Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
hobbesmaster
Jan 28, 2008

Boner Slam posted:

I was talking wind. You know, the regenerative thing.

Edit: Maybe turbine is the wrong word in English. A typical wind park with a lot of wind mills (?) is around 800MW (not h)

Oh yes, of course thats a pretty drat big wind mill. But we're judging the biomass facility as an entire plant not as a single generator in a farm.

Adbot
ADBOT LOVES YOU

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

I'm a power utility regulator.

Edit: Like I am genuinely curious what confused you guys, I'm not trying to be a smughead or anything.

A megawatt-hour is not the same as a megawatt per hour.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Schizotek posted:

e3: gently caress whoever decided KWH was the unit to use for energy.

You measure it in hour-units for some applications for a lot of reasons, but the main one is ramp-up of generating facilities.

Edit: also consumption being cumulative and having nothing to do with your max load, too.

karthun posted:

A megawatt-hour is not the same as a megawatt per hour.

I know, but I'm pretty sure I never used MW/h, just MWh. If I did I apologize for the confusion- I promise I understand the difference at work. :)

Franks Happy Place fucked around with this message at 19:47 on Apr 16, 2013

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

You measure it in hour-units for some applications for a lot of reasons, but the main one is ramp-up of generating facilities.


I know, but I'm pretty sure I never used MW/h, just MWh. If I did I apologize for the confusion- I promise I understand the difference at work. :)

Everything on the ferc site is using MW/GW/KW for generation/demand rates and demand and MWH/GWH/KWH for total usage, usually for the pricing and the energy used in comparison pages which seems to be consistent with what were explaining.

Gimby
Sep 6, 2011

Fine-able Offense posted:

I'm a power utility regulator.

Edit: Like I am genuinely curious what confused you guys, I'm not trying to be a smughead or anything.

Essentially, your units seem weird. I'm assuming that you know what you are talking about and theres some confusion here, so I'll lay it out for the peanut gallery.

Watts are a measure of power - the rate at which a source converts energy. In SI units, thats joules per second.
If we have a generator rated at 6MW, then it converts energy at a rate of six million joules per second.
We can multiply that by time to get an energy - so if we run the generator for one hour, in that hour 3600 seconds x six million joules per second = 21.6 Gj of energy.
However, we don't usually use joules in power generation, we use megawatt hours, that is, the energy converted by a one megawatt source in one hour. So in this case, if we run a 6MW generator for 1 hour, it converts 6MWh of energy.

Typically the energy converted by a power station is quoted in MWh/year - this is essentially the average power over the year.

So typically saying "something takes one hour to produce 6 megawatts" doesn't seem to make a lot of sense - power is an instantaneous measure. It might produce 6MWh of energy in an hour, but thats just a longwinded way of saying its producing energy at a rate of six megawatts

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

hobbesmaster posted:

Oh yes, of course thats a pretty drat big wind mill. But we're judging the biomass facility as an entire plant not as a single generator in a farm.

Well like I said, the "entire plant" is actually really small. It's a two-story building that would fit on an average building lot pretty much anywhere. That's part of what was so cool- you could plunk one of those down pretty much anywhere and nobody (rational) would object.

hobbesmaster
Jan 28, 2008

karthun posted:

A megawatt-hour is not the same as a megawatt per hour.

I'm curious what is energy/(time^2) commonly used for?

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

Well like I said, the "entire plant" is actually really small. It's a two-story building that would fit on an average building lot pretty much anywhere. That's part of what was so cool- you could plunk one of those down pretty much anywhere and nobody (rational) would object.

Do you have a link to an article perchance? Now that I know more or less what you tried to say (seriously the kwh thing is an awkward and unwieldy thing to grasp initially, kinda like lightyears but more so and less well known) I'd like to read about it.

Schizotek fucked around with this message at 19:58 on Apr 16, 2013

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Schizotek posted:

Everything on the ferc site is using MW/GW/KW for generation/demand rates and demand and MWH/GWH/KWH for total usage, usually for the pricing and the energy used in comparison pages which seems to be consistent with what were explaining.

Nah, take this for example: it's explicitly generation.


Gimby posted:

Essentially, your units seem weird. I'm assuming that you know what you are talking about and theres some confusion here, so I'll lay it out for the peanut gallery.

