Download How to Design a SolarEdge System

March 2015
Application Note - How to Design a SolarEdge System
Using PVsyst
As of version 5.20, PVsyst - the PV system design software - supports the design of SolarEdge systems. This application note
details the SolarEdge-specific design steps.
This application note assumes the reader has prior knowledge of the basic use of PVsyst.
This document explains the unique SolarEdge design concepts as they are realized in PVsyst and guides the user through setting
up a shading scenario using the SolarEdge system.
Introduction
The process of designing a SolarEdge system in PVsyst includes the following basic steps:
1
Project – define the location and meteorological data
2
Orientation – define module azimuth and tilt
3
System – choose the system modules, inverters and electrical design
4
Near Shadings – design the system layout in 3D and select shading calculation method
5
Simulation – view a summary of the system design and its energy output
This document briefly covers steps 1, 2 and 5 (Project, Orientation and Simulation) which are done in the same manner for
SolarEdge and non-SolarEdge systems, and describes in more detail steps 3 (System) and 4 (Near Shading), as they pertain to the
design of a SolarEdge system in PVsyst.
NOTE
While PVsyst offers basic system validation, it is recommended that you always use the SolarEdge Site Designer
software for design verification.
Step 1: Project
Choose “Project Design” and then “Grid Connected” in the main PVsyst screen. Click on “New Project” and then choose the
correct project site and meteorological file (the meteorological file will often be automatically associated with the chosen site).
In order to add a new site to PVsyst, click on “Databases” in the main PVsyst screen and then “Geographical Sites.” Choose
“New” and locate your site on the interactive map, or type the geographical location in the search box. Click “Import” and then
“Import” again (for the second Import, make sure to click on the button that has a sun icon). Click OK and then Save. When
prompted to save the hourly values, click “Yes” and then “Close” and Exit”. You can now choose the new meteorological site for
your project.
1
Step 2: Orientation
Once the geographical location and climate file have been chosen, you will define the array’s azimuth and tilt. PVsyst offers
different options to fit various types of projects, including simple fixed tilted plane, multiple orientations (up to 8 orientations),
seasonal tilt adjustments, ‘unlimited sheds’ for large systems, sun shields (modules mounted to facades of buildings) as well as
various kinds of tracking arrays, both single and double-axis.
Step 3: System
The main system parameters, including module and inverter model, system capacity, string lengths etc., are defined in this
screen. This is where the SolarEdge system advantages become apparent.
Basic Design
After choosing a PV module, select the SolarEdge inverter applicable for your project. You can then use the drop down list below
the modules to select the appropriate power optimizer. If you plan to connect more than one module per optimizer, use the
following window to specify how many modules to connect and the type of connection (series/parallel):
The Inverter Input tab in the SolarEdge interface lists the number of power optimizers per string, not the number of modules.
Thus, for example, when connecting two modules per power optimizer the string length would be 20, meaning 40 modules:
2
Uneven String Design
Unlike a traditional inverter, the SolarEdge system introduces a high level of design flexibility. A common limitation of a
traditional system is that all strings connecting to an MPP tracker must be of the same length. The SolarEdge system has no such
limitation since MPPT is performed at the module-level using the SolarEdge power optimizers. Multiple strings can be connected
to one inverter, regardless of length, module azimuth, tilt or module type. In order to accommodate this capability, PVsyst has
incorporated a feature called “fractional inputs.” This feature allows viewing strings as fractions of an inverter. By assigning
different fractions of an inverter to different strings, the sum of these fractions adds up to the whole inverter. The following 3
examples explain this concept.
NOTE
Currently PVsyst can only design homogenous strings, that is, each string must have one orientation and module
type. In practice, the SolarEdge system can be designed with non-homogenous strings.
Example 1
Consider a SolarEdge inverter with 70 modules divided into 2 strings: one with 42 modules, and one with 28 modules.
For this design, use 2 PVsyst subfields: one for each string. In the System screen, check “Uses fractional inputs” and in each
subfield’s “Nb. of inverters” fill in the relative portion of that string.
In this case, Subfield 1 has 0.6 inverters (42/70) and subfield 2 has 0.4 inverters (28/70).
In the “Global system summary” field, verify the number of inverters is 1:
Example 2
Consider a system with 3 SolarEdge inverters, each with one string of 35 modules and one string of 34 modules (69 modules per
inverter).
The system has 6 strings (3x35 + 3x34) and 207 modules. Conceptually, ‘half’ of the 3 inverters are connected to strings of 35
modules, and the other half are connected to strings of 34 modules.
In this case create 2 subfields: one for each string length. One subfield will have 1.5 inverters:
(35/69) X (the number of inverters = 3)
where each inverter has a string of 35 modules, and the other subfield will have 1.5 inverters, and each inverter has a string of
34 modules (each subfield has 1.5 inverters because PVsyst rounds the fraction to the first digit, and here the difference
3
between the string lengths is small). The total ‘amount’ of inverters will be 3:
Example 3
Consider a system with 16 SolarEdge inverters and the following design:


