Factorio Forums

mmmPI wrote: Sun Apr 05, 2026 11:44 pm Let's recap from the beginning ok ?

Gasp, you got me! I wasn't careful to distinguish between the line discussion going on in

mmmPI wrote: Sun Apr 05, 2026 8:21 am

and older comments in this thread that weren't part of that line, and left myself vulnerable to bad faith actors catching me in a technicality!

I am slain!

And out of morbid curiousity

mmmPI wrote: Sun Apr 05, 2026 11:44 pm

Hurkyl wrote: Tue Mar 31, 2026 3:08 am The situation where I've most commonly heard this use case is to add hysteresis to the on/off cycle of a backup power plant, to eliminate the extremely rapid thrashing of the system when you toggle a power switch based on accumulator levels. While SR latches get discussed in this context, of course, the discussions I've seen pretty much always have a few people say basically "there's a much easier solution: just toggle the offshore pump on and off, that's what I do".

I posted a demonstration setup where you toggle the power switch because that's what is comonly used nowadays, since what you read in other older disccusion about the offshore pump is obsolete.

Do you genuinely not realize that I was bringing this up as a supporting example of things people once did that were broken by the fluid overhaul? Or is this just more bad faith acting?

I believe this is out of topic and useless in the current discussion.

I understand there are no objections to using a power switch to control the steam power plant like in the posted setup.

mmmPI wrote: Sun Apr 05, 2026 8:21 am

macdjord wrote: Sun Apr 05, 2026 2:54 am 1)The shutdown is actually slower than cutting the steam off with a pump, since now instead of just having to fill up the heat exchanger outputs and the pipes they feed directly into,
2)I also have to fill up all turbines (each of which also stores 200 steam) and the pipes between them
3)Except that's ignoring the great honking big tank farm that's there in the middle; the whole point of this exercise is to not cool down my reactors trying to fill that thing up, remember?
4)And most importantly, if I turn off the power switch, my factory blacks out. This is what we in the business call a bad thing.

1) Not really , when the point 2) is true, then as soon as the power switch isolate the nuke plant, the steam production stops.

2) Yes that's the main point i thought, to keep everything into a single fluid box to make it simple to control the nuke plant like it was possible to control cutting water input before. That's what i show in the demo setup, how you can use a power switch instead of any sort of pumps, that's how i fixed it for my games,with the same logic as you would have attached to the water input. You need to also fill up some steam tanks anyway, so the buffer in turbines an pipes are negligible to me since it's all in the same fluidbox.

Let me explain this one more time: My nuclear power plant is the main power source for my entire base. If I cut off the turbines from the main power grid, my whole base instantly goes into a massive brownout, as 0.5GW of solar panels try to supply a 4GW base. Therefore, any solution that involves controlling steam generation using a power switch is completely effing useless to me.

mmmPI wrote: Sun Apr 05, 2026 8:21 am 3) i don't see any compelling reason to keep reactor temperature above an arbitrary threshold, i feel that's a "you" thing. To me the important part is that the electricity is reactive , that's why you have steam tank in the first place.

It's not about keeping the reactors above some arbitrary temperature. I don't care about the reactor temperature as long as it stays below 900º. It's about keeping the heat exchangers above 500º.

Let's assume I didn't shut off steam generation:

