14 ms·

I've always wondered: why exactly is ITER so expensive, and slow? Is the engineering required at such a standard that it should takes decades of planning and co

I've always wondered: why exactly is ITER so expensive, and slow? Is the engineering required at such a standard that it should takes decades of planning and construction and tens of billions of dollars? The timeline is so dilated (started in 1988, first plasma planned for 2025!) it feels like the kind of project that's expected to be cancelled from the start. It just doesn't strike me as obvious that reducing the major radius by a few meters would have such a huge impact on cost/timelines.

LeegleechN 5y ago
It's just an absolutely giant construction project. The mass and volume of the construction goes as the cube of the major radius of the reactor. Back of the envelope, SPARC is (6/1.5)^3 = 64 times smaller than ITER. The construction budget for SPARC is ~$500M, so ITER being tens of billions is in line.
- bigyikes 5y ago
  How do projects like LIGO ever get completed? I’m probably totally naive here, but I thought LIGO is physically larger and has many difficult constraints. The LHC comes to mind as well, and that absolutely dwarfs ITER in physical size. What’s the difference? Dealing with heat output? Superconductors are really hard maybe?
  - sbierwagen 5y ago
    LIGO did take decades to construct, like the LHC. According to LIGO's wikipedia article, it was the most expensive project ever funded by the NSF in 1994.
    - neltnerb 5y ago
      LIGO's size is also deceptive, the legs are kilometers long but the design is an L with the important bit being to vibrationally isolate things, maintain dimensional stability, and maintain vacuum. ITER is likely bigger in terms of volume of concrete or actual footprint.
      - lambdatronics 5y ago
        Hmm, looks like ITER and LHC both use about the same amount of superconductor: around 500 tons. [0]https://ieeexplore.ieee.org/document/1018583 https://ieeexplore.ieee.org/document/1018583 [1]https://www.iter.org/mach/Magnets https://www.iter.org/mach/Magnets
jvanderbot 5y ago
Margin? Akin's law of spacecraft #29 "To get an accurate estimate of final program requirements, multiply the initial time estimates by pi, and slide the decimal point on the cost estimates one place to the right."
dfdx 5y ago
Anecdotally, ITER was the largest of few options for a fusion researcher to run their experiment in a new tokamak. Everybody wanted to put their work into it, and as more features were added, the more funding it sucked up, leaving less money for other experiments, leading to more people wanting to put their experiment into ITER. Here's a presentation [0] that goes over why SPARC, being so much smaller and simpler than ITER could be more likely to succeed. [0] https://library.psfc.mit.edu/catalog/online_pubs/iap/iap2016/mumgaard.pdf https://library.psfc.mit.edu/catalog/online_pubs/iap/iap2016... This quote from the presentation summarizes it well: “The more money that's involved, the less risk people want to take. The less risk people want to take, the more they put into their designs, to make sure their subsystem is super-reliable. The more things they put in, the more expensive the project gets. The more expensive it gets, the more instruments the scientists want to add, because the cost is getting so high that they're afraid there won't be another opportunity later on- they figure this is the last train out of town. So little by little, the spacecraft becomes gilded. And you have these bad dreams about a spacecraft so bulky and so heavy it won't get off the ground- never mind the overblown cost.” “That boils down to the higher the cost, the more you want to protect your investment, so the more money you put into lowering your risk. It becomes a vicious cycle.” - Rob Manning, Chief spacecraft engineer, JPL
- maccam94 5y ago
  This is also why it's exciting to see the huge drop in satellite launch costs driven by SpaceX. Your satellite design is going to be much different if you have a few chances to launch on a $1B rocket per year, vs launching it for $10M anytime. Rather than one complicated reliable one, you might make 10 simpler ones and buy extra flights to do repairs if necessary.
- Gravityloss 5y ago
  For NASA science probes, this cost increase and risk averseness spiraling is called when missions become "Battlestar Galactica".
  - ksec 5y ago
    I thought this was some joke involving Bears and Beets. Turns out it is an actual thing [1]. >In 1992, Dan Goldin became the NASA Administrator. Goldin believed in a philosophy of Faster… better… cheaper—i.e., he thought NASA could do more with less. Hence, Goldin did not support the idea of having large EOS platforms in space and in fact once referred to them as “Battlestar Galactica.” He believed smaller, less expensive missions that could be built more quickly were the way to go and supported development of new programs that actually diverted funds from EOS. [1] https://eospso.nasa.gov/sites/default/files/eo_pdfs/Perspectives_EOS.pdf https://eospso.nasa.gov/sites/default/files/eo_pdfs/Perspect...
sbierwagen 5y ago
Here's a render of the completed reactor: https://www.iter.org/doc/all/content/com/gallery/media/7%20-%20technical/tkmandplant_2016_72dpi.jpg https://www.iter.org/doc/all/content/com/gallery/media/7%20-... Note human for size. It's all completely bespoke scientific equipment hand made for this project only. The cryostat will be the largest stainless steel vacuum vessel ever made-- all welded by hand. After welding, a substantial number of in-vessel components have to be installed by threading them through access ports, which is also quite a task: https://www.youtube.com/watch?v=pt70mO2nQac https://www.youtube.com/watch?v=pt70mO2nQac
- kwertyoowiyop 5y ago
  That will become a level in an FPS game for sure.
  - lambdatronics 5y ago
    Lol, I have the same though. They'll probably need to make the equipment a little sparser though, because it gets super crowded once everything is installed. Those cut-aways don't show all the detail.
- fabiospampinato 5y ago
  Here's a similar picture for SPARC from wikipedia: https://upload.wikimedia.org/wikipedia/commons/thumb/7/74/SPARC_2020.jpg/2560px-SPARC_2020.jpg https://upload.wikimedia.org/wikipedia/commons/thumb/7/74/SP...
  - pas 5y ago
    It still need a lot of the (same) support infrastructure. Plus ARC will be bigger than SPARC. That said building the first is a lot harder than scaling it up.
    - sbierwagen 5y ago
      I believe SPARC is a magnet demo, so it won't have tritium breeding blankets like ITER. It also has a shorter pulse time, 10 seconds vs 1000. (That's right. ITER, which will cost more than $65 billion and take decades to build, can't run continuously!)
