Saturday, July 14, 2012

A Machine for Light

When I was growing up in Southern California in the 1960s, my mother took me to see an exhibition of the work of the Bauhaus at the Norton Simon Museum in Pasadena. The work was strange and vibrant, by turns playful, formal, and ironic. Along with other experiences, like the films and furniture of Charles and Ray Eames, ideas from the Bauhaus made a big impression on me and must have lodged in my subconscious. Southern California at the time was ripe with experimentation- the predominant aesthetic model was modernism, expressed in the work of architects like Schindler, Neutra, and Eichler, summed up in Corbusier’s phrase “a house is a machine for living.”

Modernism never quite lost its hold on the Western American imagination, possibly because of its insistence in pure, geometrical, non representational forms and focus on the centrality of human endeavors as the primary way to create meaning. In the US, and California in particular, this was consonant with the self made, pioneer spirit, and found a voice in writers like Ayn Rand. The archetypal modernist shelter was the iconic expression of “prospect and refuge,” and nothing epitomized this aesthetic better than the above photograph by Julius Shulman of the Stahl House by Pierre Konig. Soaring over the sparkling grid of the Los Angeles basin at night, the picture of ease and elegance is cemented in my imagination by two elegantly dressed women who look precisely, exactly like my mother did in that era, down to the hairstyles and shoes. I can even smell her perfume! One of the enduring effects of images like these was to inculcate architects with the love of glazing. Sometimes it seems that they all decided, as I am fond of saying, that if it’s worth doing it’s worth overdoing. The aim wasn’t necessarily what we think of as “green” today, as in daylighting, it was more like they did it because they could, and it looked fabulous in the magazines. Still the effect was to suddenly flood homes and offices with daylight, for better or for worse (often both at once).

Lately I’ve been thinking a lot about the fundamental function of shelter through a completely different perspective, I’m beginning to understand the building not only as a strategy to stay warm (or cool) and dry, but also as a “machine for light:” a way of delivering appropriate light in appropriate, controllable levels. Architects must have understood this basic function intuitively ever since architecture has been around, but the relatively recent invention of electric manmade light ( lighting designer Jim Benya holds that all light is natural, it’s just that some of it is manmade) has had perhaps a more profound impact on building than we realize. Since manmade lighting is undergoing its next disruptive upheaval with the adoption of LEDs, we’re really beginning to look at light in a completely different way: we have an unique opportunity to reengage with it. So reimagining lighting naturally leads to reimagining architecture, yet again.

Because lighting technology is undergoing a massive disruptive change, we’re examining the technical, energy, design, and health aspects of lighting in great detail. We’re only beginning to understand the primacy of light, how it impacts circadian rhythms, cognition, mood, and behavior. If you examine lighting only, or initially, from a standpoint of energy efficiency, you realize that the impact here is on a very large scale, for instance, LEDs (in optimized systems) can eliminate 80% of lighting energy use across a wide range of technologies. Since lighting is perhaps the single biggest use of electricity in non-industrial buildings the potential scale effects of this level of efficiency are huge. Lighting is the most ubiquitous and visible use of power - lighting and electricity are inseparable. The history of electrification is the history of lighting- people’s desire for artificial light was the driving force behind the building of the electrical grid. With the emergence of whole building design practices, lighting has the potential to lead the design process, not only from an efficiency standpoint, but from a standpoint of total environmental quality.

But problems with current lighting design practice and technology are legion. For starters, it’s easy to believe that most buildings are not designed by architects (an assumption worth questioning- see this Certainly, of all buildings actually designed by architects, relatively few of them have lighting designed by lighting designers. So a very small percentage of all buildings have lighting designed by dedicated professionals. This appears to be borne out by a quick look at the built environment- most lighting really sucks. It’s glaring, too dim, too bright, poorly colored, wildly inefficient, and sometimes close to intolerable. (It’s hard to imagine any living organism surviving for long under those orange sodium vapor streetlamps). Most lighting is so bad, and we experience really good lighting in the built environment so rarely that we don’t really know what we’re missing. Fortunately, this is about to change.

Part of the reason for widespread poor quality of lighting has to do with the economics of building design practice. I know from personal experience that when building professionals put together project teams and budgets, special consultants at the end of the food chain (this unfortunately may include lighting designers) are likely to be eliminated - as architects see their profits drain away when budgets get chopped, they figure they can do the lighting themselves. Or electrical engineers do it, as they understand wiring and power use, but not necessarily good lighting design. Also, on big complicated projects, having fewer consultants to manage can be a very good thing for project managers. Electrical contractors or distributors may provide lighting design for free in order to get on projects. None of these practices necessarily incentivizes good lighting design.

The current state of design of lamps and luminaires exacerbates the problem somewhat. In view of the potential technological advances on the horizon, today’s technologies can seem somewhat primitive. In many key ways, lighting has not evolved much since the first Edison lamp. Often the cheapest, most ubiquitous lamps and luminaires are tragically inefficient or really ugly or both. There are significant color and spectral issues with many, if not most, classes of lamps. Performance varies considerably, and often falls well short of requirements across many technologies in terms of consistent lumen output, lifetime, starting, temperature, and other measures. The design of many lamps and luminaires involves what I call “provisional coping strategies”- often elaborate and inefficient schemes to overcome fundamental design flaws. To give two examples: omnidirectional lamps are routinely baffled, filtered, shaded, and otherwise diluted so that only a fraction of the light they produce actually gets delivered to us as usable light; and halogen MR16s use dichroic coating on their reflectors to filter out harmful IR and UV. Another key aspect of inefficiency is that many incandescent lamps (like the MR16) typically produce light outside of the visible spectrum, or light with highly unbalanced spectral signatures.