7 inverters, each with two strings of 34 modules
9 inverters, each with one string of 34 modules and one string of 35 modules
Total number of modules in the system is 1097:


23 strings of 34 modules = 782 modules (71.285% of the total number of modules)
9 strings of 35 modules = 315 modules (28.715% of the total number of modules)
In other words, 71.285% of the modules belong to strings of 34 and 28.715% of the modules belong to strings of 35. Thus:
71.285%*16 = 11.4 inverters connected to strings of 34 modules
28.715%*16 = 4.6 inverters connected to strings of 35 modules
The 2 PVsyst subfields should look as follows:
Step 4: Near Shadings
The SolarEdge system minimizes the energy loss due to partial shading when compared to a traditional inverter. If the system in
question has shading of any kind (trees, chimneys, inter-row shading) the 3D physical system layout should be designed to
account for the energy loss due to shade. An accurate shading simulation will ensure the SolarEdge advantage is reflected in the
annual energy output of the system. If needed, you can refer to the PVsyst help files for assistance in constructing a 3D shading
scene.
PVsyst can calculate the shading according to one of 2 methods: “linear” and “according to strings”.
1
In “Linear” shading calculation, the energy loss is linear to the PV area being directly shaded. This method does not take into
account the electric interdependency of modules in a string, nor does it allow for the electrical behavior of a partially shaded
single module (the protection of bypass diodes).
2
In shading calculation “According to module strings,” the energy loss due to shading takes into account the effects of partial
shading on the entire string. Choosing this method requires division of the PV areas into rectangles corresponding to the
size of the strings designed in the System section. This is a more realistic method and therefore more accurate. Use this
method to compare a simulation of a traditional system vs. a SolarEdge system, as described below.
4
Unlike a traditional system, in which the production of whole strings is affected by partial shading of as little as one module,
SolarEdge confines the effects of partial shading to the shaded modules only, thanks to its module-level MPPT. The way to
simulate shading in a SolarEdge system in PVsyst is to divide the PV area into strings corresponding to the size of one bypass
diode. Depending on the number of bypass diodes in the chosen module, this will usually result in 3-4 substrings per module,
running along the long side of the module, as in the image below (module graphics have been inserted for clarity purposes).
Once the shading scene is complete and the PV area has been divided into substrings, select the “According to module strings”
option in the “Near shadings definition” screen.
NOTE
When designing the same system based on a traditional inverter, define the string division rectangles based on
the number of modules per string, and not based on bypass diodes.
Step 5: Simulation
Once the system parameters have been defined correctly, click on “Simulation”. When clicking “Simulation” again the simulation
will begin, and when finished a report will be available for viewing and printing. In addition, the Simulation screen allows you to
change the time frame of the simulation (from one day to one year), as well as exporting hourly values of various parameters
including energy, inverter efficiency, PV array’s electrical behavior and more. Depending on the complexity and size of the
system, the simulation process may take from a few seconds to over an hour.
5
General Notes



SolarEdge eliminates all losses that result from mismatch, compared to 2% in a traditional system. Mismatch losses are
caused by manufacturing tolerance of the modules, temperature differences during operation, uneven soiling and other
environmental factors such as overcast weather conditions.
When selecting a SolarEdge inverter, PVsyst automatically sets the mismatch losses to zero. If more than 2% mismatch is
indicated, change this parameter by clicking the “Detailed Losses” button at the bottom of the System screen of the
traditional inverter. This parameter should always remain zero for SolarEdge.
The Module Layout section in PVsyst is not yet implemented for SolarEdge, however may be used for traditional systems.
Note that for some shading scenes, such as: very thin pole shading, light shading due to small roof protrusions, complex
commercial systems etc., the Module Layout method may provide inaccurate shading losses.
Appendix: PVsyst Loss Parameters
PVsyst calculates several loss parameters during the simulation, as shown in the losses diagram below. This diagram appears at
the end of each PVsyst report. Following is an overview of the loss parameters, calculated successively.








