(Help Needed) HUGE Heat Creation/Deletion Exploit with Conduction Panels

[Nova Starlight]

I need help creating a contraption that massively exploits this new bug that I have found, preferably to the point of requiring a hotfix from Klei. I'm pretty sure that this is novel, too.

Here it is. It is very simple. When two or more liquids are in the Conduction Panel the Panel's temperature only reflects the heat exchanged with the last liquid. This contraption takes in 5kg/s 1000C Molten Steel (all the radiant pipes do is cool the dev pump steel to that temperature) and spits it out @2550C. Works with any basically any liquid so you can also massively cool down to cryogenic temperatures using this same bug. The power shutoffs just throttle the heatant so that it has time to gain heat from the panel. Panel is thermium and near melting.

My dumbass really thought that Steel would be a good heatant because of its high temperature range and highest Thermal Conductivity of all liquids. Turns out for steam turbines Super Coolant/Nuclear Waste would theoretically be better by a factor of around 5 to 8, producing (at best) 15,000kDTU/s per conduction panel (about 15kW if put through turbines). This reflects my inexperience with the heat system and steam turbines. That's why I am asking YOU, the best players on these forums, @mathmanican@wachunga@Zarquan etc (sorry if multi tagging is not up to etiquette, I really don't know) to help invent the craziest contraption one could feasibly think of using this very bug. Something that could evaporate the whole magma planetoid into rock gas, perhaps?

World download is attached.Temperature Experiments 2.sav

[Zarquan]

On 5/24/2023 at 4:46 AM, Nova Starlight said:

Spoiler

I need help creating a contraption that massively exploits this new bug that I have found, preferably to the point of requiring a hotfix from Klei. I'm pretty sure that this is novel, too.

Here it is. It is very simple. When two or more liquids are in the Conduction Panel the Panel's temperature only reflects the heat exchanged with the last liquid. This contraption takes in 5kg/s 1000C Molten Steel (all the radiant pipes do is cool the dev pump steel to that temperature) and spits it out @2550C. Works with any basically any liquid so you can also massively cool down to cryogenic temperatures using this same bug. The power shutoffs just throttle the heatant so that it has time to gain heat from the panel. Panel is thermium and near melting.

My dumbass really thought that Steel would be a good heatant because of its high temperature range and highest Thermal Conductivity of all liquids. Turns out for steam turbines Super Coolant/Nuclear Waste would theoretically be better by a factor of around 5 to 8, producing (at best) 15,000kDTU/s per conduction panel (about 15kW if put through turbines). This reflects my inexperience with the heat system and steam turbines. That's why I am asking YOU, the best players on these forums, @mathmanican@wachunga@Zarquan etc (sorry if multi tagging is not up to etiquette, I really don't know) to help invent the craziest contraption one could feasibly think of using this very bug. Something that could evaporate the whole magma planetoid into rock gas, perhaps?

World download is attached.Temperature Experiments 2.sav

TL;DR:  With uranium, this bug can create enough heat to run 9.25 steam turbines for free.

Good find!  I made a modification that doesn't use automation.  Using this mechanism, I can control the output using a liquid valve (10 - 3 kg/s = 7 kg/s output in picture).  As far as I can tell, when set to 5 kg/s, it behaves approximately the same as your original design thermally, but does not rely on automation and is easier to define the flow.  The idea behind the valve is similar to the idea behind your timers, by increasing the mass of liquid running through the valve, it limits the flow of liquid out of the conduction plate, which liquid to stay in to be heated, then averaged out when a new packet enters.

I suspect the optimal flow rate for maximum heat generation would change based on the incoming temperature and the temperature of the plate.

I spent a little time experimenting with this machine.

I have determined that with an input of 186.8 C uranium and a preset of 2676.8 C using a thermium conduction plate and 0.1 g water in the conduction plate, 7 kg/s produces the most power out of the multiples of 1000 g/s without any temperature loss to the plate, with an output uranium temperature of 857.7 C, resulting in a free heat energy creation of 9392600 (math error) 8114197 J/s (or DTU/s).  That is enough to run 9.25 steam turbines full tilt for free for each conduction plate system built, which is honestly obscene.

I have discovered a few useful things about this machine:

1. If you select the conduction plate, you will see a lot of temperatures flying around on the uranium.  When you see 10 kg uranium, the next tick shows you the eventual average temperature of this system (assuming constant input temperature).
2. As expected, higher masses of liquid leaving are, the lower the output temperature is, meaning if your application requires higher temperatures, you can turn down the system to get the higher temperatures, but you create less thermal energy total due to the lower thermal mass that has been heated.
3. Buildings behind the conduction plate appear to have no effect on the results (not surprising).
4. Tiles on top of the conduction plate appear to have no effect on the results (mildly surprising to me, but I haven't really cared to look at thee behavior of these before).

