Factorio Forums

www.factorio.com
https://forums.factorio.com/

The perfect OCD compliant nuclear power plant! --- Now v2.4!

https://forums.factorio.com/viewtopic.php?t=47556

Page 1 of 8

The perfect OCD compliant nuclear power plant! --- Now v2.4!

Posted: Sun May 14, 2017 5:12 pm
by Distelzombie
Hi there,

NOW MODULAR! (With limitations)

I didnt really like all the other designs I could find in this Forum. :P
They either dont have storage tanks, their ratio was off, the combinators didnt really work or they looked weird.
Im going to change stuff in this build. Therefore I use a version number in Blueprint name and here.

Most designs only read if empty cells are extracted. Also most of the time the inserters were inserting 5 fuel cells at once. I searched for a good logic but couldnt find one. So I made my own:
Taipion provided me with a very good logic for this. It uses only two combinators and one is a constant. It fits perfectly because too many combinators or not perfect -> triggers OCD. But with version 2.4 I completely changed the logic again.

NOW WITH NEW IMPROVED LOGIC! (Im a salesman! Yay!)

Properties:
- I followed the calculations from Dimanper: 0.15 Reactor ratio - Thread
- Four different version: 2, 4, 6 and 8 reactors
- Does not waste fuel
- Can store a full load of energy. (Like, what is produced when all the reactors use one fuel cell without any power demand)
- Can actually live off of the saved steam without starvation! Meaning...
- No brownouts ever! (Means: Power production not enough for demand)
- And no Blackouts! (As long as there is fuel)
- Small shore line
- Has its own fail safe power source for all the necessary parts to produce power! (For all pumps, inserters, combinators...)
- Easy initial start! Just turn the fuel-cell constant-combinator on.
- Alarm if not enough fuel cells for the next two runs in the chest for every reactor
- Alarm if backup power runs out of energy.
- Alarms are NOT HORRIBLE when going off! xD (More like chimes)


Usage:
Once all the chests have atleast one fuel cell in them, turn the constant combinator on that has the fuel cell symbol on it!
This way you have one fuel cell inserted automatically whenever you turn this constant-combinator on again.
- You can change the minimum-steam-threshold using the value T. You can find it in a constant combinator in the control section. I fiddled with it: When power consumption was very low before and there is steam in system, the build can partially starve on steam if value too low. (Brownout) Therefore a good value is greater than 1000k. If you dont except HUGE fluctuation (+-500MW all of a sudden) in power consumption it doesnt really matter much. If you set the value too low the heat transfer will lag too much behind and this causes a brownout.
- Do not connect the upper part to your base power network! It is seperate to act as a fail safe in case of fuel shortages.
- You can place a bigger blueprint on a smaller on without problems - as long as you didnt change anything.


Now OCD compliant! :D
The logic does work exceptionally good! Thank you again, Taipion.
I tested this setup with various power demands: 100MW, 300MW, 600MW and full power and got no brownouts in between or at all.


-------------
This Blueprint book has four version in it: 2, 4, 6 and 8 reactors. Biggest one uses 8 reactors, 112 heat exchangers, 192 Turbines (one less than perfect because of the symmetry) and 108 tanks. It produces 1114.2 MW or power. (Because one turbine is missing for symmetry reasons it is not a perfect 1120MW.)
The logic can be extracted and used for any other build. Use provided Blueprint.

Due to a 60,000 charackter limitation in posts I have to provide you with this link instead of a string. (About 155,000 charackters)
The perfect OCD compliant power plant! v2.4 BLUEPRINT BOOK
-------------


Just the 8-reactor build:

If you find a good place for a radar, please let me know!
8 Reactor build v2.4
Extracted Reactor Logic v2.4

Re: Perfect reactor design! Completely controlled by combinators

Posted: Sun May 14, 2017 5:12 pm
by Distelzombie
Reserved for version log.
Im going to highlight the newest change.

