Pipeless Counterflow Liquid Heat Exchanger -- 90 kg/s Petroleum Boiler

[Zarquan]

I decided I wanted to build a petroleum boiler that could theoretically process as much crude oil as I could throw at it without any redesign after I set it up.  The issue is that pipes can only carry 10 kg/s crude oil in a heat exchanger, making an exchanger that handles more than 10 kg/s crude oil tricky.  To resolve this limitation, I created a pipeless heat exchanger.  To demonstrate this, I have provided a magma powered petroleum boiler capable of boiling 90 kg/s of crude oil with no thermal transfer pipes.

Spoiler

I did build this in Debug mode, and the insulated tiles are insulation here.  However, they could just as easily be ceramic or mafic rock with little change to the effectiveness of the system.  Other than this, no space resources were used.

The boiler itself is crude and not really of note, but I will put details about it in a spoiler. 

Spoiler

All the metal tiles are made of aluminum and there are a few spots of 1000 kg steam for extra thermal capacity.  The thermal sensor at the bottom is set to 425 C for stability.  The magma source in this example is infinite, as the item I am showcasing is the heat exchanger, not the boiler itself.  However, the magma delivery system is a fairly standard and works like it would with a normal volcano. 

It used 10.2 tons of magma in a cycle, but there is still a lot of heat in the igneous rock which could be extracted, so this number could easily be reduced. 

There is liquid aluminum to protect the igneous rock pile from the magma and prevent tiles.  The automation is essentially as follows:  If the temperature is above 1300 on the thermal sensor next to the door, don't open.  If there is magma on the tile above the liquid aluminum, don't open.  If it is below 1300 and there is no magma, open for 1.2, then close for 5 seconds.  This SHOULD prevent blocks from forming, but I haven't tested it over a long term.

This heat exchanger works by having the crude oil flow down and the petroleum flow up via escher waterfalls.  By doing this, I am able to make the heat exchanger operate on 90 kg of crude oil every second.  It could potentially go higher, but I have not seen a map with much more than 90 kg/s crude oil.

The crude oil is piped in to the top, and there are 9 liquid vents bringing in crude oil at about 87 C.  The crude oil falls down to the first level of the heat exchanger, where it flows right, then falls and flows left, etc, until it reaches the bottom level, where it pours in to the boiler at around 373 C.

At this point, the crude oil flashes in to petroleum in the boiler, and is poured through an escher watercall to be forced up and to the left, thereby running the hottest petroleum past the hottest crude oil.  On the left side, there is another escher waterfall pushing the petroleum up and to the right.  When it reaches the top, it is at about 120-125 C.  This is almost in the range where it could be picked up with a gold pump, but I would stick to steel. 

To prevent the system from becoming clogged with too much crude oil, I added a liquid element sensor to turn off the crude oil input if it detects crude oil 2 tiles high on the lowest level.

Pros:

• Smaller than a 90 kg/s heat exchanger with similar thermal properties.
• Can easily be scaled up without altering the pipes
• It was fun to build
• Broken pipes can no longer happen, so significantly reduced maintenance requirements if something goes wrong.
• Minor errors in the liquid flows correct themselves.

Cons:

• Most people don't need a petroleum boiler with this much capacity.
• Building it is far more difficult than a standard boiler, as there are many escher waterfalls
• Priming it is more difficult than a standard boiler, as there is a lot of petroleum in the heat exchanger.

Save Game:  Pipeless heat exchanger Cycle 767.sav .  You can find the boiler on the far right of the screen.

[TheMule]

This is brilliant.

I've been toying with the same idea for a while, but deemed too complicated for survival. 

[0xFADE]

Oh it goes up instead of down without pipes?  Very nice.  it is much larger than a single 10kg pipe version but not larger than 9 of them.

[suicide commando]

This is a Brilliant piece of engineering! AWEsome built, lots of kudo's.

 

[Yalp]

Awesome build. You might want to consider spacing the TSP 2 tiles apart, so their exchange zones don't overlap. It should result in a higher horizontal temperature gradient for every level.

