1. Please make sure you are posting in the correct place. Server ads go here and modpack bugs go here

A Guide To: Running Your Fusion Reaction Indefinitely (1.4.6/1.4.7)

Discussion in 'General FTB chat' started by ff255away, Jan 14, 2013.

  1. ff255away

    ff255away Active Member

    Several forumers have completed designs, based more or less around the ideas in this post. Here are links to their builds!


    additionally, masterventis built one that had problems but would be easily fixed (I believe woddorl achieved this).

    I finally have my fusion reactor set up and running on helium-3 and deuterium, but wanted a way to fully automate it. No buttons to press, no restarting the reactor, no hands on. Helium-3 needs something to mine and will eventually run out. This requires continuous effort, no bueno! Lets try the deuterium+tritium reaction, which can be fully automated.
    Note: I had issues getting both cells out of my fusion reactor, meaning this costs me something along the lines of 1 tin cell every 51.2 seconds, or 1 tin every 102.4 seconds. This isn't the case for everybody. You may want to babysit your reactor for several cycles before going totally hands off.

    this requires redpower to operate efficiently, and uses some of their tools to make things as cheap as possible. This is all theorycrafting, so bear with me.

    We need 7 industrial electrolyzers; 18 industrial centrifuges; the ability to fill 5 tin cells with water every 6 seconds; an extractor; and we need a fusion reactor.

    1) 6 industrial electrolyzers turning water cells into electrolyzed water.
    2) 1 industrial electrolyzer turning electrolyzed water into hydrogen (extract the compressed air cell, and send it back to be filled with water)
    3) 15 industrial centrifuges turning hydrogen into deuterium
    4) 3 industrial centrifuges turning deuterium into tritium

    6) extractor to get the cell back from the helium product of H2+H3, OR adding 1 tin cell every 51.2 seconds.

    The explanation:
    You may be asking yourself, "What does a 22 year old guy who just had his nails painted bright blue by a 13 year old girl know about nuclear fusion?" Your answer would be, "jack shit." What this guy does know is how to automate.

    We're going to work backwards, first.
    1) A deuterium+tritium reaction lasts 1024 ticks, or 51.2 seconds.
    2) It takes 150 seconds to turn 4 deuterium into 1 tritium
    3) It takes 150 seconds to turn 4 hydrogen into 1 deuterium
    4) it takes 5 seconds to turn 6 electrolyzed water into 4 hydrogen
    5) it takes 6 seconds to turn 1 water cell into 1 electrolyzed water

    So this is the time cost. The time cost is infinitely more important than the eu cost, for reasons I'm about to explain.
    1) It takes 150 seconds for 1 tritium, and we need 1 every 51.2 seconds. If we have 3 industrial electrolyzers, we can use 1 tritium every 50 seconds, leaving us with a 1.2 second window.
    3) It takes 150 seconds for 1 deuterium, and we need 5 every 51.2 seconds (1 for the reaction, 4 for making 1 tritium). This means we need 15 electrolyzers producing deuterium, resulting in 5 deuterium every 50 seconds.
    4) Now we need to skip back.
    -We need to be producing enough hydrogen to making 5 deuterium every 50 seconds.
    -4 Hydrogen (needed for 1 deuterium) costs 6 electrolyzed water and 5 seconds.
    -1 electrolyzed water costs 6 seconds.
    -We need 5 deuterium every 50 seconds, which would mean 4 hydrogen and 6 electrolyzed water every 10 seconds. Easy!

    So, lets build this thing:
    1) Start with your water feeding machine. I'm not designing an automated water cell filler for you, do that yourself! You need to be able to produce 6 water cells every 10 seconds.
    2) You need to add 2 tin cells from an outside source every 50 seconds. This can either be via uu-matter or an outside source. If you extract the helium cell that results from the H3+H2 reaction instead of storing it away, you only need 1 tin cell every 50 seconds.
    3) Have your first 6 electrolyzers constantly running water cells, to output electrolyzed water.
    4) Set the retriever to pull every 10 seconds. Pull the electrolyzed water and place it in your next electrolyzer.
    5) retrieve the compressed air and put it in an extractor
    6) this is where things change a little. Your going to need rp2 managers. You need to keep 4 hydrogen cells in each of the next 15 centrifuges. Thats a lot of managers.
    7) same deal as 6), you need to keep 4 deuterium in each of the next 3 centrifuges.
    8) I'm going to recommend keeping 2 deuterium and 2 tritium in the fusion reactor at any given moment, because I'm paranoid as hell.
    9) suck out the helium cell, and either extract it or store it away to be used in a different project.
    10) you need a retriever pulling empty cells from all electrolyzers output side, to be filled back up with water.