I actually don't disagree it's weird, but part of the confusion is that when you are talking about the aggregate power system as a whole, the peak production capacity of an individual generating source is far less relevant than the production over a given unit of time, accounting for things like ramp-up etc. Since system-wide demand is always going to be higher than the production capacity of that one plant, you want to know what that plant can contribute to the system as a whole in a base load kind of way over time. So you say, "in one hour this source will increase and decrease due to varying load requirements, but it averages 6MWh", because (in my line of work) that's what really matters.

And yes, the term is (probably) horribly misused due to the fact that it is explicitly tied to MWh per dollar, aka the only thing people care about at the end of the day.

hobbesmaster posted:

I'm curious what is energy/(time^2) commonly used for?

In my experience (which I am starting to realize is pretty weird and specific!) it is used for ramp-up factor.

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

I know, but I'm pretty sure I never used MW/h, just MWh. If I did I apologize for the confusion- I promise I understand the difference at work. :)

quote:

Again, you're going to have to explain how you got confused when I said the facility puts out 6MWh. Because if I said that in the office, everybody would understand what I meant: it makes six megawatts of heat in an hour.

A plant doesn't produce six megawatts in an hour, it produces six megawatt-hour in an hour. It will produce 12 megawatt-hours in 2 hours and 144 megawatt-hours in a day.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

karthun posted:

A plant doesn't produce six megawatts in an hour, it produces six megawatt-hour in an hour. It will produce 12 megawatt-hours in 2 hours and 144 megawatt-hours in a day.

Uh, unless I have suffered a stroke, I'm pretty sure that a plant rated 1 MWh produces one megawatt in an hour?

Gimby
Sep 6, 2011

Fine-able Offense posted:

In my experience (which I am starting to realize is pretty weird and specific!) it is used for ramp-up factor.

Righto, so explictly the load response rate - acceleration of energy generation? So high for something like pumped storage, low for something like baseload coal?

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

Nah, take this for example: it's explicitly generation.


Total generation for a week. Not rate of generation. Your getting close to understanding the difference I think. Just need that click. This reminds me of when we learned about this in college. My study group spent 4 goddamn hours on this and half the group still really never got it, but could parrot the equations enough to pass the test.

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

Uh, unless I have suffered a stroke, I'm pretty sure that a plant rated 1 MWh produces one megawatt in an hour?

A megawatt is a rate in and of itself, not a unit. We need to give a reverse Nobel Prize to whoever named these loving units.

So a plant rated 1 MWh would produce 1 MWh per year or however long they're measuring. But usually I see a year. That means it would produce enough electricity in a year to fuel a lightbulb that needs a rate of power of 1 MW for one hour.

Schizotek fucked around with this message at 20:15 on Apr 16, 2013

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Schizotek posted:

Total generation for a week. Not rate of generation. Your getting close to understanding the difference I think. Just need that click. This remind me of when we learned about this in college. My study group spent 4 goddamn hours on this and half the group still really never got it, but could parrot the equations enough to pass the test.

I think I get it more than you're giving me credit for, it's more that this is basically the base measurement unit of my work, so we're talking two slightly different languages here.

I just talked to an engineer I work with who went on the tour with me, he can't remember either if they talked about values in MWh or MW. He had just as much trouble expressing why we care so much about the hourly rating as I did, which is a relief. :)

As for the plant info: here. They seem to express units in strict max MW capacity on there, but I am, like, 100% certain they talked about MWh on the tour. That could well be a mistake on their part rather than ours, which would make sense since like I said, all we deal with are the MWh due to their link to cost, so we wouldn't have noticed it being odd until I sat here trying to explain it to a bunch of engineers on the internet.

The more I think about it, the more I think they meant peak capacity being 6MW, but that they said MWh and we didn't notice because that's a perfectly normal value unit to us, just misused in that context.

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

Uh, unless I have suffered a stroke, I'm pretty sure that a plant rated 1 MWh produces one megawatt in an hour?

1 MWh over what period of time? Plants are rated at POWER, J/s also known as watts. A MWh is a measure of ENERGY.

http://www.amazon.com/All-Power-America-APG3004-1000-Watt/dp/B0017YV188/ref=sr_1_1?s=lawn-garden&ie=UTF8&qid=1366139494&sr=1-1

This 1 kW generator will produce 1 MW-h over 1000 hours.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

karthun posted:

1 MWh over what period of time?

Over an hour. That's what the H means.

