Solar Farm Drone Inspections
How Drones Make Your Solar Operations Better
Drones In Solar PDF - Download
A guide to Inspecting Solar Fields with Thermal imaging drones - Download
Budgeting for Solar PV Plant Ops & Maintenance - Download
Aerial Thermography Inspections in Large-Scale PV Plants - Download
Local Weather Conditions - Link here (Click Map at right to select location)
How Can Drones Make Your Operations Better?
Using drones in the solar industry both opens up new possibilities and replaces existing work. Without drones, inspections are typically completed manually. For utility-scale solar farms, this means either traversing hundreds of acres and conducting the painstaking process of inspecting thousands of panels by hand, or, more commonly, inspecting only a sampling of panels in an effort to identify systemic issues. In some cases, high-cost inspections by small plane may be used. Inspecting rooftop systems, of course, involves the hazards of sending workers onto rooftops.
With drones, you can complete inspections in a fraction of the time, saving costs while avoiding hazardous man-hours and getting better data.
“We are reducing high-risk activities by using new technologies to improve safety, increase efficiencies, and enhance overall company asset management,“ explains Assel Ayapova, Global Drone Program Manager for AES Corporation.
Save Time, Save Money
Last year, Measure conducted a study to compare the time, costs, and results of its 100% IR drone inspections to relevant manual inspection scenarios across four sites.
SITE 1: 100% IR drone inspection for maintenance of 74MW in Sunrall, MS; compared to clamp testing with 20% IV curve tracing testing
SITE 2: 100% IR drone inspection for maintenance of 30MW in Spargue, CT; compared to Voc/Isc testing at the combiner box with visual inspection
SITE 3: 100% IR drone inspection for commissioning of 21MW in Rincon, GA; compared to 100% hand-held IR scanning with 15% IV curve tracing testing
SITE 4: 100% IR drone inspection for commissioning of 12.5MW in Herald, CA; compared to 100% IV curve tracing testing
Comparing drone inspection time to relevant manual inspections across the 4 sites, we saw an increase in inspection efficiency of 97%.
For the amount of data that is processed with a typical inspection (approximately 800 images per MW), turn-around time is surprisingly quick. These inspections each took less than 5 business days to analyze and deliver.
More Accurate Data
Saving time and money is great, but when it comes to drones in solar, it’s often the data that tells the best story.
To test the accuracy of Measure’s drone data, AES took the results of a solar inspection by drone and sent out manual inspection crews to run the same inspection on the same plants. The results from the manual inspection mirrored the results from the drone data with 99 percent accuracy, but the manual inspection took two days compared to two hours with the drone.
Due to the time and tedium of manually inspecting large solar farms, it’s often not realistic or cost-effective to do a manual inspection of 100% of the site. That’s not the case with drones. You’ll quickly and easily get a full inspection, identifying defects that manual inspections might miss, at the string, module, and sub-module level.
Large sites where only a percentage (e.g. 20%) of the facility would normally undergo IV curve tracing each year will benefit from enhanced revenue opportunities realized through 100% IR scanning. For example, Site 1, noted on the previous page, saw an additional $91 per MW in enhanced revenue opportunity due to incremental issues discovered by the drone inspection. A drone inspection can enable companies to proactively identify more defects, which could lead to increased power production.
Drones can also offer more than solar module inspection data. While drones are on-site, you have the opportunity to get accurate information on other site infrastructure as well. You’ll learn more about this in the next section.
Easy to Use
Following processing and analysis, Measure delivers drone data to clients in a user-friendly, actionable format. Reading and consuming the information is surprisingly easy. “The reports are pretty consolidated,” states Nick McKee, Solar Operations Manager at AES. “I have a PDF snapshot and a digital snapshot that I can move around and customize depending on what I want to look at.” Data can even be delivered through a smart phone app, which allows maintenance personnel to proceed directly to damages, improving efficiency and reducing costly repair hours.
Three ways to receive drone data:
PDF Inspection Reports: High level, punch list summaries of inspection results.
Online Webmaps: Interactive data hosting platform that includes a visual representation of aerial inspection results and a dataset that is exportable, georeferenced, and interactive.
Field Repair Application: Smart phone app that routes technicians to malfunctioning modules and allows corrective maintenance or field validation results to be input while on-site.
With drone inspection data in an online portal (or integrated into your internal data systems), you’ll also have data that can be compared over time and across sites. Tracking system issues over time will drive better and better decisions for operations, maintainance, and system design teams.
