My working LZH-condensator reactor, results on efficiency, and effective power generation.

  • Soo..... Lets just say I spent a long time making a project that I thought would be the future of "CASUC" reactors in 1.3.
    The Idea has been thought of ever since IC2 for 1.3 came out, but I was only convinced to make it after what Talonius stated in his post HERE.


    In short, I spent most of my time trying to figure out how to implement the different parts without Redpower, as it is not updated for 1.3 yet....



    So here is my solution: (If you just want to see my numbers and efficiency calculations, and Thoughts on the matter, skip to right before the last image.)


    1. My reactor setup. (note that we could go much more powerful than this, But since it was easier to not go past 2048 EU/t because of cable limitations, and for a reason I explain elsewhere, I went with this design.)

    2040 EU/t (before loss)
    also, note that this design puts 2 LZH's next to each cell. This gives me a bit of fail-safe in case a new LZH doesn't come in fast enough, that it won't start heating up and exploding imediently.



    Next, I used a simple Logistic pipes setup to automatically extract any LZH's that had ran out, since logistics pipes honors data values, this worked perfectly.



    My computer was having some issues with a large setup of logistics pipes, so I couldn't use them for the rest of the project, as they were VERY unreliable in terms of speed, so I used a sort of buffer to transfer the items to normal BC pipes, to transport to the cooling room.


    Then, the cooling room. the hard part. In the cooling room, there were three(3) main parts to the process. First, all broken LZH's come through the wall into a storage chest that is used as a buffer. This is so that in case the reactor goes through many LZH's faster than we can refill them, it doesn't mess up our refilling process.


    As items are pumped out of the chest by the BC gate, several things happen. First, the gate is an and gate, that is set up so that it will not pump out a new LZH until the old one is completely past it, and secondly, it sends a signal down the red pipe wire to a separate part of the machine to send 3 lapiz to craft.



    Next, the LZH, and the three lapiz arive at the auto crafting table, are crafted, and sent back on their way to a buffer right behind the reactor.



    Here, I once again used BC gates. This is the place where I most wished I had Redpower (besides cobble/scrap generation.) So instead, we have a complex series of gates that basically pumps a full LZH into the reactor if and only if:
    1. The reactor has empty space.
    2. There is not currently a LZH in the pipe
    3. There is a LZH in the chest to insert.


    However, this is not reliable, and sometimes pumps several, even though there is only space for one in the reactor. Hence, I placed obsidian pipes to catch the overflow, and put it back into the chest.


    I will skip over cobble generation to be used to make scrap, as I have not actually implemented this. I will, once RP is out, be using something like this:
    http://www.13-thirtyseven.com/assets/img/cobble%20gen.png using RB block breakers.
    but for now, all the methods of cobble generation won't fit nicely on my wall, so I just spawned scrap for these tests.


    and now... for the Lapiz generation.



    Here, I used more BC gates (prefer a lot of RP's stuff, but again, that isn't out yet.) to control the flow of scrap, and prevent inefficiencies.
    Firstly, the UU-generators are set up to only generate when there is some item in them, using the BC gates. As UU is pulled out as soon as it is generated, this means it will only be generating whenever scrap is present inside them.
    I have gates set up to only pump scrap when:
    1. the lapiz storage is not full.
    2. There is scrap inside the scrap buffer.
    3. There is room for the Scrap inside of the UU-gen.


    then it generates UU, and then pumpes it into the autocrafting table, and them puts the lapiz into the lapiz buffer.


    You will notice that the lapiz buffer has 3 pipes connected to it. Each one of these will pump one lapiz out of the chest for each empty LZH brought in to be refilled.






    Now, the big point. Results:


    So Talonius pointed out the following math:
    "...a simple design with one LZH-Condenser next to a Quad uranium cell."
    You get 12,000,000 EU
    The cost to remake the lapis: 1,905,778 EU
    The cost to remake the copper plates / tin cells used in the quad uranium cell: 544,700 EU
    Profit!: 9,549,522 EU


    So thats around 80% efficiency, not bad at all, and you could probably come up with a design using single uranium cells, we can get almost 85%, so not all that bad, considering how much EU/t we could still be making.


    So here were my Real-world, actually tested results:



    2040 EU/t (before loss)
    100,888,576 EU created in one full cycle
    26,240,148 EU used to regenerate the Lapiz during that cycle.


    So I got Less than 75% Efficiency.


    that being said, there are a few optimizations / things I would like to point out before everyone disregards these kinds of reactors.


    1. All EU readings were AFTER LOSS. And I was not very safe on loss at all. I used HV cable from the reactor to the MFSU's and glass fiber cable everywhere else. With more careful wiring, much more EU could be produced.
    After calculating what my loss should be by examining my cable connections, I can account for where about 4 million power was lost via bad cabling habits. However, that still leaves us with barely higher than 75% efficiency.


