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

Power Plants, Energy Storage and Reliable Energy Supply. All about efficient energy production. Turning parts of your factory off. Reliable and self-repairing energy.
mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by mrvn »

Distelzombie wrote:
mrvn wrote:Fail prooving the circuits:

1) monitor fuel available for all reactors. It is better to wait for all to have fuel than to start only some. The output will be a fraction when the neighbour bonus is lost. I would add a buffer chest to each reactor that holds only 1-2 fuel cells if fuel is sparse.

2) remove the spend fuel cells into a buffer chest and use the chest contents as trigger. cells > number-of-reactors can be used as trigger. Use a flip flop or a token to remember that spend fuel cells had been present. I think a token works best because you can put it on a belt to add a delay before the token can be returned.

3) if steam < some-large-number and token present and fuel present then insert one fuel cell in each reactor.

4) If steam < 100 then either it is startup time or something went wrong. Move the token from 2 in this case. This will probably insert a few fuel cells (if available) but you probably need a few to get the reactor started and steam buffered. Increase the token return delay to keep the number low. If too many fuel cells are inserted they will be wasted but it should only happen once. If you are realy OCD set the delay to just below the burn time of a single cell (a timer might be better there than belts for this long).
1: Seperation was done to prevent multiple fuel inputs because one reactor was faster and had send a signal. Alo you manually put the fuel in to start the power plant. I dont know how you get to this described situation without a fuel shortages. And there is a pesking alarm to remind you to act fast.
But you're maybe right. This will need a second combinator isolating the other one from the circuit though. On the other hand it could result in blackouts instead of brownouts. Need more opinions on that.

2: This is not necessary and would result in me simply using my old circuit with 12 combinators again. Because you cant control robots themselves you have to control the requester chest.

3: This will get covered with the combinator from 1, the disadvantage too.

4: I dont see what that gives in fail proovability. If the steam gets under a certain amount it will put in a fuel cell anyway. No need to have a second combinator telling the circuit the same thing again.
In my design point 1 means that all reactor are started together or none at all. There might be a slight difference because of inserter speeds but that should be at most one tick.

My point 2 then waits for enough empty full cells. If all reactors have only one fuel cell that waits for all of them. The removal of the spend cell doesn't trigger refuleing, only the removal of all spend cells. If multiple fuel cells got inserted and one reactor is slow (which should be at most one tick and not enough time for a reactor to spend 2 fuel cells anyway, but lets assume) then potentially this could trigger early and insert the next fuel cell before the slow reactor is finished. The token will be moved early but if there is steam then no fuel gets inserted till the steam is used up. The token will keep trigger the removal of additional spend fuel cells as long as it is pending. So any extras will get removed as long as there is enough steam. And if there isn't enough steam them you want the reactors running anyway. This is self correcting as it never inserts more fuel cells than it removes spend cells but it can remove extra spend cells if there are too many. Point 2 corrects any over fueling downwards.

My point 4 acts in the opposite way. Any under fueling, which should only be at startup, will be corrected upwards by adding extra fuel.

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

mrvn wrote:In my design point 1 means that all reactor are started together or none at all. There might be a slight difference because of inserter speeds but that should be at most one tick.

My point 2 then waits for enough empty full cells. If all reactors have only one fuel cell that waits for all of them. The removal of the spend cell doesn't trigger refuleing, only the removal of all spend cells. If multiple fuel cells got inserted and one reactor is slow (which should be at most one tick and not enough time for a reactor to spend 2 fuel cells anyway, but lets assume) then potentially this could trigger early and insert the next fuel cell before the slow reactor is finished. The token will be moved early but if there is steam then no fuel gets inserted till the steam is used up. The token will keep trigger the removal of additional spend fuel cells as long as it is pending. So any extras will get removed as long as there is enough steam. And if there isn't enough steam them you want the reactors running anyway. This is self correcting as it never inserts more fuel cells than it removes spend cells but it can remove extra spend cells if there are too many. Point 2 corrects any over fueling downwards.