• We start with the reactors fueled and at a nice steady state. That means that the reactors are at ~700º, while the outermost heat exchanger is at 500º. (I'd need exactly 9.6 heat exchangers per column to consume all the heat produced by the reactors. I have 10, so the last one in each row only operates a little over half the time.)
• The reactor fuel runs out. The steam buffer is reasonably full, so we don't insert new one just yet. The reactors stop generating heat.
• The heat exchangers, however, continue generating steam. They need to consume heat to do this, so they start to cool down.
• As each heat exchanger hits 500º, it stops producing steam. For the last one in the row, already hovering on the border of being too cold to function, this happens almost immediately, but the ones closer in have more of a buffer.
• Additionally, the heat exchangers can draw on the remaining heat in the heat pipes, and through them, the reactors themselves.
• Eventually, all the heat exchangers drop to 500º and stop working. If the Factorio heat transfer simulation were perfect, then at this point all the heat pipes and reactors would also be at exactly 500º. However, instead Factorio requires a minimum temperature difference of 1º for any heat transfer to occur, so the heat pipes directly connected to each exchanger will be at ~501º, those between the exchangers at ~502º, and the ones between the first heat exchanger and the reactor will be steadily hotter, until the reactors themselves are ~515º.
• Now, let's say power demand increases. I just cued up a new research and all the idle bits of the factory are now running full tilt.
• The steam buffer quickly hits its target minimum, and the circuit controls dutifully refuel all the reactors.
• The reactors start producing heat again... but the heat exchangers can't produce any steam. They're all at 500º, and it takes time for the heat to get from the reactors to the exchangers.
• The rate at which heat flows between two object depends on the temperature differential between them. The first heat exchanger will start to produce some steam almost immediately, but it won't hit full production - 10MJ of heat into 989.7 units of steam per second - until the temperature gradient between it and the reactor gets steep enough for 10MJ of heat to transfer into it ever second. Worse, the second heat exchanger won't even start to heat up until then either, because the first exchanger will be consuming ~all the heat from the heat pipe before it can get any father down the row.
• As the reactors heat up, more and more heat exchangers come online, but each subsequent heat exchanger takes longer to do so, since more of the reactor's heat output is getting turned into steam instead of heating buildings up further. And meanwhile, the steam buffer is still dropping...

See the problem? Now, if I instead shut off the heat exchangers:

• We start in the same steady state as before.
• Again, the fuel runs out and the reactors stop producing heat. This time, though, the heat exchangers are shut down at the same time, and so stop consuming heat.
• Since the heat exchangers aren't consuming heat, they don't get colder, overall.
• The heat pipes do try to balance out heat, though. This means the reactors and the exchangers closest to them cool down, while the ones father away heat up.
• Eventually, it all reaches a new steady state where everything is ~600º.
• Now power demand increases, and the buffer hits it target minimum.
• The reactors get new fuel, and simultaneously the heat exchangers are turned on.
• Since all the heat exchangers are above 500º, they all start producing steam at full speed immediately. Indeed, they're now producing steam faster than they would normally, since the last exchanger in the row is running at 100% instead of only 60% capacity.
• This heat has to come from somewhere, and the heat exchangers start to cool down. But each heat exchanger and each heat pipe stores 1 MJ of energy for every degree of heat. With 3 heat pipes for every 2 heat exchanger and a starting temperature of 600º, that means that every heat exchanger can run for ~25s before hitting 500º and shutting down.
• Meanwhile, the reactors are all producing heat. No heat gradient means no heat transfer, so they get hotter even as the heat exchangers get cooler. By the time the heat exchangers start to run out of heat, the proper heat gradient has been established, so they keep running. (Except, of course, for the last one, which is only designed to run at 60%. It'll actually run at less than that for a while; running at 100% earlier created a deficit which needs to be made up.)

Thus, shutting off the heat exchangers allows the system to start up quickly on demand.

mmmPI wrote: Sun Apr 05, 2026 8:21 am 4) No, considering the scope of your suggestion to alter the buffer of steam engine and boilers and heat exchanger , i feel there are plenty cases when it's just fine to use a power switch, you mentionned one and i made a demo setup.

Yes, there are plenty of cases where a power switch will solve the problem. However, this is not one of them.

mmmPI wrote: Sun Apr 05, 2026 8:21 am

macdjord wrote: Sun Apr 05, 2026 2:54 am But that has nothing to do with my desire to control when the heat exchangers generate steam.

Yes, imo your desire would be better fulfilled by just asking to wire the heat exchanger and/or boiler instead of that indirect way of altering the entity buffer. Both aren't things i see negative consequences about, i would be in favor of both suggestion as mentionned. But if the reasonning is the one you have , i don't expect it to be implemented. I feel like you are saying : "i can't do that" ( and some people disagree), when really you mean "i can't do it this way" ( and some other people disagree), and you could be saying " i would like to be doing it this way" . Since there is no downside i can see , i have no objection and i won't be one the disagreeing person then.

I feel the positive arguments about how the gameplay can be enriched by adding wire on heat exchanger are stronger because it better suit what you want to achieve in practice in the way of controlling steam and temperature.

Yes, adding circuit control to heat exchangers would solve my use-case. That's why I linked the thread suggesting just that both in the OP and in my previous reply to you. However, historically, the reactions of the devs to 'Can you add $CAPABILITY to $BUILDING?' has often been 'We could, but the fact that $BUILDING can't do that is intentional game design', so I am also pursuing the angle of asking them to restore an element of game balance that was thrown out of whack by a mechanics change.

mmmPI wrote: Sun Apr 05, 2026 11:55 pm I believe this is out of topic and useless in the current discussion.

