Decentralized Distribution of Steam (DDoS)

You can also use ComputerCraft computer to run the whole system. Instead of worrying about all the logic gates, you can run bundled wires. Set the filters attached to the reactors themselves to color A, the filters attached to the chests as color B, and the reactor to color C

The program "transfer" reads:

-- This subprogram will pulse filter 24 times
-- in sequential order. This will transfer cooling
-- cells properly.
-- Bundled Color Code:
--  Orange = Filter on towers
--  Magenta = Filter on holding chests
for i = 1, 24 do
  rs.setBundledOutput("back", color.orange)
  sleep(0.1)
  rs.setBundledOutput("back", 0)
  sleep(0.5)
end
sleep(1)
for i = 1, 24 do
  rs.setBundledOutput("back", color.magenta)
  sleep(0.1)
  rs.setBundledOutput("back", 0)
  sleep(0.5)
end

This will effectively pulse the reactor filters 24 times, emptying them of the cooling cells and putting the cells into the chest. Then it will pulse the chest's filters 24 times, completing the transfer.

From there, the logic is simple:

for i=1, <number of micro-cycles per cycle> do
-- number of micro-cycles per cycle is a pre-defined number rather than a variable, although I could make a local variable to input here
   rs.setBundledOutput("back",  colors.white) 
   -- white is the redstone color that turns on the reactor
   Sleep(<micro-cycle duration>)
   -- micro-cycle duration pre-defined, not a variable, although I could always create a local variable to input here
   rs.setBundledOutput("back", 0)
   -- turns off reactor
   transfer
   -- the previous program I just got done debugging, swaps nearly empty cooling cells with fresh ones from the cooling tower
end

Since a ComputerCraft computer is just a couple of redstone and some smooth stone, it's fairly cheap to run.

This program should also run any number of chambers in serial, as long as all the reactor filters are on one color, all the chest filters on a second, and the power-generating reactor on a third.

I have encountered a problem. Apparently, filters with a full cooling cell don't want to pull out cooling cells that aren't full... this poses significant problems for my design...

Quote from brumster

Ok, I think I've got this sorted now - will leave it for a few cycles just to verify things.

Did things a little differently, but only slightly. For starters, I kept my main reactor to only a single effective chamber so there are only 9 cooling cells in there - this has allowed me to use a RP Regulator to feed it with coolant cells. Fill the input and output buffers with 9xcells and the regulator will swap all 9 straight into the reactor in one quick step, negating the need to pause the restart of the reactor to allow the pipes to clear.

A filter on the other side of the reactor (with a single part-used(!)) cell pulls them out, and this needs pulsing 9 times of course. This pipes into your cooling reactors with the same layout of components as you've used - just a grid of component heat vents and cells.

Being pulsed by the same pulse generator is a retriever connected to all of the cooling towers. This pulls remotely out of the reactors and just saved lots of filter, but I guess you could do it with filters just as easily (remember to use RECHARGED coolants in your retrieve/filter buffers so it only pulls out fully cooled cells). Downside of the retriever is the obvious need for blutricity.

Anyway, that then pipes the cells back to the reactor - to the above-mentioned regulator.

I put enough coolant cells to fill everything bar 9 slots across the cooling reactor - so start the system off with a full main reactor, *a full regulator* (so that it's ready to go filling the reactor on cycle), then fill as much of the cooling reactors as you can but leaving 9 empty slots - otherwise you obviously won't have space to dump out the used cells from the main reactor.

I'll do a vid later, once I've tested a few cycles and made sure it can cool enough to stand infinite of them

Display More

Ahh, so any partially-used cooling cell will pull any other partially-used cooling cell? It doesn't require the exact same metadata? Excellent!

Quote

edit: Ha, no was the answer - that'll teach me - to busy writing this forum entry and it went boomski on me

I think I may have a reason why, and it is rather disheartening...

In effect, I may have made a miscalculation based on a confusion surrounding the terminology of the word 'tick'. Normally, a 'tick' is 1/20th of a second. However, the reactors apparently consider a tick to be a full second. in other words, the cooling towers are some twenty times less efficient than originally calculated, with a cooling cycle time of some 80 minutes. Nuts.

Another option, I'm not certain how viable it would be, is to simply destroy the used cells and make new ones. It would consume 49 Tin and 8 copper per cell doing so... which would be 1176 tin and 192 Copper per micro-cycle consumed. I don't know what the UUM costs on that would be, if it would be worth it in the long run, but it's at least an option.

UPDATE: According to my calculations, replacing 24 60k cooling cells using UUM to replace tin and copper (assuming consistent scrap supply) is 292,800 per micro-cycle. Using this reactor as a test-bed, each micro-cycle would generate over 9 million EU. Thus estimated energy loss is somewhere around the 3% mark. This would make it very viable indeed.

Quote from brumster

Yes, thankfully the metadata isn't important other than it's either 0 or non-zero.

