Both. New turbine designs are more efficient, partly because new turbine blades are made of better materials. 50 years is a lo----ng time in turbine development: compare modern jet engines to jet engines from 1969.
I mean, fair enough, but a nuclear reactor in principle means you don’t have to eke out every last drop of efficiency.
Yes, in practice, current in use designs burn quite a bit of expensive uranium fuel and create dangerous waste that has almost all of the energy still remaining in it. A good power reactor design would involve a core that never needs to be refueled - once it runs out of usable fuel in 50 years, you’d just let the core cool, immersed in water for a decade, and then store the still assembled core in in a dry cask underground.
I thought there were some designs in principle that could work this way.
Advances in gas turbines has been mainly about high pressure turbine blades that allow a higher turbine inlet temperature, hence higher pressure ratio (from intake air to combustion) which is the key driver of their thermal efficiency, which is central to their competitiveness. That goes not only for a/c engines but even more so in the more relevant direct competitor to nuclear power plants: gas fired gas turbine combined cycle electric power plants. Among the reasons nuclear power is now economically noncompetitive in US conditions is improvement in the cycle efficiency of combined cycle GT plant in the last few decades, though cheaper nat gas is also a huge reason, combined with lack of demonstrated progress on overcoming the huge financial risk (not what construction costs are ‘supposed’ to be in white papers, but the risk of what they can end up being in reality).
Advances in turbine materials are irrelevant to nuclear plants, which operate on saturated steam (ie not ‘superheated’ beyond the boiling point at the reactor’s secondary loop pressure) at temperatures which allow non-exotic steel alloys. Advances in turbine blade geometry and design (computational tools etc) apply but are a very small factor in the overall cost equation for nuclear. Improvements in reactors would have a bigger impact. But, fuel (as opposed to fuel waste disposal) cost is something like 1/5 of total nuclear operating costs (not counting the huge capital cost which must in theory be offset by a lower absolute operating cost) compared something like 3/4 for fossil plants.
The key to competitive nuclear is lowering (or just controlling) capital and non-fuel operating costs with an extremely high level of operating safety. Which has been the key failure of nuclear in the US. The latest round of upfront cost blow outs give IMO little reasonable hope for nuclear as a major power source in US conditions, wishes to the contrary.
Nuclear plants are upgraded constantly over their life with modern technology.
A very common upgrade is a power uprate. Better tech allows making more electricity (thus more money) out of the same reactor. Efficiency improvements of a few % also happen usually.
Here is an interesting PDF with nice pictures describing a steam turbine modernization in a nuclear power plant:
Turbine Replacement with 54-inch Blade Low Pressure Turbine at KRSKO Nuclear Power Plant (PDF link)
Nuclear is expensive, but even if you assume 10 billion for every new plant, we’d need about 300 new plants to make the electricity grid 100% non-polluting (we’d be at 80% nuclear and 20% renewables). That’d cost about 3 trillion dollars. A lot of money, but still only about 15% of GDP (during WW2 we were spending 40% of GDP every year on the military, so spending 15% of GDP for one year isn’t impossible).
Anyway, other than nuclear what else can provide baseload power and doesn’t contribute to climate change? I like solar, but until battery technology comes along it can’t really be used as baseload power. Not only that but are there enough raw materials on earth to make solar panels as a major source of energy? Solar keeps getting cheaper, but what if solar production goes up by a factor of 20, do we have the raw materials to make enough panels to keep up with demand?
Also China is investing heavily in nuclear power. I believe they want to be at 120-150 GW of nuclear capacity by 2030.
Yes, solar panels require “rare earth elements”, and yes, those are currently produced mostly in China. But the term “rare earth” is actually a misnomer: Those elements are actually reasonably common; they’re just not very concentrated, so it’s tougher to mine them. So they tend to be mined where labor costs are cheap, but if needed, they could be mined almost anywhere.
And of course, those rare earths make up a relatively small percentage of a solar panel: By far the largest part of them is silicon, which is super-abundant and easy to mine everywhere.
-
The US has around 1100 GW of installed electric power capacity. The only example of nuclear plants to (possibly) be completed lately are Vogtle units 3 and 4, estimated $25 billion for 2*1.1 GW units. So it would be more like 800 units of that size at a total of ~$10 trillion given the recent actual cost example. Also if considering to counter that there’s a learning curve and future costs won’t be that high, could be, but also consider that Vogtle and the completely failed Sumner project ($10 bil spent to get zero GW) were in a way low hanging fruit along the lines of the OP question. Building new plants on the grounds of already operating ones as both those projects did cuts down a lot of cost and particularly risk of failing to get permitted. Most plants in a huge expansion of nuclear could not benefit from that, and it’s actually questionable if that would ever be possible, socio-politically besides economically. Society and politics is a reality just like engineering and economics.
