Water vs Ice (2020 vs 2070 EU/t CASUC Reactors)

Warning: if you don't like side-notes (I would call This a side-note) then don't read this text. It will drive you mad.

Today I thought I'd have another go at CASUC reactors. I wanted to know if the fact that you can now attach pipes to the chambers itself changed the winner in the water vs ice race. For me it always was fairly simple. I need two different types of pipes to insert and extract buckets in and out of the reactor. Redpower tubes and it's awesome retrievers (to extract many empty buckets out of the reactor through only 1 pipe and thus only using the space of 1 chamber) deployers and filters. And I needed Buildcraft (with the additional pipes insertion and redstone pipe) to insert the filled buckets by creating a little circle (remember, it's Minecraft, so when I say circle I basicly mean a square) with an insertion pipes, an iron pipe, a golden pipe and a redstone pipe to create an "infinite" sized buffer which could dump out its entire contents in only a fraction of a second (which could be 5 or 500 buckets (or ice)). Because I could not make a "kind of a bit stable" reactor with only one of the 2 kind of pipes I stopped with looking in the possibilities of the water cooled reactors and build a gigantic ice cooled nuclear power-plant using ice cooled reactors and breeders. The result was a 15000 EU/t power-plant which was so efficient I felt wrong about asking the other players of my server 1 uranium per 10 million EU (delivered for free using teleport-pipes and power-converters (I kept track of how many EU every one used with a energy counter I designed myself and could also let them pay by means of teleport-pipes)) because it was much cheaper for me to make (it also had some ridicules number of MFSU's, good for 10 billion EU of storage (in case I didn't log in for a few days)).

I'm wasting your time by not getting to the point. Because you can now attach pipes to the chambers itself I thought I'd give the bucket design another go. The result is in the crappy quality screenshots (only 450 kB for all your attachments). It's this design and it doesn't use an absurd amount of redstone logic-circuits (yes, it does use 1 million vanilla repeaters too many). As you can see it produces 2020 EU/t and I ran a complete cycle without any problems. I could not make this design work with Redpower to insert the buckets, not even with the now 10 relays right next to the reactor or attached directly to the chambers themselves (I tried 1, 2 and 8 attached to the reactor, but they still got "out of sync" and missed 1 bucket every 10 seconds or so, enough for the reactor to overheat and blow up (luckily I did have 7 backups (I learned the hard way how important backups are when doing nuclear reactor stuff))). It's counterpart these days is this 2070 EU/t design which is only 50 EU/t more. So this means that a long as you don't use EE to get the ice, it will be very difficult to get more EU/t out of an ice reactor vs a water reactor (I know someone set himself the challenge to do this with only 50 EU/t (it was mijnboogje:"i just love the advanced machine's add ons, gues its time for me to build an industrial monster... goal: dont let the coolants eat more then 50 eu/tick now THAT wil be an chalange :P")).

Lets for argument sake his cooling design is awesome and uses exactly 50 EU/t (which it probably will) and that it is just as large as a good empty bucket extraction system (which it might not be, but who cares, it's Minecraft, so we have a lot of space to build something awesome). Then there is no difference in EU/t between the two designs, so then the biggest thing for me is: ice stacks, buckets don't. This has always been a huge reason why i like ice better then buckets. When you detect (with the system you can see in the screenshot of the reactor itself) that the reactor is filled, with ice it means you can safely let the reactor run for a minute without supplying any more ice. With a buckets cooled reactor the only thing you know is that there was no free slot in the reactor when the buckets tried to get in. This could also mean that the empty bucket extraction system failed to pull out all 8 empty buckets within 1 second and if it failed altogether then you are now relying on some other system to shut down the reactor for you. I never liked the idea that you need to react within 5 seconds, or the reactor is very likely to blow up. With ice cooled reactors you can have no ice coming into the reactor for a whole minute and still have 4 ice in every single slot and the reactor at a maximum of 299 heat.

So, what do you like? Water, or ice.

I know, but I like the disco effect.

I have a 4.5 Ghz 2500k. I don't really care.