[SamLogan]

Nice build.

[mathmanican]

12 hours ago, Zarquan said:

capable of boiling 90 kg/s of crude oil

The theoretical max is 250kg/s (with a viscosity rating of 50kg/0.2s).  I love it. Nice work. Maybe I should return to magically creating extra heat via phase shifts.  Life has been pretty busy for the last 7 months. The next 5 won't be nearly as crazy. 

 

[suxkar]

2 hours ago, mathmanican said:

The theoretical max is 250kg/s (with a viscosity rating of 50kg/0.2s).  I love it. Nice work. Maybe I should return to magically creating extra heat via phase shifts.  Life has been pretty busy for the last 7 months. The next 5 won't be nearly as crazy. 

 

Please daddy Mathmanican come out of retirement, we miss you mercilessly abusing Oni mechanics! At least I do!

[TheMule]

3 hours ago, SamLogan said:

To cook 1 Kg of crude oil at 390°C (according you transfer most of the petroleum's heat) to 399.85°C, you will need 194 764 DTU/s

Isn't it (403-390) * 1.69 * 1000 = 21 970 DTU/s? (oil turns into petroleum at 402.85C).

For 90kg/s it's 1 977 300 DTU/s. Am I missing something?

[SamLogan]

10 minutes ago, TheMule said:

Isn't it (403-390) * 1.69 * 1000 = 21 970 DTU/s? (oil turns into petroleum at 402.85C).

For 90kg/s it's 1 977 300 DTU/s. Am I missing something?

Yes, I did a miscalculation.

[Zarquan]

4 hours ago, SamLogan said:

Nice build, but I see two limitations :

1. The thermal consumption

To cook 1 Kg of crude oil at 390°C (according you transfer most of the petroleum's heat) to 399.85°C, you will need 194 764 DTU/s so for 90 Kg/sec, it represents 17 528 760 DTU / sec.

One Volcano can generate around 2 400 000 DTU/sec, so you need 7-8 volcano to sustain your boiler or 14-15 Aquatuners (thermium / super coolant).

It seems it only possible in Volcanea or with a lot of AT.

 

2. The crude oil production

To sustain, your setup, you need 37 oil well, it merely impossible even with the traits.

 

But it's an interesting concept.

The point is the capacity of the heat exchanger moreso than the boiler.  I can make it taller for more heat conversion, but I just want to be able to ramp up production without having to lay new pipes or fundamentally change the exchanger itself or without building more of them.  The fact that my heat sources are the limiting factor is a win in my book, as usually I have more magma than I know what to do with in my boiler setups and excess crude oil.

Also, you only need 27 oil wells, as each oil well produces 3333.33.  OASIS-A-2113944774-0 has 23 oil wells, so having around 90 kg/s isn't completely absurd.

4 hours ago, Yalp said:

Awesome build. You might want to consider spacing the TSP 2 tiles apart, so their exchange zones don't overlap. It should result in a higher horizontal temperature gradient for every level.

I did that and I also changed most of the tiles in the upflow from the escher waterfall in to aluminum tiles.  Now the crude oil coming out is at around 380, which let me reduce the temperature of the boiler (which would let me get more heat out of my igneous rock).  However, the heat the crude oil goes in is not as important as the heat at which the petroleum comes out, which I haven't been reducing very effectively. 

I'm going to try making it taller and narrower version.  Perhaps with two columns.  I kind of want to try one with the crude oil flowing up, but I can't find a way to do it without the chambers being 4 tall.

[SamLogan]

2 minutes ago, Zarquan said:

Also, you only need 27 oil wells, as each oil well produces 3333.33.  OASIS-A-2113944774-0 has 23 oil wells, so having around 90 kg/s isn't completely absurd.

Ok thanks for the share.

[Zarquan]

3 hours ago, TheMule said:

Isn't it (403-390) * 1.69 * 1000 = 21 970 DTU/s? (oil turns into petroleum at 402.85C).

