How much electricity do spam emails waste?

SPAM emails. Everyone with an e-mail box gets them. They are quite annoying, but that is not all: they waste huge amounts of energy.

How much energy is wasted?
According to anti-virus software specialist McAfee: “More than 80% of the world's email traffic is now spam, and the transmission and receipt of unwanted email gobbles up 33 TWh of electricity a year”, which is the equivalent of the electricity used by 2.7 million US homes, if we take 2016 wiki statistics.

As of 2016, the average annual electricity consumption for a U.S. residential utility customer was 12,077 kilowatt-hours (kWh), an average of 1006 kWh per month.

Considering that many countries in Europe and the world use less than half of that amount, it is obvious that SPAM-wasted electricity could power significantly more homes.
The same amount of electricity could satisfy needs countries like Belarus (9.6 million people), Bulgaria (7.1 million people), Denmark (5.6 million people) or Qatar (2.3 million people).

The above numbers consider email transmission using internet networks, servers running on each node, CPU processing power for SPAM filtering software, taking storage on hard drives, etc.

Additionally, there is the energy used by all people that have email, taking the time to check the junk box for mails that could have wrongly ended up in the junk box, and, at the end, time to empty junk mail, for those that like to do it manually.

Out of 7.5 billion people on the planet, 3.8 billion people use the internet on a regular base, and we can suppose that, on average, everyone has at least one email address.

Also, we could approximate that checking each junk box takes at least 15 sec a day, at least every third day (121.75 days a year).

(((121.75 * 15 sec * 3.8 billion) / 60 sec / 60 min) * 120 watt) = 0.231 TWh, which is enough to power a small town of 50 thousand people.

Seeing that this is a huge issue, what could be a solution?
Maybe we could fix emails to have additional handshakes, so the first email would be just a request for message reception, very similar how chat works now. And, if the handshake is accepted, in the future, the sender will have the ability to send messages while the recipient’s approval is valid. If the user invalidates reception, the sender will get an alert about “unsuccessful message reception,” and transmission nodes will get warnings about this refusal. Too many similar refusals would be a signal to blacklist domain of the SPAM-ing mail server.

The above is a relatively naive solution that does not underestimate the fact that SPAM mails are a difficult problem to solve, but, above everything else, solving it would save us staggering amounts of energy and prevent equally-large emissions of CO2.