orbital-decay 5y ago
Expensive because it's a custom built physics lab, not a commercial power plant. Slow because it's an international project. Not just that, but it also requires lots of infrastructure to be built and entire industries to develop in multiple countries, before it can be useful. ITER is massive, but it's also just a tip of the iceberg. >It just doesn't strike me as obvious that reducing the major radius by a few meters would have such a huge impact on cost/timelines. It would, easily. Past a certain size, production costs rise exponentially and require one-off tech.
- onlyrealcuzzo 5y ago
  Is it massive because it's 6 meters? Like - a 6 meter diamond would be "massive" - but a 6 meter boat isn't that impressive. Or is it massive like the tokamak is a 6 meter engine to a 100 km collider? Like there's a ton of other stuff being built in a massive structure?
  - tibyat 5y ago
    I agree. The responses in this thread are weak. Throw out any statement and claim it supports the known conclusion, how can you be wrong? "Well the volume (of a sphere) is proportional to the cube of its radius, so of course the cost of these physics experiments does as well" -- Why waste our time
  - dodobirdlord 5y ago
    A 6 meter superconducting magnetic containment vessel is massive, as such things go. But ITER is also an entire facility built around the containment vessel.
  - lambdatronics 5y ago
    The magnets are so large they had to build a factory on-site to put them together. If you can make the magnets small enough to transport from a centralized factory, that's a huge win.
- lumost 5y ago
  If production costs rise exponentially with reactor size then the exponential power gain with size isn’t very impressive.
thescriptkiddie 5y ago
I have a pet theory that the major cause for cost/time overruns on large projects is the cost constraint itself. In other words, attempting to do things on too small of a budget results in the overall cost increasing, not decreasing. I suspect that there are a few main mechanisms for this: 1. As budget constraints tighten, the number of man-hours spent wrestling with bean counters (and/or waiting around with nothing to do until the bean counter wrestling completes) increases exponentially. 2. "Cheap solutions" often end up being unfit for purpose, and have to be reworked later at great expense. 3. Budget overruns lead to time overruns which lead to more budget overruns, ad infinitum.
- TheSpiceIsLife 5y ago
  Also, it’s a multi-national government funded program that was never intended to return a profit. So there’s very little incentive to constrain costs.
- ncmncm 5y ago
  The budget overruns are the whole point of the project. Any sort of apparent science or engineering goal is a smokescreen.
  - IntrepidWorm 5y ago
    Citations? Or just idle theorizing?
    - ncmncm 5y ago
      It is the same process we see everywhere that massive cost overruns and unlimited delays manifest. Modern management methods can control cost and schedule where that is an actual goal. Where they don't demonstrates the actual goal. Every penny of cost overrun goes into some pocket. Smart management ensures many, many pocketholders are allied to keep the gravy train running.
      - craftinator 5y ago
ncmncm 5y ago
ITER is a victim of a process seen frequently in large public-works projects, but particularly those where no one understands what anything really should cost, and where there is no practical deadline for completion. It is worst when there is no expected utility when it is done. We see it lately in the numerous military procurements (particularly the F-35 program), in NASA's SLS rocket, in California's bullet train to nowhere, and urban tunnels such as New York's 2nd Avenue subway extension. It is why nuke plants are invariably so expensive and late. In a word, corruption. Lately, this corruption has been arranged to be wholly legal, so there is no possibility of prosecution. The majority of the money spent is funneled into myriad private pockets without moving the project toward completion. Nobody involved, at the monetary level, has any desire for it ever to be completed, because that is when the gravy train stops. Fusion projects represent the worst case of this phenomenon. Nobody knows what it should cost, and nobody in control of spending wants it over with, ever. The chance that anything of any practical use could come out at the end was openly foreclosed before it ever started: it was never promised to produce any electrical power, and no turbines, or space for any, appear in any site plan. Any sort of practically useful Tokamak plant would need to be overwhelmingly bigger and more expensive than ITER, and could never come anywhere near producing commercially competitive power, so the project is a known dead end, to be milked until it is finally cancelled in shame. What is tragic is that each euro diverted to this boondoggle brings climate disaster terrifyingly closer.
- stjohnswarts 5y ago
  This sounds like a whole lot of speculation to me. Can you cite actual sources and documents that prove it?
  - ncmncm 5y ago
    You write like someone unfamiliar with the history of the other projects cited. Copious materials are readily found online. Read up on them. Try to identify any reason to imagine ITER is different. I will wait.
- derac 5y ago
  Actually you can scale magnetic field stength or size to produce more power. SPARC (and later ARC) aims for the former with modern superconductor tech and is on track to potentially produce energy at Q~=11 by 2025.
sgt101 5y ago
Think cubically! A 6m device occupies 666 (say) --216 m^3 a 10m device occupies 10 10 6 (say) -- 600 m^3 The scale of volume means that you have to build a much bigger facility to put it in (in order for the electronics to be kept dry and for people to be able to get around it to keep birds off it and things. But worse - the weight. Concrete is 2400kg m ^3 so the small device might weigh 518 tonnes, but the bigger device is 1440 tonnes, so moving parts of it round becomes 3 * harder, the floors have to be 3* stronger, the supply chain has to be 3* better. And then time - 3* scale, 3* engineering challenge -> many times more time to deliver, many more $$$ -> risk -> planning -> admin... the less capital at risk the less it's worth spending on avoiding the risk.. the less the overhead of the project is. FWIW ITER is a science experiment - it's designed to find out more about fusion and that data will be very valuable for future reactor designs.

