More efficient/completely safe nuclear reactor setups?

  • Please make sure you are posting in the correct place. Server ads go here and modpack bugs go here
Status
Not open for further replies.

Omicron

New Member
Jul 29, 2019
2,974
0
0
Wow, nice. I like the ME interface with different cable colors to create two individual networks. Does that work on-demand though? Meaning, do you just breed and centrifuge all the isotopes you can make, or only those that are necessary to craft the next reactor cycle's worth of fuel?

My system does the latter. It will breed exactly 40 isotopes per reactor cycle, because that is how much it needs to finish up the five quad thorium and five plutonium cells. The rest of the time, it simply waits out. Obviously that "wastes" quite some breeding potential, but I don't really expect to have any other reactors to fuel.

Overall though my system is much more crude than yours :p Have a look at this screenshot.

I got rid of the ME chest, and instead use a regular chest with a storage bus. The storage bus has exactly the same functionality as a preformatted disk, but you save the energy draw of the ME chest. Plus, the regular chest can be more easily kept stocked with items from a different network - for example, using a redpower regulator, or an advanced golem.

The storage bus on the GregTech advanced regulator (nearly invisible under all the stuff connected to it) will accept plutonium cells and quad thorium cells. The storage bus on the chest will accept everything else that can possibly crop up in this network except those two. This would be copper ingots, tin ingots, uranium ingots, coal dust, tin cells, near-depleted uranium cells, depleted isotope cells, re-enriched uranium cells, single thorium cells and double thorium cells. Because the two storage buses are configured that way, it means that the GregTech regulator acts as the "chest" where quad thorium and plutonium cells are deposited. Which is convenient, because then they are instantly available when the regulator needs to insert them.

The export bus on the regulator exports quad thorium cells, and is set to "always craft". The level emitter next to it will shut it off as soon as it detects 5 quad thorium cells in the network (which works because the storage bus on the regulator makes the regulator a part of the network). I never explicitly craft plutonium cells anywhere, but they come out of the centrifuge as a side product and are forced to go into the regulator by the configuration of the storage buses.

The interface on top of the regulator has the "depleted isotope -> re-enriched uranium" crafting pattern. The interface on top of the centrifuge has the "8x re-enriched uranium -> 4x thorium cell" crafting pattern.

I am standing on top of the autocrafter module. It is hooked up via a dark cable with a level emitter that will disconnect the ME crafter when enough quad thorium cells have been produced. Unfortunately the level emitter itself takes so much energy that you don't end up saving anything :( But it was a neat concept. The autocrafter module has the remaining patterns: tin ingots -> empty cells; uranium + empty cells -> near-depleted cells; near-depleted cells + coal dust -> depleted isotopes; and the two plans necessary to craft a quad thorium cell out of copper and single cells.

The reactor has a control unit on it with four level emitters and a NOT gate (plus a manual off switch). The four level emitters each monitor one of the basic four resources (copper, tin, uranium, coal dust), and will emit a redstone signal if one of them falls below the amount necessary to craft a single quad thorium cell, which shuts the reactor off. As soon as more resources are provided, the reactor will turn on again. Sadly, as said before, level emitters cost quite a bit of power... those four will draw 2 EU/t, meaning they eat 2 million EU per reactor cycle. Not sure if that's worth it.

I also have a crafting monitor and an access terminal hooked up via a dark cable for on-demand status checking.

EDIT: Managed to save 2-3 cables with a bit of cleanup after taking the screenshot.
 
  • Like
Reactions: noah_wolfe

Peppe

New Member
Jul 29, 2019
836
0
1
Ae is so new to us it helps a great deal to see how others do. Did not see that always craft mode on export bus-- interesting.

I saw a video with ae connecting directly to auto crafting tables. They must implement the BC item insert/pull on the busses. Since we only need the thorium crafted these would probably be less power than the big crafter.

In my first system I had the centrifuge production limited to the need... I think. In my last one it was processing all available enriched cells. Torn on how I would want a real system to run -- the breeder can make enough to run itself and the three reactor set of the 400 reactors. So in my head the system should handle 4 reactors.

So many options to play with...

