Atomic Science Fuel Life Config?

[Digdug83]

I realized a few days ago that my fusion plant was absolutely chewing my tin supply alive and as such my AE solution for feeding it is clearly inadequate. I have pretty much zero experience with constructing any kind of timing mechanism so if anyone has any tips I'd appreciate it.

[TokiWartooth]

A few pages back I explain how I have a rednet controller setup that controls a timered input of only 4 cells every 10 minutes or so.

[Flextt]

Hey guys,

first of all: Thanks for the great design ideas to Charliechop. It is a huge difference in power gen if you use the liquiducts. However, I am having some issues with my rather simple design.

The design has space for 2 more Fusion Reactors (without any regulation setup like Thermometers btw^^). After doing some testing, I decided to go with 7 Large Turbines per Reactor, because of several reasons:

- The Reactors need huge amounts of energy, effectively cutting net MJ production in half

- The buffer system often seems useless, because sometimes the liquiducts fill up to 70000 mB with 100 mB/t throughput and feed excess into it and in another run, they do not fill it all and seem to be in a constant state of equilibrium with a steam amount of 500-1000 mB inside the liquiducts, 0 in the tanks and a low throughput of 50-60 mB/t

- The Fusion Reactor runs in a batch mode. A continuous setup is currently not feasible for me due to Tin constraints and simply because I do not need such huge amounts of energy.

- The turbines feed energy back to the Reactors, which are MJ-independent for a while by filled RECs connected solely to them

- The design works, but needs optimization, so any help is appreciated.

My questions are:

- Is there an improved geometry for the Reactor design? (Pretty sure there is, because 2 plasma spawns are basically unused)

- Is there a viable, functioning buffer system? I tried it with TE Portable Tanks but did have even more problems with constant equilibrium. Is it even necessary? What is the limit of the liquiducts inherent storage capacity? Should I cut down on piping to improve distribution?

- Some Turbines produce less than 50 MJ / t even though steam is available, which I observed by the respective REC not filling up with an I / O MJ-ratio higher than 1. Does someone have an approx. value of steam consumption / t of a large turbine?

- Should I increase the number of Turbines per Reactor?

- Does the Reactor increase steam production over time, like a ramp-up time?

- Do I need to take care of heat even without plasma leakages?

I know it's a lot but I hope I have given all the information someone needs to answer them.

edit: did 2 16 cell runs, in one run the liquiducts were below 1000 mB, in the next I am exceeding 82500 mB in the ducts and feeding excess in the buffer... switched to using the diagonal setup for better plasma usage and did it in quick succession. luck or heat related?

edit2: I increased the Large Turbine per Reactor ratio to approx. 13-14 with a small excess of steam. The buffer system also works fine. The magical number seems to be 16 cells, after which the steam production either massively increases or the turbines are saturated. Given the 200 MJ/t energy requirement of the Reactor, things are looking up now.

[jakalth]

For 1, your piping system and duct work is spot on. Looks to be a good setup for catching and using all the steam produced by your current reactor design. Increasing the output of your reactors will require an increase in the number of ducts needed to move the steam from the reactor. Liquaducts running in parallel will help a lot for this.

- Is there an improved geometry for the Reactor design? (Pretty sure there is, because 2 plasma spawns are basically unused)

you have 2 primary options here. More efficient looped reactor, or a flood type reactor. Both types of reactors work slightly better for different uses. Flood types seem to work better for constant running reactors and scale up better with more reactor cores. Ring types seem more efficient as single core reactors that run in pulses, not continually.

Flood type: Placing your reactors in the same configuration you have them now, with the same distances between them, is a start. Have a single electromagnet touching each face of the reactor cores 4 sides. Now make a single open cavity, with a layer of electro magnets. This cavity can be up to 7 blocks out from the side of each reactor, but with spacing between the reactors, you can shrink this down to only 5 blocks out for efficiency, but the larger the area, the more potential steam can be created. and line the outside of the reactor with electromagnets as well, to keep the plasma inside. It should look like a giant sandwich when done. Note: Diamond shape seems to be one of the better simple shapes to use with this design. with the sides of the plasma containment area having diagonal walls. It matches how the plasma flows this way. Look at DigDug's design for an example of a large design.