LeegleechN 5y ago

It's just an absolutely giant construction project. The mass and volume of the construction goes as the cube of the major radius of the reactor. Back of the envelope, SPARC is (6/1.5)^3 = 64 times smaller than ITER. The construction budget for SPARC is ~$500M, so ITER being tens of billions is in line.

bigyikes 5y ago
How do projects like LIGO ever get completed? I’m probably totally naive here, but I thought LIGO is physically larger and has many difficult constraints. The LHC comes to mind as well, and that absolutely dwarfs ITER in physical size. What’s the difference? Dealing with heat output? Superconductors are really hard maybe?
- sbierwagen 5y ago
  LIGO did take decades to construct, like the LHC. According to LIGO's wikipedia article, it was the most expensive project ever funded by the NSF in 1994.
  - neltnerb 5y ago
    LIGO's size is also deceptive, the legs are kilometers long but the design is an L with the important bit being to vibrationally isolate things, maintain dimensional stability, and maintain vacuum. ITER is likely bigger in terms of volume of concrete or actual footprint.
    - lambdatronics 5y ago
      Hmm, looks like ITER and LHC both use about the same amount of superconductor: around 500 tons. [0]https://ieeexplore.ieee.org/document/1018583 https://ieeexplore.ieee.org/document/1018583 [1]https://www.iter.org/mach/Magnets https://www.iter.org/mach/Magnets

jvanderbot 5y ago

Margin? Akin's law of spacecraft #29 "To get an accurate estimate of final program requirements, multiply the initial time estimates by pi, and slide the decimal point on the cost estimates one place to the right."