We use it because it allows us to average out generation and demand, which fluctuate wildly in a grid, over a convenient unit of time.

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

Over an hour. That's what the H means.

No, there is no "over an hour" A MW-H is 3,600,000,000 joules. A watt is 1 joule/second. A megawatt is 1,000,000 joules/second.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

karthun posted:

No, there is no "over an hour" A MW-H is 3,600,000,000 joules. A watt is 1 joule/second. A megawatt is 1,000,000 joules/second.

You do understand that electricity is sold to you at a price that is established by taking your kilowatts of consumption and then multiplying those by units of time, right? Like, capacity over a given unit of time (hours) is basically just as important as the actual quantity of energy measured in joules.

Gimby
Sep 6, 2011

Fine-able Offense posted:

You do understand that electricity is sold to you at a price that is established by taking your kilowatts of consumption and then multiplying those by units of time, right? Like, capacity over a given unit of time (hours) is basically just as important as the actual quantity of energy measured in joules.

Those two are the same thing?

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

You do understand that electricity is sold to you at a price that is established by taking your kilowatts of consumption and then multiplying those by units of time, right? Like, capacity over a given unit of time (hours) is basically just as important as the actual quantity of energy measured in joules.

One does not have kilowatts of consumption. They have kilowatt-hours of consumption. That is not the same as kilowatt per hour.

kilowatt = 1000 j/s
kilowatt-hour = 3.6 Mj
kilowatt per hour = 0.2777 j/s^2

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Gimby posted:

Those two are the same thing?

No, they aren't. Take a look at your utility bill if you don't believe me.

Or to put it another way: you use straight kilowatts to measure the output of something like a solar panel right now, or what its maximum theoretical output is at any given moment. So if it can produce ten kilowatts when it's sunny, you could rely on it to run your toaster and your TV, but not your vacuum too, because that would require (say 15 kilowatts).

Now, you turn the toaster off, so your kilowatt consumption drops, but you are still consuming 8 kilowatts from the TV. At the end of an hour, the power company will measure your total aggregate consumption (which varied in terms of kilowatts over the course of that hour), and assign it a KWh value as an average.

Watts being a power over time unit in physics has nothing to do with how they are measured by a utility, which is measuring both the amount of power they sold you, and the bandwidth you used in consuming it.

karthun posted:

One does not have kilowatts of consumption. They have kilowatt-hours of consumption. That is not the same as kilowatt per hour.

kilowatt = 1000 j/s
kilowatt-hour = 3.6 Mj
kilowatt per hour = 0.2777 j/s^2

At this point I have no idea what point you are trying to make. And no, that doesn't mean I don't understand WHAT you are saying- I don't understand the WHY.

Basically, if you are bringing joules into the conversation, you don't understand how utilities work or how power is sold.

Franks Happy Place fucked around with this message at 20:48 on Apr 16, 2013

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

No, they aren't. Take a look at your utility bill if you don't believe me.

Or to put it another way: you use straight kilowatts to measure the output of something like a solar panel right now, or what its maximum theoretical output is at any given moment. So if it can produce ten kilowatts when it's sunny, you could rely on it to run your toaster and your TV, but not your vacuum too, because that would require (say 15 kilowatts).

Now, you turn the toaster off, so your kilowatt consumption drops, but you are still consuming 8 kilowatts from the TV. At the end of an hour, the power company will measure your total aggregate consumption (which varied in terms of kilowatts over the course of that hour), and assign it a KWh value as an average.

Watts being a power over time unit in physics has nothing to do with how they are measured by a utility, which is measuring both the amount of power they sold you, and the bandwidth you used in consuming it.



This mostly accurate understanding seems totally at odds with this statement:

Fine-able Offense posted:

Again, you're going to have to explain how you got confused when I said the facility puts out 6MWh. Because if I said that in the office, everybody would understand what I meant: it makes six megawatts of heat in an hour.

So I really have no idea if you do understand and just hosed up earlier or you have some sort of bizarre mental shorthand that lets you not understand your work but able to perform it.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Schizotek posted:

This mostly accurate understanding seems totally at odds with this statement:


So I really have no idea if you do understand and just hosed up earlier or you have some sort of bizarre mental shorthand that lets you not understand your work but able to perform it.

