Google Solve for X Retreat

That is brave choice for 'X'. Bravo.

first session was on Water scarcity:

In 18 years, we will have 40% less water than we need.
To produce 1MWh of energy from coal, you need 50,000 gallons of water. From a nuclear plant: 60,000 gallons, from biofuels: 67,000 gallons, although some argue that is improving.
Solutions described using forward osmosis with recycled ammonium salts, the simple ones that are used in gingerbread cookies, and powered by low-grade waste heat.

Neal Stephenson responds (He's a fascinating fellow, who inspired some photo captions in years past — Snow Crash and Diamond Age):

P^2 - Paul (4 months ago | reply)

1MW of a coal plant costs about $1,000,000 (i.e., 1000MW plant is about a gigabuck)
50,000 gallons of water costs me $600 (at my utility's extortionist rate). I'll bet Con Ed or PG&E can get it for much less.
The water cost is a drop in the bucket compared to the cost of the plant. What point are they trying to get across here? That water is too cheap?

flytrue2 (4 months ago | reply)

If you generate nuclear power with a Liquid-Fluoride Thorium Reactor, (LFTR), water consumption is negligible. Other benefits include dramatic gains in efficiency over uranium fuels as well as significantly less harmful waste, relatively inexpensive plant construction, and neither this fuel nor it's byproducts readily lend themselves to use in nuclear weaponry. More information at energyfromthorium.com

solerena (4 months ago | reply)

Intriguing... designing for ... superstition and religion...Love solve for X part:):)

jurvetson (4 months ago | reply)

- yes - go new nukes!

- The key issue is the unit of time. It's not one-time water use in construction, but ongoing gallons/hr in operation.

And the numbers are all over the map out there:

"Power generation has been estimated to be second only to agriculture in being the largest domestic user of water. To produce and burn the 1 billion tons of coal America uses each year, the mining and utility industries withdraw 55 trillion to 75 trillion gallons of water annually, according to the US Geological Survey." (source), with details on steam, cooling, processing, and mining).

Note that is water drawn, and it could be seawater in some cases. If you look at net freshwater consumed, you get much lower numbers (wikipedia)

vennettaj (4 months ago | reply)

funny the break before the wine tasting reception..it's all about the wine, isn't it :p

P^2 - Paul (4 months ago | reply)

This doesn't add up. Scaling that a bit to understandable numbers:65 trillion gallons = 246 billion tons.

To burn a ton of coal requires 246 tons of water? That does not pass the sniff test. Even if you count the water within the steam loop in the plant, which is ultra-pure and 100% recycled within the plant. Even if you count the water used in the cooling towers, which is 100% recycled (returned to the source a bit warmer, but unpolluted and with some coming back as nicely distilled rain). Even if you add in the water sprayed in the mines to keep the dust down and to wash the trucks, it still doesn't pass the sniff test.

I love the idea of getting coal off the energy buffet -- it's stinky, dirty and radioactive, and is much better reserved for use as a chemical feedstock. But Sourcewatch and others are discrediting themselves and doing a disservice to the whole effort throwing around numbers like that. It's sensationalist and irresponsible.

P^2 - Paul (4 months ago | reply)

I see the revisionist edit from MW to MWh. OK, that looks like it makes sense. Let's see: 1 MWh of electricity from coal, at wholesale prices, costs about $25 (retail, say $150). And it "needs" 50,000 gallons ($600 retail) of water. Again, somebody has to be cooking numbers somewhere, or that water is just going round-and-round, and they're counting it over and over.

My car's engine produces 100 kW of power, and it needs 5000 gallons per hour of coolant to do it. But I only count it as the 2 gallons I put in it.

jurvetson (4 months ago | reply)

Yup, and that's why I commented on the difference between drawn and consumed. You may like that wikipedia link more.

on to other topics... so much to share... I just had a fascinating brain scientist who has lost her pituitary gland altogether to a tumor. The hormone regimen is entirely external and oh, boy, does it question your sense of what it means to be you. She boosted testosterone as an experiment - aggression, and recurring thoughts of sex were among the effects. She gained and then lost 100 lbs of weight during one set of experiments, with no change in diet or exercise.

Her presentation: reading the mind. From many fMRI scans, where seeing and imaging something is fundamentally similar in brain blood flow effects, they can create a "mental map" of correlates with elements of images seen. From this composite, and a bit of machine learning, they can then present a new image, and from the fMRI data alone recreate a fairly spooky rendition of what is being seen.

I would love to run the tape while dreaming.

