Solar Pyranometer 2026: Design Guide for Accurate Irradiance Measurement

A solar pyranometer is the single most important sensor in any solar performance monitoring system. It turns sunlight into a measurable electrical signal and gives designers, operators, and lenders the irradiance data they need to calculate energy yield, performance ratio, and warranty claims. Get the pyranometer selection or placement wrong, and every downstream calculation inherits that error.

In 2026, the standards are clear but the field choices are wider than ever. You can buy a $300 silicon-cell sensor, a $3,000 Class A thermopile, or a heated reference instrument built for polar deserts. Each has a place, but only a few configurations meet the bankability requirements that IEC 61724-1 and project financiers expect. This guide explains how to specify, mount, wire, and maintain a solar pyranometer for residential, commercial, and utility-scale projects.

In this guide:

• What a solar pyranometer measures and why it matters for PV design
• Thermopile vs silicon-cell sensors: accuracy, cost, and spectral limits
• ISO 9060 Class A/B/C and IEC 61724-1 monitoring requirements
• How to design pyranometer placement for GHI, POA, trackers, and bifacial arrays
• Outputs, data loggers, wiring, and Modbus integration
• Calibration intervals, cleaning schedules, and drift management
• Common design mistakes that corrupt irradiance data
• How SurgePV turns measured and satellite irradiance into accurate system designs
• 2026 cost ranges by project size

What a Solar Pyranometer Measures

A solar pyranometer measures solar irradiance on a planar surface. Irradiance is the instantaneous power density of sunlight, expressed in watts per square meter. The sensor has a 180-degree field of view, so it sees the entire sky hemisphere above it. That measurement includes direct beam radiation from the solar disk plus diffuse radiation scattered by clouds, aerosols, and the atmosphere.

The most common measurement is global horizontal irradiance, or GHI. A GHI pyranometer sits level and records the total shortwave radiation arriving on a horizontal plane. GHI is the starting point for most yield models and the reference value used to compare one site against another.

A pyranometer can also measure plane-of-array irradiance, or POA. In that case the sensor is mounted on a bracket tilted and oriented exactly like the PV modules. POA is the irradiance the modules actually see. It is the correct input for performance ratio calculations and the value that should be used when validating production against a design model. For a deeper explanation of GHI, DNI, and DHI, see the solar irradiance GHI, DNI, and DHI guide.

Other pyranometer arrangements include:

• Diffuse horizontal irradiance, measured with a shaded pyranometer that blocks the direct beam using a shadow band or tracking disk.
• Albedo, measured with an upward-facing and a downward-facing pyranometer mounted back-to-back in an albedometer.
• Rear-side irradiance, used on bifacial plants to quantify the light captured by the back of the modules.

The unit of measurement is always W/m² for irradiance. Energy, which is what you bill or bank, is the time integral of irradiance and is expressed in kWh/m². Think of irradiance as the speed on a speedometer and energy as the distance on the odometer.

Thermopile vs Silicon-Cell Pyranometers

There are two dominant pyranometer technologies in solar projects: thermopile and silicon-cell.

Thermopile pyranometers use a blackened thermopile detector under one or two glass domes. The absorber heats up when exposed to solar radiation, and the temperature difference between the hot junction and a reference junction produces a voltage. Because the black surface responds to the full solar spectrum, thermopile sensors cover roughly 300 nm to 2800 nm. They are the reference technology for meteorological and bankable solar measurements.

Silicon-cell pyranometers use a silicon photodiode or a small PV cell as the detector. They convert photons directly into current. The response is fast, but the spectral range is limited to about 350 nm to 1100 nm. Under clear skies this is acceptable. Under cloudy skies, low sun angles, or high aerosol loads, the spectral mismatch can add several percent error.

From Class C to Class B and from Class B to Class A, the practically achievable irradiance measurement accuracy improves by roughly a factor of two, according to Hukseflux (2026). That step change is why lenders and independent engineers insist on Class A thermopile sensors for utility-scale projects.

Silicon-cell sensors are not ISO 9060 spectrally flat. They are not accepted as the sole irradiance sensor for IEC 61724 Class A monitoring. The common best practice is to use a Class A thermopile pyranometer for the bankable GHI or POA reading and optionally add a silicon-cell sensor as a fast-response module-matched reference.

ISO 9060 Classes and IEC 61724-1 Monitoring Requirements

ISO 9060:2018 is the global standard that classifies pyranometers into three accuracy classes: Class A, Class B, and Class C. The classification tests response time, zero offsets, non-stability, non-linearity, directional response, spectral error, temperature response, tilt response, and signal-processing errors. A sensor must meet every limit for a class to carry that label.

