Fusion Power Economic Impact: An Overview

[BWV]

It's a rough and in exact analogy to Moores law, but the thing that really matters is cost per KWh, which is declining rapidly as production scales combined with perhaps a doubling or tripling of efficiency. No one really doubts that 20 years from now, Solar Pv will be far cheaper and more efficient than it is today where it already is at about grid parity. The issue with fusion is no one has any idea if it will be a viable source of energy in 20, 50 or 100 years. No one has come close to generating more power from fusion than is input into the reaction, let alone at an economical cost. The most optimistic projections of working fusion plants 20 or 30 years from now propose capitalized power costs of 2-3 cents per KWh - currently where combined cycle gas is in the U.S. and where solar will be in a few years. There are higher and better uses for the tens of billions of dollars it will take to gamble on these speculative projects.

 

[gleem]

2012 energy usage for the world was estimated at 424 TWH/day. Current PV cells produce about 18 W/sq ft. To meet that demand would currently require 4.67 10¹² sq ft of cells. A current but cheap 135 w 20% efficient panel is about 7.5 sq ft and costs around 200 USD. This gives 622 billion panels at a cost of 124 T USD. (US GDP = $17 T). I think more efficient panel are a ways off and will cost more. Installation cost will increase too. Along with inverters, charge controllers and storage batteries ( the technology of which must also improve). will surely double the cost of materials. So with the installation cost land acquisition cost we're talking about a chunk of change.

A typical totally independent household installation with 100% yearly irradiation using about 50 KW H/day would currently require 9 KW array to generate at least 50 KWH of electricity plus at least enough storage capacity for the low light hours. It would use on average 4000 AH per day (12V storage system) and need a battery bank delivering at least 166 A on average for 24 hours . Typical large capacity storage batteries have a max AH rating around 250 AH at cost around 500 USD. You would need twice the capacity to store 4000 Ah to maximize battery like so you would need 32 such batteries almost equal to the cost of the PV array. In real life most would need more PV cells and storage due to clouds, seasonal variation of the declination of the sun and length of the day.

One final note the Ivanpah Thermal Electric Solar Generator in California is currently only able to reach about 40% of its one million MWH /yr capacity after 15 month of operation. It cost 2.2B USD and uses 4000 acres of prime solar territory in the Mohave desert. You would have thought estimating the capacity would have been a no brainer, so much for estimates.

So I'm thinking efficiency in energy usage.

 

[BWV]

No one energy source will ever generate 100% of global power, but the area requirements you describe are not unachievable - rooftops could be a third or more of the area notwithstanding future increases in efficiency
(http://www.nrel.gov/docs/fy14osti/60593.pdf)

Solar is not ready for prime time yet, it will be another generation before it can generate a meaningful percentage of global electricity, but its problems are dwarfed by those in developing viable fusion reactors

 

[rootone]

As far as I know the ITER project is the first actual attempt so far to build a fusion reactor at an economically viable scale.
It is based entirely on known physics and engineering knowledge gleaned as a result of earlier experimental reactors, it's no shot in the dark based on guesswork.
We can't say that fusion is not economically viable end-of, because the whole point of the project is to prove that it is (or isn't) viable.
Designed to produce 500 megawatts of output power while needing 50 megawatts to operate it's comparable to a small-medium size fission plant.
It's not intended as a commercial generating station though, rather as a proven prototype on which commercial designs could be based.

I'm certainly not opposed to improving the effectiveness of solar collection methods, but if economically viable fusion reactors are achievable then why not do so?
Do both, - each has it's own unique advantages, and best area of applicability, and they both pose less environmental hazard than much of currently existing power plants.

 

[mfb]

It's a rough and in exact analogy to Moores law, but the thing that really matters is cost per KWh, which is declining rapidly as production scales combined with perhaps a doubling or tripling of efficiency.

Yes it is declining, but in no way comparable to the millionfold increase in number of transistors Moore's law gave in the last decades.

No one really doubts that 20 years from now, Solar Pv will be far cheaper and more efficient than it is today where it already is at about grid parity.

You have to make very optimistic calculations to get grid parity today:
- assume that solar power is always needs exactly as produced, so storage does not matter. This assumption is highly unrealistic, and storages are expensive
- compare the production costs to electricity price households pay, instead of comparing the production costs to production costs of other power sources. This means you have to ignore the costs for the grid infrastructure - exactly the point you need more with photovoltaics to balance load and production. You also ignore taxes.
- assume that subsidies continue to flow as they do now but do not account for them. Germany alone (!) invested more than 100 billion euros in photovoltaics already, and commited to invest at least 100 billions more. That is ten times the costs of ITER.

