[Official] New Reactors design thread.

  • Efficiency after factoring reflector costs: 2.97...


    Rule of the thumb: consumable reflectors are never worth it. GregTech iridium reflectors almost never.

  • Yes, it's bad. Expensive, low-efficiency, space-wasting, and unsafe (will explode after an hour or two due to heat accrued during swapping of condensators). You could never use this reactor ingame...


    I recommend you familiarize yourself with the stickied reactor design guidelines, as well as the designs on the first page of this thread, and try to understand why they are built like they are. Nuclear reactors are kind of a separate minigame inside IC2, and making a good reactor from scratch requires knowledge of the rules you need to play the game by.

  • After much effort and a few craters, I finally managed to produce a fully automated, stable, high-T, LZH-cooled breeder. It requires a few extra mods, namely:


    Nuclear Control add-on 1.5.1e (but you're a fool to muck with reactors without it. Even Chernobyl had a thermostat)
    Gregtech add-on 3.11
    Applied Energistics rv13.dev1


    NB: I'm still using Minecaft 1.5.2 so all this is being done under IC2 1.116.373 but it should be pretty similar to the latest, greatest version.


    The following breeder design has a high through-put if you can manage to automate it:


    http://www.talonfiremage.pwp.b…qyidz5t5569m2f9ju7n3yw1t9


    Output values are listed as theoretical (measured) where the measured values were obtained over the course of one uranium reactor cycle (10,000 s).

    • 240 EU/tick (216 EU/tick after taxes)
    • Efficiency 3.00
    • Overall Efficiency ? I'll let somebody else call this one.
    • Running costs: 160 copper, 240 lapis (235 lapis)
    • Speed 11.16 s/uranium for 896 recharged cells/cycle (11.49s/uranium of 470 charged cells/cycle after lag)

    This type of reactor requires automation unless you really enjoy standing around for just short of 3 hours popping out charged uranium cells and spent condensators. Simple automation typically works poorly because it becomes very difficult to keep the coolant cells and the isotope cells in their proper locations.


    The solution that I employed was to use 6 Advanced Buffers from the GregTech add-on. The buffers allow you to specify which cell of the inventory (reactor core in this case) the item should be placed in. This removes the problem of getting the wrong reactor components in the wrong locations. There's one buffer dedicated to each of the condensators and isotope cells and one for the empty cell to the right of the fuel.


    The buffers are supplied by an Applied Energistics ME network which also does the job of removing recharged U cells and spent LZH condensators from the reactor and recharging the condensators with lapis using the network's auto-craft function. Both the Advanced Buffers and the ME network are pricey pieces of equipment but they're one-time costs and the ME network can do a host of other tasks while nursing the reactor.


    The final challenge faced by automation of reactors using coolant cells is that of progressive heat rise due to lack of coolant while cells are being changed. To correct for this a reactor heat vent was added and removed from the empty cell in the design using an Advanced Buffer and ME import and export buses regulated by the redstone signal from a remote thermal monitor.


    The reactor was initially run up to a temperature of 81,000. You can use heating cells in the empty slot if you like but I prefer to just leave the coolant out and set the thermal monitor to kill the reactor at the target temperature. Two remote thermal monitor were then employed. One was used in the usual manner and set to send a SCRAM signal to the reactor at 82,300 heat (critical-1000). The second thermal monitor was set to 81,020 and the redstone signal from that was sent to the export bus feeding the reactor heat vent to the Advanced Buffer dedicated to the empty reactor cell and to an import bus that removed the same component. At T>81,020 the export bus activated while the import bus deactivated, placing the cooling fan in the reactor. Once temperatures fell to T<81,020 the export bus switched off and the import bus switched on, removing the reactor heat vent back to the ME network.


    This maintained core temperatures between 81,000 and 82,300 for the entire fuel cycle. The first remote thermal monitor never sent a SCRAM signal. I can provide video if anybody really wants but suspect it'd be a pretty dull watch.


    With the addition of four more Advanced Buffers the following reactor, which is a good bit more efficient, should be able to produce 1200 charged cells/cycle at temperatures between 72,000 and 75,000.

    http://www.talonfiremage.pwp.b…4i2aimvj1nyi3dejhobtd</a>


    I have not yet tested this design but will post further developments.
    -S

  • And yet with all that needless complication, it still can't really compare to any of the breeder designs listed so far. Heck, I just cobbled together a Breeder in about five minutes that tops this, as shown here.

