Guest Post by Thomas Fuller
There are a lot of people concerned about the pace of innovation as it relates to climate change and energy efficiency, because of fears about global warming.
Innovation has led to energy efficiency gains of between 1% and 1.5% for a very long time–perhaps as long as three centuries. For short periods within those three centuries, innovation has been even more robust.
However, every time somebody comes up with a way of saving energy, we end up finding ways to use even more energy with the money we’ve saved. This has become known as Jevon’s Paradox, and it has been discussed by economists since 1865, probably because economists weren’t interested in the invention of barbed wire.
In constructing strategies for defeating the dread global warming, the 1% – 1.5% rate of innovation is ‘baked in’ to adaptation and mitigation strategies. In some scenarios, they assume more. As Roger Pielke Jr. and his friends at the estimable Breakthrough Institute have repeatedly pointed out, it ain’t enough. To make a real difference on global warming, our energy efficiency would need to increase by between 4% and 6%, something that seems close to absurd.
But is it? Let’s talk about a subject dear to the hearts of global warming activists–tipping points. They use it to talk about points of no return for our atmosphere, something more sober scientists think is highly unlikely. But it gets them headlines.
But there are tipping points in technology, as well–witness the striking lack of horse manure on the city streets of New York and London. And the paucity of buggy whips, for that matter.
There are about 16 billion artificial lights in the world today, and about 13 billion of them get replaced every year. CFLs were supposed to change that, but everybody hates them–I think 15 billion of the world’s lightbulbs may well be CFLs stored unused in everybody’s closet.
CFLs could save 75% of the energy used for lighting. But they won’t, because they suck. (That’s a technical phrase meant to cover poor light quality, premature failure, inability to work in many settings and environments–they just suck.)
But LEDs are coming that can save 90% of the energy used for lighting, and they may work better than CFLs. (Anthony, you said you filled your house with them. How do they work?)
Stanley Jevons thought that if we saved 90% of the energy used on lighting, we would find some other use for that energy. And he might well be right. But as with other laws that have passed into obscurity, Jevons did not plan for a future that is almost within our sight, but was 150 years away from him. He couldn’t see a level of saturation that would cause energy use to plateau.
Energy use in the developed world is projected to increase by 0.3% per year through 2050. All of the growth will come in the developing world. But they will develop. They will reach the point where we are today by 2075. And regardless of whether innovation comes in strong or weak, their energy use will plateau, and then decline gently with innovation, stable population and social changes–do you know how much less energy a retired person consumes than someone in the work force? It’s a lot, and the number of retired people is going to skyrocket.
You can leave the lights on. You can buy more lights. But eventually you have enough. You can own three cars. But you can only drive one at a time. And houses will start getting smaller, not bigger, as demographic changes work through the population. And that means that eventually, innovations that improve energy efficiency will reduce energy usage. But, what are we talking about–another century? Another millenium?
How about before mid-century?
Can we achieve step change innovation in all types of energy use? That’s immediately followed by another key question–even if we can, will we?
Those who study energy use break it into several large sectors, with the largest being industrial, which consumes about half of all energy. Transportation accounts for 22%, and residential and commercial fall in between at about 30%. (Technically, the second largest use of energy worldwide is waste during generation and delivery of electricity, something that could be improved on…)
We know step change is possible for transportation. Audi had a car that got 80 miles per gallon on the market a few years ago. The U.S. fleet had an average of 22 mpg a couple years back. Ford is coming out with a model that gets 40 mpg right now. New commercial jet aircraft are at least 20% more fuel efficient than older models.
Half of all new windows sold are energy efficient, and energy efficient windows, doors and insulation could reduce waste by at least 35%. The same is true for new appliances. If we had a cash for clinkers instead of a cash for clunkers… well, you get the idea.
Industry could get a lot more mileage out of the energy it uses. In Denmark, 40% of their primary energy is delivered through combined heat and power at 85% efficiency, compared to the 35% efficiency of old fashioned power plants. In America, we get 9% of our power from CHP. (And how come nobody has thought of using the heat generated by nuclear power plants?)
