Wednesday, 9 December 2009

Osram LCA study

The recently released OSRAM Life Cycle Assessment study (executive summary) appears to be one of the most comprehensive studies so far. However, it still isn't done properly.

The 3 base-cases which the whole study hinges upon have all the usual flaws:
"For comparability reasons in the study, it was assumed that all three lamps would have a light output between 345 to 420 lm during their whole lifetime, and then burn out."
a) Very low lumen lamps were chosen in this study most likely because Osram still can't make an LED produce more light than a the equivalent of a 25W incandescent, which is good for absolutely nothing. A 25W incandescent can be used for mood lighting, an 8W LED can't be used for anything but a scary night light.

b) A 40W standard incandescent bulb gives 410 lumen (xenon-filled bulbs an extra 10%) whereas a typical high quality 7-8W CFL gives around 350 lumen and a 7-8W LED around the same, or less.

c) And then CFLs and LEDs lose output with age - more the longer they last. For example, in a Swedish consumer test (Råd & Rön 3/2008) the 7W Osram Superstar classic A CFLs had lost 27% of their initial output after 6000 hours.
"An extra analysis was done that took the gradual reduction of brightness into account. The difference was too small to impact results, though."
Ah, but that's only because very low lumen lamps were chosen for the study. The higher output lamps you compare, the more CFLs and LEDs tend to lose with age. And as they were already weaker than an equivalent 40W to begin with, this adds to the difference. A 27% decrease is not negligible!

Thus, it is not a correct comparison and all following numbers and conclusions in the study equally faulty. 
"To ensure comparability of the three lamp types a lifetime of 25,000 hours was taken a reference parameter which was evened out by the number of lamps used. This way, the lifetime of 25 incandescent bulbs (25,000 hours) equals the lifetime of 2.5 compact fluorescent lamps, which equals the lifetime of one Parathom LED lamp."
a) Halogen Energy Savers, which last 2000-3000 hours and use 25-45% less energy, were not included!

b) Maybe some of Osram's long-life CFL models last 10 000 hours if just left burning, but as it is explicitly stated that "turn-on-and-off cycles were excluded from the study" and an Osram representative confirmed the earlier study that showed frequent switching may shorten life with up to 85%...

c) We only have Osram's word for their LED lasting 25 000 hours...
"..a correlated color temperature between 2700 - 3000 K (warm white), a colour rendering index of ≥ 80 and a Classic A shape with E27 socket. All lamps provide comparable luminous flux and all are warm white lamps but the fact of a cold perception of the light from different emission spectra of the lamp types is not considered."
In other words: besides not being equal in output, these 3 lamp base-cases are not comparable quality-wise, or even appearance-wise, only in bulb-shape & socket.

a) CRI ≥ 80 = mediocre colour rendering capacity in CFL and LED, to be compared with CRI 99-100 for incandescent and halogen.

b) As Osram does not state exactly which model CFL and LED they have used, it's hard to know what that specific lamp looks like. The ones I've seen so far (see my energy savers review) have not been comparable to incandescent light colour, radiance and quality, even if CFLs have improved markedly over the last decade and Osram makes some of the most incandescent-like.
"The production of the GLS and CFL takes place in Europe. For the Parathom LED lamp, production of the Golden Dragon LEDs is located in Germany (frontend) and Malaysia (backend) and the production of the LED lamp in China. The location of the use phase, end of life, and any other processes was Europe."
This may be true of Osram CFLs, but most CFLs on the market are either low-budget lamps imported from China or manufacturers have their factories there. Sylvania says their ballasts are manufacturered in different places and sent to be assembled into CFLs in another, sometimes another country.
"The heating benefit of a GLS always leads to a discussion. This chapter estimates the actual benefit of heating losses during usage. By assuming 250 heating days, 1000 hours of GLS burning time per year, 75% GLS lighting during heating days, and heating with natural gas, it would lead to a reduction of 17 kg CO2 over 25.000 hours. This saved amount is negligible in comparison to the whole life cycle, and there is no reason to hold on that argument. Furthermore, the heating benefit could also serve as a disadvantage when cooling is necessary."
At least they mention the heat replacement effect, even if they consider it "negligible" - except when it needs to be air-conditioned away, then it suddenly counts (as always).

Poor power factor is not even mentioned.

With correction for the above factors, the outcome would be less advantageous for the LED and CFL.

Yes, they may still save something when compared with the poorest performing incandescents - a matter of quantity vs quality - but not nearly as much as claimed in this biased study, and even less when compared to the best halogen energy savers.
What could be expected? Osram is a lamp manufacturer well known to want to sell more CFLs and LEDs and get rid of the unprofitable incandescent lamps altogether.

Sorry Osram, it was an admirable effort, but do try to get your basic numbers right the next time and the results may be more belieavble, even if not as spectacular.

P.S. When do we get to see the whole study and not just the summary?


  1. yes saw your comment on

    As I said there,

    RE energy used in Osram manufacturing
    There's a lot more to it than that

    1. That's the Osram assembly of already complex parts -
    there's the manufacture of all the parts to take into account too

    2. Transport (from China!), transport to recycling, recycling process, transport of mercury (etc) back to China, as under negotiation,
    for reuse in new CFLs

    3. CFL power factor (not mentioned, as you say)
    halving supposed energy savings in terms of energy needed at the power plant

    4. Lifespan issues
    CFLs lab tested in 3 hour cycles, not corresponding to real life use, on-off switching substantially cutting life
    Also brightness testing not equivalent to practical use,
    heat factor of ordinary bulbs in temperate climates,
    and a lot more from onwards

  2. Right. Thanks for the additional details.

  3. btw did you see:

    hey don't want to ruin your lussekatter and coffee,
    but you know this kind of stuff by now :-)

    - and they keep suggesting halogens as replacement, when they'll be banned too (frosted ones immediately of course)
    haven't read the details, maybe something new there...

  4. I find it rather alarming that they keep ignoring people with serious skin conditions, with the excuse that sufferers are so few.

    But I have to agree with Scenihr on the epilepsy/migraine thing: with electronic ballasts, it cannot possibly be caused by flicker. However, if people with these conditions report experiencing problems with CFLs, it cannot be ruled out that there may be some other, yet-to-be-investigated feature of fluorescent light that is having the negative effect. I think people know best what they experience, even if science fails to explain it.

    The same goes for anecdotal reports from other groups e.g. people with Sensory Processing Disorder, autism spectrum conditions, ADHD, Scotopic Sensitivity etc. of which many report problems with fluorescent light. Just because no one bothers investigating why, or can't think of a plausible cause, it doesn't mean these reported problems don't exist.

    I should blog more about this. Or finish commenting that FAQ.

  5. Yes would be great to be able to link to all your FAQ comments here,
    as also posted on EU environment forum :-)

  6. I've been meaning to, but besides the prohibitive lenght of the FAQ itself, and my @&$#%@ computer constantly playing tricks on me, I now also have the problem that since Blogger 'uppgraded' a while back, formating is even more of a nightmare than it used to be, with arbitrarily dislocation of whole paragraphs and formating buttons not doing what they say they will but just creating a complete mess in html, so making a long post with many quotes is now an even more trying and nerve-wrecking experience, even for my long patience. :-((

    But I'll see what I can do.

  7. This comment has been removed by a blog administrator.

  8. This comment has been removed by a blog administrator.