Improved ring type: Placing a single reactor core in the middle, make 4 rings around it. Start out with the reactor. Wrap the reactor in a 3x3 ring of electromagnets. Now from each corner, extend a single line of 4 electromagnets, like 4 spokes on a wheel. turn 90 degrees and add 4 more electromagnets to the end of each of these spokes. At the end of these, turn 90 degrees again and add 4 more, then turning 90 again, add enough to connect each ring to the one next to it. This gives you the outer dimensions of the 4 rings for the design. Finishing it off is done the same as building the single ring around your current design. An example can be found Here: Note: the magnetic glass and center cobblestone are only to make it easier to see the parts of the layout. Do not use magnetic glass on the top of the operational rings, they produce no steam.

- Is there a viable, functioning buffer system? I tried it with TE Portable Tanks but did have even more problems with constant equilibrium. Is it even necessary? What is the limit of the liquiducts inherent storage capacity? Should I cut down on piping to improve distribution?

Simple things first. Buffer is not completely necessary if you have a way to use 100% of the reactors output on demand. otherwise, it does help. But does not need to be very big. One of the best things to use for the buffer, size wise, are the Liquid managers. Each one holds as much liquid(steam) as 16 BC tanks. But each one needs it's own input and output to work. As for the liquaducts, decreasing piping will only limit how much steam can be moved at any given time. They are rather smart pipes and manage the liquid they are moving quite well.

- Some Turbines produce less than 50 MJ / t even though steam is available, which I observed by the respective REC not filling up with an I / O MJ-ratio higher than 1. Does someone have an approx. value of steam consumption / t of a large turbine?

Don't have much of an answer for this. but it does seem like the turbines do not always share the steam evenly.

- Should I increase the number of Turbines per Reactor?

This all depends on the output of your whole setup more then on a per reactor basis. If the liquaducts are always 100% full, you might want to add more large turbines so no steam goes to waste, or use the buffer to balance out steam usage preventing steam waste. Buffering only really works well with a pulse running reactor. A continually running reactor will just fill the buffer if you can't use all the steam right away, wasting the rest.

- Does the Reactor increase steam production over time, like a ramp-up time?

Yes. There is a ramp-up time on a fusion reactor. It can be as short as 2 seconds, or as long as 6, from a stone cold reactor to full steam output. So Ramp-up time is normally not much of a concern when a single cell can keep the whole thing warm for 7+ minutes. warm = producing steam

- Do I need to take care of heat even without plasma leakages?

I have never had issues with heat buildup under or around my reactor. But I always build my reactor so I can walk under them to add cells manually when needed.

As a note, I have a reactor I built on a server that has been running 24/7 for 2 weeks strait now. I have never had any issues with it over heating or having any kind of malfunctions at all. It runs on a 7 minute timer, being fed deuterium cells out of an endermen chest who's matched pair is hooked up to my AE network. A precision export bus keeps the chest full of cells and auto crafts them when they run out in the network. I use an autarchic gate on a wooden pipe to pull the cells out of the chest and feed them into the reactor. This gate only runs(energy pulsar) when a redstone signal is given to it by a computercraft computer. The computer has a repeater program set to output a redstone signal for 1 second, every 420 seconds. Each pulse from the computer causes the autarchic gate to pull 2 cells. To fix this I have the cells pass through a distribution pipe. This sends 1 cell into the reactor and the other is send to an item tesseract that goes back to my AE network. Not the most powerful way to run it, but it is highly reliable, efficient, and I can keep up with the tin supply using either a single quarry or a MFR laser drill.

[Flextt]

Thanks for the great reply!