There is not one thing I’ve talked about above that is not commercially available for sale today. There is not one thing above that would not save money over the long haul for the people who buy it. The average time for technology improvements to spread through a fleet of equipment is between 13 and 25 years. Certainly, if we moved on these available, off the shelf improvements now, they would be in place and reaping benefits before 2050.
People are reluctant to give up perfectly good refrigerators and cars before they are used up. Companies are reluctant to retire coal plants early, and to make capital investments in things like CHP or Waste to Energy without prodding. But we could redirect some of the subsidies we’re giving wind power companies…
Here in America we use 323 million btus per person per year. In Denmark they use 161 million btus per year. (We drive about twice as much as they do, on average, but that’s only a small part of the equation.) We could change that almost painlessly in fairly short order.
We don’t need any new toys to show Stanley Jevons is wrong. We just need to use the tools we have.
James Sexton, sorry–I was off line. I’ll check out that thread and see if I can join you there. I do appreciate what you have to say, even where we may seem to disagree.
My last comment appears to have been eaten 🙁
http://en.wikipedia.org/wiki/Jevons_paradox
Here is an artical that demonstrates the greenies interest in Jevons Paradox. It also contains some counter points.
Thanks again for correcting my misconceptions. I will need to read more about William Jevon. He seems to be very interesting person. It would have been interesting what he would have done with the tools we have today (e.g. Ricci Flow). Then again, his pioneering work probably was responsible for providing their foundation.
Tom Fuller says:
October 12, 2010 at 7:36 pm
James Sexton, sorry–I was off line. I’ll check out that thread and see if I can join you there. I do appreciate what you have to say, even where we may seem to disagree.
========================================================
No worries. Apparently the mods are eating or something. I too appreciate civil discourse.
The morning comes early.
George E. Smith says:
October 12, 2010 at 12:17 pm
“It’s an interesting question as to how efficient it is possible to get in the production of “white” light; and by “white light” we mean light that the human eye perceives as “white”, which is worth keeping in mind since “light” by definition is visible to the human eye.”
It seems to me that the problem here is that we insist on measuring the wrong parameter. The lumen is based on the eye’s sensitivity to light. That is, how bright the light appears. But we don’t want to look at the light; we want the light to illuminate surfaces, which is a very different thing. To see a light, what counts is the intensity of light in the brightest parts of the spectrum; whereas for illumination, what counts is the intensity of light in the dimmest parts of the spectrum. No matter how powerful the light, if it doesn’t have a complete spectrum it will never give enough illumination to compete with an ordinary incandescent. That’s why the subjective illumination from CFLs and LEDs is nothing like as good as the claimed equivalents (even with the better designs of CFL you need at least 40W to match a 100W tungsten bulb, and even then the appearance is noticeably inferior). The maximum possible illumination efficiency is nothing like as big an improvement on incandescents as the ~25 times implied by the ratio of visible brightness (400lumen/W/15lumen/W).
I am slowly working through a renovation and recently installed new lighting in our kitchen using LED downlights. They are GU10 240V fittings, although run from 85-240V, and draw 10W of power. They use the new CREE LEDs. They are to all intents and purposes equaly to the 50W halogen incandesents they are intended to replace. I am very pleased with them. I will be fitting these in the rest of the house as the renovation moves forward.
The CFLs that were in place did indeed ‘suck’!
If you are interested, http://www.cree.com for the LEDs themselves, and I got the lights from http://www.ledcentral.com.au
I have no associations with either of these companies other than a customer!
