What’s the most environmentally friendly way to watch your favorite show? Environmental impact of streaming vs DVDs

Next time you feel guilty bingeing on House of Cards on Netflix or catching up on episodes of The Daily Show on Hulu, take solace in the fact that you’re doing a small part to reduce greenhouse gas emissions.  A new study in Environmental Research Letters  has shown that viewing streaming videos online leads to significantly less greenhouse gas emissions compared to watching DVDs, taking into account factors such as DVD manufacturing, data transmission for streaming, and driving distances for purchasing DVDs.

The study used a methodology known as life cycle assessment to examine the total energy costs and CO2 emissions equivalent (CO2(e)) of both DVD viewing and streaming.  Life cycle assessments measure the environmental impact of a product by measuring the energy used and pollutants generated by every step of its life, from manufacturing to delivery/transmission, use, and disposal.  The method is broken into four major steps shown above to quantify 1) what types of pollutants, etc. one wishes to quantify, 2) taking inventory of all possibles sources and collecting data for each source, 3) quantifying the impact of pollution/energy use from each source, and 4) drawing conclusions from and interpreting the data.  This provides an objective measure of the benefits and costs of different products to ensure that all possible origins of pollution can be determined.  The same methodology is commonly used in energy research, for example, in determining the environmental impact of different types of solar cells or comparing between solar cells, wind, and fossil fuels.

Figure courtesy of [1]

Figure courtesy of [1]

The diagram above illustrates the life cycle used in the study to determine energy cost and CO2(e) created from DVD viewing.  You can see that the assessment begins with the DVD creation (first column), transitions towards transportation and delivery (second, third, and forth columns), and ends with consumer use (last two columns).

The study takes average or median 2011 data regarding DVD manufacturing, use, and replacement from previous literature.  I thought some of these numbers are interesting from a societal standpoint: we (the US) purchase 1.2 billion DVDs per year, equally divided between mail orders and store bought.  About 14 million are mail-rented and 30 million are store-rented.  For determining costs of replacing DVDs, the study assumed rented DVDs have a 3 year lifespan.  Consumer-bought DVDs are assumed to be viewed once per year with a five-year lifetime.   All this data averaged together leaves us with a per capita DVD viewing of five hours per month.  Read into this whatever you wish regarding our society’s obsession with TV, but personally this doesn’t seem too high to me.

The study goes on to make more assumptions from previous literature data regarding distance traveled for mail-purchased and store-bought DVDs and the embodied energy of DVDs, which is the sum of all energy required to create and provide the product.  There are a lot of details, so feel free to look into the article.  The end result, as we’ll see below, is that transport of DVDs ends up being a huge cost and emissions source in their life cycle.  Of particular note, a purchased DVD on average travels 1825 km from manufacturer to store!  That’s a lot of fuel burned!

Figure courtesy of [1]

Figure courtesy of [1]

A similar life cycle diagram for streaming videos reveals the major difference at the core of this comparison.  No transportation costs or emissions come into play here, but rather most energy is used in supporting data severs (first column) and transmission lines for data access (second through fourth column).  The consumer use categories (last two columns) are fairly similar to the DVD case.  (One data point I found interesting – USA citizens spend 3.2 billion hours watching streaming video!)  So the comparison comes down to this.   What costs more or pollutes more – transportation or data transmission?

Figure courtesy of [1]

Figure courtesy of [1]

The head-to-head comparison is shown here.  On the left, we have energy use per viewing hour comparing the various viewing methods along the x-axis.  On the right, we have CO2(e) per viewing hour.  Furthermore, each bar is broken down by color into which parts of the life cycle contribute to energy use or emissions.  A couple points:

1) Energy use and CO2 emissions follow almost exactly the same pattern.  For example, see how the streaming bar on both the left and right have the exact same proportions of green (data transmission cost and emissions), light blue (client devices) and dark blue (device operation)?.  This is mainly because we create most of our energy from fossil fuels, which release CO2(e), so they’ll be highly correlated.  

2) Data transmission energy costs and subsequent emission are a considerable environmental impact from streaming videos – so don’t think that streaming is somehow carbon neutral!  It takes a lot of energy (probably electrical energy, which is generated from coal plants, largely) to keep those data servers up and running.  In fact, streaming has essentially the same impact as mail-ordered DVDs.  It is the transportation costs of store-purchased DVDs (pink color in bar graphs) that lead to the much higher environmental impact, as measured by total energy input on the left and CO2(e) on the right.

3) Most of the energy use in either case comes form the end user devices (computers, tablets, etc.), which are the same for both streaming and DVD viewing.  So streaming does not help in this department at all, and we need to find energy efficient, cleaner ways of producing these devices!

Figure courtesy of [1]

Figure courtesy of [1]

To bring the main message home, here’s a final figure showing the total energy use and emissions (summing up the data per viewing hour from the previous figure over the whole year).  The lefthand bar in each graph is the energy use – 25 Petajoules (PJ) and emissions – 1.3 billion kg of CO2(e) – from streaming in 2011.  The middle graph shows the same energy use and emissions from DVD viewing – 167 PJ and 9.2 billion kg of CO2(e), respectively.  The final bar in each column shows a projected energy use if we moved to entirely streaming videos – this is the interesting data because it indicates what policy promoting this direction would provide in terms of energy conservation and reduced CO2 pollution.  We see that 100% streaming would lead to 162 PJ of energy use ( about 30 PJ less than streaming and DVD viewing combined), and 0.6 billion kg less CO2(e).  This is equivalent to the energy used to power 200,000 households in the USA each year.

So there you have it – streaming is the clear winner, mainly due to the transportation costs of moving manufactured DVDs to store locations for purchase.  One note – the study did not appear to consider disposal of DVDs as part of the life cycle, which it naturally should!  This would further increase the environmental impact (if incineration is used for disposal), but may also provide some energy back, reducing the overall energy costs.  I’m not sure about this, but future work should probably consider it.  Nevertheless, this provides some interesting information for policy decisions about how to incentivize streaming and also finding energy-efficient alternatives to our current manufacturing methods of IPads, laptops, and the like.  In the meantime, happy streaming!



Shehabi, A., Walker, B., & Masanet, E. (2014). The energy and greenhouse-gas implications of internet video streaming in the United States Environmental Research Letters, 9 (5) DOI: 10.1088/1748-9326/9/5/054007

This entry was posted in Article Reviews, Energy, Energy and Society. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s