Wind farms are usually run to squeeze out as much electricity as possible, but a new control strategy argues that the smartest way to operate a farm is to maximize profit over its entire life while explicitly guaranteeing that the turbines reach a target lifetime. Researchers at the Wind Energy Institute of the Technical University of Munich, led by Carlo L. Bottasso, have introduced what they call economic lifetime-aware wind farm control, published in the journal Wind Energy Science. In simulation case studies, their formulation delivered the highest profit among six tested control strategies and was the only approach that, together with a revenue-only variant, reliably ensured the farm could operate through its full desired 20-year life.
The method rests on two innovations. First, the optimization objective is an economic model of profit rather than power yield. Profit accounts not only for revenue from selling electricity, potentially at variable spot-market prices, but also for the costs of operation and maintenance that arise as turbines accumulate fatigue damage. Second, the formulation treats fatigue damage as a finite budget: because damage accrues over decades, any control action taken today changes how much room remains for future actions. The optimization therefore includes explicit constraints on accumulated damage, ensuring that critical components such as blade roots and tower bases can reach a desired goal duration without exceeding a predefined damage threshold.
This coupling is what distinguishes the approach from most existing wind farm control. Standard strategies optimize set points separately for each combination of wind speed, direction, and turbulence intensity, treating every condition as independent. The Munich team instead bins the ambient conditions with their probabilities of occurrence and solves one large optimization that considers all bins simultaneously. That single-shot formulation is computationally heavier, but it captures how aggressive steering in one wind condition consumes fatigue budget that would otherwise be available in another. Once solved offline, the optimal yaw set points are stored as look-up tables and dispatched at run time in an open loop based on detected conditions, keeping real-time execution fast and verifiable for safety.
The framework links several models. A steady-state engineering flow model, built on the open-source FLORIS toolbox with a Gaussian wake deficit model and an upgraded wake-added turbulence formulation, estimates power and the inflow at each rotor. A neural-network load surrogate, trained on 1,908 aeroservoelastic FAST.Farm simulations of three IEA 3.4 MW reference turbines, converts rotor-averaged wind speed, turbulence intensity, and horizontal shear into damage-equivalent loads for the blade root and tower base. Those loads are aggregated into fatigue damage, which feeds a simplified economic model combining electricity revenue, priced through a polynomial correlation with wind speed fitted to German spot-market data, with maintenance costs amortized over the damage budget.
The first case study is deliberately idealized: three turbines in an aligned row with six-rotor-diameter spacing, exposed to a synthetic wind rose spanning a narrow 40-degree sector that produces very frequent waking. The scenario assumes the farm has already run for half of its 20-year target life in standard greedy mode, and that one turbine has accumulated 25 percent more damage than anticipated. Against a greedy baseline and five alternative formulations, including classical power-boosting wake steering and load-constrained and load-balance variants, the researchers evaluated power, loads, lifetime consumption, and discounted lifetime profit.
The results were striking. Under greedy, max-power, and load-constrained control, the farm failed to reach its goal end of life, losing operational years because the over-damaged turbine kept experiencing high loading. The max-power strategy yielded the shortest life and only a 1.7 percent profit gain over greedy, while the load-constrained strategy gained 7.7 percent. The load-balance strategy extended life beyond the goal, but the authors caution that because it does not enforce a lifetime constraint, this outcome cannot be expected in general. Only the lifetime-aware formulations guaranteed full operation through the desired duration. The proposed max-profit strategy achieved roughly 30 percent higher discounted profit than the greedy baseline over the 20-year horizon, exceeding the revenue-only lifetime-aware variant by about 4 percent, and by an additional 13 percent when extended operation beyond the goal end of life was included.
To check the reliability of the framework, the team dispatched the computed set points to the higher-fidelity FAST.Farm simulator, treated as ground truth. Predicted power matched simulated power closely, and damage-equivalent load trends were captured with good accuracy, with tower loads slightly underpredicted in partial-wake and overpredicted in full-wake conditions. These modest mismatches had only a minimal effect on predicted lifetime consumption and profit. A Monte Carlo uncertainty analysis varying thirteen economic parameters by 8 percent shifted lifetime profit by only 3 to 4 percent, with electricity market price uncertainty the largest single contributor, followed by mean time to failure.
The second case study scaled up to a nine-turbine offshore cluster from the CL-Windcon project, arranged in a regular grid and driven by reanalysis wind data from the Hollandse Kust Nord site, with wind speeds between 6 and 10 meters per second and an approximate 5 percent wake loss in greedy mode. Here one turbine was assumed to have accrued 75 percent excess damage after ten years. Again, only the lifetime-revenue and max-profit strategies ensured operation through the entire 20-year goal, with max-profit delivering an 8.6 percent profit gain over greedy, of which 1.3 percent was attributed specifically to accounting for maintenance expenses in the optimization. The max-power, load-constrained, and load-balance strategies produced only minor gains because of lost lifetime.
The authors are candid about limitations. The economic model is a simplified proof of concept with a crude link between loads and maintenance costs, the study used wake steering alone rather than also allowing induction control or derating, and components were treated as irreplaceable for simplicity. A preliminary test of time discounting across two five-year epochs showed the formulation can accommodate evolving control policies, but produced negligible behavioral changes in this particular case, and the authors stress that broader studies are needed before general conclusions. The one-shot optimization is also computationally expensive, with cost growing steeply with turbine count, although it runs offline on a standard desktop and could be accelerated through parallelization and newer software versions.
Even so, the study makes a clear quantitative case: control strategies that ignore how fatigue damage couples across wind conditions and across years can silently burn through a turbine’s structural budget, shortening a farm’s life and eroding revenue, while a profit-based objective with an explicit lifetime constraint both protects the asset and, in these simulations, earns the most money. The researchers suggest the framework could also serve wind farm developers before deployment, weighing, for example, a shorter, higher-yield operating life against a longer, gentler one. An experimental wind-tunnel validation of the underlying strategy with three model turbines has already been reported by the group, and the simulation data from this study are openly available.
Subject of Research: A simulation study of an economic lifetime-aware wind farm control method that optimizes profit while enforcing fatigue damage constraints.
Article Title: Economic lifetime-aware wind farm control
Article References: Anand, A., Braunbehrens, R., Guilloré, A., & Bottasso, C. L. (2026). Economic lifetime-aware wind farm control. Wind Energy Science, 11(10), 3949-3981. https://doi.org/10.5194/wes-11-3949-2026
Image Credits: AI Generated
Keywords: wind farm control, wake steering, fatigue damage, lifetime constraint, profit optimization, operation and maintenance, wind energy, control set points, spot market prices, simulation study, Economic, lifetime-aware
News Source: Alan Morgan. (October 28, 2026). Wind Farm Controller Targets Profit and Guarantees 20-Year Turbine Lifetime. Scienmag.