Tim
Spectral (colour) output sometime is as important as intensity. For human eye (and brain) it is important that colour perception in the artificial indoor lighting is as close as possible to the one in outdoor daylight; reference is overcast northern light at Greenwich, i.e colour temperature 5600K. Colour diagram: http://en.wikipedia.org/wiki/File:PlanckianLocus.png
“”” Paul Birch says:
October 13, 2010 at 3:01 am
George E. Smith says:
October 12, 2010 at 12:17 pm
“It’s an interesting question as to how efficient it is possible to get in the production of “white” light; and by “white light” we mean light that the human eye perceives as “white”, which is worth keeping in mind since “light” by definition is visible to the human eye.” “””
Well Paul I already explained that they call it light because it is visible to the human eye; and that is the only thing that matters.
And since we are talking about “Illumination”, it is rightly presumed that we are talking about Photopic Vision, and not the Scotopic Vision that would be appropriate for the dark adapted eye that was looking for the arrival of a single photon; which the human eye can easily see under those conditions. Well mine can’t; but I can detect a nuclear blast when it happens.
And the Lumen is specifically defined to reflect the behaviort of the average of normal human eyes.
Moreover, the relationships, beween Photometric Luminance, Intensity, Luminous Emittance, and Illuminance, are all precisely defined, so if those parameters are all known, than the real world application results are quite predictable.
LEDs do have some of the same problems that bare filament Incandescent lamps have in that the Luminance of the source is far too high for comfortable viewing by eye; which is why they make frosted incandescent laamps in the fist place; and why incandescent light fixtures typically employ light “shades” to prevent accidental exposure to the high Luminance of the bare filament. Of course those light “shades” absorb and scatter even more light and make them even less efficient than the bare bulb is.
LEDs at least have rather well controlled emitting surfaces; and also well controlled integrated Optics; so the lighting engineer can redistribute the flux in some effective radiation pattern without excessive additional light. The phosphor type of white LEDs in particular have a somewhat larger source area and lower peak Luminance than the bare die; but it is still a small enough sorce area to be efficiently redirected into an efficient illumination pattern.
The LED companies have already discovered that a 1:1 replacement of Incandescent/fluorescent lamps by an LED lamp is not an effective strategy for LED illumination; so new application strategies are being employed. Lighting a whole building interior from 8 to 10 feet off the floor to get say 200 foot Candle or equivalent desk top illuminance; is a totally dumb way to distribute light. You can light floors and passageways at much lower illumination levels; and then apply local area LED lighting, on the desk or other work space for a hell of a lot less total power than by any other means.
When the internal combustion engined automobile was invented they did not try to develop some new powerful buggy whip to cause it to accelerate; they used something more appropriate to the new technology; the accelerator pedal or throttle control.
The issue of the spectral composition of LED or other lighting, is a question of “Colorimetry” and not one of “Photometry”; and yes it certainly is true that in work environments attention has to be paid to the colorimetry of light sources. Standard Colorimetry theory says that any illuminant point on the color triangle space, can be exactly matched by just three monochromatic sources; but that relates only to the visual coclor appearance of the source to the human eye. To the extent that ordinary colored surfaces might not react normally to any such set of three monochromatic sources; the reflected light color of real objects can be tainted by improper lamp colorimetrical design.
Interesting post. I think I “got it” all except the barbed wire allusion. Now I feel really thick; for us dunces — can you drop a hint?
You dance around it, but there is a well-known caveat to Jevon’s paradox: it does not work in the context of inelastic demand. You are essentially arguing we are approaching a point at which demand for energy is rendered inelastic, or that elasticity will fade. I think you’re right, but we must be cautious. The same thing was commonly said about Moore’s Law a couple of decades ago. It was expected to fail because of the so-called Von Neumann limit within the first 3 decades or so, but innovation has outpaced intuition, and the Law has held strong for half a century.
Cautious visionaries today postulate that Moore’s Law might be good for another decade at the most but … who knows? While the Von Neumann limit can be demonstrated to exist in principle, nobody knows its precise order of magnitude. The same is true for the Jevon paradox: in principle demand for power cannot increase beyond bound, but is that bound really in sight? Last I checked nobody has yet fired up the engines on the first interstellar freighter…
George E. Smith says:
October 13, 2010 at 5:26 pm
“Well Paul I already explained that they call it light because it is visible to the human eye; and that is the only thing that matters.”
