06 Apr 2021
06 Apr 2021
A Reference OpenSource Controller for Fixed and Floating Offshore Wind Turbines
 ^{1}University of Colorado Boulder, Boulder, CO, 80309, USA
 ^{2}National Renewable Energy Laboratory, Boulder, CO 80303, USA
 ^{1}University of Colorado Boulder, Boulder, CO, 80309, USA
 ^{2}National Renewable Energy Laboratory, Boulder, CO 80303, USA
Abstract. This paper describes the development of a new reference controller framework for fixed and floating offshore wind turbines that greatly facilitates controller tuning and represents standard industry practices. The reference wind turbine controllers that are most commonly cited in the literature have been developed to work with specific reference wind turbines. Although these controllers have provided standard control functionalities, they are often not easy to modify for use on other turbines, so it has been challenging for researchers to run representative, fully dynamic simulations of other wind turbine designs. The Reference OpenSource Controller (ROSCO) has been developed to provide a modular reference wind turbine controller that represents industry standards and performs comparably to or better than existing reference controllers. The formulation of the ROSCO controller logic and tuning processes is presented in this paper. Control capabilities such as tipspeed ratio tracking generator torque control, minimum pitch saturation, wind speed estimation, and a smoothing algorithm at nearrated operation are included to provide a controller that is comparable to industry standards. A floating offshore wind turbine feedback module is also included to facilitate growing research in the floating offshore arena. All the standard controller implementations and control modules are automatically tuned such that a noncontrols engineer or automated optimization routine can easily improve the controller performance. This article provides the framework and theoretical basis for the ROSCO controller modules and generic tuning processes. Simulations of the National Renewable Energy Laboratory (NREL) 5MW reference wind turbine and International Energy Agency 15MW reference turbine on the University of Maine semisubmersible platform are analyzed to demonstrate the controller's performance in both fixed and floating configurations, respectively. The simulation results demonstrate ROSCO's peak shaving routine to reduce maximum rotor thrusts by nearly 14 % compared to the NREL 5MW reference wind turbine controller on the landbased turbine and to reduce maximum platform pitch angles by slightly more than 35 % when using the platform feedback routine instead of a more traditional lowbandwidth controller.
Nikhar Abbas et al.
Status: closed

RC1: 'Comment on wes202119', Anonymous Referee #1, 04 Jul 2021
Dear Authors,
First of all, I would like to congratulate you on this very nice manuscript! The ROSCO controller and toolbox seems to me very helpful for the wind community. The paper is very nicely written and very well organized.
Please find below some more important issues, some mathematical issues and some minor issues which hopefully help to further improve the manuscript. Due to lack of time, I could not double check my points, so please forgive me, if I missed or misunderstood something.
There are some more important issues:
 Page 6, p. about Region 2.5: You mention that “due to the PI and controllers and the setpoint smoother, there is no specific range for Region 2.5”. Since the PI controller and the setpoint smoother as implemented should not change the steady states, it should be always possible to calculate, at which wind speed the steady state of the rotor speed reaches rated rotor speed (depending on k of your k*Omega^2) and at which wind speed the static pitch is larger than the fine pitch. This should also allow to calculate the limits of region 2.5 even with the peak shaving. Please consider this.
 Figure 2 left: TSR seems to continue to be constant at the end of region 2, but rated rotor speed is already reached (which is not possible). For the 5MW, rated rotor speed is also reached before rated power. It would be further helpful to add region 2.5 in the plot (see comment above).