IEC 61724-1 then maps those sensor classes into three monitoring classes for PV systems:

For Class A monitoring, IEC 61724-1 requires the use of ISO 9060 Class A pyranometers with dew and frost mitigation unless the site rarely experiences dew or frost, according to Hukseflux (2026). That is why heated domes are common on bankable met stations in humid or cold climates.

The standard also sets data intervals. Class A systems normally log irradiance every 1 minute. Class B systems may log every 5 to 15 minutes. The faster interval matters because short-term variability from passing clouds is smoothed out at coarser resolutions, which can bias performance ratio calculations.

Not every project needs Class A. A residential portfolio can operate well with Class C sensors or even satellite-derived irradiance. A 5 MW commercial roof under a PPA guarantee usually needs Class B. A utility plant with lender scrutiny needs Class A. The mistake is to assume that buying a Class A sensor automatically makes the whole monitoring system Class A; the logger, wiring, cleaning schedule, and temperature sensors must also meet the standard.

Designing Pyranometer Placement

Placement is where good hardware becomes good data. A pyranometer must see an unobstructed sky hemisphere and must be oriented exactly like the surface it is meant to represent.

GHI pyranometer placement

A GHI pyranometer must be perfectly horizontal. A tilt error of 1 degree can introduce a cosine error of about 1.5% at high solar zenith angles. Use a precision bubble level during installation and recheck alignment seasonally. Thermal expansion and wind vibration can loosen mounts over time.

The sensor should sit at least 2 meters above finished ground level. It must be farther from any obstruction than 10 times the height of that obstruction. A 3-meter inverter skid, for example, should be at least 30 meters away from the GHI sensor. This rule comes from the OTT HydroMet Solar Weather Stations Specification Guide (2026).

The pyranometer should also be far from reflective surfaces that can bounce light onto the dome. Nearby white gravel, metal roofs, or glass facades can raise the reading and corrupt PR calculations.

POA pyranometer placement

For fixed-tilt arrays, mount the POA pyranometer on a representative module or a dedicated bracket that matches the array tilt and azimuth exactly. The sensor surface must be parallel to the module surface. A 5-degree azimuth error is enough to bias POA by 1% or more at high latitudes.

For tracker arrays, the POA pyranometer should ride on a tracker unit that follows the same tracking algorithm as the production trackers. Do not mount it on a fixed structure and calculate the expected tracker angle. Mechanical backlash, wind stow logic, and algorithm differences can introduce 1-3% error between measured and modeled POA.

Bifacial and albedo measurements

Bifacial plants need a rear-side irradiance sensor, usually a downward-facing Class B or Class C pyranometer mounted in the array. Without rear irradiance measurement, the bifacial gain cannot be verified against the design assumption. For a full albedo measurement, use an albedometer with two matched pyranometers mounted back-to-back.

GHI is measured with a horizontal pyranometer, while POA is measured on an inclined plane parallel to the modules, according to Seven Sensor (2026). That distinction is the most common source of error in field PR calculations.

Outputs, Data Loggers, and Wiring

Pyranometers produce either analog or digital signals. The choice affects wiring, noise immunity, commissioning time, and long-term data quality.

Analog outputs include millivolt signals and 4-20 mA current loops. Millivolt signals are simple but susceptible to electromagnetic interference on long cable runs. A 4-20 mA loop is more noise-resistant but requires a precision shunt resistor at the logger. Both analog types need the logger to know the sensor calibration factor, which is usually printed on the calibration certificate in microvolts per W/m².

Digital outputs use RS-485 with Modbus RTU or SDI-12. Digital sensors transmit calibrated irradiance values directly, eliminating analog-to-digital conversion errors and reducing wiring complexity. Most modern data loggers, including Campbell CR1000X, CR6, and SMA Data Manager M, support Modbus RTU natively. The shift to digital is now the default for new Class A installations.

The data logger must sample fast enough to capture short-term irradiance variation. For Class A monitoring, 1-minute averages are standard. The logger should also record module temperature, ambient temperature, and wind speed because these variables feed into temperature-corrected performance models.

Cable runs from the pyranometer to the logger should use shielded twisted-pair cable for analog signals and properly terminated RS-485 cable for digital networks. Keep cable lengths as short as practical. Route cables away from AC power conductors to reduce induced noise. Label both ends of every cable with the sensor serial number so that future recalibration can be traced back to the correct channel.