No, I definitely misspoke up there. I left out "of capacity" or some other load-delineating term that would have clarified what I meant.

karthun
Nov 16, 2006

I forgot to post my food for USPOL Thanksgiving but that's okay too!

Fine-able Offense posted:

At this point I have no idea what point you are trying to make. And no, that doesn't mean I don't understand WHAT you are saying- I don't understand the WHY.

Basically, if you are bringing joules into the conversation, you don't understand how utilities work or how power is sold.

A kilowatt-hour is not the same drat thing as a kilowatt per hour. A 6 MW plant will not produce 6 MW per hour, it will produce 6 MW-h per hour. A 6 MW-h plant is meaningless because a 7 Kw generator will produce 6 MW-h per year. A 14 Kw generator will produce 6 MW-h in 6 months and 12 MW-h per year.

Gimby
Sep 6, 2011

Fine-able Offense posted:

No, they aren't. Take a look at your utility bill if you don't believe me.

Or to put it another way: you use straight kilowatts to measure the output of something like a solar panel right now, or what its maximum theoretical output is at any given moment. So if it can produce ten kilowatts when it's sunny, you could rely on it to run your toaster and your TV, but not your vacuum too, because that would require (say 15 kilowatts).

Now, you turn the toaster off, so your kilowatt consumption drops, but you are still consuming 8 kilowatts from the TV. At the end of an hour, the power company will measure your total aggregate consumption (which varied in terms of kilowatts over the course of that hour), and assign it a KWh value as an average.

Watts being a power over time unit in physics has nothing to do with how they are measured by a utility, which is measuring both the amount of power they sold you, and the bandwidth you used in consuming it.

Basically, if you are bringing joules into the conversation, you don't understand how utilities work or how power is sold.

No, I understand how electricity is sold - I've got a background in the generation side of it. What you seem to be saying is that the average power consumption of the user over time multiplied by the amount of time used (and yes, this is done on time increments as the cost of energy varies over time - or can do, depending on your tariff) is the amount charged for. Thing is, this is the total energy used. Thats how its defined. For a typical household, the cost per kWh is independant of the power demand (or at least at the household level, costs throughout the day can vary, but that's based on national averages, at least in the UK). Industrial power bought on wholesale contracts can be different if energy demands are different to the expected, but its still sold in units of kWh, that is energy.

Joules are only in the conversation because, like kWh, they are a measure of energy. The kWh you see on your utility bill can be expressed in joules (or something more oldschool like BTUs) if you fancy it because they are measuring the same thing - energy.

This is a pointless derail however, so I'm happy to drop it.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Gimby posted:

This is a pointless derail however, so I'm happy to drop it.

Yeah I agree, it seems to have arisen because I was unspecific/wrong with my terms and the terms in question are pretty dumb to begin with. I'm also going to drop it.

Taerkar
Dec 7, 2002

kind of into it, really

Fine-able Offense posted:

Car insurance is not the same as calculating the risk of something like a pipeline rupture or nuclear powerplant meltdown, just FYI.

Like, nobody is using such a simple formula when the downside risk is effectively measured by ∞ instead of a dollar figure.

My example was a simplification because there's far more going into it in terms of variables, but probabilities of failure are extensively calculated for such situations. Hell, it's why Life Insurance is one of the safest bets out there for the companies involved.

The risk of a pipeline rupture, for example, would be based upon the observed failure rate of the various components of the system, the environment and likely weather patterns that the pipeline would be exposed to, and the caustic properties of what the pipeline would carry, just to name a few conditions. There's likely at least hundreds of more factors involved, but they're all part of that risk analysis.

Boner Slam posted:

We are not talking about simplifications or their worth.
It is not factually correct, and the poster before me pointed that out.

First off, you don't know the probability. You know the frequence of events in the past. It is not the same loving thing. That you assume you can _estimate_ the "non quote" probability shows you have no idea what you are talking about.

Second and more importantly, you equate a binary event with a continous value of "that much radiation by mining" by taking the expected value of a binary distribution. By that you assume that a barrel can leak or not leak, always with the same probability and always the same amount of leakage. And THEN you go ahead and claim that obviously this equates to a number value of radiation during mining. Without talking about the actual error severity of such an estimation, it is not correct in so many ways.