 Section 2.5: The authors write that the integral gains are in general negative for standard horizontalaxis wind turbines. However, this depends on the definition of your speed error (reference – measurement or measurementreference). From my perspective, the positive gains are more common (see e.g. Jonkman 2009). Please revise this part.
 Equation (15): Equation (15) is usually obtained setting the rotor motion from Equation (2) to zero and integrating the aerodynamic torque from Equation (3), see e.g. Bossanyi 2000. However, the efficiencies of the generator and gearbox are not part of Equation (2) and (3) of your paper. From my perspective, the efficiency of the gearbox should be part of Equation (2) and thus also (15), but the generator efficiency is only important to calculate the electrical power from the generator torque and thus should not be a part of Equation (2) and thus also not of Equation (15). Please make this part more consistent.
 Equation (21): If “constant power” is used, one can also include the partial derivative of the generator torque with respect to the generator speed in Equation (5). Neglecting this usually causes a large deviation from my experience, also for the 5 MW reference wind turbine. For the ROSCO controller and for the paper it would be nice, if you could include this part or provide some investigation that in your case this is neglectable.
 Section 5.5: Usually (in the Bladed interface), the tower top foreaft acceleration is a translational degreeoffreedom and thus the integrated signal is the tower top foreaft speed. This has been used in your reference (van der Veen 2012). However, you use the “tower top pitch angle” (i.e. rotational DOF), which is also possible (but much harder to measure/estimate in reality) and would provide similar results I assume. But since ROSCO is using the Bladed interface and aims to reflect the industrial stateoftheart, please consider changing to the translational DOF.
Further, the paper can be improved using consistent and accurate mathematical expressions:
 Equation (5) etc.: Please use consider using $\delta$ instead of $d$ for $\beta$, $v$ etc. and introducing that $\Delta$ is the deviation from the steady state. Using simply $d$ might cause confusion with the operator “d”.
 Equation (9) etc: Please consider that “d” is an operator and thus using $\mathrm{d}$ instead of simply $d$ would be more appropriate.
 Equation (6) etc: the tip speed ration might be good to introduce. And here, using $\partial$ as in Equation (4) would be more appropriate, since the tip speed ratio depends on both wind speed and generator speed.
 Equation (12): The transfer function $H(s)$ connects the Laplace transform of the input to the one of the output. The Laplace transforms themselves do not depend on $s$. Thus, the fraction with d\Omega_g(s) is a bit sloppy. Best might be to simply remove this and explain that the transfer function is obtained by using the Laplace transform and Equation (5) and (9).
Minor issues:
 Figure 2 and Figure 5 caption, Appendix A etc.: Units are in nonitalic in the rest of the paper (which makes sense, since they are not variables), but here you have $kNm$, $MNm$, etc.. Please consider changing them.
 Figure 3: setpoint smoother has more inputs than only the generator speed.
 l 297: “but the power is much more consistent” is not clear to me. Maybe just remove or add something to better explain it. Maybe you mean “more consistent compared to the constant torque case”?
 in Equation (20) you use “rat” as subscript, but in the rest of the paper “rated”. Please consider to have this consistent.
 Section 3.1, last sentence: From my perspective, the proportional and integral gains for the torque PI controller are often chosen to be constant for simplicity, since applying Equation (13) usually does not provide significantly differences over the considered operation points. Please check if this could be also helpful here. The reason provided in the paper (“less erratic control actuation…”) seems to be a bit vague for a Journal paper.
 l 377: Equation (17) should be included here since Equation (16) is TSR tracking only.

