CoolRod - An Early Game Cooled Rodriguez SPOM

[Supraluminal]

10 hours ago, psusi said:

14 hours ago, biopon said:

If you want to be a bit fancier: use a thermo sensor in the steam so your turbine only turns on at 200C to maximize efficiency. If you do this, you'll probably have to find a way to cool the turbine itself. If instead you let it auto-regulate at 125C, it can probably self-cool with its own output.

I just asked about this the other day in another thread and it seems that since they revamped the turbine, this no longer provides any benefit.

In fact I think you want to do more or less the opposite - cut off the flow of hot coolant to the steam chamber if the steam temp exceeds 200C.  My understanding from the wiki is that the turbine will consume steam at a steady 2kg/s and delete 90% of the heat it contains, with the output water pegged at 95C. It generates power at a fixed ratio to the heat deleted (about 1J per kDTU), up to its max output of 850W. Any heat above that is "wasted," i.e. deleted without increasing the amount of power produced.

If you're mainly using the turbine as a cooler that's fine, but if you want to maximize the power you're generating it's something to watch out for.

(Running the turbine at lower temps is fine BTW, since the ratio of power produced to heat deleted is not dependent on the temperature of the steam.)

[psusi]

So it took me about 50 damn cycles to build a slightly scaled down version of this beast for real.  I only used 3 electrolyzers and meant to put in 6 oxygen pumps for 3 full vent lines coming out, but only put in 5. Oh well. Unfortunately, gold amalgam isn't working for the aquatuner.  It just heats up too fast to dissipate it into the steam room before it overheats.  Even though I plastered the area with tempshift plates.  I think what I should have done is set up a timer circuit to limit the AT so that it can only run for every other second. With 1 second on and 1 second off, it should be able to spread the heat during its down time and avoid overheating.  Unfortunately, to change that now would require tearing the thing down again and starting almost from the beginning.

Oh, but the heat exchanger works just fine using a checkerboard of dirt and igneous rock tempshift plates instead of metal tiles or polluted water.

13 hours ago, Supraluminal said:

If you're mainly using the turbine as a cooler that's fine, but if you want to maximize the power you're generating it's something to watch out for.

Agreed.  If the turbine is hitting its max power then you want to limit the coolant going into it to maximize power, but if the goal is just to get rid of heat and get some power in the process, then you're better off not limiting it and just keep refining metal faster.

 

[psusi]

So when I put in the recirculating loop to detect back pressure and stop the oxygen pumps, it seems I found a major bug with pipes because the oxygen started flowing for 1 second, then taking 4-5 seconds off before flowing for another second.  Reported in the bug tracker.  Nevermind... I'm a derp... it just refuses to flow past that first vent because it sees both a source and a sink downstream.  I added another bridge and it fixed it.  It just happened that the first vent was letting some gas out every few seconds.

 

[Tobruk]

Early game xD Yeah, with aquatuners.

[FSharpDotNet]

I've had a lot of issues with this when either of the outputs back up the other can easily start getting the wrong element. I adjusted the gas pipes to include filtering. On the brighter side I think it was powering itself and a desalinator/sieve combo without (or with very little support). In my case I turned the tuner down a bunch since most of the oxygen just goes to electrolyzers (and had one of the three outputs skip the cooling block entirely). It's still painful to build as most things that require a vacuum (or multiple). But less of a pain than trying to get exactly 2000g of liquid onto tile(s).

[psusi]

44 minutes ago, FSharpDotNet said:

I've had a lot of issues with this when either of the outputs back up the other can easily start getting the wrong element. I adjusted the gas pipes to include filtering.

Did you forgot the pressure sensors to control the pumps?  Once the system is primed, you should have a constant layer of hydrogen at the top flowing over the one tile gap into where the hydrogen gas pump is, and it should only ever get hydrogen ( again, after an initial startup period for the gasses to stabilize ) and you certainly should never get hydrogen in the oxygen pumps.

[FSharpDotNet]

This happened after priming, and yes the pressure sensors were in. I was fairly certain that the pressure settings are too low to prevent gasses from unlayering each other. I've just reproduced it in sandbox, I ran it for a few cycles, blocking one, then releasing, releasing both, then just letting the o2 pumps run. I thought for awhile it must have been me, because it never looked out of place, then I stopped paying attention for awhile, and it finally happened. The whole thing was only running for ~4 cycles.

 

[psusi]

Yea, you need to get the pressure settings right for it to remain stable.  Try 750 on the hydrogen and 850 on the oxygen.

[psusi]

So I was thinking; rather than have a closed cooling loop from the aquatuner to the oxygen heat exchanger, why not have the output of the aquatuner split in two with valves, and feed into a pair of liquid reservoirs.  Their output is each split by valves and half from each is recombined with the half from the other.  You now have one full pipe that can feed back to the aquatuner, and one that can be sent downstream.  Downstream means a trunk pipe that first goes to the oxygen heat exchanger via a shutoff, but once that is cool enough ( it doesn't need to be 0 C, 20 or so seems just fine for sending out < 25 C oxygen ), and when that shutoff closes because the heat exchanger is cool enough, the flow can be redirected to other areas of your base that need cooling, each on their own local loop with a sensor/shutoff so each only needs to be cooled to the appropriate temperature.  Eventually the coolant loop returns to to the reservoirs where it is first divided in half and combined with the half pipe each is getting fed from the aquatuner so each reservoir always has a full pipe coming in and a full pipe flowing out.

