Rebalancing the fission reactors

Some time has passed since I implemented the reactor expansion and I’ve had enough time to experiment and think about the current balance of the system. There are some glaring issues I now see and can’t unsee.

The big bad balance blunder:

The biggest contender for “breaking the balance” or simply being overpowered as hell is the good ol 1x1 reactor core. The problem is simple: You can surround a 1x1 core with 4 2x2 cores. That mean the rod in the 1x1 core will be adjacent to eight other rods. So lets go through the scenarios of what this allows:

  • Surrounded by reflectors: Since the reflectors only receive half a side of neutrons each to double, the result is the same as if it were surrounded by just four reflectors, each receiving the full power of a side.
  • Surrounded by absorbers: Same as reflectors, since the output of the 1x1 core gets split over eight rods.
  • Surrounded by fuel rods: The rod in the 1x1 core would receive more neutrons, since there are twice as many rods surrounding it that provide these. This allows using the 1x1 core to quickly fill up breeder rods. That makes the breeding slightly more efficient, making the 1x1 core much better than the 2x2 for it, but could still be viewed as only a slight imbalance.
  • A moderator rod in the 1x1 surrounded by fuel rods: This is where the fun begins. Since the moderator rod reflects more energy based on the number of fuel rods surrounding it, there is a definite advantage in surrounding it in a 1x1, so it has eight neighbors instead of the usual four, effectively doubling the neutron output. But what about neutron maximum, shouldn’t the “output speed” be irrelevant since going over the maximum would reduce efficiency drastically? The maximum only cares about the output of the fuel rod, not about the extra neutrons created at reflectors or moderators. But doesn’t moderation from the moderator rod also reduce efficiency drastically, shouldn’t that balance it out? It reduces efficiency by 4 times. So, from a moderator you get 8 times more neutrons when surrounded by eight rods. Well, however see, that also means a fuel rod can only be adjacent to two of these, because it is in a 2x2 filled with fuel rods, meaning that only 4.5 times the neutrons get reflected (8 times 2 from the moderators plus 2 from adjacent fuel rods divided by four because that is per side, so total of 18/4), meaning it would almost be balanced out by the 4 times lower efficiency, only being an eight more efficient! Have you ever heard of surrounding a moderator in a 1x1 core with fuel rods in 1x1 cores? Did you know that in that case, the 1x1 reactor with the one fuel rod still counts as two sides to the moderator? Well, that means we speak of 8 times the neutrons reflected, even with the efficiency reduction from moderation that is still twice as efficient! And the solution can’t be to make moderation reduce efficiency even more, because that would punish the non-exploiting ways of using the moderator (in a 2x2 reactor for example) and water based reactors far too much.

So combing the 1x1 core and moderation is very problematic, especially because of how incomprehensible this behavior is for players. So how to fix this?

  • Make the neutron maximum take reflected/moderated neutrons into consideration. This would be a fine solution, the maximum would have to be based of the neutrons on the rod instead of the emitted ones. This would also make measuring if you go over the limit much more intuitive. The problem of the unintuitive behavior with moderators and fuel rods in 1x1 cores next to each other is however still unresolved.
    The balance of the whole system also gets shaken a but by this, because suddenly absorber rods will be much more efficient, since they effectively lower the neutrons on the fuel rod (since nothing gets reflected) and doubling the heat output of these, as well as reflectors now being nerfed comparatively, since their extra neutrons now count towards the neutron maximum. This would give absorbers the role of distinctly raising efficiency while reflectors (and moderators) are used for reaching criticality. Currently absorbers are kinda useless in critical reactors, because they are effectively equivalent to reflectors since reflectors also double the power output by doubling the neutrons instead of the heat, with no effect on the efficiency while also allowing to reach criticality.
    Generally this change would also count as a nerf to reactor efficiency overall, since neutron maximums are now reached earlier. This is also probably a good change, since reactors are pretty powerful right now anyway, so effectively cutting the efficiency in half (or more with moderation) isn’t that bad.
  • Removing the 1x1 reactor core (crafting recipe). This change would get rid of the weird interactions between it and moderations completely. The 1x1 core is generally more powerful than the 2x2 in any conceivable way: Easier to automate, more space to automate, less resource intensive, better for breeding, easier to reach criticality with moderation.
    Disabling it would tighten the balance and make reactors a bit more engaging an complex to build, since automation is much more difficult, having to use robot arms to interact with specific slots, while having to deal with less space for the automation (but certainly not impossible, because extenders exist).
  • Removing moderator rods (crafting recipe). This would also solve the problem with problematic behavior, but the superiority of the 1x1 reactor core over the 2x2 would still remain and water based reactors would be severely nerfed as well as the reactor system loosing a (in my totally unopinionated, not because I added this mechanic in the first place, opinion) really cool gameplay mechanic.

