Experimental Changes to Nuclear Engineering

  • Yeah, it's only EU that's ramping up. If heat output was variable, it would be near impossible to hit the equilibrium point without accidentally slipping over or under it.


    I've had a similar idea for a potential reactor as you did, though I'm not sure if I'll actually build it anytime soon. As "cheap and easy" as these look, the MOX fuel itself is actually insanely expensive. For the 16 cells in that setup, you would have to run 432 normal uranium fuel rods through a reactor. Even with large reactors, that takes quite a long time. By the time you can actually build the cells, the cost of the reactor is probably entirely irrelevant (at minimum you're going to have tons of components from your other reactors that you could be reusing).


    For comparison, in a legit world I have so far run two cycles of this and one cycle of this, and I do not even have enough plutonium for one MOX rod, much less 16.


    Thankfully the MOX rods, once you have them, will kind of recycle into themselves so long as you have enough U-238. That 2x2 quad MOX setup will consume roughly one hundred U-238 over one and a half hours.

  • As "cheap and easy" as these look, the MOX fuel itself is actually insanely expensive. For the 16 cells in that setup, you would have to run 432 normal uranium fuel rods through a reactor.

    You can offset this cost somewhat by using your MOX as soon as you get it. Build the reactor according to Shneekey's specs but build it as a standard 235U reactor. As soon as you get enough Pu for your first quad MOX cell use it in place of one of the quad U cells. Both MOX and 235U yield a small wad of Plutonium when recycled but MOX has half the burn time so it produces Pu twice as fast as 235U. You'll have to shut down the reactor half way through a cycle to recharge your MOX though. I'd also be tempted to build a single-block reactor for the purpose of burning MOX that was just hanging around but wasn't enough to make a quad cell out of. Using those methods you'd have enough to switch your reactor over to MOX entirely after 200 fuel rods, taking only 19 cycles and requiring a mere 450 Uranium ore. Still a hefty price tag but better than half off the initial 432 fuel rod figure.


    Of course, if you want to jump-start your MOX program, you can just build some brutal beast like this and have enough for a MOX quad core in one cycle.

    Even with large reactors, that takes quite a long time. By the time you can actually build the cells, the cost of the reactor is probably entirely irrelevant.

    Yeah, that's the same problem I had with breeders. Sure I can build one and run it until I have fuel coming out of my ears but what am I going to use it for? By the time I'm done making the stuff I'm going to have a produced few GEU and where am I going to store that and what will I ever need more for? Short of building castle out of Iridium or teleporting to one of the Jovian moons that sort of energy doesn't have much of a use.


    -S

  • Hey, making a full quantum suit for yourself and possibly a few others will eat quite a bit.


    Now if only I could actually find some iridium ore myself... *piles 8 music disks, 12 diamond horse armor, 10 golden horse armor, 11 iron horse armor, 7 name tags, 15 saddles and 3 valiant drones from dungeon chests in the corner*

  • Better than: Start new world, get into dungeon, finds Iridium ore, brings home, creeper invasion, chest gone, iridium gone D:

  • Actually, a CRCS style reactor would work phenomonally well with MOX fuel, as the actual temperature of the hull never changes during use.

    You are correct Shneekey. I currently have a very slight modification of your MOX reactor design running in the background cranking out 3378 EU/tick from four quad MOX cells. I was feeling lazy so I have the whole setup being cooled by a battery of your CoolMasters. It isn't graceful but I'm getting almost as much power and nearly half the energy that the largest possible Uranium reactor will generate.


    Quote

    Of course, your automation system would have to be pretty tight to avoid explosions...

    On this point I am happy to be able to tell you that you are mistaken. I've run that reactor for a full MOX cycle at 4 heat below the absolute critical threshold using nothing more than an ME import and export bus to swap coolant. As I'm sure you know, ME networks are not designed for exceptional speed of component transfer and frequently have multi-second lags between removal of depleted coolant and insertion of fresh cells. With a very simple modification, your reactor design becomes extremely robust with respect to automation lag. So much so that when I forgot to swith the coolant export bus back on I had a good half minute to muck about finding the problem, all with the reactor running at 4 heat below BOOM level.


    The trick is to cook half the coolant cells until they're half done. You start your reactor with no coolant to allow it to bake up to the desired operating temperature. Then you insert half the coolant, making sure to get one coolant cell per fuel cell. Switch your reactor back on and let it cook the coolant to about half the point at which it gets swapped out. Now add the remaining coolant and continue as normal. The result is staggered coolant swapping the leaves your reactor with coolant 100% of the time unless you hit automation lag exceeding a minute or so. When the first batch of coolant reaches the end of its duty cycle and swaps out the second half is still only half-baked. It remains in the reactor and takes up any slack during swap-out. Similarly, when the second batch swaps out, the first batch has been refreshed and is still at 50% or so. This means that no cell is ever without coolant so no heat ever accumulates due to coolant swapping lag.