I've got 6 cooler reactors cooling a 9-quad-uranium (and 9 60k coolant) 780EU/t reactor. There's slight over-capacity on the coolers I think, because 1 is always full of recharged coolant cells and another is partially used, so it probably only needs 5. It also needs (25*6)-9 cells for the cooling towers, 9 for the main reactor, and 9 in the loop ready to slot in - 165 cooling cells.
By my reckoning, then, the whole of my setup needs :-

3428 copper, 7642 tin, 36 bronze, 1726 iron, 49 rubber, 56 redstone, 14 glowstone, 14 lapis and 990 water cells (so about another 250 tin).

This is quite depressing really, I figure you can probably make a few smaller Mark 1 reactors pumping out a constant ~100EU/t and it would be far more cost effective. Damn the loss of CASUC! However it is quite a pretty system, I think - I like this concept of it.

edit: But when there's a Mk.I EB design in the reactor designs thread putting out a low-risk 420EU/t for just 1031/65/0/316 copper/tin/bronze/iron, it's just not worth the hassle is it? Jesus the reactor design is fubar'd in IC2

Display More

You are thinking too small, I think. The advantage of this system is only achievable on the macro scale. When you have something like a five-chamber operation producing 2700 EU/t, then it might be worth it to have a dozen or so cooling towers spaced around in tandem.

Anyways, I'm still trying to work a couple of angles to it, but it doesn't seem to really be resource-efficient, even if it is a rather novel idea.

Your limitation to your regulator's capacity can be bypassed simply:

Filter attached to reactor, set to pull out expended 60k cooling cells (will not pull out full ones)
Regulator attached to Reactor, set to pull full 60k cooling cells (will not pull partially charged or empty ones) into the reactor
They are both on the same redstone wire, so every time the filter pulses, the regulator pulses, replacing cells on a one for one basis

Then you have a State Cell and TImer set to pulse that redstone wire x times, where x is the number of cells in your reactor

The state cell is triggered by a wire with two possible inputs:

1) MFSU set to Emit If Full (if all of your energy storage is full)
2) if Heat Sensor detects 100 heat it will emit redstone signal

If either of these emits a signal to the state cell, it will turn off the reactor and trigger a transfer and will keep it off until the latch is thrown. That way, every time the reactor turns back on, it is full of clean cells. At some point in your MFSU line (wherever you feel it would be appropriate) is a Structure Pipe with a Gate set to emit a redstone signal if energy is empty. This triggers the latch to reset the system. However, it still won't trigger if there is any heat in the system.

You could also use ComputerCraft to control the system, putting both the filter and the retriever on one color and the reactor on a second.

rp.setBundledOutput("back", colors.<nuclear>)
-- turns on the reactor, <reactor> is the color of wire hooked up to the reactor itself
Sleep( # seconds to run for micro-cycle)
rp.setBundledOutput("back", 0)
--turns off reactor
for i=1, 24 do
rp.setBundledOutput("back", colors.<filter/retriever>
-- where <filter/retriever> is the color of wire hooked up to both the filter and retriever
sleep(0.1)
rp.setBundledOutput("back", 0)
sleep(0.5)
-- This pulses a redstone signal for .1 seconds, then waits .5 seconds to pulse again.
-- Basically, a Timer set to .5 Seconds, adjust second sleep number as necessary
end
-- This loop functions as your state cell, telling it how many times to pulse
sleep(2)
-- time buffer to make sure that all of your cells make it into your reactor before pulsing. 
-- feel free to adjust as necessary, you do NOT want your reactor turning on without all
-- cells in the reactor

That is one full micro-cycle. Then just run it on repeat until you want it turned off. You can use a While loop with a redstone wire set as your double-fault to turn off reactor and shuffle cells when it turns on.

I had a reactor able to output 1120 EU/t using the method you proposed with the retriever, however it required 10 cooling towers to function optimally. There was one I had with an output of 2700 EU/t, but it required mid-cycle cooldown periods in addition to all those towers in order to not explode. There was also a 1920 EU/t variant on that which used reflectors on the 'caps' to generate an Efficiency of 6.0.

In other words... this is likely not an economically viable build. Still a lot of fun, though.

Quote from brumster

Absolutely

I suspect I'm like many - I have a creative test world that I muck about with ideas on (fine-tuning turtle programs at the moment) with an idea to implementing the good ones in my play-it-by-the-book survival world. I think this is one of those ideas that's won't migrate across, but has been interesting to do all the same

You might want to look into viability of a true CASUC using cells and just voiding them and making new ones rather than having multiple cooling towers and see how effective it is. I think it's like 3 UUM for 10 Tin or something like that, so each cooling cell is probably something like 15 UUM all told. If the EU output is something like 2700 EU/t, it might be viable, depending on how the numbers work out...

Rather than going for raw power output, I tried for efficiency. In the Optimal Reactor List, only ONE reactor has an Efficiency higher than 5... and while it does hit an Eff7, it only produces 140 Eu/t and a total EU output of 28m.

I wanted to do better than that...

Now, a simple one-chamber ractor can crank out something like this.

960 Eu/t, Efficiency a whopping 6.0, and a total EU output of 192m.

Only problem is that it has a micro-cycle time of just under three minutes. Not cool. However, there is at least partial compensation here...