Version 1
UPDATE: The turbine-segments needed to be connected. Now the setup can produce a perfect flat line of whatever MW your base needs. It was fluctuating before because uneven heat distribution. (game bugs) (Version 1.1)
UPDATE: Changed the empty-fuel-cell value to prevent power plant getting stuck. (Version 1.2)
UPDATE: Included a constant combinator in the control section to make the initial start easier. Instruction are spelled out in constant combinators. (Version 1.3) I also updated the turbine module to include the steam balancing pipes
UPDATE: Changed some combinators. The tick generator for the requester chest circuit is gone and the first combinator of the memory cell wasnt needed. So now three combinators less. (Version 1.4)
UPDATE: Changed the look of it dramatically! Now it doesnt trigger my OCD anymore. (Version 1.5)
UPDATE: Could get rid of two more combinators. Yay! (Version 1.6)

Version 2
UPDATE: Since heatpipes work totally different now the design did no longer work at all. Took nine trys but now I made have a working one! xD
UPDATE: Ups. Had a combinator wired wrong. Fixed it (Version 2.1)
UPDATE: Totally new combinator setup. No only two are needed! Independant power grid! Also other optimizations like less pumps. (Version 2.2)
UPDATE: Added three more versions with 2, 4 and 6 reactors.
UPDATE: Minor OCD related changes. Added roboports on the sides for tiling. Added tanks to 2, 4 and 6 reactor build to go slightly over perfect ratio.
UPDATE: Huge combinator improvements! Now with more useful Alerts and a start-combinator you can use to input fuel initially or in testing (Version 2.4)

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 5:48 pm
by MicFac
Awesome setup, but wouldnt the inserters put more than 1 fuel into a reactor in the time where bots take the empty cells out of the active providers?

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 6:00 pm
by Distelzombie
MicFac wrote:Awesome setup, but wouldnt the inserters put more than 1 fuel into a reactor in the time where bots take the empty cells out of the active providers?
No. Beacuse they react to N and N is only one tick positive. This activates them only one time. And the inserter-stacksize is limited to one.
The time where bots can take the fuel cells out has to be higher than the actual time they need to do that. I chose 10 seconds.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 6:32 pm
by MicFac
I looked at the combinators again, but technically, the combinator "when A under 1 then A" should turn off in the same tick as "when F and S then A" turns on. That would mean that "when A over 1 the N" shouldnt turn on at all.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 7:02 pm
by Distelzombie
MicFac wrote:I looked at the combinators again, but technically, the combinator "when A under 1 then A" should turn off in the same tick as "when F and S then A" turns on. That would mean that "when A over 1 the N" shouldnt turn on at all.
They dont work that way:
X = "A < 1 then A" combi
Y = "A > 1 then N" combi
INPUT = "F and S then A"

1.tick: INPUT = off, X = on, Y = off // GREENa=1, REDa=0, n=0
2.tick: INPUT = on, X = on, Y = on // GREENa=1, REDa=1, n=1 (F and S = true)
3.tick: INPUT = on, X = off, Y = off // GREENa=0, REDa=1, n=0
...
100.tick: INPUT = off, X = off, Y = off // GREENa=0, REDa=0, n=0 (Because empty fuel cell < 8)
101.tick: INPUT = off, X = on, Y = off // GREENa=1, REDa=0, n=0
...
n.tick: restart

The "when A > 1 then N" combinator works together with the "when A < 1 then A" to form a tick generator. Meaning when input A is high it outputs a one tick pulse - regardless of how long A was high.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 7:14 pm
by MicFac
ok, at least I understand that part then :D

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 7:34 pm
by Distelzombie
Here. I highlighted stuff a little bit

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 7:55 pm
by Aeternus
You might want to save some room by eliminating some heat exchangers from this design. You need 3 heat exchangers per 5 turbines, and you'll still have a little extra left in terms of steam generation to refill the buffer. The reactor logic is a bit more complex then it needs to be I think, and I don't understand the need for isolating the logistics grid.

