Growing Pain #3: On-site measurements in large-scale solar

Designing and operating a large-scale solar project without fully understanding its potential output inevitably increases risks throughout its lifecycle. One part of the solution for developers is validated solar resource data calculated through satellite-based models, helping produce accurate energy yield calculations.

However, to reduce uncertainty in these calculations – and financial risk – there is another key component of a solar data strategy: on-site measurements.

For solar projects smaller than 10MW, the cost of conducting a best practice on-site measurement campaign may be uneconomical. For large-scale projects, however, investment into solar measurements brings important improvements in technical design and financial decisions.

Recognising this, solar project developers are increasingly using on-site measurements to bolster the satellite-model data and mitigate financial and technical risk. The next step for the industry is to ensure that best practice is always followed for on-site measurement campaigns, so this investment is not wasted through poor quality data outputs.

This instalment of Growing Pains will explain why ensuring the accuracy of on-site measurements is vital, and outline how the two most common challenges can be overcome: inadequate equipment and incorrect implementation.

An introduction to accurate on-site measurement campaigns

Developing a large-scale PV project requires a long history of solar and meteorological data, optimally 20 years or more. The models provide historical data, tracking variables such as temperature, wind and solar radiation, but their spatial and temporal granularity is limited.

Global models cover large swathes of territory, but may not be detailed enough to capture local climate, which can vary considerably over a distance of kilometres or even hundreds of metres. For example, temperature, wind and rainfall at a coastal city will be different to a site slightly further inland. Significant differences are also observed in hilly terrain.

Additionally, the standard temporal resolution of global datasets is one hour, but for detailed technical design modelling, data may be needed on a minute-by-minute basis. This is especially crucial for solar plus storage projects.

The solution is to complement historical data from models with site-specific and detailed local measurements taken over the course of at least a year. This enables developers to recalibrate the modelled data and make it more representative of the microclimate of the site, while reducing data uncertainty. The locally adapted historical model data are then used in accurate and reliable prediction of PV power production for the next 20 to 25 years. The adapted model data reliably represents not only long-term climate averages, but also weather extremes, important for understanding risk patterns.

Components of a best practice measurement campaign

One year of on-site measurements requires investment into pyranometers or other solar instruments, meteorological sensors, data loggers and other hardware. The equipment installation and operation should include regular cleaning and maintenance, following the best practices and standards defined by the World Meteorological Organisation.

The objective is maintaining low uncertainty of on-site measurements throughout the duration of the campaign. This requires that:

• At least 2x Class A pyranometers, measuring Global Horizontal Irradiance with a ventilation unit are used to build in redundancy.
• In more challenging climates, or for larger projects (more than 100MW), other components of solar resource such as Direct Normal Irradiance and Diffuse Horizontal Irradiance are measured. These extra solar parameters can be measured by various solar trackers, Rotating Shadowband Radiometer (RSR) or SPN1 pyranometer.
• Additional meteorological parameters such as air temperature, wind speed, wind direction, air pressure, and precipitation are measured. For projects utilising bifacial PV module technology, albedo is also measured.
• Best practice cleaning & maintenance of the pyranometers and meteorological equipment are followed.

The three critical challenges for on-site measurement campaigns

This section outlines three common challenges for acquisition and use of measurement data from solar sites. These are:

1. Using unsuitable instrumentation and equipment
2. An incorrect implementation process
3. Reconciling ground-based measurements with satellite-model time series.

Challenge 1: Using inadequate equipment

One important benefit of measurement campaigns is reducing systematic deviations in solar model datasets. If this is not achieved due to inaccurate data from pyranometers and other meteorological instrumentation, the entire effort was an ineffective use of resources.

As the upfront costs of a stringent, scientifically credible measurement campaign are not insignificant – about USD 20,000 a year or more – stakeholders within a solar project may look for cost-efficiencies by reducing the quantity of pyranometers used or opting for the cheap instrumentation.

Through inadequate investment, resulting in missing sensors, use of substandard and poorly designed equipment or lack of measurement redundancy, large-scale project developers are exposed to the following risks:

• Unreliable data readings leading to incomplete and skewed statistics.
• Systematic underprediction / overprediction of long-term energy yield, causing financial underperformance or undervaluation.
• Technical design flaws for the PV plant through not considering specific microclimate features, including extreme weather.
• Increased degradation of PV components and higher operation and maintenance costs, due to lack of understanding of local environmental risks such as strong winds, increased pollution, dust, corrosion, high temperature variations or high humidity.

For more information on financial and technical risks from inaccurate solar data, please read Growing Pain #1: Project Design

To mitigate these risks, Solargis recommends that developers and their measurement partners:

• Invest in meteorological stations, which include high standard and complementary instruments to enable systematic check of the data consistency and stability.
• Build redundancy into the campaign. This reduces the impact of any erroneous data points due to malfunction or incorrect maintenance.
• Invest with the long-term in mind. A good pyranometer can cost upwards of €1500 – but the initial savings from purchasing cheaper equipment can be far less than the costs of using inaccurate data to design a large-scale solar project.