    Swapping all the coolant at once gets you a ~2000 heat spike each time. Swapping them staggered lets me run the reactor for a full cycle without a single uptick in heat. This will work for any reactor design with at least two coolant cells per fuel cell, though it's easiest and safest if all coolant cells are heating at the same rate.


    One word of caution. A slow drift induced by random swap lag may lead to one set of coolant gradually catching up to the other. In several reactor cycles this could lead to a synchronous coolant swap that would produce a nasty crater. It's advisable to resynchronize your coolant staggering at the beginning of each reactor cycle to avoid this.


    -S

  • So I have a very nice reactor cranking out 3378EU/tick. That's more than enough power to blow an unprotected MFSU to fine powder. I'm currently running three MFSUs off of an EV transformer connected to the reactor but I suspect that I'm not capturing all the energy. Does anybody have any suggestions?


    More worryingly, I could build this horror using MOX (mwah-ha-ha-ha). By my estimate it wold generate just shy of 19,000 EU/tick, more than twice the load an EV transformer can handle.


    Thots? Other than "Don't do it!"?
    -S

  • So I have a very nice reactor cranking out 3378EU/tick. That's more than enough power to blow an unprotected MFSU to fine powder. I'm currently running three MFSUs off of an EV transformer connected to the reactor but I suspect that I'm not capturing all the energy. Does anybody have any suggestions?


    More worryingly, I could build this horror using MOX (mwah-ha-ha-ha). By my estimate it wold generate just shy of 19,000 EU/tick, more than twice the load an EV transformer can handle.


    Thots? Other than "Don't do it!"?
    -S


    Don't start it!

  • Rgarding capturing all the energy: remember that the transformers no longer output multiple packets down a line. The EV transformer can only send 2048 EU/t on the output side even if you're using a 8192 EU/t cable and multiple receiving MFSUs. It might possibly work that you can have multiple independent lines coming from the same transformer so it'll output more, but potentially you'll need multiple transformers as well. I haven't tested it yet.

  • Rgarding capturing all the energy: remember that the transformers no longer output multiple packets down a line. The EV transformer can only send 2048 EU/t on the output side even if you're using a 8192 EU/t cable and multiple receiving MFSUs. It might possibly work that you can have multiple independent lines coming from the same transformer so it'll output more, but potentially you'll need multiple transformers as well. I haven't tested it yet.

    For his setup, he will need at least 2 EV transformers, and 2 MFSUs as to not cause a bottleneck.

  • Maybe they should reintroduce passive reactor cooling? For each 1000 of reactor hull heat it will increase cooling by one point. So reactor heated to 65000 will cool down by 65 heat each tick. This way we can project automatic heating reactors which are auto regulate them self. Or some other way to make reactors auto regulate.

  • Sirus I believe said that the reactor system as a whole is pending a complete overhaul... so who knows where it'll go?

  • Just a thought - with MOX, if you can somehow measure how many EU/t the reactor is giving off, you effectively measure the temperature of it. Something that emits a redstone signal in the presence of too much EU/t could control a reactor temperature.
    Something like an MFSU (set to emit EU to another MFSU with plenty of spare space) set to "emit when partially full" would be one way to do it.... Although I've seen those things lose their settings on reload sometimes.

  • I Suggest ComputerCraft with OpenPeripherals. With a little programming you can not only measure temperature and output, but even automate the whole thing. Like removal/Insertion of rods and cooling cells at defined temperatures, emergency shutdown or pausing if energy would be wasted. You wont ever need to get near the danger zone anymore ;)

  • Lots of the brawnier setups here require some degree of automation.


    For those who can't (or won't) do that, here is the best I can come up with.


    It's 100% thermally stable (+-0 heating), so once you have it heated up to 85%, it STAYS there with no automation required.
    Heat-up is swift if you remove the overclocked vents, takes me about 40 seconds or so to boot it up from zero.


    Output COLD: 270 EU/t
    Output HOT: 1,177 EU/t (with MOX instead of Uranium, of course.)
    Efficiency 3.86



    The empty slot in the lower right is meant for a Reactor Heat Exchanger, just incase you overshoot the 85% heat during warmup.


    I know, this thread is kinda half a year old now, but I believe the design fits the topic best.
    If you know of a thread that could use my design better, feel free to post it there or let me know. :0)