Don’t Overlook Tree Shadows! Leveraging 3D Models from LRTK for PVsyst Simulations

Shadow analysis for solar power plants and its importance

In designing solar power plants, it is critically important to accurately assess the impact of shadows cast by the surrounding trees, buildings, and terrain. Shadows on photovoltaic panels not only reduce energy output but can also create additional losses (mismatch losses) by upsetting the output balance of strings (panels connected in series).

Especially in locations like Japan, where forests and buildings are often close by, generation losses from shadows cast by nearby woods or elevated terrain cannot be ignored. If surrounding shadow effects are overlooked in simulations, actual power generation may fall far short of predictions, potentially jeopardizing the business plan. Therefore, shadow analysis requires meticulous consideration that “does not overlook tree shadows.”

Challenges in conventional shadow analysis

There have been several challenges in traditional shadow analysis for solar power. First, collecting information about objects that cause shadows is not easy. Accurately determining the height and position of nearby trees and buildings at a plant site traditionally required field surveys with surveying instruments or estimates from design drawings and topographic maps. These methods are labor-intensive and may fail to cover all obstacles. For example, some small trees or structures outside the site boundary may not be reflected on drawings and can be overlooked.

Second, even if data can be obtained, reflecting it in simulations is cumbersome. Inputting shading factors into simulation software like PVsyst generally requires 3D modeling or specifying coordinates. Manually creating 3D models of nearby objects is tedious, and oversimplification reduces accuracy. As a result, conventional shadow analyses often considered only distant mountain shadows from simplified horizon data while overlooking nearby tree shadows, or focused on panel-to-panel shading but did not account for the influence of surrounding structures. To solve these problems, a new approach was needed to “digitize shadow causes comprehensively and with high accuracy” and reliably incorporate that data into simulations.

Detailed shadow simulation with PVsyst

PVsyst, a widely used solar power simulation software, provides powerful functions to evaluate shadow effects in detail. PVsyst can build a project’s 3D scene (layout) and calculate the shading on panel receiving surfaces at each time along the sun’s path. Therefore, with accurate 3D models as input, it can quantitatively compute losses from surrounding structures and trees. In detailed settings, PVsyst can perform not only calculation of the fraction of incident irradiance blocked but also electrical mismatch calculations that account for imbalances in string voltages and currents caused by shading. This enables evaluation of additional losses arising from shifts in the maximum power point of each string due to shading, yielding simulations that more closely reflect actual generation behavior.

However, to fully utilize PVsyst’s sophisticated shading analysis features, commensurate high-accuracy input data are essential. If panel layouts, terrain data, and the positions and dimensions of surrounding obstacles (trees and buildings) are not accurate, simulation results will deviate from reality. In other words, “simulation accuracy depends on input data accuracy.” So how can such high-precision data be obtained? One answer is the use of 3D point cloud data acquired by LRTK, described next.

Advantages of high-precision point cloud surveying with LRTK

LRTK is a surveying solution centered on a very small RTK-GNSS receiver that attaches to a smartphone, enabling anyone to easily perform high-precision 3D surveying (point cloud measurement). RTK-GNSS is a method that combines satellite positioning with correction data to determine positions with centimeter-level accuracy, and when LRTK is attached to a smartphone it allows the smartphone’s position to be tracked with cm-level accuracy (cm-level accuracy (half-inch accuracy)). Meanwhile, recent smartphones (e.g., models equipped with LiDAR sensors) have built-in capabilities to 3D-scan environments several meters ahead (several ft). LRTK’s combination of the smartphone’s 3D scanning function and high-precision positioning makes it possible to “precisely capture wide areas as undistorted point clouds.” Tasks that used to require specialized laser scanners or drone surveys can be completed by a single person with a smartphone in a short time, greatly reducing on-site burden.

The concrete advantages of point cloud surveying with LRTK are as follows:

• High precision (centimeter-level): Point clouds obtained with RTK-GNSS are assigned geographic coordinates (absolute coordinates), and positional errors of each point are kept to around a few cm (a few in). This enables accurate understanding of the heights and positions of measured trees and terrain in a real coordinate system, dramatically improving the accuracy of models used in simulations.

• Coverage and level of detail: Point cloud data capture terrain undulations and structure shapes down to fine details. They can record tree canopy spread and complex ground surface irregularities that are difficult to measure manually, reducing oversights. With the idea of “digitizing the entire site,” you can review data later in the office and check dimensions of items of interest.

• Immediacy and efficiency: Because only a smartphone and a small device are required, surveying can begin whenever needed. The measurement itself completes quickly, and the point cloud is generated on the smartphone screen in real time so you can check for omissions on the spot. You can also measure required dimensions (e.g., tree height or distance) on site, enabling part of the analysis to be done concurrently and speeding up on-site decision-making.

• Low cost: Dedicated 3D laser scanners and survey drones are very expensive (costing as much as luxury cars) and are large and cumbersome to transport and install. In contrast, with an LRTK device and a compatible smartphone, you can deploy the solution at a fraction of the cost of dedicated equipment. If you can use an existing smartphone, the cost advantage is even greater. Reducing the frequency of outsourcing as-built surveys or earthwork volume calculations to external survey companies can further cut costs. This makes adoption easier for small and medium enterprises, and enables building a “one smartphone per person” surveying system.

As described above, using LRTK makes it easy to obtain high-precision, high-density point cloud data, making the “digitalization of surrounding environments” required for solar plant shadow analysis dramatically easier.

Flow from LRTK point cloud data to 3D models and PVsyst import

How exactly can data obtained by LRTK be utilized in PVsyst simulations? The general steps are as follows.

• On-site point cloud acquisition: Bring a smartphone + LRTK to the planned site and its surroundings and walk around the entire planned plant area while performing 3D scanning. For example, scan evenly the trees near site boundaries, existing buildings, and ground elevation changes—objects that could cause shading. Point clouds obtained with LRTK can be checked instantly on the smartphone to verify there are no omissions on site.

• Processing point cloud data and creating 3D models: From the acquired point cloud data, create 3D models in formats suitable for simulation. Remove unnecessary noise points and, if needed, convert point clouds of trees and buildings into polygon meshes. For wide-area terrain, extract only ground surface point clouds to generate a Digital Terrain Model (DTM). By using services such as the LRTK cloud, you can upload point cloud data for browser-based 3D viewing and simple editing. When 3D model data of the surrounding environment (terrain, trees, and structures) are ready, export them in a format readable by PVsyst (for example, Collada format (.dae)).

• Importing the model into PVsyst: In the PVsyst project, open the “Near Shadings” 3D scene editor and import the 3D model you generated. Thanks to LRTK’s absolute coordinate data, model scale and placement should reflect reality, though you can adjust positions within PVsyst if necessary. Also set the solar panel layout on the site (either place arrays within PVsyst or import a layout created in external CAD). With panel layout and surrounding environment objects combined into one 3D scene, you have recreated the real plant in virtual space.

• Running shading simulation and analyzing results: Once the 3D scene is complete, run the generation simulation in PVsyst. The software calculates the sun position for each hour of the year (8760 hours) and determines the fraction of each panel that is partially or fully shaded. It then computes energy forecasts that consider both irradiance losses due to shading and electrical losses. The simulation yields annual and monthly energy generation figures, system performance (PR), and most importantly, the amount of loss due to shading is specified in the report. For example, the report might state “Shading loss from nearby objects: △△ kWh/year (▲▲%)”, quantitatively showing how much generation loss trees or structures cause. These results form the basis for the next design optimization steps.

Example of design optimization based on simulation results

With high-precision shading simulation results, you can optimize plant design accordingly. Because simulations clarify “when,” “where,” and “how much” shading occurs, prioritization and strategy for countermeasures become clear. The following is an example of how shading analysis results were used to improve design.

Example: In one megasolar project, a forest about 15 m (49.2 ft) tall was adjacent just outside the site’s eastern boundary. From LRTK-acquired point clouds, the heights and positions of these trees were accurately modeled in 3D, and PVsyst simulation revealed that nearly half of the east-facing arrays were covered by tree shadows every morning. This caused not only reduced irradiance on the affected strings but also mismatch losses, resulting in an annual generation loss of several percent. Initially, the forest’s shading was underestimated at the design stage, but detailed simulation results prompted consideration of countermeasures.

The design team first tested a physical measure: relocating several panel rows near the site boundary. Specifically, they reduced panels in the most affected area and redistributed them to less shaded areas, altering the layout. This reduced the number of panels that became fully shaded in the morning and improved predicted annual generation. As a technical countermeasure, they also considered installing power optimizers on some strings that unavoidably experience shading to reduce local mismatch losses. Simulation of this solution showed that optimizers mitigated output differences between shaded and unshaded strings and recovered some generation loss. All of these optimization studies were possible thanks to the detailed data provided by LRTK. As a result, an arrangement and equipment configuration that minimized shading impact was identified, and the plant’s annual predicted generation was improved to a more reliable value.

As this example shows, quantitative data from precise shading analysis provide strong evidence to support design decision-making. Judgments that used to rely on experience or intuition about “how much shading to expect” can now be made rationally based on PVsyst numerical results. For example, you can estimate “felling a certain tree will increase annual generation by ○○ kWh” or “optimizing panel layout can mitigate monthly PR declines by ▲▲%,” enabling design optimizations that consider investment return.

Conclusion: High-precision shading analysis anyone can do with smartphone-based LRTK surveying

In solar power plant shading analysis, accurately simulating the effects of surrounding trees, structures, and terrain without overlooking anything has become indispensable. By importing high-precision 3D models created with LRTK into PVsyst, you can run simulations that faithfully reproduce the real environment and dramatically improve the accuracy of generation forecasts. This allows identification of potential risks and losses at the planning stage and implementation of appropriate design changes and countermeasures.

Notably, this advanced analysis workflow can be completed using only a smartphone. The advent of LRTK has made 3D surveying—previously requiring specialized vendors or expensive equipment—accessible to on-site technicians. By simply walking the site with a smartphone, you can acquire high-precision point cloud data and then feed it into software for simulation. LRTK, a simple surveying tool anyone can use, brings a new DX (digital transformation) to the solar power design process.

Going forward, the combination of smartphone-based LRTK surveying and PVsyst simulation is likely to become a new standard in solar power plant design. With meticulous simulations that “don’t overlook tree shadows,” you can robustly support the reliability and profitability of power generation projects. Armed with high-precision 3D models, please realize optimal design and operation in your projects.