Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Make sure the floor between the turbine and the steam room is the best insulation you have access to (most likely ceramic in your case, or capital-i Insulation if you're late-game).
Colder coolant will indeed pull heat out faster. If you find that your coolant (water) is boiling off, consider swapping to either polluted water (higher boiling point) or oil (much higher boiling point, lower conductivity).
If you're tapping a super-hot source like magma, you can use thermal sensors to limit the turbine's intake (https://oxygennotincluded.fandom.com/wiki/Steam_Turbine under the "mechanics" section and then the "usage" section: "Variable number of inlets at high temperatures"). This will reduce the amount of heat the turbine generates while still cranking out the full 850 W output. In short: if your steam is hot enough, you can use less of it, meaning less heat being pushed into your turbine.
Some general thoughts:
Use more turbines, more water, and doors on the turbines' ports to control how much heat any given turbine is dealing with. Add a separate cooling loop for your turbines. Flood the turbine chamber with high-pressure (5kg+ per tile) hydrogen for better thermal conductivity between the turbines and the cooling loop(s), and don't forget the tempshift plates. Use intermediate materials to control how much heat you're dealing with at any given time, rather than trying to cool the magma directly. Use better materials to build your insulated tiles. Use vacuum between insulated tiles to prevent uncontrolled heat transfer. Use tempshift plates to spread the heat in your steam chamber(s) so the flow is more predictable and less "spiky".
When all else fails check reddit and YouTube to see what others have done.
Use the key word "geothermal" to find magma biome taming instead of volcano tamers; a "lava blade" is a completely different beast than a gigantic pool of molten rock.
A quick Google search gave me this "simple" tamer that might be helpful to look over: https://imgur.com/gallery/OPVKyf3
As far as the turbine is concerned, 160 degrees is completely fine, and you're getting that water more than cool enough. As long as the turbine doesn't hit the boiling point of water, you don't have to care about it. You can even let it get to juuuuuust shy of the boiling point and use the not-quite-boiling water it outputs to absorb a surprising chunk of the heat, meaning you only have to delete the residual heat, but this does get finicky pretty quickly (and I often just add more aquatuners and throw more power at cooling to avoid worying about getting the build juuust right.)
Others have given advice for the turbines intake.
So as long as your able to keep your resevoir at its current temperature, you have enough cooling. Could maybe even let it heat up another 20-30 degrees and see if existing heat deletion passively ramps up enough to compensate. (a surprising number of processes in oni delete a little bit of heat and then delete more heat the warmer it gets, as long as it's still cool enough for them to actually operate.)
Given the lack of experience that implies, I feel like OP is perhaps not in a position at this point to actually succeed with this project. I intend no malice with this statement, I am merely pointing out that the physics and mechanics involved in this particular task do not lend themselves well to "winging it" without sacrificing dupes (and perhaps the entire colony), and a large portion of the difficulty is in understanding the underlying concepts and mechanics involved. If nothing else, "save early, save often" is probably a hard and fast rule in this scenario.
It might be best to wall off the area with a double row of insulated tiles and come back to this project in a few hundred cycles with more information and practice under one's belt; perhaps tame a hot steam geyser first, and/or build a volcano-powered petroleum boiler to get a firmer grasp on some of the more intricate mechanical interactions involved in a project of this size and scope.
Sandbox mode could be extremely helpful in determining a solution without endangering an otherwise solid colony... and if the colony is not otherwise completely stable, I would urge caution and restraint even more fervently. As the saying goes, "Measure twice, cut once."
https://www.youtube.com/watch?v=lLIP-Tpm7-k
Though for different reasons. The metal used in a steam turbine only matters insofar as its conductivity and heat capacity; a metal with a low heat capacity in a steam turbine could overheat faster if there's a surge of heat, but a low-conductivity turbine needs to be kept in a cooler atmosphere to help it exchange heat better. The overheat temp of a steam turbine should only ever become a problem if the volcano itself starts to leak onto the steam turbine.
But...yeah....if their water is only 160 degrees farenheight, aka more than 40 degrees below the shutoff temp, how is the steam turbine getting so hot as to overheat? It's possible to almost-run steam turbines using a mere 5 degrees of margin room, so that should be PLENTY...
It does seem strange, and likely points to them not looking at the correct temperatures for acutally diagnosing their problems, since what matters isn't the temp of the water as it gets all the way back to the resevoir, but rather the temp of the water going in and the temp of the water immediately after the steam turbine.
