WC Archive

Wouldn't centralized power production (whether by sustainable means or not) be just a recontextualized version of the oil problem? i.e.: centralized power production would put the global grid at the mercy of local politics, natural disasters, whatever.

The "heavily distributed smart grid" has never made economic sense. Economies of scale are real; location factors in energy production (such as sunlight over Africa) are just as real. And energy alternatives will never become significant without being truly economically competitive, so these factors are important.

Buckminster Fuller proposed a "Global Energy Grid" some time ago - there even seems to be a group working to make it happen: http://www.geni.org/energy/issues/overview/english/grid.html
The problem is electric transmission technologies aren't efficient enough yet over such long distances.

It's a good idea - the energy equivalent of the internet. In the long run, given abundant world energy supplies, the global grid should be able to route around failures caused by local politics, natural disasters, etc.

I'm surprised that WorldChanging seems so okay with massive damming of a huge wild river!

the consequences for regional ecology and biodiversity would be mind-boggelingly profound-- look at the total ecological transformation of the Columbia River in the Pacific Northwest for example, then overlay the invaluable biodiversity and ecological richness of the Sub-Saharan jungle. Flooding that out is not okay!

The consequences for regional communities and indigenous peoples would be huge as well, because of dislocation and ruination of the regional environmental systems. And there are examples all over the developing world of huge dams going in with the usual fanfare of promises, the de riguor environment-trashing and dislocation, and then promised distribution beyond the cities never quite happening and many communities left more impovershed than before.

Much as I love the heavily distributed production grid model, i think global demand will require at least some intensive centralized production for several more decades. But dams are not the way to go... Instead how about pebble-bed fission in five years and magnetic confinement fusion in twenty.

Under some circumstances, hydroelectric dams are the right choice. This is not one of them, for reasons you lay out (as well as for the greenhouse gas problem). Frankly, I think the plan is highly unlikely to move forward at all, which is why I focused on the more abstract question of Africa as a global renewable energy producer.

I can see how my phrasing would leave a reader with the conclusion that I was encouraging this project. I'll change the wording to make it more clear. Thanks!

The "heavily distributed smart grid" has never made economic sense.

A lot of things didn't make economic sense at some point in time. But because of advances in computers and communication technology, distributed energy is a reality and growing quickly these days.

The economic nonsense of distributed generation is unrelated to computer and communications technology, but rather to the cost of the energy technology itself. If the NorthEast gets 3 times less usable solar energy in a year than the SouthWest of the US, then the same capital expended on solar installations on NorthEast rooftops would produce 3 times as much payback if put into centralized solar production facilities in the SW. Wind power is similarly location dependent. Forcing everybody to generate their own power locally is a huge waste of resources - something pro-environment people should be opposed to.

There's sufficient diversity in "centralized" power generation for it to have all the putative benefits of distributed power anyway - we have thousands of generators already interconnected in North American electric grids and it seems to work pretty well.

Ah, the Congo and the fantasy.
Congo has had its fair share of megalomaniacs wanting to exploit the huge, abstract potential of the river.
-Mobutu built Inga upon Inga (I, II, III and IV), of which only 2% is used today
-Khadaffi still wants the pan-African Congo channel, bringing fresh water all the way to North Africa and South Africa, over the mountains, through the jungles and accross the Sahara
-The Saoudis wanted a water pipe-line + electricity...

The problem with all these plans is the reality on the ground. It's a catch-22: as long as there's political instability on the ground (basically the worst war since WWII), none of those plans will leave the drawing board; and as long as there's no revenue for Congo, the resource wars will continue.

Mega-plans are definitely not the way to go. They're top down scenario's, while nobody's investing in the small bottom up initiatives.

As they say in Congo: the water in the kettle starts to boil from the bottom up.
The megalomaniancs better remember this wisdom.

Check out this other Congo fantasy:
A gigantic fresh water pipe-line bringing Congo water all the way to the Middle East.

No comment on whether this would be a good idea or not... However, the fact that Mr Khoza mentions selling the "surplus" to southern Europe seems a little odd. California uses over 60,000 megawatts. How is Africa going to industrialize and provide power to all of its people with 40,000 megawatts?