    Final Thoughts:
    If you don't choose to extract the helium cell, you can add one tin cell every 51.2 seconds, or what would equate to costing you 1 tin every 102.4 seconds. That is an awfully low price to achieve infinite run time in a fusion reactor. The eu cost can be broken down to each reaction, costing 1 tritium and 1 deuterium. So, you have [4 deuterium -> tritium]15000eu+[20 hydrogen -> 5 deuterium]75000eu+[30 electrolyzed water -> 20 hydrogen]15000eu+[30 water cells -> 30 electrolyzed water cells]491520eu=596520eu to run the reactor, with an output of 33,554,432 eu, with a net output of 32,957,912eu every 51.2 seconds. Even simpler, that means you're running at approximately 32,189eu a tick.

    If you're worried about capacity, just run the fuel generation (all of the industrial electrolyzers) several times before turning on the fusion reactor.

    It should be noted that this guide is flawed. The theory is sound, but the theoretical construction was only loosely thought out. If you're going to use managers, you're going to need tons of them; I used almost a whole stack because every input and every output requires a manager. I used 4 industrial electrolyzers with water pumped into their tops- this was a safe number. 3 extractors, with 2 overclockers each were enough to keep up with everything, with plenty of time to kill. Always keep at least 2 tritium and 2 deuterium in your reactor- you will be very unhappy otherwise. Finally, it doesn't seem like anybody has engineered a solution to preventing the tritium cells from stealing all of your cells. Make sure to have a number of tin cells feeding into your system constantly.
  2. Greytusks

    Greytusks Well-Known Member

    You actually built one of these?

    I'm not even sure I will ever get that far up the tech tree. (especially in Gegtech hard mode)

    Seems like a nice setup, reading about it.
  3. TheLoneWolfling

    TheLoneWolfling Well-Known Member

    I suggest adding extra capacity throughout - you REALLY don't want to have to restart the reactor.
  4. TruculentMC

    TruculentMC Active Member

    Scrapboxes are a MUCH cheaper source of the needed Tin than UUM.
  5. Celestialphoenix

    Celestialphoenix Too Much Free Time

    Now all we need is a frame machine/turtle to automatically set up one of these, and a factory to produce the components....
  6. Honza8D

    Honza8D Active Member

    i dont get it, why do you need extra tin? You dont lose any tin in the process, in fusion reactor you get both tin cells back, and the compressed air/helium can be exctracted in extractor.
  7. ff255away

    ff255away Active Member

    You get both back? It doesn't seem to be working properly for me, then. I'm only getting the helium cell back. I'll make a modification to the title post, then.
  8. DoctorOr

    DoctorOr Popular Member

    1 in 100 scrap boxes will generate a tin dust, which is 900 scrap, which is 7200 inputs to a recycler.

    Every 100 seconds.

    Which means you need over 150 recyclers.
  9. TruculentMC

    TruculentMC Active Member

    Did you check all the Scrapbox output recipes? You can see these from the Scrapboxinator UI by clicking the arrow. Tin Dust isn't the only source as you can get various cells from it now.

    Edit: And powering even 150 recyclers is still MUCH cheaper than generating the Tin with default UUM rates in GregTech so I stand by my statement... though it's not even needed if you get both the cells back from the Fusion Reactor as it's intended to work.
  10. Honza8D

    Honza8D Active Member

    well it works for me (mindcrack pack, no modifications)
  11. masterventris

    masterventris Active Member

    Ok I built this. It works but occasionally the fuel only just arrives before the reactor shuts down, but that is probably as I cheated and used logistics pipes instead of precise timing using RP.

    Btw, you need centrifuges not electrolyzers to produce deut and trit.

    Anyway, screenshots.
    Here is the whole setup, building is incomplete.