I understand there are no objections to using a power switch to control the steam power plant like in the posted setup.

Then you haven't been reading my posts, because I've been objecting to it ever since you first suggested it.

macdjord wrote: Mon Apr 06, 2026 8:19 am Let me explain this one more time: My nuclear power plant is the main power source for my entire base. If I cut off the turbines from the main power grid, my whole base instantly goes into a massive brownout, as 0.5GW of solar panels try to supply a 4GW base. Therefore, any solution that involves controlling steam generation using a power switch is completely effing useless to me.

Thank you for patience, i do understand/feel that you mention here a more specific case than the scope of the suggestion whereas when i mentionned the use of power switch it was more related to general use cases when the offshore pump doesn't provide the same ability for control, like for boilers / steam engine. It can also be adapted to nuclear for back up solar, you mentionned it first, and i just made a demo setup for it.

macdjord wrote: Mon Apr 06, 2026 8:19 am Yes, there are plenty of cases where a power switch will solve the problem. However, this is not one of them.

Indeed the case you mention is specific, it is what brought me to the second part of my reasonning, why do you (imo) put yourself the constraint of keeping reactor hot "at all time" ? I understand it's not exactly that from your explanation, rather you want to keep the heat exchanger able to resume steam production "fast" from your conclusion :

macdjord wrote: Mon Apr 06, 2026 8:19 am Thus, shutting off the heat exchangers allows the system to start up quickly on demand.

The conclusion is true imo, i agree, shutting off the heat exchanger allows the system to start up quickly and on demand. But also it's not "sufficient" to justify implementing the suggestion in a way, because you can "already" achieve that with method like Tertius shown : viewtopic.php?p=690766#p690766 It doesn't use a power switch, it's a different method that is more suited than power switch for the case you mention but still doesn't require pumps.

In my understanding this part is the one where we diverge :

macdjord wrote: Mon Apr 06, 2026 8:19 am [*]The reactors start producing heat again... but the heat exchangers can't produce any steam. They're all at 500º, and it takes time for the heat to get from the reactors to the exchangers.
[*]The rate at which heat flows between two object depends on the temperature differential between them. The first heat exchanger will start to produce some steam almost immediately, but it won't hit full production - 10MJ of heat into 989.7 units of steam per second - until the temperature gradient between it and the reactor gets steep enough for 10MJ of heat to transfer into it ever second. Worse, the second heat exchanger won't even start to heat up until then either, because the first exchanger will be consuming ~all the heat from the heat pipe before it can get any father down the row.
[*]As the reactors heat up, more and more heat exchangers come online, but each subsequent heat exchanger takes longer to do so, since more of the reactor's heat output is getting turned into steam instead of heating buildings up further. And meanwhile, the steam buffer is still dropping...
[/list]
See the problem? Now, if I instead shut off the heat exchangers:

to me this isn't a problem or i don't see it :

1) one way to deal with that is to have a steam buffer consequent enough that it will not run dry before all the heat exchanger can produce steam at full capacity again, it is easier to achieve if you "undersize" just a little the heat exchanger number compared to the reactor, to ease the propagation of heat to the last heat exchanger of the array. ( another would be to use a setup like Tertius posted ).

2)Another way to think of this to me is why ? why do you want to control when the heat exchanger generate steam ? isn''t that enough to control fuel in reactor ? sorry if you feel like you've explained this a lot, and i called it a "desire" rather than a "need" , because imo the wish to be able to control when the heat exchanger generate steam is not for the sake of it, it is rather because it allows fast and reactive nuclear power plant but i'm unsure of why if that is your only source of power. This is why i asked for the "compelling reason to keep reactor hot", i understand it's not reactor, but heat exchanger, but i still do not see the reasons as compelling, to me it feels like a part of a more general method to control the nuclear power plant, and such higher level of control can be achieved without it necessarily taking the form in practice of the kind of precise control on steam generation as you wish. And why would it be necessary to further control when steam is produced in a nuclear plant, if that is your only source of power ?