dfdx 5y ago

Anecdotally, ITER was the largest of few options for a fusion researcher to run their experiment in a new tokamak. Everybody wanted to put their work into it, and as more features were added, the more funding it sucked up, leaving less money for other experiments, leading to more people wanting to put their experiment into ITER. Here's a presentation [0] that goes over why SPARC, being so much smaller and simpler than ITER could be more likely to succeed. [0] https://library.psfc.mit.edu/catalog/online_pubs/iap/iap2016/mumgaard.pdf https://library.psfc.mit.edu/catalog/online_pubs/iap/iap2016... This quote from the presentation summarizes it well: “The more money that's involved, the less risk people want to take. The less risk people want to take, the more they put into their designs, to make sure their subsystem is super-reliable. The more things they put in, the more expensive the project gets. The more expensive it gets, the more instruments the scientists want to add, because the cost is getting so high that they're afraid there won't be another opportunity later on- they figure this is the last train out of town. So little by little, the spacecraft becomes gilded. And you have these bad dreams about a spacecraft so bulky and so heavy it won't get off the ground- never mind the overblown cost.” “That boils down to the higher the cost, the more you want to protect your investment, so the more money you put into lowering your risk. It becomes a vicious cycle.” - Rob Manning, Chief spacecraft engineer, JPL

maccam94 5y ago
This is also why it's exciting to see the huge drop in satellite launch costs driven by SpaceX. Your satellite design is going to be much different if you have a few chances to launch on a $1B rocket per year, vs launching it for $10M anytime. Rather than one complicated reliable one, you might make 10 simpler ones and buy extra flights to do repairs if necessary.
Gravityloss 5y ago
For NASA science probes, this cost increase and risk averseness spiraling is called when missions become "Battlestar Galactica".
- ksec 5y ago
  I thought this was some joke involving Bears and Beets. Turns out it is an actual thing [1]. >In 1992, Dan Goldin became the NASA Administrator. Goldin believed in a philosophy of Faster… better… cheaper—i.e., he thought NASA could do more with less. Hence, Goldin did not support the idea of having large EOS platforms in space and in fact once referred to them as “Battlestar Galactica.” He believed smaller, less expensive missions that could be built more quickly were the way to go and supported development of new programs that actually diverted funds from EOS. [1] https://eospso.nasa.gov/sites/default/files/eo_pdfs/Perspectives_EOS.pdf https://eospso.nasa.gov/sites/default/files/eo_pdfs/Perspect...

sbierwagen 5y ago

Here's a render of the completed reactor: https://www.iter.org/doc/all/content/com/gallery/media/7%20-%20technical/tkmandplant_2016_72dpi.jpg https://www.iter.org/doc/all/content/com/gallery/media/7%20-... Note human for size. It's all completely bespoke scientific equipment hand made for this project only. The cryostat will be the largest stainless steel vacuum vessel ever made-- all welded by hand. After welding, a substantial number of in-vessel components have to be installed by threading them through access ports, which is also quite a task: https://www.youtube.com/watch?v=pt70mO2nQac https://www.youtube.com/watch?v=pt70mO2nQac

kwertyoowiyop 5y ago
That will become a level in an FPS game for sure.
- lambdatronics 5y ago
  Lol, I have the same though. They'll probably need to make the equipment a little sparser though, because it gets super crowded once everything is installed. Those cut-aways don't show all the detail.
fabiospampinato 5y ago
Here's a similar picture for SPARC from wikipedia: https://upload.wikimedia.org/wikipedia/commons/thumb/7/74/SPARC_2020.jpg/2560px-SPARC_2020.jpg https://upload.wikimedia.org/wikipedia/commons/thumb/7/74/SP...
- pas 5y ago
  It still need a lot of the (same) support infrastructure. Plus ARC will be bigger than SPARC. That said building the first is a lot harder than scaling it up.
  - sbierwagen 5y ago
    I believe SPARC is a magnet demo, so it won't have tritium breeding blankets like ITER. It also has a shorter pulse time, 10 seconds vs 1000. (That's right. ITER, which will cost more than $65 billion and take decades to build, can't run continuously!)

orbital-decay 5y ago

Expensive because it's a custom built physics lab, not a commercial power plant. Slow because it's an international project. Not just that, but it also requires lots of infrastructure to be built and entire industries to develop in multiple countries, before it can be useful. ITER is massive, but it's also just a tip of the iceberg. >It just doesn't strike me as obvious that reducing the major radius by a few meters would have such a huge impact on cost/timelines. It would, easily. Past a certain size, production costs rise exponentially and require one-off tech.

