The turbine can only control
what it can measure
Accurate nacelle wind measurement, non-contact pitch and yaw position feedback, and drivetrain condition monitoring for wind turbine control.
A small error, applied continuously, is a large number
A turbine controller yaws and pitches against the wind measurement on the nacelle roof. It has no independent way to know whether that measurement is right, so it acts on it whatever it says.
That makes the failure mode quiet rather than dramatic. A sensor reading two degrees off does not trip an alarm. It produces a turbine that spends its life pointing slightly the wrong way, losing a small percentage of available energy every hour it runs.
The other failure is worse and equally quiet. A cup and vane anemometer that ices or seizes stops rotating and reports calm. The controller believes it, derates or shuts down, and the loss is only visible if somebody compares the machine against its neighbours.
What the controller is actually asking the sensor
Nacelle wind measurement is not a general-purpose measurement problem. The controller cares most about accuracy close to the datum, because that is where yaw decisions are made, and about the reading remaining valid through icing, salt and a decade between crane visits.
The WindObserver 75 states a turbine-control accuracy of 1 percent within 25 degrees of the datum, with 1 degree on direction, in a heated 316 stainless open head. The open head is the reason: it holds accuracy as the angle of flow increases, which is the condition a yawing turbine spends its life in.
Around that sit the other control inputs. Blade pitch and yaw position feedback, which run continuously and cannot be a wear item. Drivetrain oil condition, where the failure that needs a crane announces itself in the oil first. And a site mast that provides the only independent check on whether the nacelle sensor is still honest.
Where a turbine controller gets its picture
A turbine controller acts on measurements taken from the nacelle roof, the rotor, the yaw ring and the drivetrain below, with a site mast as the only independent reference.
Nacelle anemometer
This is the reading the controller yaws and pitches to. An error here is not a data problem, it is lost generation on every hour the turbine runs.
- WindObserver 75 turbine-control accuracy 1 percent within 25 degrees of datum, 1 degree direction
- WindObserver 90 0 to 90 m/s where the site sees extreme gusts
- WindSonic M heated aluminium, operating to -40 degrees C
Secondary and backup sensing
A single nacelle sensor has no way to tell you it is wrong. A second one disagreeing with it does.
- WindObserver 70 heated head, RS422 and RS485
- WindSonic 60 0 to 60 m/s, 0.5 kg, low additional nacelle loading
- WindMaster Pro 3-axis where inflow angle also matters
Rotor hub and pitch system
Pitch actuators run millions of cycles across a turbine life, in a hub nobody reaches without stopping the machine.
- Blade position sensors non-contact inductive, no wearing parts
- Blade60 contactless linear position feedback
- Blade Dual Output a second independent channel where redundancy is required
Blade root and pitch bearing
Blade pitch angle is a control input and a load-shedding mechanism, so the feedback has to be right in the conditions that make it matter.
Yaw ring and drive
Yaw is the slowest and most expensive control action a turbine takes, so it is only taken when the measurement says it is worth it.
- Blade position sensors non-contact rotary feedback on the yaw drive
- WearDetect oil debris sensing in the yaw and pitch gearboxes
- LevelPro lubrication level with no moving parts
Tower base controller
The measurement is only useful once the controller has it, in the protocol the controller already speaks.
- WindObserver 75 RS422 and RS485, 1200 to 38400 baud, Modbus RTU variant
- MetConnect One Modbus RTU and ASCII, SDI-12 v1.3, NMEA 0183
- DD1615 level display local indication at the tower base
Site met mast reference
The mast is what tells you whether the nacelle sensor is still telling the truth, which is the only independent check the turbine has.
- WindObserver 75 the long-duration reference at hub height
- WindMaster Pro 3-axis to 32 Hz for turbulence intensity
- MetConnect THP air density from temperature, humidity and pressure
Calibration and verification
Power curve verification and availability warranties are argued from measured wind. Traceable calibration is what makes that measurement admissible.
