Streaming is an energy hog and causes a lot of CO2. I have often thought about this topic, but today I would like to write down my thoughts in this post and provide food for thought to find a solution.
As it is, you can choose which study to trust, because studies are made by people, make calculation errors, others may only have the understanding of CO2, but not the deeper understanding of IT to add up all the points correctly.
I remain neutral here and will not call any of the studies less relevant or wrong, but will use for calculations, if necessary, data from the International Energy Agency (IEA) use
Everything that needs energy, causes as of 2023, unfortunately also CO2, because we are technically not able to cover streaming completely by renewable energies, even if we only had nuclear power to support the renewable ones at night, every data center that is connected to the power grid receives the power mix, that is, every power plant that is operated with fossil energies has a share in the produced kWh. That means, in every produced kWh you will still find a few grams of CO2.
The problem here is now that we would have to concentrate with the Streaming not only on the server, but quasi also the way up to the client, over Proxyserver, Edge server, Caching etc. all this has an energy consumption and since these are internationally connected, however in each country another electricity mix receive, the 1. server can be operated indeed for example by Google, with regenerativ energies and cause no CO2, however as soon as one has then for example, in Germany a server for forwarding, then this uses the electricity mix from Germany for forwarding, this CO2 portion results additionally.
In this screenshot I've done a route trace of where everywhere a request is routed to YouTube, of course this is related to Google's international data centers and Edge servers, but shows where everywhere a request is routed to YouTube.
It can be seen that a request from the Czech Republic goes through several countries and each country has a different CO2 footprint, which is added to the CO2 footprint:
The attempt shows YouTube, of course streaming services like Netflix or Amazon Prime Video would make more sense, but unfortunately the route tracking did not work there. For this reason, I leave it with the YouTube example.
The current status
There are several organizations that produce studies, I have chosen here the International Energy Agency, because they have the know-how around the international energy production and therefore, in my opinion, can best calculate the amount of CO2.
Update 11/12/2020: The energy intensity figures for data centers and data transmission networks were updated to reflect more recent data and research. As a result, the central IEA estimate for one hour of streaming video in 2019 is now 36gCO2The updated charts and comparisons also include the corrected values published by The Shift Project in June 2020, as well as other recent estimates quoted by the media.
According to the IEA, it is estimated that per hour of video streaming in 2019, you get 36g of CO2. Even if other popular media speak of other figures, their articles are based on a study by The Shift Project, but these have had a bit/byte conversion error and started a momentous mistake, because 1 byte corresponds to 8 bits, or 8 bits correspond to 1 byte and that led to the calculation of 8 times higher Co2 imprint has assumed, one of 1.6kg Co2 for 30 minutes of streaming, after the correction comes to 0.2kg Co2 per 30 minutes of streaming in 2018.
A number of recent media articles, including in the New York Post, CBC, Yahoo, DW, Gizmodo, Phys.org and BigThink, have repeated a claim that "the emissions generated by watching 30 minutes of Netflix [1.6 kg of CO2] is the same as driving almost 4 miles."
The figures come from a July 2019 report by the Shift Projecta French thinktank, on the "unsustainable and growing impact" of online video. The report said streaming was responsible for more than 300m tonnes of CO2 (MtCO2) in 2018, equivalent to emissions from France. The Shift Project published a follow-up article in June 2020 to correct a bit/byte conversion error, revising the original "1.6kg per half hour" quote downwards by 8-fold to 0.2kg per half hour.
0.2kg ~ 200g CO2 for 30 minutes of streaming is still quite a lot, but through modernization and better and more efficient technologies, such as better processors, more efficient storage to store more data per storage module, be it HDD, SSD, or even NVMe, this also reduces the power consumption. So that one is now at 36g CO2 per 1 hour of streaming.
Unfortunately, you can't rest on this small value, because the consumption of streaming services is increasing, Netflix alone has had a large influx of users in the US during the Covid quarantine/lockdown:
Netflix surpassed YouTube as the most-watched service in 2018, with an average viewing time of 23.2 minutes daily.
In 2019, the platform's users spent two hours per day on average watching Netflix. On top of that, the corona lockdowns the following year caused a 61% increase in streaming. Netflix viewership statistics show that the average user streamed around 3.2 hours per day during the quarantine. This equals 203,840,000 hours per day watched if we take into account all users. And these are solely statistics for US users.