Personally, I would remove the 1x1 core (maybe leave a optional config setting to leave it in) and also change to neutron maximum system, as I think it would generally really improve gameplay and provide some necessary streamlining to the nuclear system while providing more complexity in reactor automation. The other really nice thing with this change: Not breaking backwards compatibility (much). Since the block would still exist and heat/neutron mechanics wouldn’t be changed, old reactor designs would still work and output the same amount of heat, only their efficiency would have changed.

Coolant balance:

The big issue out of the way, there are also other changes I would make, more specifically to the balance of the coolants. While I’m generally happy with most of them, I feel like a few coolants don’t fit their intended role: The molten metal and gas reactors.

The molten metal coolants are supposed to fill the role for breeder reactors, however they fall short in comparison to the carbon dioxide coolant. Carbon dioxide has the huge advantage of raising the factor of the fuel rod. This allows using one adjacent reflector less to reach criticality, freeing the side up for breeder rods.

This allows a critical carbon dioxide breeder reactor with U235. U235 has a factor of 1/4, which gets increased to 1/3, which means only 3 sides need to have reflectors instead of 4 to reach criticality, allowing the remaining free sided to be used by the breeder rod. In a molten metal reactor however, since molten metal coolant doesn’t increase the factor, the rod would need all 4 sides to reach criticality and thus can’t be used for breeding, only allowing for high tier fuels with a factor of 1/3 to even be used there for critical breeding.

Since the stats of the molten metal coolant reactors are generally only useful in a critical breeder anyway, this is unacceptable. I would personally give the molten metals the bonus of increasing the factor, while taking it away from carbon dioxide coolant, probably with some smaller additional balance changes to the molten metal and gas coolants in general to make them better fit their niche (or establish one for the gases) while keeping their general balance in check.

This change would however be breaking and could in the worst case lead to explosion (noises and power outages, the explosions of the cores are still disabled iirc). The potential boomers would however only be the molten metal reactors, which I doubt anybody currently uses anyway, carbon dioxide would only be getting nerfed, so power outage in the worst case but no boom.

3 Likes

Removing the 1x1 seems like a good idea overall, considering how exploitable it can get.

3 Likes

For me byggest problem with nuclear energy is too much nuclear fuel in worldgen.
So I disabled all uranium/thorium ores and byproducts (no more uranium from azurite/rutile byproducts). Also made refined cobalt (Co-60 source) only available from copper in aqua regia (100 times lower chance, I think. I’ts 0.5% currently).
Also added 1% chance to get tiny refined thorium and 0.33% chance to get tiny refined uraninite from centrifuging 2 dark ashes. Also a way to get nuclear fuel from bedrock-mining coal.

Result is interesting. First reactor was moderated Co-60 “bomb”. 3 1x cores, 3 universal extenders, 1 moderator rod, 2 Co-60 rods, coolant: CO2. Enough for large invar boiler. I’m currently upgrading it to use 4 Th rods and feed 2 large invar boilers (currently stuck because of not enough iron).

I have another similar reactor running. 5 1x cores, 4 universal extenders, 1 moderated rod, 4 Co-60 rods, coolant: industrial coolant. So my base is currently powered :slight_smile:

I need to have 20 of tiny refined uraninite to get my first U-235, currently I have 18. I get most of my dark ashes from centrifuging komatiite dust. Only then I can start breeding Pu-239 (real fuel).

Yes, I like extracting nuclear fuel from crap :slight_smile:

4 Likes

I want Pu-239 mostly because it will allow me to build stable reactor (good for powering large centrifuge).

3 Likes

Hmm, that is a very good point. I’d argue the easy availability of good nuclear fuel (i.e. U235) is the biggest issue. While small, low power reactors like the ones you describe are still useful, they are really not that powerful. With the 1x1 reactor core being removed, that first reactor would (when rebuild using a single 2x2 core) output almost half as much power anyway (24HU instead of 44HU, though that design, since it is subcritical, works much better in a industrial coolant reactor, producing 52HU).

Cobalt-60 and Uranium 238 however can just go critical in a moderated reactor setup (using only 2x2 cores), though since the neutron maximum on those is quite low it may not be entirely worth it, also considering you need 6 2x2 reactor cores of space to get a single rod critical.