    2. I noticed that LZH's do not refill evenly. The WIKI (link) lists that a LZH can store 100k heat, while each lapiz repairs 40k of that.
    Therefore, the first 2 lapiz's give us 80k heat, while the third one is being half wasted, and only repairing 20k heat.
    We could never repair them up to full, and be forced to replace them more often as a result, but we would then use roughly 20% less energy to refill the LZH's, saving us about 5.2 Million EU in lapiz regeneration in this example, bringing us down to only about 20% total EU loss.


    3. Heat (and therefore EU required to regenerate goes up exponentially as you pack Uranium cells closer together, and as it stands currently (whatever it is about the quad uranium cell lifetime bug, could someone please confirm it is a bug?) I think we begin to lose efficiency as we pack cells closer together. (and therefor make more power per tick)




    Overall, Assuming the Quad uranium cell lifetime glitch is a bug, and will be fixed, this is a VERY viable layout.
    However, with the high losses I had in this experiment, and with the bug, It is only semi effective, netting us about 1400 EU/t with the layout I was using, after using UU to refill all related components.

    The post was edited 3 times, last by 1n5aN1aC: spelling fixes ().

  • nicely done setup. I may have to copy it once I start playing minecraft again. Waiting on RP2 for 1.3.2 before I restart the (private) server.
    also, on the scrap: are you using pieces or boxes? you can put boxes in the uu fab now.

  • nicely done setup. I may have to copy it once I start playing minecraft again. Waiting on RP2 for 1.3.2 before I restart the (private) server.
    also, on the scrap: are you using pieces or boxes? you can put boxes in the uu fab now.

    Really? Is boxes more efficient in any way? or just fit more in the slot at a time?

  • Really? Is boxes more efficient in any way? or just fit more in the slot at a time?


    Don't read the changelogs at all I take it?
    anyway, boxes provide 9x the scrap power as they're worth 9 scrap.


  • Don't read the changelogs at all I take it?
    anyway, boxes provide 9x the scrap power as they're worth 9 scrap.

    Ahh, so no efficiency bonus. I've switched over to using scrap boxes instead, just for ease of use, and not having to pump quite as many items into the UU-gen.


    Also, I've expanded my reactor design to accommodate up to 4096 EU/t,


    I have also essentially gotten rid of my EU loss due to cable fails, and am doing another test run with the optimization I mentioned with only partially refilling the LZH-condensators.

  • Not worth it. The reason is that scrap generation does not scale. It isn't possible to make enough scrap to keep up with very high energy power sources (10,000 or more EU per tick) without creating absurd amounts of machinery and therefore creating server lag. (if this were multiplayer, and if you aren't playing SMP, what's the point?)


    Scrap generators all create lots of lag because they have to create and destroy huge numbers of items.


    Anyways, you should base your efficiency numbers off of 0 scrap input - which means that this "casuc" is junk.


    Not your fault - your design is good. However, the game just isn't meant to be played this way any more.

  • On the base, I agree.


    However, practically, it's not hard to create that much scrap without any considerable lag.
    The design that I'll try as soon as RP comes out can easily feed 2 recyclers with 5 overclockers each, and keep them running 99% of the time, and although there are considerable block changes, you encase it in blocks, so there are no lighting updates, which stops most of the client-side lag, and the items travel through pipes for very short distance into a recycler, and if the recycler is full, the scrap generation stops.


    Although it does make some lag, Its not really much that I care about, even being the admin of the server I'm building it on. And It can easily make enough lapiz to fuel at least a 3000 EU/t reactor of this design for at least a full cycle, with the 1/3 chest of lapiz buffer i've been using

  • You won't lose any EU if you hook up multiple HV transformers directly to the >EV reactor. So you can get the full 4k EU if you use two HV transformers and if it goes past the 4096, use three.

  • This design looks very nice indeed. :)


    But consider this:


    9 Lapis cost 4 UUM and cool for 40k each. 1 UUM is one million EU / 6 for scrap amplification.


    So, at best, you cool for 1,85 EU/heat.


    Now the efficiency depends on your reactor design. Using chains of quad cells as shown you get efficiency around 5.


    One quad cell at efficiency 5 produces 100 EU/Tick at 240 heat/second, using 22.22 EU/tick for cooling (77.78% left) - which makes an overall efficiency of 3.88.


    If you put just one LZH condenser next to each quad cell you get them to efficiency 6, then they each give 120 EU/Tick at 336 heat/second, using 31.11 EU/tick for cooling. 74.07% of the EU left, overall efficiency 4.44.


    Considering you can get efficiency 4 by putting two quad cells next to each other, and efficiency 4.5 by surrounding a quad cell with single cells, both of which can be cooled in a non-CASUC-way... your reactor certainly generates one huge amount of EU/tick, but it's not as efficient as the old CASUCs.


