RealClimate logo

Technical Note: Sorry for the recent unanticipated down-time, we had to perform some necessary updates. Please let us know if you have any problems.

Wired Magazine’s Incoherent Truths

Filed under: — raypierre @ 15 June 2008

Many of our tech-savvy friends — the kind of folks who nurse along the beowulf clusters our climate models run on — are scratching their heads over some cheeky shrieking that recently appeared in a WIRED magazine article on Rethinking What it Means to be Green . Crank up the A/C! Kill the Spotted Owl! Keep the SUV! What’s all that supposed to be about?

Let’s take air conditioning for starters. Basically WIRED took a look at the carbon footprint of New England heating vs. Arizona cooling and jumped to the conclusion that air conditioning was intrinsically more efficient than heating. To see where they were led astray let’s consider a house sitting where you need to cool it by 20 degrees to be comfortable. The heat leaks into the house at a rate that is approximately proportional to this temperature difference, and the heat leaking in needs to be removed. Now, in order to move that heat from inside to outside, energy has to be expended. Given a fixed electric power usage (in watts), a better air conditioner can remove more heat per day than a worse one, but every air conditioner needs to expend some energy to move the heat. That’s just thermodynamics.

Efficiency of air conditioners is measured by a SEER rating, which is the ratio of heat moved to the outside (in BTU/hr) to the electric power consumption (in Watts). A typical modern air conditioner has a SEER rating of 10, We can convert this into nicer units by converting BTU/hr into Watts, which means dividing the SEER rating by 3.413, which then gives us a Coefficient of Performance, in units of Watts of heat moved per Watt of electricity used. For the aforementioned efficiency, we move heat at a rate of 2.92 Watts if we expend 1 Watt of electric energy. An air conditioner is just a heat engine run in reverse: instead of making use of a temperature differential to use heat flow from hot to cold to do work, we expend mechanical work in order to move heat from a colder place to a hotter place. Thus, an efficient heat engine is an inefficient air conditioner. That’s basically why the Coefficient of Performance gets smaller when the temperature difference between indoors and outdoors is greater — with bigger temperature difference heat engine cycles tend to get more efficient, which means that air conditioner cycles tend to get less efficient. That’s also where the “S” in SEER comes from. It stands for “Seasonal,” and reflects the fact that efficiency must be averaged over the range of actual temperature differentials experienced in a “typical” climate. Your mileage may vary.

This situation can be contrasted with heating. If that same house were in an environment that were too cold instead of too warm, so that it had to be kept 20 degrees warmer than the environment, then the amount of heat leaking out of the house each day would be about the same as the amount leaking into the house in the previous case. That heat loss needs to be replaced by burning fuel. Now, generating heat is the only thing that can be done with 100% efficiency. Old furnaces lose a lot of heat up the chimney, but modern sealed-combustion burners– the kind that can use PVC pipes instead of a chimney — lose virtually nothing. With a heat exchanger between the air intake and the exhaust, they could closely approach the ideal. But still, in this case we are generating heat rather than just moving it, so it takes 1 watt of heat power from fuel burning to make up 1 watt of heat loss. That would seem to make heating a factor of 2.92 less efficient than air conditioning.

But wait, the story doesn’t stop there. First, there’s the fact that air conditioning almost invariably runs off of electricity, and the increased electricity demand is a big source of the pressure to build more coal-fired power plants. A house can be heated by burning natural gas, and right there air conditioning becomes 1.8 times worse than heating, because natural gas emits only 55% of the carbon of coal, per unit of heat energy produced. And it gets even worse: Coal fired power plants are only 30% efficient at converting heat into electricity, on average, so there you get another factor of 3.3 in carbon emissions per unit of energy transferred between the house and its environment. Finally, figure in a typical electric line transmission loss of 7% and you get another factor 1.075. Put it all together with the energy efficiency of the air conditioner itself and air conditioning comes in at a whopping 2.19 times less efficient than heating. for a given amount of temperature difference between house and environment. That means that so far as carbon emissions go, heating a house to 70 degrees when the outside temperature is 40 degrees is like cooling the same house to 70 degrees when the outside temperature is 83.7 degrees.

