Powering a Nation: An Insane MOX CRCS Reactor

  • Edit: So I discovered something today while discussing numbers with Omicron...looks like my config files generated a little funky (promise, I didn't realize it!), and I'm getting 4x more EU from nuclear fuels than I should have been...so the numbers below are wrong! :( Divide all EU values by 4.



    So I saw all these awesome reactors...Tower of Power, Decentralized Distribution of Steam (DDoS), the Official Reactor Design Thread...there's so many creative and interesting builds out there. Well, I wanted to one up them. I wanted to power a nation.


    And this is what I came up with. It's nothing fancy, and not really creative...but it's a power horse. Putting out a whopping 6,400eu/t at zero heat, this bad baby is just crying to be abused. And abuse her I did. I pumped her up to 55,500 heat, and with a max heat of 60,000, that placed her at a multiplier of 4.7. Then, when you do the math, that's a reactor putting out over 30,000eu/t. That's no typo! THIRTY THOUSAND! And that's living on the edge, because without cooling, this reactor produces 4,300 heat/second. Given the operating temperature I had her running at, that left me with a 200 heat buffer before exploding...scary.


    Anyways, how about you feast your eyes on my beast! She's running a full 6-chamber reactor, as shown in the design above. To maintain a constant cooling temperature, 20 coolant cooling reactors (designed by Someone Else 37) are needed. On top of that, a special (and probably badly designed, but I wasn't looking for efficiency or speed, just something that'd help them cool off a little bit faster) reactor designed to cool Overclocked Heat Vents had to be produced. The whole thing is controlled by a single computer, which monitors the status of the reactor (heat levels, and components) every 8/10ths of a second. This computer forces the reactor to shut down and inserts an Overclocked Heat Vent to cool it off when it exceeds 55,500 heat, removes coolant cells when it's below 54,500 heat to allow the reactor to reach critical levels. It replaces Coolant cells and Heat Vents when they're around 80% broken, so they can be recharged in the cooling reactors. It replaces the fuel when it depletes as well. It does all this using Open Peripherals...other than the ME Interface to keep a supply of fuel rods handy in an "inventory", and exporting coolant cells and heat vents into obsidian chests (vanilla chests kept breaking around the reactor) surrounding the base of the reactor, AE has nothing to do with the actual maintenance of components in the reactor. It does, however, maintain the transfer of components into and out of the cooling reactors.


    All this power runs underground via a single glass fibre cable. It goes into 15 EV Transformers, wired in parallel and spaced a meter apart. These transformers output 2048eu/t into two MFSUs wired in a serial fashion. The intent is that the first MFSU remains nearly empty while it keeps the second MFSU charged. Wireless Redstone Transmitters are wired to every first MFSU, and the first MFSU's mode is changed to "Emit when Empty." This in turn is received by a Wireless Redstone Receiver in front of the Computer, which detects that an MFSU is requesting power as part of the conditions required to turn on the reactor.


    To test stability, I ran the reactor for a complete cycle before classifying it as "stable enough". You can see the bank of MFSUs I had to place in order to store as much power as that reactor generated. I calculated that the reactor output is in the neighborhood of 3 billion EU per cycle (assuming MOX lasts 5000 seconds, and there's 20 ticks in a second each producing 30,000eu/t). I got a few ideas on how I can tweak around another 1,000 eu out of this reactor, but I wanted to share the progress I made thus far.


    Mod List:
    Computer Craft
    IC2 Experimental (tested in build 299)
    Open Peripherals
    Applied Energistics
    Wireless Redstone
    Iron Chests


    Some things I learned:
    Wireless Redstone doesn't like nuclear explosions...they'll crash your game...or at least they did mine.
    ComputerCraft sometimes was listed in the crash logs as well when a reactor I was testing blew on me.
    Even if you turn the reactor off while you swap coolant cells, occasionally a reactor tick will get by and add between a few hundred and a thousand heat to the reactor. The control program detects this, immediately shuts off the reactor and places the Overclocked Heat Vent in it to cool it back down to a safer temperature. And it's never anything a Overclocked Heat Vent (or two) can't handle, and it tends to keep the reactor running at the top end of the desired range of temperatures you have set in the Control Program.