Horizontal global irradiation: a combination of the global diffuse irradiance and the global beam irradiance calculated
hourly over one year on a horizontal surface.
Global incident in collection plane: using a transposition model (either Perez or Hay), PVsyst increases or decreases the
horizontal global irradiation depending on the azimuth and tilt. This value represents the actual irradiation reaching the
module.
Near Shading: irradiance loss: the loss of energy due to direct shading (affecting the beam component of the irradiance)
and diffuse shading (affecting the diffuse component of the irradiance. For example: nearby objects such as the next row of
modules in a ground mounted system decrease the diffused light hitting the modules even if they do not cause direct
shading).
IAM factor on global: IAM (Incidence Angle Modifier) relates to the decrease of irradiance reaching the PV cell due to the
sun rays’ refraction when passing through the PV module antireflective coating and glass. The higher the incident angle
(with respect to the sun’s position) the higher the loss.
Effective irradiance on collectors: the remaining irradiation following the previously detailed losses, multiplied by the PV
area (the module area as defined in the *.PAN file).
PV Conversion: the module efficiency at STC.
Array nominal energy (at STC efficiency): the PV conversion efficiency multiplied by the effective irradiance on collectors.
PV loss due to irradiance level: calculates the reduced module efficiency in low light conditions.
6












PV loss due to temperature: represents the reduced module performance due to temperature change. The module’s
temperature coefficients as well as the thermal loss factor parameter affect this loss. The thermal loss factor is not a
scientifically determined value; rather it is set by the system designer according to experience and previous estimations. A
value of 20 W/m²k is acceptable for most systems. If the modules are embedded into a roof structure, the value can be set
to 15. When the modules are free-standing in a cool and windy location, the value can be set to 29.
Shading: electrical loss according to strings: in addition to the Near Shading irradiance loss, this represents the energy lost
due to the electrical effect of shading. For example, in a system with strings connected in parallel, one shaded module can
cause a whole string to be bypassed due to voltage mismatch.
Module quality loss: This parameter makes use of the manufacturing tolerance values in the *.PAN file (i.e. ±2% tolerance).
The formula used is: The lower tolerance plus a quarter of the difference between lower and higher tolerance. So for
example, for a module with a ±2% tolerance, the quality loss is
[-2%+(0.25 x 4%)] = -1%
Module array mismatch loss: The energy loss due to mismatch between modules in a string. The 2% mismatch for
traditional string inverters stems from manufacturing tolerance, un-even soiling, temperature differences between modules,
tilt or azimuth differences within the string, etc.
Ohmic wiring loss: The voltage drop due to wiring resistance is calculated as one value for the whole system. If there are
differing conditions such as strings and combiner boxes with varying lengths of cable, it is not accounted for in the
simulation. The default value of 1.5% loss from STC is recommended. This translates to a reasonable actual loss of around
0.6% on the DC side for most systems.
Inverter loss during operation (efficiency): The inverter’s DC to AC conversion efficiency, weighted for variance in power
levels over the year.
Inverter loss over nominal inv. power: The power clipping in overloading conditions (where the array produces more DC
power than the maximum AC output of the inverter).
Inverter loss due to power threshold: The loss of energy when the array operates below the inverter’s minimum power
threshold (defined in the *.OND file).
Inverter loss over nominal inverter voltage: The energy loss when the array is producing voltage below the inverter’s MPP
voltage range.
Inverter loss due to voltage threshold: The energy loss when the array is producing voltage above the inverter’s MPP
voltage range.
AC ohmic loss: As with the DC wiring loss, a recommended value of 1% AC loss with respect to the STC value will produce
around 0.5% actual energy loss.
External transformer loss: Unless actual parameters are available, the default 0.1% iron loss and 1% resistive loss are
recommended
7