Calibration and Maintenance Schedule

A pyranometer drifts over time. Dome coatings age, desiccants saturate, and thermopile responsivity shifts. Without a calibration schedule, the data becomes technically non-compliant even if the sensor still produces a signal.

IEC 61724-1 requires pyranometers in Class A monitoring systems to be recalibrated every 2 years, according to Hukseflux (2026). Many manufacturers recommend the same interval for Class B sensors. A typical Class A recalibration, including shipping, costs $300-$700 per sensor.

Calibration follows ISO 9847, which describes comparison to a reference pyranometer, or ISO 9846, which uses a pyrheliometer as the primary reference. Traceability ultimately leads back to the World Radiometric Reference maintained at the World Radiation Center in Davos. The NREL Broadband Outdoor Radiometer Calibration, or BORCAL, process is one widely used implementation of the outdoor comparison method, as described by Hukseflux (2026). Independent work at NREL found that different calibration methods can produce responsivity differences of 1-2%, which directly translates into irradiance measurement differences of the same size, according to a NREL study (2016).

Cleaning is just as important as calibration. Soiled domes can read low by 5-20% during dry seasons. IEC 61724-1 says Class A irradiance sensors should be cleaned every week unless site data proves that longer intervals are acceptable. In dusty regions, that interval may need to shorten to twice a week. Heated domes help prevent dew and frost but do not remove dust.

A simple maintenance checklist keeps data bankable:

• Clean the dome at the interval defined by the monitoring class and site soiling rate.
• Inspect the desiccant window monthly for moisture or fogging.
• Verify level with a bubble level quarterly for GHI sensors.
• Compare the sensor against a freshly calibrated portable reference at least once per year.
• Recalibrate every 2 years and update the logger sensitivity factor immediately.

Common Solar Pyranometer Design Mistakes

Even experienced engineers make the same mistakes. Here are the errors that most often corrupt irradiance data.

1. Mounting a POA sensor on a fixed structure at a tracker site

Trackers do not always sit at the ideal angle. Wind stow, backtracking, and mechanical tolerances mean the real module plane differs from the modeled plane. A fixed POA sensor cannot capture that difference. The fix is to mount the sensor on a tracker unit that follows the same algorithm.

2. Using GHI in place of POA for performance ratio

A 30-degree south-facing array at mid-latitudes receives 10-20% more POA irradiance than horizontal GHI over a year, depending on climate and season. Substituting GHI for POA makes the denominator too small and the PR too high. Always use a co-planar pyranometer or a properly transposed irradiance value.

3. Assuming all Class A sensors perform the same

ISO 9060 sets minimum limits, but some Class A sensors beat those limits by a wide margin while others barely pass. Independent testing, such as NREL BORCAL campaigns, regularly shows differences among Class A models. The fix is to specify not just the class but also the model, and to request calibration history and independent validation data.

4. Ignoring soiling

A dirty dome is a low-reading dome. If POA reads low, PR reads high. The system looks healthy even when it is underperforming. In arid regions, weekly cleaning may not be enough. Add a soiling station or use a cleaning schedule tied to measured soiling loss.

5. Skipping the heated dome in dew-prone climates

Dew and frost block the dome and create data gaps exactly when morning production is starting. IEC 61724-1 mandates dew and frost mitigation for Class A systems unless the site is proven dry. Heated domes or ventilated radiometers solve this.

From Measurement to Design: Using Irradiance Data in SurgePV

Measured irradiance is valuable, but most projects do not have 12 months of ground data at the design stage. That is where validated satellite irradiance and design automation come in.

SurgePV uses transposed POA data from verified satellite sources for energy simulations in the generation and financial tool. The platform applies site-specific weather, shading, and soiling losses to produce production and financial models that match the way systems actually perform. For sites with on-site pyranometer data, you can cross-check the satellite model and tighten the uncertainty band before financial close.

The solar design software lets you lay out arrays, run shadow analysis, and simulate hourly production without manual spreadsheets. When you reach the proposal stage, the solar proposals tool turns the design into a branded, finance-ready document. For large or complex sites, Clara AI accelerates layout and yield optimization by automating repetitive design decisions.

The workflow is straightforward:

1. Import the site location and pull satellite GHI, DNI, and DHI data.
2. Add array geometry, module specifications, and inverter sizing.
3. Run shading analysis and soiling adjustments.
4. Compare modeled POA against any on-site pyranometer measurements.
5. Export the production forecast and financial model directly into the proposal.

For a deeper look at the broader monitoring stack, see the guide to solar weather monitoring tools. If you need to explain performance ratio to a customer, the post on how to calculate solar performance ratio walks through the formula step by step.