And that’s still not the end of the story. A house in need of air conditioning has other heat inputs besides the heat leaking in from outside, and all that extra heat needs to be gotten rid of as well. For example, heat is a waste-product of all energy use going on in the house. Four people produce 400W that needs to be gotten rid of, and then there’s the heat from hot water, lighting, the TV, cooking and what have you — all the energy usage within the house, plus 100W of biological heat per person needs to be gotten rid of. On top of that, you’ve got direct radiative heating from the sun, both from the sunllight getting through windows and solar heating of the exterior surfaces of the house, some of which will leak in through the insulation. Energy must be expended to remove all this heat. In contrast, in the heating season waste heat is subtracted from the energy needed for home heating.

So, WIRED got the story egregiously wrong, and not just because they did the arithmetic wrong. In their rush to be cute, they didn’t even make a half-baked attempt to do the arithmetic. But what if they had been right and air conditioning really were intrinsically more efficient than heating. Would that justify their conclusion that you can just "crank up the A/C?" without worry? No, of course not, because cranking up the A/C would still use additional energy and still lead to the emission of additional carbon. For the conclusion to be justified, it wouldn’t be enough for A/C to be more efficient than heating; it would have to be so much more efficient that the incremental energy usage from cranking it up were trivial. WIRED didn’t even try to make that case. If they had, they might have spotted their errors.

Is there any real conclusion that could have been drawn from more clear thinking about the heating vs. air conditioning issues danced around in the article? Yes, in fact. The conclusion is that it makes a lot of sense to build houses in places where the environment requires neither much heating nor much cooling. This is in fact why Los Angeles scores pretty well in carbon footprint per capita, despite all the driving (as noted recently in The Economist.). Another conclusion to be drawn from the carbon footprint of New England heating is that there are probably a lot of leaky homes up there heated by inefficient oil-fired furnaces. Fixing that situation represents a huge untapped virtual energy source.

What’s more, for a magazine that purports to be written by and for tech geeks, WIRED missed the biggest and most interesting part of the story: the same intrinsic efficiences of heat pumps can be run in reverse to give you the same economies for home heating as you get for air conditioning. To do this effectively, you’d have to run the heat pump off of natural gas rather than electricity (or perhaps run it off of locally generated solar power or wind). You’d also have to deal with the fact that heat pumps become less efficient when working across large temperature gradients, but that’s where geothermal heat storage systems come in, making use of the fact that the deep subsurface temperature remains near a nice 55F all year around. Now that would have been a nice story for a tech magazine to cover. And by the way, the decrease in efficiency of heat pumps as the temperature differential increases has another implication that WIRED missed: not only does global warming increase the basic demand for air conditioning, with all the attendant pressures on electricity demand, but it exacerbates the situation by decreasing the efficiency of the entire installed base of air conditioners.

Now about that spotted owl. This refers to a claim that industrial tree plantations take up carbon faster than old growth forests; Since spotted owls require the large trees found only in old-growth, the supposed implication is that if we want to soak up carbon we ought to damn the spotted owl and cut down all the old growth. WIRED really committed serial stupidities on this one. First of all, the article they cited in support of their claim was about carbon emissions from Canada’s managed forests, not from old growth. Now, it’s true that a rapidly growing young tree takes carbon out of the atmosphere more rapidly than a mature forest which more slowly transfers carbon to long term storage in soil. However, to figure out how much net carbon sequestration you get out of that young tree once it’s chopped down, you need to figure what happens to it. Lots of trees wind up in paper, carboard boxes, shipping palettes and other things that rapidly sit around decomposing or get burned off (or worse, turn into methane in landfills). Even the part that turns into houses has a relatively short residence time before being oxidized. Anybody who has maintained an old Victorian house knows about the constant battle against rot, and the amount of wood that needs to be replaced even if (knock wood) the thing doesn’t burn down or turn into a tear-down. So, WIRED is totally off the mark there, unless, to use the colorful language of my colleague Dave Archer, they can get trees to "drop diamonds instead of leaves."