The issue with fusion is no one has any idea if it will be a viable source of energy in 20, 50 or 100 years.

That's exactly why we should investigate it.

capitalized power costs of 2-3 cents per KWh - currently where combined cycle gas is in the U.S. and where solar will be in a few years.

This is speculation.

20 billion euros, distributed over 3 billion people and 10 years, is 60 cents per average person and year. $0.62 per US citizen, taking into account the actual distribution of the costs and the current exchange rate (and the very crude total cost estimate). I'd happily spend twice that amount to have a second ITER.
As comparison: 200 billions over 20 years for 80 million in Germany are 125 euros per year for photovoltaics.

 

[gleem]

One wonder how implementation of fusion will unfold. Fission reactors where suppose to provide clean, almost limitless energy when introduced. Not much was talked about the inherent hazards of a fission reactor or waste disposal. We eventually settled on a technology that only used 4% of the fuel in the rods was used before they where "poisoned" and had to be replaced loosing 96% of the promised energy. Reactor safety was (is?) not totally understood.

What promises for fusion will go unfulfilled, or overstated. Will we just create another "mess" to clean up?

 

[rootone]

... Will we just create another "mess" to clean up?

Apart from the fact they are both 'nuclear', a fusion reactor is an entirely different technology to a fission reactor.
Fusion reactors produce no appreciable amount of waste by products at all.
(What it does produce is Helium which is in fact rather useful stuff)

 

[gleem]

Fusion reactors produce no appreciable amount of waste by products at all.
(What it does produce is Helium which is in fact rather useful stuff)

Not directly but It will make lots of neutrons which will activate the containment vessel. ITER is suppose to obtain data on this issue. ITER is a breeder for tritium so it is not just a simple "Sun in a bottle". How long will the bottle last? Sure I say go for it but don't put all your energy concerns "eggs" in one basket for now. I am doubtful I will see ITER go on line in my lifetime and I am doubtful that most of you will see a viable commercial reactor in yours.

 

[mfb]

The vessel will become activated. It is studied how that effect can be mitigated. Medium lifetimes are the most problematic. Anything living shorter than about a year (half-life) can be stored until it is decayed, everything living billions of years has a tiny activity and no relevant heat production, and can be stored underground forever.

Fusion power plants won't work completely without problematic waste, but they do not produce all the problematic (trans)actinides fission power plants generate and the amount will be significantly smaller.

 

[mheslep]

... PV cells are a semiconductor technology that has a Moore's law dynamic .. it

Moore's law derives from ever increasing circuit density and complexity on a given size of semiconductor. PV cells have no such connection. Neither does the storage technology required for when the big fusion reactor in sky lights the other side of our sphere.

 

[mheslep]

...but they do not produce all the problematic (trans)actinides fission power plants generate and the amount will be significantly smaller.

Not necessarily so for all fission, but for thermal spectrum U235 fission which happens to be the dominate form of fission reactor at the moment. It need not be so. The Russians are currently operation two fast reactors that burn actinides, and the U.S. has a couple startups with low volume waste designs.

 

[Burnerjack]

What I see is little to no change from the consumer's point of view. If Fusion were indeed made exceptionally cheap, massive build out effort to ensue, increased efforts to break distribution bottlenecks. All made possible by a wider profit margin. The price for delivered power will remain at what the market will bear. However, much change in the balance of delivered energy as far as point of use fuels vs. grid supplied energy goes. Morale of the story: Investor Benefit. Cleaner environment.

 

[mheslep]

The price for delivered power will remain

In the US, some 2/3 to 3/4 of the price of grid delivered electric power is due to construction and maintenance of the grid. In this sense the promises of "too cheap to meter" was never true. So even zero fuel cost to the utility would not much reduce the cost to deliver residential power.

 

[mfb]

The infrastructure price can go down if losses become less important. Probably not a large effect, but large enough to be considered.

 

[Iain53]

I'm not in a position to critique the pros and cons of competing technologies such as Edit: link removed
but I would think that there needs to be a shift from the approach that's been used in the past. Large, continental power grids are far too complex and vulnerable, I like the idea of developing power sources that are much smaller and thus much safer. Local generation and consumption makes more sense than multibillion dollar, huge plants, whether conventional or nuclear.

