Google opens up on seven years of its data center history | Hacker News Reader

Google opens up on seven years of its data center history

gigaom.com

77 pointsaphtab37 comments

From the linked post [1], 1/4 the power to run pumps instead of fans is a major difference!

Was curious if using local water sources would have any environmental impact from the heat or flow, and found an article talking about the first seawater-cooled center, from only May 2011 [2]. Google apparently creates a 30-year thermal model of the effect on the environment. I hope other corporations take their consideration and transparency as an example.

[1] http://www.greenm3.com/gdcblog/2012/11/13/google-adopts-wate...

[2] http://gigaom.com/cleantech/google-to-switch-on-worlds-first...

I wonder how much global warming gets prevented just through everyone adopting these practices for datacenters (either directly or by switching to cloud providers with an economic incentive to do so).

Maybe a carbon tax wouldn't be so unreasonable, if it shifts more innovation like this. There's probably a "Google of Cement Production", agriculture (arguably this is Monsanto or ADM or Cargill, as much as they are reviled over IP, they are quite efficient), etc.

There already is an energy tax. You have to pay for energy.

A carbon tax on top of that wouldn't effect efficiency.

The argument is that the cost of extracting the energy and transforming it into a usable source does not match the cost it does to the environment once that energy is used. That's what the carbon tax is for. It's for the carbon (pollutants) you produce not for the energy you use. Hence why it's called a carbon tax and not an energy tax.

Exactly. By capturing externalities correctly you'd encourage use of inconveniently located less-polluting energy sources in some cases, in preference to easily transported (or conveniently located) more polluting fuels.

(or, my favorite, easily transported, non polluting nuclear...)

It's not so simple. Until the EROI of other fuels improves dramatically a carbon dioxide tax is just going to make every fuel more expensive, since non-carbon fuels need carbon fuels as input.

Once the EROI goes up this isn't a problem. But until it does a carbon dioxide tax is pointless - it just makes everything more expensive.

And once the EROI does go up people will switch to those energy sources without prompting (they will simply be cheaper), so again a carbon dioxide tax does nothing.

A correctly-structured carbon tax would make a new NG plant relatively more attractive vs. continuing to operate a coal plant without the tax. (both fossil fuels, but NG is more efficient).

Also I think it shouldn't be a straight carbon tax, but an "environmental externalities tax". I personally care far more about other forms of pollution (especially localized) than CO2; I don't think global warming is a non-issue, but it's not the #1 environmental issue. Point sources of particulates, NO2, etc. should be taxed as well as maximum emissions specified (obviously a steel mill should be allowed to emit more, and pay for it, at full production in an efficient way, than a badly adjusted oil-fired school furnace, even if the school chooses to ignore economic rationality.)

Agreed!

And interestingly coal plants are already being shuttered due to the low price of natural gas. All that without any taxes - just new technology.

If you make all energy more expensive, people will use less of it than they would otherwise. That's already a win. You're also creating a pain point that people will innovate around, and improving the ROI on investing in better energy sources.

The fact is, once alternative energy sources are at par, or even slightly above par, with fossil fuels, it will be a net win to switch to them. But the economy won't do so unless the price of fossil fuels reflects that.

> people will use less of it than they would otherwise

The trouble is that the economy is directly tied to energy usage. Force lower energy usage, and the economy suffers.

The cost of virtually everything is directly tied to the cost of energy. What costs money in lettuce? Mainly the trucks used to plant, harvest, and ship it.

What's your cost with trucks? Energy. What costs money there? The metal, and the assembly. The metal is free - it's in the ground. What costs money is the energy used to get it out of the ground and refine it.

Assembly? That's more machines, trace it back and it boils down to the cost of energy.

Salary? Also energy, since people need money to buy stuff.

What costs money in making alternative energy? Energy again. So raise the cost of conventional fuel, and alternative also costs more. It's impossible to win!

Use the carbon dioxide tax to fund alternative energy and people will fake it, like in Germany. They outsource the carbon dioxide costs to China, then ship that energy to the country in the form of solar panels. Those solar panels will not pay back their energy costs for decades. (Although if they do last that long it will be a win. But in the medium term it's a loss.)

You just can't force this stuff. You have to let the energy sources develop naturally, and let people use them due to price, not because you force them.

http://en.wikipedia.org/wiki/Pigovian_tax

A fair number of libertarians actually believe in this (modulo implementation issues, regulatory capture, corrupt politicians, etc.)

> You have to let the energy sources develop naturally, and let people use them due to price

That's exactly what a carbon tax does, except it allows the prices to reflect the externalities of CO2 emissions, rather than not doing so.

People will find ways of doing the same thing with less energy. What happened when oil prices went up? SUV's went out of fashion.

The most important feature of a carbon tax is that it start small and get progressively larger over the next thirty to fifty years.

Carbon tax isn't about fairness or justice. It's about making investors feel safe allocating long term dollars into alternative energy r&d over a timeline long enough to allow a real transition. A gradually increasing carbon tax is the simplest and surest way for the government to catalyze innovation.

