counterflow heat exchange calculations

[el_turista]

how does one go about calculating the pipe length needed for counterflow heat exchangers? thanks

[NewWorldDan]

Trial and error, mostly. Longer is more efficient, better thermal conductivity is more efficient. Best case, you'll still have a few degrees difference between input and output. If you're making a petroleum boiler and a minor volcano is your heat source, I'd look at Francis John's tutorials. Otherwise, I just make it up as I go. If you've got aluminum pipes, there isn't much to worry about.

[MilesLong4200]

Just now, NewWorldDan said:

Trial and error, mostly. Longer is more efficient, better thermal conductivity is more efficient. Best case, you'll still have a few degrees difference between input and output. If you're making a petroleum boiler and a minor volcano is your heat source, I'd look at Francis John's tutorials. Otherwise, I just make it up as I go. If you've got aluminum pipes, there isn't much to worry about.

Do you ever find that making them too long will actually overheat your oil at the end and break the pipes? I'm trying to figure out why this has been happening to me

 

[NewWorldDan]

Just now, MilesLong4200 said:

Do you ever find that making them too long will actually overheat your oil at the end and break the pipes? I'm trying to figure out why this has been happening to me

 

I've had that issue. Don't run radiant pipes all the way to the spout. 

Look at the wiki here:

https://oxygen-not-included-guide.fandom.com/wiki/Petroleum_Boiler

There's a 1 tile drop after the petrol leaves the boiling chamber. That's where the radiant pipe stops and changes to insulated pipe. If you continue radiant pipes to the boiling chamber, there's a small chance of overheating and breaking.

[el_turista]

30 minutes ago, NewWorldDan said:

Trial and error, mostly. Longer is more efficient, better thermal conductivity is more efficient. Best case, you'll still have a few degrees difference between input and output. If you're making a petroleum boiler and a minor volcano is your heat source, I'd look at Francis John's tutorials. Otherwise, I just make it up as I go. If you've got aluminum pipes, there isn't much to worry about.

ive used that tutorial before, however this time my boiler wont fit the way he built it... but as long as i keep the same number of radiant pipes it should work regardless of how many "floors" it has? 

im not so good at trial and error with magma, and im trying to avoid sandboxing a copy of my map. i thought that this time i should actually figure out how its done properly. 

[Jann5s]

More floors tends to be slightly more efficient, have a look a waterfall heat exchangers if you are curious, but I wouldn't worry too much. It's as Dan says, just build one, if it's not perfect itvwill still work.

[Fradow]

You test them, mostly. I did that a while ago, for a range of length and designs (and there are a lot of useful comments in there too)

That can give you some idea of how long your exchanger needs to be according to your target efficiency, provided you use the same parameters.

2 hours ago, MilesLong4200 said:

Do you ever find that making them too long will actually overheat your oil at the end and break the pipes? I'm trying to figure out why this has been happening to me

That happens because (1) your counterflow is very efficient and push the oil very close to your boiling chamber temperature and (2) your boiling chamber isn't super efficient and heats the Petroleum too much above it's boiling point. You can either make your counterflow less efficient or make your boiling chamber more efficient (there is a lot of theory about that, but a simple improvement is to use normal ceramic tiles for the petroleum floor to have a more gradual heat injection rather than heat conductive tiles that are too spiky).

[LadenSwallow]

Is there a difference in 'natural' ceramic and ceramic tiles for this purpose?

[Prince Mandor]

43 minutes ago, LadenSwallow said:

Is there a difference in 'natural' ceramic and ceramic tiles for this purpose?

Yes and no. 'Natural' tile of 100 kg ceramic is equal to normal ceramic tile. But you can create 'natural' ceramic tile from 1g to 2000 g in mass, without fear of glitches, and some tons more with risks on save/load.

So there are no difference if you use same mass, but mass may be different

[Gurgel]

14 hours ago, el_turista said:

im not so good at trial and error with magma, and im trying to avoid sandboxing a copy of my map. i thought that this time i should actually figure out how its done properly. 

I think if you do it without experiments, you will need to go into some not so simple mathematics. Basically ONI is not very different from a real counterflow heat exchanger and you can find how to calculate them via Google. You will probably have to dig into some differential equations (the only thing in mathematics I never really understood and never needed so far).