Our Turbo V6 Cooling Tower can handle 24 cooling cells. Which means one cooling tower can handle three such cycles. Still, at just under 3m each micro-cycle, I'd STILL need about 8 Cooling Towers to keep up with cooling. And that comes out to only 120/tower. Not good enough.

Let's see if we can do better still...

If we employ GregTech Helium Coolant Cells, worth 360K cooling, we end up with a micro-cycle of just under 18 minutes. Normally, that wouldn't really help us, because the time it takes to cool them back down is also multiplied six times, but now we've managed to get 10% of a full cycle in a micro-cycle. So now we only need 10 micro-cycles per cycle rather than significantly more. And at 3 micro-cycles per cooling chamber, you'll only need 3 chambers to complete a full cycle! This brings our effective power generation up to 240/generator, so we've beaten the 7 Efficiency tower in terms of power per tic output, and we've got -way- more total EU per cycle.

But what if we wanted to do even better?

Thorium is an interesting fissionable material. While it produces less EU/t, it does have the advantage of also producing less heat per tic, in exchange for a longer run time. So lets swap out the Uranium for Thorium and see what we get.

At nearly 12 minute cycle, using standard 60k cooling cells, our cells cool off in roughly 6 and 2/3 a cycle. Which means only two or three cooling towers are needed. Of course, the 192 EU/t is fairly lackluster for such a system, but it's got an efficiency of 6 and a total output of 192m. It's designed more as a workhorse reactor, producing a constant amount over time, with a high efficiency rating one would expect from a long-term reactor.

Helium cells wouldn't change these numbers much, unfortunately. Because Thorium runs 5x longer than Uranium, even the very impressive 71.5 minute micro-cycle time would still only represent about 1/12th of a total cycle. So again, you'd need about three cooling towers, just to be safe, and your cooling cycle would hit before your end cycle would. However, it would cut down the number of micro-cycles by a factor of 6, so that's probably worth it, considering Helium is generally a byproduct of producing other materials.

So, let's see if we can do even better...

Plutonium is a high-energy fissionable material which is a royal pain to produce. Which means you *probably* want a high efficiency rating out of it. Unfortunately, it also produces an insane amount of heat buildup. So lets see if our system works well here...

Using the same setup of 8 quad-cells flanked by 8 cooling cells and a couple of reflector plates to bring up the efficiency to 6, we see a rather disappointing micro-cycle time of 80 seconds. Since Plutonium has twice the lifetime that Uranium has, you're looking at two HUNDRED fifty-ish micro-cycles per cycle... yeesh. However, we're -only- looking at a stunning 62.5 microcycles in a cooling cycle. That means about 20 cooling reactors.

Having said that, you are looking at 1920 EU/t, so comparable to old-school CASUC level power generation, and Efficiency of 6, and a total EU output of some 768m.

Now, Helium cells will reduce the number of micro-cycles per full cycle by a factor of six, meaning suddenly you're only looking at a total of 42ish cycles. Which means only 14 cooling towers needed for a full cycle.

There's another advantage to this system that I wasn't leveraging with previous designs. As i am only utilizing 8 cells in the active reactor, I can use a filter and a retriever with a total of 8 cells, meaning I only need to pulse them once to get all eight to transfer. This means that Bumbster's setup is more than enough, if he just adds a few more cooling towers to his design.

It's probably not the most cost effective method yet, but it's at least quite efficient

Quote from Shakie666

Hi, i'm new to the forum, and all I can say is WOW, whoever came up with this is a genius! I never used tekkit, so I never had a chance to play around with the old ice-cooled CASUC reactors, but I've been looking for a way for get more oomph out of my nuclear setup for quite some time. I especially like the look of the plutonium setup, (the one that gives 5440eu/t), but I don't know how many cooling towers i'd need. I don't really mind how many it takes, since they don't degrade or anything, but I don't want to build more than necessary. Also, just thought i'd ask: would using helium cooling cells really be any better than the normal ones? Because they store 6 times more heat, but they also take 6 times longer to cool down. Or am I missing something?

Another small question: what does CASUC stand for? Was just wondering.

The 5440 EU/t reactor has an efficiency of only 5.66, as opposed to several of the later designs with an eff6 rating. It also requires a full compliment of 24 cooling cells and a five-chamber reactor (as opposed to a single-chamber), which means each cooling tower can only handle a single cycle rather than three of them.

In the reactor you are looking at, assuming using the 360k He Cooling Cells, you have 42 micro-cycles per full cycle. Each cooling cycle is somewhere between 500 minutes and 375 minutes, depending on where in the cooling chamber they are. Since both of these times exceed the full cycle time by a good margin, you will want 42 cooling towers, plus a cooldown period of 500 minutes between full cycles.

The smaller reactors are more efficient. For a single chamber reactor with plutonium and the reflectors to make it an eff6 rated reactor, you're looking at 1920 EU/t with the same micro-cycle time of just under eight minutes, but since each cooling tower can handle THREE micro-cycles, you would only need 14 cooling towers.

Having said that, the smaller reactor with good ol' fashioned quad-uranium cells produces 960 EU/t and only needs three of cooling towers, so you are looking at a higher EU/tic/tower ratio (looking at about 320 EU/t per tower average at efficiency 6, which actually isn't too bad a setup)