A guy pointed this out.
And instead of saying "I think in this case the difference is not important to the hypothetical argument" (which would have been fine) you guys went ahead and jumped on him like "just because you don't understand it dumbo". Literally.
It is a perfectly reasonable thing to point out and it is also correct. And this kind of posting makes you look like an rear end

We can make pretty good guesses at the probability, sure they're estimates of estimates, but a lot of it is based around some very strenuous stress testing and past evaluations, along with some pretty drat complicated computer modeling that gets tossed around today.

Those estimates are (or at least should be) reevaluated from time to time based upon both an improvement in testing procedures as well as an analysis of real-world performance. It's why we know of the potential lifespan of various parts of a plane, for example.

These are also more than just binary events of "poo poo goes bad" and "poo poo doesn't go bad". A properly done analysis will build a likely range of possible outcomes and then build out from there. And yes, one of those is the worst possible circumstance.

The problem becomes when those analysis meet up with other factors; Shareholders, governments, the voting public, greed, corruption, etc. Something that's evaluated to be a 'distinctively remote possibility' may become much more likely if, say, the item in question is used for twice as long as it was originally expected.

The risks should be reevaluated in that sort of situation but hey, that costs money, and that money would cut into the bottom line. Many of the risks that contributed to the failure at Fukushima are things that were known but not addressed in no small part because of the cost.

Aureon
Jul 11, 2012

by Y Kant Ozma Post

Fine-able Offense posted:

This is the goddamned stupidest thing I have yet read in this thread, and that's (sadly) saying something. I know that goons are positively obsessed with white-knighting nuclear power no matter what contortions it requires, but calling the consequences of Fukushima "not severe" is just... do you expect anybody to take that statement seriously? Come on, man.

As of late last year, TEPCO estimated the cost of the Fukushima incident + cleanup at around $11 TRILLION yen. So there alone is a "severe consequence" in the context we were speaking about (i.e., effectively uninsurable costs), never mind the actual real effects in terms of the ongoing dangerous and expensive containment protocols, the exclusion zone, etc. etc., all of which is real and necessary no matter how many sperg warriors make quips about TEH ATOMZ.

I'm not opposed to nuclear power per se as part of a rational energy mix, but the way some people talk about it in this thread is just the gooniest, spergiest poo poo. I posted just last week about a tour I went on of a biomass generator where they literally just throw wood garbage into a pile and squeeze 6MWh out of it with no pollutants, noise, or other negative impact whatsoever, in a tiny unassuming building you could plop down in the middle of a dense urban environment without anybody noticing. Nobody gives a poo poo about that, but lets all go on and on and on about nuclear power being the Mary Sue of power sources and try to pretend like a $137 billion insurance bill isn't a valid policy concern for people in my line of work to deal with.

Good-drat-job taking me out of context! Of course the fukushima events are "Severe".
They are not severe enough to make risk calculation invalid.
Oh, and we'll need a source on that 110t Yen damage, since the whole Earthquake damage was ~350b
Especially since (you love that word, don't you?) it's an externality of perhaps the worst natural disaster humanity has suffered in recent times.
Also good job on using BIG! SCARY! NUMBERS!

Also, good job burning biofuels, higher food prices are exactly what we need! (And no, i'm not saying that the current crop prices rises depend on biofuels: I'm saying that if a substantial % of all crop land was converted to biofuels, or to solar panels (in places lacking deserts and such, such as more or less all Europe), they would rise and create a real disaster.)

We all here understand the political, economical and social limitations of the implementation of Nuclear Power.
That doesn't mean we also acknowledge that, from an engineering viewpoint, it's the best option we have at the present. If in 20 years we get solar orbital, miracolous solar cells that somehow break the physical limit of 27% or something equally incredible that'll actually solve our problems, good. That may include a variety of options, and all of those should get research funding. Which is not industry subsides, though - and if you want to see how these "Feel-Good" bandaids have worked, look no further than Germany and how it's building coal plants to switch out Nuclear plants.
The very moment there's a better option than Nuclear, you won't see a single nuclear supporter here. Until then, for a compendium of roughly every single talking point, we've got the aptly-named thread "The issues affecting the implementation of Nuclear Power" (Which has now been archived, sadly)

shrike82
Jun 11, 2005

Taerkar posted:

My example was a simplification because there's far more going into it in terms of variables, but probabilities of failure are extensively calculated for such situations. Hell, it's why Life Insurance is one of the safest bets out there for the companies involved.