Worse, they ignore the abundant literature indicating that old growth forests can be a net sink of carbon even in equilibrium, whereas the soil disturbance of clear cutting and industrial forestry can lead to large soil carbon releases. A classic article in the genre is "Effects on carbon storage of conversion of old-growth forests to young forests" (Harmon et al. Science 1990) . They state "Simulations of carbon storage suggest that conversion of old-growth forests to young fast-growing forests will not decrease atmospheric carbon dioxide (CO2) in general, as has been suggested recently.". For more recent work, take a look at what Leighty et al. (ECOSYSTEMS Volume: 9 Issue: 7 Pages: 1051-1065. 2006 ) have to say about the Tongass:.

  • "The Tongass National Forest (Tongass) is the largest national forest and largest area of old-growth forest in the United States. Spatial geographic information system data for the Tongass were combined with forest inventory data to estimate and map total carbon stock in the Tongass; the result was 2.8 +/- 0.5 Pg C, or 8% of the total carbon in the forests of the conterminous USA and 0.25% of the carbon in global forest vegetation and soils. Cumulative net carbon loss from the Tongass due to management of the forest for the period 1900-95 was estimated at 6.4-17.2 Tg C. Using our spatially explicit data for carbon stock and net flux, we modeled the potential effect of five management regimes on future net carbon flux. Estimates of net carbon flux were sensitive to projections of the rate of carbon accumulation in second-growth forests and to the amount of carbon left in standing biomass after harvest. Projections of net carbon flux in the Tongass range from 0.33 Tg C annual sequestration to 2.3 Tg C annual emission for the period 1995-2095. For the period 1995-2195, net flux estimates range from 0.19 Tg C annual sequestration to 1.6 Tg C annual emission. If all timber harvesting in the Tongass were halted from 1995 to 2095, the economic value of the net carbon sequestered during the 100-year hiatus, assuming $20/Mg C, would be $4 to $7 million/y (1995 US dollars). If a prohibition on logging were extended to 2195, the annual economic value of the carbon sequestered would be largely unaffected ($3 to $6 million/y). The potential annual economic value of carbon sequestration with management maximizing carbon storage in the Tongass is comparable to revenue from annual timber sales historically authorized for the forest."

So, it looks like that old Spotted Owl and its kindred old-growth denizens are in fact sitting not just on a nest, but on a treasure trove of carbon credits worth potentially more than the timber harvest.

And should you keep that SUV? This blurb in fact contains some useful advice, buried amidst some fuzzy reasoning and published over a witless tag line stating that "pound for pound" a Prius takes more energy to manufacture than a Hummer. The apparent implication of that tag line is rebutted in the article itself, but why give the reader that as a 32-point type take-home point when the WIRED editors don’t even themselves believe it’s an important statistic? This factoid refers to the energy used in the nickel component of Prius batteries, but it’s irrelevant because "pound for pound" doesn’t count if your point is moving 4 people from point A to point B. What transport value do you get from transporting four people plus the weight of the Hummer? Now, the rest of the fuzziness in the logic is a bit more subtle. The author notes quite rightly that there is a very significant carbon emission from manufacturing a car, which is indeed more for a Prius (at least for the moment) than it is for comparable sized non-hybrids.. Thus, if you are faced with ditching your existing car (whatever it may be) and buying a Prius, you need to consider how much you drive per year and see how long it takes to "pay back" the carbon emission from manufacturing the Prius. So far so good. But this is more a statement about the transition to more efficient cars, and how to deal with mistakes of the past, rather than a statement about what is intrinsically desirable in the fleet. As far as carbon emissions go, we’d still be better off if everybody who needed a car were in a Prius, except maybe for people who drive very little per year — who should then be into shared hybrids via iGO or ZipCars, Maybe if you drive very little and live out in a rural area where there are not going to be any shared cars, getting a compact non-Hybrid might make sense. There must be at least a dozen or two people out there in that category, I guess.

The rest of the advice WIRED gives makes even less sense. They say that if you want to be green, you ought to buy a used Civic or something like that, not a Prius. That’s because the used car already has the manufacturing carbon emissions "written down" (or, I guess at least the carbon guilt accrues to the original owner, not that the atmospheric radiative forcing is going to care much about that). However, this advice, sensible-sounding though it is — ignores the fact that to make that used car available to you, the original owner almost certainly had to buy something else, and probably that was a new car, or at least a newer one. So, for the scheme to work, you’d have to buy your used Civic from somebody who was giving up driving altogether. I no longer own a car myself, but I’m sorry I wasn’t able to participate in a scheme like this; by the time I gave up our remaining car ten years ago, it was suitable only for the crusher, and in fact had to be towed there.

