The Bead Flaker - A Crude To Petro Boiler Utilizing Partial Evaporation Mechanics

[mathmanican]

This is a follow up post (for a history, see here) from last summer's discussion about the gas heat deletion bug. The topic of crude to petro boilers seems to gather attention regularly.  This goal to this thread is to give a simple reference point for those interested in pursuing flaking mechanics (thanks for making a wiki page @Fradow). Here are two versions of the boiler, one attached to a staircase heat exchanger, and the other attached to a bead heat exchanger, introduced by @wachunga, with a blob of naphtha below the petro to form the beads.

   

Why would I build this? What advantages does it have?

When you use partial evaporation (flaking) to transform crude to petro, you get three gains that help conserve your heat source (simulated by the steam room in the picture above).  

1. The exact amount of heat needed to raise the crude oil to 3C above it's vaporization point is used. No more, no less, so you get perfect boiling with no overshooting. 
2. You only loose a single tick of conduction worth of heat between the igneous rock and outgoing petroleum.  So your heat bleed can be minimized by quite a bit.  You can keep your boiler at 500C, and you still only loose 1 tick of conduction between an igneous rock tile and a 5kg blob of petroleum.  With my room set at 430C, the petro measures a steady 402.9C at the top of the bead boiler (so basically non visible heat bleed to conduction).  
3. The phase transition does not suffer from a 1.5C loss of heat, which regular boilers all suffer from (amplified by the fact that this 1.5C occurs for each contact surface that participated in the exhange, so a possible total loss of 6C with a poorly designed boiler). 

How does flaking work?

Let's examine the requirements to enable flaking.

1. The piping requires valves to be set at exactly 5010g. Any less, and flaking won't occur. Any higher, and you may end up with unwanted crude oil clogging the works. Upon flaking, the 5010g of crude turns to 5000g of petro, and several blobs of crude with size less than 10g. These tiny blobs disappear in the next tick. Here is an example piping overlay (it could be improved). 
   
   Note that I have a valve set at 5010g feeding crude into the heat exchanger as well, so that the liquid reservoir maintains a constant volume. An improvement wiould be a shutoff attached to the automation port on the liquid reservoir, or a door underneath it, to make sure you always have liquid in the reservoir. You must have at least 5010g flow past the valve, so take care to ensure you never have a single packet smaller than this.
2. The incoming temperature of the crude oil must be large enough to prevent the flaking point from dropping more than 10C upon flaking. With an igneous rock tile kept near 425C (the tempshift plate along with conveyor bridge transfer heat into the tile extremely fast), then the oil temp needs to be above around 170C (make a copy of this spreadsheet to perform your own computations). 
   
3. The requirement of 170C incoming crude poses a problem only when staring the boiler. Once the process begins, even poorly designed heat exchangers should keep incoming crude well above 300C. Start by filling a liquid reservoir with some hot crude oil, for example a few runs in a metal refinery is more than enough. Make sure the reservoir never runs dry and you can start and stop this boiler without hiccups (unless you decide to toggle it on/off rapidly).  You have to let it run long enough to recover from any initial dip in temperature as the first batch of oil makes its way through and cools off the heat exchanger.

OK, but you don't have a boiler, just a steam chamber! Cheater!#@&

The steam chamber in the middle can be any heat source you want.  You don't have to have both boilers (of course), but for showing off the design, that was the easiest. To enable flaking, just ensure that the incoming crude oil is hot enough to prevent the igneous rock from trying to drop 10C upon flaking. So set a thermosensor next to the igneous rock to 430C (or 500 if you want), and it just works.  The higher you set your thermostat, the more heat you bleed into your outgoing petro, but it literally is only a single tick of conduction, once every second, between a 200kg tile of igneous and 5kg of petro.  Even with the room set at 500C, the heat bleed is pretty negligible. 

What heat exchanger should I use after this?

It doesn't matter. This thread isn't about heat exchangers. Pick your favorite. I chose these two heat exchanger designs based off the recent tests done here and on Discord by several great folks (such as @Fradow, @Saturnus, @wachunga, @nakomaru). Left-to-right staircase designs are affected by a current bug. The bead design can be mirrored to form a waterfall or bead heat exchanger, with slight modifications (see either here or here for examples). Use a @JohnFrancis design, grab one from @Tonyroid, or make your own messy contraption.

The bead flaker is a boiler that will transform crude to petro very efficiently, nothing more, nothing less.

With a 50 segment radiant pipe heat exchanger, here are two examples. 