With either the fuel cell ejection -or- a 200 second delay as an option to signal when the reactors are ready for more fuel, isn't it easier to simply save a "fuel cell loaded" signal, clearing it after either 200 seconds, or when "spent fuel cell >0" is true, then check if your steam buffer is below a treshold, and insert more fuel as needed? That way the logistics grid can be connected, or you could even belt in/out the fuel.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 8:07 pm
by MicFac
Why are "when O > 0 then input count Everything"s and "when O < 2 then input count everything"s inputs connected to a green wire that isnt connected to anything else?

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 8:12 pm
by MicFac
Distelzombie wrote:Here. I highlighted stuff a little bit
ok, some off topic questions: where did you get all of that combinator knowledge and which program did you use to edit that picture?

Re: Perfect reactor design? Completely controlled by combinators

Posted: Sun May 14, 2017 8:59 pm
by Distelzombie
Aeternus wrote:You might want to save some room by eliminating some heat exchangers from this design. You need 3 heat exchangers per 5 turbines, and you'll still have a little extra left in terms of steam generation to refill the buffer. The reactor logic is a bit more complex then it needs to be I think, and I don't understand the need for isolating the logistics grid.

With either the fuel cell ejection -or- a 200 second delay as an option to signal when the reactors are ready for more fuel, isn't it easier to simply save a "fuel cell loaded" signal, clearing it after either 200 seconds, or when "spent fuel cell >0" is true, then check if your steam buffer is below a treshold, and insert more fuel as needed? That way the logistics grid can be connected, or you could even belt in/out the fuel.
The heat exchangers are all needed: viewtopic.php?f=5&t=44778&start=80#p269934 (You cant subtract an exchanger, but you can add one turbine. So good is the ratio)
The logistics net is isolated because I need the value of the empty fuel cells for a certain amount of time. I couldnt find an easier solution to have this one saved for the exact time it has to be there. You can connect your base-logistics network to the two chests in the control section. I dont see the big issue with that.
I fiddled with 200s delays for hours.

You're right. You can probably save one or two combinators. You still need a clock, a memory cell, a tick generator and control stuff thought. You wouldnt save much. Can you provide a blueprint or description?
MicFac wrote:Why are "when O > 0 then input count Everything"s and "when O < 2 then input count everything"s inputs connected to a green wire that isnt connected to anything else?
It is there to create the memory cells first part. You can make a memory cell with only one combinator, but you cant delete the value externally. Therefore you need to split it in two. The green wire is connected to the input AND output of "when O > 0 then input count Everything". Just like the red wire in the other combinator of the memory cell.
MicFac wrote:
Distelzombie wrote:Here. I highlighted stuff a little bit
ok, some off topic questions: where did you get all of that combinator knowledge and which program did you use to edit that picture?
I either use Gimp (If it has to be more controlable) or Paint for quick an dirty (I did this with paint) :lol:
I do some programming from time to time. This really helps getting the logic behind all of it. Combinators are essentially the same as the most common programming command: "IF". Of course I also watched a couple of tutorials that i cant link because I forgot all of them. Except Gepwin. Search youtube for Gepwin. :)

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 1:47 am
by Aeternus
Distelzombie wrote:
Aeternus wrote:You might want to save some room by eliminating some heat exchangers from this design. You need 3 heat exchangers per 5 turbines, and you'll still have a little extra left in terms of steam generation to refill the buffer. The reactor logic is a bit more complex then it needs to be I think, and I don't understand the need for isolating the logistics grid.