This mind-reading requires a clunky fMRI today, but others are working on a radically different way to export the image: running neurons in reverse... and expressing the image on the retina to be read externally.

This led the guy next to me to say "guys are screwed." And that's what led her so share her hormonal cocktail experiment story.

We sensed a TED talk in the making.

seatonsnet (4 months ago | reply)

Wow! "What it means to be you"... Probably the central philosophical and spiritual question of all. I would really like to hear more about this lady and her experiments.

Pieces of Eye (4 months ago | reply)

What held back thorium from being used in new reactors?

obskura (4 months ago | reply)

that sounds interesting...
My "X" would probably be large-scale energy storage, which would allow to go 100% renewable from intermittent sources. My preferred solution to this would be decentralised mechanical systems, e.g. high-energy composite flywheels running magnetically suspended in a vacuum, or maybe compressed air (although there are probably thermodynamic losses). I think the best way to deploy is to extend the concept of feed-in tariff from solar/wind production to clean storage as well, or to allow individual households to trade electricity at market price. Surely the grid operators wouldn't like it much though... but a flywheel in every second household would create a pretty robust system.

vennettaj (4 months ago | reply)

SJ, forget the mind reading..did you say she lost 100 lbs without a diet..oh, i hope you wrote down everything

energy..i'm for the gym powered buildings
.

flytrue2 (4 months ago | reply)

surprisingly, nothing at all. In the 1950's when the possibility of using fission as a source of electricity was first being considered, most of the existing knowledge was an outgrowth of weapons programs. The government funded several alternatives including a working thorium based reactor at Oak Ridge laboratories but when the chief advocate for thorium based reactors passed away, simple politics determined the projects which received funding. Despite all of it's benefits thorium fell by the wayside.

For anyone interested in clean energy or energy independence from fossil fuels the first 5 minutes of this video contrasts todays technology with thorium technology. The rest of the video is an in-depth examination of both technologies and how we got to where we are today. This video as well as others are also available at www.energyfromthorium.com

My interest in this technology is entirely environmental. Apart from living on this planet I hold no personal stake in this technology of any kind.

nhr (4 months ago | reply)

> To burn a ton of coal requires 246 tons of water?
> That does not pass the sniff test.

The value actually seems reasonable.

A typical coal plant has a theromdynamic efficiency of about 30%.
So, a 1GW (electric) power plant needs a power input of about 1GW/.3 ~= 3.3GW (thermal)

Energy density of 1kg of coal: about 6.67 KWh
The energy content of a ton of coal is thus about 6.67 MWh, of which 30% — i.e. about 2 MWh — leaves the plant as electrical power, and the rest — i.e. about 4.47 MWh — must be evacuated by the cooling circuit.

4.47 MWh ~= 16.1 Gigajoules ~= 3.844 Gigacalories

246 tons of water = 2.46 10^8 grams of water.

Energy required to raise the temperature of 1 gram of water by 1 degree Celsius: 1 calorie.

With 3.844 Gcal, one can thus theoretically raise the temperature of 246 tons of water by 15.63 degrees Celsius.

In practice, if cooling towers with sprinklers are used, the enthalpy of evaporation ensures that a significant part of the heat load will be carried away by water vapor. Thus, the liquid water's temperature will change by much less than 15.63 degrees. This is a good thing, as environmental regulations typically limit the temperature difference between the cooling water inlet and outlet of a power plant using a natural body of water (river etc.) to a few degrees Celsius.

Hence the need to use a lot of water so as to minimize the heat load and impact on the surrounding river's biotope, and the reason why power plants are generally sited near an abundant source of cooling water.

> What held back thorium from being used in new reactors?

A nuclear power plant, like a coal power plant, works by generating heat. That heat is typically used to boil water, and the resulting high-pressure steam is used to drive the turbines to which the generators are attached.

A conventional nuclear reactor uses solid pellets of uranium (and some plutonium), assembled in fuel rods.
OTOH, a thorium reactor would have a "core" that is liquid — molten lithium fluoride and beryllium salts carrying thorium and uranium.

A decent-sized — e.g. 1GWe — thorium reactor would need to circulate several tons of molten salt per second to the heat exchanger, where the steam driving the turbines would be generated.

Circulating several tons per second of potentially reactive metal salts, heated to about 400 degrees, is a serious engineering challenge.

Fast breeder reactors, which typically used molten sodium as the reactor core's coolant, present similar difficulties. Water slows down the neutrons too much, which hinders the operation of a breeding reactor, hence the selection of sodium — with its low neutron absorption cross-section — as the primary circuit's coolant.