2026 Cost Summary by Project Size

Pyranometer costs scale with accuracy class, number of sensors, and integration level. The table below shows typical 2026 pricing for hardware only. Installation, calibration, and data services add 15-30% on top.

For context, a full bankable-grade utility met station typically represents 0.3-0.5% of total plant CAPEX. The cost is small compared with the value of the data it protects. A single undetected inverter string fault can cost tens of thousands of euros in lost revenue, and a biased PR report can trigger warranty disputes or lender covenant reviews.

A silicon-cell pyranometer costs $300-$800, a Class B thermopile costs $1,200-$2,500, and a Class A thermopile costs $2,500-$5,000, based on 2026 distributor pricing. Budget $300-$700 per sensor for biennial recalibration and $500-$2,000 per year for cleaning and maintenance, depending on site soiling.

Frequently Asked Questions

What is a solar pyranometer and how does it work?

A solar pyranometer is a radiometer that measures global solar irradiance on a flat surface, reported in watts per square meter. A thermopile sensor under a glass dome absorbs shortwave radiation and produces a voltage signal proportional to irradiance. The sensor sees a 180-degree hemisphere, so it captures both direct beam and diffuse sky radiation.

What is the difference between a thermopile and a silicon-cell pyranometer?

A thermopile pyranometer responds to the full solar spectrum from roughly 300 nm to 2800 nm and meets ISO 9060 Class A/B/C requirements. A silicon-cell pyranometer uses a photodiode that only sees about 350 nm to 1100 nm. Silicon-cell models are cheaper and faster, but their spectral mismatch adds error under cloudy skies and low sun angles.

What ISO 9060 class do I need for my solar project?

Choose Class A for utility-scale or bankable commercial projects where lenders review performance ratio calculations. Choose Class B for standard commercial monitoring and warranty tracking. Choose Class C or silicon-cell sensors only for residential portfolios or basic operational checks where lower cost matters more than traceable accuracy.

How do you mount a pyranometer for PV monitoring?

Mount a GHI pyranometer perfectly horizontal at least 2 meters above ground, with no obstructions above the horizon and a distance from objects of at least 10 times their height. Mount a POA pyranometer on a representative module or bracket at the exact tilt and azimuth of the array. On tracker sites, mount the POA sensor on a tracker unit that follows the same algorithm.

How often should a pyranometer be calibrated?

IEC 61724-1 requires pyranometers in Class A monitoring systems to be recalibrated every 2 years. Many manufacturers and calibration labs recommend the same interval for Class B sensors. Clean the dome at least weekly unless site data proves that soiling is negligible.

Can I use a silicon-cell pyranometer for bankable performance data?

No. Silicon-cell pyranometers are not ISO 9060 spectrally flat and are not accepted as the sole irradiance sensor for IEC 61724 Class A monitoring. An independent engineer will flag silicon-cell data in a bankable yield report. Use a Class A thermopile pyranometer for lender-grade work.

What is plane-of-array irradiance and why does it matter?

Plane-of-array irradiance, or POA, is the total solar radiation striking the tilted surface of the PV modules. It is the denominator in the performance ratio equation. Using horizontal GHI instead of POA can overstate PR by 10-20% for a typical fixed-tilt array, so a co-planar pyranometer is needed for accurate reporting.

How much does a solar pyranometer cost?

A silicon-cell pyranometer costs $300-$800. A Class B thermopile pyranometer costs $1,200-$2,500. A Class A thermopile pyranometer costs $2,500-$5,000. A complete bankable weather station with redundant Class A sensors, temperature sensors, and a data logger ranges from $15,000 to $30,000.

Conclusion

A solar pyranometer is a simple device with an outsized impact on project economics. The right class, placement, and maintenance schedule turn irradiance data into a reliable foundation for design, performance tracking, and financing. The wrong choices turn every yield and PR calculation into a guess.

Three actions to take next:

1. Match the sensor class to the contract. Specify ISO 9060 Class A for bankable projects, Class B for standard commercial work, and Class C only where traceability is not required.
2. Design placement into the array layout. Reserve a clear, level, obstruction-free location for GHI sensors and a tracker-following mount for POA sensors on tracking sites.
3. Lock in calibration and cleaning budgets before commissioning. Biennial recalibration and a written cleaning schedule are non-negotiable for Class A data.

If you are ready to move from raw irradiance data to accurate system design and proposals, run your next project through SurgePV and see how verified weather data, shading analysis, and financial automation fit together.