Edit: Just looking at the crafter you think it uses a lot of power, but only 1 unit /t (for the single provider size)... So as soon as you have more than one recipe it is cheaper to maintain than autocrafting...
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Unsure of that. My 3x3x3 autocrafter only takes like 2 energy (I can see the difference when the dark cable toggles). Of course each crafting process also consumes a fixed amount of energy, but it depends on how much energy the import/export buses require. I have 6 recipes being taken care of by the ME autocrafter, so I would require 6 autocrafting tables and therefore 6 export and 6 import buses. That's 12 things requiring energy to sustain. I'm not convinced that it's more efficient than the single ME autocrafter.

As for a setup where one reactor breeds for two others (according to the planner it's just a few breeding cycles short of sustaining 4 total hybrid reactors), I'm not sure how I'd properly sort the correct amount of each fuel cell to each reactor. I mean, I suppose I could ask a handful of golems to always keep each advanced regulator stocked from a ME interface providing the fuel cells, but that would be... I don't know... unfitting. I feel like I should use a technological solution for automated nuclear reactors, not a magical one. Plus, would you want to leave a group of evil brain'd golems that are plotting a golem uprising alone with your nuclear reactors of all things? :eek:
 

Peppe

New Member
Jul 29, 2019
836
0
1
Unsure of that. My 3x3x3 autocrafter only takes like 2 energy (I can see the difference when the dark cable toggles). Of course each crafting process also consumes a fixed amount of energy, but it depends on how much energy the import/export buses require. I have 6 recipes being taken care of by the ME autocrafter, so I would require 6 autocrafting tables and therefore 6 export and 6 import buses. That's 12 things requiring energy to sustain. I'm not convinced that it's more efficient than the single ME autocrafter.

As for a setup where one reactor breeds for two others (according to the planner it's just a few breeding cycles short of sustaining 4 total hybrid reactors), I'm not sure how I'd properly sort the correct amount of each fuel cell to each reactor. I mean, I suppose I could ask a handful of golems to always keep each advanced regulator stocked from a ME interface providing the fuel cells, but that would be... I don't know... unfitting. I feel like I should use a technological solution for automated nuclear reactors, not a magical one. Plus, would you want to leave a group of evil brain'd golems that are plotting a golem uprising alone with your nuclear reactors of all things? :eek:

Yeah tested it in my world, the crafting machine is big, but uses energy based on what is in the core. Right now everything in the mod basically uses one energy a tick, so even though it looks big it only passively uses 1 energy with just the single core block in the 3x3.

Just spit balling here.

Can't we use redpower 2 to do a reliable round robin of the fuel? Have an ME interface exporting some quad cells and plutonium cells. Use a sorting machine, probably overkill, so maybe even just a transposer + timer to pull fuel out of the ME interface. At each GT regulator have a redpower 2 regulator -- which i think can be set to stop accepting input when the GT regulator has the right fuel amount.
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Okay. Tested the keeping-exact-numbers-in-storage part a bit.

Test 1: Thaumcraft Golems. Result: does not work because thorium cells do not stack and the golems either cannot handle it or glitch out when trying to take more than one non-stacking object from a chest at once.

Test 2: Redpower regulators. Result: working proof-of-concept build. Pay no mind to the silly backwards construction, I was just tossing things on the ground at random with no clear plan. It can be made much more compact if you intentionally plan for it.

You need three source chests (one for each item) and three transposers to make this work. They're being driven by one timer on a 10 second cycle, which is more than fast enough... also, the regulators emit a redstone signal when they are happy, which I have hooked up to an AND gate that turns the timer off entirely when there's no work to be done. Just a neatness measure.

So is it all good? Well... no. Take a look at this regulator. This is the "I am fully satisfied" steady state. In the grid on the left, you tell the regulator what items to accept from the network, and it also dictates how many to buffer at any time. In the grid on the right, you specify what the chest is supposed to contain. In the middle is the buffer, which the regulator draws from to fulfill the condition mapped on the right. If the condition is filled, the buffer will fill up to match the left side grid. Then the regulator is happy.

