Skip to main content

Power Costs Pinch Internet Companies

From today's Wall Street Journal (subscription required):
Some Internet executives say electricity has become a closely watched expense and can even be a factor when they consider rolling out new services, While always a concern, the cost of power has become more important amid a recent run-up in energy prices and increased use at data centers.

To satisfy their power needs, Internet companies are exploring options ranging from building facilities in former defense bunkers -- which already have rugged grid connections -- to plunking themselves near hydroelectric plants to get a slice of inexpensive power. Anticipating demand a decade from now, some executives are mulling whether proximity to nuclear power plants could be a plus.
An item like this isn't news to our readers here at NEI Nuclear Notes. This past December, we pointed to a study that appeared in ACM's Queue, which included the following projection made by the folks at Google:
"If performance per watt is to remain constant over the next few years, power costs could easily overtake hardware costs, possibly by a large margin," Luiz Andre Barroso, who previously designed processors for Digital Equipment Corp., said in a September paper published in the Association for Computing Machinery's Queue. "The possibility of computer equipment power consumption spiraling out of control could have serious consequences for the overall affordability of computing, not to mention the overall health of the planet."

(snip)

If server power consumption grows 20 percent per year, the four-year cost of a server's electricity bill will be larger than the $3,000 initial price of a typical low-end server with x86 processors. Google's data center is populated chiefly with such machines. But if power consumption grows at 50 percent per year, "power costs by the end of the decade would dwarf server prices," even without power increasing beyond its current 9 cents per kilowatt-hour cost, Barroso said.
The WSJ reports on how a number of companies are expanding their operations in the Pacific Northwest to position themselves to purchase power from hydroelectric plants, however, as our readers know, there are practical limits to that course, as hydroelectric capacity in North America is beginning to reach its practical limits.

In other words, there aren't a whole lot of sources left to dam. From the Wikipedia:
Apart from a few countries with an abundance of it, hydro power is normally applied to peak-load demand, because it is readily stopped and started. Nevertheless, hydroelectric power is probably not a major option for the future of energy production in the developed nations because most major sites within these nations with the potential for harnessing gravity in this way are either already being exploited or are unavailable for other reasons such as environmental considerations.
Which is why locating near a nuclear power plant might be the next option.

If I were a local politician concerned with economic development, I'd be paying close attention. Support jobs at data centers pay well, and attract a well-educated and affluent taxpayer to a local community.

Technorati tags: , , , , , Microprocessors, , ,

Comments

Rod Adams said…
Eric:

There was a recent article in Business Week that described the server farm developments in a small town in Washington. Not only is that course of action limited from a power point of view, but it is also limited from a network point of view. The Internet needs distributed information sources, it would not function if too many services are concentrated in a particular location.

Nuclear power is reliable power and has the potential for being cheaper than other alternatives. Costs need to be controlled, but it is doable.

Rod Adams

Popular posts from this blog

Activists' Claims Distort Facts about Advanced Reactor Design

Below is from our rapid response team . Yesterday, regional anti-nuclear organizations asked federal nuclear energy regulators to launch an investigation into what it claims are “newly identified flaws” in Westinghouse’s advanced reactor design, the AP1000. During a teleconference releasing a report on the subject, participants urged the Nuclear Regulatory Commission to suspend license reviews of proposed AP1000 reactors. In its news release, even the groups making these allegations provide conflicting information on its findings. In one instance, the groups cite “dozens of corrosion holes” at reactor vessels and in another says that eight holes have been documented. In all cases, there is another containment mechanism that would provide a barrier to radiation release. Below, we examine why these claims are unwarranted and why the AP1000 design certification process should continue as designated by the NRC. Myth: In the AP1000 reactor design, the gap between the shield bu...

How many nuclear plants does it take to meet the world's energy needs?

Several weeks ago Joshua Pearce at Clarion University in Pennsylvania released a study titled “ Thermodynamic limitations to nuclear energy deployment as a greenhouse gas mitigation technology .” In the study he stated... nuclear energy production would have to increase by 10.5% per year from 2010 to 2050 to both replace fossil-fuel-energy use and meet the future energy demands. This line, of course, made the headlines and has been picked up by several outlets and blogs . When looking into his calculations for this statement, he made one assumption error that overstated the above sentence by nearly a factor of three. Page 121, Section 4.1 of the study states: Richard Smalley pointed out that in 2004, the global economy consumed the equivalent of 220 million barrels of oil per day, which converted into electricity terms is the equivalent of 14.5 TeraWatts (TW), or 14,500,000 MegaWatts (MW) (2005). … With a nuclear plant having about 1000 MW (1 GW) of capacity, we would need 14,500...

What Happens During a Refueling Outage?

You may have noticed over the past few weeks that a number of nuclear plants are shut down for refueling outages or are resuming operations after just returning from one. This type of routine outage usually occurs in the spring or fall when electricity demand is low so that nuclear reactors can replace about one-third of the spent fuel rods with new fuel and conduct other routine maintenance and repairs at the plant. To get a better sense of how refueling works at a nuclear energy facility, I spoke with Marcus Nichol, NEI’s senior project manager for used fuel storage and transportation, and asked him to explain the basics. Why does a nuclear plant need to replace one-third of its fuel? Nichol: The main purpose of a refueling outage is to replace older fuel that is depleted—meaning it can no longer efficiently produce energy from nuclear fission reactions—with new fuel. This “used fuel” has typically been used in the reactor for four-and-a-half to six years before it is pe...