We need to cook!

[biopon]

In light of the new (to me) aluminium pipes, I wanted to optimize the heat exchanger on my petroleum boiler build. 

The heat exchanger is a crucial part of the petroleum machine. If it's good enough, you can minimize magma used (good for minor volcanoes or ground-based heat) and it doesn't have to take up half your base either. Aluminium helps a ton. However, you shouldn't make it TOO good either. There are two issues: when oil phases into petroleum, its temperature stays the same but its SHC goes up, creating heat in the process. And since this occurs at or over 402C, if your heat exchanger is too good, you'll get broken pipes (unless you try to avoid this specifically with shutoffs and temp sensors, which is possible but let's try to keep things simple.) Compounding this problem is the fact that the temperature in the boiling chamber tends to be spiky. You try to keep it as close to 402C as possible, but temp sensors don't react instantly, the petroleum in the chamber and whatever's communicating heat both have a buffering effect. So you'll see temperatures vary between, say, 400 and 404C in a decent boiler, but maybe much more. 

Putting the spiking issue completely aside, I wanted to find the perfect heat exchanger, one that doesn't gobble up half a volcano but isn't too large, and won't need constant repairs either.

So I cooked some petroleum.

I built this:

And this:

And every variation between.

I let them warm up, fixed pipes wherever needed, ran everything for a few cycles, and then reset everything to the same 90C oil, and 1000 tonnes of 2000K magma.

After running it for a fair while, I stopped everything, and noted the temperature of the magma in each of the supply tiles. Almost half of the exchangers proved to be "too good" and stopped due to a broken pipe, these were ignored. (The boilers themselves are a very naive design and were spiking up to 406C, way above the tolerance of the more efficient exchangers, but the boilers weren't what I wanted to find out about.) The ones that did not break were still very good.

This one came out at the top:

(It is showing pipe damage, but I'll let that slide. It's the only one that took any damage but survived.) It decreased the 1000 ton magma blob's temperature from 1726.9C to 1719.7C, which means a total of 7200 mDTU was transferred to the oil. The petroleum comes out at 105C (1848kDTU/s), the incoming oil is 90C (1521kDTU/s), total gain per 10kg is 327kDTU/s - this means this must have boiled about 220 tons of oil, which sounds about right since I ran this for ~37 cycles. Putting this in "amount of magma used" terms, if your input magma is a perfect 1726.9C, and you use it down to 500C before grabbing a new chunk, you will go through 279 grams per second.

In comparison, shorty here used 19100mDTU total (magma blob temp 1707.5C), which would give it the performance of 741g/s.

Aluminium is just amazing. I wouldn't have thought something like this heat exchanger would even be able to cook petroleum, let alone do it fairly "efficiently". The output fluctuates around 134C, which means you'd need a steel pump, but still, it fits anywhere.

So while size does matter, what is optimal? Clearly the winner above is overkill, unless you have an earth hugging build. Shorty will get the job done as long as you have a decent volcano, don't care about your output temperature, or about having magma left over. However, just by making it two tiles wider, and going from an area of 56 to 72 tiles, magma use drops to 605g.

Spoiler

Or you can go 2 taller instead, for 70 tiles used, and use 613g/s magma.

Spoiler

Is it better to go wider or taller? Taller, but it doesn't much matter. This is the widest and shortest one I tested, it has an area of 120 tiles, and uses 463g/s magma:

Spoiler

A tall one with similar size (112 tiles) is calculated at 462g/s:

Spoiler

As a general relationship between the area of the heat exchanger and its efficiency, this graph should help:

The X axis is the area of the heat exchanger in number of tiles, the Y axis is the mDTU taken from magma divided by the area of the heat exchanger. Shorty is at the top left, and the winner I showed first is in the bottom right. It seems there's a sweet spot around 150 tiles, and while there are data points that share the same X axis point but differ on Y, the difference is not significant (hence tall vs. wide doesn't matter much). 

Above 100 tiles the graph becomes linear, which means there's a linear correlation between size and efficiency. Anything less than 100 tiles seems like a bad use of space though. 

This is 100 tiles, 466g/s calculated efficiency:

Spoiler

From the sweet spot around 150, this is 156 tiles with 376g/s calculated efficiency:

Spoiler

To cut this short: aluminium is amazing, and if you build a heat exchanger larger than 100 tiles, you're not doing it wrong.

Here's my data:

https://docs.google.com/spreadsheets/d/1BQ4lFuobrmYnYRyKb8F9BbnMItVSSKcc_gqAiCLExaA/edit?usp=sharing

And here's the save.

The Antigravity Sewer 3.sav

 

 

 

[Yalp]

8 hours ago, biopon said:

Compounding this problem is the fact that the temperature in the boiling chamber tends to be spiky.

Just a thought:

an in between buffer at a somewhat fix temperature could smooth things.

Something like a steam chamber (with relatively high thermal capacity), where the magma added in on demand.

For testing just replace the magma with steam (temperature to be determined).

All that will obviously require extra space and probably won't be worth the effort, just something to consider.

[biopon]

2 minutes ago, Yalp said:

an in between buffer at a somewhat fix temperature could smooth things.

Something like a steam chamber (with relatively high thermal capacity), where the magma added in on demand.

For testing just replace the magma with steam (temperature to be determined).