With either the fuel cell ejection -or- a 200 second delay as an option to signal when the reactors are ready for more fuel, isn't it easier to simply save a "fuel cell loaded" signal, clearing it after either 200 seconds, or when "spent fuel cell >0" is true, then check if your steam buffer is below a treshold, and insert more fuel as needed? That way the logistics grid can be connected, or you could even belt in/out the fuel.
The heat exchangers are all needed: viewtopic.php?f=5&t=44778&start=80#p269934 (You cant subtract an exchanger, but you can add one turbine. So good is the ratio)
The logistics net is isolated because I need the value of the empty fuel cells for a certain amount of time. I couldnt find an easier solution to have this one saved for the exact time it has to be there. You can connect your base-logistics network to the two chests in the control section. I dont see the big issue with that.
I fiddled with 200s delays for hours.
You're right. You can probably save one or two combinators. You still need a clock, a memory cell, a tick generator and control stuff thought. You wouldnt save much. Can you provide a blueprint or description?
My current reactor control design uses the "Eject spent fuelcell" to find out when the reactors can handle more fuel. Based on a simple latch. It only became feasable at 0.15.10 since before then, reactors ejected their spent cartridge immediately. The latch has only 4 combinators. It's described here:
viewtopic.php?f=208&t=47300
Before that I used a timer, described here:
viewtopic.php?f=16&t=45784&p=264704#p264704
Both designs use the logistics grid, but that grid is part of my main base's grid. It can be isolated but does not need to be, so long as there's some bots available from time to time to restock fuel and move spent cells to reprocessing.
Can't generate the blueprint strings at the moment due to computer issues on the system the save's on.

As for the exchangers: I got my math screwed up, you're right, the ratio is correct. 69.7 MW draw, 70MW production. Means the steam tanks will always fill up if the reactor is on, even if the plant is on max load.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 12:31 pm
by MicFac
Distelzombie wrote:The green wire is connected to the input AND output of "when O > 0 then input count Everything".
Sorry, but it doesnt seem to be like that in the blueprint. Image
The green wire is connected to the inputs of the two combinators, not to the input and output of one combinator.

I also dont understand the combinator "when R < 600 then input count R". It iterates fromm one to 600 which doesnt make sense to me. I would understand it like that:
1.tick: 1R input from "when O = 1 then R", 1R ouput because the 1R input is < 600
2.tick: 1R ouput because of a 1R input because "when O = 1 then R" doesnt output anything anymore, however the combinator has 1R input because of its own output from 1.tick.
3.tick: 1R input from its own output, so 1R output
4.tick: 1R input again, so 1R ouput.
This would go on forever, so its basically like the resetable memory, but not resetable.

Why is "when O > 0 then inout couts everything"'s input connected to a tick generator? I think its unnecessary because the signals the tick generator gets should both only be 1 tick already.

You need to put 1 fuel into each reactor to make the setup start up, right? Also, the ractors dont have fuel while the inserters need to put 1 in each reactor. As the setup is ratioed perfectly, in a base that permanently uses up all of the setup's power, wouldnt that mean that the base would be turned off for that very short preiod when the inserters are putting fuel into the reactors?

I could just copy your design, but I wanna understand it first :D

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 3:14 pm
by Distelzombie
MicFac wrote:
Distelzombie wrote:The green wire is connected to the input AND output of "when O > 0 then input count Everything".
Sorry, but it doesnt seem to be like that in the blueprint. Image
The green wire is connected to the inputs of the two combinators, not to the input and output of one combinator.

I also dont understand the combinator "when R < 600 then input count R". It iterates fromm one to 600 which doesnt make sense to me. I would understand it like that:
1.tick: 1R input from "when O = 1 then R", 1R ouput because the 1R input is < 600
2.tick: 1R ouput because of a 1R input because "when O = 1 then R" doesnt output anything anymore, however the combinator has 1R input because of its own output from 1.tick.
3.tick: 1R input from its own output, so 1R output
4.tick: 1R input again, so 1R ouput.
This would go on forever, so its basically like the resetable memory, but not resetable.

Why is "when O > 0 then inout couts everything"'s input connected to a tick generator? I think its unnecessary because the signals the tick generator gets should both only be 1 tick already.

You need to put 1 fuel into each reactor to make the setup start up, right? Also, the ractors dont have fuel while the inserters need to put 1 in each reactor. As the setup is ratioed perfectly, in a base that permanently uses up all of the setup's power, wouldnt that mean that the base would be turned off for that very short preiod when the inserters are putting fuel into the reactors?

I could just copy your design, but I wanna understand it first :D
Youre right. Apparently the green wire is not connected, huh. So that was the issue with reproducing the setup. I build it several times, but it would only work in the first build. If you build it from top to bottom it looks like it is connected then. That must have confused me.
Anyway this is correct, I was wrong. It is still a part of the resetsable memory.