My point 4 acts in the opposite way. Any under fueling, which should only be at startup, will be corrected upwards by adding extra fuel.
But it would cause blackouts when there is even only one fuel too few. Better have brownouts. Then you may have enough time fixing the fuel production while you're getting annoyed by the alarm. I said that before.
It does not matter if fuel is inserted in one reactor before the other. The bonus will increase when both have fuel. Also if you dont use all the power of the plant you'll always have a leveling factor in fuel insertion since they all wait for steam to empty first.
And underfueling at the start cant be corrected by inserting more and more into them. I mean how DO you unintentionally underfuel when there is enough fuel to add to the reactors? This doesnt make sense. But I may be misunderstanding this. :D
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by mrvn »

Distelzombie wrote:But it would cause blackouts when there is even only one fuel too few. Better have brownouts. Then you may have enough time fixing the fuel production while you're getting annoyed by the alarm. I said that before.
It does not matter if fuel is inserted in one reactor before the other. The bonus will increase when both have fuel. Also if you dont use all the power of the plant you'll always have a leveling factor in fuel insertion since they all wait for steam to empty first.
And underfueling at the start cant be corrected by inserting more and more into them. I mean how DO you unintentionally underfuel when there is enough fuel to add to the reactors? This doesnt make sense. But I may be misunderstanding this. :D
The problem is that if you don't fuel all the reactors then you only get a fraction of heat. With a 8 reactor setup you get 28 times the base heat. If you even miss one reactor that drops by 5-7. Miss two and it drops by as much as 12. That is nearly half the output. I think it makes a lot of sense to synchronize heating unless the steam gets real low.

As for underfueling at the start: The problem is the reactors aren't fueled at all. So nothing will trigger fueling them.

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

mrvn wrote:The problem is that if you don't fuel all the reactors then you only get a fraction of heat. With a 8 reactor setup you get 28 times the base heat. If you even miss one reactor that drops by 5-7. Miss two and it drops by as much as 12. That is nearly half the output. I think it makes a lot of sense to synchronize heating unless the steam gets real low.

As for underfueling at the start: The problem is the reactors aren't fueled at all. So nothing will trigger fueling them.
4: Ah now you understand what you mean with point 4: an Autostart! But I dont like it inserting 6 Fuel in all reactor before it stops: The first fuel is just heating up the reactors, the second can maybe produce the steam to stop the circuit, but until then there has been pressed 6 fuel into all of them. This will also mess with the rest of the circuit which reacts on the extraction of spend fuel cells since there are always at least 4 fuel in the reactors. Better just put one fuel in manually.

1: I think its better to have at least some power in the network instead going full blackout. Have fun trying to get fuel cells made then.
Also most people will have a backup power plant that uses coal - it is probably not able to fulfill all the need. Then the partially fueled reactor comes to rescue.
And dont forget the alarm does pesker you early enough to fill the reactor again.

If I put all inserters together I need more circuitry to handle a lagging behind reactor - for whatever reason he might do that. If I read the chests I have to control the robots so that they dont accidentially are faster than the circuit - or I need a memory cell that I delete when 8 spend fuel cells have been taken out while I, in both cases, also need a pulser to send a one tick signal to the inserters that insert fuel.
If I want to include an autostart that inserts only one fuel cell and starts the circuit, then I need three combinators. One constant that you have to set to on and two decider that act as a pulser so that the inserters dont insert more than one.

Hm. I thought about that. Having awkward manual start is not very good. Having too many combinators (+3) just because of this isnt too. It just is the best compromise as it is now. I had 12 combinators before and the circuit did exactly what you describe. But it just was an unnecessary amount.
I cant please all of you. It is easier to add combinators instead of removing them.
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

Taipion
Long Handed Inserter
Long Handed Inserter
Posts: 81
Joined: Tue May 02, 2017 6:58 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Taipion »

Thanks for mentioning me. :-)

I was unhappy with how the reactor is started, even though you ever only need to do this once, you may still need to do it several times if you play with designs, so I thought of this:

In my setup, I have each inserter for inserting fuel only connected to the respective inserter that is taking out used fuel.

Instead, you could:
- connect those inputting inserters also to the main circuit network of the reactor setup
- set them to activate if the "pulse hand content" signal (used fuel) from the outputting-used-fuel-inserters is bigger than zero. (if used fule > 0, then activate).
(- this ofc requires the taking-out-used-fuel-inserters to be set to "read hand content" and "pulse", but I guess that's the case already)
(!)(!)(!) For safety reasons, I HIGHLY recommend using yellow inserters for outputting used fuel, and fast(er) inserters for inputting fuel,
(!)(!)(!) For this case though, you may want to use yellow inserters for both, BUT if you plan to have more than 1 fuel in a reactor at a time, like, fueling 2 or 3 to each when they activate, so they burn for 2-3 cycles at a time, do NOT use same speed inserters for both, but slow for taking out and fast for putting in

- use a setup like the following (there may be simpler things, also things with just chests/inserters and no combinators to generate the signal, but this looks more sophisticated :D)

I'll add a string for copy/paste on the bottom.