I did some numbers crunching. Due to massive steam surges from time to time - I highly suspect the turbines' saturation here is the culprit, if there isn't something like a heat build-up period longer than a few seconds - the turbine steam consumption is incalcuable. I will try to find some sort of entry in the code, but the classes have foreign (chinese?) names...

FR = Fusion Reactor

LT = Large Turbine

T = small Turbine

Anyhow, using a multimeter and little testing setups, these are the results:

FR power consumption: 200 MJ / t

LT power production: roughly 20 MJ / t. It doesn't matter whether the liquiduct is full or half-full or if you attach 9 conduits to it.

T power production: roughly 2 MJ / t. Since an LT consists of 9 Ts, an LT is roughly 10% more efficient.

That also means, to have a meaningful way of obtaining MJ, you have to build at least 10 LTs and 1 T ( == 91 Turbines == 364 Bronze Plates == 1456 Bronze == 2184 Copper and 728 Tin, not including RECs and the Reactor itself), which net you exactly 2 MJ/t in the end + the small surplus production when the liquiducts are still filled.

It is also meaningful to point out, that many video guides put a REC below the FR. This is a great tactic to make sure, your FR starts after a down-time, if you are running it in batch-mode. Unfortunately, RECs can only output 100 MJ / t at maximum, which means you need at least 2 per Fusion Reactor. I cannot imagine a 50% downtime being great, although finding out exact numbers is difficult. Due to highly probable network saturation, the 2 RECs should be dedicated solely to the FR.

I will break down my existing setup in a more efficient electromagnet setup for higher steam yield, since I basically only net energy from 4 Turbines in the end^^

[jakalth]

Well, here is the key, in the current version of atomic science in Tekkit. You only need to feed 50+Mj/tick to the reactor core to get it operating effectively. It WANTS 200, but only needs 50+ to operate. That is where a single REC comes into play, it acts as a power regulator for the reactor core. This way, you can add, up to, another 150Mj/tick to your usable output without loosing functionality. Simple tricks like this are usually not mentioned in videos. Also, using magnetic glass on the top of the reactor is another common mistake that many video makers make. Magnetic glass produces no steam, hence no power when next to or under water. TokiWartooth brought this to my attention so I try to share it as well.

[Flextt]

We should try and collect the information. Atomic Science documentation is lackluster at best. This whole thread helped me more than any video guide

[Digdug83]

More atomic tests, huzzah! I started wondering how much power x mB steam produces. I also thought I'd answer a few questions I had about the turbines operate. I will preface this by saying that I have altered the turbine output multiplier to 10 (I think default is 4.5 but correct me if I'm off). I don't think I changed any other variables (steam ratio is 40 which I believe is default).

Before I get to the results I will state a few things that may or may not be known to everyone about turbine behavior that I discovered:

-The fewer inputs you have into your turbines the better. This isn't really an issue with small turbines obviously but with large ones it is. You will lose power if you have more than one steam input underneath it. The steam basically leaves the pipe/tesseract and vanishes for all intents and purposes.

-In the vein of the above note, it does not matter which face you attach the steam input to, so long as it's underneath the turbine as steam only travels upward

-For large turbines, you can only output the power from the top center face.

-Conduit will output the power from turbines regardless of its input/output state.

In the first test I ran some liquiducts from a portable tank filled with 8000 mB (max value, all subsequent tanks have this amount) of steam to a single small turbine which then had an REC sitting on top of it set to maximum input. I consistently generated 1246 MJ from the tanks I used (5) so a small turbine generates 155.75 MJ with a turbine efficiency of 10.

The second test was a large turbine with 2 connections to the tank. The output was 6664 MJ. This was not repeatedly tested due to my discovery of input connections and output.

The third test was a large turbine with 9 connections to the tank. The output was fluctuating between 5700-5850 MJ.

The fourth test was a large turbine with 1 connection to the tank. The output was consistently 10969 MJ. This means that a large turbine with only one input connection generates 1371.125 MJ with a turbine efficiency of 10.