    Hi Shneekey,


    That depends entirely on how you're making your comparison. If I wanted a simple set-it-and-forget-it breeder with great efficiency for cells charged/U burnt then your design wins hands down. If I wanted a reactor that charges cells fast, albeit at a reduced charge/U efficiency, then my first design wins by more than an order of magnitude. My design also outputs significantly more EU/tick and has better generating efficiency but those aren't really factors that I'm interested in since my objective is a fast breeder.


    Please also take into consideration that the complication is not necessarily needless. If you want a dead simple design then yes, it's worse than needless, it's a waste of resources. If I want to solve an interesting and challenging problem like swapping out multiple reactor components including coolant without thermal rise that would make my shiny high-T reactor into a low-elevation glass bowl :cursing: then the complexity is rather the point of the exercise. It's simple enough to go to the Reactor Planner, plunk in fuel and isotope cells, divide the heat by five and add that many reactor heat vents. Here is a design that's based off the one that you posted but using Th in place of U, using 12 fewer heat vents and getting much better efficiency. It also runs five times longer. It took me all of about as long to design it as you did with yours but was a pretty joyless exercise in remedial math MinecraftZombie. It's great if you want dead simple but really, where's the fun in that ? :D

  • Hi Shneekey,


    That depends entirely on how you're making your comparison. If I wanted a simple set-it-and-forget-it breeder with great efficiency for cells charged/U burnt then your design wins hands down. If I wanted a reactor that charges cells fast, albeit at a reduced charge/U efficiency, then my first design wins by more than an order of magnitude. My design also outputs significantly more EU/tick and has better generating efficiency but those aren't really factors that I'm interested in since my objective is a fast breeder.


    Please also take into consideration that the complication is not necessarily needless. If you want a dead simple design then yes, it's worse than needless, it's a waste of resources. If I want to solve an interesting and challenging problem like swapping out multiple reactor components including coolant without thermal rise that would make my shiny high-T reactor into a low-elevation glass bowl :cursing: then the complexity is rather the point of the exercise. It's simple enough to go to the Reactor Planner, plunk in fuel and isotope cells, divide the heat by five and add that many reactor heat vents. Here is a design that's based off the one that you posted but using Th in place of U, using 12 fewer heat vents and getting much better efficiency. It also runs five times longer. It took me all of about as long to design it as you did with yours but was a pretty joyless exercise in remedial math MinecraftZombie. It's great if you want dead simple but really, where's the fun in that ? :D


    I think you miss the meaning of my post completely.


    Do recall whom you are talking to. I invented CRCS reactor design, which is heavily based on the transfer of components, including nuclear and coolant material. I can appreciate the complexity involved, I followed how, where, and why you did what you did. However none of that makes LHZ's any less horrible under any situation, PARTICULARLY the one you've tried to apply it to.


    First off, it's not going to break even on power, even WITH a couple of quad cells. Not with the amount of Lapis you'll be chewing through, needing to be replaced with UUM, which requires EU to create. Hell, you're talking GregTech here... you know, some ten to hundred TIMES the EU cost of 'core' IC2... you're ridiculously energy inefficient. You'd have to build another battery of reactors just to keep it supplied! This DEFEATS the purpose of a breeder... to be efficient in your use of uranium.


    The breeder I cobbled together in less than a minute was a sarcastic example thrown in your face to demonstrate just how ludicrous the design is. The one advantage the breeder I linked (fully admitting it's probably one of the worst designs ever submitted on these forums... which is saying quite a bit) is that it at least doesn't cost energy to maintain. Yours does.


    You claim it generates more EU, but it costs you more than you generate in keeping those blasted LHZ's filled. If you were to use those same item-transferring methods with cooling cells and swapping them out, you'd obviate and negate a large chunk of my problem. In fact, it's how I got started on the CRCS design to begin with... I saw the huge trap that LHZ's presented, and I sought a more efficient solution. Same principle would apply in your case.


    Also, Thorium stinks as a breeding material these days, don't trust everything the planner says, particularly not when it comes to GT components. The formulae have changed significantly since the planner was last updated.

  • I think you miss the meaning of my post completely.

    That's entirely possible, but I suspect what we have here is more a difference of perspectives.


    Quote

    Do recall whom you are talking to.

    No. Kinda tricksy to recall something I don't know. No offence intended. I'm a noob. Does it show? :D


    Quote

    I invented CRCS reactor design

    Congrats. What's CRCS? No, seriously. Like I said, I'm a noob.


    Quote

    However none of that makes LHZ's any less horrible under any situation, PARTICULARLY the one you've tried to apply it to. <snip> First off, it's not going to break even on power, even WITH a couple of quad cells. Not with the amount of Lapis you'll be chewing through, needing to be replaced with UUM, which requires EU to create.

