Maybe it meant the rate of energy consumption while the servers are rendering and sending the page. You'd have to multiply 30w with the time spent by the server doing all this stuff (probably less than a second) to get the total consumed energy. A web server consuming 30w of power per connected user sounds exaggerated to me. A high-end server CPU at full load probably pulls 200w and can handle at least 1000 users.
But this is an art project and I have no idea what expertise does the artist have, and I'm not an expert either.
Yes, watt hours or joules would work there but not watts. In general the figures all seem to assume 30 watts of power consumption at near-idle, which is completely ridiculous unless you happen to be using either a high end workstation or an antique.
The first page states an estimate of 50g of CO2 per visit, but I'm pretty sure reality is nowhere near that. At ~1000lbs CO2 per MWh for home electricity in the US (https://www.epa.gov/energy/greenhouse-gases-equivalencies-ca...), that's ~0.45g per Wh. My laptop only has a ~40 Wh battery, so I can't even be using anywhere near that - and that's already two orders of magnitude smaller.
Scrolling down the project page about half way, it's displaying 0.15 kcal for me. That's ~0.17 Wh, which would equate to only being able to do that ~235 times on a single charge for my laptop. That's very obviously nowhere near reality.
It's supposed to be "sciency", not science. Look to scientists for meaningful
energy to accounting. This a political art project, where accurate facts are not important.
So now it says 30wh, which seems just wrong. If the transaction took 10 seconds or so, during that short time it would have had to consume over 10 kilowatts of energy.
Neither journalists nor USians are very careful with their use of units. Imagine my dismay to find a Nordic person (who are usually so careful, precise and diligent!) forgetting to make sure she (?) make sure that power and energy were each separately accounted for.
The good news is that in 4 billion years, the sun will explode, and we'll have to teach unit-accounting to sluggards in the dark. (Wait. That's the good news?)
What is that supposed to mean? A watt(= 1 J/s) is a measure power so energy per unit of time. Shouldn't this rather be given in joule?