The only input is 95C crude going in at 5010g/sec. After reaching equilibrium, the temp inside both liquid reservoirs stays at 387.8C (with less than 0.1C variation). The reservoir instantly equalizes any minor fluctuations (the Z design is not needed for stability).The outgoing petro pools at 121.2C (for both designs). 

OK, but I need a 10kg/s boiler.  This 5010g thing is garbage. 

You can easily modify this to meet your needs.  Here is an example. where I spread out the piping to make it easier to see what is going on. 

The piping on the left is a simple loop that guarantees only 10kg packets flow past the top valve. Every second this 10kg/s boiler runs, you loose 20g of crude. So the shutoff valve turns OFF for 2 seconds every 500 seconds, to account for the lost oil.  There is a pipe element sensor set to detect crude oil, based off @Saturnus's inline packet stacker, which guarantees that only 10kg packets ever flow past the top valve. This isn't quite a 10kg/s boiler, but if we round to 3 significant digits, then it is. 

Edit: Warning, because of how saves work, the tick sequence of this particular boiler can become desynchronized upon load (resulting in the crude not flaking).  Use this particular version at your own risk.  To fix this, I would just have two different contact plates for each 5010 g, not much more) and then make sure any newly formed petro drips down, rather than beads down. 

You're telling me that I'm going to loose 20g/s of crude oil if use this!  #$%!%&*%*#%^

  

You understood perfectly.  Yep.  That is the cost. 

You will lose 0.2% of your possible petroleum to gain efficiencies from your boiler.  You can conserve magma for much longer.  Of course, once you have thermium, then who cares. Power is cheap and who cares if you loose a tad bit of heat energy to a subpar boiler. 

Have fun all. 

 

Oh, for those who want a save file, here you go. 

Happy Gang Petro Cycle 39.sav

[Hokaeru]

This is way over my head but somehow I feel that I have to build it.

[mathmanican]

1 minute ago, Hokaeru said:

somehow I feel that I have to build it.

My hope is that this inspires people to learn more mechanics of the game (beads, flaking, waterfalls, heat exchanges, boilders, etc.)  It's definitely not a beginners guide, but you can load the save file, play with it, and see what makes it work and makes it break.  

[Saturnus]

A very practical implementation to get full 10kg/s output would be to loop the secondary vent until the reservoir is filled at which point the 20g/s is effectively drained very very slowly from the reservoir. A reservoir takes 1.67 cycles to fill with 4990g/s but takes 416.67 cycles to drain fully at 20g/s. So basically you'd have it running at 5kg/s output while starting up and temperatures settle, and then run full tilt 10kg/s for 400 cycles after.

[mathmanican]

13 minutes ago, Saturnus said:

very practical implementation to get full 10kg/s output

Yep.  That works as well. Another is to just top off the liquid from an external source that doesn't run up the line, either fed into the reservoir, or directly into the line. I got the idea of topping it off with with a single pump from @nakomaru, and I like the simplicity of a single pump (but maybe the addition of an in-line packet stacker is worse for complexity).  Oh well, if it gets people interested in alternate ideas, then it's a win.  

[0xFADE]

Dripping it for exact heat is super nice.  Though I don't quite like the stair style for the rest.  The flat ones are more compact.  Compact is nice.

[wachunga]

An alternate solution to the startup problem of needing somewhat warm crude is to make a more massive donor tile. Using an insulated tile works but is problematic with the low conductivity. But how about cooking debris into a solid tile? I totally forgot we can do that.

500kg of coal cooked into a refined carbon tile has enough thermal mass to flake 0C crude. Dirt into sand and clay into ceramic work as well but I think coal is the easiest. Only takes 1-2 cycles to heat that amount of coal debris, depending on your method. Placing the debris in steam (convenient if a steam chamber is already part of your design) or a bit of crude are both good and simple enough to be reasonable IMO.

Also is the option of saving a natural tile of sufficient mass or using a dispenser to increase the mass of a natural tile. Interestingly it is possible to load a dupe built tile with more mass using a dispenser, but it doesn't survive a reload and seems to require a natural tile above the built tile. Pretty pointless since you could have used the natural tile instead.

[0xFADE]

Interesting.  I was looking at the build to see the materials and a bit of crude oil made it through the dripper while running at normal speed.  I guess it is a little more delicate than I thought.

[babba]

On 5/22/2021 at 8:22 PM, Hokaeru said:

This is way over my head but somehow I feel that I have to build it.

Its the mathmanican addiction...its...addictive. Someday I will prepare special meme !

[mathmanican]

1 hour ago, 0xFADE said:

a bit of crude oil made it through the dripper

There are many reasons this could happen. I'll provide an update soon with various ways to deal with it. I'm hoping to increase the flaking speed to 25kg/s.