The engineering reality is that it's horrendously difficult to create a cooling circuit (including pumps and pipings) able to withstand for years the circulation of several tons per second of molten sodium metal.

I'm afraid there hasn't been a single FBR in the world that didn't experience leaks in its sodium circuit.

Once leaks happen, sodium is much more complicated to deal with than the water used in normal nuclear reactors, for it reacts explosively with water. And unfortunately, some water is generally present near the sodium in a FBR used to generate power: how else would the steam used to drive the turbines be generated ?

These difficulties explain why the number of actually operational FBRs in the world, even after decades of development work, is exactly zero.

High-temperature lithium molten salt circuits carrying thorium and uranium, compared with molten sodium, would be even more "challenging", especially if a leak happens, as that molten salt would be transporting very radioactive materials, including the nasty actinides generated by the fuel's fission reactions.

P^2 - Paul (4 months ago | reply)

All true and correct (I had done the same calculation), but it is disingenuous of organizations like Sourcewatch to imply that 246 tons is used to burn a ton of coal. The vast majority (except for the tiny fraction that evaporates in the cooling tower, if one is used) is returned to the source, or re-used entirely within the plant. It's not "used" and removed from inventory in the sense that the coal is. They do a disservice implying that it is.

Steve Kirsch (4 months ago | reply)

It would be nice if we all stuck to the facts.

The BN-600 is a fast reactor that has been in operation for more than 30 years. It is so trouble free that experts like "nhr" above didn't even knew it existed, despite being mentioned on the wikipedia page that nhr used to "prove" his erroneous assertion that there are zero FBRs in operation.

Furthermore, the EBR-II ran for 30 years without any significant mishaps even after running chernobyl and TMI simulations. When they disassembled it, they found that the materials were as "clean" as they were on day 1 with no corrosion.

The problem with FBRs is really the misinformation being spread about them, not the technology itself.

Steve Kirsch (4 months ago | reply)

...and the BN-600 is STILL in operation AFAIK.

They extended the license for another 10 years recently, so it will have been operating so trouble free for nearly 40 years that nobody knew about it.

See BN-600 fast reactor to operate until 2020

Steve Kirsch (4 months ago | reply)

And the only reason we shut down the FBR in the US is for poltical reasons, not for technical reasons. See Steve Kirsch's letter to Heather Zichal regarding the Integral Fast Reactor (IFR) for an extensive writeup as to why it was cancelled and why we aren't restarting it.

The reason we aren't restarting it is nobody is putting it on the priority list because nobody knows the real facts about it and when you send the White House stuff like my memo that explains the facts, nobody reads it because they think if it were true than more people beyond Bill Gates, myself, our nation's most qualified fast nuclear scientists, GE, the President of the American Nuclear Society, Congressman John Garamendi, James Hansen, Mark Lynas, Stewart Brand, Barry Brook, Tom Blees, etc. would be supporting it.

So it's a chicken-egg problem. If you ask Al Gore why he isn't supporting it, he'll say "I'm technology neutral." If you ask the Sierra Club, they'll say, "we are not opposed to it."

nhr (4 months ago | reply)

> The BN-600 is a fast reactor that has been in operation for more than 30 years
> It is so trouble free that experts like "nhr" above didn't even knew it exist

My bad.
The BN-600 had, indeed, escaped my attention.
Still, one operational reactor, compared with the total deployed number of more conventional nuclear reactors, be it in Russia or elsewhere, is hardly indicative that circulating liquid sodium or radioactive molten salts is seen as cost-effective or safe or advantageous by most decision-makers in the nuclear power industry, be it in Russia, the US or elsewhere.

Note that I'm just a disinterested observer, and I don't have a dog in this fight.
If the challenges I outlined in my previous comment can be shown to have found effective solutions, then I'm sure the support base for these fuel-efficient reactor technologies would grow quickly.

The number of cycles the water goes through the cooling circuit is variable, but not a factor in the flow calculation: regardless of the number of cycles, at some point in time, the water will be too hot to be useful for cooling the reactor.
At that point, the hot water must be replaced with cold water.

The power plant, considered as a thermodynamic entity, cannot maintain its temperature — i.e. prevent its entropy from rising — if it can't discharge heat to the external environment at the rate its reactor produces it.

The only way to discharge the heat is by pumping in cold water so that the rate of heat production and the rate of negative calories — a.k.a. frigories — provided by the cold water are in equilibrium, and this is the principle underlying my calculation.