It's a bit of the same point, but 1) is more "can't you do that ?" (isn't that a working solution) whereas 2) is more why do you want the solution to be the way you want it to be ? ( what makes 'other solutions' to control power plant not satisfying').

macdjord wrote: Sun Apr 05, 2026 2:54 am Yes, adding circuit control to heat exchangers would solve my use-case. That's why I linked the thread suggesting just that both in the OP and in my previous reply to you. However, historically, the reactions of the devs to 'Can you add $CAPABILITY to $BUILDING?' has often been 'We could, but the fact that $BUILDING can't do that is intentional game design', so I am also pursuing the angle of asking them to restore an element of game balance that was thrown out of whack by a mechanics change.

I do feel adding circuit control to heat exchanger is exactly what you want to achieve in the most straightforward way. I see no objection to this. As you say maybe devs made it intentionnally this way for gameplay , or maybe it's a technical reason, maybe they havent considered it too much. I can't tell. I don't necessarily see objection to this suggestion like "no it should NOT be because .....", it's rather that i don't share the reasonning that leads to the conclusion as flawless i feel like 'hum" "this can be solved this way currently". or "but this is a self imposed constraint on the method to achieve a result that you can achieve in other way".

macdjord wrote: Mon Apr 06, 2026 8:19 am

mmmPI wrote: Sun Apr 05, 2026 11:55 pm I believe this is out of topic and useless in the current discussion.

I understand there are no objections to using a power switch to control the steam power plant like in the posted setup.

Then you haven't been reading my posts, because I've been objecting to it ever since you first suggested it.

What you say doesn't appear off topic or useless, on the contrary, this wasn't adressed to you.

I have been reading in details the posts ! You have said that there are plenty of cases where using a power switch is fine, i posted a setup that illustrate it i believe. I dont believe there are objections to using a power switch to control the steam power plant LIKE IN THE POSTED SETUP. The setup works fine right ? it's not the same use case as you mention, but then it is the subject of my previous question, why ? why do you 'force yourself' into willing to control steam when more comonly people control just the fuel. (i'm not surprised players will come and say : but you don't need that as imo) There are "solutions" to replace the offshore pump logic, and to have reactive nuclear power plant. ( if you only have 1 source of power, obviously power switch leads to black out, but then what's the need to control steam generation ? and not just fuel ? ).

This is why i think more strongly of the other other suggestion and thank you for linking it. If you say "i want to control heat exchanger with circuit" it's only devs that are going to say wether or not it is intentionnal for the gameplay, and you can mention all the new puzzle possible as positive things to explore if it is implemented, whereas currently i don't share the reasonning that lead to suggesting reducing the buffer of the entity, i feel the use case isn't really convincing compared to the simpler : it would be like before for the duration of the buffered water. whereas you seem to say : "it's not possible", "there is only one method that involve pumps", which prompt players to post their setup.

Really no matter what setup people post, the suggestion has merits on its own. But if the posted setup "invalid" the reasonning that claim you have to use pumps, then it doesn't serve the suggestion properly.

I care maybe too much about those things, sorry ^^

mmmPI wrote: Mon Apr 06, 2026 9:50 am

macdjord wrote: Mon Apr 06, 2026 8:19 am Thus, shutting off the heat exchangers allows the system to start up quickly on demand.

The conclusion is true imo, i agree, shutting off the heat exchanger allows the system to start up quickly and on demand. But also it's not "sufficient" to justify implementing the suggestion in a way, because you can "already" achieve that with method like Tertius shown : viewtopic.php?p=690766#p690766 It doesn't use a power switch, it's a different method that is more suited than power switch for the case you mention but still doesn't require pumps.

And as I said when he proposed it, his solution is basically what I'm using now, just with somewhat different control logic. However:

Tertius wrote: Mon Mar 30, 2026 8:18 am There is a valve (pumps) between heat exchangers and steam turbines+steam buffer. As default, it's closed.

Getting rid of the need for that pump is exactly what this proposal would allow me to do.

mmmPI wrote: Mon Apr 06, 2026 9:50 am 1) one way to deal with that is to have a steam buffer consequent enough that it will not run dry before all the heat exchanger can produce steam at full capacity again, it is easier to achieve if you "undersize" just a little the heat exchanger number compared to the reactor, to ease the propagation of heat to the last heat exchanger of the array. ( another would be to use a setup like Tertius posted ).