The fifth test was 9 small turbines grouped together but not wrenched to make a large and connected to an REC via conduit. The output ranged from 5579-5678 MJ across 10 tests.

[phazeonphoenix]

Could you continue running tests but increase the "large turbine with one connection" count? I'm curious how the power will scale.

[Digdug83]

As requested, yet more steam tests for you. Again, these were done with a turbine efficiency multiplier of 10 and portable tanks containing 8000 mB each.

The first test is one tank hooked up to 2 LT's with one connecting pipe each. With one exception out of 5 (likely a tick rate thing) all of the outputs were 12939 MJ. The first test yielded 15142 MJ.

The second test is two tanks separately connected to one LT each with one connecting pipe and then linked via conduit to an REC. As with the first test, the first result varied but the next 4 yielded 21939 MJ. The first test yielded 21849 MJ.

The third test is two tanks jointly connected to two LT's with one pipe and then linked via conduit to an REC. This one I tested 7 times as I wasn't sure if I was somehow going dyslexic reading the numbers lol. I ended up with 4 results being 21894 MJ, 2 results being 21984 MJ, and one result being 21939 MJ. It's entirely possible I misread the two 21984's or that they were simply tick rate variances like the above tests.

[Flextt]

Great observation Digdug83, my steam consumption was significantly reduced thanks to your tests and reducing inputting liquiducts.

One question though: Do liquiducts fully ignore the only-upward-movement of steam?

[jakalth]

when transferring steam around, yes, they seem to ignore steam direction limitations. I believe it is more that turbines only accept steam vertically. You can add and remove steam from tanks from any side the tanks will accept/output from.

[jakalth]

ok... Here I thought the 4 torus design did not scale well. I scaled up the design a bit. Increased the size of the center rings from 3x3 to 5x5 and added a fusion reactor to the connecting corners of the rings.(the 4 outside corners). Also added two 3x3 rings to each of those added reactors. Ended up with 12 total rings and 5 reactors. Piped all this using steam funnels and liquiducts into 105 large turbines. Yep, that is a lot of turbines. And stored the power in 60 empty redstone energy cells. After 5 tests I found that

1: a steam buffer does not work with this design, lost 30% of the reactor's output to the buffer for some reason, when I added one...

2: When you add deuterium cells to the 5 reactors, it creates so much lag that for several seconds, nothing else can be done in game and it nearly causes the game to run out of memory.(memory usage peaked at 117%... not even possible, but that's how bad the lag got)

3: It has the capability to output a total of about 2,300Mj/tick

4: After 10 minutes running on a single cell in each reactor, yeah it's crazy I know... it output between 25,000,000 and 31,000,000 Mj after running 4 non-buffered tests.

5: it is not completely efficient. it was still producing steam for 3 minutes longer, but not enough to generate power out of the turbines.

The single reactor core design with just the 4 rings, was outputting 6,300,000 Mj after 8 minutes run time consistently. So the scaling is not smooth, but a lot better then I thought it would do. between 5-6 million per core... about an 85% power scaling?

How well does your reactor scale Digdug? CharlieChop? Trying to see if the type of reactor you two use really does scale up better. I have a feeling it should...

[Digdug83]

I use a single reactor in a diamond shape (3x3 ring around reactor and then the diamond ring at the edge of potential plasma generation). It's currently feeding 80 LT's from a liquid manager buffer system I heavily augmented with additional pipes and tesseracts. I'm dead certain I could add more LT's to the design but at the moment I'm just too lazy to bother with it. My cell distribution isn't as efficient as pretty much everyone else's in this thread but it gets one about every two minutes. I use a thermometer block set to activate at 100 and then an export bus set to move single items in the absence of a signal. I just wish the temperature sensor was more consistent as it depends on plasma touching the magnet it's attached to.

