Today we’re barraged by claims of “efficient lighting” or criticisms of “inefficient lighting”, but what does that actually mean, or what should we actually be concerned about as designers?
In casual terms, we think of “efficient” lighting as using less energy to produce a given amount of light, or as producing more light for a given amount of energy. Technically, the term used to relate visible light produced to overall power consumed is “efficacy”. This is typically expressed as the ratio of visible light to electric power, or lumens per watt. But for practical purposes, efficiency means providing the useful light we seek for as little energy consumption as possible. Useless or wasted light doesn’t count. Or even more importantly, it should mean satisfying our visual needs using as little energy as possible. And that can’t be measured with a light meter.
With today’s emphasis on energy-efficiency, too often evaluating “efficiency” based strictly on light meter readings (or on calculated predicted meter readings) results in visual environments of poor quality. So the key question we need to ask about efficiency is: “efficient at what?” A bare light bulb hanging in your living room could be very efficient at registering on a light meter, but very inefficient at creating a comfortable visual environment.
If we do limit ourselves to what can be measured with a meter, for architectural lighting there are really four components to efficiency:
Lamp efficacy: how much visible light is our lamp (“bulb”) producing for each watt of electricity?
Control gear efficiency: with the exception of incandescent (including halogen), all modern light sources require some electrical components to get the lamp started and to provide the proper operating voltage and current. These ballasts, transformers, and LED drivers consume energy, sometimes a lot of it – they can use 10% or more as much energy as the lamp they serve. So we need to include this energy consumption in the overall lighting efficiency evaluation.
Luminaire efficiency: rarely does all the light from a lamp manage to get out of its light fixture. There are almost always shields, reflectors, lenses, etc. to shape and baffle the light output, and these block some of the light from escaping. Luminaire efficiency can range widely: for a good linear fluorescent indirect-direct pendant it might be over 90%; for a good compact fluorescent downlight it hovers only around 50%. (One of the advantages of LEDs is that, although their efficacy is not particularly high, because LED light output is intrinsically directional, luminaire efficiencies can be higher for direct, controlled beam applications).
Utilization: related to the antiquated “CU”, or coefficient of utilization, this is basically the fraction of light coming out of a luminaire which actually ends up doing something useful – lighting a surface we want to light. A good (or rather, bad) example is the typical dropped-lens cobra-head streetlight. What we want to light is the roadway and maybe the surrounding area or sidewalks. But, as anyone who has ever looked out of an airplane window knows, an awful lot of the light from cobra-heads goes right into the sky. This isn’t useful (in fact the opposite), so it doesn’t count in the “utilization” coefficient.
So we need to multiply all these four factors together to get even a simple numerical evaluation of a lighting system’s “efficiency”. There’s also a fifth, very important factor affecting energy use: controls – the most efficient light can be the one that’s turned off when it’s not needed.
But lastly and very importantly, we need to consider whether a good design can achieve an equally good, or better, visual environment while registering “less” on the light meter. There is no question that this is possible – it happens all the time. A study by the GSA of recently completed federal courtrooms (see link below) found that measured light levels had little to do with actual user satisfaction with the lighting. As another example, an environment with a substantial indirect lighting component can have lower measured light levels while actually providing better visibility and a greater sense of brightness and comfortable seeing. So let’s design for true efficiency: satisfaction per watt.
Photo Credit: SwamiStream