    If you want to look at the code I used for controlling the reactor...it's sloppy, but it works, and can be found here. Couple that with the following startup program, and you'll be as safe as you can be while completely automating the reactor. And trust me, you want it automated because at the heat levels it obtains, you can't get within 3 blocks of it without starting to take damage, and everything within a block of it will randomly catch fire. Anyways, the startup program:

    Code
    shell.run("MOXControl")
    rs.setOutput("back", false)
    sleep(1.1)
    os.reboot()


    That will run your control program, if you terminate it or it crashes, it'll turn the reactor off instantly, and you have a delay to terminate the startup program before it reboots, running the control program again. This means your control program will almost -always- be running, even if it crashes out for some reason.


    Nothing really new or mind blowing in this reactor...other than the massive power output. It takes tried and true concepts, and applies them in a dangerous situation. Living on the edge man, living on the edge. Just wait till you see my ideas on how to tweak around another 1,000 eu out of this girl.

    • Official Post

    Puts old ice cooled reactors to shame :P

    145 Mods isn't too many. 9 types of copper and 8 types of tin aren't too many. 3 types of coffee though?

    I know that you believe that you understood what you think I said, but I am not sure you realise that what you read was not what I meant.


    ---- Minecraft Crash Report ----
    // I just don't know what went wrong :(


    I see this too much.

  • Very impressive system you've got there. A single GT fusionreactor produces about 46000 EU/t, after subtracting the power draw of the processing machines- not even twice what your probably much cheaper (at least in Chrome) MOX setup produces.


    Your heat vent cooling tower isn't as impressive as the rest of your system, though, since the component vents cannot transfer any heat. You have several components that aren't actually doing anything.


    A better plan is what KenKen originally used in his HVC rig- an otherwise empty reactor stuffed with heat vents. Such a 1-chamber reactor can cool 480 heat per second, using nothing but 24 overclocked vents. Since the overclocked vent has the highest heat dissipation rate of any component, there's just no way to dissipate more heat per second per slot. As for cooling them faster, just use a few more- they probably won't cost as much as your component exchangers.


    Also, just to make sure, you are accounting for the fact that my cooling towers will never completely cool any cells, correct? I see Fuzzy Busses in your screenshot, so I can only assume that you did. I take it that they're set to extract 75% cooled cells?

    If you stare at my avatar hard enough, you'll notice that it consists of three triangular rings, interlocked in such a way that if you were to remove any one of them, the other two would be free to float apart.

  • Very impressive system you've got there. A single GT fusionreactor produces about 46000 EU/t, after subtracting the power draw of the processing machines- not even twice what your probably much cheaper (at least in Chrome) MOX setup produces.


    I've never played with GT, so I'm not sure about his machines...but indeed, that's a lot of power.


    Quote

    Your heat vent cooling tower isn't as impressive as the rest of your system, though, since the component vents cannot transfer any heat. You have several components that aren't actually doing anything.


    A better plan is what KenKen originally used in his HVC rig- an otherwise empty reactor stuffed with heat vents. Such a 1-chamber reactor can cool 480 heat per second, using nothing but 24 overclocked vents. Since the overclocked vent has the highest heat dissipation rate of any component, there's just no way to dissipate more heat per second per slot. As for cooling them faster, just use a few more- they probably won't cost as much as your component exchangers.


    I thought component heat vents supplies an additional 4 cooling to surrounding components (ie the overclocked heat vent). I'm sure if I looked at things I could design a much better tower...I did say it was probably poorly designed. :S The reason I didn't use component heat exchangers and more overclocked heat vents to cool it is the ME Network would see those heat vents that are supposed to remain in the reactor and pull them out as well. I know overclocked heat vents will eventually cool off on their own, I was just attempting to speed up the process. Would a setup like this be better: http://www.talonfiremage.pwp.b…je0cmlunrby25af8pp5zz2tc0 ?


    Quote

    Also, just to make sure, you are accounting for the fact that my cooling towers will never completely cool any cells, correct? I see Fuzzy Busses in your screenshot, so I can only assume that you did. I take it that they're set to extract 75% cooled cells?