To make an astronautical analogy, heat can be considered a pollutant making water unfit for use in a cooling system, just like the carbon dioxide exhaled by human respiration is a pollutant as CO2 can, at high enough concentrations, become toxic and make the air unfit for respiration.

Hence the need for CO2-scrubbing equipment in environments where air must be reused, like in the Apollo spaceships.

For a power plant, the "scrubbing system" acting as a heatsink and providing the power plant with "scrubbed" water from which the "heat" pollutant has been removed is the Earth's geophysical processes.

Steve Kirsch (4 months ago | reply)

nhr: The soviets consider the BN-600 one of their best performing reactors.

The reason there is only one fast reactor is because it is HARD TO GET RIGHT.

It took our smartest people DECADES to figure it out. Nobody in the world has better technology than we do. So after developing this amazing technology and solving all the hard problems to get it to work totally reliably for 30 years, the government said, after 30 years of bi-partisan support of the effort, that we NO LONGER NEED CLEAN ADVANCED NUCLEAR POWER. That is the justification Clinton used to cancel it, not because it wasn't cost effective or didn't work.

General Electric believes IFRs are very cost effective. So do the scientists who designed it.

Of course, some people don't believe that. These are all people who do NOT have first hand experience with fast reactors. I don't know of anyone who worked on the IFR who says it is not cost effective. Do you?

The only way to resolve it is to build them and ride down the cost curve.

Bill Gates wouldn't be investing the money he's investing in his fast reactor (an IFR derivative) if he didn't believe these reactors are cost effective. He's very very smart.

Had we not cancelled the IFR program (which was cancelled for political reasons), we would be far down the learning curve.

Our technology is much more cost effective than the Soviet reactor because we use metal fuel.

And we should never assume the stuff we invented 50 years ago can't be cost reduced even more. That's like saying computers built 50 years ago can't be made cheaper and most cost effectively. Once these reactors are a national priority, it is amazing what people will think of.

nhr (4 months ago | reply)

[ Irony alert ]

To summarize:

1. The best way forward as far as Gen.4 nuclear reactors are concerned, according e.g. to the scientists who worked on the development of the IFR is, well, the IFR.
Let me point out that the IFR team specifically selected a sodium "pool" design, instead of the sodium "loop" design.
In a "pool" design, the reactor core is immersed in a pool of molten sodium, which is, in turn, cooled by a secondary sodium circuit (or "loop") connected to the heat exchanger / steam generator.
The envisioned fuel, once enriched by the IFR, is reprocessed on-site e.g. by electrorefining and then re-used – hence the name, "Integrated" Fast Reactor.
2. Bill Gates invested in TerraPower, which specifically choose to go with a sodium "loop" design, instead of the pool design. TerraPower's envisioned fuel is also "once-through", without reprocessing.
3. Thorium reactor proponents argue that instead of the conventional, solid U or Pu nuclear fuels, a circuit of high-temperature molten salts carrying radioactive isotopes should be preferred. The fuel in the circuit presumably needs periodic reprocessing to remove the "nuclear poisons" — i.e. the fission by-products that would absorb the neutrons and impede the Th to U fertilization process.

So, even if we ignore the implicit choices made by most decision makers in the overwhelming majority of the production nuclear reactors across the globe, we have the aforementioned three groups of presumably smart people who already have divergent opinions as to what the preferred design of a safe and cost-effective nuclear reactor should be.

About the only commonality these designs have is that they don't use an established technology — water coooling circuit, — but instead rely on ideas that have, er, a "proven" worldwide track record of reliability and accumulated engineering experience like circuits of liquid sodium or radioactive molten salts.

We also have various unspoken assumptions, e.g.:

• Leaks in the sodium or radioactive molten salt circuit would never happen; that's a mere engineering and manufacturing and plant monitoring and operational and maintenance problem that's been solved.
• Loss of electrical power leading to the shutting down of coolant pumps is something that would never happen.
• In the IFR, natural convection in the pool is supposed to be sufficient to cool the reactor core; sodium is kept at a low pressure so as to minimize the risk of leaks and keep vessel and piping costs low, but said sodium would never boil (trust us). The fact that the IFR has a positive void coefficient, leading to increased heat output from the nuclear fuel if bubbles were to form in the sodium can be, well, ignored.
• The IFR's secondary sodium cooling circuit is presumably an add-on, as the IFR sodium pool is presumably capable of eliminating heat autonomously so that internal convection takes place, perhaps by externally radiating tens of megawatts of heat away if the secondary circuit is shut down e.g. due to a leak or a loss of electrical power. It must be quite toasty near the IFR's primary sodium pool.
• Even if leaks e.g. to a Th+U+actinides molten salt circuit were to happen, decontamination of the radiation caused by nuclear fuel is something that's proven to be, er, quick and inexpensive.
• Even if radioactive contamination due to leaks were to happen, it should, er, never lead to an event such that it would make more economic sense to shut down, write off and decomission the plant.