[jakalth]

Just wondering if you'd get a better power output increase if you added more reactor cores and upsized the diamond to go with it. Mainly, would it increase by more then 85% for each additional reactor added when making the reactor larger? I might have to try this out myself though, now that I have a better mental image of how you've made your reactor.

[jakalth]

Hmmm. interesting results here people. Made a diamond shaped reactor like Digdug83 has been using. tested it with 1 deuterium cell and 25 large turbines(needs more then just 25...). 8,300,000 Mj approximately off just the 1 cell. Ok, I have to admit, that is a bit more power then my 4 torus design can output. O_o 2 million more. I'll see what numbers I can get from a multi core reactor using this design and test the single core design with the correct amount of large turbines.

the reactor layout has the outside of the diamond being 6 blocks away from the reactor core. It is a flood type, and the area between the outer diamond and the central core is a 1 block high cavity with magnets above and below.

[jakalth]

Ok, found two things here, my atomic science config file got corrupted so the setting were all wrong... so the numbers I have been giving are also all wrong. and the diamond reactor is still far more effective then I first though, as well as my big reactor being as bad as I thought.

firstly, my big reactor was not actually getting 85% increase per reactor core, it was only getting about 50%... so instead of 25-31 Million, the numbers should be 17 - 21 million Mj. far less then I said earlier. it DOES scale as badly as I thought.

Secondly, after fixing the config file, I was still getting approximately 10,500,000 Mj per cell out of the single core diamond shaped reactor. compaired to the 6,300,000 Mj I get from my 4 torus design. about 60% more. Used 35 large turbines to handle all the steam as well, as apposed to only 25.

[Digdug83]

Wait, 10.5 million per cell? How in the hell did you manage that with only 25 LT's? I have my turbine efficiency just over double the default and with 80 LT's and assuming one cell holds for 2 minutes (which isn't very reliable) before the temperature drops to 0 I get around 8.45M at best (assuming my tick rate is 20/s). Are the magnets still generating steam after temperature drops to 0?

[jakalth]

I am using 35 turbines with the design, would take a massive buffer to hold the amount of extra steam this thing can dump out. I have found that the reactor can dump out steam for somewhere between 8 and 9 minutes using default configuration settings. After that it's all buffer, so about another 30 seconds, maybe. For at least 7 minutes, the turbines are running at max output. So the 10.5 Million is over ~9 minutes time.

Also, it doesn't seem to run at full efficiency unless I place 2 cells into the reactor. That is when starting it up from a completely cold state. Not sure about if I keep it running warm. Haven't gotten the reactor dialed in yet. But it seems to be pointing towards an 8 minute delay between deuterium cells. I won't know for sure about that until I can let it run for an hour at least. Might have to turn the delay back down to 7 minutes like with my 4 ring design.

Current results from setting up one on a server are 8.2 million with only 25 LT's and 32 liquid managers as buffers. There is a lot of wasted steam since the buffers are full within the first 2-3 minutes, and the 25 turbines can not use all the steam produced, even with the buffer. I'll be adding another 10 LT's to this setup as well and I can give better readings once this is done.

:Edit: Tested it with 35 LT's and the 32 liquid managers as buffers. Still filled the buffers within the first 2 minutes. Still used 2 deuterium cells to make sure it primed the whole reactor properly. This time got 14,040,000 Mj after 10 minute run time on those two cells. Looks like I'll have to add 10 more again, and make it 45 LT's on the reactor. Almost scared to see how much power it'll produce after doing that. O_o

[Digdug83]

Alright I changed my feeding system to something similar to jakalth's setup listed earlier in the thread.

I have an Iron OR gate set next to my steam storage which reads one of the liquid managers. When it detects the tank is empty it emits a red wire signal along a pipe to my reactor. At the reactor end I have another iron OR gate set to broadcast a redstone signal when it receives a red wire signal. It then activates a precision export bus which is set to turn on and single item/craft with signal. The chemical extractor is fed empty cells with an interface and imported back into the network. This way I don't have to mess with more ender chests and the system activates when storage runs dry. Given how many REC's I have as a buffer and the fact that they''re inconveniently located I did this instead of hooking one of those up to a gate. It makes it a continuous system rather than on demand but I also guarantee I'll never run out of power.