For 90kg/s it's 1 977 300 DTU/s. Am I missing something?

3 hours ago, SamLogan said:

Yes, I did a miscalculation.

So with that number, it should only take one or two volcanos to power this, right?  Also, rockets are a good heat sources that shouldn't be ignored.

[mathmanican]

Hi @Zarquan. Yesterday I couldn't stop thinking about your heat exchanger, and came up with an alternative. I read through lots (lost) of heat transfer threads, such as @biopon's thread on counterflow heat exchangers and @socooo's 10kg crude oil in, 10kg oil out thread (a few of my personal favorites). Here is what I came up with. Have you seen anything like it before (I haven't)?  

A series of escher falls basically create vacuum like separation horizontally. You can make this as wide as you want, and you can feed it with pretty much endless liquid (limited only by the natural flow rates of liquids). It's not optimized yet at all. For example, would 3 tall towers be better? Where to place tempshift plates? Large blobs of crude on top interact with small amounts of petro below, so should I put the petro above with an escher fall at the end to wrap the crude onto the cooker? Do I want gravestones in each layer to even heat vertically? Would conveyor bridges help or hinder? Alternating diamond/metal ties, or solid blocks? The list goes on. I've wasted a day playing with it, and would love help from any interested parties. Some improvements are massive, others not so much. The nice part is that this is extremely simple to make, can be placed under a 2 stable wide patch of ground anywhere, with metal refinery to provide heat early on. As you already mentioned in your original post, this thing can be scaled up from 1kg/s to 90kg/s, pick your flavor, with no extra upkeep. At 10kg/s, I can get the temp on the crude anywhere from 350 to 390 safely, with petro leaving in the 120s. At 90kg/s, petro leaves in 130s. A few more sections would probably help this.  You could even have the thing wrap around, with two more layers below or above this, doubling the height and number of interaction chambers. 

The gases are 100mg CO2 and 100mg chlorine, so basically vacuum, between separate compartments. After this is turned off, the liquids equalize above and below, and don't change temp anymore between compartments (or if they do, it will take thousands of cycles to even everything out).

It's super easy to install in survival. For those who are not sure how to build this (I'm pretty sure this is cake walk for you @Zarquan), here are some steps. For now, no save file attached (if we optimize it, we'll make a new post and add the save file there). Feel free to message me personally, or we can play with it publicly here. 

Spoiler

1. You need to trap the gasses first, so build the surrounding tiles only around the gas enclosures. You can put ladders in place of the other things for now, and remove those later.
2. Now drop 30g petro/crude on each end of the gas traps, making sure you use the correct liquid.  
3. To get the petro there, build liquid pipes that pass through each spot you need petro (these can be constructed before you build tiles). Then use a liquid valve set at 30g or so and fill all the pipes in a loop with this amount.  Deconstruct the feed pipe that filled the initial loop, and then deconstruct the liquid valve.  All the liquid will stop flowing completely, get stuck in place, and then you just deconstruct the pipes in the spots you need them. 
4. Now place the gasses.  Do the same thing as the liquid, but set your air filter as low as it goes (100mg). After you ill the loop, deconstruct the feed pipe and then the valve, then deconstruct the pipes over the needed spots.  
5. Once you've got your gas traps made, build the rest of tiles, clean up debris, and seal it off.  Then start the machine.

Here are a few pictures. 

Then build the walls surrounding the gas rooms, deconstruct the appropriate gas pipes to get your 100mg in there. Then reconnect the pipes, let everything you didn't use flow away. Clean up the debris, enclose the area, and you're done. 

 

[Zarquan]

I think this is a superior design.  The increase in the number of zones in such a small space should be a significant improvement.  Thanks! 

 

[mathmanican]

18 minutes ago, Zarquan said:

I think this is a superior design.