RC2: 'Comment on wes202119', Anonymous Referee #2, 08 Jul 2021
Congratulations for the nice work and the fine manuscript! The wind community will definitely benefit from the ROSCO controller and toolbox. The paper has a clear and logic structure and is nicely written.
The detailed comments from the first reviewer are very pertinent and I agree with them but will not repeat them here.
My main comment after going through the very extensive manuscript is: while undoubtedly the ROSCO controller and toolbox are a significant contribution to the community and represent a very large effort, the authors should also consider highlighting the scientific contributions of the paper, avoiding reading the article as a “ROSCO user manual”. It seems that most of the effort has been towards implementing well known methods and approaches, and automatizing some of the processes, which is of course a great effort that merits appreciation, however it does not automatically and necessarily entail publication in a scientific journal.
 AC1: 'Response to Reviewer Comments', Nikhar Abbas, 17 Aug 2021
Status: closed

RC1: 'Comment on wes202119', Anonymous Referee #1, 04 Jul 2021
Dear Authors,
First of all, I would like to congratulate you on this very nice manuscript! The ROSCO controller and toolbox seems to me very helpful for the wind community. The paper is very nicely written and very well organized.
Please find below some more important issues, some mathematical issues and some minor issues which hopefully help to further improve the manuscript. Due to lack of time, I could not double check my points, so please forgive me, if I missed or misunderstood something.
There are some more important issues:
 Page 6, p. about Region 2.5: You mention that “due to the PI and controllers and the setpoint smoother, there is no specific range for Region 2.5”. Since the PI controller and the setpoint smoother as implemented should not change the steady states, it should be always possible to calculate, at which wind speed the steady state of the rotor speed reaches rated rotor speed (depending on k of your k*Omega^2) and at which wind speed the static pitch is larger than the fine pitch. This should also allow to calculate the limits of region 2.5 even with the peak shaving. Please consider this.
 Figure 2 left: TSR seems to continue to be constant at the end of region 2, but rated rotor speed is already reached (which is not possible). For the 5MW, rated rotor speed is also reached before rated power. It would be further helpful to add region 2.5 in the plot (see comment above).
 Section 2.5: The authors write that the integral gains are in general negative for standard horizontalaxis wind turbines. However, this depends on the definition of your speed error (reference – measurement or measurementreference). From my perspective, the positive gains are more common (see e.g. Jonkman 2009). Please revise this part.
 Equation (15): Equation (15) is usually obtained setting the rotor motion from Equation (2) to zero and integrating the aerodynamic torque from Equation (3), see e.g. Bossanyi 2000. However, the efficiencies of the generator and gearbox are not part of Equation (2) and (3) of your paper. From my perspective, the efficiency of the gearbox should be part of Equation (2) and thus also (15), but the generator efficiency is only important to calculate the electrical power from the generator torque and thus should not be a part of Equation (2) and thus also not of Equation (15). Please make this part more consistent.
 Equation (21): If “constant power” is used, one can also include the partial derivative of the generator torque with respect to the generator speed in Equation (5). Neglecting this usually causes a large deviation from my experience, also for the 5 MW reference wind turbine. For the ROSCO controller and for the paper it would be nice, if you could include this part or provide some investigation that in your case this is neglectable.
 Section 5.5: Usually (in the Bladed interface), the tower top foreaft acceleration is a translational degreeoffreedom and thus the integrated signal is the tower top foreaft speed. This has been used in your reference (van der Veen 2012). However, you use the “tower top pitch angle” (i.e. rotational DOF), which is also possible (but much harder to measure/estimate in reality) and would provide similar results I assume. But since ROSCO is using the Bladed interface and aims to reflect the industrial stateoftheart, please consider changing to the translational DOF.
Further, the paper can be improved using consistent and accurate mathematical expressions:
 Equation (5) etc.: Please use consider using $\delta$ instead of $d$ for $\beta$, $v$ etc. and introducing that $\Delta$ is the deviation from the steady state. Using simply $d$ might cause confusion with the operator “d”.
 Equation (9) etc: Please consider that “d” is an operator and thus using $\mathrm{d}$ instead of simply $d$ would be more appropriate.
 Equation (6) etc: the tip speed ration might be good to introduce. And here, using $\partial$ as in Equation (4) would be more appropriate, since the tip speed ratio depends on both wind speed and generator speed.
 Equation (12): The transfer function $H(s)$ connects the Laplace transform of the input to the one of the output. The Laplace transforms themselves do not depend on $s$. Thus, the fraction with d\Omega_g(s) is a bit sloppy. Best might be to simply remove this and explain that the transfer function is obtained by using the Laplace transform and Equation (5) and (9).
Minor issues:
 Figure 2 and Figure 5 caption, Appendix A etc.: Units are in nonitalic in the rest of the paper (which makes sense, since they are not variables), but here you have $kNm$, $MNm$, etc.. Please consider changing them.
 Figure 3: setpoint smoother has more inputs than only the generator speed.
 l 297: “but the power is much more consistent” is not clear to me. Maybe just remove or add something to better explain it. Maybe you mean “more consistent compared to the constant torque case”?
 in Equation (20) you use “rat” as subscript, but in the rest of the paper “rated”. Please consider to have this consistent.
 Section 3.1, last sentence: From my perspective, the proportional and integral gains for the torque PI controller are often chosen to be constant for simplicity, since applying Equation (13) usually does not provide significantly differences over the considered operation points. Please check if this could be also helpful here. The reason provided in the paper (“less erratic control actuation…”) seems to be a bit vague for a Journal paper.
 l 377: Equation (17) should be included here since Equation (16) is TSR tracking only.

RC2: 'Comment on wes202119', Anonymous Referee #2, 08 Jul 2021
Congratulations for the nice work and the fine manuscript! The wind community will definitely benefit from the ROSCO controller and toolbox. The paper has a clear and logic structure and is nicely written.
The detailed comments from the first reviewer are very pertinent and I agree with them but will not repeat them here.
My main comment after going through the very extensive manuscript is: while undoubtedly the ROSCO controller and toolbox are a significant contribution to the community and represent a very large effort, the authors should also consider highlighting the scientific contributions of the paper, avoiding reading the article as a “ROSCO user manual”. It seems that most of the effort has been towards implementing well known methods and approaches, and automatizing some of the processes, which is of course a great effort that merits appreciation, however it does not automatically and necessarily entail publication in a scientific journal.
 AC1: 'Response to Reviewer Comments', Nikhar Abbas, 17 Aug 2021
Nikhar Abbas et al.
Nikhar Abbas et al.
Viewed
HTML  XML  Total  BibTeX  EndNote  

831  705  13  1,549  23  7 
 HTML: 831
 PDF: 705
 XML: 13
 Total: 1,549
 BibTeX: 23
 EndNote: 7
Viewed (geographical distribution)
Country  #  Views  % 

Total:  0 
HTML:  0 
PDF:  0 
XML:  0 
 1