Comparing the risk of a Fukushima-level nuclear plant failure with life insurance companies measuring mortality risk is a terrible analogy which really underscores a lack of understanding about probability by you and other posters ITT. The reason why life insurance works so well is that there's a huge pool of participants (past and present) that help in deriving statistically accurate mortality tables. We don't have enough nuke plants operating in the world to give a similar confidence level.

If anything, it's probably better to make an analogy with the financial markets. We've seen enough hedge funds blow themselves up over the past few decades due to misjudgments about six sigma events.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Aureon posted:

Oh, and we'll need a source on that 110t Yen damage, since the whole Earthquake damage was ~350b

I linked the citation in my post. :allears:

Aureon posted:

Also, good job burning biofuels, higher food prices are exactly what we need!

Biomass is not biofuels. :allears:

Aureon posted:

The very moment there's a better option than Nuclear, you won't see a single nuclear supporter here.

Given the above two notes, I am hardly surprised that you don't understand the distinction between an attack on you for supporting nuclear power vs. a critique of the apparently ideological means with which you do so, especially considering the discussion was related to something as mild as "Hey, you can't treat all power systems equally when it comes to the risk-cost of critical failure, especially from an actuarial standpoint."

Taerkar
Dec 7, 2002

kind of into it, really

shrike82 posted:

Comparing the risk of a Fukushima-level nuclear plant failure with life insurance companies measuring mortality risk is a terrible analogy which really underscores a lack of understanding about probability by you and other posters ITT. The reason why life insurance works so well is that there's a huge pool of participants (past and present) that help in deriving statistically accurate mortality tables. We don't have enough nuke plants operating in the world to give a similar confidence level.

If anything, it's probably better to make an analogy with the financial markets. We've seen enough hedge funds blow themselves up over the past few decades due to misjudgments about six sigma events.

As of the start of the year there are 435 NPRs in operation around the world. Overall those reactors, along with decommissioned ones, account for a bit under 15,000 years of combined operational experience. Source: IAEA Report

If you can't make a decent estimate out of that, you shouldn't do stats.

And I'm not positive, but I suspect that number does not include military reactors.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Aureon posted:

Oh, and we'll need a source on that 110t Yen damage, since the whole Earthquake damage was ~350b

I think he misplaced a decimal, or performed the USD -> Yen conversion twice, or something, because the American Nuclear Society puts the cleanup cost at 1.15T Yen, or about $15b USD.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Paul MaudDib posted:

I think he misplaced a decimal, or performed the USD -> Yen conversion twice, or something, because the American Nuclear Society puts the cleanup cost at 1.15T Yen, or about $15b USD.

Seriously how hard is it to click the loving link in the post? I literally hyperlinked the words "around $11 TRILLION yen" to make it easy. And if anything that number is still low, that's just the amount TEPCO has requested.

Also did you even read what you cited?

American Nuclear Society posted:

Decommissioning the four reactors is estimated to cost at least 1.15 trillion yen [$15 billion (USD)]

Decommissioning the four reactors is not the same as the entire cleanup, come the gently caress on. This is exactly the kind of lovely argumentation I was referring to. You guys are nuclear power's own worst enemy at this point.

Franks Happy Place fucked around with this message at 22:01 on Apr 16, 2013

shrike82
Jun 11, 2005

Taerkar posted:

As of the start of the year there are 435 NPRs in operation around the world. Overall those reactors, along with decommissioned ones, account for a bit under 15,000 years of combined operational experience. Source: IAEA Report

If you can't make a decent estimate out of that, you shouldn't do stats.

And I'm not positive, but I suspect that number does not include military reactors.

Whoops...

http://www.modelisation-prospective...ents_entail.pdf

CDF: Observed frequency is 7.6E-04/r.y versus the estimated 1.0E-04/r.y
LERF: Observed frequency is 2.8E-04/r.y versus the estimated 1.0E-06/r.y

Boner Slam
May 9, 2005

Taerkar posted:

My example was a simplification because there's far more going into it in terms of variables, but probabilities of failure are extensively calculated for such situations. Hell, it's why Life Insurance is one of the safest bets out there for the companies involved.

