All the different ways to increase the efficiency of a nuclear power plant

[foamy]

Also modded games can make correcting the supply chain a very long thing ! I know buffer are not popular but i like them.

I don't mind buffers in continuous-flow systems where everything that goes into a buffer will eventually be consumed. Some of it's necessary (and unavoidable) in simply processing materials. And in some kinds of production -- gears for belts comes to mind -- actual chest stockpiles are useful because demand can be very spikey and the assemblers can easily outrun the capacity of a belt to feed them. Power supply and uranium processing are other good ones.

But the characteristics of the ideal buffer system are:

1. No dead zone. Any material that goes into the buffer can be extracted without requiring additional material to be added.
2. The things consuming that material have priority over filling the buffer further (there's no point in filling a buffer if something that could use those materials is being starved).
3. The buffer is sized such that filling it doesn't take forever. Train stops with unlimited chests, for example, can consume extraordinary quantities of material if you're not careful.

Requiring the entire belt to be backed up before starting a nuclear reactor breaks principles 1 and 2, and potentially 3 depending on how your U-235 production is doing. It works okay if you start from a full belt to begin with, but it isn't really properly a buffer, because if your fuel production gets interrupted all that fuel on the belts is useless because the circuits will see you don't have a full belt and thus not enable the inserters.



(Note: I have encountered people whose 'solution' to any given production issue is 'eh, stick a chest in there somewhere', including the sciences. Baffles me. Science is the one thing you have absolutely no need to buffer whatsoever (excepting space, I suppose), because it's produced continuously and consumed at exactly known rates!)

[mrvn]

(Note: I have encountered people whose 'solution' to any given production issue is 'eh, stick a chest in there somewhere', including the sciences. Baffles me. Science is the one thing you have absolutely no need to buffer whatsoever (excepting space, I suppose), because it's produced continuously and consumed at exactly known rates!)

But the rates are not constant. Nor do you always need all science packs. One tech takes 10s per red + green pack. The next takes 120s but uses 2 red + 1 green. So buffering some science packs when consumption is slow for the next tech where consumption exceeds production is a good thing.

I also like to select a tech and have it finish quickly using 2, 3 or 4 times the number of labs I can support uninterrupted. Then while I build stuff for whatever I just researched I collect more science packs for the next research, whatever that might be. Means I can delay picking what to research till I see what's best next.


So in conclusion: buffer chests for science packs can be a good thing. Unless you are low on raw materials. Then you want the science to stop using resources when you pause research. But you can do that with some wires at the input of the factory instead of pausing only the science labs.

[foamy]

(Note: I have encountered people whose 'solution' to any given production issue is 'eh, stick a chest in there somewhere', including the sciences. Baffles me. Science is the one thing you have absolutely no need to buffer whatsoever (excepting space, I suppose), because it's produced continuously and consumed at exactly known rates!)

But the rates are not constant. Nor do you always need all science packs. One tech takes 10s per red + green pack. The next takes 120s but uses 2 red + 1 green. So buffering some science packs when consumption is slow for the next tech where consumption exceeds production is a good thing.

I think this question's getting a bit afield :v

[mmmPI]

But the characteristics of the ideal buffer system are:

1. No dead zone. Any material that goes into the buffer can be extracted without requiring additional material to be added.
2. The things consuming that material have priority over filling the buffer further (there's no point in filling a buffer if something that could use those materials is being starved).
3. The buffer is sized such that filling it doesn't take forever. Train stops with unlimited chests, for example, can consume extraordinary quantities of material if you're not careful.

Requiring the entire belt to be backed up before starting a nuclear reactor breaks principles 1 and 2, and potentially 3 depending on how your U-235 production is doing. It works okay if you start from a full belt to begin with, but it isn't really properly a buffer, because if your fuel production gets interrupted all that fuel on the belts is useless because the circuits will see you don't have a full belt and thus not enable the inserters.

maybe i should have used another word, like reserve or stockpile ? for the case of the fuel cell. Because it is not going anywhere else than to the nuke plant. but no need to add anything, it's just some fuel in advance, like don't open a store if you don't have at least a week of things to sell

for the case of buffering science, it's a bit off topic but i'm ok with doing it, especially military science, i generally under-build it and don't chain research that require it. I would put a chest or use the lengh of the belt, with more labs than i can feed continuously, but when i want some particular important tech, i can go 3 4 research in a row an then let it build up/stockpile/buffer (??) while i'm constructing the new building that just unlocked.

please note that 10 uranium fuel cell = 1 uranium235, and i was only using half a belt, only 1 side. It was good enough for 3) in my mind

Also i didn't used fuel check circuitry, so in case the belt wasn't full , i was running the risk of an arbitrary number of reactor not getting fed and being disconnected until manual intervention. It's worse as time goes, but if the belt is 85% full instead of 100% which is already quite some time running in deficit, you have 15% bad luck, the inserter doesn't grab anything. It's statistically very unplausible that you get a full shut down by surprise. It didn't happened to me because the belt was always full anyway:)

The fuel was unloaded to the power plant coming from several places via train, the chest for unloading was where my alarm was connected to monitor if i had the "buffer/reserve", then this chest unloaded on half a belt to feed the power plant and collect the used cell on the other half.