Why is this a problem? Because none of these things are ghost slots. They are all physical items. That means with three regulators set up this way, you will automatically have 18 quad thorium cells and 9 plutonium cells (plus 2 depleted isotopes if you let one reactor breed, or 6 if you let all of them breed) tied up in the sorting grids. At efficiency 9 which these reactors run at, that is almost one billion EU in fuel cells sitting there just to specify the sorting. One billion EU you'll never get out of your system. It's also an uneven amount, while the centrifuge produces (and the reactors are using) even amounts. Meaning you have to either produce 9 quad thorium cells somewhere without generating plutonium, or contend with 9 plutonium cells (another 324 million EU) lying around unused in storage.

I will take another look at GregTech automation devices and see if they can do what is needed.
 

Peppe

New Member
Jul 29, 2019
836
0
1
Okay. Tested the keeping-exact-numbers-in-storage part a bit.

Test 1: Thaumcraft Golems. Result: does not work because thorium cells do not stack and the golems either cannot handle it or glitch out when trying to take more than one non-stacking object from a chest at once.

Test 2: Redpower regulators. Result: working proof-of-concept build. Pay no mind to the silly backwards construction, I was just tossing things on the ground at random with no clear plan. It can be made much more compact if you intentionally plan for it.

You need three source chests (one for each item) and three transposers to make this work. They're being driven by one timer on a 10 second cycle, which is more than fast enough... also, the regulators emit a redstone signal when they are happy, which I have hooked up to an AND gate that turns the timer off entirely when there's no work to be done. Just a neatness measure.

So is it all good? Well... no. Take a look at this regulator. This is the "I am fully satisfied" steady state. In the grid on the left, you tell the regulator what items to accept from the network, and it also dictates how many to buffer at any time. In the grid on the right, you specify what the chest is supposed to contain. In the middle is the buffer, which the regulator draws from to fulfill the condition mapped on the right. If the condition is filled, the buffer will fill up to match the left side grid. Then the regulator is happy.

Why is this a problem? Because none of these things are ghost slots. They are all physical items. That means with three regulators set up this way, you will automatically have 18 quad thorium cells and 9 plutonium cells (plus 2 depleted isotopes if you let one reactor breed, or 6 if you let all of them breed) tied up in the sorting grids. At efficiency 9 which these reactors run at, that is almost one billion EU in fuel cells sitting there just to specify the sorting. One billion EU you'll never get out of your system. It's also an uneven amount, while the centrifuge produces (and the reactors are using) even amounts. Meaning you have to either produce 9 quad thorium cells somewhere without generating plutonium, or contend with 9 plutonium cells (another 324 million EU) lying around unused in storage.

I will take another look at GregTech automation devices and see if they can do what is needed.

Yeah wish redpower used ghost images. I've got some time to play, going to try rail craft loaders/unloaders. If i recall they allowed stacking and supply specific counts of items -- before redpower 2 was released i had a test breeder setup using them...
 

Shakie666

New Member
Jul 29, 2019
768
0
0
Wow, hasn't this thread come a long way? Who would've thought people would be so enthusiastic to generate EU in the unsafest manner possible eh?

Anyway, after many long hours of tinkering i've finally built a fully automated (well, apart from the fuel), highly powerful DDOS reactor setup, giving off a truly incredible 5760eu/t (beat that CASUC :p)! You can see it in the first attachment below. The 2-chamber reactors (which really ought to be 1-chamber but I was just testing it) are configured like this (the reactor plating is only there to ensure the cooling cells end up in the correct slots). Each reactor needs just under 3.5 cooling towers, for which this is the configuration. This means that you theoretically get away with 28 coolers, but be on the safe side I am using 30.

Now, if I could've done this with CC, I would've, but I don't have the time nor the inclination to learn a programming language I won't use in 99.999% of my life. So instead, I used logic gates to control the whole thing, which you can see in the second attachment. That lever in the top-left is the on/off switch for the whole thing (however, the cell cooling setup will carry on running, but it won't make any difference if all the cells are cool anyway). I could probably also get away with only using 7 repeaters, but if you're on a laggy server it probably isn't a risk worth taking. The state cell is set to 1050s, the timer is set to 2.5s, and the counter is set to a maximum count of 48, an increment of 1, and a decrement of 48 - the total number of cooling cells in the reactors. That wrathlamp is only there to avoid any lag caused by lighting updates.

The main disadvantage of logic gates is that they're a lot less compact than computers. However, they are immune to server resets, don't require a special language to do, and is (in my opinion) more visually impressive.
 

