Yet another tokamak

[georgepds]

"On Friday, Sept. 30, at 9:25 p.m. EDT, scientists and engineers at MIT’s Plasma Science and Fusion Center made a leap forward in the pursuit of clean energy. The team set a new world record for plasma pressure in the Institute’s Alcator C-Mod tokamak nuclear fusion reactor. "


http://news.mit.edu/2016/alcator-c-...cord-1014?mc_cid=4f05fb7619&mc_eid=43ddca35db

Some where in the article they say they expect a commercial product in ~20 years

 

[iamlucky13]

Not "another" Tokamak as in a new one, but one of several existing in the US. It's been around for over 20 years, focusing on the plasma confinement challenge. Final plasma was last month. I guess the plan is to mothball it for a few years in case additional experiments are warranted in support of ITER, and then eventually dismantle it. The current DOE budget doesn't provide enough funding to continue research on Alcator C-mod and meet our ITER commitments.

 

[DBoon]

Fusion is always 20 years away from commercialization. When we finally get there, solar+wind with storage will be ubiquitous and cheaper than fusion, for sure. Time to throw in the towel on fusion.

 

[iamlucky13]

The 20 year quip is a comedy line, not accurate history. The reality is we've never purposefully tried to commercialize it. It's been stuck in the research phase because the funding has never been sufficient to work through all the open research questions to actually transition to a development phase oriented specifically at commercialization. The ITER project is the first actual attempt to do so, but we've been underfunding that, too, which has screwed up the schedule, in turning driving up the cost, and caused a cyclic brain drain, which worsens the schedule problems.

Aside from the problems of underfunding ITER, shutting down non-ITER facilities like Alcator means research they were doing to support the design of ITER doesn't happen. As a result, ITER will end up being less refined than it should have been, and require time and mid-life upgrades to achieve all of its goals. Penny wise and pound foolish.

The way people talk about fusion research, you'd think we were throwing unfathomable sums of money into it, yet when one of the main fusion labs in the country shuts down, nobody even notices. I just looked up the annual budget for Alcator C-mod before it was shut down: $22 million.

Anything million sounds like a lot to most people, but that's a bargain for a year's worth of successful research in the mid-cycle development of a new energy technology. My local electric utility recently spent than that on a couple cargo containers filled with batteries. To address one of your other claims, the price they spent and the listed performance for those batteries works out to around $0.54/kWh (on top of the cost of generating the electricity in the first place) assuming they last 20 years with zero maintenance, operating, or financing costs, none of which are accurate assumptions. We're talking about relatively straight-forward upscaling of a technology that has been on the commercial market for 20 years, but still needs another 90% cost reduction to be cost-effective for much more than load balancing.

The federal government will spend roughly as much in the next two years on mature technologies like wind and solar as we've spent on fusion power over the entire history of the technology, but 1/3 of that was actually on weapons-oriented fusion research like the Inertial Confinement Facility (accounted in the DOE budget because a long time ago a decision was made to consolidate responsibility for all things nuclear in the DOE...even the bombs).

 

[woodgeek]

Anyone hear anything about the SkunkWorks-o-mak?

 

[iamlucky13]

This is the latest I know of, from this spring:
http://www.defensenews.com/story/de...nuclear-fusion-generator-investment/83870398/

There's a vague reference to having achieved initial plasma. I thought they had talked about proving they could contain plasma during their big PR event in 2014, but perhaps I misunderstood and they were referring to modelling containment in simulations.

Regardless, having plasma in the prototype to actually measure the behavior of is a positive step, if a very early one. I'm not seeing any word if they're externally heating it yet, or how long they sustain their plasma for. It's hard to read much into their statements so far.

I also haven't heard of them securing any government grants, which could help accelerate their work. However, the note that Lockheed has increased their funding suggests they haven't run into any major unexpected challenges.

 

[DBoon]

Storage doesn't have to be batteries. Load shifting through appropriate and common-sense time of use electric metering, peak energy storage in thermal form (heat or ice) can make use of renewable production peaks at a much lower cost than batteries, and in 20 years, electric battery storage has a much better chance of large scale commercialization at competitive prices than fusion energy has of achieving a sustained production fusion reaction. And a fusion reactor, with all it's complexity and requirements of scale, is a technology suited for the electric grid of 50 years ago - big, centralized base-load plants as opposed to more distributed, smaller-scale renewable generation with fast reacting peaker plants.

 

[peakbagger]

I was at a presentation once on fusion power and the speaker was speculating that the ultimate fusion power plants would be regional at least 5 gigawatts for them to be economic.