What is most important to lower however is the availability of Uranium 235, which would ideally be the gateway to more critical reactors and breeding reactors, allowing getting better fuels.

However currently U235 is obtainable in several ways:

  • Processing uranium bearing ores by centrifuging. While the output of U235 here is not too high, the process itself is quite easy and the ore readily available.
  • Processing uraninite using the uranium fluorite chain. The output of U235 is a bit bigger, but since the chain itself is so much more work compared to simply centrifuging it, it’s probably not used by anyone.
  • Breeding with lower tier fuels. Ok, while not providing U235, it provides even better fuels like U233 or Pu239. Currently breeding seems to be too easy, since it is totally feasible by just using low tier fuels, taking just some hours. And with the 1x1 reactor enabled it would be even quicker.
  • Extracting it from depleted U238. I think this is also much more powerful than I originally intended. You get one third the materials you need for a fuel rod from a depleted one, making just “burning” the fuel in a critical reactor extremely viable.

So here is what I would do considering these things:

  • Make breeding without U235 unfeasible. This could simply be done by increasing the neutrons required. This would of course require careful examination, so the “intended” way doesn’t become to tedious. Another way to archive this would be to rework the breeding efficiency system. Instead of becoming exponentially more efficient with higher neutron counts per tick, a “neutron barrier” would take its place. This mechanic for efficiency would be very simple: When adding neutrons onto a breeder rod every tick, a “neutron barrier” value gets subtracted from it. This value would be so high, that its only even possible to breed fuel with critical reactors. Obviously more neutrons per tick would mean more efficiency, as the breeder rod would then require less ticks and therefore less subtractions to convert, but it would no longer be exponential.
  • Less “decay products” from depleted fuel rods. Quite simple, I would probably reduce it to two 1/72 of a pile of dust. Makes it not completely unfeasible to gain good fuels this way, but definitely less viable.
  • Making uranium bearing ores bedrock drill exclusive. Gating uranium behind it makes thorium and cobalt-60 the starter fuels that could remain fairly available. Since regular U238 is now gated, decaying it into U235 is would now require first attaining it by decaying Th232, which would make the process much, much slower and less feasible, although still possible.
  • Removing the centrifuging method of extracting U235. This will force players to setup the uranium fluorite processing chain, adding another gate to U235.

This would mean the fission progression would look something like this:

  1. Subcritical thorium and cobalt-60 reactors. Moderated critical cobalt-60 reactor possible.
  2. Unlock U238 by bedrock mining. Allows for U238 subcritical and moderated critical reactors, better than cobalt-60.
  3. Unlock U235 by uranium fluorite processing chain or burner reactor. Allows for unmoderated critical reactors and thus breeder reactors.
  4. Unlock U233 and Pu239 with a breeder reactor. High tier fuels which allow getting even better fuels in burner reactors.

All in all, good fuels should be harder to obtain, taking multiple steps, but once you have the necessary infrastructure, they remain readily available. Though I would really appreciate feedback on this, since it isn’t really my area of expertise. How easy are cobalt-60 and thorium to obtain for example?

2 Likes

Co-60 was very extremely abundant, until Greg made it just extremely abundant fairly recently. Uranium is more common than thorium, (because of uraninite and pitchblende ores).

Be careful about nerfing nuclear reactors. Some people may just not bother, especially if they have captured bedrock natural gas.

Also, currently I use Co-60 for large boiler, using moderated reactor, which outputs increasing counts of neutrons (autostops at 4000/tick). I’m not sure if it’s critical, or supercritical.

Uraninite processing chain is not that hard. I finished it… today, I think. Started at one point yesterday. 6 mixers, 2 smelters, 1 shredder, 1 dust funnel, 1 titanium centrifuge, pipes, drums, etc. All electric, because I don’t have that much uraninite :slight_smile:

3 Likes

Hmm, Co-60 seems to be the somewhat unbalanced due to its overabundance and ability to create critical reactors. Personally I think, instead of nerfing them significantly by taking away their ability to go critical, much lower abundance, preferably even lower than thorium, would probably the best way to handle this. But maybe it isn’t and nerfing the factor to be 1/32 like thorium would be the way to go, making it the worst but also most abundant fuel, while locking critical reactors to uranium.

