Size the inverter and the strings, checked on every change
pvSpline checks each string against the voltages of its inverter and its modules, and the inverter against the array it serves. The checks run every time the design changes, and the losses they warn about appear in the simulation, line by line.
A starter inverter for the array
Create initial design writes a starter design straight away: one generic inverter with two MPPT inputs and a 1000 V DC limit, rated for the installed power drawn on the map at a DC/AC ratio of 1.15. A 36 kWp roof gets a 31.4 kW starter. Replace it with your inverter, picked from your favourites or imported from its .OND datasheet, which brings its MPPT window, its DC voltage and current limits and its efficiency curve.
String length, checked at the site’s temperatures
Set the modules per string and the site’s minimum and maximum cell temperatures. The validator then holds each string’s open-circuit voltage at the coldest temperature below the inverter’s maximum DC voltage and the module’s own maximum system voltage, and its maximum-power voltage at the hottest temperature inside the inverter’s MPPT window. With mixed modules, such as a roof and a carport of another technology, Limit string length to MPPT window caps the length to the strictest window.
The DC/AC ratio
The validator warns when the array is too small for the inverter, “inverter heavily oversized”, or too large, “significant clipping expected”. The default band runs from 0.50 to 1.50; bifacial and tracker plants often run at 1.5 to 1.7 by design, so widen it for them in the diagram settings.
MPPT inputs and routing
The strings are spread over the combiner boxes and balanced over the MPPT inputs automatically; route any of them by hand, in a table or by dragging its card on the diagram. The validator flags an MPPT input given more strings than it takes and a combiner box with too many inputs.
What the inverter costs in the yield
- Clipping: the energy lost above the inverter’s nominal power has its own line in the loss cascade, as do the losses below its start power and outside its MPPT voltage window.
- Each MPPT input, hour by hour: solved with the strings routed to it, each at its own length, so a string too short or too long for the window shows in the yield.
- Down to the minute: a time step from 60 minutes down to 1 captures the clipping that an hourly average smooths away.
- Staging: identical inverters can share an array in master/slave operation, the slaves starting one by one as the power rises.
Inverters compared in a batch
A batch study scores each candidate inverter on strings laid out for it, at the series length the module’s voltages and the inverter’s MPPT window allow, through the same simulation as your report.
Size an inverter and its strings in five steps
- On Diagram, click Create initial design, or add your inverter with Add inverter.
- Select the inverter and pick it in Pick from favourites, or click Import a .OND file….
- Under DC array sizing, set Modules / string, Site Tmin (cells) and Site Tmax (cells).
- Read the Issues tab and adjust until no string and no ratio is flagged.
- Run the simulation and read the inverter lines of the loss cascade.
Questions
Which maximum system voltage does a string answer to?
The module’s IEC rating under UTE C15-712-1 and IEC 60364, its UL listing under the NEC; a module that states only one rating is held to that one. The electrical code setting decides.
Why do the voltage checks not run?
Both site temperatures must be set first: each reads “unspecified” at its minimum, and a line under the fields says which state you are in.
Does a string painted on the map count?
In the simulation, yes: a string wired on the map is simulated at its own length. The voltage checks test the modules-per-string value, so a string painted longer escapes them; keep the two consistent.
The fastest way to judge a simulation tool is to run it on a project you already know.