Reconstructing eleven years of near-surface wind conditions along the Caribbean coastline of Magdalena and Atlántico in Colombia reveals that improving a small vertical-axis wind turbine’s low-speed operating threshold expands operational availability more effectively than raising tower heights. Conducted by Carlos Robles-AlgarÃn, VÃctor Olivero-Ortiz, and Adalberto Ospino Castro and published in Results in Engineering, the study established an inverse, duration-aware framework using hourly ERA5-Land reanalysis data from 2015 through 2025 across twenty-four coastal land nodes within twenty kilometers of the shoreline. The researchers determined that for a baseline operating threshold of 3.0 meters per second at a ten-meter height, reducing the threshold by half a meter per second yielded an 8.66 percentage point median gain in operational wind availability across the domain. In contrast, elevating hub height from ten to thirty meters provided a median gain of only 6.55 percentage points, demonstrating that low-speed turbine responsiveness can outweigh structural height investments in low-elevation coastal settings.
Small wind turbines occupy an engineering space distinct from utility-scale installations because they operate within the lowest tens of meters of the atmospheric boundary layer. In this near-surface zone, encompassing coastal, peri-urban, and rural environments, atmospheric flow features lower mean speeds, stronger turbulence, and rapid directional shifts. While horizontal-axis wind turbines dominate centralized power generation, vertical-axis wind turbines offer practical advantages through omnidirectional operation and compact installation geometry. Drag-driven Savonius rotors are especially relevant for low-speed distributed applications due to favorable self-starting behavior, despite relatively modest peak power coefficients and torque fluctuations. Existing literature frequently emphasizes rotor geometry, tip-speed ratios, and peak aerodynamic efficiency. However, the authors emphasized that peak efficiency alone does not ensure field effectiveness; inertia, control, and responsiveness to rapidly fluctuating inflows determine whether assembled machines can self-start, sustain rotation, and deliver useful energy under marginal winds.
The Colombian Caribbean possesses notable wind resources, yet previous assessments primarily emphasized utility-scale horizontal turbines, aggregate energy potential, and coarse climatologies. Although earlier studies documented seasonal variability across the region, they overlooked the low-height operating constraints of small vertical-axis systems. A major limitation of conventional assessments is their reliance on mean wind speeds and standard frequency distributions, which fail to capture the persistence of unfavorable operating states. For small decentralized installations and battery-charging systems, multi-hour or multi-day intervals below operational cut-in thresholds cause repeated start-stop cycles, mechanical stress, and prolonged power interruptions. To address this gap, the researchers developed a duration-aware screening framework that translates atmospheric records into explicit, testable operating thresholds. Crucially, the methodology remains turbine-agnostic, defining climate-side operating requirements before imposing specific commercial power curves, cut-in velocities, or prototype electromechanical characteristics.
The spatial domain comprised twenty-four coastal nodes across Atlántico and Magdalena, selected using a twenty-kilometer marine coastline mask and an effective elevation limit of three hundred meters to exclude the mountainous Sierra Nevada de Santa Marta. Hourly ten-meter horizontal wind components from ERA5-Land provided eleven years of continuous near-surface data from 2015 to 2025. To reconstruct wind speeds at twenty and thirty meters without assuming fixed wind-shear power laws, the team extracted hourly ten-meter and one-hundred-meter wind fields from ERA5 reanalysis. Monthly power-law shear exponents were computed from monthly mean wind speeds at both heights to avoid mathematical instability during near-calm periods. These monthly exponents were bilinearly interpolated to ERA5-Land node coordinates to reconstruct hourly speeds at twenty and thirty meters, capturing seasonal shear variations while retaining the fine spatial resolution of the near-surface ERA5-Land dataset.
The operational framework inverted empirical wind distributions to determine the maximum operating threshold compatible with prescribed availability targets of 50, 60, 70, 80, and 90 percent. Across the twenty-four nodes, long-term mean wind speed rose from 3.77 meters per second at ten meters to 4.11 meters per second at twenty meters and 4.33 meters per second at thirty meters. At a reference threshold of 3.0 meters per second, median availability increased from 58.30 percent at ten meters to 62.02 percent at twenty meters and 64.16 percent at thirty meters. Evaluating a 70 percent availability target, a typical median coastal site required a maximum compatible operating threshold of 2.33 meters per second at ten meters, 2.52 meters per second at twenty meters, and 2.63 meters per second at thirty meters. To ensure 70 percent availability across approximately 90 percent of analyzed locations, conservative threshold targets dropped to 1.35, 1.48, and 1.57 meters per second across the respective heights.
Comparing design interventions revealed that lowering the operating threshold outperformed moderate height increases in both availability gains and low-wind spell mitigation. Relative to a ten-meter installation operating at 3.0 meters per second, raising height to twenty meters produced a median site-specific availability gain of 4.19 percentage points and reduced the median annual maximum low-wind spell from 118 to 87 hours, a 30.5-hour reduction. Raising height to thirty meters yielded a 6.55 percentage point gain and reduced maximum spell duration by 44.5 hours. In contrast, reducing the ten-meter threshold to 2.5 meters per second generated an 8.66 percentage point availability gain and a 53.5-hour spell reduction. Lowering the threshold to 2.0 and 1.5 meters per second expanded availability by 17.98 and 27.43 percentage points while cutting maximum spells by 78.5 and 96.5 hours. Equivalent-threshold calculations showed that elevating a turbine to twenty or thirty meters matched threshold reductions of merely 0.24 and 0.37 meters per second.
The advantage of threshold reduction proved especially pronounced during the low-resource autumn season. Median monthly ten-meter wind speeds reached 5.90 to 6.51 meters per second during January through March but fell to 2.23 and 2.21 meters per second in September and October. During September and October, raising installation height from ten to thirty meters increased median availability by 6.65 and 6.82 percentage points, whereas lowering the ten-meter threshold from 3.0 to 2.5 meters per second delivered gains of 12.00 and 12.10 percentage points, with a 2.0 meters per second threshold exceeding 26 percentage points. Furthermore, the threshold advantage persisted across all eleven individual years between 2015 and 2025. Across annual domain-average wind speeds ranging from 3.43 meters per second in 2024 to 4.16 meters per second in 2015, the 2.5 meters per second threshold outperformed the thirty-meter tower increase in every annual realization by 1.51 to 2.49 percentage points.
Spatial clustering using k-means identified three recurring operational regimes across the coastal corridor. Six nodes located entirely in Magdalena formed a low-resource, high-persistence regime characterized by a mean 20-meter wind speed of 2.37 meters per second, 28.89 percent availability above 3.0 meters per second, and an extensive average annual maximum low-wind spell of 579.5 hours. Nine nodes, primarily in Atlántico, constituted an intermediate-resource regime with a mean 20-meter speed of 4.06 meters per second, 57.40 percent availability, and the highest relative height sensitivity, exhibiting a 17.92 percent wind speed gain between ten and thirty meters. The remaining nine nodes formed a medium-high-resource regime distributed across both departments, featuring a mean speed of 5.34 meters per second, 71.54 percent availability, and a 64.2-hour maximum calm spell. These groupings indicate that low-resource environments require prioritized low-speed operating capability, whereas intermediate zones derive the greatest operational returns from taller towers.
Methodological sensitivity assessments confirmed that core trade-offs were robust to analytical choices. Modifying the domain elevation cutoff between 100, 200, 300, and 500 meters shifted median ten-meter availability by less than 1.2 percentage points, validating the three-hundred-meter filter as a reliable definition of coastal screening conditions. Testing four alternative vertical-transfer schemes—combining monthly or long-term shear with nearest-cell or bilinear ERA5 interpolation—produced domain median availabilities of 61.49 to 62.32 percent at twenty meters and 63.26 to 64.66 percent at thirty meters. Median availability gains for a ten-to-thirty-meter height increase ranged from 5.00 to 6.90 percentage points across the four transfer options. Because the ten-meter ERA5-Land baseline is independent of vertical interpolation, the 8.66 percentage point gain achieved by lowering the threshold to 2.5 meters per second remained larger than the thirty-meter height benefit across all four vertical-profile formulations.
Ground evaluation at Isla Salamanca and downscaled Vortex TIMES benchmarks established the operational boundaries of the regional framework. Across 5,925 paired hours, the Isla Salamanca 10-meter station recorded a mean wind speed of 2.48 meters per second and 37.72 percent availability above 3.0 meters per second, revealing substantial positive reanalysis biases of +3.02 meters per second for ERA5-Land and +2.43 meters per second for ERA5. Nevertheless, strong temporal co-variability emerged, with daily and monthly Pearson correlations exceeding 0.92 and 0.96. High-resolution Vortex benchmarks demonstrated that hourly aggregation preserved annual availability within one percentage point while muting transition counts. Independent prototype testing of a project-specific Weibull-Savonius rotor yielded a maximum power coefficient of 0.202 near a tip-speed ratio of 0.663, paired with bench generator testing reaching 235 watts. The authors concluded that while their duration-aware framework provides robust screening criteria, site-specific installations require local anemometer measurements and assembled-system testing to account for microscale exposure and electromechanical conversion losses.
Subject of Research: Small vertical-axis wind turbine operating threshold requirements and coastal atmospheric dynamics in Colombia
Article Title: From coastal wind profiles to small-VAWT operating requirements: A duration- and uncertainty-aware framework for Magdalena and Atlántico, Colombia
Article References: Robles-AlgarÃn, C., Olivero-Ortiz, V., & Ospino Castro, A. (2026). From coastal wind profiles to small-VAWT operating requirements: A duration- and uncertainty-aware framework for Magdalena and Atlántico, Colombia. Results in Engineering, 32, Article 114192. https://doi.org/10.1016/j.rineng.2026.114192
Image Credits: AI Generated
DOI: 10.1016/j.rineng.2026.114192
Keywords: small vertical-axis wind turbines, Savonius rotor, coastal wind profiles, Colombian Caribbean, Magdalena, Atlántico, ERA5-Land reanalysis, wind availability, low-wind persistence, atmospheric boundary layer, wind shear exponent, operating threshold, Vortex TIMES downscaling, Isla Salamanca meteorological station, tower height trade-offs, renewable energy screening, power-law wind profile
News Source: Denise Maddox. (November 20, 2026). Lower Operating Speeds Outperform Tower Height for Coastal Small Wind Turbines. Scienmag.