It works like this: whenever you switch the constant combinator on, it pulses one "used fuel", but only then, not when you turn it off, and only once.
So this little contraption (people here like to say "contraption" I heard^^) will fill in one fuel whenever you feel like it, so you can do the inital fuel-loading with one click,
and there is no timing needed, and nothing that can go wrong, just one click, you though should turn it off again within due time as the number otherwise increases till it flips over again. ^^

If you, however, use this after there has been the first fuel inserted, you will have to remove the 2nd fuel IF you wanted just one in.
On the other hand, this could be used to add one more fuel (and more if you want) to your setup, so that whenever the reactor activates it gets 2 fuel, or however much you inputted, in case you want that for whatever reason.

So here is how it goes:
(ofc, use whatever unused letters you have instead of signal "A" and "B")

- add one constant combinator that is getting input and output from the circuit network, that outputs "1 used fuel" when signal "A" is exactly = 1
- add a second combinator that says "when signal "B" > 0", then output "A" and select "input count"
- add one constant combinator, giving it's output to the network, that will output "A" = 1, and "B" = 1, when activated

Now, if you activate the constant combinator, it will output "A" = 1, which will trigger the combinator that then outputs "used fuel" = 1, but only for one tick, as the other combinator will add +1 to "A" every single tick.
If you turn off the constant combinator, all goes to sleep again and nothing happens, and no "used fuel" signal is outputted again when turning it off.

Sidenotes:
- the second combinator will flash and the number will increase until you turn it off again
- if you set the second combinator to ouput "A" = 1 and not "A" = input count, it will send a "used fuel" = 1 signal both when you turn the constant combinator on and off, and nothing will flash, and no number keeps running away, but I kind of prefer the "A" = input count solution, your choice
String
[edit] And only now I read that you figured out a manual start yourself already. :D

[edit2:] I am actually working on a nice little reactor setup for the last version myself, I just can't seem to find the time to get it done. If I ever do, I'll add it here though.

[edit3:] Having a "initial reactor start" blueprint that you add, use once, and then remove, is actually a nice option, too.

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

Ok, when even you say it is too awkward this way...
I'll take a look at it. Problem with connecting all together is they always have to fuel at once or they fuel twice if they react on used cells.
Well I think this time Im going for TOTAL FAILPROVE!! AAAAH
Forget the number of combinators. Im going to imagine any possible failure and have a way to prevent it. Maybe even a box of backup-fuel! Who knows what I'll come up with! MUAHAHAHA!! :D :D
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

AcolyteOfRocket
Fast Inserter
Fast Inserter
Posts: 124
Joined: Sun Mar 06, 2016 9:58 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by AcolyteOfRocket »

My nuclear plant needs a re-vamp, so I came online and found this thread - Yay !

I have some rather general (noob-ish) questions.

1. Your 8-reactor block has 96 heat exchangers heated by 8 reactors, equivalent to 1 reactor per 12 exchangers. Isn't this equivalent to the reactors having 200% adjacency bonus ? Don't 4 of the reactors have a 300% bonus ? If so, doesn't your design sacrifice the extra bonus, and if so why ? I realise that the 12 exchangers is the limit of supply for 1 offshore pump.

2. Why the control circuitry for fuel feeding ? Is it to use less fuel when the plant is under less-than-full load ?

3. Why the steam storage at all ? Is it to allow the plant to be supplemented by solar during the day ? You don't seem to have the extra turbines that idea would suggest, so I must be guessing wrong.

BTW, I like the 10x10 minus 4 = 96 turbine arrangement - pure genius :D

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

AcolyteOfRocket wrote:My nuclear plant needs a re-vamp, so I came online and found this thread - Yay !

I have some rather general (noob-ish) questions.

1. Your 8-reactor block has 96 heat exchangers heated by 8 reactors, equivalent to 1 reactor per 12 exchangers. Isn't this equivalent to the reactors having 200% adjacency bonus ? Don't 4 of the reactors have a 300% bonus ? If so, doesn't your design sacrifice the extra bonus, and if so why ? I realise that the 12 exchangers is the limit of supply for 1 offshore pump.