    See, there's the fundamental difference in perspective that's causing this. You're measuring the efficiency of this design based on the energy cost of the UU-matter needed to replace the lapis that the LZH condensators burn through. From that view-point you are entirely correct. That reactor design would consume a hideous amount of energy to keep running.


    I on the other hand have no intention whatsoever of wasting valuable UU-matter to make lapis. I understand that UU-matter is a convenient, if somewhat flawed, conversion for various forms of matter/energy but why would I actually exchange the one for the other? I am in the position of having little use for lapis and great gobs of the stuff taking up space in my storage. If I use LZH condensators then I can effectively convert lapis that is little more than mine waste into large volumes of reactor cells. At the end of the day I still have great gobs of lapis, marginally more storage space and a whole lot more reactor fuel. Can you point out the downside to me? I'm having a hard time seeing it.


    Quote

    This DEFEATS the purpose of a breeder... to be efficient in your use of uranium.

    That's precisely what I'm doing. 75 Uranium input, 470 uranium cells output at the cost of 235 lapis burnt and gone forever. That may not be terribly efficient use of lapis but I don't care about it and would have to have my head checked if I stared converting Uranium to energy to UU-matter and then to lapis. Hell, I'd probably need my head amputated for reasoning that defective.


    Quote

    The breeder I cobbled together in less than a minute was a sarcastic example thrown in your face to demonstrate just how ludicrous the design is.

    And how was that useful? Constructive criticism would have been much more effective. If you'd mentioned your objection to my design, that LZH burns through lapis, we could have cut straight to a meaningful discussion of the pros and cons of LZH coolant and its potential applications in a lapis-rich economy.


    Quote

    If you were to use those same item-transferring methods with cooling cells and swapping them out...

    I considered that method but you need some means of cooling the cells back down again. That more or less requires a dummy reactor, heat vents, etc. All expensive components to build. Granted, it's a one-time cost but given the relative rates of heating in the reactor (480 heat/s/cell in my design) and cooling in the dummies (can you cool faster than 16 heat/s/cell with efficient use of space?) that's a 30:1 ratio. Assuming you can use half the space in a dummy for cooling cells back down, the total tally comes out to two full blown 6-chamber reactors, one smaller single-block reactor (damn those extra six cells!), 60 or so component heat vents and 64 coolant cells in order to maintain thermal equilibrium for a full cycle. It might be worse given poor cooling efficiency around the margins of the dummy reactors. Let me know if there's a better way to chill the coolant back down.


    You could reduce that by storing heated coolant cells for later cooling but you can't run your reactor while you do. In terms of time you'd probably save less that way than by simply using a Mark III (or higher) with inactive periods to allow the reactor to cool itself and you'd certainly spend a whole lot more resources so it's a non-starter. The only way that would make sense is if you had short-term demand for large quantities of Uranium followed by longer hiatuses but I don't see that as being terribly likely.


    At the end of the day I decided that I'd rather burn the lapis than build a whole lot more equipment. The latter is a more sustainable choice given that lapis is theoretically non-renewable but given my existing stockpile of the stuff, its lack of significant alternative uses and the rate at which I'll find more I'll probably run out of Uranium to make into DI cells first.


    Quote

    Thorium stinks as a breeding material these days

    Thanx. I was tinkering with it earlier today and noticed that it'd been nerfed. Plutonium looks like it might have some promise but the heat becomes trickier to manage. All in all I'm not hugely impressed by GT so far.

  • All in all I'm not hugely impressed by GT so far.

    It got revamped and Pu was awesome, but it got changed bacl because it causes issues with the Reactor Planner and some other things. I hope it'll be overhauled again a day ^^


    Soon with Molten Salt Reactors, right ? :D
    NERF THA FUSION REACTOR!

  • Congrats. What's CRCS? No, seriously. Like I said, I'm a noob.


    Okay, you might want to check out this thread for the blind stumbling around to get to CRCS, with this being a much better explaination of the theory and design, even if it is still written back in the day when RP2 was pretty much the best way of moving things around.


    Quote

    See, there's the fundamental difference in perspective that's causing this. You're measuring the efficiency of this design based on the energy cost of the UU-matter needed to replace the lapis that the LZH condensators burn through. From that view-point you are entirely correct. That reactor design would consume a hideous amount of energy to keep running.


    I on the other hand have no intention whatsoever of wasting valuable UU-matter to make lapis. I understand that UU-matter is a convenient, if somewhat flawed, conversion for various forms of matter/energy but why would I actually exchange the one for the other? I am in the position of having little use for lapis and great gobs of the stuff taking up space in my storage. If I use LZH condensators then I can effectively convert lapis that is little more than mine waste into large volumes of reactor cells. At the end of the day I still have great gobs of lapis, marginally more storage space and a whole lot more reactor fuel. Can you point out the downside to me? I'm having a hard time seeing it.