- UKAS-accredited air velocity calibration traceable certification for anemometers
- Multi-parameter calibration six parameters, equipment from any manufacturer
- Recalibration scheduling intervals, reminders and turnaround
Accuracy where the control decision is made
Yaw and pitch decisions are taken close to the datum, so accuracy at small misalignment angles is worth more than headline range. The WindObserver 75 carries a turbine-control accuracy of 1 percent within 25 degrees of the datum and 1 degree on direction.
It measures 0 to 75 m/s in a 316 stainless open head at IP66, heated, with no moving parts anywhere in the measurement.
1 percent within 25 degrees of datum
Explore wind sensorsPitch and yaw feedback with nothing in contact
Pitch and yaw actuators cycle continuously across a twenty-year life, in a nacelle and hub that come out of service only when a crane is already booked.
The Blade range is non-contact inductive position sensing in linear and rotary formats, with a dual-output variant where redundancy is required. There are no wearing parts in the measurement, so the feedback is not on the maintenance schedule.
Non-contact. Nothing wears in the measurement.
Explore position sensorsOne nacelle sensor cannot tell you it is wrong
A single measurement has no way to report that it has drifted, iced or seized. It reports something plausible, and the controller acts on it.
Cross-checking is what turns that from an invisible loss into a detectable fault. That can be a second nacelle sensor, a comparison against the site met mast, or a comparison against neighbouring turbines in the same conditions. The right architecture depends on the turbine, the site and what the operator can act on.
Our role is to supply measurement that can form part of that architecture, chosen for the environment, the accuracy the control loop needs and how the data has to be integrated.
More than the nacelle anemometer
Wind measurement, position feedback and machinery condition sensing for the turbine and the site around it.
Wind measurement
WindObserver
Heated 316 stainless open-head 2-axis anemometer
The nacelle and mast workhorse, with a stated turbine-control accuracy and no moving parts in the measurement.
WindMaster
3-axis ultrasonic anemometer
For the site mast, where turbulence intensity and inflow angle matter as well as speed and direction.
WindSonic
Lightweight 2-axis ultrasonic anemometer
For a second nacelle reference or a temporary works mast, where 0.5 kg and simple mounting matter.
Position and condition
Blade position sensors
Non-contact inductive position sensing
Linear and rotary feedback with nothing in contact, for actuators that cycle for the life of the turbine.
WearDetect
Oil debris sensor
Monitors gearbox and bearing oil for ferrous wear particles, so the failure that needs a crane gives warning first.
Level sensing
Capacitive liquid level sensors
Lube and hydraulic level in the nacelle and tower base, with no float to stick between service visits.
Where turbine control sensing does the work
Keeping the rotor on the wind
Continuous yaw alignment against the nacelle measurement, where accuracy close to the datum decides how much of the available energy is captured.
Load shedding and power regulation
Blade pitch as both a power regulation and a load-shedding mechanism, with position feedback that has to stay valid for the life of the machine.
Protecting the machine
High wind cut-out and restart decisions taken on a measurement that must not report a false calm or a false gust.
Operating through icing
Heated sensing and icing-tested construction, so the control loop keeps a valid reading in the conditions where a mechanical sensor stops turning.
Proving performance
Measured wind against measured output, traceable enough to support a warranty claim or defend against one.
Planning the crane
Drivetrain oil condition and fluid level, moving a gearbox failure from an unplanned outage into a planned one.
UKAS-accredited calibration for six measurement parameters
Labcal is an independent UKAS-accredited calibration laboratory, Lab 0625, operating to ISO/IEC 17025:2017. It provides calibration across gas and air flow, air velocity, liquid flow, pressure, temperature and humidity, all under one roof.
For turbine measurement that includes anemometers, Pitot tubes and thermal probes, alongside pressure gauges and transducers, thermometers, temperature and humidity probes and data loggers. Labcal supports equipment from Gill and other manufacturers, so a site can consolidate an instrument schedule that was never bought from one supplier.