Challenge 2: Incorrect implementation of measurement campaigns

Even with the best instrumentation available on the market, the measurement campaign will not deliver useful data if the deployment and maintenance strategies are not also in line with best practice.

For large-scale solar projects in particular, there are two key questions to consider:

1. Is the cleaning, operations and maintenance strategy suitable?
2. Do the measurements taken reflect the actual conditions of the site?

Is the cleaning, operations and maintenance strategy suitable?

For the large PV projects, tens of pyranometers are often installed. However, numbers alone are not sufficient to deliver bankable data. Suboptimal cleaning & maintenance, and postponed re-calibration, can quickly decrease the value of data obtained from a measurement campaign.

To reduce costs, there is a tendency within the solar industry to implement measurement campaigns in-house rather than instructing experienced solar measurement contractors.

Unlike temperature measurements, however, solar radiation equipment requires careful placement and a systematic maintenance regime. The majority of meteorological equipment at development sites is installed in sub-optimal conditions, and are subject to:

• Shading from PV modules, objects or terrain obstructing sunlight
• Reflections from PV modules, bright terrain surfaces or water.
• Incorrect position, such as at excessive height, which doesn’t allow easy maintenance.
• Insufficient cleaning, allowing the build-up of dust, soil or dew, obscuring the sensors.
• Battery, internet or data logger issues, void calibration certificates, missing or insufficient data quality control.

To tackle this issue, Solargis recommends:

• Using a specialist partner with the knowledge of good practice deployment and layout of instrumentation, and experience protecting against theft and animals.
• Install a limited number of higher-quality and well maintained pyranometers, optimally coupled with RSR or SPN1 instrumentation.
• When budgeting for the campaign, dedicate a significant part of the overall budget for cleaning, operations and maintenance, rather than purchasing additional equipment which cannot subsequently be maintained properly.

Do the measurements taken reflect the actual conditions of the site?

Large-scale solar projects are often located in remote areas. The security and surveillance of valuable meteorological equipment at a remote site incentivises organising the measurement campaign near human settlements, where the equipment can be deployed and maintained at a lower cost.

However, the local microclimate may be highly variable, meaning that the measurements taken are inaccurate for the intended project site. This is especially the case in mountainous or coastal areas.

Additionally, large-scale projects increasingly use bifacial modules, which capture solar irradiance reflected from the surface. The albedo measurements need to be taken from the same or similar surface type planned for the project operation.

To mitigate this issue, Solargis recommends that:

• For the case of large solar plants the measurements are taken in several locations to ensure that local climate/surface variation is fully accounted for.
• Seasonality is accounted for – with albedo measurements a surface has different characteristics during winter and summer, or during dry and humid seasons, so measuring for at least 12 months is critical.
• Measurement data is quality controlled, validated and certified by knowledgeable and trustable solar resource experts.

Challenge 3: Reconciling ground-based measurements with satellite-model time series

Solar developers are often faced with significant challenges in integrating data from on-site measurement campaigns with existing satellite-based models. This is particularly the case when the data from on-site measurement campaigns is not controlled for quality.

Data from on-site measurement campaigns may be:

• Unreliable – containing anomalous and suspicious data points, which differ from each other, and/or are different from the model values.
• Incomplete – missing data for extended periods due to equipment failure or miscalibration.
• Inconsistently formatted – with different time stamps and metadata.

This can lead to significant inefficiencies for developers, particularly when data experts need to manually correct and align the data using inadequate software tools, such as spreadsheets, which do not offer domain-specific functionality.

To manage data quality efficiently, Solargis recommends that developers invest in a data visualisation and analysis tool designed for use in the solar sector with the capability to:

• Easily import and visualise data to identify areas that need correcting.
• Create harmonized time series by filling the gaps with data from alternative measurements or from model time series.
• Align time series data and fix heterogeneous time stamps.
• Add metadata to the level of detail needed.
• Verify data using tools based on solar physics.
• Validate measurements with satellite-model time series data.

Ultimately, measurement time series work best when used to complement verified and accurate satellite-derived solar datasets. Both satellite-model data and measurements have uncertainty involved. Combining the data sources helps reach the lowest possible level of uncertainty when assessing the solar performance of a site.

Based on rich experience with processing solar measurements and model data, Solargis has recently released a software tool to help solar developers and operators manage the quality of their data efficiently. Solargis Analyst enables straightforward comparisons of data sets, from on-site measurements or models, and ensures that data quality is consistent.

Get in touch to find out more about how Solargis can advise on implementing and validating on-site measurement campaigns for large-scale solar projects: https://solargis.com/about-us/contact