onlyrealcuzzo 5y ago
Is it massive because it's 6 meters? Like - a 6 meter diamond would be "massive" - but a 6 meter boat isn't that impressive. Or is it massive like the tokamak is a 6 meter engine to a 100 km collider? Like there's a ton of other stuff being built in a massive structure?
- tibyat 5y ago
  I agree. The responses in this thread are weak. Throw out any statement and claim it supports the known conclusion, how can you be wrong? "Well the volume (of a sphere) is proportional to the cube of its radius, so of course the cost of these physics experiments does as well" -- Why waste our time
- dodobirdlord 5y ago
  A 6 meter superconducting magnetic containment vessel is massive, as such things go. But ITER is also an entire facility built around the containment vessel.
- lambdatronics 5y ago
  The magnets are so large they had to build a factory on-site to put them together. If you can make the magnets small enough to transport from a centralized factory, that's a huge win.
lumost 5y ago
If production costs rise exponentially with reactor size then the exponential power gain with size isn’t very impressive.

thescriptkiddie 5y ago

I have a pet theory that the major cause for cost/time overruns on large projects is the cost constraint itself. In other words, attempting to do things on too small of a budget results in the overall cost increasing, not decreasing. I suspect that there are a few main mechanisms for this: 1. As budget constraints tighten, the number of man-hours spent wrestling with bean counters (and/or waiting around with nothing to do until the bean counter wrestling completes) increases exponentially. 2. "Cheap solutions" often end up being unfit for purpose, and have to be reworked later at great expense. 3. Budget overruns lead to time overruns which lead to more budget overruns, ad infinitum.

TheSpiceIsLife 5y ago
Also, it’s a multi-national government funded program that was never intended to return a profit. So there’s very little incentive to constrain costs.
ncmncm 5y ago
The budget overruns are the whole point of the project. Any sort of apparent science or engineering goal is a smokescreen.
- IntrepidWorm 5y ago
  Citations? Or just idle theorizing?
  - ncmncm 5y ago
    It is the same process we see everywhere that massive cost overruns and unlimited delays manifest. Modern management methods can control cost and schedule where that is an actual goal. Where they don't demonstrates the actual goal. Every penny of cost overrun goes into some pocket. Smart management ensures many, many pocketholders are allied to keep the gravy train running.
    - craftinator 5y ago

ncmncm 5y ago

ITER is a victim of a process seen frequently in large public-works projects, but particularly those where no one understands what anything really should cost, and where there is no practical deadline for completion. It is worst when there is no expected utility when it is done. We see it lately in the numerous military procurements (particularly the F-35 program), in NASA's SLS rocket, in California's bullet train to nowhere, and urban tunnels such as New York's 2nd Avenue subway extension. It is why nuke plants are invariably so expensive and late. In a word, corruption. Lately, this corruption has been arranged to be wholly legal, so there is no possibility of prosecution. The majority of the money spent is funneled into myriad private pockets without moving the project toward completion. Nobody involved, at the monetary level, has any desire for it ever to be completed, because that is when the gravy train stops. Fusion projects represent the worst case of this phenomenon. Nobody knows what it should cost, and nobody in control of spending wants it over with, ever. The chance that anything of any practical use could come out at the end was openly foreclosed before it ever started: it was never promised to produce any electrical power, and no turbines, or space for any, appear in any site plan. Any sort of practically useful Tokamak plant would need to be overwhelmingly bigger and more expensive than ITER, and could never come anywhere near producing commercially competitive power, so the project is a known dead end, to be milked until it is finally cancelled in shame. What is tragic is that each euro diverted to this boondoggle brings climate disaster terrifyingly closer.

stjohnswarts 5y ago
This sounds like a whole lot of speculation to me. Can you cite actual sources and documents that prove it?
- ncmncm 5y ago
  You write like someone unfamiliar with the history of the other projects cited. Copious materials are readily found online. Read up on them. Try to identify any reason to imagine ITER is different. I will wait.
derac 5y ago
Actually you can scale magnetic field stength or size to produce more power. SPARC (and later ARC) aims for the former with modern superconductor tech and is on track to potentially produce energy at Q~=11 by 2025.

sgt101 5y ago

Think cubically! A 6m device occupies 666 (say) --216 m^3 a 10m device occupies 10 10 6 (say) -- 600 m^3 The scale of volume means that you have to build a much bigger facility to put it in (in order for the electronics to be kept dry and for people to be able to get around it to keep birds off it and things. But worse - the weight. Concrete is 2400kg m ^3 so the small device might weigh 518 tonnes, but the bigger device is 1440 tonnes, so moving parts of it round becomes 3 * harder, the floors have to be 3* stronger, the supply chain has to be 3* better. And then time - 3* scale, 3* engineering challenge -> many times more time to deliver, many more $$$ -> risk -> planning -> admin... the less capital at risk the less it's worth spending on avoiding the risk.. the less the overhead of the project is. FWIW ITER is a science experiment - it's designed to find out more about fusion and that data will be very valuable for future reactor designs.