    Reactor and feed pipes, power goes along superconductors out through the wall. MFSUs and AESU are for ignition power.


    Aqueous accumulator and 2 liquid transposers to fill empty cells with water. This is far faster than necessary with an upgraded electrical engine running each transposer.


    Row of 6 electrolyzers producing electrolyzed water, single electrolyzer producing hydrogen.


    15 centrifuges producing deuterium. This the bottleneck as the logi pipes tend to fill a machine, it starts running, then they fill it again before using the next, meaning they arent running at the same time.


    3 centrifuges at the top producing tritrium. Similar issue as before, except as these are closer they actually steal fuel from the reactor. These are full of deut with none in the reactor.


    HV transformers taking the output of the reactor into the IDSU. This handles 8192 eu/t which is less than the capacity of the reactor as stated above. About a quarter of it apparently.


    These 2 extractors empty the compressed air and helium which are byproducts of the system. i could use the helium as coolant for a nuke reactor, maybe when can use 35 million eu in a minute...


    Any questions about this? maybe ill make it with RP for the challenge.
  12. TheLoneWolfling

    TheLoneWolfling Well-Known Member

    What type of log pipes are you using for machine inputs?
  13. ff255away

    ff255away Active Member

    @masterventris You have the basics of my proposed design down, but you're absolutely right about it needing rp2 tubes. Redpower is the only mod at the moment that can time things as they are needed to be timed for the fusion reactor. You'd need a whole lot more electrolyzers, centrifuges, buffers and eu cost to get things functioning properly with logistic pipes. I'll be fixing the main post about the centrifuge information.
  14. masterventris

    masterventris Active Member

    They have supplier pipes on top set to maintain the required number of resources for 1 operation. Extraction is done by a logistics chassis with a provider module so the machines are only emptied when another calls for resources.

    I can also confirm that all cells are not lost in this setup. The fusion reactor takes the 2 inputs and immediately outputs 1 empty cell, then 1 helium cell at the end of the cycle. I think if you dont remove the empty then it gets replaced by the helium, requiring it to be replaced.
  15. ff255away

    ff255away Active Member

    Ah, makes sense. Must have been my issue then, only set the retriever on my previous set up to pull the product out and not any empty cells. I've gotta hold off on adding your pictures to the main post, working from my phone for the next little while.
  16. Korenn

    Korenn Well-Known Member

    Awesome setup! nice endgame device :)

    I have to wonder though, why don't you just set the pipes to maintain resources for 2 operations, including the reactor? It'll mean some extra cells present in the system but that should hardly be a problem.
  17. masterventris

    masterventris Active Member

    I had, as ff255away suggested, however it takes longer than a cycle to produce 1 set of fuel, so after about 15 cycles of testing the reactor overtook the supply and it shut down.

    The bottleneck is deuterium, I would have to see if the electrolyzers can produce more hydrogen. Maybe a few more electrolyzers is easier than building the managers and setting it up with RP.
  18. TheLoneWolfling

    TheLoneWolfling Well-Known Member

    I am a little confused as to what you mean by that... So it isn't self-sustaining with resources for 2 operations but is with resources for 1?
  19. Squigie

    Squigie Active Member

    Is there a reason to make hydrogen from electrolized water cells given that water can now be directly piped into the ind. electrolizer for a 1-step process? I did a bit of calculation with numbers from NEI.

    For 4 H cells:
    6 water - 38s, 93000 EU
    6 elec. cells - 5s, 3000 EU
    6 water cells - 36s, 98304 EU
    Total - 41s, 101304 EU

    Unless I made a mistake, piping water is cheaper, quicker and simpler than using water cells.
  20. silenos

    silenos Member

    I figured out something similar, at the very least its easier to automate :)

    Here is my system I threw together in creative:

    The complete system (don't mind the stuff in the background, it's a testworld)

    4 ind. electrolyzers to produce Hydrogen

    the reactor with inputfilters for Trit and Deut cells (you can't use managers here because they emit a redstone signal when satisfied, preventing the reactor from outputting energy)

    and finally the wall of centrifuges to get Deuterium and Tritium

    so far it is a pretty stable system, but it will over produce a very slight amount of Deuterium, Tritium and Hydrogen cells, but you could just add additional extractors to empty them.

Share This Page