Of course, despite the many shortcomings of lamps and luminaires, the best lighting designers work wonders with current technology, and better design is certainly not all about new technology. This is very much the case with the HVAC industry, which also hasn’t really seen innovative new technical developments in a long time, and better design practices are perhaps needed there more than new technology. The lighting industry, however, is very different, as the LED revolution is now in full swing, and the big changes are based on a fundamental shift in technology.

There are really 2 classes of LEDs- what I call 1.0 and 2.0 (first generation and next generation). Many LED 1.0 products are just as bad, or worse than the technologies they are meant to replace in terms of light quality and performance, and LED 2.0 products are much better in those areas. While  LED 2.0 products are mostly still rather expensive compared to other lighting technologies, they’re going through what we might call a rapid maturation process, and because they’re based on semiconductors, we will see significant advancements in performance and a steady drop in price in the next few years, may be sooner. Although it won’t be anywhere near as dramatic as price/performance curves seen in semiconductors used in computers, forces are aligning to put the manufacturing technology and infrastructure into place to allow global scale production of affordable, high quality LEDS in the near future. We may not know what hit us until it’s well underway. LEDs are already utterly disrupting lighting, and the questions facing designers now are how to design with them, and how to make architecture more responsive to high quality lighting - how to expand the services delivered by the “machine for living” to include better light than we ever dreamed possible.

We need to see past the temporary high cost of this new technology, which will change very soon, and get our minds around what may be counterintuitive for most of us – the fact that LED 2.0 products can offer everything at once: energy efficiency, controllability, longevity, and quality of light. We’re used to too many tradeoffs and shortcomings in new technologies, so it’s kind of hard for us to believe that we can get it all: if ten or fifteen years ago someone were to describe to you today’s iPhone and all the functions it provides at its current cost, you couldn’t possibly imagine it, yet it doesn’t compromise on anything important, there are no real tradeoffs in quality, it’s all there. The same is becoming true with LED lighting (with the exception of high “blue spike” LEDs based on non-native substrates, which have had rapid and widespread market penetration). I’ll outline some of the ways I see this happening.

For starters, a lot of LED technology can be “plug and play,” that is, it can fit into the existing lighting infrastructure and work immediately, delivering better light at significant energy savings . This is not true of other classes of building services, like HVAC. Highly inefficient installed infrastructure is a pervasive problem today, as capital for financing building projects is still hard to come by. It’s much easier to build a really energy efficient new building than it is to retrofit an older one with deeply embedded inefficient HVAC equipment, a low performing building skin, too much glazing, and bad siting. Even if retrofitting is preferred to “greenfield” development, the overall costs of extensive HVAC retrofits are too high for most owners to bear. Lighting retrofits on the other hand are usually much easier to do. This is not as simple as it sounds, however- even though LEDs in optimized systems can use something like 70-80% less energy, the average building’s electrical infrastructure, like most systems, is overbuilt for the older incandescent or fluorescent technology, and overbuilt in general beyond that. Eventually, electrical equipment designed specifically for LED technology will work its way into the market and the infrastructure.

Next, LED technology allows for exquisite control of all the elements we care about in lighting: color, temperature, direction, starting, dimming, size of source, long life, and stable output. The efficiencies in LEDs come from many factors: they produce only visible light, meaning that energy is not wasted in delivering wavelengths that we don’t need and can in fact be harmful; they produce considerably less heat than other light sources, which means more energy goes into producing light; GaN on GaN LEDs (definitely 2.0) can now behave essentially like “point sources,” very bright and very small, so that beams can be controlled and directed much better, wasting much less light in spill and other non- beam areas; what is essentially digital light is much more responsive to many different control parameters, including dimming, temperature, occupancy sensing, and intelligent energy management. Since the spectral signatures of LEDS are also now much more controllable, we will be able to design sources specifically with better light for human health, friendly to our circadian rhythms, eyeballs, brains, and souls.

Smaller, more powerful, more controllable sources will aid in the decentralization of lighting, improving efficiency and light quality. Point source performance allows for much smaller, lighter lamps and luminaires, meaning that lighting can be considerably more flexible, less intrusive, and require a lot fewer accessories to control light – this aspect of LED 2.0 technology can have an amazing scale effect, where all kinds of things will change dramatically, starting with one little dot the size and color of the yolk of a hummingbird’s egg. Smaller sources mean smaller, lighter lamps, smaller fixtures, less material used for construction, more ceiling space, more usable space per floor, more floors per building, more building per construction dollar, smaller heat load from lighting, smaller HVAC systems, less energy use per building, fewer power plants, and on and on.

Since we’ve had very little time on an evolutionary scale to adapt to manmade light, we don’t yet know much about how it affects our genetic structure and physiology. We still envision daylight as the ultimate light source to replicate in manmade light, but daylight itself varies widely depending on time of day, direction, and a host of other factors. Sunlight contains lots of harmful UV radiation that we need protection from, hence architecture. We’re as intimately connected to our shelters as are bees to their hives, and clearly shelter provides the key way in which we manage our light intake. Ancient forms of architecture devised many ways to manage light without manmade electric sources: in modernism, suddenly many of these practices were forgotten, ignored or overridden, especially with the advent of overglazing. Part of making better, more sustainable buildings involves the rediscovery of older design practices for managing the delivery of light and other services. Decentralizing lighting and controls will help designers to effect better daylight balance in buildings and provide higher quality manmade light where it is most effective.

New, dramatically better lighting technology, the increased awareness of the primacy of light, and more intelligent design practices will redefine architecture, making it more adaptive to climate, energy use, shifting patterns of settlement, and human needs for health, wellness and security. Better buildings are essential for better light, and better light is essential to life.