Regarding the subcritical, critical, supercritical wording: My way of describing the reactors for GT6 is not entirely realistic, technically all reactors sustain at least a stable chain reaction, so every reactor would at least be critical and the ones where the neutron counts rise would be supercritical (while they are rising, when stopped they would be subcritical for a while again until they become critical, i.e. stable, at some lower neutron count). I however like to call the ones with a stable neutron count subcritical and the ones with a rising one critical as categories for reactors rather than descriptions of the current state of the neutron reaction for simplicity. So calling them critical or supercritical is both fine imo.

Overnerfing the reactors is genuinely something I’m worried about. I suggest turning a lot of balance screws at the same time, but I think reactors currently are powerful enough to warrant such heavy measures. The ability to create critical reactors that can theoretically put out any arbitrary amount of power (if there weren’t fluid throughput limits, power conversion and reactor fuel efficiency) is still present after the changes, but it would just be a little harder. You just wouldn’t be able to archive everything with the most abundant and worst fuel available.

I hope the critical Co-60 reactor isn’t the one with just 2 or 4 fuel rods, that shouldn’t be able to go critical, even with the 1x1 reactor core, especially in industrial coolant.

The uraninite chain isn’t supposed to be very hard, but it represents at least some special effort setup that is required to get the better fuel.

2 Likes

so im not sure how to respond to your idea’s of change, first i felt the 2x2 core is annoying to automate with gregs robotic arms. i love some complex automation but i felt it sometimes worked and sometimes didnt work the way it should. not sure how to explain but ive seen it pull from wrong slots even when set correctly. maybe thats been fixed by now. but if your having issues with the 1x1 reactor why not hard cap it so your neutrons cant go above 5k? i feel that would fix the balancing issues and still allow people to make small powered reactors, ive only ever used it for a breeder, i prefer to avoid making a critical/super crit reactor all together other then for breeding anyways, sure you can get some stupid power out of it. but ive made a ton of neat designs for 2x2 reactor parts that already output enough power to run two MB heat exchanger/boilers. ive never felt the need to make over powered reactors other then those. and they are “Safe” reactors, not once do they go critical or need turned off once started. well untill they run out of fuel. also if you debuff the co2 then it wont have a use imo. i barely use any other coolants then co2 and industrial coolant. ive never seen the need. if anything maybe buff the other ones, they are not exactly the easiest to get, or maybe rebalance the whole thing from scratch. irl nuclear reactors are massive power generators. they are suppose to be powerful. if you need to make crafting recipes cost more or not thats up to you, but i also never use basic fuels and only use bred fuels. basic ones are easy too, its just i never felt the need in my gameplay to use them other then initial breeding. with or without the 1x1 core for my breeder reactor its not hard to make another kind of it, it will just make automating it even more of a choir. its hard enough as is to get item sensors to tell you items are in it or not. and then which ones are in there or not at that. we dont have easy item sorting and with bred fuels never being a “perfect” rod missing atleast one point of fuel etc, you cant exactly set filters. im not sure the neutron maximum you would go for, but i can get 12-20k neutrons per 2x2 core in a safe reactor so i hope that wouldnt be messing with the “max neutron” system you speak of. each rod in my safe reactors atleast hit 6k neutrons. and in a 2x2 thats two per core hitting 6k+ ive not even tested americium yet. and these are not going critical at all and only use industrial coolant…

if you watch the video of bear989 touring my base, you can see the old initial design on my safe reactor and the max neutrons im pulling out of my fuel. sure its my older world and ive made progress on automating and setup on another world. but that is the basic idea of a safe reactor with massive power output, im not sure if you should just remove the benefit of critical stuff or not then. maybe make it skip over neutrons for critical reactors if its outputting too many too fast? as in if a reactor that outputs 20k per 2x2 core just started up it would lose out on “heat” to fluid from the neutrons till it “stabilized” so you can still do high output safe reactors?

ok there is way to much to read for me here but reactors are not an early game thing, hell xarses made 3 fusion reactors before even touching fission do to the complexity and your wanting to nerf it? especially your idea of making Uranium bedrock drill only, automating the Uranium processing line is complex enuff i dont think many will attempt it. i still havent done it and making power from large gas turbines and diesel gennies or large solar grids why would anyone setup a reactor? and your wanting to nerf it to where we have to search the world over for a bedrock vein? that will just make less people bother in my opinion. when i setup a reactor it will be be small constant power, i probably wont even mess with critical reactors.

also i found this, Cobalt-60 - Wikipedia so i would say Co-60 should be more of a byproduct that would make a long low running reaction and doesnt come from Co itself. i would say safe, low neutrons, non critical.