1 hour ago, wachunga said:

But how about cooking debris

You made me think of @Zarquan. I like this option as it eliminates the liquid reservoir completely. It does add another fun skill hurdle of baking your own natural tiles.

[mathmanican]

3 hours ago, 0xFADE said:

a bit of crude oil made it through

Something as simple as the following will catch any overflow. The one tile vertical gap between airflow tiles is there to force any beads to start dripping. This way you don't get a bead of crude oil swapping the order of all the materials in the materials as it descends. 

See here for some other options. I decided to add a bit of naphtha, thanks to @nakomaru, precisely to provide a thermal buffer between the 402.9C petro and the crude (which never gets that hot).  

My attempt to get 25kg/s flaking here with one tile failed miserably. 

[Zarquan]

On 5/26/2021 at 2:01 PM, mathmanican said:

Something as simple as the following will catch any overflow. The one tile vertical gap between airflow tiles is there to force any beads to start dripping. This way you don't get a bead of crude oil swapping the order of all the materials in the materials as it descends. 

See here for some other options. I decided to add a bit of naphtha, thanks to @nakomaru, precisely to provide a thermal buffer between the 402.9C petro and the crude (which never gets that hot).  

My attempt to get 25kg/s flaking here with one tile failed miserably. 

Out of curiosity, is there a particular benefit to flaking beads here?  I would think one of the liquid flaking machines we created back when it was buggy would get a higher throughput.  Also, I thought flaking beads in a confined space like this deleted a lot of mass arbitrarily.  Did this get fixed?

On 5/26/2021 at 10:08 AM, wachunga said:

An alternate solution to the startup problem of needing somewhat warm crude is to make a more massive donor tile. Using an insulated tile works but is problematic with the low conductivity. But how about cooking debris into a solid tile? I totally forgot we can do that.

500kg of coal cooked into a refined carbon tile has enough thermal mass to flake 0C crude. Dirt into sand and clay into ceramic work as well but I think coal is the easiest. Only takes 1-2 cycles to heat that amount of coal debris, depending on your method. Placing the debris in steam (convenient if a steam chamber is already part of your design) or a bit of crude are both good and simple enough to be reasonable IMO.

Also is the option of saving a natural tile of sufficient mass or using a dispenser to increase the mass of a natural tile. Interestingly it is possible to load a dupe built tile with more mass using a dispenser, but it doesn't survive a reload and seems to require a natural tile above the built tile. Pretty pointless since you could have used the natural tile instead.

I like liquids or gas for my donor tile.  You can get tiles of arbitrary mass using infinite compression.  If you use a gas, you can put it in an airflow tile and be immune to pressure damage at that point.  Alternatively, you can infinite compress a liquid metal to get a higher conductivity, but I believe we generally found that lower conductivities are better as long as we can transfer heat in to the tile fast enough, as it reduces the amount of heat added to the petroleum.

[mathmanican]

5 hours ago, Zarquan said:

Out of curiosity, is there a particular benefit to flaking beads here?

Flaking crude to petro is unique (the only liquid to liquid transformation). This is just a follow up post that brings to the front the bead flaker (designed last year).

When flaking any other liquid where you get gas output, then other designs are much better that don't use beads. It's the liquid to liquid issue that makes beads useful.

[JRup]

40 minutes ago, mathmanican said:

Flaking crude to petro is unique (the only liquid to liquid transformation).

This is notable because we're talking about a transition to higher temperatures in mind when talking about flaking... If we gave ourselves the leeway to say any kind of transition is valid then we'd be able to include salt water to brine in the list of liquid to liquid transformations (+ ice as byproduct). Just sayin'

[Zarquan]

14 hours ago, mathmanican said:

Flaking crude to petro is unique (the only liquid to liquid transformation). This is just a follow up post that brings to the front the bead flaker (designed last year).

When flaking any other liquid where you get gas output, then other designs are much better that don't use beads. It's the liquid to liquid issue that makes beads useful.

But what advantages does the bead design have?    Because I created a compact lossless 25 kg/s petroleum boiler using flaking back in the day using the design we used for gasses.   The main benefit I see of the bead design is being able to turn it off easily, but are there others?

[mathmanican]

13 minutes ago, Zarquan said:

are there others?

Thermal isolation. Minimal heat bleed.

[Zarquan]

26 minutes ago, mathmanican said:

Thermal isolation. Minimal heat bleed.

Ah, so the boiler is only in contact with the petroleum for 1 tick.  That makes sense. 