The risk of a pipeline rupture, for example, would be based upon the observed failure rate of the various components of the system, the environment and likely weather patterns that the pipeline would be exposed to, and the caustic properties of what the pipeline would carry, just to name a few conditions. There's likely at least hundreds of more factors involved, but they're all part of that risk analysis.
You still don't understand my initial criticism. Now you are saying models are more complicated. Well no poo poo. Also the example you give is still no analogy to the method I was attacking. If you are working with an estimator of any kind of complexity, you are not doing what was claimed to be correct in the earlier posts. If you know what an estimator is and what properties are associated, you know you are technically absolutely wrong when saying expected value is literally equal to constant rate as risk. And it would only be in this one single case when reality, REALITY was based on a binary distribution as such. In all other cases the cricisim of the method is completely justified and beeing sassy about it was inappropriate.

Taerkar posted:

We can make pretty good guesses at the probability, sure they're estimates of estimates, but a lot of it is based around some very strenuous stress testing and past evaluations, along with some pretty drat complicated computer modeling that gets tossed around today.

Those estimates are (or at least should be) reevaluated from time to time based upon both an improvement in testing procedures as well as an analysis of real-world performance. It's why we know of the potential lifespan of various parts of a plane, for example.

Statistical inference is more complicated than you are apparently willing to admit and I can only reiterate that you must know that even experimental testing is based on assumptions. Statistical inference about leakage of individual barrels dropped randomly in the loving sea, are you kidding me?

Taerkar posted:

These are also more than just binary events of "poo poo goes bad" and "poo poo doesn't go bad". A properly done analysis will build a likely range of possible outcomes and then build out from there. And yes, one of those is the worst possible circumstance.
Yes, which is why I attacked the post that claimed exactly such a binary event.

Taerkar posted:

The problem becomes when those analysis meet up with other factors; Shareholders, governments, the voting public, greed, corruption, etc. Something that's evaluated to be a 'distinctively remote possibility' may become much more likely if, say, the item in question is used for twice as long as it was originally expected.

No the problem is that the expected value does not characterize every estimator completely, especially when it is complex. It is therefore incorrect to say the "risk" is equal to a constant value which is the expected value. It is not correct. This is not the risk. The risk includes the properties of the estimator as based on the assumptions of the underlying probabilities and this can be quite simply illustrated by a "catastrophic" and by a "mild" "risk", both with the same expected value like with the nuclear power plant example or what have you.
This was the initial point you attacked as wrong and it is not wrong. The person who made this argument is correct. It is not "as simple as that".



But there is another thing you seem to not understand or adress, which is that the variable "amount of radiation leaked" is not binary but metric.
By this alone the argument you make is wrong. You can come out and say "well on average we _expect_ this to be the correct amount based on our estimated properties of a barrel dropped in the sea" but this is not what was claimed. What was claimed was that this approach was 100% correct and characterized "the risk", the end. No it does loving not. Alone by all these assumptions you are probably making every estimating error in the book.

And as far as financial and economical or just monetary risk goes, you know loving well that between "what is a traditional way to do it since the 80's" and what is even close to acedemically sound is a gigantometric fuckton fo space which, yes, did have real consequences in the past.

And even then, in many fields even academics make horrible, horrible estimations, such as not even understanding when things are not a controlled experiment. I can give you gigantic datasets and you'd still be no closer to a good estimation in many questions. And I am pretty sure that amount of leakage of a barrel dropped in the sea would be a pretty tough one. Maybe not "Total radiation leakage in a barrel dump as determined by multiple experiments in the same area", but that is after all not what we are loving talking about.