"wouldnt that mean that the base would be turned off for that very short preiod when the inserters are putting fuel into the reactors?"
No that cant happen. There is always steam in the tanks. I guess, since turbines use 60 steam per second and the period you mention is one tick, even 200 steam in the system is enough to not get a blackout.
Even if power plant is under full load it builds up VERY slowly. I just changed my build to include a switch you can press to initially start the power plant. (Because inserting fuel manually sucks)

The "R < 600 then output value of R" combinator does only count up exactly 10 seconds. (Depending on the UPS of your base or game.speed) This is the timer I use to specify the time the bots can empty the active provider chests.
I'll use the picture below as help to clarify it.
  • N=1
    combinator A is outputting O
    tick generator does its stuff
    tick generator (B) ouputs O
    O value gets saved in C
    D outputs R
    Clock is ticking up to 600
    As long as R is bigger than 9 the output combinator ins outputting a empty-fuel-cell signal with value 100 into the requester chest
    E reads R signal, outputs O if R=598
    This O value goes through tick generator (I think thats not really necessary, but it doesnt hurt)
    Now memory cell has a second O input, making the O value on the second combinator 2. It stops outputting O=1
    D stops outputting R because O is now 0
    Clock stops working because no R input
    R is no longer bigger than 9 and output combinator stops the requester chest.
So, the clock would go on forever, but im cutting the R signal input from it after its finished - making it a timer.

------------------------
I just updated the blueprint to include an easy start of the system. It was a hazzle to manually input one fuel cell in each reactor.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 3:35 pm
by mophydeen
I like the look of it.
I'll test it out in a few hours.
I'm interested in the combinators.

I'd like the 4 outer pipes to go through the middle (or at least a few of them)

I might even add it to my book :)


I've made a linear reactor setup that you can extend endlessly in symmetry.
But this design(2x4) looks good and has better ratio.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 3:46 pm
by Distelzombie
mophydeen wrote:I like the look of it.

I'd like the 4 outer pipes to go through the middle (or at least a few of them)
Thank you :)
If you use the segment/module thing and turn them around then arrange them in the same positions, the build would be a little wider. (Well i tested this before I added the necessary steam balancer pipes... Oh i need to update the blueprint.)
Also you can use the module in any other reactor build.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 5:41 pm
by MicFac
Distelzombie wrote:
MicFac wrote:
Distelzombie wrote:The green wire is connected to the input AND output of "when O > 0 then input count Everything".
Sorry, but it doesnt seem to be like that in the blueprint. Image
The green wire is connected to the inputs of the two combinators, not to the input and output of one combinator.

I also dont understand the combinator "when R < 600 then input count R". It iterates fromm one to 600 which doesnt make sense to me. I would understand it like that:
1.tick: 1R input from "when O = 1 then R", 1R ouput because the 1R input is < 600
2.tick: 1R ouput because of a 1R input because "when O = 1 then R" doesnt output anything anymore, however the combinator has 1R input because of its own output from 1.tick.
3.tick: 1R input from its own output, so 1R output
4.tick: 1R input again, so 1R ouput.
This would go on forever, so its basically like the resetable memory, but not resetable.

Why is "when O > 0 then inout couts everything"'s input connected to a tick generator? I think its unnecessary because the signals the tick generator gets should both only be 1 tick already.

You need to put 1 fuel into each reactor to make the setup start up, right? Also, the ractors dont have fuel while the inserters need to put 1 in each reactor. As the setup is ratioed perfectly, in a base that permanently uses up all of the setup's power, wouldnt that mean that the base would be turned off for that very short preiod when the inserters are putting fuel into the reactors?

I could just copy your design, but I wanna understand it first :D
Youre right. Apparently the green wire is not connected, huh. So that was the issue with reproducing the setup. I build it several times, but it would only work in the first build. If you build it from top to bottom it looks like it is connected then. That must have confused me.
Anyway this is correct, I was wrong. It is still a part of the resetsable memory.