 

[russ_watters]

Large, continental power grids are far too complex and vulnerable, I like the idea of developing power sources that are much smaller and thus much safer. Local generation and consumption makes more sense than multibillion dollar, huge plants, whether conventional or nuclear.

I don't agree with any of those statements. The complexity of power grids makes them less vulnerable and more reliable than local power generation because there are multiple pathways for the power to get from one place to another. More, local power plants means more failure points with less back-up and distributed risk (smaller but much more frequent accidents).

We have the grid that we have for real reasons; it didn't happen by accident.

 

[mfb]

In general, larger power plants are also more efficient, cheaper and cleaner. A 1 GW coal power plant produces more electricity per coal than a 1 MW coal power plant, does not require 1000 times the manpower to operate, can afford better filtering of its emissions, and so on. Photovoltaics is probably the only thing where the scaling effects are not that significant.

 

[Ryan_m_b]

Photovoltaics is probably the only thing where the scaling effects are not that significant.

Correct me if I'm wrong but it seems like another advantage of a grid, especially a large continental one, is so that things like PV can be used to power places where it would be less efficient.

 

[mfb]

Yes, but I don't see how that is related to my post.

Building ten 1 MW photovoltaics power plants is not much more expensive than building one 10 MW power plant - the bigger plant is mainly just the same thing 10 times next to each other. That is the scaling I was talking about.
Building ten 100 MW nuclear power plants is much more expensive than building one 1 GW power plant.

 

[russ_watters]

Solar is a bit of an odd duck on this issue. It is very unreliable (second only to wind) and as a result is almost always grid connected. I'm not sure about the economics of local vs centralized because the economics are so heavily manipulated and therefore they aren't really real (in addition to being not well publicized and being variable). The main benefit I see on grid vs distributed for solar is that buildings have roofs, which are somewhat wasted real-estate. .

[edit: Looks like I went to a different level of "local" than mfb was referring to]

 

[Iain53]

I don't agree with any of those statements. The complexity of power grids makes them less vulnerable and more reliable than local power generation because there are multiple pathways for the power to get from one place to another. More, local power plants means more failure points with less back-up and distributed risk (smaller but much more frequent accidents).

We have the grid that we have for real reasons; it didn't happen by accident.

With all due respect, and I readily acknowledge that you have earned respect, the North American power grid is in grim shape. It is a hodgepodge collection of independent utilities with risks associated with command, control, and communication, aging infrastructure, and growing demand. The initial ideas that were the foundation for the grid were from Edison's time. The lack of adequate network security has been documented widely. I'm sorry, but I don't see how a fragmented, aging system such as the power grid can be considered robust. You make good points regarding the ability to switch loads to deal with outages, however, the record shows that brown-outs and black-outs are becoming quite common, and they are lasting longer. I don't think that the benefits of networking such as those we see in the internet are mirrored in the power grid although efforts are ongoing to emulate that kind of redundancy.

 

[Evo]

With all due respect, and I readily acknowledge that you have earned respect, the North American power grid is in grim shape. It is a hodgepodge collection of independent utilities with risks associated with command, control, and communication, aging infrastructure, and growing demand. The initial ideas that were the foundation for the grid were from Edison's time. The lack of adequate network security has been documented widely. I'm sorry, but I don't see how a fragmented, aging system such as the power grid can be considered robust. You make good points regarding the ability to switch loads to deal with outages, however, the record shows that brown-outs and black-outs are becoming quite common, and they are lasting longer. I don't think that the benefits of networking such as those we see in the internet are mirrored in the power grid although efforts are ongoing to emulate that kind of redundancy.

You say this is all documented widely, yet you failed to post this documentation, please post the sources. Thank you.