    Indeed. I have 2 powering chambers of ShneekeyTheLost's exact design from the Tower of Power (so half of the 4-reactor build) on my survival world running for about a week now. I just didn't set them up in a pretty square stackable fashion...you can see how I laid it out in my basement here. Only modification is I added 2 more transformers since I learned transformers output 2048 total, not 2048 per side.

  • On top of that, a special (and probably badly designed, but I wasn't looking for efficiency or speed, just something that'd help them cool off a little bit faster) reactor designed to cool Overclocked Heat Vents had to be produced.


    I thought component heat vents supplies an additional 4 cooling to surrounding components (ie the overclocked heat vent). (...) Would a setup like this be better: http://www.talonfiremage.pwp.b…je0cmlunrby25af8pp5zz2tc0 ?


    Yes, that setup would definitely better. You see, your original setup is essentially this one with a whole bunch of components thrown in that do nothing at all. The component vents do not spread heat around, you need actual exchangers for that.


    I calculated that the reactor output is in the neighborhood of 3 billion EU per cycle (assuming MOX lasts 5000 seconds, and there's 20 ticks in a second each producing 30,000eu/t).


    Runtimes were changed in build #288. MOX fuel lasts 10,000 seconds and Uranium 20,000 now. ;)


    All this power runs underground via a single glass fibre cable.


    Just so long as you're aware that the moment they implement cable behavior, your reactor will instantly melt any single cable... when the e-net gets finalized, you'll need four individual lines of glass fibre coming from your reactor without touching each other, each leading into four transformers. Might be a good idea to come up with a design idea for that, just in case.

  • Yes, that setup would definitely better. You see, your original setup is essentially this one with a whole bunch of components thrown in that do nothing at all. The component vents do not spread heat around, you need actual exchangers for that.


    Already changed :) Not that it makes much of a difference with the amount of Overclocked Heat Exchangers I have in the buffer, and how little they're used.

    Runtimes were changed in build #288. MOX fuel lasts 10,000 seconds and Uranium 20,000 now.


    Then up the estimate to 6billion EU...I was wondering why I was filling more MFSUs then I calculated I'd needed, by a good number.

    Just so long as you're aware that the moment they implement cable behavior, your reactor will instantly melt any single cable... when the e-net gets finalized, you'll need four individual lines of glass fibre coming from your reactor without touching each other, each leading into four transformers. Might be a good idea to come up with a design idea for that, just in case.


    Now that you mention that I should have a plan, I've already come up with one. Instead of coming out the bottom, I'll come out the 4 sides of the reactor chambers, and go underground from there...and split the transformers to the 4 cardinal directions. And won't I still need 15 to 16 transformers to handle the 30-31k output?

  • Now that you mention that I should have a plan, I've already come up with one. Instead of coming out the bottom, I'll come out the 4 sides of the reactor chambers, and go underground from there...and split the transformers to the 4 cardinal directions. And won't I still need 15 to 16 transformers to handle the 30-31k output?


    Yes, but only 4 per cable, since glass fibre is supposed to hold up to 8192 EU/t. If your reactor outputs (just shy of) 32k, then you split it into 4 lines of 8k each, and each of those 4 lines has 4 transformers (for a total of 16) to break the 8k further down into 4x 2k so that it can be accepted by MFSUs.

  • Could you upload the reactor arrangement as an image file instead of the planner for those of us who can't view the planner?

  • Yes, but only 4 per cable, since glass fibre is supposed to hold up to 8192 EU/t. If your reactor outputs (just shy of) 32k, then you split it into 4 lines of 8k each, and each of those 4 lines has 4 transformers (for a total of 16) to break the 8k further down into 4x 2k so that it can be accepted by MFSUs.


    Yeah, and if I'm going to split it, I'm have to split it symmetrically...otherwise, it's just going to look bad :P


    Currently testing a MOX reactor with an output of 41,000eu/t...and I have some blueprints for some which should out do Gregtech. >: ) Who needs Gregtech for endgame power? NOT THIS GUY!


    Expect an update later today!