203,840,000 hours, 203 million 840 thousand hours. This number of hours were consumed in 2019 by US Netflix users only, PER DAY!
203,840,000 hours x 36g CO2 per streaming hour = 7,338,240 kg ~ 7.3 million kg CO2 ~ 7338.24 tons CO2 per day
As you can see, even with this small CO2 footprint per hour of streaming, it adds up to a hefty sum of CO2 if many people use it.
Can streaming scale in this way?
Let's say we have 8 billion people, and each person consumes streaming services, now let's say in the example that each of those 8 billion people watches a movie for 1 hour a day.
8,000,000,000 hours x 36g CO2 (0.036kg) per streaming hour = 288 million Kg CO2 per day
Okay, that's pretty high, let's calculate the same more optimistically and say that we use more renewable energies and nuclear power so that enough electricity is stable at night and the use of fossil energies decreases. Let's optimistically assume 5g CO2 per streaming hour.
8,000,000 hours x 5g CO2 (0.005kg) per streaming hour = 40 million Kg CO2 per day (At 1gCO2 per hour per day = 8 million Kg CO2 per day)
You see, even if we were to drastically reduce the Co2 footprint, streaming will make it impossible to not cause CO2 and thus still impact the international CO2 footprint on a daily basis and that is a problem.
Streaming abolish I consider unrealistic, and whether worldwide each person daily a stream of 1 hour, is also not realistic, however, that balances out again, if individual people watch several hours a day streams, but shows the problem, because consumption will not stop and a ban or abolition will not work.
I see the possibility here that these prices will be increased to such an extent that only the better-off will be able to afford streaming, which will reduce the number of users.
The problem is, this assumption is also unrealistic, because people will then find illegal ways to get movies and will accordingly look for and find illegal sites for consumption. Already today, the number of illegal sites is increasing because people are not willing to pay for more than one provider on the whole, for streaming movies or series.
And it is also unrealistic to believe that you will be able to take all of these sites offline.
Downloads as a solution?
One solution to solve this problem would be the download of video content, the advantage is that the computing power and processing on the server of the provider would be omitted, since virtually only a data download would take place, and the resource consumption of electricity is then omitted on the client side, be it the desktop computer, smartphone or SmartTV.
The servers would only be used to provide the downloads and a simple web interface with an index would make the downloads findable. However, the problem is that many people only watch a film/series once, so you would need more local storage space to store these downloads for a later time if necessary.
Monetization in the download model
One problem you have, you pay a subscription today to get quasi access to thousands of streams, and the more you pay, the more quality etc. you get. Models like Amazon Prime, where you already have a subscription, and you still have to pay extra for individual movies, or Google Play Movies & Series to pay for the individual watch, is not very attractive for me personally, for one, there is no pure download of the file and the purchase is assigned to a user account, you can not watch it offline, at least that was my last stand (have not bought anything for years at Google Play Movies & Series).
One idea would be to keep the subscription model, but the user can only watch the download if a token/password is stored in the system, which contacts the provider via the Internet and the provider checks the validity and the device. In this way, one could at least add the new Netflix account when changing the computer and then play the download in the offline player.
But you could also create pirate copies here, as is the case with streams, just activate a screen recorder and record the video 🤷♂️🤷♂️🤷♂️
BitTorrent - more than just pirated copies
Even in my younger years I was fascinated by BitTorrent, as this technology is actually the closest thing to the Internet as a decentralized system than what we call the Internet today.
BitTorrent is a protocol for efficiently sharing large files over the Internet. It was developed by Bram Cohen in 2001 and has since established itself as a popular peer-to-peer protocol.
The BitTorrent protocol is based on the concept of peer-to-peer connection, where every person who downloads a file also shares the file with other people at the same time. This makes it possible to download faster, since the load is distributed among several people and not just on a single server.
A torrent file is a small file that contains information about the file to be shared and the tracker. The tracker is a server that coordinates the connection between peers and stores information about the status of the download.
- User A wants to download a movie, but the file is very large, so he decides to use BitTorrent to speed up the download process.
- A downloads the corresponding torrent file, which contains information about the movie and the tracker.
- A opens the torrent file in a BitTorrent client program, which communicates with the tracker and searches for peers who share the file.
- The client finds several peers from which it can download small pieces of the file. A starts downloading different pieces from different peers, which speeds up the download process.