It has a flaw in terms of high flow applications, namely that you have to deal with both vertical compression and horizontal flow limitations.  When I set 90kg/s loose on this thing, you have to be careful where you place the liquid, otherwise it will start to back up on the infeed line (which I guess naturally jams the vents and it self regulates (about 72kg/s Edit:Have to do more testing, but it eventually blocks vents).  the chambers pretty much hit equilibrium mass in a few moments, as you change flow rates.  It takes a tad longer for changes in flow rates to reach new equilibrium temps in each chamber, but less than a cycle.  Oh, and the outflow rate pretty quickly matches the inflow rate. No waiting a long time for stuff to make it all the way.  Escher falls are magical.  

[Zarquan]

1 hour ago, mathmanican said:

When I set 90kg/s loose on this thing, you have to be careful where you place the liquid, otherwise it will start to back up on the infeed line (which I guess naturally jams the vents and it self regulates (about 72kg/s). 

I had a similar problem with my original design.  If I put too much in at once, it backed up and didn't work right.  For example, sometimes crude oil would block the petroleum output.  When I started it, I stepped it up in increments of 10 kg/s, waiting for it to approach an equilibrium before bumping it up again.  It didn't take too long.

I'm surprised that it would back up to the point that it blocks the liquid vents though.  I never had an issue like that.  But then again, my vents are just sitting around 5 tiles above the liquid.  Could we see a picture of how the vents are set up?

Quote

For example, would 3 tall towers be better? Where to place tempshift plates? Large blobs of crude on top interact with small amounts of petro below, so should I put the petro above with an escher fall at the end to wrap the crude onto the cooker? Do I want gravestones in each layer to even heat vertically? Would conveyor bridges help or hinder? Alternating diamond/metal ties, or solid blocks?

I don't see why making the chambers taller would help.  That would mean more mass and more pressure (I think).  I built mine tall because I was emulating classic heat exchangers and the petroleum had to be moved to above the crude oil.  Since you don't have this problem, I am thinking smaller is better, as all the tiles in each segment are in thermal contact with a temp shift plate that is directly moving heat to the diamond tiles.  Plus, if it is smaller, then you can stack them more efficiently and get even more efficient transfer.

I'm not sure the insulated tiles in contact with the liquids helps.  I think you want to transfer as much heat as possible in those three tiles, so throwing in bridges and anything else that can help is best.  My feeling is that the 3 tile area is small enough that we shouldn't worry about heat backflow and treat it as one unit of heat transfer.

To this end, your 3 tile high diamond areas could be separated and have a temp shift plate in between.  This would be good because it is in direct contact with both the petroleum and crude oil.  In that gap, you could put water (steam) or phosphorus or something to increase thermal transfer.  If you are feeling adventurous, you could try putting a natural aluminum tile there on top of the temp shift plate, but that is dangerous.

I definitely think you want the insulated tiles where the escher waterfalls are, as that is backflow we should avoid.

Did you encounter any pressure damage in your setup?

[mathmanican]

45 minutes ago, Zarquan said:

Did you encounter any pressure damage in your setup?

Nope. not even at 90kg/s.  Basically liquid leaves each region as fast as it enters.  The only backup I had was at the start, when I tried to extend the entrance to squeeze out a little more heat.  Basically, I was hoping that I could drop 130C oil down to under 115, but failed when things backed up. 

47 minutes ago, Zarquan said:

Could we see a picture of how the vents are set up?

The problem is clearly the narrow tube that I forced the crude through.  I can fix that, and it shouldn't be a problem anymore. The flow rate should be up to 250kg/s (theoretically), though at this rate I'm not sure the upwards flow will be fast enough to prevent structural damage, nor do I think anyone ever needs 250kg/s (but science dictates we try). 

This is currently running at 20kg/s.  I put 1000kg/s of steam in the chamber you suggested, and plastered tempshift plates everywhere except the gas tiles (that would defeat the build). It takes a lot longer to stabilize (having to wait for 1000kg of steam to reach temp, but the results are AMAZING! The crude oil is entering the cooker at over 395C.  The petroleum is leaving the exit at about 113C. Both temps are rising very slowly, so I'll let it run for an hour and see what it stabilizes at.  I have a temp sensor on the first blob of petro after the first waterfall, that keeps the petro from ever passing 401.  I had the conveyor rails instant boil some crude on the top once when I let too hot petro in, so that temp sensor prevents extra heat from getting added. (Just had to drop it to 401, and lowered the cooker from 425 to 420, otherwise this was gonna blow up. 

