Jump to content

Introducing the FLiCLiRR: A hyper-efficient mid game coolant-limited research reactor


Recommended Posts

tl;dr CLRRs are cool, I made a hyper efficient one, load up the save file below (no mods needed) and play with it. If using debug mode (recommended) don't alt-z while the reactors are running.

Before I start, I wanted to be sure to highlight this forum post (CLRR: Mega Research Reactor Tutorial) from @Nedigo. It inspired everything below, and I recommend reading it as it will make things a lot easier to understand.

Once I finished my initial read of Nedigo's post, I came to three important conclusions:

  1. Conventional research reactors (RRs) are bad
  2. Fuel-limited (FL) RRs are worse
  3. Coolant-limited (CL) RRs are super awesome and really cool

These nuanced insights motivated me to build a CLRR in a survival game, but I got sad when I realized how wasteful they are in practice. CLRRs are highly efficient in terms of the power they can produce per kg of enriched uranium and accordingly can generate a LOT of power (~30kW), but the majority of it is not going to be used in an actual base. FLRRs can be left off when power demand is low and therefore kind of solve this problem, but in a very inefficent way (see point 2 above). These facts present an interesting question: Can we combine the benefits of a CLRR with those of a FLRR while avoiding their respective pitfalls? Before answering this question we need to cover some RR basics.

RRs will accept up to 180kg of enriched uranium, 60kg of which is heated constantly during the operation of the RR. We will refer to these 60kg as "active fuel" going forward, and the remaining 120kg as "inactive fuel". Over time the active fuel is converted to nuclear waste at the same temperature, creating massive amounts of heat. Whenever the active fuel mass drops to 59.5kg, 0.5kg are transferred into it from the inactive fuel. This occurs once every ~30s, which translates to ~10kg/cycle of fuel consumption.

The temperature of the active fuel is determined by the ratio of active fuel mass to coolant flow. If both of these are held constant, the fuel temperature will be constant (see red line below), and thus the heat production of the RR will be constant. Limiting coolant flow will cause the fuel temperature to rise to a new equilibrium. This is the fundamental principle underlying a CLRR. On the other hand, A FLRR fed a single bolus of 60kg will see its active fuel temperature, and thus heat production, decrease over time (see blue line below). The area between the blue and red lines represents the heat (and therefore fuel) that a FLRR wastes.

image.png.ede9b6e7796848337abad132c6c1c3bc.png

Our first challenge then is to design a method to reduce coolant flow as active fuel mass decreases, maintaining a higher active fuel temperature. This can be accomplished using a series of valves connected to liquid shutoffs, as seen below (6 valves are depicted below but in theory one could use a higher number of valves to achieve a higher average active fuel temperature) When each valve closes, the overall coolant flow decreases by the amount the valve was set to. This causes the fuel temperature to rise. Because the absolute rate of fuel consumption is constant, as fuel mass decreases the relative fuel consumption rate increases, leading to lower trough fuel temperatures.

image.png.bf3d81ffc740549e5dacf92f51d8c255.png

Now that we have combined the benefits of a CLRR (high active fuel temps) with those of a FLRR (can idle until needed), we need to solve the main problem of CLRRs: boiling nuclear waste. The nuclear waste inside a CLRR will typically be >2000°C which is obviously way above its boiling point. CLRRs solve this by using a pool of nuclear waste (into which the hot RR waste is dropped) that is kept cool by constantly running a ton of steam turbines (see Nedigo's post). However as mentioned above this wastes a lot of power, and ideally we only want to turn on the steam turbines when we need them. Geothermal plants are faced with a vaguely similar problem, and typically solve it by using heat injectors to transfer heat from very hot magma (>1410°C) into a much cooler steam room (~200°C). Thus, to mimic this solution we need to create a high-temp chamber separated from steam rooms by heat injectors. The chamber will need a very high heat capacity in order to prevent the nuclear waste from boiling. Conceptually it simplifies things to think of this chamber as a high-capacity heat battery.