Figure 1.1 - Drone Data Delivered by Webmap and Field Application
Measure’s Drone Inspection Time:
10 min Per MW of Solar
Measure has seen the best results from flying each site twice, using different sensors and at different altitudes. One flight collects RGB imagery, while a second flight collects thermal imagery. New technology is driving a shift to collect both RGB and thermal imagery on a single flight for even greater efficiency.
In order to ensure that the data collected meets the project needs, Measure pilots also conduct quality checks with the data team, typically at several intervals, during complex jobs. This helps to reduce the chance of having to return and re-fly a job site due to poor data quality. See figures 3.2 - 3.4 for examples of quality data.
Example Inspection Images
Figure 3.2 - Good Solar RGB Image - Panels in focus. Panels at relatively low angles. Clear panel boundaries. No noticeable glare.
Figure 3.4 - Mapping and 3D Modeling - If using data for mapping or 3D modeling, ensure data has proper overlap by using a program such as Pix4D to load images and perform a rapid quality check
Unless you are doing a simple site check (like taking a quick look after a storm event) or taking basic aerial imagery (such as for marketing use), your drone data will require processing and analysis in order to turn it into something that can be put to use by your organization.
Measure’s Data Team Processes & Analyzes:
800 Photos per MW of Solar
Once data has been captured and transferred, it is loaded and prepared for processing and analysis. Because Measure uses a high image overlap, thermal images can be compiled into orthomosaics that remove artifacts such as blur and glare prior to final analysis. Each individual panel is identified and outlined using automated processing techniques. Defects are identified based on thermal data and geo-located within the orthomosaic. All defects flagged are verified manually through a review of several individual thermal images of the damange location. Extensive thermal data capture combined with manual quality control allows Measure to mitigate false positives and deliver highly accurate results.
To visualize the results of solar inspections, Measure uses ArcGIS to produce an interactive webmap that can be annotated. ArcGIS data can also be delivered through a convenient app, making it easy for field personnel to locate issues and provide updates right from their smart phone.
Hardware & Software
These are the Tools Required for Every Drone Program
Drones and Sensors
The next obvious part of any drone program is, of course, the drones. When choosing the right equipment for a job, there are a number of factors to consider: type of operations, data requirements, security requirements, and cost. Common drone platforms and sensor payloads are shown in Figures 4.1 and 4.2. Standardizing on one drone manufacturer, or, if possible, one drone airframe, will simplify aircraft maintenance and repair for an in-house drone program.
Drone aircraft come in two major physical configurations: multi-rotor and fixed-wing. Multi-rotor drones tend to be easier to use and more maneuverable, while fixed-wing drones typically offer longer flight times.
In the multi-rotor drone marketplace, DJI dominates with over 70% market share. DJI products offer quality and reliability at an affordable price, and they cover a wide range of applications and levels of sophistication. DJI’s market dominance and expansive equipment selection makes it an attractive choice for residential and small-to-mid commercial solar applications.
Fixed-wing drones tend to be a better match for sites larger than 5MW. Fixed-wing platforms lack the maneuverability and ease-of-use associated with multi-rotors, but offer superior endurance, key for large-scale operations as frequent battery changes slow data collection. Measure currently uses the senseFly eBee for utility-scale solar power plant inspections and has seen flight times of up to 50 minutes, depending on weather conditions. Measure also prefers the higher frequency thermal image capture offered by the eBee, compared to common multi-rotor airframes. This fixed-wing drone packs a lot of power in a small platform and accepts a variety of payloads, including visual, thermal, and multispectral sensors. However, because fixed wings drones tend to be slightly more challenging to fly (or, more specifically, to land), Measure typically recommends that large-scale operations are outsourced to a highly experienced and specialized drone operator.
Program Management Software
As discussed in the previous section, managing a drone program is a complex operation, covering many functions (see Fig 3.1). Looking across the drone software market, you will find a plethora of products targeted at one or a few of these functions. For example, there are popular software products focused only on flight logging or only on equipment management.
However, using a single software solution for as many functions as possible - work ordering, resource management, flight planning and tracking, program reporting, compliance, and data management - will help you streamline your operations and manage your program more efficiently. You’ll also have the program oversight that many large corporations need to ensure consistently safe and compliant execution of all aspects of their drone program.
Measure was searching for a comprehensive software platform to manage its own extensive drone operations. Unable to find a platform that met all of its needs, Measure built one, based on the experience of managing thousands of flights across myriad applications. That product is Measure Ground Control. A basic overview of Ground Control’s management portal is provided in Appendix C. Ground Control also includes an integrated flight application.
Figure 5.1 - Common Commercial Drone Hardware
FIGURE 5.2 - Common Commercial Drone Sensors
A flight application is used to control the drone during flight. You can find many drone flight applications available in the various iOS or Android app stores, and several of them are available for free. DJI offers a free application, DJI GO, that works with all DJI drones and is available at no additional charge. Given DJI’s dominance in the drone market, their flight application also has the highest number of users. However, this application serves the needs of a very wide range of drone users, including hobbyists who want to take aerial pictures and upload them to social media.
A flight application designed specifically for commercial use is typically the better choice. A commercial flight application should offer a simplified interface with only the functions required, and it should be oriented toward safety and security with features such as pre- and post- flight checklists, integrated airspace advisories and authorization, and local data mode to block data sharing with DJI.
Having a flight application that is integrated with your program management software offers additional benefits, such as automatic uploading of all flight data. Flights along with completed checklists and screen shots can be easily and consistently captured, tracked, and reported on, along with notifications of any flight activities that do not adhere to safety best practices. Measure Ground Control, which integrates a program management platform with a flight application, also includes features such as flight playback, where the flight is recreated on-screen, as well the ability to translate flight logs into drone and battery usage data that informs maintenance decisions.
Measure Ground Control Flight Application: LAANC airspace authorization
Measure Ground Control Flight Application: Pre-flight checklist and automated grid flight pattern
Measure Ground Control Flight Application: Waypoint flight setup
Drone Data Software
Drone data is often uploaded into various software platforms for processing, analysis, and visualization. Figure 4.3 summarizes some of the different software products used by Measure’s Data Engineering team. As shown in the Figure, different software platforms are best suited for different types of applications and data needs. Some tools are designed to be used only for raw dataset processing, while others are only useful for analysis or visualization of processed data. Analytical tools in particular are often targeted to a specific industrial application. ArcGIS is most commonly used by Measure for analyzing and viewing data for utility scale solar inspections, while ArcGIS or Scopito may be used for related inspections of electric transmission lines and poles, vegetation assessments, etc..
What data software you need will depend on whether you will be processing and analyzing drone data internally, or whether you will be outsourcing this function. If you don’t plan to do internal data processing and analysis, you only need to worry about how to best integrate the final drone data product into your operations. Some organizations will want to integrate ArcGIS data, for example, while others will use online drone data platforms separately. If you will be using an outsource model, access to these platforms will typically be provided by your vendor.
Always keep in mind who the stakeholders of your program are and who will need access to the data. Data that is difficult for asset managers or O&M teams to use, for example, is not likely to maximize the return on your investment. Make sure that the processing, analyzing, and visualization of your drone data results in a data deliverable that can drive better decisions for your business operations.
ArcGIS data of a utlity pole inspection delivered through a smartphone application for use by field personnel
Figure 5.3 - Common Drone Data Software
Appendix A: Insource vs. Outsource Decision Guide
8.25 hectars per day
1MW per Hectar
8.5MW per day
5 hours per day 10am to 3pm
Sensefly covers 270MW per day
65ha per 42 minute flight
s well as being quicker, drones could be around half the price of ground-based inspection techniques, assuming one megawatt of panels per hectare and rates of around USD$250 (€225) per megawatt of panels for manual inspections.
On a site generating 21 MW, for example, Measure can complete an inspection in 7 hours
5. Solar plant drone inspections
A global consulting firm estimates drones cut the cost of thermographic inspections of utility-scale solar farms between 30 and 40 percent. Data collection using infrared imaging for a 75 MW, 500-acre facility lowers inspection time from about a month to about a week, with fewer workers.
71.4 acers per day
15.4 MW per 5 hour day
Pricing Solar Farm Inspections
Drone service providers think in terms of acres, but solar asset managers pay for services by the megawatt (MW). Older solar installations typically require 7-9 acres/MW, while newer installations only need 5 acres/MW. You should price according to the nameplate capacity, which will typically be in DC power at peak wattage (MWDC).
According to a country manager at Enertis, a global solar O&M service provider, “Carrying out the site data collection for quality control using infrared imaging across a 75 MW, 200-hectare plant would usually take two or three people at least a month.”
Aerospec was invited to inspect a 100MW solar farm in the U.S. which consist of approximately 450,000 PV modules. With its proprietary image processing technology and detection algorithm, Aerospec successfully identified and located all defective panels while use one-tenth of the time it would take for manual inspections. At these speeds, we can operate 10 times faster than what other technologies in the market can achieve.