    So, judge for yourself if you think it's worth it.



    P.S. Woah, I just did another quick calculation... generating the scrap is much more expensive than I first thought. Even with non-overclocked recyclers you use about 72k EU to generate the scrap for 1 UU-Matter, getting the base cooling cost of this system up to 2.65 EU/heat! Diminishing efficiency by another 10%... :S


    This brings the overall efficiency of your design down to 3.4 (68.18%), and the variant with one condenser per quad cell to 3.77 (62.87%) once you start generating your own scrap.


    Using non-overclocked recyclers. Woah again.


    Okay, it seems that while your design is a good proof of concept, it's definitely not worth it. Sorry to say so...


    EDIT: Miscalculated this.

    The post was edited 1 time, last by inwerno ().

  • what numbers?
    the cooling wont cool anything
    and you need to replace the LSZ about 33 times


    all is included in the reaktor planner.
    just play a little bit more with it and you will know.

    Change the scheme, alter the mood. Electrify the boys and girls if you'd be so kind.


    [b][i][u][url=' [url='http://forum.industrial-craft.net/index.php?page=Thread&threadID=7745']HAYO CORP: Nuclear Power (FREE: Reactor Blueprints)

  • what numbers?
    the cooling wont cool anything
    and you need to replace the LSZ about 33 times


    all is included in the reaktor planner.
    just play a little bit more with it and you will know.

    Just read the OP...then take a look at the scheme again and if you still don't get it...don't post again.
    And next time before you get all smug...think/look twice...because if you're wrong this happens and you loo like a jackass/noob.

    mcmz4e aka MaryuZ aka 2.muCh.Pride

  • ok tell me the numbers you want
    I still don’t get which you want,
    I will apologies then.


    Also with the new breeders the cooling system needs to be very optimized to the Reactor,
    You can’t just swap reactors, sometimes the cooling isn’t enough and you have to make some changes.


    Also looking at your blueprint
    I can tell that you are not very experienced with the new reactor system.


    The cooling you have provided will do nothing unless you remove one of the LHZ from the uranium.
    Also it is a great waste of copper with that.


    When you want to have a safe CRCS with cooldown if one of the LHZ get full and your system isn’t fast enough:
    http://www.talonfiremage.pwp.b…4w2hjwalcz8pu25i9irc3fsao


    if you want to cooldown some of the heat, so it won’t cost you that much Lapis.
    there is somewhere a good design, can’t find it now.


    the problme is that all the heat from the uranium gets stored first in the LHZ and after they are dead, then the reaktor will heat up and your cooling will start to work...
    also if enough heat is in the hull the cooling will melt cause the OC vents will pull more then thy can cool...

    Change the scheme, alter the mood. Electrify the boys and girls if you'd be so kind.


    [b][i][u][url=' [url='http://forum.industrial-craft.net/index.php?page=Thread&threadID=7745']HAYO CORP: Nuclear Power (FREE: Reactor Blueprints)

  • Sorry I haven't replied.


    I've actually thought of several optimizations to the design to get probably a few more % efficiency out of it,


    but I haven't had any time to work on it, between college and updating my public server to 1.4.

  • I used this for testing. The nice thing about this one is that all the LZH-condensators are taking damage at the same rate. This means you can detect if they are at 1 second away from being empty so that you never get any heat in you hull. Also it is more efficient at about 5.5. After all the EU/t the recyclers and mass-fabricator use up to make everything work as a Mk.1 without any (ofcourse you need to take components out in put them in, but other then that the system can run 24/7) cooldowns or waiting for heat to be removed from the hull it still produces 1825.5 EU/t (not the 2000+ EU/t we were used to, but still very nice in my opinion).

  • OK...Lets clear the air a little...I see some genuine nuclear engineers have posted...but are wrong...I blame lack of experience with logistic pipes so I will explain now.
    Why my design has active cooling...vents and such...:
    In order to get that 6 eff rating i have just one lzh touching each quad-cell...now the reactor is always on...if you would have more carefully read the OP you would have noticed the nuclear facility described uses logistics pipes to automate the swapping of lzh...and logistics pipes...on a really short 2 pipe setup still take 2-3 seconds to do the swapping...it's just how they work...items travel through pipes at a speed and the fresh lzh is only sent 1 second after the empty one is extracted.
    So in those 2-3 seconds that the lzh slot is empty 336 heat is transfered to the hull...thus the cooling...to take care of that aprox 1k heat in the next interval so the thing won't blow up after a while.
    with a 3 second gap for switching LZH condensators 8064 heat is transfered to the hull every 297.61 seconds at a rate of 1008 every time a lzh runs out.
    those 4 OC vents provide 128 cooling/s...that is a little much but it's better to be safe than sorry...the plating is also just to be on the safer side of things...
    The OP did not have this problem in his setup as he has 2 condensators touching each cell so when one is pulled the other just takes double load.
    now...the numbers i was asking for:
    I want to know with my design how much EU is generated per uranium used after the lapiz cost is substracted...so just total eu generated in one run and the cost of the lapiz.
    @the bard.....it's not equivalent...you design blows up after 300 seconds.
    Sidenote...for my design to work a little more safely the initial LZH condesators that are used to start the reactor with would have to be already damaged...each with a different dmg in them...best case would be from 12500 to 100000 in 12500 steps...this would ensure a minimum heatload on the hull...if the reactor is started with 8 brand new condensators after 298 seconds the hull would get a instant load of 2688 * seconds it takes logistics pipes to swap.