But it isn’t the only thing that matters; the colour of illuminated surfaces also matters.
“And the Lumen is specifically defined to reflect the behaviort of the average of normal human eyes.”
Yes, the eye, not the surfaces being illuminated. Which makes it the wrong measure for the illumination of surfaces, for many of which the spectral response will be nothing like that of the human eye.
“Moreover, the relationships, beween Photometric Luminance, Intensity, Luminous Emittance, and Illuminance, are all precisely defined, so if those parameters are all known, than the real world application results are quite predictable.”
Well, no, they’re not, because the spectral responses of the surfaces being illuminated are wildly variable.
“Lighting a whole building interior from 8 to 10 feet off the floor to get say 200 foot Candle or equivalent desk top illuminance; is a totally dumb way to distribute light.”
I don’t agree. For most domestic purposes I want a single light, controlled by a single switch. Multiple lights would be enormously more expensive (costing thousands of pounds to install, the more so since we’re no longer allowed to do it ourselves). Personally, I have never liked desk lamps or standard lamps or wall lamps or spotlamps anyway, so I wouldn’t use them even if they were there.
“The issue of the spectral composition of LED or other lighting, is a question of “Colorimetry” and not one of “Photometry””
Sorry, but it isn’t (or only in part). It’s a question of the spectral response of the illuminated surfaces, not the eye.
“Standard Colorimetry theory says that any illuminant point on the color triangle space, can be exactly matched by just three monochromatic sources; but that relates only to the visual coclor appearance of the source to the human eye. To the extent that ordinary colored surfaces might not react normally to any such set of three monochromatic sources; the reflected light color of real objects can be tainted by improper lamp colorimetrical design.”
Precisely. And the absolute requirement for proper general purpose design is that the lamp’s emitted radiation shall cover the entire visible spectrum continuously and smoothly. If it doesn’t, then no matter how “white” the light looks, the colour of (some of the) illuminated surfaces will be wrong. The illumination will be inferior to that of sunlight, daylight, or incandescent light. That is why CFLs and LEDs are and will remain fundamentally inferior (at least until they can be given a genuinely continuous spectrum).
R. Craigen says:
October 13, 2010 at 6:53 pm
“… the Jevon paradox: in principle demand for power cannot increase beyond bound, but is that bound really in sight? Last I checked nobody has yet fired up the engines on the first interstellar freighter…”
On the contrary, in principle the demand for power can increase without limit. Interstellar freighters – or personal relativistic starships – would be only the beginning. How about a fireworks display with real supernovas? Nah, that’s for pansies. Let’s do it with quasars. For now. We can move on to Big Bangs later.
“”” Paul Birch says:
October 14, 2010 at 4:30 am
George E. Smith says:
October 13, 2010 at 5:26 pm
“Well Paul I already explained that they call it light because it is visible to the human eye; and that is the only thing that matters.”
But it isn’t the only thing that matters; the colour of illuminated surfaces also matters. “””
Well the color of illuminated surfaces is once again something that is again determined by the human eye viewing it; and the variability in human eyes is at least as great as the variability in the emission spectrum of even incandescent lamps.
And the human eye is also able to make the color look whatever it wants it to look like within reason.
For example it is a well known; one might say incontrovertible fact that sea water selectively absorbs spectral colors with depth, starting with the absorption of reds first; so as you go deeper in the ocean water, the spectral composition of the available light (from sunlight keeps getting narrowerea nd narrower with depth ending up with just a ble green .
The State fish of Californaia is the Garibaldi; a bright red/Orange fish of generally snapper like shape, and they are all over the place in the favorite soutehrn California scuba diving areas.
And if you dive in those areas, you will find that the Garibaldi loos exactly the same color in two fe4et of water as it does in 100 feet of water. There isn’t any red light to speak of at 100 feet; but the eye knows that the fish is supposed to look red, so that’s the way you see it. You can measure the actual light spectrum down there, and find almost no red light yet the Garibaldi still looks the way it is supposed to. And the human eye red light spectral fall off is very steep, so it will see only a pittance of the red illumination that might remain; yet the eye insists that the fish is still red, and not blue; even thoguh blue light is ablut all that is left.
