Findings from 14GW of thermography in 17 countries

4 min read

Hot-spots in PV modules are a commonly encountered defect that can significantly impact performance on sites large and small. Hotspots can be caused by a number of underlying defects that present with a wide variety of thermal signatures. While remediation is essential, the severity of the underlying defect is the primary determinant of the appropriate remediation action.

Hot-spots can be broadly categorized into four types based on their cause

    1. Soiling/Shading: Hotspots caused by shading from vegetation, adjacent structures, module racks, debris on the surface, bird droppings and more
    2. Mechanical damage: Hotspots caused by cracked and broken glass, improper handling and installation, physical impact damage, 
    3. Module/Cell defects: Hotspots caused by cell defects, cell cracks,de-lamination, bad solder joints, stuck bypass diodes, diode failures and more
    4. Installation defects: Hotspots caused by unconnected strings and modules, reverse polarity

The focus of investigation and remediation action is always in the reverse order of defect types listed above, with the most bang for the buck delivered by correcting installation defects and module/cell defects. Beyond operational performance, hotspots also impact safety on sites as they can lead to runaway thermal events and fires. Hotspot detection and correction is therefore an important O&M activity on sites of all sizes. An annual hotspot check is usually recommended, but 100% coverage has been an elusive goal as it was often too expensive to check a site completely to identify defects. But this has changed with the development of drone based thermal scanning procedures. What was once a manual job that took weeks is now completed in a matter of hours with drones enabling single operators to thermally image over 50MWp of installed solar PV in a single day.

While data collection has become relatively simple, the onus is now on accurately processing collected information to identify and classify hotspots. As operators start to rely solely on drone based scans, accuracy of the results becomes the primary determinant of site performance and safety.

At SenseHawk, we have processed over 14GW of thermal scans. Starting with what was then the world’s largest single owner site (750MWp+), we have worked on sites of all sizes and technologies in 17 countries around the world and have perfected our detection and classification engines  to ensure we do not miss anything. We are also constantly at work examining the data we collect, to infer trends and spot areas of improvement. In the following section, we have attempted to provide you with a summary of our findings

  1. Asset owners lose 2% of generation on average due to defects that can be detected with thermal scans
    • The median loss observed was 1.4%, while the range was wide with the best performing site losing as little as 0.1% while the worst performing site was losing as much as 20%. Without considering outliers, the loss range was between .5 and 3%
  2. Thermal scans at commissioning pay back instantly and should be a part of the HOTO process
    • On newly constructed sites, we observed an average loss of 2% just from strings that were offline.
    • Additionally, an average of 2200 modules qualified for warranty replacements per 100MWp of installed capacity
    • An average of 4 potentially hazardous defects (Reversed strings/ short circuits) were detected per 100MWp
    • Minor defects with low delta T at commissioning often resulted in failures in the first year of operation, with the most common failure being burnt out interconnect solder
    • Broken/damaged modules are easily found at commissioning and can be flagged for replacement by the EPC
  3. Single cell hotspots are a mixed bag
    • Single cell hotspots display high delta T and are often flagged for inspection. However, causes vary and a majority are caused by bird droppings, dust or vegetation shading. About 20-35% of single cell hotspots are caused by cell defects
  4. Stuck open diodes are a common occurrence
    • We observed between 400-3500 modules with stuck open diodes per 100MWp on sites we scanned. Of these, 80% were warranty replaceable with diode/junction box failures
  5. Thermal defects occur as often on large sites as on small sites
    • Defect density was is not correlated to size of installation and small sites between 1 and 20MWp, with limited maintenance fared as well as sites that were 100MWp or larger
    • However, rooftops did not fare as well
  6. Small rooftops fare poorly compared to larger sites
    • The average rooftop site loses 4% of production due to thermal defects. Defect density correlates with age of the site
  7. Thermal scans pay back in as little as 3 months
    • On utility scale sites, thermal scans pay back within 3-5 months and are a function of the PPA rate and installation size
  8. A map based app with a defect location and management system improves outcomes
    • In a number of sites, thermal scan results do not get the attention they deserve as field teams find it difficult to accurately locate and fix problems
    • With our desk and app companion products, we observed resolution rates increase from 46% to 91% while resolution time per 100MWp reduced by 60%

Based on our findings, we recommend the following best practices

  1. Thermal drones should become an integral part of the solar PV commissioning and O&M toolkit. 
  2. An annual thermal scan should be a part ot the PM plan
  3. Annual scans can be coupled with on-demand scans based on specific production problems. These will need in-house drone capability
  4. Surveys should be scheduled immediately after module cleaning to eliminate dust and bird dropping related hotspots
  5. A platform approach to managing scan data with cloud based software and analytics tools is critical
  6. Issue tracking software coupled with a map based field app closely integrated with the analytics platform is critical to tracking, remediation, warranty claims, and long term monitoring


Standard Test Conditions: normalized to 1000 W/m2 irradiance

Interesting? Get in touch to learn more and talk about how we can help you maximize performance. 

SenseHawk

SenseHawk

SenseHawk, Inc. is a provider of cloud based drone data analytics and productivity tools for the solar industry. With customers in 15 countries, SenseHawk solutions are used by solar & EPC companies for terrain analytics, construction monitoring, thermography and task management on solar sites

Sign up for insights and opinions from the best in the solar.

Leave a Comment