To summarize, the ideal RR design needs two main features:

  1. Variable coolant flow
  2. A high-capacity heat battery

After many, many hours I have achieved a hyper-efficient mid game coolant-limited research reactor design which can be built in survival mode without space materials. Oh and it also makes mutant seeds super fast. Behold, the FLiCLiRR:

image.png.882a578ca8e2fa23942697d71fedbc02.png

As you can see, there is large central chamber that uses a huge pool of nuclear waste as a heat battery. The heat is moved into 6 steam chambers using heat injectors. Above the RR you will see 7 valves, 6 of which are connected to a liquid shutoff. Each steam chamber is paired with 4 tuned-up steam turbines for maximum efficiency. When maxed out, this build will consume ~7kg of enriched uranium per cycle while generating ~23kW. Compared to a traditional tuned-up RR this represents a ~285% increase in efficiency (space materials allows efficiency to increase to ~450%, and peak power to 29kW), before you even account for the power a traditional RR build wastes. Overall it's a massive increase in efficiency. The main advantage of space materials is that the valves, shutoffs, liquid pump, and aquatuners can be made from thermium to tolerate higher temperatures, allowing you to run higher fuel temperatures. You also can get away with 2 aquatuners instead of 4 due to the magic of super coolant. Please note the important caveat that the exact efficiency numbers you can achieve depend on the gamespeed you use and the performance of your PC. Each user is going to need to calibrate their FLiCLiRR according to these parameters. In general, higher gamespeed means lower efficiency, as RRs behave differently at high speeds. This is really only relevant in debug mode or if you use mods to increase gamespeed. However, it is possible that slower PCs may have issues even at vanilla speeds. FLiCLiRR calibration (mostly) involves adjusting the valves to ensure that fuel temps don't exceed ~2500°C.

And now, the important overlays:

image.png.4fdc00f46cbc5677e31fb478bceaf39a.png

Plumbing is relatively simple (compared to automation at least). The valves/shutoffs are fed by 3 intakes of a steam turbine (right-most turbine in the upper left chamber, aka the feed turbine) plumbed in a way that allows the steam turbine output to cycle around the main chamber, preventing pipe damage. A single radiant pipe segment on the output of the turbine cools the output to further decrease this risk. As mentioned above, this setup can be tuned by adjusting the flow rate of each valve. For the space materials build (aka the premium build) I aim for ~875 g/s total coolant flow with all valves open, decreasing stepwise to ~60 g/s over time. The no space materials version (aka the basic build) uses much higher coolant flow rates to prevent overheat damage of the steel components. The steam turbines are cooled by 4 nuclear waste aquatuners. Each steam room contains steel lights, to speed up tune ups. These lights are cooled with conduction panels fed by the steam turbine output.  

image.png.43482c93b96943efee07d9f34743f968.png

The automation is very complicated and by far was the hardest part to get right. It incorporates several failsafes and thus the FLiCLiRR is basically meltdown proof (see known issues below). Key features:

  1. Main Clock (upper right steam chamber) composed of buffer gates looped back into a not gate. This allows a red signal to be sent every 600 seconds (configurable by changing the buffer gate values) which closes each liquid shutoff. The 6th shutoff (right most) is closed 300 seconds after the 5th by the two buffer gates within the main chamber (see graph 2 above).
  2. The Fuel Request System is governed by a single thermosensor (upper one near the liquid pump) which tells the system to turn on the RR when the temperature of the nuclear waste drops below the set temp. When this occurs, the system will send a pulse to the left of two conveyor meters to the right of the RR. This causes 60kg of enriched uranium to be dropped into the chamber and loaded into the RR. If power demand is above a certain threshold (more on this later) the signal is sent to the right most conveyor meter, which is set to 120kg. The 120kg mode, which I call ultra mode, is only available in the premium version. The advantage of ultra mode is that it ensures that the reactor has a constant active fuel mass for the first 6 cycles, leading to increased efficiency when power demand is high. This is achieved by holding all the valves open for the first 6 cycles, while allowing them to close as normal for the second 6. Ultra mode is triggered when the bottom right bank of steam turbines runs for at least 30s straight. Basically if the reactor is exceeding 2/3 capacity ultra mode turns on when the nuclear waste pool temp drops below the set temperature. Side note: I didn't use the automation port on the RR itself due to a potential bug I encountered where the RR would not start up again after this port was used. It was pretty inconsistent and maybe imaginary, as I didn't fully rule out other factors. However, I didn't want to design a complex automation system around a potentially buggy port so I skipped it.
  3. The Seed Mutation System controls the left most conveyor meter, which is set to 2.5kg. If the reactor is off, and the plants are ready to be harvested, the reactor will turn on long enough for harvesting to occur. The signal counters should be set to a number close to the growth rate of the plant (rounded up), and the timer sensors should be set up so that one pulse is sent per cycle. Hydro farm tiles are used because they can be made from gold amalgam, which has low radiation absorption. This system can be adapted to whatever plant you want. I used it in my survival game to make exuberant waterweeds, which should give you some sense of how powerful it is.
  4. The steam turbines are regulated by smart batteries, while the steam chamber temp is regulated by basic heat injector automation. The power control stations are only enabled if their associated steam turbines are in use, which cuts down on dupe labor. The lights in each steam room are controlled by basic automation. Side note: dupes will enter the upper left steam turbine chamber to make power chips and thus turn on the lights, but the light above the feed turbine only turns on if the feed turbine is running. This prevents pipe damage that can occur if the light temp exceeds 100°C, which is necessary because the feed turbine is only used (and thus the light is only cooled) if the reactor is on or if the central chamber temp exceeds 275°C (this last part is not needed but I wanted to do it because I like seeing full green bars on every turbine).
  5. Key failsafes include the lower of the two thermo sensors, which opens all the valves if temperature exceeds the set temp. This ensures that the pump doesn't overheat (basic version) or that the nuclear waste doesn't boil (premium version). The hydro sensor controls a liquid shutoff below the feed steam turbine, which prevents the nuclear waste from overflowing. The basic idea is to keep the hydrosensor tile as full as possible without overflowing, which is why the liquid pump is right below it. This mitigates the risk of nuclear waste boiling. Other key failsafes include the radiation sensors. The right one ensures that the valves open and the feed turbine turns on as soon as the reactor starts up, while the left one ensures that the seed mutation system does not add fuel to the reactor while it is already running.

Known issues: The conveyor meters can be buggy sometimes. For whatever reason they will occasionally drop an extra packet, but happily this behavior can be remedied by loading an earlier save as it will not happen every time. In my survival game this never happened, but I have seen it occur when testing in debug mode. Another issue is then when tuning the reactor, sometimes it behaves differently on game load. You can spend a lot of time calibrating things perfectly just to immediately meltdown the next time you load up your save.

Calibration tips: Be conservative. If you try to ride the edge you will get meltdowns due to the issues mentioned above. You will need to adjust the the flow rate of the valves and optionally the shutoff timing (main clock buffer gate values) to maximize efficiency while preventing meltdowns. I use longer shutoff timings for the first couple valves (when active fuel mass is high) and faster timings towards the end. This leads to slightly higher average active fuel temperatures. I strongly recommend loading up the save file attached below, which contains both the basic version (upper) and the premium version (lower). This will allow you to play with the calibration (turn on debug mode to set valves without dupe interaction) before attempting anything in survival mode. Speaking of survival mode...

Survival mode tips: Startup can be a bit tricky, but works best when you run the RR at max coolant flow by bridging in a separate water line. Be sure to disconnect the pipe leading from the RR input to the liquid vent so you don't dump cold water into the steam chamber. Remember that when filling the chamber you need to account for the eventual height of the waste pool. Do not exceed ~50kg/tile of steam the main chamber during startup, as it will compress as waste accumulates. It takes a long time to accumulate the needed amount of nuclear waste, but in the meantime you will have a function RR that generates a fair amount of power. For the steam rooms, fill the lower tiles with heated water (tepidizers work well) so that the eventual steam pressure is around 500kg/tile. It makes things way easier to build the whole thing in vaccuum, which can take a while to achieve. Make sure you start on research early, as it can take a while to unlock everything you need. Anticipate these factors and you can get one of these up and running wihtin a few hundred cycles. Once you get space materials you can swap out the main components, no need to rebuild anything. The main chamber can be accessed by deconstructing the power control station in the middle left steam turbine chamber and setting up a two-tile vertical liquid lock that touches the ceiling. You can then deconstruct the ceiling tile and place a single ladder segment to allow access.

