System Efficacy: Driver Efficiency Meets Luminous Efficacy

The efficiency numbers on a lighting product appear in two separate places, and combining them wrongly produces a misleading picture. Luminous efficacy, in lumens per watt, describes how well the LED converts electrical power into light. Driver efficiency, as a percentage, describes how well the driver converts mains power into DC power for the LED. The system efficacy is the product of the two, and it is the number that actually matters for energy cost.

Luminous efficacy is a property of the LED and its phosphor. A modern white LED achieves 130 to 180 lumens per watt at its rated current, with the exact figure depending on the chip, the colour temperature, and the bin. This is the headline number on the strip or module datasheet, and it is measured at the LED, before the driver's losses are counted.

Driver efficiency is a separate percentage, typically 88 to 93 percent for a well-built constant-voltage driver. It describes the ratio of DC output power to AC input power. The loss, a few percent, becomes heat inside the driver. At 90 percent efficiency, a 100 W driver draws about 111 W from the mains and delivers 100 W to the load.

System efficacy multiplies the two. A strip with 150 lm/W and a driver at 90 percent efficiency delivers 135 system lumens per watt from the wall. If a second driver is only 82 percent efficient, the same strip delivers about 123 lm/W from the wall, a visible difference in operating cost even though the LED's own efficacy is unchanged. The comparison that buyers should make is the system number, not the LED's lumens per watt in isolation.

The interaction becomes more interesting with dimming. As discussed in an earlier article, a driver's efficiency falls at light loads because its control circuits draw a fixed quiescent current. The system efficacy at 20 percent dimming is therefore lower than at full load, both because the driver is less efficient and because the LED's efficacy at a low current may shift. An energy model that uses only the full-load system efficacy will overestimate the savings of a heavily dimmed installation.

The buyer's rule is to compare like with like. Two products should be compared on system efficacy at the dimming level they will actually run, not on the LED's lumens-per-watt headline. A driver that trades a point of peak efficiency for better part-load performance can deliver a higher real-world system efficacy on a dimmed installation. The datasheet's efficiency curve, combined with the strip's efficacy at the operating current, is the pair of numbers that produces an honest energy estimate.

The same logic applies to the mains voltage at the installation. A universal-input driver draws a different input current at 120 V, 230 V, and 277 V, and its efficiency can differ slightly across those inputs, with a small penalty at the extremes. Two drivers that quote the same peak efficiency may diverge by a point or two at the actual mains voltage of the project, and that difference shows up in the system efficacy and the operating cost. The rigorous comparison therefore states the mains voltage, the dimming level, and the load, then reads the system efficacy off both datasheets under those exact conditions. A number quoted without those three conditions is a headline, not an estimate.