[wachunga]

I started up a new game just to try out this style of boiler. Working well so far with no issues. We'll see how it goes after a few hundred cycles. I think I'll dejankify the piping at some point as I changed my mind partway through and am still pondering a final layout. I also took the opportunity to go old school with diamond on rails for heat control because doors have been done to death and are boring. The rails are also good for transporting heat from an inconvenient source location to a preferred boiler/exchanger location.

 

Spoiler

Attention to detail is certainly necessary but the build is very realistic. Not anything as crazy as I initially thought when @mathmanicantalked about it way back when. Crude is quite cold on Rime so I cooked 600kg coal into refined carbon for the donor cell. Cooking the donor also allows placing a thermo sensor where the refined carbon will be, which is handy. I went with piping flow that advances 10kg in discrete steps as opposed to a continuous 5.01kg (with full pipes). The heat exchanger performs a bit better that way. Partially full pipes of 5.01kg is the best, but is ruled out by the potential of flow stopping from the shutoff at the very end. And speaking of exchangers, with aluminum you have to be careful as the exchanger can get so efficient that the crude is too hot to flake. Need to fiddle some more, including stopping and starting, to see where the sweet spot is. Math says the copper volcano is more than sufficient for the 5kg/s boiling and the turbine will be needed eventually. Or if I turn the boiler off.

[W.J.]

This build is simply brilliant! I think I haven't understand it completely since it's bleeding bit of crude oil sometimes but still, constant prox 402.5C of petroleum is BEAUTIFUL. I did added some failsafe that boils extra crude directly through conveyor bridge cause I just couldn't figure out why it's bleeding crude (I hate calculating stuffs and not that familiar with all niche bugs so I'm just crossing fingers here it's not bleeding lots of heat off) but it's working for nearly for 400 cycles without any complications. Much easier to build, can turn off, and no unexpected temperature spikes. Love it.

 

Thermo sensor behind refined carbon block is set at 450C and temperature fluctuates abit between 480C~433C but it starts dripping crude around 460C or so sometimes. I wonder why

[Prince Mandor]

2 hours ago, W.J. said:

 

Thermo sensor behind refined carbon block is set at 450C and temperature fluctuates abit between 480C~433C but it starts dripping crude around 460C or so sometimes. I wonder why

How large carbon block is? possibly you need much larger block to keep temperature less volatile

 

Another question. What is the meaning of two horizontal zig-zag shelves? Why not start waterfall nearer to boiling zone?

[mathmanican]

9 hours ago, W.J. said:

why it's bleeding crude

I'd happily look into this one, if you want to provide a save. There are several reasons that could be the cause.  The most likely is that you have a packet that is not exactly 5010g getting through.  I have found that green outputs, when withing 1 tile of a white input, mess up liquid valves.  I like to make sure I have a full 10kg pipe segment flow past my valve, and I always design for the excess to pass by it to an overflow area.  This prevents piping bugs from hitting me. This also makes sure that I can start/stop things without issue.  Do your valves have an overflow planned in?  Or does the bleeding crude happen when you stop the flow into one of the valves because you opened the door below one of the liquid tanks (as you can't have 10020g  flowing in with just a single pump).  

I see the close proximity between the output of the left liquid reservoir, and the input of the liquid valve, to be the most likely cause.  It's an annoying "feature" of valves. 

[W.J.]

On 6/17/2021 at 8:01 PM, Prince Mandor said:

How large carbon block is? possibly you need much larger block to keep temperature less volatile

Another question. What is the meaning of two horizontal zig-zag shelves? Why not start waterfall nearer to boiling zone?

It's 800kg, Just slapped coal tempshift plate there. Zig-zag is actually a design failure... It should start where bead is but it kept wrecked by crude so had to improvise.

On 6/18/2021 at 3:13 AM, mathmanican said:

I'd happily look into this one, if you want to provide a save.

I'd be glad to! Thanks for the input.

 

I don't have a overflow (guess i'll make it now) but certainly isn't about the flow cause it happens when door is shut. Actually it didn't happen when only one vent is operating, so i was suspecting it's somewhat related to that. 

 

[mathmanican]

14 hours ago, W.J. said:

...

I was unable to test things. With all mods uninstalled, here is what I see when i load your game. 

So some blocks don't even appear, your liquid lava is all gone, and I can't test anything.  It may be mod related, or might be something else.

Sorry I can't help more.  If you only have a single 10kg pump supplying crude, then running both vents will definitely cause problems, unless you find a way to make sure the other line is always topped up to at least 5010g. One pump cannot supply 10020g, and so you will have crude spill in and mess stuff up. 

[sakura_sk]

21 minutes ago, mathmanican said:

It may be mod related

Ehmmmm... I don't think it is mod (only) related 

 

[mathmanican]

Haha. It's probably my slow potato computer.