Bloomberg[1] says a current discussion in Washington is for a $20/ton carbon tax increasing by 6% a year. That's $33/ton in 10 years, $60/ton in 20, and $350/ton in 50 years.

[1] http://www.bloomberg.com/news/2012-11-07/obama-may-levy-carb...

Exactly -- slow and steady increases are fine (in inflation, taxes, etc.); it's risk which might strand capital assets which destroys and economy. Unfortunately, modern democracy (at least in the US) doesn't seem to do long-term sustained slow and steady anything; it's always wait wait wait then overcorrecting (or correcting more in the wrong direction). The Fed, which is quasi-independent, is one of the few organizations which seems to exert constant slight pressure, and that's complicated by the dual mandate.

Energy prices only reflect the immediate traceable costs of that energy, not the costs of the externalities of carbon production. Making something more expensive changes cost-benefits and changes incentives.

Data centers are 1-2% of the world's energy usage.

Yeah, but it is unclear what fuels power them. If it's 2% powered by nuclear/hydro, making them use 1/4 the electrical energy is less of a win than if there are big savings in places powered by coal or diesel or fuel oil.

At first glance the big datacenters are going in places with cheap hydro (TVA powers NoVA, right? and OR/WA, although they do use more coal than I would have thought).

Transportation remains one of the biggest potential wins, and some weird things like cement production which are huge outliers relative to economic value created.

It makes no difference what powers them, energy is interchangeable - if the data center doesn't use it, someone else will.

Energy isn't interchangeable across the globe. A lot of alternative fuel sources are in remote and underserved locations (Iceland has lots of geothermal); it's economically worthwhile to ship bauxite there, but not worthwhile to build a HVDC circuit to power a plant in downtown London from Icelandic power.

An even more clear example is that natural gas is about 30% the price in the US Midwest as it is in Europe or the rest of the world market. That's the whole point of the Keystone XL pipeline. Even oil has a 20% spread right now between central USA and coastal USA/world.

Sorry, a carbon tax will do absolutely nothing for the environment - the money will just go into the general fund, that is, the same slush fund that politicians already use for buying votes and paying off contributors to their re-election fund.

Taxes both raise revenue and "distort" the economy.

In some cases, these distortions are accidental (no one really wants capital investment reduced, which is the effect of even a 0.1% capital gains tax; they just consider the revenue raised to be more important than the distortion).

In some cases, the distortions are themselves the goal. Cigarette taxes are a great example -- I think the economy overall would be better off if there were 0% revenue on a $50/pack cigarette tax, as it would save health care and other costs far higher than the revenue on a $0.50/pack cigarette tax. Or, a local tax per-bag of trash produced, in excess of direct disposal taxes, in an attempt to capture the external costs of waste. (in Zurich, I think the bags are something like $5 each)

Carbon taxes could be either, but a steadily increasing carbon tax is much more like a cigarette tax than a capital gains tax.

Taxing something makes it more expensive, which reduces it. First law of economics: people respond to incentives.

Or drives it to black market. Canada implemented a VAT-like GST - in a country of 30 million, the first year it was implemented an estimated $1 billion underground (non-GST-paying) economy was formed.

Only if the costs of operating a black market are less than the costs of the tax. Coal and petroleum aren't like liquor and cocaine, though. It would be massively impractical to set up a black market for energy.

The closest thing is dodgy energy accounting -- making high energy-embodied or pollution-intensive products in places outside the regime, like Chinese steel mills.

They innovated on three frontiers - search engine algorithm, intent based advertising, and data center/infrastructure technology. A startup has to be really good at one thing to succeed. Google was really good at 3 three things right from it's early days.

As much as I like google, I doubt they're as good as that. Their search and advertising is essentially a monopoly because they have such a data lead over anyone else. That's why facebook freaks them out because they know how having more data is the secret sauce. Their data center innovations don't seem all that incredibly innovative. If they had robots whirring around I'd be much more impressed.

> If they had robots whirring around I'd be much more impressed.

They kinda do. See their self driving cars. http://en.wikipedia.org/wiki/Google_driverless_car

I mean in their data center. Also, I'd be more impressed if they actually got the self-driving cars out to the world. It's been 5 years now. Google has a problem executing, they're lulled by their one success.

I'm impressed because people have been trying to nail this problem for decades. Try convincing people that a few million automated half-ton canisters of metal going 60 miles an hour are safe. An early success will win investors or make a nice press release, but developing a rigorous, safe product of that complexity and risk is a long slog. It's got to be so well understood it's boring - then it's ready for massive public use.

If you live in the Bay Area you see the automated cars driving Google Employees to work all the time. I used to get an SMS any time somebody saw one (they are very distinctive) - but not as often, because people are starting to get more used to them. Though, they are still pretty cool to run across.

One success being Search, AdSense, GMail, Maps, Analytics, Chrome, or Android?

Success as in what earns their profits.

You got marhallp-trolled. Check the post history.

Dude, if you exclude all of its successes, Google is essentially a failure. #sarcasm

It has only been publicly announced two years ago.

Might wanna read this one too:

http://gigaom.com/cleantech/google-backs-iowa-wind-farm-has-...

Reply on news.ycombinator.com