  • The second link in his post (under "30,000 EU/t") is an ingame screenshot.


    Derp, didn't see that.


    Is there any other way to control the heat though, e.g. via Nuclear Control? I really don't want to use ComputerCraft.


  • Is there any other way to control the heat though, e.g. via Nuclear Control? I really don't want to use ComputerCraft.


    I'm not sure how well nuclear control would handle the micro-managing control I have implemented via my Computer. All I've ever used Nuclear Control for was to get information about the reactor TO my computer in 1.4.7. Since 1.5, I've replaced Nuclear Control functionality with Open Peripherals. Also the computer is completely handling the replacing of components when they need replaced before they are destroyed or when they are depleted, as well as inserting them and extracting them based on multiple sets of heat levels, not just on and off. On the 41k reactor I'm testing right now, I'm keeping the heat in the reactor between two values which are 100 units apart...and are less then 1000 units from max heat. I also have 4 modes which detect 4 different heat levels, the first, when the reactor is really cold, is heatup mode where all coolant is removed from the reactor to ensure it reaches heat, and does it as fast as possible. Once I get within a certain margin of my operating temperature, it switches to slowdown mode, where it places all but one coolant cell in the reactor to slow the heatup process so it doesn't overshoot the operating temperature and explode. This reduces the heat generation to the reactor hull from 4000ish to 240 or something like that. Then it has an operating mode, as I stated above. Finally, there's a cooldown mode where it makes sure all coolant cells are in place in the reactor, a overclocked heating unit is placed in the reactor to draw heat from the reactor (at a rate of around 100 heat a tick, for some reason...I thought they only were able to exchange 36 heat at a time?), and the reactor remains shut down so it's impossible to generate any more heat. This is useful because for some reason, when coolant cells are swapped, the reactor ticks over and does its calculations before realizing there's a new cell in the reactor, so it calculates that heat to the reactor hull as if there were no coolant cell present. I've tried to mitigate this by turning the reactor off during coolant swaps, as well as adding a delay before and after the coolant cell is applied before turning the reactor back on. Neither seem to reduce the occurrence of this problem.


    So yeah, I don't know if Nuclear Control has the fine tuning to do that, I don't have that much experience with it. You'd probably need multiple (remote?) thermal monitors, and your control system would be much larger, not to mention the system that swaps components in and out of the reactor when appropriate. You can always go into a creative world, test and find out.

  • By the way, if anyone wants to use this reactor in legit survival gameplay... you really have your work cut out for you :P


    Let's say you take an uranium reactor with a large number of fuel rods so you can generate plutonium as quickly as possible. Something like this one. for example, because after all you still want a modicum of efficiency.


    Every 5 hours and 33 minutes, you'll generate 28 tiny plutonium. You will need 1,512 tiny plutonium to stock that CRCS monster. Therefore, you'll be running that reactor for... just a few minutes short of 300 hours. In other words, just under two RL weeks, 24/7. And if you don't want to accept efficiency 3.00, you'll probably have to double that number - reactors running efficiencies above 4 generally can't fit more than 12-14 fuel rods.


    So maybe you want to build many reactors to cut that time short by a lot? Well, there's another issue: you're also going to need 1,134 blocks of uranium ore to pull it off. That's almost 18 stacks... two thirds of a chest filled with nothing but uranium ore. Since uranium spawns anywhere below sea level, you can't focus on tunnel boring specific heights. Good luck with your mining endeavours! ;)

  • Thanks for doing that math on resources Omicron...I still think it's completely feasible, though it'd take a lot of work. I might try for it...after I'm done having fun in creative seeing just how OP I can make these things. :D


    edit: That's under 11 days of UU Generation for me and my setup...minus the Mark I's I have scattered around my base/mining rig using single fuel rods...I wonder how much it'll make...I already have enough uranium (you'll need 4,436 small piles of uranium 235...and I got over 5k in my storage network, and it's increasing every day)...I just need to mine more for the components and reactors themselves.


    edit2: I just checked my plutonium reserves in my storage network...I have 98 full size plutonium, 8 small, with 54 more small ones being refined as we speak, or 105 full size total...so that's 8 of the 14 quad MOX rods I need for my first design. :) Not that I'm going to build it...with machines unable to take more than 8,192 eu/t, I think less ambitious Mark I reactors would be more advantageous in the long run...or at least, for now.