The economic nonsense of distributed generation is unrelated to computer and communications technology, but rather to the cost of the energy technology itself. If the NorthEast gets 3 times less usable solar energy in a year than the SouthWest of the US, then the same capital expended on solar installations on NorthEast rooftops would produce 3 times as much payback if put into centralized solar production facilities in the SW. Wind power is similarly location dependent. Forcing everybody to generate their own power locally is a huge waste of resources - something pro-environment people should be opposed to.

The end-use cost of electricity isn't simply a factor of efficiency at the production point. That's the whole reason why distributed generation can often make greater economic sense. Not only does it cost a considerable amount of money to build and maintain long-distance transmission networks, you also get substantial transmission losses along the way as a function of distance.

There's also a cost to reliability, and this is why some energy-intensive locations (like server farms) might be better served by having an independent and proximate power source, simply because the costs of losing power could be substantial.

It's not some simplistic matter of $/kWh at the point of production.

And distributed energy isn't simply about electricity. You can also gain efficiencies from co-generation.

Then there's also pollution considerations, which is the reason you see a lot of smaller natural gas plants popping up to make electricity instead of more and more large coal or nuclear plants.

One thing that pisses me off about this page is the constant talk about nanotubes. They aren't really feasible to mass produce, not even in the near future.

Regardless of the detriment of damming up a river the size of the Congo, but what about the tranmission costs? I've been googling around and I can't fine any reliable numbers (anyone?) but common sense tells me that a huge percent of anything generated in the Congo, by whatever means, is going to be lost in transmission.

Losses in transmission are certainly an issues for large scale distribution - one that deserves more R&D investment, but that was part of the initial discussion and my comments as well. Capital cost of transmission is also going to be an issue, I agree. For typical 90% efficient 1000 km-scales though it's far less than the cost of solar photovoltaics...

If you have reason to pay for reliability, you HAVE to pay for redundancy - because if both local and grid sources can't independently carry the load, you don't have reliability. And off-grid locations receive power much less reliably than on-grid. It makes economic sense for some applications, but all you're really doing is building a "fancier grid" at a local scale - the main grid is reliable enough for most applications, just as the internet is quite good enough for most networking applications, you don't need to build your own (as many companies did in the 1980's and 90's).

$/kWh at the point of production is an excellent benchmark because it really is how you measure how much coal production each $ expended can replace (assuming typcial low transmission costs and losses).

Co-generation's an interesting point - but it's not clear why it's anything new. Why don't we have more heat-intensive processes running off coal-fired power plants right now? After all about 2/3 of the energy's going to waste! There are lots of industrial processes that need heat - but my guess is there's a mismatch between the quality supplied and what's needed, and any kind of co-generation is going to run into the same problem. True you could heat your building in winter using the waste heat from your electric generation - but you need more electricity to cool it in summer, and what's the use of that waste heat then? Co-generation may help a tiny bit, but it's unlikely ever to be a big component of the solution, as far as I can tell.

Why don't we have more heat-intensive processes running off coal-fired power plants right now? After all about 2/3 of the energy's going to waste!

Because the biggest coal-fired plants tend to be in the middle of nowhere.

Co-generation may help a tiny bit, but it's unlikely ever to be a big component of the solution, as far as I can tell.

If hydrogen were the energy carrier of choice and we had more neighborhood level power production from it, then the waste heat could certainly be used to heat homes. That's just one example.

I think we're kind of speaking across each other here, because I believe I'm speaking of "distributed energy" in the most literal sense -- meaning not centralized. We already do this with most heating applications -- there aren't giant heat plants in the country piping in hot liquid to the cities. That's because of the heat loss in covering distance, as well as the cost of pumping that liquid increasing with distance. Electricity behaves in a similar way, economically speaking, but obviously with less loss per unit of distance.

We could also be discussing the issue of the cost of incremental cost of additional capacity and the costs associated with maintaining a more centralized infrastructue given variability in demand throughout a given day and variations in demand throughout the year.

Certainly that system could be made more efficient, but the room for efficiency gains is far greater with a distributed approach. And certainly the cost of communication and computing technology relative to performance is a factor in this. For example, twenty years ago, a power company that would have had solar power generating from 100,000 rooftops would have had an unbelievable cost in simply maintaining that system. Nowadays, such a system can be remotely monitored at an instant. That's just one simple example.