- Once a piece of the file has been downloaded, A can make that piece available to other peers who have not yet completely downloaded the file.
- The client periodically checks if more pieces of the file are available and automatically downloads them to speed up the download process.
- When the download is complete, A can watch the movie and at the same time has also helped other users to download the file faster.
This example illustrates the advantages of BitTorrent for downloading large files, especially when many users share the file, allowing for faster transfers.
What's the point?
According to the study Consume Local: Towards Carbon Free Content Delivery (PDF), the energy demand is reduced
Abstract - P2P sharing among consumers has been proposed as a way to decrease load on Content Delivery Networks. This paper develops an analytical model that shows an additional benefit of sharing content locally: Selecting close by peers to
share content from leads to shorter paths compared to traditionalCDNs, decreasing the overall carbon footprint of the system.
Using data from a month-long trace over 3 million monthly users in London accessing TV shows online, we show that local sharing can result in a decrease of 24-48% in the system-wide carbon footprint of online video streaming, despite various obstacle factors that can restrict swarm sizes. We confirm the robustness of the savings by using realistic energy parameters drawn from two widely used settings. We also show that if the energy savings of the CDN servers are transferred as carbon credits to the end users, over 70% of users can become carbon positive, i.e., are able to support their content consumption without incurring any carbon footprint, and are able to offset their other carbon consumption. We suggest carbon credit transfers from CDNs to end users as a novel way to incentivise participation in peerassisted content delivery.
It is therefore possible to reduce the CO2 footprint between 24-48% through P2P sharing technologies such as BitTorrent, but the control instance to monetize a film at the individual customer is omitted, or one would have to tie the playability to a user account and the respective shared device, as with the download.
I wonder if the film and music industry would get involved with the arch-enemy BitTorrent for the sake of climate protection? I would rather bet that I would become a millionaire tomorrow than that I would be willing to make this bet....
Back to the Roots - the DVD as a solution?
Streaming is not good, downloads will probably not be used for that and BitTorrent will not even be touched by Hollywood. So what solution is left? The Digital Video Disc / Digital versatile Disc (depending on who you ask and interpret it). Yes, the DVD or its modern counterpart, the Blu-ray Disc would be the better solution.
The DVD has a long life of 10-100 years, depending on storage, can store between 4.7-8.5 gigabytes.
Blu-ray disc has a life span of 30-50 years, can store 50-128 gigabytes, and in the laboratory even 500 gigabytes.
Would make more sense if you can have multiple movies burned to one Blu-ray disc, saves resources and power when you want to watch them again.
The film Avatar - Departure to Pandora* from 2009 is available on Blu-Ray Disc, for an average of 7-12,99€. This has it also as DVD* but not in the 4,7 GB but in the 8,5 GB DVD version, of which 7.2 GB were used for the film.
Now, if you do the math and stream this movie, you have to expect the following quality levels:
HD 720p video at 4,000kbps (4Mbps)
4,000kbps => 4Mbps => 0.4768 MB/second * 3600 = 1.676 GB/hour
HD 1080p video at 8,000kbps (8Mbps)
8,000kbps => 8Mbps => 0.9537 MB/second * 3600 = 3.353 GB/hour
4K Ultra HD at 14,000kbps (14Mbps)
14,000kbps => 14Mbps => 1.6689 MB/second * 3600 = 5.867 GB/hour
The film runs 178 minutes in the extended cut version, based on the dates of dollarpeak.com, the following results:
HD 720p -> 178 minutes x 60 = 10,680 seconds x 0.4768 MB = 5.09 gigabytes
HD 1080p -> 178 minutes x 60 = 10,680 seconds x 0.9537 MB = 10.185 gigabytes
4K Ultra HD -> 178 minutes x 60 = 10,680 seconds x 1.6689 MB = 17.823 gigabytes
You can see that with each higher resolution, the data consumption for streaming doubles or more than triples compared to 720p to 4K.
However, the IEA has also presented corresponding figures for bit rate and resolution:
The original "1.6kg per half hour" claim overestimated bitrate, the amount of data transferred each second during streaming, apparently assuming a figure of 24 megabits per second (Mbps), equivalent to 10.8 gigabytes (GB) per hour. This was six times higher than the global average bitrate for Netflix in 2019 (around 4.1 Mbps or 1.9 GB/hr, excluding cellular networks) and more than triple the transfer rate of high-definition (HD, 3 GB/hr). Other typical transfer rates are 7 GB/hr for ultra-high definition (UHD/4K), 0.7 GB/hr for standard definition (SD) and 0.25 GB/hr for mobile.