Attachments

  • 2013-03-23_21.25.27.png
    2013-03-23_21.25.27.png
    838.5 KB · Views: 171
  • 2013-03-23_21.36.00.png
    2013-03-23_21.36.00.png
    201.5 KB · Views: 142

Omicron

New Member
Jul 29, 2019
2,974
0
0
Now that is a properly impressive power plant! :)

There's something you should know, though... those logic cells aren't quite as impervious to outside influences as you think they are. The state cell, in particular, resets to zero everytime the global time changes for any reason, such as:

- Sleeping in beds
- Using the /time set console command
- Using NEI's cheat mode to limit the day-night-cycle
- And quite possibly server resets

The timer is affected just the same, but with a 2.5 second cycle time I suppose it doesn't matter as much.
 

Shakie666

New Member
Jul 29, 2019
768
0
0
Now that is a properly impressive power plant! :)

There's something you should know, though... those logic cells aren't quite as impervious to outside influences as you think they are. The state cell, in particular, resets to zero everytime the global time changes for any reason, such as:

- Sleeping in beds
- Using the /time set console command
- Using NEI's cheat mode to limit the day-night-cycle
- And quite possibly server resets

The timer is affected just the same, but with a 2.5 second cycle time I suppose it doesn't matter as much.
Oh, I didn't know that. Well that's bad. Not because it'll lead to a big explosion, since the state cell resetting will just cause the cooling cell replacement cycle. However, that would lower the effective eu/t quite a bit. Is there a way to avoid this (other than computers)?
 

Peppe

New Member
Jul 29, 2019
836
0
1
Wow, hasn't this thread come a long way? Who would've thought people would be so enthusiastic to generate EU in the unsafest manner possible eh?

Anyway, after many long hours of tinkering i've finally built a fully automated (well, apart from the fuel), highly powerful DDOS reactor setup, giving off a truly incredible 5760eu/t (beat that CASUC :p)! You can see it in the first attachment below. The 2-chamber reactors (which really ought to be 1-chamber but I was just testing it) are configured like this (the reactor plating is only there to ensure the cooling cells end up in the correct slots). Each reactor needs just under 3.5 cooling towers, for which this is the configuration. This means that you theoretically get away with 28 coolers, but be on the safe side I am using 30.

Now, if I could've done this with CC, I would've, but I don't have the time nor the inclination to learn a programming language I won't use in 99.999% of my life. So instead, I used logic gates to gontrol the whole thing, which you can see in the second attachment. That lever in the top-left is the on/off switch for the whole thing (however, the cell cooling setup will carry on running, but it won't make any difference if all the cells are cool anyway). I could probably also get away with only using 7 repeaters, but if you're on a laggy server it probably isn't a risk worth taking. The state cell is set to 1050s, the timer is set to 2.5s, and the counter is set to a maximum count of 48, an increment of 1, and a decrement of 48 - the total number of cooling cells in the reactors. That wrathlamp is only there to avoid any lag caused by lighting updates.

The main disadvantage of logic gates is that they're a lot less compact than computers. However, they are immune to server resets. don't require a special language to do, and is (in my opinion) more visually impressive.

Nice setup. Any estimate on the copper used for that? Trippy looking wiring there on the left with the cover strips. You might want to look up insulated wire - switch to alloy at the reactor ;)


@omicron
Looked at some power draws on ME networks. If you trust the display of the ME controller, cables and dark cable use 1/16 an energy rounded down -- like loss in IC2 cables. Thinking about keeping the ME network off until the centrifuge has output. Should do all it needs to do and then be offline most of the time.

Played with railcraft loaders and they are quite nice. Unloaders are giving me a little headache, but sure its just a placement issue. Has the same interface as the loader. Be pretty cool to keep a reactor supplied through rail carts :p

Edit:
Got an advanced unloader working where i wanted -- checking the wiki guess the regular unloader goes under the track. Anyway finding carts are surprisingly nice way to supply a certain level of inventory.

Little system mock up:
http://imgur.com/a/gPE6B

You can add reactors onto the track and carts if you need more storage. Could even pickup depleted cells, but think it would be better to pull those out using any other system, so it is out faster.
 