The temperature delta between the last bit of crude, and first bit of petro is under 6C, and shrinking slightly.  Input crude 85C, output petro near 113C I could improve the last tiny heat exchanger a bit, maybe. It would probably be better to just add a couple more chambers?

Here is a save (I guess I'll share it). It has a stalled spiral boiler in it, that sucked itself to vacuum, and will restart when ready for more crude. At some point I want to see the potentially catastrophic startup, as I need to reset the temp sensor on the supercoolant pool to make sure the liquid methane never freezes.... Oops. 

Lucky Colony.sav

 

The 72kg/s I mentioned earlier, that was wrong.  What you see is 72kg(r rather 71.9 because the code displays truncated 71.999999 as 71.9) in each tile because of liquid flow rules.  Note that 72/4*5=90. With 1/4th of the liquid moving left each tick, and each tick happening 5 times a second, that is where the 72 came from.  So the actual flow rate of 90kg/s will cause all your crude oil ledges, right before the escher falls, to display 71.9kg. Once you see this on each ledge, you know that full flow has been achieved. 

Similar computations will let you know when any flow is achieved. If you use 20kg/s, then you should see 20*4/5=16 (or probably 15.9(9999....)).  Just multiply to 4/5th, and you know when full flow is achieved. 

[mathmanican]

I decided to crank it up to 180kg/s.  Unfortunately, if the crude doesn't vaporize fast enough, then when it changes to petro, a high mass single liquid bead explosion occurs, and that has huge ripple effects, stopping the whole process, causing backups and pressure backlogs. This would be an argument for making a three or four tall tower for the first esher waterfall (maybe even higher). If you could keep the boiler plate hot enough, It could process 180kg/s.  I'm pretty sure this thing could handle 250kg/s.  Haha!!! Now, where will I get enough heat to process that much oil (let alone that much oil).  We could burn through the entire oil biome in less than a cycle..... I wonder how many aquautners I would have to keep running at above 900C to make this work, and is that even hot enough....  

Yep. 250kg/s is the max flow.  You can see the 199.9kg in the picture below, which is what you expect as 200 = 250*4/5. 

Once I dropped the cooker down a tad, and mad the first tower 4 tall (the rest are 2 tall, it worked without a hitch.  I have 9999kg 85C oil pressing in on the inflow, to force as much liquid through as possible. No pressure damage anywhere along the exhanger. 

It will eventually have a problem (which can be worked around). The bubbling up crude causes less than 250g/s occasonally to flow through the intake.  As such, the pool will slowly build. This seems to fix the issue.

 

[mathmanican]

Now I just need to make a bead dropper (cannot accidentally form waterfalls) that causes the crude to bead into the cooking pool, as we already know that dripping liquid enables extra heat deletion. I'm thinking a bit of chlorine gas, along with the same bead pump from the spiral boiler, will do the trick. If we can bead the crude into the hot petro pool (without getting the petro to bubble back up), we'll get way better use of our boiler plate (fewer volcanos needed).  Problem is preventing the petro from bubbling up the beads...

 

[Zarquan]

1 hour ago, mathmanican said:

Now I just need to make a bead dropper (cannot accidentally form waterfalls) that causes the crude to bead into the cooking pool, as we already know that dripping liquid enables extra heat deletion

I thought they fixed dripping liquids deleting heat.  I know that boiling can delete heat if the 5 kg boiling occurs, but I think this can be resolved with the shape of the boiler.

EDIT 2:  FINAL RESULT:  Beads do cause heat deletion in this context (even in my boiler design).  Beadless pouring off of a ledge did not result in boiling until the crude oil took up two tiles.