Boner Slam fucked around with this message at 22:15 on Apr 16, 2013

Adenoid Dan
Mar 8, 2012

The Hobo Serenader
Lipstick Apathy

Fine-able Offense posted:

Again, some pro-nuclear people in this thread keep making this argument like it's some kind of trump card, because apparently they know dick all about human psychology, politics, and economics. News flash: the costs of pollution from coal plants are what we call "externalities", and they are essentially socially-acceptable ones at that. By way of analogy, they fall into the same category as heart disease, obesity, and diabetes from eating lovely food- sure, everybody knows eating lovely food is bad, and the government could easily mandate everyone become vegetarian or whatever (thereby saving society enormous costs), but that's only a rational argument to make if you are somehow pig-loving-ignorant of all of the vested interests at play in making that decision, up to and including the average dude's love of Cheetos. If you want to make an argument to convince somebody to drop the trans fats and eat some celery, you can't do it by being Captain Smugboat McSperglord and trying to somehow stand athwart human nature.

The problems with nuclear power are also externalities, in fact that's pretty much all this discussion is about. I fail to see why people being bad at judging risk (consistently overestimating the danger from spectacular but rare events while underestimating common causes of death) is a mark against nuclear and in favor of coal. They are socially acceptable because risk is not intuitive and people make decisions based on emotions. That's why we need to approach problems like this scientifically.

And yes, of course psychological damage is part of the damage caused by nuclear accidents, one of the biggest, in fact.

Schizotek
Nov 8, 2011

I say, hey, listen to me!
Stay sane inside insanity!!!

Fine-able Offense posted:

Seriously how hard is it to click the loving link in the post? I literally hyperlinked the words "around $11 TRILLION yen" to make it easy. And if anything that number is still low, that's just the amount TEPCO has requested.

"A private think tank says" "The figure includes 54 billion to buy up all land within 20 kilometers of the plant, 8 billion for compensation payments to local residents, and 9 to 188 billion to scrap the plant's reactors."

There's no way this is just a bullshit figure pulled out of TEPCO's rear end to rip off the government.

Franks Happy Place
Mar 15, 2011

It is by weed alone I set my mind in motion. It is by the dank of Sapho that thoughts acquire speed, the lips acquire stains, stains become a warning. It is by weed alone I set my mind in motion.

Adenoid Dan posted:

The problems with nuclear power are also externalities, in fact that's pretty much all this discussion is about. I fail to see why people being bad at judging risk (consistently overestimating the danger from spectacular but rare events while underestimating common causes of death) is a mark against nuclear and in favor of coal.

My point was simply that stating the fact that coal is harmful because of a variety of very real, but very diffuse risks doesn't automatically make Joe Public want to sign on for one singular great big risk, even if that risk is a lot lower. Human nature is part of it, but so are the very real insurance factors we've been talking about. Nobody running a coal plant has to insure for, or mitigate against* air pollution, acid rain, global warming, etc., which puts nuclear at a real disadvantage.

Nuclear advocates would be better off finding common cause with green energy advocates and pushing for carbon taxes or whatever, and returning some of those externalities back where they belong, which would go a long way to making nuclear power competitive.

*In any meaningful way I mean.

Schizotek posted:

"A private think tank says" "The figure includes 54 billion to buy up all land within 20 kilometers of the plant, 8 billion for compensation payments to local residents, and 9 to 188 billion to scrap the plant's reactors."

There's no way this is just a bullshit figure pulled out of TEPCO's rear end to rip off the government.

Hey, I presented my evidence: the actual dollar amount TEPCO has told Japanese authorities they will need to pay for this mess.

You have presented... nothing. So, if you want to make the argument that the cost of a nuclear accident is actually really not that bad blah blah blah, well: [Citation needed]

Adbot
ADBOT LOVES YOU

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Schizotek posted:

"A private think tank says" "The figure includes 54 billion to buy up all land within 20 kilometers of the plant, 8 billion for compensation payments to local residents, and 9 to 188 billion to scrap the plant's reactors."

There's no way this is just a bullshit figure pulled out of TEPCO's rear end to rip off the government.

Their decommissioning costs ("scrap the plant's reactors") are off by an order of magnitude. The ANS puts that cost at $15b, not $188. That part is where TEPCO is gouging the government, or they're trying to add in some improbable what-if horror scenario that is ballooning the high end of that estimate.

I also doubt that area is really sufficiently contaminated to warrant excluding anything other than maybe livestock grazing, making the land purchases more of a comfort thing than actually necessary. If I remember they were hard at work scooping up the topsoil in those areas too. Anyone have an up-to-date soil contamination map?

Paul MaudDib fucked around with this message at 22:25 on Apr 16, 2013

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply