Every row turns on
the gust you measured
A single-axis tracker stows on one number. Wind sensing for the stow decision, the rows the wind reaches first and the feedback that confirms every row arrived.
Stow too early and the sun is wasted.
Stow too late and the steel is.
A single-axis tracker turns its rows to follow the sun and, when the wind rises past a threshold, drives them to a stow angle the structure can survive. The threshold is written by the tracker manufacturer’s structural engineers. The decision to use it is made by a controller comparing that threshold with a wind measurement, many times a minute, all year.
Get it wrong one way and generation disappears. A study by DNV for Array Technologies put the average annual energy lost to active wind stow at 2.8 per cent, and up to 4.3 per cent on some sites. Every hour in stow is an hour the rows are not following the sun.
Get it wrong the other way and the cost is structural. When wind drives a row into torsional galloping, the motion grows with every cycle and stops only when the wind falls well below the speed that started it. Both errors trace back to one place: a controller has no independent way of knowing whether the wind number it was given is right.
The threshold is engineering. The measurement is what makes it real.
A tracker controller compares measured wind speed with a stow threshold set by the tracker manufacturer, usually defined as a short gust or a short average rather than an instantaneous reading. When the measured value crosses the threshold, the controller drives the rows it governs to their stow angle and holds them there until the wind has stayed below a lower exit threshold for a set dwell time.
The definition of the gust decides the outcome. A threshold written against a 3-second gust means nothing if the sensor reports a 1-minute average, or a different gust each time. The averaging has to match what the structural engineer specified, the same way every time.
Where the wind is measured decides who is protected. The outer rows and site edges catch the wind first, and on sloping or uneven ground an exposed corner can see a very different wind from the middle of the array. A sensor has to represent the rows it controls.
What happens when the sensor fails decides everything else. A seized cup anemometer does not report a fault. It reports calm, and a controller that believes calm keeps the rows tracking into a storm. Snow, hail and flood stow are increasingly triggered by forecasts and alerts as well; wind is the case where the site’s own sensor makes the call in real time.
Where a tracker site decides to stow
The stow decision is made on one wind number, confirmed by position feedback and explained afterwards by the site station. Each point below shows what is measured there and which sensing fits.
Irradiance reference
This is the denominator of the performance ratio. Measured badly, it moves the reported performance of the entire asset without a single module changing.
- MaxiMet GMX301 solar radiation only, 0 to 1600 W/m², 0.6 kg, for a dedicated reference node
- MaxiMet GMX501 solar radiation alongside wind, temperature, humidity and pressure in one unit
- MaxiMet GMX551 the same again with rainfall, where soiling recovery has to be explained
Site weather station
One unit, one cable, one output string. The fewer separate instruments on the mast, the fewer things to align, power, calibrate and climb to.
- MaxiMet GMX501 wind, temperature, humidity, pressure and solar radiation, 0.8 kg, IP66
- MaxiMet GMX551 adds rainfall by optical or tipping bucket input
- MetConnect where sensors have to sit at separate heights for WMO-compliant siting
Tracker row wind and the stow decision
The controller stows on this number and has no independent way to know it is wrong. A seized cup and vane does not report a fault, it reports calm.
- WindSonic 60 0 to 60 m/s, 0.5 kg, WMO-compliant gust measurement
- WindSonic 75 0 to 75 m/s where the site sees higher extremes
- WindSonic M heated hard-anodised aluminium, operating to -40 degrees C
Exposed rows and site edges
A single sensor in the middle of a sloping site does not represent the corner that catches the wind first. How many sensors a site needs depends on its size and its terrain.
- WindObserver 75 0 to 75 m/s in a heated 316 stainless open head, IP66, for the most exposed position
- WindObserver 90 0 to 90 m/s where the extremes justify it
- WindUltra 270 g, IP69k, under 100 mW without heating in calm conditions, for self-powered rows
Tracker angle feedback
A stow command is only a command. Position feedback is what confirms the row reached the angle, and a contacting sensor on a mechanism that cycles daily is a wear item.
- Blade360 full 360 degree rotary, configurable range, IP67, sealed for temporary immersion (1 m for 30 min)
- Blade60 60 mm linear or 90 degree angular, analogue and digital outputs
- Blade Dual Output single and dual channel output where the feedback has to be redundant
Rainfall and soiling
Soiling losses are only arguable if you know when it last rained and how hard. A tipping bucket with a moving part is one more thing on the site that needs a visit.
- TruMet PW100 infrared optical rain gauge, no moving parts, pairs with GMX550 and GMX551
- MaxiMet GMX551 rainfall by optical or tipping bucket alongside solar radiation
- MaxiMet GMX603 optical rainfall 0 to 200 mm/hr at 5 per cent or better, integrated
The wind half of a hybrid site
On a hybrid connection the wind sensor stops being only a safety input and starts explaining generation, which raises what the measurement has to support.