2. Why the control circuitry for fuel feeding ? Is it to use less fuel when the plant is under less-than-full load ?

3. Why the steam storage at all ? Is it to allow the plant to be supplemented by solar during the day ? You don't seem to have the extra turbines that idea would suggest, so I must be guessing wrong.

BTW, I like the 10x10 minus 4 = 96 turbine arrangement - pure genius :D
Thank you :D

1. I followed the calculations from Dimanper: viewtopic.php?f=5&t=44778&p=269934&hili ... io#p269934
I just recalculated it and came up with a good visualization. Im going to use a layered approach where bonuses are just stacked reactors on top of the others:
You have 8 basic reactors without bonus. (We'll come to it)
xx
xx
xx
xx

Every reactor is adjacent to another reactor. So we can draw a vertical line...
x|x
x|x
x|x
x|x

... now you add the bonuses of each: All of them get a 100% bonus. So you have another 8 reactors on top of them = 16.
And now lets do horizontal lines:
(xx)
-----
(xx)
-----
(xx)
-----
(xx)

Lets look at the first line: It is surrounded by 4 reactors that also all have an adjacent one:
(xx)
-----
(xx)

So you add 100% to every reactor again, thats another 4. So now 8+16=20 reactors (Remember: there are still some steps ahead and some done before)
We do that three times since there are three sets of four reactors seperated by a line. Here highlighted by the colors red, blue and green:
(xx) ------------- (xx)
-----
(xx) and this set (xx)
------------------- -----
(xx) ------------- (xx)
-----
(xx) ------------- (xx)
16 + 4 + 4 + 4 = 28
And since we know we need 4 heat exchanger per reactor we simply multiply it: 4 * 28 = 112

2. Yes it is because reactors use fuel even if the heat is not used. Thats also why...
3. ... there is steam storage. Actually there is so much storage that the heat from 28 fuel cells can be stored. (The reactor bonus)

PS: I still havent been working on the combinators. It was too late yesterday. Im gonna do it today.
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

Ok, Taipion and the guys, what about this:
Example
I got a little all out on this one. 20 combinators. I would love to add 8 more to be able to insert fuel into a certain reactor alone after one had run out. But thats not REALLY necessary.
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by mrvn »

Distelzombie wrote:
mrvn wrote:The problem is that if you don't fuel all the reactors then you only get a fraction of heat. With a 8 reactor setup you get 28 times the base heat. If you even miss one reactor that drops by 5-7. Miss two and it drops by as much as 12. That is nearly half the output. I think it makes a lot of sense to synchronize heating unless the steam gets real low.

As for underfueling at the start: The problem is the reactors aren't fueled at all. So nothing will trigger fueling them.
4: Ah now you understand what you mean with point 4: an Autostart! But I dont like it inserting 6 Fuel in all reactor before it stops: The first fuel is just heating up the reactors, the second can maybe produce the steam to stop the circuit, but until then there has been pressed 6 fuel into all of them. This will also mess with the rest of the circuit which reacts on the extraction of spend fuel cells since there are always at least 4 fuel in the reactors. Better just put one fuel in manually.
Yes, an autostart. Note that the autostart should move the refuel token and not refuel in itself. So a few things factor into this:
A) at the start fuel is just coming out of the centrifuge so you might not have 6 fuel per reactor to drop in. A new fuel would be dropped in every time a set of fuel for all reactors is ready.
B) After dropping in fuel the refuel token has to be returned. I mentioned adding in some belts to act as a delay. You can use a counter instead that counts down 12000 ticks and then moves the token.
C) This won't mess with the rest of the circuit in my design since the token can only be moved to the refuel position once. So as soon as the presence of steam stops refueling any extra spend cells will be removed.
Distelzombie wrote: 1: I think its better to have at least some power in the network instead going full blackout. Have fun trying to get fuel cells made then.
Also most people will have a backup power plant that uses coal - it is probably not able to fulfill all the need. Then the partially fueled reactor comes to rescue.
And dont forget the alarm does pesker you early enough to fill the reactor again.
My experience with brown outs was that they are self promoting. You don't have enough coal to fuel all the boilers so you get less power. This makes the coal miners go slower. So even less coal and less power and you are in a death spiral to total blackout with 0 coal left to restart the boilers. Same with nuclear fuel. Your centrifuges will slow down making things worse.