    You aren't going to be able to pull up that much Lapis constantly. I flat guarantee it. Sooner or later, you will run out. And when you do, you'll be running on UUM for your Lapis needs. Furthermore, Lapis is by no means 'mine waste' if you ever bother with anything in IC2. Doubly so if you use Tinkers Construct (which you probably don't, since you use GregTech).


    Quote

    That's precisely what I'm doing. 75 Uranium input, 470 uranium cells output at the cost of 235 lapis burnt and gone forever. That may not be terribly efficient use of lapis but I don't care about it and would have to have my head checked if I stared converting Uranium to energy to UU-matter and then to lapis. Hell, I'd probably need my head amputated for reasoning that defective.

    Efficient in your uranium USAGE does not mean maximum uranium output, it means most economical use of your uranium. And once your Lapis runs out... and it will... you'll be so far behind on your power budget that you'll never recover.



    Quote

    I considered that method but you need some means of cooling the cells back down again. That more or less requires a dummy reactor, heat vents, etc. All expensive components to build. Granted, it's a one-time cost but given the relative rates of heating in the reactor (480 heat/s/cell in my design) and cooling in the dummies (can you cool faster than 16 heat/s/cell with efficient use of space?) that's a 30:1 ratio. Assuming you can use half the space in a dummy for cooling cells back down, the total tally comes out to two full blown 6-chamber reactors, one smaller single-block reactor (damn those extra six cells!), 60 or so component heat vents and 64 coolant cells in order to maintain thermal equilibrium for a full cycle. It might be worse given poor cooling efficiency around the margins of the dummy reactors. Let me know if there's a better way to chill the coolant back down.

    Check that second thread I mentioned. There's a simple one-chamber coolant reactor which has a very good cooling micro-cycle.


    Quote

    You could reduce that by storing heated coolant cells for later cooling but you can't run your reactor while you do. In terms of time you'd probably save less that way than by simply using a Mark III (or higher) with inactive periods to allow the reactor to cool itself and you'd certainly spend a whole lot more resources so it's a non-starter. The only way that would make sense is if you had short-term demand for large quantities of Uranium followed by longer hiatuses but I don't see that as being terribly likely.

    No, I'm suggesting using AE network to literally decentralize the cooling of cells to multiple simultaneous cooling reactors.


    Quote

    At the end of the day I decided that I'd rather burn the lapis than build a whole lot more equipment. The latter is a more sustainable choice given that lapis is theoretically non-renewable but given my existing stockpile of the stuff, its lack of significant alternative uses and the rate at which I'll find more I'll probably run out of Uranium to make into DI cells first.

    You're going to go through that stockpile a LOT faster than you think you will.


    You seem to think that your Lapis will never run out. I know better.

  • Thanx for the links. I'd actually read your CRCS for Utter Noobs earlier in the day. A few numbers that I crunched:


    Kenken244's coolant tower design sheds an impressive 568 heat/s but only holds 6 cells, so it's cooling the cells just shy of 95 heat/s. I fumbled around with a few designs similar to it last night and found that they work wonderfully while your cells are cooling down from red heat but once you approach the last few hundred heat in the cell the heat stored in the OCHVs and heat exchangers starts flowing back and forth through the system and actual cooling slows and approaches nil. Normally I'd say to blazes with it, the cell's 99% cooled, and ship it back out but my antiquated version of AE has that bugged import bus that can't distinguish between damage levels. But that's a separate issue.

    Your Coolmaster
    is the type of checkerboard design I was contemplating. Incidentally, why did you block the cell in row 4, column 9? Once you've excluded the two low-efficiency cells in the bottom corners, that design has space for 25 cells but only cools 14 of them at 16 heat/s. The 11 cells around the edges only shed 12 heat/s. Total cooling rate is 356 heat/s but there's no heat stored in components so my ME network works with it.


    My reactor design puts out 960 heat/s, so it'll evaporate a pair of 60k coolant cells in exactly 125 s. Instead, let's swap them out at a nice even 120 s with 57600 heat each.


    Kenken's cooling tower would cool that pair of cells off in 609 s. In that time, the reactor will have swapped out another 10 heated cells that require cooling so I'll need two cooling towers and 14 60k coolant cells to keep the reactor running constantly. There's an extra 9 seconds building up there but those cells can cool back to zero while I'm refueling the reactor.