01 · Gas and air flow
Gas meters · Mass flow controllers · Dispensing systems · Low-flow laboratory to higher-flow industrial
02 · Air velocity
Anemometers · Pitot tubes · Thermal probes
03 · Liquid flow
Flow meters and sensors · Process control equipment · Water used for the accredited capability
04 · Pressure
Gauges · Transducers · Manometers · Transmitters
05 · Temperature
Thermometers · Probes · Thermocouples · Data loggers
06 · Humidity
Hygrometers · Probes · Transmitters
Measurement only matters when the controller can use it
A nacelle measurement is read by a control loop, not by a person.
Gill supports environmental monitoring, industrial sensing, embedded electronics and systems integration to help turbine OEMs, operators and control suppliers bring reliable measurement into turbine controllers and SCADA.
Whether the requirement is Modbus into an existing controller, serial into a logger, or a bespoke embedded solution, Gill engineers can support integration from specification through commissioning.
Outputs and options
- RS422 / RS485 : Serial into turbine controllers
- Modbus RTU : Into SCADA and asset management
- Heated variants : Cold climate and icing conditions
- Analogue & Pt100 : Bringing existing instruments into one station
Why turbine programmes specify Gill
Accuracy where it earns money
The WindObserver 75 states a turbine-control accuracy of 1 percent within 25 degrees of the datum, with 1 degree on direction.
No moving parts to seize
Ultrasonic measurement removes the bearings that ice, foul and produce a false calm on a cup and vane sensor.
Built for the crane problem
316 stainless at IP66, heated, icing tested to MIL-STD-810F, for a sensor that is only reachable when the machine is stopped.
More than the anemometer
Non-contact pitch and yaw position feedback, oil debris monitoring for the drivetrain, and capacitive level for the fluids.
Designed to integrate
RS422, RS485 and a Modbus RTU variant, so the reading arrives in the format the controller already accepts.
Evidence when the yield is questioned
Independent UKAS-accredited calibration across six parameters, for equipment from any manufacturer.
Questions about sensing for wind turbine control
The WindObserver 75 is the model with a stated turbine-control accuracy: 1 percent within 25 degrees of the datum, and 1 degree on direction. It measures 0 to 75 m/s in a heated 316 stainless open head.
It holds accuracy as the angle of the flow increases. A yawing turbine is almost never exactly aligned, so the sensor is measuring off-axis flow most of the time.
A cup and vane sensor stops rotating and reports calm. The controller has no way to know, so it derates or shuts down while the wind is still blowing. An ultrasonic sensor has no rotating element, and the heated WindObserver models keep the transducer paths clear.
WindObserver heating draws up to 7 A at 24 V AC or DC. That is a cabling and supply decision at design stage rather than a footnote, particularly on a retrofit.
A single sensor cannot report that it is wrong. A second sensor, or a comparison against the site mast or neighbouring turbines, is what turns a silent drift into a detectable fault. The right architecture depends on the machine and what the operator can act on.
Because pitch and yaw actuators cycle continuously for twenty years in a location that needs a crane to reach. The Blade range is non-contact inductive in linear and rotary formats, with no wearing parts in the measurement.
Earlier detection. Vibration signatures change once a surface has degraded. Ferrous wear particles are in the oil before that point, which is the difference between planning a crane and being given a date.
Most take Modbus or a serial input. The WindObserver range outputs RS422 and RS485, full or half duplex, at 1200 to 38400 baud, with a Modbus RTU variant available.
Usually not for control. Two axes give the speed and direction the controller acts on. Three axes matter on the site mast, where turbulence intensity and inflow angle explain why the site is or is not meeting its yield model.
Set by the operator and by what the measurement supports. Where a reading feeds power curve verification or an availability warranty, annual traceable calibration is common.
Labcal is UKAS-accredited to ISO/IEC 17025:2017 and calibrates air velocity, temperature, humidity, pressure, and gas and liquid flow.
Speak to our team
Tell us what you need to measure, where and under what conditions. We’ll help you work out the appropriate approach.
We aim to respond in one working day.