Jump to content

Smokestack: mk17

Recommended Posts

Hoo boy I have rebuilt this thing a few times, to find something somewhat compact, with working automation, and with less frame loss. This one is a little bit cleaner & more compact.

This build deletes heat by venting hot exhaust to space.

First, I need a way to generate *lots* of exhaust material. There are quite a lot of different things you could use. Any solid or liquid that boils easily will do the trick. Then, heat it up as much as possible before venting overboard.

But, I've settled on an ethanol power plant. It creates CO2 when distilling, we need ethanol for our cooler anyway, in large quantities; the generators themselves create even more CO2, and also produce a considerable amount of water. Seems up to the task.



I'm only teched up to steel so far, so working temp is just below 325C. My automation prefers to make use of CO2 exhaust that's in the room (see that 440 kg pressure) but when the room pressure gets low it switches to water, which is also coming from the power plant.Plenty of extra, in fact, I'll wind up just pumping extra water up to the surface eventually:



At thermium temperatures the most efficient possible thing to do is actually pipe crude oil from that oil well directly up into the boiler. As crude oil heats & phase changes to thermium working temps, it can reject about 40% more heat than the *equivalent* amount of water would. If this one makes your face screw up a little bit, remember that 1 kg of water pumped into an oil well creates 3.33 kg of crude oil.

But at high temps we'll be making much less exhaust anyway, so, I suspect that by the time I get thermium I won't need the 2nd source of exhaust material anymore. I.E. water, OR something better.

Why ethanol as secondary coolant?

Ethanol carries more potential energy from its melting point to working temps than PH20, about 40% more. So while a single liquid cooling loop filled with PH2O can only pipe about 2 A/Ts worth of cooling, A single pipe full of ethanol at -115C can serve 3 A/Ts.

There's a 4th spot for an A/T but there's no A/T in that 4th spot: That one is going to be for condensing LOX & LH2. Coming soon.

Build industrial equipment on metal tiles; use conveyors full of lumber to cool the power plant. Moderates the room temp, increases overall cooling. Exhaust & PH2O get generated at building temp clamped to 40C, so you really want those generators not to be much warmer than that. Lumber is coming in at 20C, so while it's waiting to get used I sink some heat with it first.




400kg or so of super coolant is enough. You'll notice my A/T bypasses are a little imperfect. The bypass path is 1 segment shorter than the cooling path, so, issues. I have to leave a 1 packet gap in each cooling loop or it binds up when the A/Ts turn on. Bother. Open to suggestions, but for now this is o.k. It works out to one missing input packet & one packet slip each time an A/T starts up.




Working temperatures



Moderating temps with wood on conveyors is working well. My petroleum generators are idling at about 33C. Sometimes under extended uptime they peak up to around 60C but that's okay.

What the heck are these vacuumed rooms with reservoirs

It's the design I settled on for distributing cold ethanol without keeping long stretches of pipe full, so, minimizing inefficiency. What I want to do is send ethanol until all the "cooling zones" (1 reservoir per zone) are full, and then stop. The idea is that the pipes stay emptied out, and I use a "full pipe" sensor to detect when I should stop sending coolant.



So for instance, between the power plant room and my main base, there's a rather long pipe for sending cold ethanol down to the base, connecting from that 2nd reservoir room to the 3rd. Room 2 sends coolant until all the "zones" in room 3 are full, then the pipes empty out. Then, room 1 sends coolant until all the "zones" in room 2 are full.

I'm not going to break this down in detail, just, think about it. Some way to transmit cooling fluid across large distances without heat leaking in is needed, otherwise you need the cooling plant to be next to base. Not ideal. (My original design had the plant right next to the main base - it wasn't as clean)


This is a state machine and it's not easy to look at. But it works okay.


Every day at sunrise I open the door on the left for just 1 second, then close it & reset the machine's state. 30 seconds later I sample gas pressures on the left using a pair of S-R latches. The latch on the top represents "high pressure" and the latch on the bottom represents "low pressure".

This is a trick for measuring very high pressures. I want the room to stay above 100 kg or so, and I want to switch back to using my alternate exhaust source if pressure goes below that. But pressure sensors can only see 20 kg, so, I sample a small amount of gas & create a gradient with it instead.

In the image below:

Red boxes are manual overrides. They're for overriding the various outputs manually while setting something like this up. For difficult designs in survival mode always include manual overrides. You might need them when it's time to send a dupe in to repair something.

Blue is my special extra-high-pressure sensing circuit

Green is the timer circuit for controlling CO2 inlet/outlet

Purple is the door control circuit (timing to close all 3 rows in sequence and then re-open the center door to create a vacuum without deleting mass)

And then the various outputs, I'll let you trace them out.


State 1 - high gas pressure - When boiler temp gets high, uses a timer cycle to first open the inlet doors, then lock the inlet doors shut for a time (allowing exhaust to absorb heat), then open the outlet doors. Repeat. When boiler temp lowers, locks the outlet doors open and the inlet doors closed.

State 2 - low gas pressure - locks the inlet doors closed & outlet doors open. When boiler temps get high, sends a green signal to a liquid shutoff sending water to the boiler.

There are some significant improvements that should be made to this control circuit. Mostly, CO2 gets exhausted before it's fully heated up to working temp. That's wasteful. But, my attempts to time things more carefully turned the control circuit into a very large state machine, with multiple parallel states that needed really careful isolation. It wasn't worth the hassle. So for now I've settled on this timer-based approach. 


Bringing it all together

Just a glimpse of what one operation in the preferred state (CO2 exhaust) looks like.




Link to comment
Share on other sites

I love all the thought that went into this. And I love that you found a use for all the CO2 the generators belch out. I'm up to about 3000 tons of CO2 in my power room, and despite 280 slicksters, the pressure just keeps growing.

However. One question for you: Steam Turbines? Granted they're boring but you also get some power back, so it's not all bad...

Link to comment
Share on other sites

13 minutes ago, biopon said:

However. One question for you: Steam Turbines? Granted they're boring but you also get some power back, so it's not all bad...

How many steam turbines & A/T loops would you need for an equivalent amount of cooling? It's a lot. Steam turbines aren't a dense cooling solution.

This contraption cools as much as a chill pill without using any temp reset glitches, & consolidates all cooling into a single secondary liquid loop. The 4 A/Ts I put up there each get to run at full capacity and they all pour work into the same central cooling loop. It's a very dense solution.

I also don't have to salvage heat that's leaked back out. (You could run a steam turbine on 1000C steam, but, good luck with that - it's a terrible idea)

But, mostly, I've been interested in this idea for a long time, and I wanted to do it. Which is why I didn't bother with an apples to apples heat rate comparison vs steam turbines. If someone felt like doing one, they should account for the amount of heat that leaks across the turbine itself, not just that gets rejected by the steam turbine as it does work.

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

  • Create New...