To build this, first preheat the conduction plate to 2676 C, build the right side pipes and valves of the system, populate the pipes with your liquid of choice, then add the water (can be added using a liquid valve set to 0.1 g/s).  Then add your liquid to the left input of the conduction plate and set the liquid valve to ten minus the target output.  Other than the preheating, this should be very easy to accomplish.

Things I want to test:

1. How does the thermal conductivity of the plate and liquid affect this?  Theoretically, super coolant from 150 C to 436 C at 2.6 kg/s contains as much heat energy, so with a higher flow, we could get more turbines, but that would be finicky to set up.

I have a few machines in mind that I want to try out:

1. Liquid hydrogen machine:  I think it won't be great at this due to the smaller deltas
2. Rust/Regolith melter:  As heat energy intensive processes, it would be interesting to see the production per second.

This could easily power a sour gas boiler or a petroleum boiler as well, but it would behave the same as any other generic heat source, but be enough to easily boil how ever much you want.

I will say, I've never considered vaporizing the magma on the magma planet as an option to get in to it, so that could be funny.

[Nova Starlight]

4 hours ago, Zarquan said:

TL;DR:  With uranium, this bug can create enough heat to run 10.7 steam turbines for free.

Good find!  I made a modification that doesn't use automation.  Using this mechanism, I can control the output using a liquid valve (10 - 3 kg/s = 7 kg/s output in picture).  As far as I can tell, when set to 5 kg/s, it behaves approximately the same as your original design thermally, but does not rely on automation and is easier to define the flow.  The idea behind the valve is similar to the idea behind your timers, by increasing the mass of liquid running through the valve, it limits the flow of liquid out of the conduction plate, which liquid to stay in to be heated, then averaged out when a new packet enters.

I suspect the optimal flow rate for maximum heat generation would change based on the incoming temperature and the temperature of the plate.

Nice! Although I was expecting you could make the steam turbine setup too. Last time I used turbines in survival I tried to use SCST to process 660C+ steam for power generation, 5 inlets too. Needless to say, that blew up in my face pretty fast.

Quote

Things I want to test:

1. How does the thermal conductivity of the plate and liquid affect this?  Theoretically, super coolant from 150 C to 436 C at 2.6 kg/s contains as much heat energy, so with a higher flow, we could get more turbines, but that would be finicky to set up.

Heat transfer with the content follows the same equation as pipe content and normal/radiant pipe. So q = kave * TempDelta * 50 * 0.2 per tick. Using this is math intensive though (there will be exponentials and such), I'd prefer just to experiment. But I'm lazy though so don't take this too seriously.

Quote

I have a few machines in mind that I want to try out:

1. Liquid hydrogen machine:  I think it won't be great at this due to the smaller deltas
2. Rust/Regolith melter:  As heat energy intensive processes, it would be interesting to see the production per second.

Don't forget sand melters too. They need some love.

Quote

This could easily power a sour gas boiler or a petroleum boiler as well, but it would behave the same as any other generic heat source, but be enough to easily boil how ever much you want.

Oh yeah, this completely trivalizes sour gas boilers. Would be good to make a super simple super compact pre steel one (no heat exchanger!) just to show off how broken this is.

[Zarquan]

4 hours ago, Nova Starlight said:

Nice! Although I was expecting you could make the steam turbine setup too. Last time I used turbines in survival I tried to use SCST to process 660C+ steam for power generation, 5 inlets too. Needless to say, that blew up in my face pretty fast.

And here it is:

10 steam turbines at almost full bore and two partially running turbines.  This cools the uranium from 820.8 C to 134.8 C.  It appears to produce an average of 7733 W (subtracting the aquatuners cost), which is about 9.1 full steam turbines worth.  It is possible there is heat deletion in the setup or the system hasn't stabilized or that a multi-cycle average would be needed to see the expected 9.25.

And, in case anyone is just looking at this here, this heat and energy comes from the conduction panel without draining temperature from anything and it appears that it will run forever.

The molten steel on the left is not involved.

[Nova Starlight]

https://forums.kleientertainment.com/forums/topic/149687-prototype-exploit-based-sour-gas-boiler-constructive-criticism-needed/

[Nova Starlight]

Update: The new mechanics in the test branch fixed this particular exploit. However! The conduction panel is still broken. I'll update with a bug report soon enough.

EDIT: