Solargis Evaluate PV simulation and losses

Overview

Our Evaluate PV simulator employs raytracing (in combination with Perez’s all-weather sky model) and complex Electrical component models to provide unparalleled accuracy in estimating the energy yield of PV systems. Leveraging high-resolution Solargis Time Series (TS) or Typical Meteorological Year (TMY) data, this simulation offers comprehensive loss calculations at every stage that surpass current market standards. For a better understanding of the concept, processes, and methods involved, we have segregated the information into three main sections:

• Simulation inputs: This section will cover both technical and site parameters. Technical parameters include the configuration of the energy system, such as module and inverter specifications provided by the PV Components Catalog (PVCC), while site parameters encompass sun geometry, terrain, and solar radiation inputs (Global horizontal irradiation (GHI) and Direct normal irradiation (DNI)) from Solargis data sources. These inputs are crucial for setting up the simulation environment.

• Optical simulation: This part will delve into the optical aspects of the simulation, utilizing raytracing and Perez’s all-weather sky model. These models are used to calculate the Global tilted irradiance (GTI) per PV module (cell) and the related losses, such as those caused by far/near horizon shading or soiling.

• Electrical simulation: This section will explain how the electrical configuration of the PV system is simulated, including energy losses in the system and how they are accounted for in the final energy yield calculation. The electrical simulation ensures that all components, from PV modules to grid connections, are accurately modeled.

This simulation is incorporated in the Solargis Evaluate solution, and the overall simulation chain and sequence of processing steps are illustrated in the Evaluate simulation diagram below. Each step in the simulation chain yields intermediate results, which are described in the sections below. These intermediate parameters can be used for detailed analysis of the PV power plant’s expected performance, and can be exported in Solargis Evaluate application.

Simulation inputs

Simulation inputs in Solargis Evaluate are categorized into two main types: Site parameters and Energy system configuration. Each plays a crucial role in accurately modeling the PV system's performance.

Site parameters

Site parameters encompass environmental and meteorological conditions that affect the PV system's performance.

Energy system configuration

Any change performed by the user on the energy system within the Energy system designer affects the simulation:

• Physical layout (segments, mounting system, spacings, table layout, shading objects)

• Electrical layout (PV module, strings, inverter and transformer specifications, grid connection)

Shading objects

All elements of the energy system configuration are considered in shading calculations. Users can also specify additional shading objects like fences, tree lines, or buildings by defining them in the Energy system designer or importing them via KML files.

PV module & inverter specifications

These are provided by the PV components catalog (PVCC), which offers verified specifications in the required format. The PVCC ensures high-quality inputs, leading to consistent and reliable simulation results.

Optical simulation

The optical simulation in Solargis Evaluate is the first part of the energy yield calculation process, calculating the Global tilted irradiance (GTI) on PV modules. This section will delve into the detailed steps and models used in the optical simulation process.

Sky irradiance model

In Solargis Evaluate, satellite-derived Global horizontal irradiation (GHI) and Direct normal irradiation (DNI) variables are used to calculate the distribution of diffuse radiance on the sky dome. This is achieved through the implementation of Perez all-weather sky model. From these inputs, the Sun position and the configuration of the power plant, the theoretical Global tilted irradiation (GTI) without any losses is also calculated

Solar position calculation

The solar position is calculated using the PSA model from the geographical coordinates of the site and date-time information.

Far horizon shading

Far horizon shading effects are simulated using View Factor model. The default horizon from Solargis data is used at the project reference point. For energy systems with segments defined, the far horizon is pre-loaded for each segment’s reference point. Far horizon is defined by azimuth and height and can be edited in the Energy system designer.

Near shading

Near shading simulation is the most complex step in the Evaluate PV Simulator pipeline. To quantify near shading losses, all objects from the Energy system design and the surrounding terrain are put into a 3D calculation scene, in which direct and diffuse light simulation is run using backward raytracing.

3D calculation scene

The 3D calculation scene is an important component of the near shading simulation in Solargis Evaluate. It is constructed using inputs from the Energy system designer, terrain data, horizon data, and albedo values. The purpose of this scene is to accurately model the surfaces of PV modules and all objects in the area that can cast shadows or reflect solar radiation.

Components of the 3D scene:

• PV modules: These are the primary surfaces for which incident solar radiation is calculated.

• Objects: Include support structures, inverters, transformers, and any additional shading objects specified by the user, such as fences or buildings.

• Terrain: The terrain model is integrated to account for its impact on shading and reflections.

Simulation dynamics

The lighting and shading within the 3D scene are dynamic, depending on the solar position and solar radiation values. These factors vary over time and are recalculated for each time step of the simulation, ensuring that the simulation accurately reflects real-world conditions.

For a visual representation, refer to image below. This image illustrates how PV modules, objects, and terrain are integrated into the simulation environment.

Backward raytracing

Backward raytracing, specifically unbiased Monte Carlo path tracing, is a key method used in Solargis Evaluate to calculate the irradiation on PV modules. This process involves the following steps:

1. Direct illumination calculation: Determines whether sample points on PV cells are in direct sunlight or shaded, calculating a shadow ratio for partially shaded cells.

2. Diffuse radiation calculation: Generates random rays and traces them through a 3D simulation scene, recording ray direction and treating intersections as Lambertian reflections.

3. Post-processing: Denoises diffuse radiation values and resamples them per cell, summing them with direct radiation to obtain the final Global tilted irradiance (GTI).

For bifacial Energy systems, the simulation is done separately for front and rear side of the PV modules. This high accuracy method allows detailed simulation of light to an extent that even gaps between PV cells are considered.

Soiling losses

Soiling losses are an essential factor in the energy yield simulation, as they affect the amount of solar radiation that reaches the PV cells. In Solargis Evaluate, these losses are applied to the GTI calculated in the previous step. For bifacial energy systems, the simulation is done separately for the front and rear sides, while 15% of the front side loss value is applied to the rear side.

• Method of application: Soiling losses are typically applied as average monthly figures or as a single yearly figure. Users have the option to specify these values if needed.

Angular reflection losses

Angular reflection losses occur due to the angle of incidence effects on the surface of PV modules. These losses are significant because they affect the resulting radiation reaching the PV cell.

Solargis Evaluate employs the Martin and Ruiz model to estimate angular reflection losses. This model uses an angular loss coefficient, which is estimated by Solargis based on the properties of the PV module surface, particularly its soiling. For bifacial energy systems, the simulation is done separately for the front and rear sides.

Spectral correction

Spectral correction is an essential step in the simulation process, and Solargis Evaluate uses the Lee & Panchula model for this purpose.

The specific intensity of the spectral responsivity correction depends on two key atmospheric factors:

• Air mass: This represents the optical path length of sunlight through the Earth's atmosphere. It increases as the Sun's position moves closer to the horizon, affecting the spectral distribution of sunlight.

• Precipitable water content: This refers to the total amount of water vapor present in a column of the atmosphere.

Electrical simulation

Conversion of irradiation to DC electricity

The Single diode equivalent circuit model, also known as De Soto's "Five Parameter" model, is used in Solargis Evaluate to simulate the conversion of solar irradiance into electricity within PV cells.

The Single diode model requires five key parameters to describe the current-voltage (IV) curves of PV cells. These parameters are typically acquired at Standard Test Conditions (STC):

• Modified ideality factor

• Diode saturation current

• Light current (photocurrent)

• Series resistance

• Parallel resistance

These parameters, along with the Global tilted irradiance (GTI) and cell temperature, are used to generate the IV curves for each PV cell.

Note: When calculating the IV curves of bifacial PV modules, the bifaciality factor (as specified in the PV components catalog) is applied to the rear-side GTI to account for the decreased efficiency of the PV module’s rear side.

Advantages of the model

The use of the Single diode model in conjunction with raytracing allows for detailed analysis of shading conditions at the level of individual PV cells. This enables precise simulation of the electrical performance of PV modules under various environmental conditions.

Transient thermal correction model

To calculate the cell temperature, we use the Transient thermal correction model to account for the thermal inertia of PV modules by smoothing 1-minute temperature data with a weighted average over the previous 20 minutes. In the case of 15-minute data, only the current and previous time slots are considered. As a result, module temperature changes more gradually, reflecting real-world behavior and improving the accuracy of performance predictions.

IV curve addition

Solargis Evaluate PV simulator calculates the IV curve for each PV cell in the power plant, based on De Soto’s model. The IV curves of the PV cells are then summed together for the submodules, PV modules, and strings, following the power plant electrical layout and electrical circuit physics. The string’s IV curves then enter the next stage of the PV yield simulation.

Using this approach, the performance of each string in the PV power plant is accurately calculated, reflecting the real operating conditions such as partly-shaded PV modules. This means that no assumptions on partial shading performance have to be made, which is sometimes a requirement in other PV yield simulation software.

Inverter power limitation losses

Inverter power limitation losses, also known as clipping losses, occur when a PV array generates more DC electricity than the inverter's maximum rated AC output or more than the specified grid limit. The inverter then "clips" or discards this excess power, meaning the potential energy that could have been produced is lost.

Clipping may arise from two sources:

• Inverter clipping: Internal hardware limits prevent conversion beyond the rated capacity of the device.

• Grid power limitation: Output of the power plant capped to meet grid stability requirements.

DC losses

DC losses in the direct current path from PV modules to inverters are a crucial factor in energy yield simulations. These losses occur due to the electrical resistance in the DC cables, and other DC components such as combiner boxes.

Default value used in Solargis Evaluate is 2%.

Inverter DC/AC conversion

An inverter is an electronic device that converts direct current (DC) generated by PV modules into alternating current (AC). The output can be either a one-phase or a three-phase voltage. Simulation of inverters is based on inverter parameters taken from PV components catalog.

In Solargis Evaluate, the inverter conversion calculation consists of two main stages:

• Maximum power point (MPP) calculation: The Maximum power point (MPP) calculation utilizes the IV curve at the inverter input, which is derived from the previous simulation steps.

• DC/AC conversion model: Solargis Evaluate uses the Sandia inverter model which considers operational and environmental conditions in determining the inverter efficiency and its AC output. It also calculates the active and reactive power components, considering the user-specified power factor and inverter capabilities.

Auxiliary losses

Auxiliary losses in an energy system are caused by various equipment that consumes energy, including systems for protection, monitoring, heating or cooling (depending on the climate zone), lighting, module tracking, and other energy-consuming devices.

The auxiliary losses can be divided into two categories:

• Constant losses: These are continuous losses measured in watts, and are split into night and day constant losses.

• Proportional losses: These depend on the power produced by the power plant in each moment, and are expressed as watts per kilowatt of installed power.

Default values used in Solargis Evaluate:

Transformer losses

Transformers are essential devices in energy systems, used to change the voltage level from the AC side of inverters to the desired voltage level for connection to the utility grid. In Solargis Evaluate, we utilize our proprietary Transformer model, which can be segregated into two sub-models:

• Variable losses model: Includes iron (no-load) losses and copper (load) losses.

• Constant losses model: Transformer losses are represented as a percentage reduction of electrical power at the primary side of the transformer.

If the difference between the AC voltage on the inverter output and the grid connection is sufficiently large, Solargis Evaluate considers several transformer stages - inverter and power transformer. In such case, the transformer losses are considered separately for each transformer and AC cable losses are considered in between the transformer stages, all depending on the respective settings.

Default values used in Solargis Evaluate:

Variable losses model

Constant losses model

AC losses

AC losses in an energy system occur in the AC cabling, affecting the transmission of electricity from the inverters to the grid connection point.

In Solargis Evaluate, required AC cable losses are set as a percentage value. This percentage represents the total AC electrical loss across the AC electrical network at reference conditions, typically Standard Test Conditions (STC). The AC losses are applied in several stages (low, medium, and high voltage) depending on the number of transformer stages (inverter and power transformers) in the power plant.

Default values used in Solargis Evaluate:

Post processing

After calculating the power output of the PV power plant using the Solargis Evaluate PV simulator, Solargis Evaluate accounts for further system losses in the post processing step. These losses are quantified in the Solargis Evaluate Analysis section and reports, but are not included in the exported data.

Unavailability and snow losses

System unavailability losses quantify the electricity losses incurred due to the shutdown or power output limitation of the energy system or its components. These losses can be categorized into two main types:

• Technical events: Incurred due to internal reasons (equipment failures or scheduled maintenance work) and external reasons (Grid connection issues, curtailment).

• Weather-incurred events: Caused by the snow coverage of PV modules.

Default values used in Solargis Evaluate:

Note: The implementation of technical losses is based on IEC Technical Specification 61724-3, which outlines energy evaluation methods for photovoltaic systems.

Long-term degradation

The performance of PV modules and other components decreases over time, and long-term degradation serves as a measure of this performance reduction. Typically, PV components experience more rapid degradation in the initial years of their lifespan.

Solargis Evaluate models long-term degradation looking 25 years forward into the PV power plant operation. We use the long-term average yearly power production (PVOUT specific and total) and Performance Ratio (PR) from the simulated historical data, and apply the specified degradation rates to these figures to estimate the expected future PVOUT and PR.

Degradation rates

Based on existing in-field experiences from commercial projects, the long-term annual performance degradation for well-manufactured PV modules may be approximately:

• 0.8% for the first year

• 0.5% for subsequent years

Note: This assumption includes initial degradation of the modules.

Simulation outputs

The result of the simulation is the power output of the energy system, referred to as PVOUT. It is quantified both in absolute numbers and as specific PVOUT, normalized to the installed capacity of the power plant.

The simulation results are visualized in numerous charts and tables in the Solargis Evaluate Analysis section. The data is categorized and segregated into different units with a data interpretation guide included for every presented value, making the data easy to understand and interpret. The power conversion losses are specifically quantified and visualized as average yearly figures (including the Sankey diagram shown below), and monthly breakdowns.

Further reading

• A new simplified version of the Perez diffuse irradiance model for tilted surfaces. Richard Perez, Robert Seals, Pierre Ineichen, Ronald Stewart, and David Menicucci.

• All-weather model for sky luminance distribution—Preliminary configuration and validation. R. Perez, R. Seals, and J. Michalsky.

• Updating the PSA sun position algorithm. Manuel J. Blanco, Kypros Milidonis, and Aristides M. Bonanos.

• Comparison of ray tracing rendering technique with ground measurements for improved solar radiation modeling.  L. Dvonc, P. Orosi, T. Cebecauer, and B. Schnierer.

• Simple Model for Predicting Time Series Soiling of Photovoltaic Panels. M. Coello and L. Boyle.

• Calculation of the PV modules angular losses under field conditions by means of an analytical model. Martin N. and Ruiz J.M.

• Spectral correction for photovoltaic module performance based on air mass and precipitable water. Lee, M., & Panchula, A.

• Improvement and validation of a model for photovoltaic array performance. W. De Soto, S.A. Klein, and W.A. Beckman.

• Lambert W function for applications in physics. D.Veberič.

• Transient Weighted Moving-Average Model of Photovoltaic Module Back-Surface Temperature. M. Prilliman, J. S. Stein, D. Riley, G. Tamizhmani