Mod suggestions: Suppress Notifications and True Tiles are recommended. True Tiles will make it much easier to see what materials I used, while Suppress Notifications makes things look a lot cleaner. Fast Track might allow you to achieve better calibrations by increasing FPS. I would guess that RRs behave strangely below 30 FPS, I haven't tested this though.

Save file: I have included a save file here, as mentioned above. Please use it, it will answer your questions about what materials are used etc. I set up a bank of aquatuners (hooked up to cycle timers to simulate the variable power draw of an actual base) to the right of each build. Once you load the save file, hook up the aquatuners and watch the FLiCLiRRs go!

FLiCLiRR.sav

7 hours ago, Pedro_L said:

quick tip, make the heat battery taller and thinner for better heat transfer

Are you talking about transfer through the heat injectors into the steam rooms? IMO heat distribution within the battery is pretty good due to the nuclear waste loop. Heat injector performance is definitely a limiting factor in the basic version.

I had an earlier design with a thinner reservoir of nuclear waste, but found that the heat capacity was too low and couldn’t tolerate high fuel temperatures (~2400 °C). I wanted to keep everything on one screen at normal zoom, so extending the reservoir downward wasn’t an option. Arbitrary I know, but it’s fun to watch the whole thing at once and I didn’t want to rely on screenshot mode to do so :)

Nice build! I really like the on-demand radiation idea for mutating plants. I wanna try it.

On 10/4/2024 at 10:21 PM, Charletrom said:

I would guess that RRs behave strangely below 30 FPS

Well.. a lot of things suffer from the same problem.

So, since I'm the proud owner of a potato PC, I'm gonna try and build something like this with somewhat stripped down automation. Straight with space materials and no ultra mode. And if it goes boom, or my pc goes boom, whatever. Back to boring reactors and wasted power.

I'd also remove the tempshift plates inside the nuclear waste pool, bar the ones touching the metal walls. A big square of tempshift plates has pitiful thermal mass compared to a pool of nuclear waste and liquid/liquid heat transfer is brutally fast, I think the pool would reach a decent equilibrium even without them. More space for silly pixel art, like a big radiation symbol. Unless you have already tried that... and it results in waste boiling.

Kudos!

Thanks! A few tips for the plant mutator:

1. Set up the farm so that a single dupe (use door permissions) can only enter when all the plants are ready to harvest. This can be accomplished with timers and a signal counter that controls a door. Make sure the signal counters are set to a value above the growth time for the plants. For example, mushrooms need 7.5 cycles to grow, so set the counter to 8. Set the harvest priority to 9, and the timing so that the doors open at the start of the dupe’s work time. Don’t forget to include dupe sensors inside the farm so no one gets stuck.

2. Use a conveyor meter to control how much fuel gets loaded, making sure the RR only turns on long enough for the dupes to finish harvesting. If you are swimming in uranium you can just use 10kg so the RR is on for the full cycle. Make sure you set up the automation so that the autosweeper is disabled before any more fuel is added to the chamber.

3. Use gold amalgam for the farm tiles, and gold metal tiles for the reactor chamber. These materials have low radiation resistance and will allow a ton of radiation to reach the plants.

4. Mitigate radiation sickness. There can be ~3k or more rads in the farm when the RR is on. At high difficulty settings w/o lead suits you will have major problems. Lead suits are a good idea, and you can also put some rad pills inside the farm. That way only the farmer will eat them, and only during the cycle that they harvest the plants.

 

 

Archived

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...