22.5% Efficiency TOPCon Half-Cut Modules: Spec Sheet Guide

Published: October 19, 2026 | Author: Weltrus Energy Team | Reading Time: 12 minutes

When 22.5% Is Copied Off the Banner

Module efficiency is a measured ratio, not a product name. Teams still paste “TOPCon half-cut, up to 22.5%” into a comparison as if every module in the family produced that ratio. The lower bin on the same outline does not. A different length does not. A cell-efficiency number from a wafer slide does not. The yield model then starts from a wattage the factory will not ship.

TOPCon (tunnel oxide passivated contact) is the cell stack. Half-cut means each cell is divided and rewired, usually so the module current and the shade path change. Neither phrase is the efficiency. Efficiency is nameplate power divided by the sunlight power on the module area, at standard test conditions: 1000 watts per square meter, 25 degrees Celsius cell temperature, and the AM1.5 spectrum. How TOPCon compares with PERC in a bid is a separate procurement note: TOPCon versus PERC bifacial procurement. This page is how to read the efficiency line once you are on a TOPCon half-cut sheet.

Weltrus manufactures TOPCon modules across a span of about 100 watts to about 700 watts. There is no single efficiency for that whole span. The number you may quote is the efficiency printed on the datasheet revision for the model you will buy. A family sentence such as “up to about 22.5%” is a ceiling phrase. It is not a promise that the bin on your purchase order tested at 22.5%.

What the Percentage Actually Measures

At standard test conditions the sun is taken as 1000 watts on each square meter. A module efficiency of 22.5% means the nameplate is 225 watts for each square meter of module area. The area is length times width of the module, including the frame, unless the datasheet explicitly says it used cell area. Cell area gives a higher percentage and a smaller surface. Procurement should require module efficiency, because the roof and the container see the frame, not the bare cell.

The check is short. Convert length and width to meters, multiply them, then multiply by 225 if the claim is 22.5%. The result is the nameplate watts that belong with that percentage.

Apply the same division to a real outline instead of trusting the banner. One published Weltrus lot is model FX-630HT-02, specified at 630 watts, with a pallet label module size of 2382 by 1134 by 30 millimeters. That outline is 2.701 square meters. Divide 630 watts by 2.701 square meters and you get about 233 watts per square meter, which is about 23.3% of the 1000 watt standard-test irradiance. The packing note is in the 630 W factory shipment. This division uses the published nameplate and the published outline. It is not a flash-test reprint, and it is not the efficiency of every Weltrus module. It does show why a family line of “up to 22.5%” must not be pasted onto that bin: 22.5% of the same area is about 608 watts, and the published nameplate is 630 watts.

If your datasheet prints both a percentage and a wattage, they have to agree within the power tolerance. A mismatch usually means one of four things: the percentage is cell efficiency, the watts are a different bin, the area includes or excludes the frame differently than you assumed, or the two numbers come from different revisions. Stop the comparison until the supplier says which one is on the module you will receive.

What Half-Cut Changes on the Sheet

A half-cut module starts from a full cell that is cut into two pieces. Those pieces are strung so the module is typically two internal groups side by side, with bypass diodes between sections. The efficiency definition does not change. What changes is the electrical layout you must read before you size the inverter.

Current is the line people skip. Each half-cell carries less current than the full cell it was cut from, which cuts resistive loss inside the cell. The module can still show a high short-circuit current, because the two internal groups are in parallel and large wafers already produce high current. Do not infer amps from 22.5%. Read short-circuit current and operating current on that model, then compare them with the MPPT limit, including parallel strings. How wattage, outline, and current fit together as a buying sequence is in how to spec a wattage from 100 W to 700 W.

Also read the cell count and the diode count as printed, not as a habit. A 108-half-cell module and a 144-half-cell module are different circuits even when both are called half-cut. Three bypass diodes is common and is not guaranteed. The diode splits the module into sections, so shade on one section can bypass that section while another still produces. That is a shade behavior, not an efficiency gain, and it does not raise the 22.5% figure.

Mechanical lines stay mandatory. Half-cut does not tell you the design load, the clamp zone, or the mass. A higher efficiency on a larger outline can raise watts per module and still raise the kilograms the rail must carry. Put the load and the clamp note next to the efficiency. They are not implied by the percentage.

Which Bin Owns the Percentage

Factories sort modules from one bill of materials into power bins after the flash test. On one fixed outline, efficiency moves in step with the bin, because area is the same. The top bin is the one that can print the highest efficiency. A lower bin on that outline is a lower efficiency, even though the cell type is still TOPCon and the cells are still half-cut.

Ask for the bin you will receive in volume, not the bin in the photograph. If the family table runs from a lower wattage to a higher wattage, write the efficiency beside each wattage. “Up to 22.5%” is allowed only on the row that calculates to 22.5%. If the purchase order is the bottom bin, the energy model uses the bottom-bin watts and the bottom-bin efficiency.

