Gravity-driven leaky oil boiler

[Joe Dee]

I generally don't really need a lot of petroleum, just the output from a leaky oil fissure is enough to feed my rocket, and I get more geothermal power than I need anyway. This is without the DLC.

So after some more-or-less failed experiments I found a design I liked, and haven't seen this anywhere else so I thought I'd share.

It only requires a small heat input and no other powered equipment such as pumps for the boiling process, by creating a gravity-driven counter-flow with conveyor bridges (made from iron ore) for the heat exchange. The tile with the thermosensor is regulated to 403°C, it has to be offset from the edge by at least one tile so that the oil can flash into petroleum, and then the oil from the next tile can push both petroleum portions together and overflow to the boiling tile itself. The petroleum then just drips over the edge and flows the other way, cooling down and heating the oil on the way.

This setup outputs the petroleum at 374°C with 2-3% duty cycle on the aquatuner, or about 30W on average. (I haven't done the math to figure out how badly this compares to full flow boilers using a volcano.) Initially I tried various other heat sources because I didn't really want to waste 1200kg of Thermium on a low duty cycle AT, but without a convenient volcano nearby nothing else really work well enough. With a self-cooled turbine afterwards cooling it to 100°C, and further using the coolant from the AT I can run this power-positive however, cooling the petroleum down to about 10-20 °C in the end, so the AT is actually advantageous in that regard.

What I don't understand however: Initially I built it in this size because of the adjacent space exposure, but I wanted to try and get the power requirement down a bit more. Pointless optimization FTW! So I tripled the counterflow width from 18 tiles to 54 tiles. Sure enough, once a steady state was reached after a few hundred cycles that brought down the output temperature to... 352°C. Not as much as I hoped, although I didn't expect a linear relationship I surely expected more than a 75% improvement in temperature drop from tripling the counterflow width.

However, despite the drop in output temperature, the aquatuner duty cycle increased to 5%, twice as much as before! Why??? And despite all the extra width being built in the space exposure region, so hard vacuum all around and no heat loss possible.

Is this because one of those game quirks, like that heat preferably flows sideways in one direction but not the other? If I built this in the other direction, would it have even less power draw? This oil fissure is the only one I have, and it's is too close to the map edge so I can't test this right now. Eventually I'll want to find out the optimal direction and width. And maybe someone has suggestions for a better heat source, I could get neither thermo regulator nor the tepidizer to function reliably.

Anyway, I think this will be my favorite petroleum source for the time being. I just need to remember not to build lead wiring inside the insulated tiles, because they will eventually heat up enough to melt it, and it takes forever to notice because molten lead looks almost identical to petroleum...

[mathmanican]

This might be what you're looking for. 

Special thanks to @wachungafor identifying the problem.  @klei.ruby at one point was working on this and had a tentative fix, but the fix caused more problems.  I'm guessing fixing the issue has been abandoned.  The direction of flow does matter, and you can use that to gain or lose heat. 

[Joe Dee]

3 hours ago, mathmanican said:

This might be what you're looking for. 

Most likely, yes, thanks for digging it up.

I guess that means trying it right-to-left on my next map will be worth it. Might even be able to forgo the aquatuner and use some "cheaper" heat source then. Or make it smaller.

[tuxii]

For a power-inefficient reference:

On a 175.25 g/s average fissure.
Petroleum is output at ~65°C.
Pump uptime: 3%