11 ms·

I'm really surprised no one at Facebook thought about the transmission loss problem before building that way. It seems like a surprising oversight for the count

by beat 7y ago

I'm really surprised no one at Facebook thought about the transmission loss problem before building that way. It seems like a surprising oversight for the countless scientists and experienced engineers involved with the project. Alternately, maybe there's something we don't see in the one-line project description from an internet comment. edit: A quick googling answered that - they're buying the power for a data center in Nebraska, not in California.

krobbn 7y ago
Not as bad as NYC building a 333 mile, $3B transmission line to get hydro power supplied from Canada. Rather than spending $3B to build out wind or other renewable power locally.
- beat 7y ago
  Again, don't make assumptions that it's stupid, based on little data and no math. First, hydro power is very stable, and Quebec has a surplus, so they could probably get a good price deal for an ongoing relationship. The construction cost is nothing, compared to the ongoing cost. What does that Canadian power cost, vs the cost of wind? They could be looking at tremendous savings over the next few decades. Second, this appears to be a New York City decision, not a New York state decision. Does NYC have enough viable land to build a gigawatt (or more like 3-4gw, considering load variation) of wind? Probably not.
IAmEveryone 7y ago
Transmission losses are also far lower than people keep pretending on the web. This[0] says 3% per 1000km, meaning Nebraska to California could be done at about a 5% loss, which could easily be offset by any number of factors such as lower construction costs, more favourable wind conditions, or availability of land. [0]: https://en.m.wikipedia.org/wiki/High-voltage_direct_current#Advantages https://en.m.wikipedia.org/wiki/High-voltage_direct_current#...
- owenversteeg 7y ago
  I don't think people are "pretending", just that losses are different based on what system you have (not just DC vs AC, but also the voltage, the type of insulator on your cables, the conductor [copper/aluminum, most new stuff is aluminum] and the substations.) Corona losses, specifically, are also strongly affected by the weather! Then you have people writing about it doing incorrect math on some percentage they don't understand from a technical paper. That said, I agree that 5% sounds close for the transmission losses alone (excluding distribution), and 5% could be easily made up, like you said, by many things, like being next to a wind turbine factory/cheaper land/cheaper labor/tax breaks/many things. Interesting note: I found here [0] that silver would be a feasible conductor for HVDC lines if prices were lower. On an example line, it would save $19M/yr but cost $18.5B to build, so 1000 year payback time, which is unrealistic. But! If the price of aluminum increased substantially (for example: say we discover smelting releases mega-carcinogens and it's banned) and the price of silver fell (silver is now 4x as expensive as it was in 2003, so I feel like it could fall fairly precipitously) then we could in theory have silver power lines. One thing to remember, though, is that it's only the "skin" of the conductor made of copper/aluminum/silver - the skin carries 90% of the power, but the bulk of the conductor's mass is the cheap steel core that doesn't need to be as conductive because it only carries 10% of the power. [0] http://large.stanford.edu/courses/2010/ph240/harting1/ http://large.stanford.edu/courses/2010/ph240/harting1/

by beat 7y ago

krobbn 7y ago

Not as bad as NYC building a 333 mile, $3B transmission line to get hydro power supplied from Canada. Rather than spending $3B to build out wind or other renewable power locally.

beat 7y ago
Again, don't make assumptions that it's stupid, based on little data and no math. First, hydro power is very stable, and Quebec has a surplus, so they could probably get a good price deal for an ongoing relationship. The construction cost is nothing, compared to the ongoing cost. What does that Canadian power cost, vs the cost of wind? They could be looking at tremendous savings over the next few decades. Second, this appears to be a New York City decision, not a New York state decision. Does NYC have enough viable land to build a gigawatt (or more like 3-4gw, considering load variation) of wind? Probably not.

IAmEveryone 7y ago

Transmission losses are also far lower than people keep pretending on the web. This[0] says 3% per 1000km, meaning Nebraska to California could be done at about a 5% loss, which could easily be offset by any number of factors such as lower construction costs, more favourable wind conditions, or availability of land. [0]: https://en.m.wikipedia.org/wiki/High-voltage_direct_current#Advantages https://en.m.wikipedia.org/wiki/High-voltage_direct_current#...

owenversteeg 7y ago
I don't think people are "pretending", just that losses are different based on what system you have (not just DC vs AC, but also the voltage, the type of insulator on your cables, the conductor [copper/aluminum, most new stuff is aluminum] and the substations.) Corona losses, specifically, are also strongly affected by the weather! Then you have people writing about it doing incorrect math on some percentage they don't understand from a technical paper. That said, I agree that 5% sounds close for the transmission losses alone (excluding distribution), and 5% could be easily made up, like you said, by many things, like being next to a wind turbine factory/cheaper land/cheaper labor/tax breaks/many things. Interesting note: I found here [0] that silver would be a feasible conductor for HVDC lines if prices were lower. On an example line, it would save $19M/yr but cost $18.5B to build, so 1000 year payback time, which is unrealistic. But! If the price of aluminum increased substantially (for example: say we discover smelting releases mega-carcinogens and it's banned) and the price of silver fell (silver is now 4x as expensive as it was in 2003, so I feel like it could fall fairly precipitously) then we could in theory have silver power lines. One thing to remember, though, is that it's only the "skin" of the conductor made of copper/aluminum/silver - the skin carries 90% of the power, but the bulk of the conductor's mass is the cheap steel core that doesn't need to be as conductive because it only carries 10% of the power. [0] http://large.stanford.edu/courses/2010/ph240/harting1/ http://large.stanford.edu/courses/2010/ph240/harting1/