3 of the author’s manuscript, I’d expect a 15 deg flap deflection would bring the flight speed down to around 50 m/s (180 km/h or 97 knots) – well above 30 m/s as seen in (DOI: 10.1002/we.70109). Moreover, 50 m/s is the flight speed of a small two-seater aircraft, so a 54 kg kite (even if tethered) should have no issue maintaining flight in this airspeed.
The authors mentioned ‘tether collapse’, which I suspect refers to the tether sagging (tension reaching zero). This apparent low performance may be attributed to a lack of pitch and yaw control (the authors have stated that only the rudder was used for flight control). To this end, the authors might need to either implement a more comprehensive flight control system, or at least trim the elevator to a better value that generates enough lift to keep the tether in tension. Investigating the airspeed and angle of attack at the point of instability may reveal the reason for the crash at higher flap settings.
4. The section on radius surrounding fig 4 is unclear. Are the authors comparing circular vs elliptical trajectories? If so, the baseline circular radius and power have not been stated. Furthermore, the current airframe has a 7.5 m wingspan, so being able to fly a 20 m radius loop (i.e., diameter / wingspan = 5.3) is quite impressive (given that only the rudder is used). It would be useful to show the exact trajectory and rudder movement required to achieve such a tight turn.
5. The discussion in fig 8 does not mention much beyond showing the figure and reporting the average power.
6. Flaps scheduling: this is arguably the most novel and promising aspect of the paper. This topic should be studied in more depth. Currently, the authors only show one plot of power vs time, and the exact flap scheduling law is not provided (the authors only said that flaps are scheduled with speed and move between 0 and 10 deg, without providing any further detail). I would expect a clear explanation of how flaps affect flight dynamics during circular flight, thereby explaining the power gain, plus some investigation on the aerodynamics (e.g., lift to drag at different flap deflections) in the tethered flight configuration.
7. Minimum production height: similar to the above, only one time series of the mechanical power at the best minimum production height is presented without much discussion. Readers would expect to see results from the parameter sweep and discussions on the link between the sweep parameter (min production height) and power.
8. Elevation angle: similar to point #9 above, except that no result was presented beyond a single number reporting the optimal value.
9. Figure 13 does prove that manipulating the three parameters investigated in this paper can improve power. However, the +71% gain figure has not been rigorously quantified. The authors have sequentially modified one of the three parameters, finding the optimal point after each sequence. However, the true optimal point cannot be determined this way – it’ll need either a full 3D parameter sweep or an optimisation search.
10. The main KPI for this paper is cycle power, which is heavily influenced by winch speeds (both during reel out and reel in). Please provide some information on this. Are the winch speeds fixed in all cases?