It does appear out of balance of sorts but I don’t think nerfing many of the things you say you want to do are a good path forward. I want to remind you that there is a 7 page document attempting to describe how to work with fission reactors, its not nearly complete or effective in telling the masses how to build fission reactors in the game so I want to strongly caution you that the solution is already complicated and hard to enter.

I built 3 fusion reactors, 5 large centrifuges, 5 tanks, 55 huge carbon pipes, 90 drains, 4 bronze centrifuges, a lv and mv electorizer, and supporting pipes and motors to run 3 sustained reactions because its easier to enter than fission. This produces ~24Ki Eu with about 7Ki going to generating fuel for the reactions

Fusion reactors are stable and not complicated to maintain, they offer a consistent 8192 Eu/t across every fuel which seems fair due to the low effort to maintain them. Compared to fusion, fission allowing for insane outputs in more complicated builds seems more than fair. Keep in mind that sustained usage of the large matter fabricator draws 16Ki Lu → 32Ki Eu making Fusion not suitable for sustained matter manipulation.

Other late stage Energy generation basically only includes Tungstensteel or higher turbines. Gas Turbines can produce 24Ki EU with FeW. Sustained steam generation can be accomplished with traditional fuels, but is complex and appears to consume fuel quite quickly. I use Butane/Propane/Methane High/Low tanks and a high/low for NG (fed from ~60 springs scattered thought the vicinity). and iridium diesels and HV dynamos for the fuels. This is enough to keep my base battery mains filled, but would not be enough to run my base on EU (I mostly use turbo diesel for any RU application and Gas for heat) let alone deal with the power demands from matter fabrication. These combine to make the only viable power source for late game steam generation via heatant generated from fission reactors.

So I caution the approaches taken to reform the fission system so that its still feasible to generate the enormous amounts of power required in the later stages of the game

Ore Changes

Cobalt-60 is too abundant

This is very true, but given that the fuel appears to be piratically useless, I’m not sure the correct reaction here, you imply that you can make it critical, but I’m not sure how to do that, or what the maximum power here is

Thorium is too abundant

Maybe, I’m not sure here at ~60k Uraninite ore/~85k Uraninite dust , I only have 13.5k (15%) Thorium, it seems that there is much here, but not more than would be expected

Uranium (238) is too abundant

As direct byproduct processing, (uranium small ores) and the like I have just over 2k from the above uraninite load and another 6k large refined ore to seperate, given that it has the same byproduct rate as thorium (15%), there is far too much of this form of uranium available as its easy to process at 1024 RU

Uranium-235 is too abundant

As a direct byproduct processing U-235, my production gives an actual rate of 24%, which is the same as the chance rate of 24% ( 0.03*8) per 1/9th. This seems to be far to high for the ease of getting the material this way.

No one does uranium hexafloride

raises hand I run hexaflouride, its a beast of a loop requiring at least 18 machines, but ya, given the high rates of U-238 and U-235 as byproducts, no one is going to set it up. It consumes SiO2 and H (It does byproduct Si) like a mother but seems to be over producing in general (I’ll be back later after computing the exact cycle numbers) However the Rate of U-235 appears to be nearly 40% which is quite high, given the effort it seems reasonable.

Plutonium (244)

As a byproduct of centrifuging Uranium ore, Pu-244 clocks in around 10% the rate of Uranium, I’m not sure my math here, I ended up with 195 units, which appears to be 1.5% rate.

Plutonium-241

I ended up with 22, so not a lot of gain here compared the the crap ton of uraninite I processed.

Plutonium-243

I ended up with 19 here, again quite low gains.

Ore wise we are quite heavy on Co-60 and Th, those outputs can most likely be shortened, or have the ticks of the material changed to make up for the abundance. They may also need their power reduced, but I don’t understand enough there to comment at the moment.

Uranium is outright too easy to get, and the U-235 rate from ore is very high, we need to take steps to reduce that, either making ore less potent, or requiring another step (yellow-cake) to get U-238 out. We can also do centrifuge or other lossy reduction to get early amounts of U-235 out if we are going to need it for breeding.

However moving Uranium to only work with bedrock IS NOT a solution. The in world rates of Uraninite/Pitch/Thorium are actually on the low end, the three veins ive found are quite sparse. I ended up using an IE vein for “Uranium” which Greg “fixes” to uraninite when crushed.

Putting them behind bedrock mining seems interesting, but the rate of finding bedrock veins, or even the ones you need is quite low. I’d rather see the existing ores made less potent or require reduction recipies for the higher quality reactors

[ I will be back with more comments on the other aspects proposed ]