The real implication is that manufacturing costs count, so most people should buy a small, efficient hybrid and keep it until it runs into the ground. The implication is also that durability of cars counts for nearly as much as gas mileage, since an efficient car that needs to be replaced every five years isn’t really all that efficient.

Along with all the nonsense is a certain amount of true (if by now commonplace) advice. Among this is the basic truth that urban living is inherently green, and if more people lived in cities (and if more cities were kept livable so people would want to move there). then per capita carbon emissions would go down. Even there, the Economist managed to be both more informative and more iconoclastic with its surprising analysis of the pattern of urbanism in Los Angeles. The other truism in WIRED is that nuclear power deserves a second look, and probably has an important role to play in a decarbonized energy future. Still, if you compare the cost of making all those chilly New England homes efficient with the total true cost of building more nuclear plants, well, let’s just say I’m buying stock in argon-filled low-e window manufacturers rather than Areva, much as I like their track record on nuclear electricity.

367 Responses to “Wired Magazine’s Incoherent Truths”

  1. 351
    Nick Gotts says:

    David B. Benson [344] I suggest that increasingly science is being done with less human conceptual understanding and slowly, increased computer program understanding.

    On the contrary, when used properly, increased computing capacity is increasing human conceptual understanding in science. Although I left AI ten years ago now, I’d still be pretty confident in saying no computer program is yet anywhere near understanding anything in the way a human being does. Human conceptual understanding of science is built on multiple layers of both innate (i.e. evolved) and culturally and individually learned understanding of the everyday world, and ability to manipulate it. There is interesting work in “e-science”, with which I’m somewhat involved – techniques for making the raw data-storage and computational power of computers more accessible and useful to scientists (in my case, specifically social and environmental scientists).

  2. 352
    Rod B says:

    Martin, O.K. You just have a far more loose definition of “cherry picking” than even my light definition…

  3. 353
    Chuck Booth says:

    Re # 344 David B. Benson “consider the possibility that the computer programs will determine what questions are interesting.”

    Interesting to whom, or what?

    If you mean interesting to humans, that implies an underlying theory to which the questions are relevant. If you mean interesting to computers, I would say, Who cares? What an intelligent computer thinks is important is of no relevance to me. Unless, of course, the computer has been programmed to identify questions that humans find interesting, in which case you again are dealing with an underlying theory.

  4. 354
    David B. Benson says:

    I seemed to have stirred up quite the honet’s nest via comment #344. I don’t find RealClimate to be an appropriate site to continue these observations/speculations. Those still interested might wish to learn a bit about bioinformatics, semi-automated medical diagnosis and other areas in which, increasingly, ‘the computers just do it’.

  5. 355
    Hank Roberts says:

    Rod, the reason for “… recording more data … at all the stations …”) is that we’re attempting to get the big picture without having the satellite sitting where the big picture is measured quite simply.

    We know the radiation balance for _other_ planets because we’ve backed far enough away from them to get the whole disk in the viewfinder and take one picture.

    Answer? Triana — languishing in its warehouse.

    Without Triana sitting out far enough away to image the whole visible side of the planet, continuously, the climatologists are putting together a patchwork of data. Satellite imagery at varying altitudes, at varying angles through the atmosphere, at intervals, with satellites wearing out, falling down, and being replaced by different instruments or none at all.

    All because we’re _too_close_ to measure the whole planet at one time, continuously.

    The government knows why Triana is still waiting. Don’t ask. Google search provides these interesting links, among much else, from the first page of hits:
    “… In May 2008, a Freedom of Information Act request relating to DSCOVR was rejected by the White House Office of Science and Technology Policy [2]…”

    NASA IT Security Warning Banner
    U.S. GOVERNMENT COMPUTER If not authorized to access this system, disconnect now. ..