 

This means that even though the output of the aquatuner is cycling between 0 and -14 C, the flows coming out and cooling everything else will maintain a nice constant 0 or -1 C temperature.  And you get good temperature regulation for everything, not just the oxygen.  What do you think?

[Prince Mandor]

9 hours ago, psusi said:

 

As I can imagine, you mix multiple things here. What's the purpose of so many reservoirs and valves?

First, a big loop. It is central cooling. It works well, but needs a ton of material for insulated pipe running through entire base. And you need two temperature sensors on small radiator loops – one in room to check for required temperature, and one on pipe checking for liquid in main loop really colder than room (if liquid go from previous room it may be too hot already). 

This system is mega-project, needs lot of time to build, careful routing of many pipes etc. And in normal circumstances, after initial equalizyng of temperatures, it isn't much better than just cold oxygen. Of course if you, for some misteryous purpose, want rooms of different temperature, so one room to be at 24° and next one at 17°, this is a good way to achieve it.

 

Another thing is rerouting part of water back to AquaTuner. Really, you don't even need reservoirs, just set valve at, for example, 5kg and reroute it back to AT input. This is a way to create colder result out of same AT. Instead of 10kg/s dropped 14°, you will have 5kg/s dropped 28°. But overall cooling remains same

[Supraluminal]

R1...AT...R2

R1 = reservoir before AT that receives returning liquid from cooling loop; R2 = reservoir after AT; AT = you know what it is

To get a precise temp out of R2, put a temp sensor directly after the R2 output. If the flow is too warm, route liquid from R1 -> AT -> R2. Otherwise cut out the AT and pass liquid directly from R1 -> R2. (R1 doesn't even have to be a reservoir necessarily, it could just be the return end of your cooling loop.)

You can add more temp sensors in various spots for failsafe operation around some edge cases, but in general this is a super-simple way to fine-tune the output temp from an AT. Maybe I'm missing something, but I don't think the complicated mixing scheme gets you any other benefits.

[psusi]

7 hours ago, Prince Mandor said:

As I can imagine, you mix multiple things here. What's the purpose of so many reservoirs and valves?

Like I said in the second paragraph; to maintain constant output temperature instead of swinging across a 14 degree range between when the AT runs and when it doesn't.  But maybe that isn't really a big deal.

7 hours ago, Prince Mandor said:

And in normal circumstances, after initial equalizyng of temperatures, it isn't much better than just cold oxygen

I was trying to only cool the oxygen down to 25 C since it is going to be deleted by dupes I didn't really want to waste cooling it lower.  For most of the colony that should be fine, but then my desalinator started radiating a lot of heat and I don't think even if the oxygen being dumped into the room were 10 C it would be enough to keep that thing cool.  As it was, all of the oxygen passing by in the duct was being warmed to 35-40 and taking that heat to the rest of the colony.  Likewise, when I build an ethanol plant, I'm almost certain it can't be cooled by cold oxygen; it's going to require a cold liquid loop.

42 minutes ago, Supraluminal said:

To get a precise temp out of R2, put a temp sensor directly after the R2 output. If the flow is too warm, route liquid from R1 -> AT -> R2. Otherwise cut out the AT and pass liquid directly from R1 -> R2. (R1 doesn't even have to be a reservoir necessarily, it could just be the return end of your cooling loop.)

That was my initial thought but then whenever the coolant is warm and needs sent back to the AT, you get no output to circulate around the colony.  I came up with the two split reservoirs so that you can always have constant output whether or not the AT is running.

[FSharpDotNet]

gold aquatuner can up like 25-35c past overheat point if it sits in some crude or petrol.

[psusi]

So after learning some lessons with my first attempt at this, I started the world over and explored the map in the other direction.  I found plenty of polluted water in a slime biome and a polluted water vent I haven't broke open yet to replenish it.  I built this smaller version of the spom using that instead of the hot salt water vent.  I used logic gates to alternate between which of the two aquatuners in the steam turbine is allowed to run at a time so they can both be fed by a single conductive wire battery flipper sub station, and with 1 second on, 1 second off, that should ensure that the gold AT never overheats.  For good measure I threw in some crude oil this time.  Also I did not bother with the atmo sensor for the oxygen pumps.  Instead I'm just going to use a valve to restrict the maximum flow on each gas pipe to 850 g/s, which should be a little below the maximum the two electrolyzers can produce.  I haven't even hooked up the second gas pipe or the second AT yet, but it's ready.  And I even left pipes in there to dump heat from the metal refinery once I finish melting the cold biome I started refining in.