Power tolerance sits beside that row. A nameplate with a tolerance that allows watts below the printed number does not deliver the same floor as a nameplate that may only test at or above the printed number. Efficiency printed from the nameplate inherits that floor. When you compare two 22.5% offers, compare tolerance, not only the percentage. Then confirm the datasheet revision. Two revisions can keep the same marketing efficiency after a glass, cell, or frame change. The certificate and the warranty follow the revision, not the percentage. Match IEC 61215 and IEC 61730, or the local equivalent, to the model code, using the IEC module verification checklist.

Temperature Coefficient, With the Arithmetic

Standard test conditions hold the cell at 25 degrees Celsius. A roof or a desert array runs hotter. The power temperature coefficient is the percent of nameplate you lose for each degree above 25 degrees Celsius, at the same irradiance. It is a property of that model. Half-cut and a 22.5% banner do not let you copy a coefficient from another datasheet.

Here is the method with an assumed coefficient, so the numbers are not a Weltrus rating and not a design. Assume the power coefficient is −0.29 percent per degree Celsius, and the cell temperature in operation is 65 degrees Celsius. The rise above 25 degrees Celsius is 40 degrees. The loss is 0.29 percent times 40, which is 11.6 percent. The module then delivers about 88.4 percent of its standard-test nameplate, at 1000 watts per square meter, before soiling, mismatch, and inverter limits. A 560 watt nameplate becomes about 495 watts under those assumptions. Change the coefficient to the one printed on your sheet and the result changes. A coefficient of −0.34 percent per degree Celsius on the same 40 degree rise is 13.6 percent, and the same 560 watt nameplate becomes about 484 watts. That gap is why the bid must quote the coefficient, not the cell nickname.

Nominal operating cell temperature, sometimes shown as NOCT, is a different reference point for a defined wind, ambient temperature, and irradiance. It is not 22.5%, and it is not the temperature coefficient. Use NOCT when the yield model asks for it. Use the coefficient when you are moving power from 25 degrees Celsius to a hotter cell. Voltage has its own coefficient and it sets the cold string length. Do not use the power coefficient for that check.

Bifacial Factor Is Not Energy Gain

Many TOPCon half-cut modules are offered as bifacial, with glass or a transparent rear so the back can collect light. The datasheet then prints a bifacial factor: rear-side power divided by front-side power under a stated test, often with a tolerance band. A factor of 80% means that, in that lab setup, the rear produces about 80% as much power as the front would at the same irradiance on the rear. It does not mean the plant produces 80% more energy, and it does not mean you may multiply 22.5% by 1.80.

Energy gain is a site result. It depends on how much light reaches the rear: ground reflectance, module height, row pitch, rear shading from rails and cables, and how often the rear is dirty. A rooftop with a dark membrane and a module close to the surface can show a small gain. A ground array with a reflective surface and a clear rear can show more. Put the factor into the yield model as an input. Put the gain in only after the site assumptions are written down. If the datasheet does not state bifacial at all, do not invent a factor because the cells are TOPCon.

Front-side efficiency and the bifacial factor answer different questions. The 22.5% line, when it is real, describes the front at standard test conditions. The factor describes the rear relative to the front. Keep them in separate columns. Adding them into one “effective efficiency” on a quote sheet is how two suppliers stop being comparable.

Order for Reading the Datasheet

Read in this order so a high percentage cannot hide a module you cannot install.

1. Identify the model code and the revision. One code, one date or revision letter. Reject a mix of pages from two revisions.
2. Read length, width, and the power bin you will actually receive. Compute module area. If the claim is 22.5%, the watts should be about 225 times that area, within tolerance.
3. Read tolerance and the test condition. Confirm the watts are standard test condition nameplate, not a nominal operating point and not cell efficiency.
4. Read half-cut layout. Cell count, bypass diode count, short-circuit current, and operating current. Check the current against the inverter input.
5. Read both temperature coefficients, power and voltage, plus NOCT if the yield model uses it. Run the hot-power loss with the printed power coefficient.
6. Read the bifacial factor only if the rear construction is bifacial. Keep it out of the front-side efficiency. Record the tolerance on the factor.
7. Read load, mass, clamp zone, maximum system voltage, and the series fuse. Then match the certificate and the warranty to the same model code.

If step two fails, do not continue to price. A percentage that does not survive length times width is not a buying specification.

What This Percentage Does Not Decide

• It does not rank every module in a 100 W to 700 W catalog. Efficiency is per outline and per bin. Choose the wattage class from the site limits first, then read the efficiency of the surviving model.
• It does not prove the cells are TOPCon or half-cut. Those are separate lines: cell technology, cell count, and the drawing. A high percentage can appear on more than one cell type. Confirm the words on the same revision.
• It does not replace the temperature coefficient or the bifacial factor. Hot-cell power and rear-side gain move the energy. The standard-test percentage does not include them.
• It does not set a Weltrus efficiency for modules that are not on the quotation. Use the datasheet we send for the model, the bin, and the revision. Send the target watts, the maximum module length, and whether you need the efficiency checked against area, and we will answer from that sheet rather than from a family ceiling.

Frequently Asked Questions