Come on people, I hope you can see that this is a prove of concept, you really think I mined for those items? The extraction, bluetricity and bucket filling is all just something I build in a few minutes and its just to prove that you can live on the edge with bucket cooled reactors (only as many slots of buckets as you need and not 1 more). And as a reaction on the efficiency of the uranium: what is the efficiency difference between this and this if we are talking about efficiency? Just so you know, I don't say this is a good design, I just want to know if you reconsidered ice or water or if you know the changes and are perfectly happy with what you are using and you might share your experiences with others (that is for me the idea behind a forum, but I could be wrong).

@GregoriusT I inproved on the Red and the Blue redundancies and disco-lights. I also insulated the wiring which you suggested in a tutorial which you conveniently link to in every single one of your posts. And replaced the timers with only 1 which is set at 1 second. I hope you like the changes.

@jppk1 Let us do the math and use your simple formula and the numbers given in my openings post. Your formula: "(ice casuc output-cooling consumption on average)/ice casuc output*the efficiency that the uranium setup has=the actual efficiency which you can actually compare to other rectors!"

For the Ice cooled reactor 404 mil EU(): (2070-50)/2070*4.40 = 4.294
Or a 100 EU/t cooling system (394 mil EU): (2070-100)/2070*4.40 = 4.187
Water buckets (404 mil EU): (2020)/2020*4.39 = 4.39

The system that GregoriusT uses (368 mil EU): (1840-0)1840*4.38 = 4.38

So yes, the efficiency is lower but only by a very small amount (2.24% with a 50 EU/t cooling system and 4.84% for the 100 EU/t system). But the funny thing is: at the end of your cycle you have just as much, more EU or a little bit (2.5%) less but with the added benefit that you can use a logic-circuit (a nand-gate) which can actually indirectly check if the reactor is cool enough and when it is running out of ice. With a bucket cooled reactor you can't easily check if all empty buckets are replaced with full ones. This makes you rely on an other system to shut down the reactor before it blows up in case the extraction system got out of sync with the ticking of the reactor (which has happened many times in my nuclear-engineering test worlds in the past). The worst part about this "efficiency" story is that it only takes into account the actual EU produced by the uranium. So it is not about how many EU you have left from building a reactor (with all the costs of the investment) and running it for a cycle, instead it is about how efficiently you use the uranium you mine, which in my opinion is absolutely not f-ing important. The power-plant I talked about in the opening post used 8 breeders and 8 reactors. This setup is self sustaining. It uses 336 (in the 8 reactors) and 48 (in the breeders to enrich the cells) uranium cells in 1 cycle. It also produced 336 uranium in 7 of the 8 breeders to run the 8 reactors for another cycle and 48 uranium in the last breeder to re-enrich to next batch of depleted cells in the breeders. The nice thing was that I always got about 1 out of 4 uranium cells i burned back in the form of a depleted uranium cell. This combined with the recipe for the depleted uranium cell meant I only needed 36 uranium per cycle. The output was 8*60+8*1820 = 15040 EU/t. This for 10000 seconds is more than 3 billion EU, which meant I needed 300 MFSU's just to store the energy of 1 cycle. I did this in a 10 by 3 wall which was 10 long so that I could output just a little faster than I could produce at 15360 EU/t. Yes, I did spawn everything because I promised my users a power company and I had to build it fast. But to get to the point, my users payed for the energy with uranium (1 for 10 million EU, for which I still feel bad) and they got the uranium in a legit way (I am the admin on my server and also the only one not playing legit, so I also felt entitled to use EE for the ice) and not by using a quarry (this was prevented because I failed at getting LogBlock and Buildcraft quarries working together). So the power-plant produced 12*8+364*8 = 3008 million EU with 36 uranium, so nearly 83 million per uranium, which meant an efficiency of nearly 42 (this is also why I still feel bad about my ridiculously high price for my users). My point being, uranium is not a problem when you use ice or bucket cooled reactors and breeders, so I don't waist my time with your formula. I look at safety, energy output per reactor - energy used by the cooling system per reactor and of course lag (which I never heart my users complain about lag (6 core ramdisk server with 8 gig of ram for 8 users might also have something to do with that, but that's not the point)).