-
But the assumption here is the standard IMO very questionable one that climate change is going to be addressed by big reductions in CO2 emissions with technology more or less like today’s. I think that’s really unlikely. It will be some combination of real breakthroughs in low cost low CO2 complete energy solutions (we agree, renewables’ low cost in some cases as marginal addition to grids running on baseload fossil and nuclear plants isn’t that relevant to the issue of greatly reducing CO2 emissions from electricity generation), direct climate engineering and/or economical removal of CO2 from the atmosphere, adaptation, and tough shit. Especially given the underlying issue that the projected cost of climate change for the temperate zones isn’t that high*. It’s higher in the world’s South. And while you can get average people in the temperate zones to say ‘yeah, yeah we have to do something about this, my late night comedian told me’, that support will evaporate IMO when the rubber hits the road as accepting huge costs and lowered living standards mainly to limit the climate impact on other countries.
-
As I said in the post you quoted, in some other countries nuclear will have some role. But the number quoted is order of 10% of US capacity by the time China’s economy isn’t much smaller than the US. So on to the other common theme in climate change discussions, ‘but you have to do something’, but a lot of the somethings are really quite small on a global scale, and not necessarily a huge number of them to make up for that. New nuclear electric power plants, realistically, fit in this category globally, though not a complete zero globally as is likely in the US.
*last year’s celebrated report, ‘huge’ future costs to the US economy from climate changehttps://techcrunch.com/2018/11/23/new-u-s-report-says-that-climate-change-could-cost-nearly-500-billion-per-year-by-2090/
A relative bad case was projected to cost 500bil/yr by 2090 in today's 's. IOW at 2% real growth from now to 2090, 0.6% of 2090’s ~$79tril GDP. Real GDP would be 5tril pa lower just cutting growth from 2% to 1.9% pa for the next 71 yrs by forcing use of more expensive energy. If you consider the risk of much worse (climate) outcomes and/or the effect in already hot places the numbers can change, but once and if the climate debate gets off square one where one side says 'no problem' so all the other has to do is say 'yes, problem', to really analyzing whether it's worth it in 's for the US to spend huge sums turning its economy upside down, I think it’s rough sledding for aggressive plans, or any comparison to WWII.
Just a thought. I know this wouldn’t really fly, but it seems like you could feasibly do “bulk permitting”.
You start with a plant design. Not just a reactor design, a complete plant design including physical security and preparations for weather events.
You then, ahead of time, say “this plant designed is licensed for anywhere in the United States where the following conditions are true”
a. Some maximum number of people living within x miles of the plant
b. Some maximum risk per year of earthquakes, hurricanes, other extreme weather
c. Some availability for cooling water
d. Some distance from a major highway if something goes wrong and supplies need to be rushed in
e. Some maximum number of other nuclear reactors in proximity (since one meltdown can prevent access to the others - this is why Fukishima lost 3 cores instead of just 1)
And so on with an arbitrary list of mathematically defined constraints.
The way I imagine it working, you apply all these conditions to a map of the U.S., and a computer model shades in maybe 1-5% of the all the land in the USA as viable sites for further reactors using this complete design.
Then all a would be plant builder needs to do is say “ok, this next plant is within the shaded region. We are using contractors who built another copy of this exact plant. We want another one, exactly the same. Give us our license by the end of the week, please”
There would be nothing to review since you’d have exactly the same relative risks of the next reactor, and it’s the same complete design as before. Obviously this reactor series would occasionally get “updates”, where all examples are updated to fix a problem.
Doing it this way, the costs might actually be manageable and predictable.
That approach should work, but of course should and would are different. You’d still have just as much NIMBY opposition as ever.
I used to live within sight of a nuke plant, close enough that we got mailers every year documenting the current evacuation plan - where we should head and what route we should take if the sirens went off. I always thought that one solution to the NIMBY problem would be to offer discounted electric rates based on the distance from the plant. Buy people’s approval.
A better solution is to build them away from populated areas. Set up a daily bus line to commute the workers in from 30 miles away or so.
Too long of a commute, and you’ll have a hard time finding the people with the right skills willing to work there. We’ve seen that happen with prisons in my state; built in rural areas to provide jobs, they can’t attract the trained medical/psychological/nursing/dental professionals needed to keep the place running.
