Re: [recipe] FYI: [admin-discuss] REMINDER: IETF carbon emission measurement workshops on 20 and 21 September

[Michael Welzl <[e-mail]>]

Thanks again for this pointer - will read soon!

Indeed…  Jonathan Koomey and his colleagues also did interesting work on “cleaning up the mess” … e.g., this:
https://www.cell.com/joule/fulltext/S2542-4351(21)00211-7?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS2542435121002117%3Fshowall%3Dtrue#relatedArticles <https://www.cell.com/joule/fulltext/S2542-4351(21)00211-7?_returnURL=https://linkinghub.elsevier.com/retrieve/pii/S2542435121002117?showall=true#relatedArticles>
and, in particular, this paper is *really* interesting:  https://onlinelibrary.wiley.com/doi/10.1111/jiec.12630

It would indeed be good to write up (with some kind of IETF blessing, e.g. as an RFC coming out an RG) what’s likely an exaggeration, and what’s probably at least roughly right.

Cheers,
Michael


> On Sep 24, 2022, at 4:16 PM, Hesham ElBakoury <[e-mail]> wrote:
> 
> I agree with you! We should clean up the mess.
> 
> Another example of inaccurate myths is related to data center energy consumption which Masanet addresses in his papers. Please refer to this article: The Myth of Data Center Inefficiency - Compass Datacenters Podcast <https://www.compassdatacenters.com/compass-u/podcasts/podcast-eric-masanet/>
> Hesham
> 
> On 9/24/2022 6:41 AM, Michael Welzl wrote:
>> Hi,
>> 
>> Thanks for the pointers!  These are quite interesting indeed (I knew about the second, but not the first).
>> Regarding inaccurate estimates, I also came across quite a number of them…  trying to clean up this mess (and calling out these false claims) would be a good work item for the envisioned group here.
>> 
>> Cheers,
>> Michael
>> 
>> 
>>> On Sep 23, 2022, at 9:51 AM, Hesham ElBakoury <[e-mail] <mailto:[e-mail]>> wrote:
>>> 
>>> Hi Mike,
>>> 
>>> Ericsson did some studies regarding carbon emission of ICT sector and aviation. As an example, you can refer to A quick guide to your digital carbon footprint - Ericsson <https://www.ericsson.com/en/reports-and-papers/industrylab/reports/a-quick-guide-to-your-digital-carbon-footprint> which dispels some of the myths regarding carbon emission for streaming and aviation. Other studies from Ericsson and Telia reveal that the claims of the electricity consumption of video streaming and data downloading over mobile and fixed networks are inaccurate
>>> and out of range – often by orders of magnitude.
>>> 
>>> Daniel Migault has a paper on "CO2eq: Estimating Meetings’ Air Flight CO2 Equivalent Emissions An Illustrative Example with IETF meetings".  The paper is available here: https://www.iab.org/wp-content/IAB-uploads/2021/11/Migault.pdf <https://www.iab.org/wp-content/IAB-uploads/2021/11/Migault.pdf>. 
>>> 
>>> Thanks
>>> 
>>> Hesham
>>> 
>>> On 9/21/2022 1:26 AM, Michael Welzl wrote:
>>>> 
>>>> 
>>>>> On 20 Sept 2022, at 21:45, Toerless Eckert <[e-mail] <mailto:[e-mail]>> wrote:
>>>>> 
>>>>> On Tue, Sep 20, 2022 at 06:12:51PM +0000, Cedric Westphal wrote:
>>>>>> Interesting that for a typical meeting, the airplane leg comprises 99% of CO2 emissions. I had no idea it was so much.
>>>>> 
>>>>> Yes, i wonder about that too. I think this is based on the CO2 emission multiplied by a factor
>>>>> for having the emission in the stratosphere, and that factor i think does not
>>>>> necessarily have a widely agreed upon value. But i am not quite sure. I can also
>>>>> not imagine that the whole hotel energy consumption does not factor into the CO2
>>>>> emissions significantly, especially not when its in a hot US american city
>>>>> and still freezing the attendees.
>>>> 
>>>> Difficult to say, about the building vs. flying thing….   but generally, regarding the CO2 cost of flying and what this group is trying to do, there are some interesting relationships:
>>>> 
>>>> * if the numbers above are roughly right, then we’re better off meeting online than flying from the perspective of CO2 emissions from meetings  (btw I’m not sure CO2 is even the right term here: I believe I read somewhere that, while planes do emit a significant amount of harmful greenhouse gases, different from most other things, these are not necessarily CO2). I believe in this; this paper also suggests that meeting online is better, and surely the CO2 emissions of the Internet are worth it:  https://dl.acm.org/doi/10.1145/2070562.2070571 <https://dl.acm.org/doi/10.1145/2070562.2070571>
>>>> 
>>>> * then again, considering what this group wants to do, let’s consider the GHG emissions of the Internet. I think there’s reason to believe that these are in the ballpark of half the aviation industry… this is super rough, as it’s obviously hard to get good numbers on it!  but for my rationale, see the introduction of this paper:  https://folk.universitetetioslo.no/michawe/research/publications/wons2022_authors_version.pdf <https://folk.universitetetioslo.no/michawe/research/publications/wons2022_authors_version.pdf>    it would be good to have a more in-depth analysis on where we really stand!  E.g., the HotNets paper I mentioned above is quite old… one thing the envisioned group should do, IMO, is to start a collection of material on estimates of Internet GHG emissions, perhaps in a wiki.
>>>> 
>>>> * If we really manage to pull of something that reduces Internet energy, even only a tiny bit, at a global scale…  then one can start thinking about what that would mean in terms of plane trips, in relation to the above…
>>>> 
>>>> =>   I’d say: if *anyone* should be allowed to fly for the sake of more constructive meetings, then it’s probably us   :-)
>>>> 
>>>> Cheers,
>>>> Michael
>>>> 
>>>> PS: this is all really quite tiny (but nonetheless important to reduce!) compared to other things:   https://ourworldindata.org/emissions-by-sector <https://ourworldindata.org/emissions-by-sector>
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> recipe mailing list
>>>> [e-mail] <mailto:[e-mail]>
>>>> https://www.ietf.org/mailman/listinfo/recipe <https://www.ietf.org/mailman/listinfo/recipe>
>>> _______________________________________________
>>> recipe mailing list
>>> [e-mail] <mailto:[e-mail]>
>>> https://www.ietf.org/mailman/listinfo/recipe <https://www.ietf.org/mailman/listinfo/recipe>
>>