Narration script

Unit ten. Outdoors, light level, spectrum and cell temperature change all the time. To compare products fairly, the industry fixed standard test conditions: one thousand watts per square metre, the A M one point five global spectrum, and a cell temperature of twenty-five degrees Celsius. Real cells run hotter than that, and hotter silicon produces less voltage, roughly zero point three to zero point four percent less power for every degree. Two more ratings, called NOCT and NMOT, describe more realistic operating temperatures.

You will be able to state the standard test conditions and why each was chosen, explain how current and voltage respond to light level, estimate cell temperature and the resulting power loss, and tell NOCT and NMOT apart.

Every module leaving a factory is flash tested at standard conditions and labelled with that power. Warranty, binning and price all follow from it. And temperature coefficients are a key selling point of newer cell technologies over older ones.

Here is the table of standard test conditions. One thousand watts per square metre, because it is a round number close to clear noon sunlight on a well-tilted panel. The A M one point five global spectrum, because it represents mid-latitude conditions. And twenty-five degrees cell temperature, because it is easy to hold in a lab. In factories, flash testers fire a pulse of light lasting only milliseconds, so the cell does not have time to heat up during the measurement.

How does a cell respond to light level? Two rules of thumb. Current is proportional to irradiance: twice the photons, twice the current. But voltage grows only logarithmically. Each halving of light costs only about eighteen millivolts per cell at room temperature.

These curves show both effects, from a standard cell model with illustrative parameters. On the left, I V curves from one thousand down to two hundred watts per square metre. The current, the height of each curve, scales down with light. But the voltage, where each curve crosses zero, barely moves. On the right, the same cell from zero to seventy-five degrees. Now the current hardly changes, but the voltage shrinks steadily as the cell gets hotter.

At two hundred watts per square metre, the current drops to twenty percent of its standard value, but the voltage drops by only about forty-one millivolts, around six percent. So power falls to a little under twenty percent.

Why does heat hurt? Higher temperature increases the cell's dark current exponentially, so open-circuit voltage falls, by about two point two millivolts per degree for a silicon cell. The band gap shrinks slightly, so the current rises a tiny bit, but the net effect is less power. Datasheets give three temperature coefficients: alpha for current, slightly positive; beta for voltage, negative; and gamma for power, also negative.

Here are typical power coefficients, estimated from datasheet ranges. PERC, around minus zero point three five percent per degree. TOPCon, around minus zero point three. Heterojunction, around minus zero point two five. Higher-voltage cells are less temperature sensitive.

How hot do cells actually get? NOCT, the nominal operating cell temperature, is measured on an open-circuited module under eight hundred watts per square metre, twenty degree air and a one metre per second breeze. Datasheets typically list values in the low to mid forties. A simple model then says: the cell runs above the air by an amount proportional to the irradiance. NMOT uses the same conditions, but with the module actually producing power at its maximum power point. Since some energy leaves as electricity, NMOT is a little lower than NOCT.

This graph uses that simple model at twenty-five degree air. At one thousand watts per square metre, cells run roughly twenty-five to thirty-five degrees above the air. On a hot summer day, that means cells at sixty degrees or more.

Let's work it through. A five hundred and eighty watt TOPCon module, gamma of minus zero point three percent per degree, NOCT forty-four. Thirty degree air and full sun. The cell temperature comes out at sixty degrees, thirty-five above standard. The power loss is thirty-five times zero point three, ten and a half percent. So the module makes about five hundred and nineteen watts, before any other losses.

On the same day, a PERC module at minus zero point three five loses about twelve percent, and a heterojunction module at minus zero point two five loses about nine. For equal nameplate ratings, the heterojunction module delivers about four percent more power in that heat.

Irradiance varies on every timescale: seconds, as clouds pass, sometimes briefly exceeding one thousand watts when light reflects off cloud edges; hours, as the Sun moves; and seasons and years. Because standard conditions rarely happen outdoors, other standards measure power across a whole matrix of irradiances and temperatures.

Misconceptions. Panels do not work best on hot sunny days; cool, clear, bright days are best. Twenty-five degrees means cell temperature, not air temperature. Low light does not kill voltage; it falls only logarithmically. And NOCT is not the same as NMOT.

In the factory, flash testers hold cells at twenty-five degrees and set irradiance with reference cells. Modules are binned by their measured power, current and voltage. And the temperature coefficient is a technology differentiator that process engineers influence through passivation quality and voltage.

Try the review questions. The final unit of SOL 101 previews what sunlight actually does inside a semiconductor, and draws our first map of where all the energy goes.