And 30 miles is still close enough that it’ll get NIMBYed.
I’m really curious how much you are aware of the situation in Fukushima.
The scale of problems at Fukushima are well documented, so I’m perplexed by this answer. The problems were clearly related to the number of reactors which melted down, as well as the problems with the spent fuel. The scale was much larger than had there simply been on reactor there.
Can you define what you mean by “negligible” and how the cost of the clean up relates to the being “negligible.”
Or don’t. I’m not into arguing against disingenuous tactics. If you are aware that the cost is estimated now to cost between US$198 billion (government estimates which critically do not include some substantial costs) and US$626 billion by a private, conservative think tank, then there really can’t be a discussion. The cost corresponds to the number of units involved, so the argument is absurd.
Just a side point - is 626 billion comparable to all funds spent constructing and running nuclear reactors for electric power throughout the entire history of Japan?
That is, is this retroactively doubling the price of every kilowatt hour that nuclear energy has provided?
Because if so, this is a pretty compelling anti-nuclear argument. Oh, sure, “just” build a taller tsunami wall - that statement ignores the thousands of other rare ways this accident could have happened, and how you only need it to happen once in 40 years at one of dozens of sites to wipe out all benefits of nuclear power.
As a side note, a small bomb on the switchyard, a second on one of the main cooling lines, a third on the diesel power distribution panel - it wouldn’t take much to create this same accident on purpose.
Once the core temp gets too hot, which would probably happen in an hour or so with a gaping hole in the right cooling pipe at the reactor underside - the problem becomes unfixable and a meltdown inevitable. This is because the reactor building would fill with radioactive steam, making it impossible to fix the damaged electrical panel or the pipe break.
I know nuclear plants have security - but can you guarantee that security for 80 years across 100 or more sites? At a cost you can afford to pay?
One of the things not talked about in this thread is water pollution caused by Nuclear plants. For every 1 GW a nuclear power plant puts out, it puts out 2 GW into the surrounding environment. Much of it goes through the big cooling towers but even then there is blowdown that goes to the water system.
Water is one of the limiting factors effecting the expansion of a nuclear power plant which are typically big to start with.
Heating up the water disturbs the ecosystem by growing algae / disturbing the natural balance.
Not sure this is particularly relevant. A coal plant or natural gas plant also puts a similar amount of extra heat into the environment. A hydroelectric plant disrupts entire ecologies, a solar farm takes up land where plants could have grown to support an ecosystem.
None are perfect and the immediate coal pollution and long term CO2 pollution from natural gas are probably at least 100 times more relevant to human life
Not an ecosystem, but it’s possible to combine solar panels with farming. I think the reason this works is that plants don’t use anything close to all the sunlight that falls on them. More like 1 or 2 percent. So they still have enough light even with the panels blocking a lot of it.
Your number is roughly correct, however it doesn’t consider future growth in electrical consumption which would occur over the duration of such a buildup.
Annual US electrical production is about 4,000 billion kilowatt hrs. Annual production of the 98 operating US reactors is 807 billion kilowatt hr, or 20% of the total.
This is about 8.2 billion kWh per plant per year. At 90% capacity factor, this equates to roughly 936 megawatts electrical output per plant, on average.
So for new nuclear plants to take over the remaining 3,193 billion kilowatt hrs, this would superficially take about 389 new plants of the current average size.
However annual US consumption is projected to be 5,000 billion kWh by 2030, so over the duration of such a buildup, the actual capacity required would be at least 4,192 billion kWh. This means about 511 new nuclear plants would be needed, minimum. Improvements in consumption efficiency would be offset by growth in demand, plus the actual buildout would take much longer (see below).
It wouldn’t make sense to embark on a massive buildup like that using current designs. Probably using Gen IV reactor designs and standardization would be needed, since some of those designs greatly reduce the waste problem and are much safer. However those designs are not yet available. Even if approved and funded as a hypothetical national effort, it seems unlikely that could be achieved within 40 years.
It would likely take 10 years to finalize Gen IV designs, build pilot plants and test them, then the constuction phase of 500+ plants would take at least 30 years. Generation IV reactor - Wikipedia
Cite : Thermal Water Pollution from Nuclear Power Plants
Keep in mind that the typical Nuclear power plant is in the GW capacity where’s as you can have distributed generation with Natural Gas. In fact you can have a 0 water usage Natural Gas plant running in open cycle.