Peppe

New Member
Jul 29, 2019
836
0
1
Hah quirk to RC unloader + GT advanced regulator. Railcraft counts the ghost items in the regulator as part of the inventory. Was pulling my hair out for 20 minutes on that one. Works great once you account for that :p


Edit:
Went ahead and tried the 4 reactor set.
http://imgur.com/a/KCs3Y
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
So yeah, after examining it closely, I think Peppe won nuclear engineering for 1.4.7 with that design. This thing is simply amazing.

Unfortunately it won't last long, for I'm hearing that current hybrid reactor designs overheat and blow up in the GregTech 1.5 beta, because plutonium cells pulse faster. Ironically, while this is a buff, I wonder if it won't end up making hybrid reactors so difficult that you end up having to build smaller, less effective (or not thorium neutral) setups o_O
 

Peppe

New Member
Jul 29, 2019
836
0
1
So yeah, after examining it closely, I think Peppe won nuclear engineering for 1.4.7 with that design. This thing is simply amazing.

Unfortunately it won't last long, for I'm hearing that current hybrid reactor designs overheat and blow up in the GregTech 1.5 beta, because plutonium cells pulse faster. Ironically, while this is a buff, I wonder if it won't end up making hybrid reactors so difficult that you end up having to build smaller, less effective (or not thorium neutral) setups o_O
Thanks. Was fun to engineer.

Cool to see carts loaders/unloaders actually be relevant and maybe even the optimal solution :p Doesn't really matter if the reactors change unless they get simplified down to one fuel type needed per reactor, then you no longer need to worry about inventory levels and the whole cart + advanced regulator system are not needed at all.
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Doesn't really matter if the reactors change unless they get simplified down to one fuel type needed per reactor, then you no longer need to worry about inventory levels and the whole cart + advanced regulator system are not needed at all.

Well, it would matter if, for example, the extra plutonium ticks caused so much extra heat that certain hybrid designs become entirely uncoolable. The 512 EU/t design that Requia and I threw around for a while proved to be uncoolable without a CRCS or DDoS system, and you're running a 484 EU/t design. There's very little separating these two.

But, until we get a real 1.5 release all this is speculation, really.
 

captainmjs

New Member
Jul 29, 2019
2
0
0
Hi everyone,

I'm kind of a noob at FTB but I've been playing around with a couple of thorium + uranium designs and I was hoping someone would have some feedback as to how to make them less costly/more efficient. I like using thorium cells because they last longer, are more abundant, and aren't big copper hogs. The first design is a pretty simple one that I'm actually using in my world right now to make UU matter (which is how I got into nuclear power in the first place):

172 eu/t (dual thorium/single uranium) 4 chamber
http://www.talonfiremage.pwp.blueyo...0jwu7amknzqulyeulo17k38j9edrc5tlhi5slijfkh14w

The other is one that I am planning to implement in the future (once I get more copper):

332 eu/t (quad thorium/single uranium) 6 chamber
http://www.talonfiremage.pwp.blueyo...4fhkp9ettmmuqhrqnw1752ni71lrda1foe4oga97wiyo0

This thread has been really helpful to me in getting started with nuclear power in the game so thanks. Sorry if someone has already posted designs similar to these I didn't see any of them on the thread but I may have overlooked them.
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Well, we've been discussing automating a plutonium/thorium hybrid selfbreeder on the past page and this one. An uranium/thorium hybrid however hasn't been discussed, you'd be correct there.

The second design you posted isn't half bad, with an 8.3 efficiency score (never trust the reactor planner to calculate it properly with hybrids). But the cooling system can likely be made cheaper.
 

Omicron

New Member
Jul 29, 2019
2,974
0
0
Oh, and on the topic of hybrid reactors in general I posted this over at the IC2 forums today. May be of interest to some, like jumpfight5 who wanted a tutorial of sorts...
 

CodaPDX

New Member
Jul 29, 2019
203
0
1
I haven't had a chance to tool around with it yet, but while I was setting up my tungstensteel crafting the other day I noticed that the gregtech vacuum freezer can be used to repair reactor system components. The energy draw looks pretty modest, too. Has anyone looked into the feasibility of using it to quickly repair cooling cells instead of dedicated cooling reactors?
 
Status
Not open for further replies.