I think in actual reality, engineers will mostly simulate and not calculate these designs today, i.e. experiment.

One additional complication is that if you do it for crude oil boiling into petroleum, you have point where one material changes to the other in there, with a bit different thermal properties.

 

[Prince Mandor]

15 hours ago, el_turista said:

ive used that tutorial before, however this time my boiler wont fit the way he built it... but as long as i keep the same number of radiant pipes it should work regardless of how many "floors" it has? 

im not so good at trial and error with magma, and im trying to avoid sandboxing a copy of my map. i thought that this time i should actually figure out how its done properly. 

Really it is very untrivial calculation, because every tile of petroleum continue exchange heat with neighbor tiles. And there are also some bugs in this process, making left-to-right not equal to right-to-left etc.

So, simplest way is not limit yourself on first build. Made it, as good start,  20 working tiles, and see if result is good enough. If not, add couple of pipes, if it is good and you need space, remove couple of tiles.

Usually, it is heater design where we lost most heat, not exchanger. If heater heats up petroleum after conversion, this is a place for most optimization

 

 

BTW, there was design with flaking, but horizontal flow of oil. I cannot found it anymore. Do anyone know where to look?

[Fradow]

3 hours ago, LadenSwallow said:

Is there a difference in 'natural' ceramic and ceramic tiles for this purpose?

A ceramic tile is considerably easier to build. As Prince Mandor explained, you can have a "natural" ceramic tile of any mass, but I don't think it would be worth doing: any mass should work reasonably well.

6 minutes ago, Prince Mandor said:

BTW, there was design with flaking, but horizontal flow of oil. I cannot found it anymore. Do anyone know where to look?

Are you thinking of mathmanican's Bead Flaker?

 

[Prince Mandor]

1 hour ago, Fradow said:

Are you thinking of mathmanican's Bead Flaker?

 

No, about Zarquan's version at the end of same thread. With horizontal oil flow and chlorine as 'heatplate'. Thank you! It was exactly what I was searching

[LadenSwallow]

I was just thinking you can make a 100kg natural ceramic tile by 'building' a farm tile out of clay and thus save the coal it would take to refine the clay by kiln.

[JRup]

@LadenSwallow: I believe one only needs to cook at least 1g of clay on top of a hot tile (for example) to gain access to a ceramic natural tile because any amount that is cooked will morph accordingly. One would only need to "top up" the mass of the resulting natural tile with an Automatic dispenser and the needed amount.

The difference this makes is that of time needed to cook the clay to get the natural tile...

[Padishar]

12 hours ago, Gurgel said:

I think if you do it without experiments, you will need to go into some not so simple mathematics.

 

I agree. From my undestanding first you would need to calculate in each tile how much temrnal conductivity you have, which should be an average of the one of the oil, the pipe and if any of the gas or solid tile in which the flow is happening. On top of that the amount of heat exchanged depends on the temperature difference between the oil and petroleum, which is not constant: it varies on each tile and has to be calculated accordingly.

In theory there shouldn't be any integral or differential equation, since oni world is pixelated (space is discrete, not continuos). Anyway the calculations are too complicated unless one wants to have fun and try to write an algorithm to run the calculation.

[Gurgel]

3 hours ago, Padishar said:

In theory there shouldn't be any integral or differential equation, since oni world is pixelated (space is discrete, not continuos). Anyway the calculations are too complicated unless one wants to have fun and try to write an algorithm to run the calculation.

Well, yes. But mathematically, that may make it even harder, because you get no closed-form solutions anymore, but some probably recursive things that you just need to run until they converge enough. (If they converge. But I think heat-exchange in ONI is dampened/sluggish enough that they should. Not that this is a sure way to get something to not oscillate, but often it works.) Easy to simulate, hard to solve. Now, real-world counterflow heat exchanger apparently use some differential equations that apparently have closed-form solutions and you _could_ probably use them as approximations for how things work in ONI. But I doubt it is worth the effort, when you already have a nice, accurate (!) simulator available.

That said, I do understand the desire for some nice equations you can just put some desired numbers in and it tells you exactly what you need to build. It is just that things with feedback loops often do not work that way.