  6. 356
    Nick Gotts says:

    Re 354 [David Benson]

    David, I worked (admittedly some time ago) on “semi-automatic medical diagnosis”. Computers are useful tools here, but they are not doing science, nor do they understand what they are doing, nor would any responsible clinician rely on them for a final diagnosis. While I was working in this area, I was interviewing a cardiologist. I asked him “When does your diagnostic problem-solving begin?”. He answered something like: “When the patient walks through the door. I see their sex and apparent age. I see how they walk, how they hold themselves, how they sit, their skin tone, their confidence or diffidence, whether they look anxious or depressed.” This type of perceptual expertise could be combined with reasoning about time, space and causation (the bits I was most interested in) which was far more complex and subtle than anything that can yet be automated.

  7. 357
    David B. Benson says:

    Nick Gotts (356) — I agree. Nonetheless, the progress this century so far in automated Bayesian reasoning, evolutionary computation and multi-agent sytems is rather impressive.

    More will surely come.

  8. 358
    Nick Gotts says:

    David Benson,
    I agree! It’s a question of timescale.

  9. 359
    Ray Ladbury says:

    Nick and Dave, I’ve been unable to find the original von Neumann quote, but it is something to the effect that many consider that machines are incapable of performing certain undefinable tasks that humans can. Of course, if they could ever actually define what these tasks are, we could build a machine to perform them. Perhaps the strength of human intelligence is that it can perform indefinable tasks. Witness the difficulty we have even defining the scientific method exhaustively.
    In a larger sense, though, science is inherently a human activity. Quantum mechanics showed us that the world does not break down neatly into human concepts or perceptions, and yet we still have to define our science in terms of them. I think that if you divorce the human element from science, it ceases to be science. It may be something else quite worthwhile, and we may be able to learn from it, but it will be alien to our intelligence. I have to ask what good “science” will be if it does not increase human understanding.

  10. 360
    David B. Benson says:

    Ray Ladbury (359) — That is one perspective. I’m willing to debate it, but not here. Meet me at Tamino’s Open Thread?

  11. 361
    Steven F. Scharff says:

    After reading the posts here, I’ve taken the plunge and canceled my subscription to Wired. It was becoming too much like a techie version of Playboy anyway. Thanks for helping me make up my mind.

  12. 362
    Brian Dodge says:

    RE #341
    That sounds to me like the approach taken in Multivariate Statistical Process Control in Industrial Processes, which, since the advent of computers and digitally networked controls, is being used to do things like produce plastic film of a more uniform thickness. Basically,(and that is the extent of my knowledge, VERY basic), the computer program looks at a large amount of production line data (temperatures, pressures, PID loop coefficients, flow & feed rates, power consumption, relative humidity, anything that is already being measured and/or controlled, and sees how these correlate with say, product thickness. The software may tell you that as meas. A goes up, thickness goes up after T1 minutes; as control B is increased, thickness goes down after T2(

  13. 363
    Frank Hagan says:

    I think you engage in some questionable math, at least on the heating vs. a/c portion of the debate. In the NE, the example used in the WIRED article, 38% of the homes are heated by fuel oil in boilers that are, at best, 84% efficient (and most of them are far less than that, with older boilers being 76% efficient or less).

    Even in areas where natural gas is available, consumers do not get the advertised 96% – 98% efficient operation from the modern condensing appliances if they have in-floor radiant or radiator systems (the most common methods). The efficiency you achieve is determined by the return water temperature of the system in a condensing boiler, and at a return water temp of 140°F … common for those systems … your very expensive condensing boiler is about 85% efficient, perhaps a bit better than a newer oil boiler, but obviously not the figure cited in your article. The ASHRAE manual has the chart on the efficiency, which declines as the return water temperature rises, and while I cannot buy you a subscription to their service, the chart is reproduced at the bottom of .

    The other factor you have ignored is that it is very easy, and relatively inexpensive, to upgrade an a/c compressor to a more efficient model as they become available, giving those with a/c an easier upgrade path. In contrast, replacing your 84% efficient oil boiler with a high efficiency natural gas condensing boiler … if you have access to natural gas … can cost up to $10,000. It is unlikely to happen very often at those prices, which is one reason why the dirty, smelly, inefficient fuel oil continues to be used. At a real-world gain of only a few percentage points in efficiency, the upgrade cost has a payback that stretches into decades.

    Finally, you further stack the deck by using coal as the source for energy generation. Here in California, only 3% of our electricity is generated using coal, and more environmentally friendly natural gas and nuclear take a bigger role. And those oil boilers? They are illegal here, due to our superior air quality laws.

