Hacker News Reader

Comment by goda90 | Hacker News Reader

The most fair comparison of water use would include evaporated vs soaked into the ground vs run off into rivers vs contained in the final product. As far as I'm aware, data centers are 100% evaporated while farms are going to be some percentage lower than that(though still high). Evaporating extra water can be detrimental to an area if the rainfall isn't high enough, while soaking into the ground is fairly good.

View on HN

A lot of estimates used online for datacenter usage at large scale actually calculate water dependency including power generation - to the point that they might include water used by hydropower plants. It does not mean the reports are wrong, the data is simply for specific use that is then misrepresented by people who didn't read deep enough (or who have an agenda)

As far as I'm aware, data centers are 100% evaporated while farms are going to be some percentage lower than that(though still high).

Newer data centres use closed loop electric (refrigeration) cooling instead of induced draft open loop evaporative cooling.

Do they? That's been a sticking point for many regarding this big push to build more. Evaporative is cheaper and these companies aren't going to spend more if they don't have to.

But they have to, if not spending threatens the continuity of their operations, e.g. by allowing public perception to get so bad that laws get changed due to misinformation about the ecological footprint. Data centers are investments, they take time to build and get operational.

Or they move around to someplace they get less pushback.

What misinformation?

Evaporative is cheaper and these companies aren't going to spend more if they don't have to.

There's a cost either way (electricity or water), but electricity is much easier to get to your data centre so the capex cost at the start is lower, and municipalities have this annoying tendency to ban cooling towers because they're not pretty enough.

Which uses more electricity, at peak cooling times, when wet-bulb would advantage an open evaporative system.

I would imagine plenty of data centers use both? Last time I worked at one they had both industrial heat pumps and evaporative cooling towers outside. While the heat pumps are effective, the cooling towers are way cheaper on cool dry days, so it mostly ended up being which one cost less that day vs strictly one or the other.

Plants evaporate a ton of water. So much so that corn fields create their own weather. A corn stalk is a high surface area, water infused, porous object. It's sending tons of water into the air.

Indeed: "corn sweat" https://extension.osu.edu/about/resources/corn-sweat-and-hum...

why would including evaporation make any difference to the water usage counts?

The water used, whether evaporated or returned into soil (which includes rivers) is unusable any more. Unless the runoff is collected and reused again, which is a form of recycling, and should also not count towards usage of water.

Aquifers maybe? Well water. The issues related to things like the California's central valley subsidence. Not sure data centers are at that scale though.

https://en.wikipedia.org/wiki/Central_Valley_land_subsidence

Indeed. If you evaporate all the water before it soaks into the ground, you end up with a desert

Soil water being unusable is certainly a take. I mean, my drinking straw can't access it, but every damn thing I eat can.

Reply on news.ycombinator.com