So if you are afraid of this then you better put the fuel producing machinery on solar power separate from everything else. A few solar cells will run some minors and the centrifuge so it's no big deal.

Also given the relative speed of centrifuges and reactors I really don't see not having enough fuel as a problem except for 2 cases: A) Startup. B) No more fuel is coming at all. All mined out or some biter broke the centrifuge or something. Personally I would rather have e.g. 4 of 6 fuel left to restart once I fix the fuel problem than run a few extra second on half burn.
Distelzombie wrote: If I put all inserters together I need more circuitry to handle a lagging behind reactor - for whatever reason he might do that. If I read the chests I have to control the robots so that they dont accidentially are faster than the circuit - or I need a memory cell that I delete when 8 spend fuel cells have been taken out while I, in both cases, also need a pulser to send a one tick signal to the inserters that insert fuel.
If I want to include an autostart that inserts only one fuel cell and starts the circuit, then I need three combinators. One constant that you have to set to on and two decider that act as a pulser so that the inserters dont insert more than one.

Hm. I thought about that. Having awkward manual start is not very good. Having too many combinators (+3) just because of this isnt too. It just is the best compromise as it is now. I had 12 combinators before and the circuit did exactly what you describe. But it just was an unnecessary amount.
I cant please all of you. It is easier to add combinators instead of removing them.
Combinators have their use. But so do chests, belts and items. You earlier mentioned a memory cell. If you build that out of a combinator then it is suspectible to brownout. Circuits like that also have issues with signal timing that are hard to debug. Why not use a chest and a single stone as memory cell? Put the stone in the chest and it is set. One inserter to remove it wired to the clear signal and one inserter to reinsert it wired to the set signal. A belt or more inserters to move the item from one to the other. They are much slower but that makes it easy to see what it does.

And you mentioned timing issues with your robots. Again chests come to mind. Put 4 chests in a square with 4 inserters. Add one stone. Now think of them as a state machine with 4 states. You can wire up the inserters to trigger moving the stone and each chest having the stone represents one state. First chest for burning fuel, second chest means spend cell was removed, third chest means steam is low, fourth chest fuel is ready, and back to first chest when you insert fuel. And you remove the spend cells into a buffer chest. If chest >= <num reactors> you trigger the first->second inserter. And if first-chest = 0 you move spend fuel cells from the buffer chest to the provider chest. So far you have only used wires to control the inserters. No combinator at all.

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

mrvn wrote: Yes, an autostart. Note that the autostart should move the refuel token and not refuel in itself. So a few things factor into this:
A) at the start fuel is just coming out of the centrifuge so you might not have 6 fuel per reactor to drop in. A new fuel would be dropped in every time a set of fuel for all reactors is ready.
B) After dropping in fuel the refuel token has to be returned. I mentioned adding in some belts to act as a delay. You can use a counter instead that counts down 12000 ticks and then moves the token.
C) This won't mess with the rest of the circuit in my design since the token can only be moved to the refuel position once. So as soon as the presence of steam stops refueling any extra spend cells will be removed.
Distelzombie wrote: blackouts
My experience with brown outs was that they are self promoting. You don't have enough coal to fuel all the boilers so you get less power. This makes the coal miners go slower. So even less coal and less power and you are in a death spiral to total blackout with 0 coal left to restart the boilers. Same with nuclear fuel. Your centrifuges will slow down making things worse.

So if you are afraid of this then you better put the fuel producing machinery on solar power separate from everything else. A few solar cells will run some minors and the centrifuge so it's no big deal.

Also given the relative speed of centrifuges and reactors I really don't see not having enough fuel as a problem except for 2 cases: A) Startup. B) No more fuel is coming at all. All mined out or some biter broke the centrifuge or something. Personally I would rather have e.g. 4 of 6 fuel left to restart once I fix the fuel problem than run a few extra second on half burn.
Distelzombie wrote: stuff
Combinators have their use. But so do chests, belts and items. You earlier mentioned a memory cell. If you build that out of a combinator then it is suspectible to brownout. Circuits like that also have issues with signal timing that are hard to debug. Why not use a chest and a single stone as memory cell? Put the stone in the chest and it is set. One inserter to remove it wired to the clear signal and one inserter to reinsert it wired to the set signal. A belt or more inserters to move the item from one to the other. They are much slower but that makes it easy to see what it does.

And you mentioned timing issues with your robots. Again chests come to mind. Put 4 chests in a square with 4 inserters. Add one stone. Now think of them as a state machine with 4 states. You can wire up the inserters to trigger moving the stone and each chest having the stone represents one state. First chest for burning fuel, second chest means spend cell was removed, third chest means steam is low, fourth chest fuel is ready, and back to first chest when you insert fuel. And you remove the spend cells into a buffer chest. If chest >= <num reactors> you trigger the first->second inserter. And if first-chest = 0 you move spend fuel cells from the buffer chest to the provider chest. So far you have only used wires to control the inserters. No combinator at all.
- This chest thing is interesting. Now I know what you mean when you say "token" and "moving token". But it is cleaner done using combinators and since combinators are all in the reactor control grid they're powered by solar power and therefore safe.
- Im never adding belts as a delay. This is too crude

Anyway I solved these issues while also having every reactor inserter in a seperate grid. Even if they go asynchronous for some reason they dont interfere with the logic and you can start the reactor with one flick of a combinator. Also every reactor has a seperate alarm if fuel is low. So even if the overall fuel is too big to trigger the old alarm one reactor could have starved already. Its not prevented, kind of.
Take a look at the blueprint in my postabove. in the next post
Last edited by Distelzombie on Mon May 22, 2017 12:50 pm, edited 1 time in total.
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

AcolyteOfRocket wrote:1. Your 8-reactor block has 96 heat exchangers ...
I just re-read your post and you're mistaken. My reactor has 112 heat exchangers. ;)
Still take a look at my other post after yours, it took some time.


EDIT:
Realized I can save 7 combinators. Here's the new example blueprint of the logic:
New example
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

Shokubai
Filter Inserter
Filter Inserter
Posts: 470
Joined: Mon May 02, 2016 3:17 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Shokubai »

Really appreciate this setup and the tile-ability of "as-you-grow" design.

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by mrvn »

Here is my take on reactor control. I build it with just 2 reactors because that is enough to test. The top reactor is the master and the bottom a slave. Any other reactors would be setup as additional slaves. The difference is only the buffer chest for fuel cells and when it gets filled. Make sure new fuel cells pass all slaves before they pass the master.

The reactor produces up to 162MW, as is expected for 2 reactors I think. Reactors are fueled with one fuel cell at a time when steam is below 50k. The design uses an iron plate as token and the lack of an iron plate in the control chests at the top causes one to be added and the initial fuel cells to be inserted. So this should be fully blueprintable and start on its own. My design uses one combinator per reactor to keep the fuel chests balanced. Everything else is just wires.

Ignore the 61 combinators on the top right. They just show the change in buffered steam per second and are purely informational.

I've also included my fuel cell production at the left. It should produce U235 and U238 keeping an equal number in store.
Note: In the beginning I didn't limit U235 production so the savegame has about 1k head start. Production is blocked till it equals out. 7 combinators needed for that one. I should add one more to make it keep U235 and U238 in the 1:19 ratio needed for fuel cells. Just in case fuel cell production backs up.
overview
overview
nuclear-overview.png (13.24 MiB) Viewed 8713 times
Closer look
Circuit help
Fluid statistics
Attachments
sndbox-nuclear2.zip
Savegame
(4.44 MiB) Downloaded 145 times

AcolyteOfRocket
Fast Inserter
Fast Inserter
Posts: 124
Joined: Sun Mar 06, 2016 9:58 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by AcolyteOfRocket »

The nuclear plant I am trying to design (badly) is on a base and every so often it has a few hundred prod module3's and I take time off designing power plants to set up some more mines to try and get my green circuit production up.

I was wondering how people seemed to test their design, I guessed there was some creative mode that folk were using.

Then I saw mrvn's large bank of radars - nice :lol:

Beeblebrox
Burner Inserter
Burner Inserter
Posts: 19
Joined: Tue May 24, 2016 9:57 pm
Contact:

Combinator-free design?

Post by Beeblebrox »

Seems to me you don't need the constant combinator, because you can just use the steam signal directly (as in: steam < 1,000,000).
For that matter you don't need the decider combinator, since you can enable the inserter directly using the steam signal as well.

I tested this with your blueprint (putting the inserters that unload the reactors on the green measuring circuit and removing the combinators). I replaced green with red for the network that connects loading and unloading inserters but I am not sure this is necessary. Seems to work, but maybe I am missing some nuance?

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: Combinator-free design?

Post by Distelzombie »

Beeblebrox wrote:Seems to me you don't need the constant combinator, because you can just use the steam signal directly (as in: steam < 1,000,000).
For that matter you don't need the decider combinator, since you can enable the inserter directly using the steam signal as well.

I tested this with your blueprint (putting the inserters that unload the reactors on the green measuring circuit and removing the combinators). I replaced green with red for the network that connects loading and unloading inserters but I am not sure this is necessary. Seems to work, but maybe I am missing some nuance?
Which blueprint? The one on this page, 5?
I know the constant is not needed. Before 0.15.11 I had it outside of the plant in a contral area thingy. It is more understandable when you set a value in a constant combinator instead of in a decider. Anyway, if I dont have a control area I shouldnt have such a constant combinator...
And you're right you can get rid of the decider. Have fun changing the steam value then. :)

In both blueprints, if you change any wires it leads to problems:
OP: every fuel insterter was only connected to its neighbor that rips the old fuel cell out. Connecting all ripper together can lead to multiple inserts, if for some reason one reactor was slower.
This page: Messes everything up. Dont do.

I think you mean the BP in OP though. Since then I build another combinator setup. This requires the signal from that combinator. Could you test this? I wanna know how that is percieved.
Distelzombie wrote:Here's the new example blueprint of the logic:
New example
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

AndrewIRL
Fast Inserter
Fast Inserter
Posts: 240
Joined: Fri Mar 24, 2017 2:17 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by AndrewIRL »

My single continuous feed reactor was struggling with demand so I dumped it and switched to your blueprint 2 reactor design. Turned off my coal fired boilers, power demand has roughly doubled but nuclear fuel consumption is maybe a third of what I was using with a single reactor and no steam storage. Yours is dramatically more efficient Appears to work as advertised. Nice. I have a good ore field and I'm not sure I need Kovarex with this low rate of consumption.

AndrewIRL
Fast Inserter
Fast Inserter
Posts: 240
Joined: Fri Mar 24, 2017 2:17 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by AndrewIRL »

This reactor was covered in a Xterminator video. Unfortunately they didn't do it justice, failing to get it started, not mentioning the critical features and made disparaging comments about steam storage and smart builds. From the looks of them yours was the only reactor design series which smoothly expands to 2,4,6,8 reactors and the only one that has load following. They did mention the alarm at least.

https://www.youtube.com/watch?v=Jv_0XdHUcso

If I had only watched the video I would have ranked your reactor near the bottom based on their comments. But actually having used it I know that it works perfectly and looks to be the best of what was shown.
Last edited by AndrewIRL on Tue May 30, 2017 10:46 am, edited 1 time in total.

User avatar
Distelzombie
Filter Inserter
Filter Inserter
Posts: 336
Joined: Tue May 02, 2017 4:27 pm
Contact:

Re: The perfect OCD compliant power plant! Now 0.15.11 safe

Post by Distelzombie »

AndrewIRL wrote:This reactor was covered in a Xterminator video. Unfortunately they didn't do it justice, failing to get it started, not mentioning the critical features and made disparaging comments about steam storage and smart builds. From the looks of them yours was the only reactor design series which smoothly expands to 2,4,6,8 reactors and the only one that has load following. They did mention the alarm at least.

https://www.youtube.com/watch?v=Jv_0XdHUcso

If I had only watched the video I would have ranked you reactor near the bottom based on their comments. But actually having used it I know that it works perfectly and looks to be the best of what was shown.
Yea, I know. I think I send him the link to this thread with the blueprint, but I am not sure.
I've send him three versions actually. Two from before the 0.15.11 Patch that had major differences and then this new one after I finished it. May have been 2.0 or 2.1, but I didnt want to bother him with yet another post. xD (Ah yea, thats why I included the link.)
If I had bothered him again he would've used a newer one. I know all build worked though. Probably didnt wrote the instructions right or he forgot it later on. It was the build which needed a seperate logistic robot network in order to work.

Im not mad or anything. :)
And thank you for the appreciation. :)


Also I JUST made a huge change!
I included the logic I had posted before because I think it is great and give way more control. It is now also in every module-stage-thing.
Complete 2-Lane system as a Blueprint-Book! The perfect OCD reactor? Testing chained science lab efficiency Please use real prefixes and proper rounding!

Post Reply

Return to “Energy Production”