With 3GB per hour and Avatar going almost 3 hours, we're already approaching 10GB in HD 1080p when streaming.
The High Resolution on Netflix in standard definition is logically 720p, which consumes 1GB per hour. High resolution in HD up to 3GB would be 1080p HD and 4K up to 7GB.
If 30 minutes produces 0.2kg of CO2, then 1 minute of streaming produces 0.2kg/30 minutes / 30 minutes = 0.006kg ~ 6g CO2/minute.
6g CO2 x 178 minutes = 1.068 ~ 1.1kg CO2
That means, if the numbers are correct, that for a movie like Avatar - Departure to Pandora, which goes 178 minutes, you would cause around 1kg of CO2. If you use a different way of calculating, namely CO2 per gigabyte transferredwould result in 0.49 kg of CO2.
You probably won't get more exact data, but it's enough to calculate in theory.
Let's assume you want to watch such a movie once a year, then you would do that over 40 times in your life (assuming you are 20 and watch the same movie until 60), then you would cause 44 kg of CO2 in this time by watching this movie (of course there would also be technical development and better compression codecs etc.).
According to an article from 2009 Carbon Footprint of a DVD(sadly, I have not been able to find anything more recent), they talk about 1.02 Lbs per DVD, Lbs means pounds, which would be the equivalent of 0.48kg of CO2 per DVD.
That means, if you compare the DVD in production and distribution with 0.48kg CO2 to streaming, streaming causes 0.49 - 1 Kg CO2 per call.
That means, if I base it on this data, the DVD is worth it, even before you stream, and depending on how you look at it, every time you watch the DVD, you save twice the amount of CO2 that you would otherwise cause by streaming.
To be fair, here is the link to the article Comparison between digital and physical delivery
According to Treehugger.com, it also clearly depends on what you now include, what quality you watch, but also how often and especially how the procurement looks like, if you now drive extra to buy a DVD, the carbon footprint is greater than if you order it by mail.
The German-Czechoslovak Christmas classic Three Hazelnuts for Cinderella (1973)*, you actually look at every year, and is with 915 Mb, due to the low resolution at the time, even so small that a 4.7Gb DVD, in addition to this film, could be written with 3-4 more from that time. As of March 2023, you can still get the film at the Watch MediathekWebView from ARD Mediathek, and there you can see the 915 Mb
Conclusion: What does that mean now?
I myself approached this article with an open mind, simply because I was interested in the topic and what solutions could be implemented, from the data, the DVD is the winner, but why was the DVD sent to the sidelines? Perhaps one could also have given the EcoDisc / EcoDisc can give a chance, which 52% produces less CO2.
Man is probably to blame here, because convenience has a price and it is now times more convenient everywhere and mobile to watch a movie or series in parallel, if someone blocks the TV and no one would voluntarily want to give that up to look out again DVD's on the shelf and always change by hand.
If blockbuster movie rentals had been made more modern and efficient, maybe they would have had a chance, because renting DVDs/Blurays is more ecological and saves more resources than would be needed for production if everyone wants to have a movie in their closet at home.
Maybe you should think about it again, at least if you are serious about climate protection, because the consumption of streaming content will continue to increase, and accordingly the power demand and you will not be able to further optimize the server side. And even if some sites write that streaming is now better, that may well be, it depends on the context of the consideration. I do not lease the truth and think you can certainly improve the accuracy of the calculations.
Perhaps an alternative future would have been to order a movie online for the evening and have it delivered by bicycle. Who knows how the world could develop differently. At least I do not think that streaming will continue in the long term, the resource of electricity is too precious, at least the EU with the new Eco Design Directive which came into force on March 1, 2023 already times the size of TV sets artificially restrictedThe more inches, the more electricity is needed to provide brightness, so televisions that fall into energy efficiency class G will be banned. But whether this is the right way or requires companies to invest more, we'll see.
Note/Disclaimer: This article contains affiliate links, these are marked with an asterisk *, when purchasing I receive a commission, this has no effect on the purchase price. I am not paid extra to put these links.