The station itself was disabled when the fuel level was good enough which backed-up far away and in several places the consumption of U235 for fuel.

All of it making it even worse for the amount of material in the "dead zone", in each train there was a little load of cell,plus in each belt from the fuel assembly to each train haha; In that regard the amount of fuel on the belt around the reactors was neglected.

I did that only 1 time, i wanted/needed that thing which felt like a pipeline where you can't just send 1L or 1m3, without first filling it up. but once it's filled up, it act as a bad giant buffer in case of emergency, which still felt better than nothing at all

[Khagan]

Requiring the entire belt to be backed up before starting a nuclear reactor breaks principles 1 and 2, and potentially 3 depending on how your U-235 production is doing. It works okay if you start from a full belt to begin with, but it isn't really properly a buffer, because if your fuel production gets interrupted all that fuel on the belts is useless because the circuits will see you don't have a full belt and thus not enable the inserters.

Yes, it hurts to see all that fuel sitting on the belt doing nothing. But I think it really is the simplest, cheapest and most robust way to handle the supply. Fuel cells are cheap, and if you think of the initial belt (or rather lane) full of them as part of the construction of the power station, it's a negligible part of the total cost.

[Khagan]

Note: The fuel check in my designs, or any design that uses the removal of the spend fuel cell as trigger, is necessary. If you ever remove a spend fuel cell without a fresh fuel cell to put into the reactor the setup dies. That's really the only argument I need.

I do use the removal of the spend fuel cell as trigger; I do not make a fuel check. I do start my reactors up manually (actually, I find that quite a satisfying thing to do, like a sort of ceremonial ribbon cutting), but in a cumulative total of hundreds of hours of running time across various sizes and generations of design, I have never had one 'die' and need a manual restart.

[mrvn]

Note: The fuel check in my designs, or any design that uses the removal of the spend fuel cell as trigger, is necessary. If you ever remove a spend fuel cell without a fresh fuel cell to put into the reactor the setup dies. That's really the only argument I need.

I do use the removal of the spend fuel cell as trigger; I do not make a fuel check. I do start my reactors up manually (actually, I find that quite a satisfying thing to do, like a sort of ceremonial ribbon cutting), but in a cumulative total of hundreds of hours of running time across various sizes and generations of design, I have never had one 'die' and need a manual restart.

Then you never had a fuel shortage.

[Khagan]

I do use the removal of the spent fuel cell as trigger; I do not make a fuel check. I do start my reactors up manually (actually, I find that quite a satisfying thing to do, like a sort of ceremonial ribbon cutting), but in a cumulative total of hundreds of hours of running time across various sizes and generations of design, I have never had one 'die' and need a manual restart.

Then you never had a fuel shortage.

On the contrary, I have had fuel shortages on several occasions. What I have never had is a fuel shortage that went unnoticed and unattended to for hours, completely emptying the fuel supply belt. Unless that happens, no restart is required.

[mmmPI]

Then you never had a fuel shortage.

On the contrary, I have had fuel shortages on several occasions. What I have never had is a fuel shortage that went unnoticed and unattended to for hours, completely emptying the fuel supply belt. Unless that happens, no restart is required.

That's just wording, supply disruption or total shortages. In one case your system can handle, the other case would be joining a random multiplayer server called "i need help fixing my nuke plant", you join in, and you can immediatly guess what happens, someone copied a blueprint he didn't understood, it got stuck and broke .

Could also be you join-back a server and people tell you your nuke plant doesn't work, you look at it and you realize they broke it by manually putting a leftover cell from inventory in a reactor.

"how it react when broken?" , not " how easy it is to break it ?".

Fuel cell lasting 200 sec is quite long, and it's not hard to make enough reserve for hours. so it's unlikely that it would break on normal operation. but if for whatever weird reason it does break, then it's a bit more annoying to fix. (using used fuel ejection as trigger for refuel i mean )

Then, if you build a 100 core central on a very populous server full of newcomers, the likelyhood of weird thing happening increases a lot, and the annoyance of fixing those 100 core too, nothing to compare to a 2x4 on a solo map on the other extreme, adaptation is important hehe

[mrvn]

Here is another failure scenario for reactors using logistic bots: Someone placed a roboport near the reactor so the logistic zones connected and now all the logistic bots are off halfway across the map delivering landfill, rails, belts or something.

Worse, someone then removes the roboport and the logistic bots never return because they are in the other zone now.

[Bauer]

A blue belt can ship 22.5 fuel cells / s in and out.
That'll feed a row of 22,5*200=4500 reactors
That's a disappointing 719.682 GW (taking 4500 pumps and inserters, and the 4 reactors at the end of the line into account).
But you can build 2 of these to avoid bots AND get more than a TW.

THE FACTORY MUST GROW.

[mrvn]

A blue belt can ship 22.5 fuel cells / s in and out.
That'll feed a row of 22,5*200=4500 reactors
That's a disappointing 719.638 GW (taking 4500 pumps and inserters, and the 4 reactors at the end of the line into account).
But you can build 2 of these to avoid bots AND get more than a TW.

THE FACTORY MUST GROW.

And that's why I use half a yellow belt.

Isn't the electricity for pumps and inserters a rounding error?

My modded game has pumps at 10kW. 45MW for pumps. Ok, that shows with 3 digit accuracy. But inserters are idle for 200s and then swing once. What's their energy consumption per second? Again modded game, inserters have 1kW drain and 18kW max. So something more than 13500kW for all inserters. More than I thought. Still, it's 0.004% of the power output. A rounding error.

.oO(How many roboports and bots would it take for 4500 reactors? How much power wasted is that?

[Bauer]

In vanilla, the pumps have 30 kW and inserters 13 kW/0.4 kW.
You're right we should take the 0.4 kW + a little bit, say
(30+0.5 kW) x 4500 = 158 MW + 160 MW for the corner reactors.
I corrected the figure above.

[mrvn]

In vanilla, the pumps have 30 kW and inserters 13 kW/0.4 kW.
You're right we should take the 0.4 kW + a little bit, say
(30+0.5 kW) x 4500 = 158 MW + 160 MW for the corner reactors.
I corrected the figure above.

Don't forget you have one inserter to fuel, one for spend fuel and usually one for the buffer chest. So (30kW+1.5kW)*4500 = 141_750kW. Not sure where your 158MW and 160MW come from, make no sense. For a reactor that size you basically need 1 reactor core itself just to power the pumps and inserters and not a corner one. Impressive.

Now you have me imagine a 3 level reactor: Level 1 a simple 2x2 reactor cores setup to provide 160MW power to power level 2 and 320MW for fuel production. Level 2 a 4500 reactor cores setup for the 720GW mentioned earlier to power level 3. Level 3 the actual power generator using 20 million reactor cores and 3240TW of power.

[quyxkh]

haha you can save on that rounding error with a burner and a longhand.

I make a burner inserter picking up three fuel cells off the near lane of a full blue belt a 92-tick cycle. 92/60*94200 is 144440J/swing, you need one every 600s is 240.7J/s. That's a lot less than even "just" the 400J/s drain on a yellow inserter, whose 15413J swing every 600s isn't much more than a rounding error on its energy usage. 240.7W for the burner, 425.68W for the yellow. But hey, multiply that by your 4500 reactors and you're talking over 800KW in savings!

Feed the burner with rocket fuel on the outer lane, and you need a longhand (or belt braiding but that's just insane to unload, but it's got the same drain and the 16758⅓J 50-tick unload swing every 200s for that compares to the 16320J 72-tick for a yellow inserter (one-item swing to an empty blue belt both), plus 80000J idle drain for the 200s, averaging out at 438J/200s more for the longhand. 2.19W. Nothing.

[mrvn]

Have you considered the cost of producing rocket fuel? Not sure that is still a gain if you include that.

[quyxkh]

The calc doesn't consider the energy that goes into making the fuel, you're right, and not considering that the burner setup saves about 185J/s. One rocket fuel is 100MJ, enough for 692 reloading swings (691 really, one to pick up the next fuel). That's 691 times three reactor cycles, 6910 minutes, 1382000 seconds, over which time using the burner has saved 255.67MJ. I feel pretty sure it costs less than 255MJ to make one rocket fuel, but no, I hadn't considered it beyond ballparking it (tttt my quick checks were even cruder than that).

edit: just checked with kirk's, 22 factories in a full-beaconed setup makes 492.8 rocketfuel/min and uses 156.238MW all told (not counting inserter usage I notice but still) so that winds up really not that great, about 19MJ to make each piece.

Ahh: doing it with just E1's gets us down to 2.145MW for 40/min, that's 128.7MJ/40, 3.2175MJ production cost / 100MJ rocket fuel. That's . . . that's attractive

[mmmPI]

I make a burner inserter picking up three fuel cells off the near lane of a full blue belt a 92-tick cycle. 92/60*94200 is 144440J/swing, you need one every 600s is 240.7J/s. That's a lot less than even "just" the 400J/s drain on a yellow inserter, whose 15413J swing every 600s isn't much more than a rounding error on its energy usage. 240.7W for the burner, 425.68W for the yellow. But hey, multiply that by your 4500 reactors and you're talking over 800KW in savings!

Feed the burner with rocket fuel on the outer lane, and you need a longhand (or belt braiding but that's just insane to unload, but it's got the same drain and the 16758⅓J 50-tick unload swing every 200s for that compares to the 16320J 72-tick for a yellow inserter (one-item swing to an empty blue belt both), plus 80000J idle drain for the 200s, averaging out at 438J/200s more for the longhand. 2.19W. Nothing.

Because of you I was playing around with trying to use 3 burner inserter, one for fuel injection one for fuel ejection, and another one to fuel the 2nd. If you put nuclear fuel, because that's what is done in nuke plant, you have 1210000kJ/70kJ=17285.7143 swing per nuclear fuel. Counting for fuel ejection and fuel injection every 200 second, that means 1 nuclear fuel would last 3457142.86 second in those 2, ( around 40 days ).

Which means the burner inserter that refuel the ejection inserter will do a swing every 40 days, which means it will need to confiscate one fuel to refuel itself every:

17 285.7143 x 3 457 142.86= 59 759 183 772.245 sec, 5.9E10 sec, that's 691 657 days, or 1894 years.


But that's only if you run the power plant 100% capacity all the time that you will have a refuel every 200 sec, because if you only need half capacity, and you have a system to reduce fuel consumption, instead the initial nuclear fuel placed manually in the burner inserter used to refuel the ejection inserter would last for almost 4 millenia.

Which in your case increases the potential gain, since burner inserter don't drain energy contrary to yellow inserter, the lower % of electric output you use the slower the refuel will be and the worse the drain of the yellow inserter !!

After some thought i decided to not include a system to refuel the burner inserter, considering 900 hours or so is more than i would spend on a game.

Fun fact: if you want to place 2 nuclear fuel on a burner inserter, one that get consumed a bit immediatly and create a red bar, and one for after, it works like a charm for the injection inserter and the refuel inserter that are picking from a belt, but for the ejection inserter you need to place the nuclear fuel and rotate it 360° so that it face the belt for an instant and then face the reactor again, it would then be "primed" and you can add another nuclear fuel, which would make it enough for 80 days worth of ejecting used fuel. or more if you run under full capacity.

[gGeorg]

Which means the burner inserter that refuel the ejection inserter will do a swing every 40 days, which means it will need to confiscate one fuel to refuel itself every:

17 285.7143 x 3 457 142.86= 59 759 183 772.245 sec, 5.9E10 sec, that's 691 657 days, or 1894 years.

make an independent power circuit for internal usage. Power it by one solar panel then backup by one battery. This lasts forever and can feed inserters and pumps for 4-8 block of cores.
It is way cheaper than fill half belt with fuel.

[foamy]

Which means the burner inserter that refuel the ejection inserter will do a swing every 40 days, which means it will need to confiscate one fuel to refuel itself every:

17 285.7143 x 3 457 142.86= 59 759 183 772.245 sec, 5.9E10 sec, that's 691 657 days, or 1894 years.

make an independent power circuit for internal usage. Power it by one solar panel then backup by one battery. This lasts forever and can feed inserters and pumps for 4-8 block of cores.
It is way cheaper than fill half belt with fuel.

An accumulator can drive 22.4 yellow inserters. At two inserters per reactor minimum, that's 11 reactors you can handle, but it also must be a pump free design. A solar panel provides an averaged power supply of 42kW and therefore can drive, at most, one pump continuously.

If you use a triple inserter setup for a holding chest, you can only put 7 reactors on that same accumulator -- again, assuming no pump. Each pump added to either system reduces the number of inserters that can operate at peak by slightly more than two, i.e. you lose another reactor.

If you use fast inserters, which is my own preference for other reasons, you can only operate 6.45 per accumulator, which is two or three cores tops.

In a 2xN layout those accumulator fields are somewhat awkward to manage, since if you put them internal to the reactor you stretch your material flows (water/steam/heat), which harms throughput, but if you put them outside of the reactor, you need to be very careful about not crossing power grids. My own preference would be to power 'em with a bootstrap plant of some sort, I think, since it'd be more compact.