- WindObserver 75 heated 316 stainless, 0 to 75 m/s, for a permanent site mast
- WindMaster Pro 3-axis, where turbulence intensity and inflow angle matter as well as speed
- WindUltra 0 to 75 m/s, IP69k, heated as standard, for a remote or unattended mast
Calibration and verification
When a yield number is questioned, the instrument behind it is questioned with it. Traceability is what ends that conversation rather than extending it.
- Labcal air velocity UKAS-accredited calibration of anemometers, Pitot tubes and thermal probes
- Labcal temperature and humidity thermometers, probes, transmitters and data loggers
- Labcal pressure gauges, transducers, manometers and transmitters
A gust defined the same way, every time
WindSonic measures wind ultrasonically, with no bearings to ice, foul or seize, and produces a WMO-compliant gust from a rolling average, so the number compared with the stow threshold is defined consistently. Every output carries an instrumentation status code, which is the difference between a controller that knows its sensor has failed and one that believes a dead sensor reporting calm.
Outputs suit the controller rather than the other way round: RS232, RS422 and RS485, Modbus RTU, SDI-12, and 0 to 5 V, 0 to 20 mA or 4 to 20 mA analogue, depending on model and option. Accuracy is 2 per cent RMSE at 12 m/s, with 2 degrees RMSE on direction.
WindSonic 60 0 to 60 m/s, 0.5 kg. WindSonic 75 0 to 75 m/s, for sites with higher extremes. WindSonic M heated hard-anodised aluminium, operating to -40 °C.
A status code on every output, so a failed sensor cannot read calm
Explore WindSonicThe corner that catches the wind first should be the one you measure
How many wind sensors a tracker site needs depends on its size, its terrain and how its rows are grouped into control zones. A flat, compact site may be governed well from a small number of positions. A large or sloping site, or one with an exposed edge facing the prevailing storm direction, needs sensors where the wind arrives first, and each sensor should govern only the rows it genuinely represents.
WindObserver 75 0 to 75 m/s in a heated 316 stainless steel open head, IP66, for the most exposed mast. WindObserver 90 0 to 90 m/s, where the site’s extremes justify it. WindUltra 270 g, IP69k, under 100 mW without heating in calm conditions, for row-level sensors on self-powered controllers.
A stow command is only a command
The controller knows what angle it asked for. Position feedback is what tells it every row got there, including the row with a failed motor or a jammed drive. On a mechanism that moves every day of its life, a contacting sensor is a wear item; a non-contact sensor measures the angle without touching the shaft it measures.
Blade360 full 360 degree rotary, configurable range, IP67. Blade Dual Output a redundant second channel from one sensor, where the feedback has to be proven.
Confirm the angle, not just the command
Explore position sensorsQuestions about solar tracker wind stow
A solar tracker controller compares a measured wind speed with a stow threshold set by the tracker manufacturer. The threshold is normally defined as a short gust or short average. When the measured value crosses it, the controller drives the rows it governs to a stow angle the structure is designed to survive, and returns them to tracking once the wind has stayed below a lower exit threshold for a set time. The decision is only as reliable as the wind measurement and its gust definition.
The stow wind speed for a solar tracker is set by its manufacturer and depends on the tracker’s structural design, row length, module size and stow angle. Published examples range from around 12 m/s for some designs with large modules to 16 m/s for a conventional tracker and 22 m/s for some rigid designs that stow flat. The threshold for a specific site should always come from the tracker manufacturer, and the wind sensor should report the gust in the same way the threshold defines it.
The number of wind sensors a solar tracker site needs depends on the size of the site, its terrain and how its rows are grouped into control zones. Each sensor should represent the wind reaching the rows it controls, so exposed edges, corners facing the prevailing storm direction and sloping ground usually need their own measurement. Many sites also add redundancy, so the loss of one sensor does not leave a zone without protection.
Yes. Rows in stow are not following the sun, so every hour in stow loses generation. A DNV study commissioned by Array Technologies put the average annual energy loss from active wind stow at 2.8 per cent, and up to 4.3 per cent on some sites. Modelling published by Arctech showed losses rising from under 1 per cent to over 4 per cent of projected generation as the stow threshold fell from 16 m/s to 12 m/s. A wind measurement the site can trust allows a threshold that protects the structure without stowing more often than the design requires.
An ultrasonic anemometer measures wind speed and direction from the time sound takes to cross a known path, so it has no bearings or rotating parts to ice, foul or seize. That matters for tracker stow because a seized mechanical sensor typically reports calm rather than a fault. Gill ultrasonic sensors add an instrumentation status code to every output, so a controller can tell a genuine calm from a failed sensor. Ultrasonic sensors still need correct siting, sensible mounting and periodic verification.
A solar tracker can take its stow wind from a site weather station if that station measures where the wind reaches the controlled rows, reports the gust as the stow threshold defines it and is fast enough for the control loop. On small, uniform sites one compact station can serve both reporting and stow. On larger or more exposed sites, dedicated wind sensors at the edges and in each control zone usually represent the rows better, with the weather station providing context for performance reporting.
Sizing the wind sensing for a tracker site?
Tell us the tracker, the control zones and the site’s exposure. We will tell you plainly what we would measure, where, and with which output, before the masts go up.
We aim to respond in one working day.