[Daemon_Eleuel]

Ok, I am planning my reactor right now and I already had a test up yesterday, but I'd like to get some answers to doubts I have here considering it is the most informative thread about it I have found.

The idea is a pretty big flood type fusion reactors.

My question is:

How far does flood plasma expand? And how many buckets of steam are produced by each heated water block are produced per tick?

Let me explain my doubts. I fear the liquiducts will be a bottleneck in the extraction of the steam and I want to make sure the whole area is flooded properly so the distance between the seeds should be as much as possible. At the same time I don't want to leave too much space.

At the moment the general design is this:

https://docs.google.com/spreadsheet/pub?key=0Apov5LEiNDQsdGxwYUFLTlR5eUZpM3VXMUtvU0I2aEE&single=true&gid=0&output=html

It is just a general idea for now and most probably I'll put the reactors on different axis to maximize efficiency, I was just wondering if someone had a diagram showing the plasma expansion field.

[Flextt]

Plasma expands up to 6-7 blocks in one direction.

To adress your fears: Unfortunately, calculating the exact the amount of steam per plasma has proven difficult for the think tank, that has gathered here. Reasons are the low amount of feedback AtomicScience gives about just everything, the comparably, previously low interest in Voltz and the bad documentation.

However, your only way to capture generated steam is through funnels, making this an issue of space. Hence, you have to weigh your options: Do you consider losing steam funnel for an additional parallel connected liquiduct - because that's the only way to do it with the given mechanics. (hint: don't do it)

[phazeonphoenix]

Unfortunately, calculating the exact the amount of steam per plasma has proven difficult for the think tank, that has gathered here.

It's a complicated mod that's relatively new and hasn't been fully explored yet. There are so many variables and as you said little feedback it makes agreeing on things like how much steam can one deuterium cell can create very difficult because we have to ask are you funneling it through pipes or using turbines on top or have this shape of core and etc etc.

The effectiveness of your reactor depends on how you design it. I decided not to focus so much on harnessing every little last drop of steam or producing the most steam for my buck and instead focused on continual uninterrupted long term power generation with minimal interaction.

[Digdug83]

The best you can do with steam output is cover the whole top of the area above your steam funnels with liquiducts (even over the glass) and then attach liquid tesseracts every two meters or so so they all have 4 faces being inputted by their own tubes. I think I have something like 20 tesseracts feeding steam from the reactor to storage and it's still not enough. It produces a metric load of it is all I can tell you lol.

Anyone tried using waterproof pipes to improve flow? I don't know how much they can carry but they're compatible with tesseracts at least.

Edit: What phazeon said. We've been swapping ideas back and forth and for the moment I'm pretty happy with my reactor setup, although getting all the liquid managers onto a gate network has proven impossible for me due to the lack of wiring options :(

[Flextt]

The best you can do with steam output is cover the whole top of the area above your steam funnels with liquiducts (even over the glass) and then attach liquid tesseracts every two meters or so so they all have 4 faces being inputted by their own tubes. I think I have something like 20 tesseracts feeding steam from the reactor to storage and it's still not enough. It produces a metric load of it is all I can tell you lol.

Anyone tried using waterproof pipes to improve flow? I don't know how much they can carry but they're compatible with tesseracts at least.

Edit: What phazeon said. We've been swapping ideas back and forth and for the moment I'm pretty happy with my reactor setup, although getting all the liquid managers onto a gate network has proven impossible for me due to the lack of wiring options

Hey digdug, could you clarify your tesseract setup? I tried using Tesseracts in a stupidier, easier way before, but they were a huge bottleneck and (<500 mB per Face, maybe?) since I am now considering switching to a flood type, I am not really sure how I should incorporate liquid tesseracts.