  • I'd say it's a valid endgame goal on a 24/7 server, with the proper steps for offline production taken. Singleplayer? Yeah, not going to happen.

  • So 30,000eu/t just wasn't enough for me. I had to squeeze every last once of power I could out of those MOX Fuel Rods, and so I went about it...I went about abusing the ever-living snuff out of those poor helpless fuel rods. I rewrote the MOX Control Program to add a 4th "phase" or mode to the reactor cycle. When you first start it up, it goes through the heatup phase. In this phase, all cooling mechanisms are removed from the reactor so every last Kelvin of heat goes into the reactor's hull. Then, at 54,500 heat, it enters the second phase, "slowdown". In this phase, it inserts all the coolant cells into the proper places, then removes one. This throttles the reactor heatup process by eliminating the heat of all the fuel rods except one from going into the reactor hull. This reduces the heat transferred to the reactor hull from 4,320 a tick to a measly 240. This reduces the total EU the reactor can generate in the first cycle by a small amount, but it increases the safety factor so much that it's worth it. And given the next change, we need that extra safety factor. The next phase is the "running" phase, where it reaches "optimum" operating temperature where you've balanced the risk of explosion to the gain of bonus EU to your liking. In my setup, I have the operating temperatures set between 59,000 and 59,100 heat. Remember, when the reactor reaches 60,000 heat, it explodes. What can I say, I like to push the lines! And the 900 heat safety factor is needed. There seems to be a glitch where replaced coolant cells take a reactor tick to be detected and used in the reactor's heat calculations. Every time this happens, between 240 and 336 heat is transferred to the reactor's hull, increasing the temperature above the temperature margin for the running phase. When this happens, it enters the cooldown phase, where the control program immediately shuts down the reactor, makes sure every coolant cell is in place, and inserts an overclocked heat vent to start transferring heat away from the reactor's hull. Once enough heat is transferred away that the running temperatures are reached again, the heat vent is removed from the reactor, and it is turned back on to operate.


    So, the moment you've been waiting for: At 59,100 heat it generates a total of 31,616eu/t.. You might be able to squeeze a couple hundred more EU out of her, but in doing so you increase the chance for meltdown to unacceptable levels (in my opinion). Given the stated operating temperatures, it'll generate a minimum of 31,573eu/t. I'm quite proud of that.


    But honestly, it wasn't enough for me. I know I could do more. I know I could do bigger. But I couldn't do it with my current reactor design. I'd reached the limit of what 14 Quad MOX Fuel rods could give me. A redesign of the reactor was in order. It was a simple upgrade, but it allowed me to squeeze in 4 more Quad MOX Fuel rods, though it resulted in a significant 16,000 heat units lost to the maximum hull temperature. Whatever, I knew I could adapt and overcome. Putting to work what I enacted with the 14-Quad-rod reactor, all the work was done for me. All I had to do was add 6 more cooling reactors to ensure my Coolant Cells were always available and refreshed when the reactor needed them. So simple, it doesn't even come with a picture. Because of the changed max heat variable, I had to change my control program to adapt. I decided to go with 35,000 heat units for threshold to enter the slowdown phase, as this reactor would produce 5,664 heat a tick, and I wanted a good buffer. The running phase would be in effect between 43,200 and 43,300 heat. With 700 heat units separating my reactor from a crater in the dirt, I was putting myself and my reactor at significant risk. Remember the bug where coolant cells weren't always detected and used properly? Well if that happens twice in the same tick, 360 times 2 is over 700. BOOM! Luckily the chance for that is small (it didn't occur during any of my testing, and I ran a full cycle), but the possibility is there. Lowering the operating temperature by a few hundred heat units will significantly increase the safety factor, but I wanted to reach that level for a reason. The reason? at 43,207 heat, the reactor produces a massive 41,000eu/t, on the dot. I wanted a reactor that outpaced my previous design by nearly 10,000eu/t, and I got it. At higher heat levels, which it's more prone to operating at due to the coolant cell bug, it'll produce slightly more. Here's an example at 43,272 heat where I'm producing 41,049 eu/t.