"wouldnt that mean that the base would be turned off for that very short preiod when the inserters are putting fuel into the reactors?"
No that cant happen. There is always steam in the tanks. I guess, since turbines use 60 steam per second and the period you mention is one tick, even 200 steam in the system is enough to not get a blackout.
Even if power plant is under full load it builds up VERY slowly. I just changed my build to include a switch you can press to initially start the power plant. (Because inserting fuel manually sucks)

The "R < 600 then output value of R" combinator does only count up exactly 10 seconds. (Depending on the UPS of your base or game.speed) This is the timer I use to specify the time the bots can empty the active provider chests.
I'll use the picture below as help to clarify it.
  • N=1
    combinator A is outputting O
    tick generator does its stuff
    tick generator (B) ouputs O
    O value gets saved in C
    D outputs R
    Clock is ticking up to 600
    As long as R is bigger than 9 the output combinator ins outputting a empty-fuel-cell signal with value 100 into the requester chest
    E reads R signal, outputs O if R=598
    This O value goes through tick generator (I think thats not really necessary, but it doesnt hurt)
    Now memory cell has a second O input, making the O value on the second combinator 2. It stops outputting O=1
    D stops outputting R because O is now 0
    Clock stops working because no R input
    R is no longer bigger than 9 and output combinator stops the requester chest.
So, the clock would go on forever, but im cutting the R signal input from it after its finished - making it a timer.

------------------------
I just updated the blueprint to include an easy start of the system. It was a hazzle to manually input one fuel cell in each reactor.
Thanks for all the help!
Distelzombie wrote:Clock is ticking up to 600
Yes, but I dont really understand why. I already said how I would interpret what the combinator does, so what did I interpret wrong?

Connecting input and output of " when O > 0 then inout count everything "together seems to break the system. This combinator doesnt seem to need a green wire there at all.

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 6:33 pm
by Distelzombie
MicFac wrote:Yes, but I dont really understand why. I already said how I would interpret what the combinator does, so what did I interpret wrong?

Connecting input and output of " when O > 0 then inout count everything "together seems to break the system. This combinator doesnt seem to need a green wire there at all.
I thought i was clear:
The "R < 600 then output value of R" combinator does only count up exactly 10 seconds. (Depending on the UPS of your base or game.speed) This is the timer I use to specify the time the bots can empty the active provider chests.
Ok let me add to this quote of yours...
  • 1.tick: 1R input from "when O = 1 then R", 1R ouput because the 1R input is < 600
    Yes. BTW "when O = 1 then R" is combinator D
    2.tick: 1R ouput because of a 1R input because "when O = 1 then R" doesnt output anything anymore, however the combinator has 1R input because of its own output from 1.tick.
    Wrong assumption. D does output R = 1 all the time until it is shut down, over detours, by the clock reaching R = 598. Thats why I need a memory cell to store O =
    1
    3.tick: 1R input from its own output, so 1R output
    The clock does output R into itself again, adding the R= 1 output to its R = 1 input from D - essentially adding 1 every tick
    4.tick: 1R input again, so 1R ouput.
    Now the clock outputs R= 2 and adds that to the R=1 value from D, outputting R=3 and so on ...
Now if the clock reaches R=598 the E combinator sends an O=1 signal to the memory - deleting it because now the O value is 2 in the second memory combinator - which stops the memory output of O=1 to D and D stops to output R=1 which gives the clock no value to add to its own saved R, stopping it.

But this is the same thing as I said in the post before. Sorry, I cant get more specific. ;)

Re: Perfect reactor design? Completely controlled by combinators

Posted: Mon May 15, 2017 7:38 pm
by Distelzombie
Haha, sorry I just changes this logic! :D I could get rid of three combinators. (Aeternus was right. But he didnt knew why or didnt bother with telling me.)
But maybe it is more easily understandable now:
Tick generator is gone and the first combinator of the memory cell wasnt needed.

I'll update the blueprint when I finished testing the blueprint to make sure im not uploading crap.
All times are UTC
Page 1 of 8

Powered by phpBB® Forum Software © phpBB Limited