 

[Iain53]

You say this is all documented widely, yet you failed to post this documentation, please post the sources. Thank you.

http://www.ibtimes.com/aging-us-power-grid-blacks-out-more-any-other-developed-nation-1631086
http://scitation.aip.org/content/aip/magazine/physicstoday/news/10.1063/PT.5.5020
http://blogs.scientificamerican.com...ower-grid-is-in-need-of-a-technology-upgrade/
http://www.datacenterknowledge.com/archives/2013/01/16/us-power-grid-has-issues-with-reliability/
http://www.nytimes.com/2013/11/11/booming/the-blackout-that-exposed-the-flaws-in-the-grid.html?_r=0

 

[Evo]

http://www.ibtimes.com/aging-us-power-grid-blacks-out-more-any-other-developed-nation-1631086
http://scitation.aip.org/content/aip/magazine/physicstoday/news/10.1063/PT.5.5020
http://blogs.scientificamerican.com...ower-grid-is-in-need-of-a-technology-upgrade/
http://www.datacenterknowledge.com/archives/2013/01/16/us-power-grid-has-issues-with-reliability/
http://www.nytimes.com/2013/11/11/booming/the-blackout-that-exposed-the-flaws-in-the-grid.html?_r=0

Sorry, I forgot to say that you need to also quote the pertinent parts (be careful not to violate copyright laws by quoting too much) as you cannot expect people to read an entire article and try to figure out which part you are referring to. Please quote the pertinent parts. Please be sure that they are current. And just two will be fine. Thank you.

For example

1. The electric grid is an engineering marvel but its aging infrastructure requires extensive upgrades to effectively meet the nation’s energy demands. Through the Recovery Act, the Department invested about $4.5 billion in grid modernization to enhance the reliability of the nation’s grid. Since 2010, these investments have been used to deploy a wide range of advanced devices, including more than 10,000 http://www.smartgrid.gov/glossary/term/73 , over 7,000 http://www.smartgrid.gov/glossary/term/315 and approximately 15.5 million smart meters. See a map of the Recovery Act-funded Smart Grid Investment Grant and Smart Grid Demonstration projects at smartgrid.gov:
2. SmartGrid.gov:
   The American Recovery and Reinvestment Act of 2009 (Recovery Act) provides the U.S. Department of Energy with $4.5 billion to modernize t...
   [/PLAIN]
   SMARTGRID
3. One of the key solutions for a more resilient and reliable grid is http://www.smartgrid.gov/sites/default/files/doc/files/Synchrophasor%20Report%2008%2009%202013%20DOE%20(2)%20version_0.pdf . These mailbox-size devices monitor the health of the grid at frequencies not previously possible, reporting data 30 times per second. This enhanced visibility into grid conditions helps grid operators identify and respond to deteriorating or abnormal conditions more quickly, reduce power outages and help with the integration of more renewable sources of energy into the grid. To date, nearly 900 of these devices have deployed as a result of Recovery Act investments.

]http://energy.gov/articles/top-9-things-you-didnt-know-about-americas-power-grid

Recovery[/PLAIN] Act Projects

The American Recovery and Reinvestment Act of 2009 provided the U.S. Department of Energy with $4.5 billion to modernize the electric power grid. Learn about the projects that utilized these investments to deploy smart grid technologies and their impact on grid performance.

https://www.smartgrid.gov/

 

[dlgoff]

... the North American power grid is in grim shape. It is a hodgepodge collection of independent utilities with risks associated with command, control, and communication, aging infrastructure, and growing demand. The initial ideas that were the foundation for the grid were from Edison's time.

Something must be working. How do you think over 1100 Gigawatts of energy are getting transported?
image compliments of http://www.tsp-data-portal.org

 

[mfb]

the North American power grid is in grim shape.

The European one works nicely, blackouts are rare and usually limited to local regions for local reasons (e. g. ice or heavy snow taking down the power lines going to some town). And one counterexample is sufficient to refute the claim that "Large, continental power grids are far too complex and vulnerable".

The largest blackout happened in 2003, affecting 56 million people - for three hours in the middle of the night. Uhh, scary. A small fraction of the population got power a bit later than the others.
The second largest blackout happened in 2006, with 15 millions without electricity for 2 hours.

Going by Wikipedia articles, London 2003 is the last European one that has its own article - 500,000 people for 30 minutes to 2 hours.

The longest blackout I personally experienced in my life was about 5 minutes long.

 

[gleem]

For what it is worth I heard In a news report some time ago that the US power grid is very vulnerable to severe Sunspot activity and in particular the present type of transformers that are used. Apparently there are new types that would withstand such an event. The problem being that there would not be enough spares on hand to replace the damaged one and the lead time for the manufacture of replacement transformers is 6 months. It is just another example of a neglected infrastructure issue. IMO the maintenance of the power grid is akin the the repair of highways be filling the pot holes.