EDIT:  To be clear, the "flaking" boiling mechanic can be avoided with a very narrow boiler plate (one tile wide) with no more than 1 tile deep crude oil and at least 1 tile deep petroleum on top appears to prevent the boiling mechanic.  I believe it is because the crude oil doesn't know where to go, so boiling does not take place. 

Spoiler

If you have 2 tiles wide instead of one, boiling can occur.  Though to be clear here, most of the boiling liquid is petroleum.  However, 5 kg of crude oil did boil off before it was shoved to the side.

I tested the bead scenario (with a 500 kg pool of crude sitting under 1000 kg petroleum and introducing a 50 kg bead of crude oil from above the petroleum) in the 1 tile wide boiler plate, and it boiled off 5 kg of crude oil through the flaking mechanic.  Ergo, energy lost for each bead dropped.  Or energy gained if you feel like doing some precise solid block formation.

 

[mathmanican]

2 hours ago, Zarquan said:

FINAL RESULT:

So we should always use a 1 tile wide boiler.  I'll update my design, leaving room higher up on the sides for stuff to gather after phase change. 

An alternate plan might be to force beads (they fall at a steady rate), but then boil them in mid air as they fall (no solids below them), hitting the bottom as petroleum, so that the change occurs mid flight. This isn't too hard to make happen, but does require a little more precision in cooking than just "let it slop down and add heat."  Thoughts?  This is a project I was working on to generate heat with Molten Lead, last year before work got super busy (though I was cooling stuff, rather than heating stuff, mid flight). Single tile wide boilers are definitely simpler. 

[Zarquan]

1 hour ago, mathmanican said:

An alternate plan might be to force beads (they fall at a steady rate), but then boil them in mid air as they fall (no solids below them), hitting the bottom as petroleum, so that the change occurs mid flight. This isn't too hard to make happen, but does require a little more precision in cooking than just "let it slop down and add heat."  Thoughts?  This is a project I was working on to generate heat with Molten Lead, last year before work got super busy (though I was cooling stuff, rather than heating stuff, mid flight). Single tile wide boilers are definitely simpler. 

I'm not sure.  I don't think the beads are treated any differently from any other liquid.  If the bead is in mid air without being flanked on both all sides by a non-matching liquid or solid, it will boil 5 kg on the way down.  I think the best way to deal with it is simple pouring unless you can make sure your beads are less than 5 kg.

An accompanying picture.  The flanking tiles are 750 K insulation and this is what it looked like after I dropped 500 kg of crude oil.  10 kg of crude oil from the bead boiled.  The same results occurred when the insulation was at 405 C.

I would like to point out that this interaction does not occur with temp shift plates.  If we use airflow tile or insulated tile on the boilers with a zigzag pattern and temp shift plates, we could avoid the boiling heat deletion bug from tiles.  However, petroleum can also cause the boiling mechanic, so we need to be careful.

[mathmanican]

2 minutes ago, Zarquan said:

the boiling mechanic, so we need to be careful.

So I need to make some ridiculously simple builds that mass exploit this to hope for some kind of fix (change a divide sign to multiply sign). Infinity stone time. Then we can stop being careful. 

(Rolling up my sleeves - let's break the game)

[wachunga]

If I am understanding the goal correctly, how about something like this?

 

The crude is cooked to petroleum by the conveyor bridge, which avoids the flash boiling bug by either a solid tile or by dripping into a pool of petroleum. A waterfall is avoided to prevent the petroleum from bubbling up the waterfall and causing havoc.

Flow is every other tick so two cooking points. Crude gets cooked into petroleum and prevents flow. On the next tick, the petroleum collapses and allows flow again.

It would mean having to make higher towers to get the room necessary however.

 

As for making the necessary heat, how about the rapidly cycling door bug? Couldn't say offhand how it compares to other heat producing bugs, but maybe worth fiddling with. Or perhaps the debris heat bug? I suspect that would be a pain to design, but would be a neat show off of the bug.