    I made some other small changes to the control program in the 18-cell build. One change is I reduced the console spam by a considerable margin. When it adds or removes all the Coolant cells in the reactor, instead of a line on the console for each cell, there's just one line for the mass of them. I also made it so it only displays the phase and temperature if the temperature changes. Finally, because the increase in the number of cells that can be moved at once when it places or removes them all, I felt this created a lag in the control program where a reactor tick could occur that the Control Program wouldn't detect and operate on, I told it to enact it's nuclear logic function where it detects the heat level and acts accordingly both before and after all components are checked, before pausing for 8/10ths of a second and repeating the process. I felt this increases the safety factor significantly, while harming nothing.


    And as you can see in a screenshot of my AE system mid-cycle, the system can handle and maintain adequate cooling. The only reason you don't see Overclocked Heat Vents in the buffer is I lost a few when stopping the control program during cooldown phases, and neglected to replace them. This caused a problem part-way through the cycle when it attempted to insert a heat vent, but there wasn't one in the slot in the chest it was looking for one in....and is when I first noticed the problem, and rectified it.


    Some things I learned during these builds:

    • If you calculate that you need 20.05 transformers to handle the power you're putting out...the .05 is SIGNIFICANT! Don't ignore it ;)
    • Power doesn't distribute evenly down a parallel power line. The line of MFSUs coming off the transformer closest to the reactor had more EU stored than the ones further down the line. It's so significant that over the course of the cycle, the MFSU lines at the last few transformers has a whole MFSU's less power stored than the ones at the front of the line.
    • nodding off while you have your reactor gui open when you have a tendency to click your mouse when you startle back awake isn't safe. Especially when your mouse is hovering over one of the reactor platings when you click, at least when operating this close to max heat (thus the picture of the crater).


    Hope you enjoyed these builds. I have a few blueprints I'm going to try soon that'll (theoretically, if the math's right) put the best Gregtech has to offer to shame.

  • Are you sure the coolant cell thing is a bug, and not simply related to the fact that reactor components are processed one by one, left to right, top to bottom? Maybe the slot where the coolant cell was inserted had simply been processed already, if the insertion happens during the one tick each second that the calculation is being run...

  • Are you sure the coolant cell thing is a bug, and not simply related to the fact that reactor components are processed one by one, left to right, top to bottom? Maybe the slot where the coolant cell was inserted had simply been processed already, if the insertion happens during the one tick each second that the calculation is being run...


    I doubt that's the case...in one attempt to squash the bug, I set up my code to turn off the reactor (rs.setOutput("back", false)). I then told it to wait 2/10ths of a second. Any processing going on in the reactor for that tick -should- be done at that point, unless the calculations are spread evenly across the 20 ticks in a second to reduce spikes every 20th tick. Then I'd swap the a coolant cell out, and wait another 2/10ths of a second before turning the reactor back on. Even doing this, I had the reactor randomly gain heat from the glitch.


    The -only- other option I have is to pause the program for a whole second after turning off the reactor before replacing the coolant cell and see if that works...but that'll reduce the -effective- eu/t, so I didn't want to do that...but I may give it a try just to see if it works.

  • Two questions:


    1) Does 85% heat melt (iridium/tungstensteel) reinforced stone? What is the radius that the reactor affects?


    2) How well will the design work with Advanced Reactors?


    Either turning things to lava is extremely rare, or it's not implemented yet. In all my testing, I've yet to see it happen. The worst that's happened is:


    1) Things catch within a block or two of the reactor. As long as they're not burnable blocks like wood, this isn't a problem.
    2) Vanilla chests break and vanish near the reactor
    3) Within two or three blocks of the reactor you take damage.


    As for Advanced Reactors, I don't know...but I'll look into it.


    edit: Looked into Advanced Reactors and honestly, I wouldn't want to run a high-efficiency MOX reactor with it. It has too many negative environmental effects for my taste.