 

[gleem]

The longest blackout I personally experienced in my life was about 5 minutes long.

Periodic blackout in the US are quite common due to storms, but these are typically localized affecting only a part of a state or two but sometimes will last many days. I expect several power outages in my are of Maryland several times a year lasting up to about 12 hrs. but usually only a few hours.

However the power grid in the northeast has been compromised significantly twice in the last 50 years.

In 1965 much of the northeast of the US and parts of Canada lost power for over 12 hrs due to a maintenance error. https://en.wikipedia.org/wiki/Northeast_blackout_of_1965

In 1977 New York City lightning took down the cities grid for almost 24 hrs, before totally restored.
https://en.wikipedia.org/wiki/New_York_City_blackout_of_1977. During this time there was significant civil disobedience.

In 2003 a computer "bug" took down the grid of the northeast of the US and Canada affecting 10's milliions of persons .https://en.wikipedia.org/wiki/Northeast_blackout_of_2003 with some remote areas being 2 weeks without power

The European power grid is newer than much of the US because it has been rebuilt since WWII.

 

[mheslep]

...and growing demand...

In the US electricity demand has been flat, not growing, for several years now, even in the face of a slowly growing population. Electricty use per capita has dropped to levels last visited in the 1990s.

 

[mheslep]

http://www.ibtimes.com/aging-us-power-grid-blacks-out-more-any-other-developed-nation-1631086
...

That news article references this DOE summary report which states:

Today’s electricity system is 99.97 percent reliable, yet still allows for power outages and interruptions that cost Americans at least $150 billion each year — about $500 for every man, woman and child.

Or an average outage of 2.6 hrs per year.

 

[litup]

No. A computer chip is a piece of technology, not a building. The cost of buildings and non-technology products do not follow Moore's Law, never have and never will. The materials and labor that go into them always get more expensive over time, not less.

Fundamentally, I see no reason why a fusion plant should be less expensive than a fission plant, especially since fusion is proving to so difficult to make happen (unlike fission, which was immediately commercially deployable). So the economics works like this:

1. If the fuel for fusion power were free and the plant cost the same to build, fusion power would cost 14% less than fission power.

That's not a very compelling promise for something we've been waiting for decades for and spending enormous sums of money for without success.

For the above reasons, I think fusion power is near totally irrelevant for the near and long term future of civilization. And if it ever gets in operation at all will not be any cheaper than fission power.

Remember: pretty much the only thing we know for sure about fusion power is that it is a lot more difficult to do than fission power.

Except for the end game cost where you have massive amounts of useless fission byproducts that have to be sequestered for thousands of years.
Take that into consideration in your calculus.

 

[russ_watters]

Except for the end game cost where you have massive amounts of useless fission byproducts that have to be sequestered for thousands of years.
Take that into consideration in your calculus.

I do; that is a very small cost that is already priced into the cost of fission energy...though I think that charge was temporarily suspended by the courts due to the government's violation of their end of the contract (building the facility).

By coincidence, that cost is nearly equal to what we spend on fusion research.

 

[mheslep]

useless fission byproducts that have to be sequestered for thousands of years.

The dangerous fission products decay to below the radioactivity of uranium ore in ~500 yrs. The transuranic, non-fission products, like Pu, created in reactors have much longer half-lives, though they are alpha emitters with much lower specific radioactivity than the fission products.

The heavy metal waste in, say, solar panels never disintegrates.

 

[mfb]

The dangerous fission products decay to below the radioactivity of uranium ore in ~500 yrs.

3000 years going by your graph.

Am-241 is useful as alpha emitter, if you separate it to use it elsewhere the waste falls below the level of uranium ore much faster. A factor 1000 for Sr/Cs needs about 300 years.

 

[mheslep]

3000 years going by your graph.

Am-241 is useful as alpha emitter, if you separate it to use it elsewhere the waste falls below the level of uranium ore much faster. A factor 1000 for Sr/Cs needs about 300 years.

Yes, but not a fission product. Am-241 is a transuranic produced by neutron capture and decay from other transuranics, an important difference as some of the Gen IV designs would significantly burn their transuranics. An alpha emitter with low penetration power, it is dangerous when inhaled or ingested. I'd rather have some grams of Am-241 buried in my backyard (or micrograms in my smoke detector) rather than some tons of heavy metal laden coal waste.