    mcmz4e aka MaryuZ aka 2.muCh.Pride


  • and you're also going to have to replace the 8 Thick Neutron Reflectors after every run as well.

    why?...are they bugged or something...aren't they supposed to last 4 times longer than normal ones? and a normal one lasts a full cycle?
    do mind they're only exposed to one cell each...

    mcmz4e aka MaryuZ aka 2.muCh.Pride

  • And here's a bigger 6 chamber 1320 eu/t one that should in theory still work and still be safe provided it's started with condensators having different dmg values preferably evenly damaged at 9000 steps from 9000 to 99000.
    Still eff 6 :D
    http://www.talonfiremage.pwp.b…vo6uxn8gvzlxh6y9q66q1ux34



    and an even bigger one with 12 cells for the really power hungry ones at 1440 EU/t:
    http://www.talonfiremage.pwp.b…kfz2qbjwriowhno3haqw6nk74
    this one can work ( as in not explode ) with a 4 second swap time for the LZH....it has 64 /s cooling and makes 54.3 distributed heat/s in 1344 chunks of heat each time a LZH runs out so it would even work with 12 evenly damaged LZH but it would dump 16128 heat in 4 seconds in the hull once every 297.61 seconds...the hull can take it courtesy of the plating but i don't know if the the vents can take it without melting although if my calculations are correct they should.
    you can't really go bigger without compromising that 6 eff...and that's what I'm aiming for.


    And if you're willing to compromise that 6 eff here's a monster for you:
    eff: 5.67
    output: 2720 eu/t ( uses 12 quad cells )
    LZH swap time: need a swap time of 2.3 seconds or less otherwise it will blow
    LZH damage distribution: needs to be started with pre-damaged LZHs at precise intervals otherwise it will blow (* will work with 6 LZH at 25k, 6 at 50k,
    6 at 75k and 6 full ones but don't take my word for it...do your own math...it's your base)


    http://www.talonfiremage.pwp.b…6465umf9fmsr2emiqubn951j4



    And now for the daddy of new generation Casucs:
    Eff: 5.35
    output: 4280 eu/t
    How to operate:
    1: set up as shown
    2: set logistics pipes to swap used LZH for new ones*
    3: on a pipe attached to the reactor set up a BC gate to emit redstone signal when inventory full
    4: have faith in logistic pipes that it will pull instantly**
    5: have faith in BC that it will stop redstone signal instantly**
    6: take a step back...look at the pattern and try to figure out what it reminds you of :P
    * you could alt set up the pipes to extract ALMOST used LZH to be safer, but some LZH touch 4 cells, others just 3 and others just 2, so you need to calculate the damage values and set up multiple extraction patterns...and this only works if logistics pipes extract instantly...in fact the whole design only works if logi pipes extract instantly...and I think they do...once every 20 ticks...on the full second tick...so the question you have to ask yourself before operating this monster is: Does my computer do the IC calculations or the Logistic pipes calculations first on that faithfull tick???
    ** do not let yourself be fooled by the one piece of plating...it's there just to make it possible to use a BC gate trigger...the reactor can not operate for even a second without the LZH in place...it has no cooling it will blow.


    http://www.talonfiremage.pwp.b…719m8qzk5oom8ybl96qi13vnk
    I'm such a noob...replace the plating with a thick neutron reflector for extra juice/eff.
    you get 5.43 eff and 4340 eu/t

    mcmz4e aka MaryuZ aka 2.muCh.Pride

    The post was edited 8 times, last by mcmz4e ().

  • why?...are they bugged or something...aren't they supposed to last 4 times longer than normal ones? and a normal one lasts a full cycle?
    do mind they're only exposed to one cell each...


    The one cell they are exposed to is a Quad cell. So while yes the Thick Neutron does last 4x longer than the normal one... the length of time is based upon a single cell. If you used the regular Neutron Reflectors, you'd have to replace them 4x over a single run.


    Single Cell + Regular Neutron Reflector = 1 full run
    Single Cell + Thick Neutron Reflector = 4 full runs
    Dual Cell + Thick Neutron Reflector = 2 full runs
    Quad Cell + Thick Neutron Reflector = 1 full run