    The Coolmaster fairs worse. Since the inventory fills top to bottom and left to right the first five cells land in 12 heat/s cooling slots. These take 4800 s, an appreciable portion of a reactor cycle, to cool back to zero. The coolant that lucks into 16 heat/s slots manages it in a mere 3600 s, exactly an hour. So the Coolmaster doesn't start returning fully recharged cells until 3960 s have elapsed, at which point the cell that was swapped out at T=360 and landed in slot 6 finally reaches zero heat. After that it returns two cells every 120 s with pauses for the luckless ones that landed in the 12 heat/s slots. They'll catch up eventually. In 3960 s, the reactor will have swapped out 66 hot cells so I'll need at least 3 Coolmaster towers to keep running full-time, possibly four depending on how those low-efficiency slots play out.


    So, total cost for two of Kenken244's design: 1748 Cu, 164 Sn, 552 Fe, 240 Au, 230 rubber, 88 redstone, 4 glowstone and 4 lapis for the cooling towers and another 72 Cu, 510 Sn for the 12 60k coolant cells


    Total cost for 3 Coolmasters: 876 Cu, 336 Sn, 819 Fe, 21 rubber, 24 redstone and 6 each of glowstone and lapis for the towers and a minimum of 68 60k cooling cells running 544 Cu and 3060 Sn.


    Either way it's a pretty hefty price tag. Granted, those are one-time costs but they're not small costs.


    Quote

    You aren't going to be able to pull up that much Lapis constantly.

    Two points here:

    • Are you sure?
    • Who said anything about constantly?

    Point number 2 first. In 10 cycles, about a day of game play, I'll be able to charge 8700 DI cells with that reactor. Ignoring the costs of running the reactor, that's 1088 U, 4350 Sn and 17400 coal plus refining costs just to craft the cells. And what am I going to do with 8700 U cells?!? At an efficiency of 4 that's 34.8 GigaEU, assuming you could ever build enough reactors to burn it all.


    But let's assume that I have some pressing need for effectively infinite U cells. My reactor burns 16 U and 235 lapis to produce 870 U cells. Nothing we can do about the U since we can't fabricate it from UU-matter so my starting efficiency is 6.97 cells/U. Now I need to make the lapis from UU-matter at a cost of 4 UU to 9 lapis for a total cost of 105 UU-matter/cycle, or 105 MEU/cycle. My reactor puts out 43 MEU/cycle after running costs so It has a deficit of 62 MEU/cycle. A smartly built Mark I EA using four quad cells will give you 64 MEU/cycle at a cost of 16 more U cells. So final efficiency after deducting 235 lapis in UU-matter is 854 U cells crafted using 122.75 U, 6.85 cells/U. Net difference of -0.128 cells/U.


    Assuming all costs including copper, tin and coal (who in the name of all that's unholy turns UU-matter into coal?!?) including the copper for the dense copper plate to make the quad cells for the Mark I EA are covered by the most extravagant expenditure of UU-matter the final bill is 576 U cells, leaving 294 for a net efficiency of 2.36 cells/U.


    So yes, even spending my resources like a drunken sailor I can do this constantly and continuously with room to grow.


    Much more effectively though, I can go out and mine lapis, copper, tin and coal. The breeder put out 43 MEU and a quartet of miners runs about 5 MEU after processing the ore. So 8 mine moves later I'll have burnt through that 43 MEU, give or take, and I will have extracted all the lapis in a 36 by 72 by however deep it goes volume. To break even I need just under four stacks of lapis from that effort, or a mere 8 lapis/miner/location. I'll also get the copper, tin and coal I need. Honestly, the only thing I'm likely to run out of is Uranium.


    Quote

    Furthermore, Lapis is by no means 'mine waste' if you ever bother with anything in IC2. Doubly so if you use Tinkers Construct (which you
    probably don't, since you use GregTech).


    I can spend the occasional shard on advanced circuits, lapotrons (Is it just me or do these sound like robots you'd expect to find at strip clubs? Vaguely frightening.) and lap crystals but hardly enough to even dent my inventory. Other uses? Seriously, I'd love to be able to do something more useful with it. It's one of the game's under-exploited resources. I added a scuba mod because it used lapis to craft some of the gear.


    I'm not exactly using GregTech. I'm using the advanced buffers to get the right reactor components into the right slots. The rest of it is.... not my style. I'm getting the impression that it's one of those love-it-or-hate-it mods. ME buses work fine if you're just moving one component around but you need precision for two or more if you want them in the right place in the reactor core. Any suggestions of something that might work better? I'll definitely give TC a look. Is it a standalone mod or an IC2 add-on?


    Thanx,
    -S