New electricity distribution infrastructure or distributed generation?

For the sake of this thread let us please make the assumption that we do indeed need to significantly reduce our CO2 emissions and that renewables will be a large part of how that is done. This is not the thread to debate whether or not this assumption is correct but instead is debate the how: utility grade or in a distributed manner. Incentivized it also has its own thread.

Our transmission infrastructure is in poor shape. Utility level generation of power in solar and wind farms located where those resources are best will require running new efficient transmission from those locations to where the power will be needed. One solution for the intermittency of wind and solar is to balance them out across a supergrid with a high voltage direct current grid (HVDC).

These infrastructure investments can be quite sizable. Wind farms in particular require these sorts of investments if they are to play a major role (20% plus) in the future.

Another solution is avoiding major transmission infrastructure investment by minimizing the need for as much power, especially during peak power demand periods (when the system is most overloaded), to be sent through them. A major part of that tactic would be distributed generation. Among those options is locally generated solar, on individual home and business and factory rooftops and integrated into the building itself in new construction. This approach is attractive because as a resource it is best when peak demand hits most - daytime during Summer. Utilities could even own the equipment placed on a factory’s roof, giving a discount rate to the factory as rent and selling the excess to other local users. Some more rural locales can use local wind generation and some industries can use cogeneration to good effect as well. Local biomass electricity generation is also proposed as part of the mix. And there is the possibility of storing power locally to release to handle the spikes during the peaks that most stress the transmission and production infrastructure, either in multi-megawatt sized maegabatteries at the level of local substations, or as EVs of all stripes become commonplace, in vehicle to grid (V2G) applications.

Of course the two are not mutually exclusive but it does seem that one of the advantages of the distributed approach is to limit how much we need to invest in improved transmission infrastructure and that we really should pick a direction sooner rather than later. Which direction is preferable?

solar and wind are not solutions in themselves. They suffer when it gets dark and when the wind drops respectively. You need something for baseline load. Currently that something looks as if i’s going to be nuclear power. I’ve read great things here about Pebble-bed nuclear reactors. If what I’ve read here can be commercialised, and people get over their fears of nuclear power, then you could have a distributed network of PBNRs, basically solving the problem with minimal disruption. Indeed they’d actually remove a lot of the problems of power distribution. No more pylons. Automatic fail-over. And more.

Any opinions that have to do with the question of the op, quartz?

You are going to have to have some some sort of mix. HVDC is only more efficient for hugely long runs, with no taps. For local distribution, it’s just not cost effective, and the losses incurred in transforming voltage up and down quickly overwhelm any efficiencies you gain in long run transmission. So it might work for joining centralized power sources, like your wind farm or solar panel farm.

But the majority of the grid will need to stay AC

The same disadvantages of HVDC for local distribution are also a drawback of solar. Solar generates DC, which in order to be effectively distributed needs to be transformed into high voltage. The losses incurred at low voltages add up quick. A solar panel a block away at 120V is basically usless to you.

Converting DC is much more expensive than converting AC. With AC, a simple transformer, like the green boxes or transformer poles on every block, sometimes every house, lets the transmission stay at high voltages to minimize losses. That would be cost prohibitive for DC. So while having a solar panel array on a building might work for that building alone, it really won’t be able to supply power elsewhere without some hugely expensive converters to up the voltage for distribution to the next building over.

I did some work on this in the past, as stated by **Sinaijon ** HVDC is great for long distance runs, but AC is better locally.

For an ideal network you could have intercountry/continental HVDC (assuming it was possible to get the individual nation states to find agreement as to operating it etc) and local AC distribution.

From my own (unpublished) work:

Senator Harry Reid has just introduced a bill to help create a National EHV Overlay ( Basically a Green Superhighway that would transfer renewable eerngy long distances).

Not every house and busioness is suited to rooftop solar, of course. Some are simpy too shaded, some exist in areas where sunlight is not very ideal or are not postiiosned ideally. In fact, if it weren’t for the tranmission problem this would be a no-brainer. Bulilidng solar, wind and biomass facilities in the most favorable places is economically more efficient. However, transmission IS a problem. So personally I do not see either as being a stand-alone solution. However, my opinon (and I work in the field) is that building transmission to large utility size solar and wind installations (and geothermal and biomass) is the way to go for the large part, at least out west (where I do most of my analysis).

Siniajon and Walker,

How does net metering work then?

I might be able to answer, but I am not sujure I understand what you are asking. Net Metring is for rooftop solar and such (depending on the size of your property of course it doesn’t have to be on the rooftop). THe PV DC power is converted to AC using an inverter. It is genreally done right at the site of the solar panels and energy delivery is not really an issue. Since energy cannot be stored efficiently (there are ways, but not very cheap) it must be used as it is generated. If the house is not using it then it can literally run somone’s meter backwards and return energy to the grid to be used somewhere else. net metering gerneally refers to the fact that most people do not have time-of-use meters, so folks get paid only what they return to the grid in excess of what they use between meter reads. So while you may return a lot of energy during the day because you are at work, you offset that usage with energy you take off the frid at night.

Sorry if I was unclear Gangster. I understand that. My confusion is with Sinaijon’s statement that

and how that jives with the efficacy of net metering and the US Dept of Energy’s take that feeding excess solar energy back into the grid is cost effective and helps deal with peak demand. My proposal is merely the same with the difference that the utility owns the solar panels and sells the electricity.

Umm… yes. As I said. You’re concerned with issues of power distribution and as I said, PBNRs allegedly solve this because the power is generated locally. Of course, it has yet to be put to the test.

Ah. I see your point now. Sorry.

I think Sinaijon’s point was regarding a 120V DC solar panel trying to send its power a block away via DC. With Net Metering, your solar panel power is converted to AC before tieing into the grid. Any excess power flowing back into the grid then gets its voltage raised by the same transformer that normally drops the line voltage to your house voltage. So you get the efficiency of higher voltage transmission using the already-existing system.

Well yeah but it was said to explain why excess electricity generated by a factory roof’s solar panels could not be used to power other local demands in a distributed generation system. The converters do not seem to be so “hugely expensive” and are already in use.

A pertinent article.

The quality of the transmission infrastructure limits utility grade renewable deployment but as that article explains, this aint cheap stuff to do.

That article had me got me to this executive summary of the Joint Coordinated System Plan for infrastructure development (pdf). They attempt to compare some of the costs and benefits of enough infrastructure investment to bring wind up to 5% (using resources closer to population centers and not tying together with a long distance HVDC system) vs a 20% scenario. It’s a limited value study given the nature of assumptions made and the fact that it explicitly excludes accounting for cost savings going forward but for anyone curious it is an interesting, albeit confusing read.

So this much bottom line we have predicted: $50 billion for infrastructure to get to 5% wind and $80 billion for the infrastructure to get to 20%. Both likely cost effective over the long haul without accounting for the carbon cost or for the probability of it solving the possible intermittency issue but it is a lot of money. And investments will not be made while the various players are unsure of what long term policies will be.