NB if the temperature in your base genuinely does exceed 1000C then you have bigger problems then the material used in steam turbines.
I think heat is bleeding from the magma to the turbine room. Yes cooling the water would help, or using more reactive pipe and more of it, but both of those destroy the heat you are trying to use. Better would be to build a more insulating wall between magma and turbine. Because are limited to one layer of tile below turbine, and that's likely not enough unless maybe if it's insulation, you probably don't want loose magma in the room below the turbine. Instead you want automated doors to allow entry of heat (not magma) as needed.
Given the lack of screenshot, we're all pretty much just throwing darts in a dark room, of course... but there have been several very good points made in this thread so far, so hopefully some progress can be made toward a solution at some point.
Perhaps the OP would be so kind as to share a screenshot or two, perhaps some information as to what kind of map is being played on, a rough idea of available materials, and whether or not the DLC is installed? We could make better-educated guesses as to potential solutions with a bit more background information.
-Yes, I am using Fahrenheit.
-I apologize for my sloppy language - when I said the turbine was overheating, I meant it was 212 degrees.
-This is actually my second try at this - the first attempt was made out of sandstone, melted into slag and turned my base into a sauna. Mistakes were made, lessons were learned.
-I am playing on the Volcania map. Features are Volcanoes, Frozen Core, Magma Channels and Large Glaciers.
-EDIT: I am playing vanilla.
https://steamcommunity.com/sharedfiles/filedetails/?id=2869286434
https://steamcommunity.com/sharedfiles/filedetails/?id=2869286462
In regards to the first, it's a basic layout of the pipes. As I mentioned before, I am moving water from a polluted vent, so it comes in at 85 degrees, snakes around inside gold radiant tiles, leaves at around 160 degrees, and then I ship it off to an ice biome to be turned into oxygen. As for materials being used, the insulation/ladders/non radiant pipes are all made of igneous rock to avoid melting, the turbine is made of gold.
The second map is there to give you an idea of the size of the biome - it's not very big, and the current temp ranges from around 640F (the steam in the chamber, approx density 750kg) all the way to 1300F (the rock at the very bottom of the biome). Disturbingly, this was (at the time) one of the smaller magma pools I had come in contact with, so I figured it'd be good for a trial run.
***
As for some of the comments I'm hearing, it sounds like a big mistake I'm making is not having any buffer between the steam and the hot rock, resulting in the turbine easily being overloaded. To answer a few questions on that front, my first goal is to clear the area for construction, and learn how to manage that much amount of heat. Energy is (for now) gravy, but the more I can harvest the better. Right now, the turbine is connected to a carbon skimmer (no automation).
In the future, there is a bigger magma biome that's in my way (height 75, width 52), so this is more a trial run for that. I'm thinking of using a build I saw elsewhere that uses mechanized doors to block steam if it gets too hot, so that might solve the problem at least part of the way.
If nothing else, you you should have tons of ice which can supply some cooling until it melts. At which point, you'll have plenty of extra water also....
Yeah. This smaller magma area has cooled down some, but you don't want to do without some heat management with a bigger magma sea, such as if you had a normal one instead of Frozen Core.
One design for managing heat is to have an Airlock Door connected to a thermo sensor in the Steam room. Surround it with metal tiles on either side so that it creates a vacuum when it opens.
Igneous rock is OK, but Ceramic would be better when dealing with extreme heat.
The aquatuner used in a cooling loop will get hot because it receives all the heat removed from the steam turbine, so it also needs to be cooled. Usually you will solve this problem by making the aquatuner out of steel so can operate up to 275C and placing it in the steam chamber which is already being cooled by the steam turbine's cold water output.
If your current steam chamber is very hot because its being heated directly by magma and you can't bring the temperature somewhat below 275C, then the aquatuner cant go there or it will overheat. You will have to build a second insulated steam chamber where the aquatuner will be submerged in water that eventually all turns to steam. The aquatuner cooling loop will be used to cool off both of the steam turbine rooms. The aquatuner will be submerged in water so that as it heats up it will create steam until the whole room is full of steam then the second steam turbine will cool that steam and recycle its cool water back into the second steam chamber, keeping the temperature from rising too high. Since you have no magma in the second steam chamber you should be able to regulate its heat with automation controls to keep it below 275C.