The amount of time the heat exchangers column takes to get back to full temp is substantial. It would require a significant increase in the size of the tank farm to achieve it, and I can't actually make my tank farm much bigger; I'm already very near the point of exceeding the maximum pipeline length when measuring side-to-side.

mmmPI wrote: Mon Apr 06, 2026 9:50 am2)Another way to think of this to me is why ? why do you want to control when the heat exchanger generate steam ? isn''t that enough to control fuel in reactor ? sorry if you feel like you've explained this a lot, and i called it a "desire" rather than a "need" , because imo the wish to be able to control when the heat exchanger generate steam is not for the sake of it, it is rather because it allows fast and reactive nuclear power plant but i'm unsure of why if that is your only source of power. This is why i asked for the "compelling reason to keep reactor hot", i understand it's not reactor, but heat exchanger, but i still do not see the reasons as compelling, to me it feels like a part of a more general method to control the nuclear power plant, and such higher level of control can be achieved without it necessarily taking the form in practice of the kind of precise control on steam generation as you wish. And why would it be necessary to further control when steam is produced in a nuclear plant, if that is your only source of power ?

You're right, I don't need to control steam generation. I could just use solar all the way. Or coal-burning boilers and generators. Or just uninstall Factorio and not play. I don't need to do any of this, but I want to.

As for why I want to control steam generation, I just bloody explained it in great detail. If I don't, then if I have a low steam buffer, cold heat exchangers, and high power demand, I can run out of steam and suffer a brownout.

macdjord wrote: Mon Apr 06, 2026 10:24 am

mmmPI wrote: Mon Apr 06, 2026 9:50 am

macdjord wrote: Mon Apr 06, 2026 8:19 am Thus, shutting off the heat exchangers allows the system to start up quickly on demand.

The conclusion is true imo, i agree, shutting off the heat exchanger allows the system to start up quickly and on demand. But also it's not "sufficient" to justify implementing the suggestion in a way, because you can "already" achieve that with method like Tertius shown : viewtopic.php?p=690766#p690766 It doesn't use a power switch, it's a different method that is more suited than power switch for the case you mention but still doesn't require pumps.

And as I said when he proposed it, his solution is basically what I'm using now, just with somewhat different control logic. However:

Tertius wrote: Mon Mar 30, 2026 8:18 am There is a valve (pumps) between heat exchangers and steam turbines+steam buffer. As default, it's closed.

Getting rid of the need for that pump is exactly what this proposal would allow me to do.

It is using a pump unlike what i said, correct.

But the pump is there only because of the arbitary need to have a power plant that keeps temperature hot imo. It is to show that what you request is possible to do, but you don't need the pump to make a working power plant !

macdjord wrote: Mon Apr 06, 2026 10:24 am You're right, I don't need to control steam generation. I could just use solar all the way. Or coal-burning boilers and generators. Or just uninstall Factorio and not play. I don't need to do any of this, but I want to.

As for why I want to control steam generation, I just bloody explained it in great detail. If I don't, then if I have a low steam buffer, cold heat exchangers, and high power demand, I can run out of steam and suffer a brownout.

I disagree with your explanations, you don't need to control steam generation to make a working power plant, nor do you need to use pumps. What you request is an additionnal method to do things that are already possible imo.

macdjord wrote: Mon Apr 06, 2026 10:24 am The amount of time the heat exchangers column takes to get back to full temp is substantial. It would require a significant increase in the size of the tank farm to achieve it, and I can't actually make my tank farm much bigger; I'm already very near the point of exceeding the maximum pipeline length when measuring side-to-side.

I don't buy that , can you share the design ? i can guarantee it is possible to make working power plant design that do not risk brown out in current state of the game, while being lossless.

macdjord wrote: Mon Apr 06, 2026 10:24 am

Tertius wrote: Mon Mar 30, 2026 8:18 am There is a valve (pumps) between heat exchangers and steam turbines+steam buffer. As default, it's closed.

Getting rid of the need for that pump is exactly what this proposal would allow me to do.

As far as I see the need for getting rid of the pump doesn't have enough positive impact on the game as a whole, for every player, to warrant a game engine change. It's a very specific use case.

It isn't clear if this will add build variety or remove build variety. If you followed the fff with the fluid mechanics change you will remember water supply by barrels was explicitly mentioned as benefit. If you reduce the buffer, it might be you're making water supply by barrel more unreliable for heat exchangers. You can always add a storage tank of course, but this will add additional buildings similar to adding the pump, so it's a draw.