 

[iamlucky13]

DBoon, my opinion is that's a lot of assumptions upon which to base putting all our of our scheduleable energy needs in one basket based on storage. I could say a lot about the complications, but that's an extremely involved topic that I can only scratch the surface of.

For a few short notes: I'm just going to claim as a baseline to be disproven that after 20 years on the market, lithium ion battery systems are not going to drop another 90% in price. Thermal storage is better positioned for those cases where your input is heat (solar thermal), but not for PV and wind - converting electrons to heat and then going back through the ~40% efficient heat to electrons step is a non-starter even before you start looking into the salt quantities needed to store a region's worth of energy over night or through a week long production shortfall.

Load shifting will, I'm certain, be a significant part of our future energy management, but only a part of it. Distributed production is an ideal, not a reality. So far, big centralized plants remain the most effective baseload supply we have, and work well with the inter-regional transmission lines alternative energy proponents favor for helping deal with regional supply-demand variations by averaging between regions. Peakers are fossil fuel. I'm ok with that. Others are not.

Peakbagger, I don't think that will really be known until ITER has at least a couple years worth of testing under their belt to determine how well they can actually scale heat gain in a Tokamak with increased volume, field strength, and control. Notionally, they seem to be talking about 1 GW electrical. I'm unclear how much of that is based on the research so far at much smaller reactors vs. extrapolating experience fission plants to fusion.

And that result only applies to Tokamaks. Stellarators might be able to do better. Lockheed claims their high beta concept can do a lot better, even perhaps able to be economical down to 100 MW scales.

 

[woodgeek]

For a few short notes: I'm just going to claim as a baseline to be disproven that after 20 years on the market, lithium ion battery systems are not going to drop another 90% in price. Thermal storage is better positioned for those cases where your input is heat (solar thermal), but not for PV and wind - converting electrons to heat and then going back through the ~40% efficient heat to electrons step is a non-starter even before you start looking into the salt quantities needed to store a region's worth of energy over night or through a week long production shortfall.

RE Li-ion costs....

Studies suggest that many technologies follow a learning curve, so that price falls by some percentage every time cumulative production doubles. For large-format Li-ion since 2010, the rate seems to be a 21% drop for every doubling of production (similar to the value for PV, 25%).

http://rameznaam.com/2015/10/14/how-cheap-can-energy-storage-get/

So, if we figure we grow Li-ion production by 32X while building out the EV fleet, or five doublings, the price should drop to (1-0.21)^5 = 30% of the current price, or a 70% drop. This amount of production could be expected in the next 7-10 years, and would make long-range EVs cheaper than cheap ICE cars.

AFAIK no one has reported learning curve data for thermal storage (or if it obeys that behavior).

A 70% drop in Li-ion costs would get them down to where thermal grid storage is now, but in a much larger market (not relying on thermal inputs) and would be distributable.

In a coincidence, this morning I had a 2 hour blackout, and I backfed my house off my 2013 Nissan LEAF and a 1500W sine inverter, for the first time. Lithium is cool.

 

[DBoon]

I'm just going to claim as a baseline to be disproven that after 20 years on the market, lithium ion battery systems are not going to drop another 90% in price.

Except that I never said that Li Ion batteries are going to drop 90% in price, likely ever, and certainly not in the volumes we would need if they went utility-scale. There are other utility-scale battery technologies that could scale and at lower cost, but at what cost, I don't know. Regardless, I said that solar+wind with storage would be cheaper than fusion, and I further clarified what "storage" could be considered, e.g., off-peak home thermal storage (not molten salt utility-scale) for heating, better demand moderation on the grid, etc. I don't think electrical storage will ever be the most cost-effective way to shift demand or store excess, but it will have a place and costs will come down somewhat.

Thermal storage is better positioned for those cases where your input is heat (solar thermal), but not for PV and wind - converting electrons to heat and then going back through the ~40% efficient heat to electrons step is a non-starter even before you start looking into the salt quantities needed to store a region's worth of energy over night or through a week long production shortfall.

Again, I'm not looking at this from a utility/industry perspective with a view towards 100% centralized generation, transmission line transfer of power over long distances, and simple pricing structures of xx/kWh no matter how much and when you use it. There are so many common-sense things we could do to provide cleaner power at less cost that don't have to fit into the standard industry model of "we produce, you consume it."

I will say this - it would take a lot to convince me that a Fusion plant in 20 years could be produced for less than $5 billion for 1 maybe 2 GW capacity. At this cost, new nuclear plants are not cost-competitive. Why would we expect that Fusion plants would be? And if we don't, why would we keep investing anything in researching them? Perhaps there will be a military/space niche for fusion, but for commercial power production, it just seems a little too late and a little too much money.