All that will obviously require extra space and probably won't be worth the extra effort, just something to consider.

An inbetween buffer does sound good, the unit for conductivity is (DTU/(m*s))/C, which means conductivity heavily depends on temperature. (This is intuitive too but it's good to see it written down.) So having a relatively stable, and moderately hot buffer could greatly help smooth things out.

I've tried using a natural aluminium tile in the wall between the door and the crude, with an entombed thermometer. Not as a buffer, but as a better place for triggering the door. Too much mass, too high SHC. It might work with gold, I'll test that in debug when I get the chance. Even though I doubt it, unless it's possible to somehow make low-mass natural tiles from metal. 

I think all this is worth the effort as long as it's fun, but I'm not sure if it'll be worth the space it adds to a build.

[Yoma_Nosme]

I use basically the same system in my build. But I add a tank(~6x4) to capture the petroleum. There I have radiant liquid pipes as well(first preheating for oil). The tank has tempshift plates. It adds a lot to temperature stability to the system. 

Petrol in the tank stabilizes at around ~200c. Also the tank has a mechanical overflow via auto doors for further cooling or whatever. 

[0xFADE]

Aluminum is way to overpowered compared to other early metals. Though I guess the balance is the rarity on the other maps. Care packages of it are only 100kg so they understand the value. 

[fishoutofwater]

@bioponwhat happened if I use the old counter heat exchanger using 2 liquid; petrol and ethanol, hot water comes from the left side passing through ethanol, cold polluted water comes from the right side passing through petrol, can this delete heat if the ethanol evaporate and then condense later?

I guess I just build it, don't have petrol at the moment

Spoiler

Spoiler

 

[lonelysquirrel]

I've played around with a similar boiler, and my experience was that I could stop pipes from cracking by putting a 3 tile gap at the top the heat exchanger:  the crude oil pipe bends up in your build at the lip of the boiler, this is a 0 tile gap.  Move the bend to the right 3 tiles.  This will produce a heat gradient on the flowing petrol from the boiler lip (at 406) to the tile with the radiant pipe (at 399). 

Sorry I can't provide a screenshot right now.

[biopon]

9 hours ago, fishoutofwater said:

can this delete heat if the ethanol evaporate and then condense later

Sure, that cycle erases energy, much like how cooking petrol creates it. I'm haven't really played with this concept much, my only game recently has been on Rime and I had steam turbines way before I had heat issues. I'd love to see what you come up with. TBH you don't have to have petroleum for this to work, can't you have your hot water & cold water interact with the ethanol directly?

6 hours ago, lonelysquirrel said:

putting a 3 tile gap at the top the heat exchanger

It's interesting that this worked for you - to me it sounds like just kicking the can down the road. I mean, the reason the petroleum gets cooler as it flows out of the chamber is because it has a pipe in it that cools it down; if you move the pipe out at the top left for a few tiles, you just move the point of contention a few tiles to the right. But since you say it worked, there must be something else going on.

[fishoutofwater]

1 hour ago, biopon said:

 can't you have your hot water & cold water interact with the ethanol directly?

I was thinking If I use ethanol only I will have to fill 2x2 chamber with 4000kg of it, I’m afraid it will be hard to boil, in the picture above I only put 200kg of crude oil and 200kg of ethanol, hoping that I could delete some heat from the hot water from my cool steam vent.

 

[Yalp]

8 hours ago, lonelysquirrel said:

I could stop pipes from cracking by putting a 3 tile gap at the top the heat exchanger

This might be the simplest and most elegant solution. I saw others replace the last few radiant pipes in petroleum with regular/insulated pipes which has a similar effect. This also allows for easy readjustments while running the whole thing.

1 hour ago, biopon said:

the reason the petroleum gets cooler as it flows out of the chamber is because it has a pipe in it that cools it down

For every tile a thermal spike is transmitted, its amplitude is reduced as only part of it will carry over. By the time the spike would have transferred it's maximum temperature all the way to the pipe, the incoming crude oil already counteracted it to a degree (or several °, haha - sorry, stupide pun ).
The petroleum in in contact with the pipe also cools the petroleum from the cooker, dampening the spikes at the pipe further.

 

[lonelysquirrel]

2 hours ago, biopon said:

 

It's interesting that this worked for you - to me it sounds like just kicking the can down the road. I mean, the reason the petroleum gets cooler as it flows out of the chamber is because it has a pipe in it that cools it down; if you move the pipe out at the top left for a few tiles, you just move the point of contention a few tiles to the right. But since you say it worked, there must be something else going on.

I think one of the reasons it worked for me is that you are reducing the thermal mass of the petroleum in thermal contact with the crude.  You are instead exchanging heat into the crude piping through a couple tiles of 10kg petroleum.  Whereas if you have 10kg petroleum immediately next to 700kg of 402C petrol, the single 10kg petrol tile can be overwhelmed by the boiler.

Also the crude is approaching 400C from below. When the petrol flow starts you should have crude at about 150C and petrol at 402.  That should not crack.  The top of the counterflow exchanger should fall below 400 before the boiler's heat makes that tile go above 400.  Then as long as you maintain the temp gradient it should be stable.

However I have not done extensive testing. If there is some pipe damage before it reaches equilibrium, I wouldn't be totally surprised.