What efficiency means
Efficiency is a ratio: the useful output of a device divided by the input it takes, usually written as a percentage. If a device takes in 1000 watts and delivers 300 watts of useful output, its efficiency is 30%. The rest is lost, mostly as heat.
The idea is simple, but the percentage only means something if you know what counts as input, what counts as output, and under what conditions both were measured. Most misleading claims exploit a gap in one of those three.
Ask: efficiency of what, relative to what?
The same word is used for very different quantities.
- A solar panel efficiency compares the electrical power produced with the sunlight power landing on the panel. A figure around the high teens to low twenties of a percent for common panels does not mean the rest is wasted in the sense of a fault; it is a physical limit of the technology. The sunlight itself is free, so a lower efficiency simply means you need more area.
- A heat pump is rated with a coefficient of performance, such as 3, meaning units of heat delivered per unit of electricity. This is not the same measure as the efficiency of a boiler, and it does not break any physical laws, because the extra heat is moved from outside. Comparing “300%” with “95%” without explanation is misleading either way.
- A battery has a round-trip efficiency: energy out divided by energy in across a full charge and discharge. If charging is 90% efficient and discharging is 90% efficient, the round trip is 0.9 times 0.9, which is 0.81, or 81%.
- A power supply or charger has a conversion efficiency, which may vary with load.
- A motor or vehicle might be rated per distance travelled, which blends many factors.
Percentages from different categories are not directly comparable. A fair comparison needs the same start and end points.
Common ways context goes missing
Peak versus average. A device may reach its best efficiency only at a certain load, temperature or voltage. A quoted maximum can be far from typical use. A power converter might be most efficient at moderate load and noticeably worse at very low load, which matters for always-on equipment.
Laboratory conditions. Standard test conditions are designed to allow comparison, not to predict your results. A panel on a hot roof, a heat pump in freezing weather or an electric car in cold conditions will often perform differently from the datasheet figure.
Component versus system. A highly efficient inverter or panel does not guarantee an efficient system. Wiring losses, shading, mismatched components, standby consumption and poor controls all subtract. This is why seasonal or whole-system figures are more useful than a single component number.
Efficiency versus cost-effectiveness. A more efficient product can be worth less to you if it costs far more or lasts a shorter time. Likewise, a less efficient device might be the sensible choice if it is cheap and rarely used.
Energy lost upstream. Where electricity comes from affects the real picture. Efficiency at the point of use does not include losses in generation or transmission, so “zero emissions at the tailpipe” or “100% efficient at the plug” says little about the whole chain. A simple electric resistance heater really does turn almost all its electricity into heat, but that does not make it cheap to run compared with a heat pump.
“Up to”. Phrases like “up to 40% savings” describe a best case, not a typical one. Ask what the average or median result was and what the baseline was.
Missing baseline. “30% more efficient” than what? A previous model, an industry average, an old device from decades ago? Without a named comparison, the number carries little information.
Percent improvements can be tricky
If efficiency rises from 20% to 22%, that is a gain of 2 percentage points, but a 10% relative increase (2 divided by 20). Marketing may choose whichever sounds bigger. Both are correct; you need to know which is being used.
Also check the arithmetic when claims are chained. If a system has three stages at 95%, 90% and 85%, the overall figure is their product, 0.95 times 0.90 times 0.85, which is about 0.727, or roughly 73%, not an average of the three.
Questions worth asking
- What exactly is the input and the output?
- Under what conditions was it measured, and by whom?
- Is this a peak, an average or a seasonal figure?
- What is the comparison baseline?
- Does the figure include the full system or only one part?
- Is there a recognised standard, test or label behind it, and does it apply in my region?
- What does it mean for my actual energy use and costs?
Check against measurement
The most reliable test is your own data. Home energy monitoring lets you compare what a device draws with what it delivers, and measure before and after a change. Even a plug-in meter can reveal standby draw that a headline figure ignores. Remember that measurement has errors too, especially at low loads, so use repeated readings and be modest about the conclusions.
Reading datasheets carefully, including footnotes and test conditions, is a habit worth building. When the numbers are dramatic and the conditions are absent, caution is the sensible response. This applies equally to news about new battery chemistries, fuel cells and other emerging technologies, where laboratory results often take years to reach real products, if they do.