JUNE 22, 2018
For instance, two human beings spending 30 days, doing handheld IR analysis, spot checks and IV curves on a 100 MW plant just doesn’t make sense. However – doing the same site in 24 hours, and discovering up to 3,500 anomalies in 423,000 solar panels, as occured in a recent case study, makes a whole lot of sense.
The drones commonly used, Aerospec referenced the M100 and M200, can cost between $5,000-20,000. Then you add a Flir infrared camera that starts at $5,000.
Historically, per Lance, using an airplane to do the thermal imaging cost greater than $1,000/MW. In recent times, he’s seen these services fall 60% to $400/MW, before data analysis, mainly due to competition from drone contractors at $50-75/hour. A talented operator that maps a location efficiently can cover a MW of solar in less than ten minutes – see unique zone maps below – before the battery might need be replaced. A drone Aerospec recently sent to a customer can fly for up to 38 minutes.
Before drones, there were two choices for inspecting strings, PV panels, and other equipment. PV technicians could conduct inspections on truck rolls and on foot with a handheld thermal camera, inspecting thousands of acres and tens of thousands of panels in large solar farms, a project that can take weeks and often results in flawed data. Inspections have been typically a reactive task, focused on pinpointing problems versus preventive maintenance to reduce equipment downtime. The other choice has been to conduct inspections with manned aircraft mounted with sensors, which can be faster but more expensive, at about $1,000 to $2,000 an hour.
Here are some of the ways drones aid in thorough, cost-effective solar utility inspections:
Finding malfunctioning units in an array.
Performing preventive maintenance. Panels, mounting systems, wiring, monitoring equipment, and other plant components can be visually imaged for dust and dirt, snail trails, leaves, corrosion, defects, cracks, and water or insect intrusion.
Monitoring vegetation to keep foliage from casting shade on strings.
Detecting equipment that may be in danger of overheating (combiner and junction boxes, inverters).
Assessing damage following wind, seismic, flood, fire, or electrical storm events.
Inspecting transmission lines.
Monitoring fencing for security purposes.
Piloted helicopters and planes may still make sense for inspecting facilities of more than 25 MW or covering more than 5,000 acres,
*A 2015 EPRI study, done before drones were being widely used for solar farm O&M, estimated a cost of $10 to $25/ kW-yr. for inspections and other standard preventative maintenance activities that would typically be performed at most utility-scale sites (e.g., visual/structural system inspection, wires and combiner box inspection, infrared thermography scanning, IV curve tracing, and inverter maintenance).
With an exponential level of growth, current methods of operation and maintenance are simply not sustainable nor is it economically feasible to deploy poorly trained manpower. Two field engineers spending 30 days, doing handheld IR analysis, spot checks and IV curves on a 100 MW plant just doesn’t make sense. Imagine if you could lift these sensors above the ground and do the same site in 24 man-hours, and discover up to 3,500 anomalies in 423,000 solar panels, now that makes a whole lot of sense.
This then asks the question how much detail do you need?
If you need to determine module or array level and bigger defects, then the flight height can be much higher, if you need to see cell level defects then your Inspection needs to be conducted at a much lower height ( and therefore will take longer to fly)
There are really 3 level of inspection
Overview – Altitude 300ft+-
Array & String level defects
Detail – Altitude 150ft+-
Module level and cell level defects
Comprehensive – Altitude 75ft+-
Cell level defects
We also provide a hybrid level of inspection with our M210 dual sensor mount option which carries the Flir XT2 and the DJI Z30 zoom lens/camera. With this option we can provide a thermal Images detail at a sub module level and using the zoom camera provide RGB images of specific defects on the module. This can show physical damage or lack of physical damage to help guide the repair decision process.
3 Images below are all taken at the same time with the drone in the same position
XT2 – Thermal – (R-JPG) Sub module level issue. Hotspot’s visible.
XT2 – RGB – (JPG) visual of sub module level issue, confirm physical of surrounding array.
Z30 RGB (JPG) Zoom approx 10X, close up of module showing physical damage.
International 668 951 860 76
Local: 089 518 6076
Display Name: Bangkok Video
Bangkok Video Productions Operates Aerial Filming Drones with the following licenses:
Royal Aeronautics Sports Association of Thailand ID: B15-101-0740
Thailand Film Office Coordinator license Number: 1 1014 01752 24
Civil Aviation Of Thailand Licensed Pilots & Drones
Thailand Flight - Regulations & Restricted Areas