  14. 364

    Your argument about heating vs cooling is bull. Fortunately, so is Wireds. :)

    You assume straight off that every time you turn on an airconditioner, you use coal to make the electricity. That doesn’t cut it. And you say that houses can be heated with natural gas. Well, they can in theory, but in practice, most people do not have gas pipes outside their house. So in fact, most heating will be done either with an oil heater, or, surprise, electricity. Which again will be just as coal fired as the coolers. :-) So, no, your counter argument is wrong.

    But Wired is also wrong. Yes, moving heat is more efficient than generating heat. So cooling is more efficient than heting by generating heat. But both Wired and you have missed out on a very simple fact: You can heat the house by moving energy as well! Basically, you take an air conditioner, but stick the cool bit outside and the hot bit inside, and hey presto! you got heating that is just as efficient as cooling.

    These heat exchanger systems exist, and work well down to well below freezing, and will heat your house on what may very well be clean electricity, and do it cheaply as well.

  15. 365
    Bo says:

    Wired,Vogue, and the Newyorker…as well as many other magazines are owned by the same company. Some years ago I happened upon an article in the Newyorker talking about how dangerous SUV’s where and how they would be so much safer ( for those hit by one ) if they were unibody or unitized in construction( read no frame or composed of overlapping sheet metal like a cardboard box). The funniest part was that the author claimed that the reason the auto manufactures continued to make the trucks with full frames was out of pure greediness. According to the author, the SUV manufactures keep making their trucks with full frames because building them untilizing a unitized construction would be more expensive.

    Problem with this of course is that the claim is total B.S. In the early to mid sixties most american auto manufactures had already changed over to a unibody construction. And the reason that all of our cars are made this way today is because a sheetmetal frame/body utilizes less materials, is easier to make,are lighter, and cost less. Walter P. Chrysler himself claimed this very reason for swithching over to unibody construction. Cars today are made using unitized construction for the same reason boxes for moving are made of cardboard and not 2×4′s and wood.

  16. 366
    Hank Roberts says:

    Bo, you should check what you believe with the Google.
    You’re remembering (or rather mis-remembering) this article, from the 2004 New Yorker:

    Check what he wrote in 2004 in the Google.
    Current search:

    There you’ll find, for example, that this year this change is news, and considered newsworthy:

    Surge of the Unibody Truck – Car News
    Four years after the Honda Ridgeline made a startling debut, others are wading into car-based-pickup territory.

    BY ALISA PRIDDLE March 2008

    When the Honda Ridgeline burst onto the automotive scene as a unibody compact pickup truck concept at the 2004 Detroit auto show, it turned conventional thinking on its ear. Up to that point, pickups were, by definition, body-on-frame workhorses for lugging gear and towing. The fact that cabs continued to grow in size and amenities and the trucks were increasingly being used as family cars was irrelevant when it came to the frame.

    Then Honda did the unthinkable. It engineered a pickup on a car platform, with a clever in-bed trunk. And the industry watched ….
    ——end excerpt——-
    … watched, and waited. In 2008, some possible change.

    Gladwell was entirely correct. Your memory or your source got it wrong.

    Hey, I don’t trust my own recollection or things I read on opinion blogs either. Lesson is — check.

  17. 367
    Hank Roberts says:

    Bo, see also:
    GM May Break Up SUV-Truck Marriage to Cut Fuel Use, Emissions

    By Jeff Green

    May 8 (2008) (Bloomberg) — General Motors Corp. may be forced to break up a seven-decade marriage of pickups and large sport- utility vehicles as Americans restrict the fossil-fuel diet of their transportation.

    Under pressure to produce a more fuel-efficient and cleaner- running line of vehicles, GM is investigating ways to design a lighter replacement for its biggest SUVs, such as the Chevrolet Tahoe, without relying on a heavier pickup-truck frame, according to people familiar with the effort.

    The Tahoe and its predecessors have shared the design of the Chevy Silverado pickups since the model was introduced in 1965. While no decision has been made, GM engineers are considering a shift in 2012 to a car-like construction for successors to the Tahoe and other large SUVs, including the GMC Yukon and Cadillac Escalade….

    —-end excerpt—–

Switch to our mobile site