The Low Pressure Sodium street lamps are another example of the eye laying tricks with us. In that case you only have a sharp pair at 5890, and 5896 Angstroms; and nothing else; yet you can see a lot of colored surfaces illuminated only by sodium yellow street lights. But yes some colors like my red (maroon) Ford looks brown under sodium light.
But you can light up your house anyway you like Paul; that is what freedom is all about; but the availability of newer higher efficiency lighting technologies, is going to change the way the whole wolrd provides illumination.
Many cities in California now have nothing but LED traffic lights; and they still haven’t fully exploited that technology.
When you get rid of the incandescent light bulb, you no longer need the fanct steel box with heat dissipators that incandescents have since you don’t need to be able to open them up in inclement weather to change the light bulb. In fact these cities have eliminated the guys in trucks who drove around changing light bulbs. In the future the LED traffic lights will be cheap molded plastic housings that don’t need on the street service, and cost a small fraction of the fancy metal housings with waterproof seals and the like.
The city of Sunnyvale where I live svaes millions every year in traffic light costs; and that is just the power savings. PG&E gae Sunnyvale a $360,000 bonus, when they completed their LEDification, just because of what it saved the power company in energy distribution and line service costs.
But yes it will evolve over time. Phillips Lighting (Lumileds) owns the building where I work; and they make the “Luxeon” Line of LED power lamps; and they are already selling modular lighting units for the newer more intelligent indoor office lighting designs. And they seem to have figured out how to do all that is necessary in a typical office room with just a single switch.
Things like the internet, and radio, have changed how we obtain our information. The buggy whip newspapers still try to peddle their biassed excuse for information on dead tree; even thoguh it is old hat befopre they even get their product out to the street.
But you can still buy a newspaper if you want to still live in the past; Lighting will experience the same disruptive technology shift.
By the way; for the legal disclaimer; I don’t work on any sort of lighting products; other than the light that you migth see underneath your (optical) mouse if any. Of course if it is a laser mouse you won’t see it because it is infra-red. We do sell LEDs but only as panel indicator lights; and are sepcifically barred fro doing the power LED thing; (besides not having a corporate wish to do so. So I’m not selling anything; just telling what I know about where the industry is going.
But by all means do it whatever way you like Paul; I certainly will.
George E. Smith says:
October 14, 2010 at 3:34 pm
Paul Birch says: But it isn’t the only thing that matters; the colour of illuminated surfaces also matters.
“Well the color of illuminated surfaces is once again something that is again determined by the human eye viewing it”
No, it isn’t. It’s determined by the physics of light absorption and reflection at the illuminated surfaces. Nothing to do with the human eye. The human eye doesn’t get into the picture until later, when it’s too late to change the spectral composition. If a surface’s reflective band is missing from the illumination spectrum, nothing on Earth can put it back and make the colour look right. The human brain does make subjective corrections for the colour temperature of the light, which is why things don’t look too different by sunlight, daylight or incandescant light, but it can only do so for a continuous smooth spectrum (effectively, by adjusting the gain along the curve). It cannot correct for a discrete line spectrum; the data it would require do not exist. The only reason sodium street lights are at all acceptable (and they’re not very, which is why they’re seldom used any more) is that the brain doesn’t expect to be able to use colour vision at night anyway.
“But you can light up your house anyway you like Paul; that is what freedom is all about; ”
It would be, if we had freedom. We don’t. We have people banning the bulbs we like, and bullying us to use inferior forms of lighting that would require the expenditure of huge sums of money in wiring changes.
“Many cities in California now have nothing but LED traffic lights; ”
Traffic lights are not luminaries; as I pointed out from the beginning, making a light visible is very different from illuminating surfaces. For indicators, navigation lights, etc., monochromatic emitters are fine.