Note also that:

• General Electric etc. develop and market reactors which they'd obviously describe as "cost-effective". GE etc. also stand to benefit from the awarding of nuclear reactor procurement contracts; they "privatize" the profits they earn on such engineering and construction projects. Any financial losses, be they caused by the unexpectedly low operational efficiency of the delivered "fuel-efficient & cost-effective" power plant due e.g. to technical "glitches" while the manufacturer "accumulates experience and rides down the cost curve", or the possible writing off of the not unsubstantial sunken costs of a contaminated power plant can be "socialized" and aren't GE's problem.
• For TerraPower, it's interesting that enlightened countries like China, whose government has an excellent record for transparency and social accountability, happen to be more interested in their reactors than e.g. the more socially backwards countries of the West.

vennettaj (4 months ago | reply)

wha.. cool chat..
(of which i understood not much, but the argument that bill gates is very smart and that's why the reactor must be good cracked me up..just as a choice of argument..not saying he's not smart :p)

PhOtOnQuAnTiQuE (4 months ago | reply)

"Richard DeVaul The #solveforx conference videos and other materials should be posted by the end of the day today. wesolveforx.com ".

Can' t wait to see them =)
Thanks for sharing Steve

jurvetson (4 months ago | reply)

some anticipatory buzz at Mashable

I also added some photos above. Here are the Happy Google organizers at the close:

Astro Teller, in the middle, did the opening for my tech talk.

.McCarthy (4 months ago | reply)

love to see that you are using Flickr as a forum

jurvetson (4 months ago | reply)

thanks. The site just went live with videos of some of the talks. www.wesolveforx.com

I suggest starting with Mary Lou Jepsen's talk on reading the mind (HD version).

And I added a new post with most of my notes here

traik (4 months ago | reply)

The most efficent method of addressing carbon foot prints is reducing the size of our global population. So an advocate can contribute by not having kids.

jeany777 (4 months ago | reply)

Started with Mary Lou Jepsen's talk and the brain astonishing results based on imagination

Pieces of Eye (4 months ago | reply)

When are they going to post your talk?

jurvetson (4 months ago | reply)

I don't think they will, as I was part of a couple of the responses to the talks, and I think that was off the record.

But you can see me discuss my X ad nauseum in my Google Tech Talk

jgury (3 months ago | reply)

twitter.com/#!/erick_miller/statuses/15951113 2548243456
"The vector of change, of disruptive change, always comes from new entrants, never from incumbents" -Steve Jurvetson He did not get the quote exactly right. at 6:50 in that talk. "The new, the important, the vector of change in the world, the vector of disruptive change, always comes from new entrants, never from incumbents."
For me as a math type when I hear vector I think specifics in direction and magnitude, A very specific object. All change, disruptive or otherwise, I view as much more well understood as stochastic variables in tensor spaces. That works from the subatomic up to the cosmic levels as well and even for things like big evolutionary events such as extinction epochs. Your specific company in 20 years bet question is roughly the same as doing an evolutionary organism question or a technology forecast. In other words, I would not really bet you 20 years from now what the dominant organism on this planet will be much less some stupid business the way change is progressing. Now if you want to think of change in vector epidemiology terms that is ok but it really does not give you much insight about really big historical disruptive technologies like gunpowder, nuclear weapons, and rockets. It also does not give any significant socio political insights. Indeed, Nazis are certainly disruptive non incumbent change vectors alright, especially when they get well ordered in higher German tensor space. They are also a basic counterexample to the absurd Sam Harris "some propositions are so dangerous that it may even be ethical to kill people for believing them." Have not made it to the superstion, relgion and evolution parts but the working statement is still the "all religon is voodoo" equality. That is not IFF or exclusive. So it is better stated as there are elements of voodoo in all the words religions, "what these natives are practicing is only what Father Rodrigo has taught them. Isn't that right Father? Father, Uhh, he seems a bit ill...."

jgury (3 months ago | reply)

What it means to NOT be you is much better. What it means to be MU. One of the best answers in all philosophy: 無