Ultimate-Ish Big Reactors Guide

  • Please make sure you are posting in the correct place. Server ads go here and modpack bugs go here

MikrySoft

New Member
Jul 29, 2019
14
0
0
Annoyed by repeated false statements about BigReactors designs, I decided to get on the GitHub, read the code and get my own findings. The following work is a result of almost exclusively code analysis, which may or may not miss some aspects of the mod, but the findings match limited in-game tests I did.
Feel free to correct me if I missed something.

Guide created bases on version 0.4.1A (current GitHub "master" branch)
Code:
 Reactor Moderator/Coolant data
            Abs.    Heat    Mod.    Cond.
Blocks:
Iron        0.50    0.75    1.40    0.60
Gold        0.52    0.80    1.45    2.00
Diamond     0.55    0.85    1.50    3.00
Emerald     0.55    0.85    1.50    2.50
Graphite    0.10    0.50    2.00    2.00
Glass       0.20    0.25    1.10    0.30
Ice         0.33    0.33    1.15    0.10
Snow        0.15    0.33    1.05    0.05
Copper      0.50    0.75    1.40    1.00
Bronze      0.51    0.77    1.41    1.00
Aluminum    0.50    0.78    1.42    0.60
Steel       0.50    0.78    1.14    0.60
Invar       0.50    0.79    1.43    0.60
Silver      0.51    0.79    1.43    1.50
Lead        0.75    0.75    1.75    1.50
Electrum    0.53    0.82    1.47    2.20
F.Electrum  0.54    0.83    1.48    2.40
Shiny       0.57    0.86    1.58    2.50
Enderium    0.60    0.88    1.60    3.00
Fluids:
Water       0.33    0.50    1.33    0.10
Redstone    0.75    0.55    1.60    2.50
Glowstone   0.20    0.60    1.75    1.00
Cryotheum   0.66    0.95    6.00    3.00
Ender       0.90    0.75    2.00    2.00
Pyrotheum   0.66    0.90    1.00    0.60

Abs. - Neutron Absorption - How much radiation is converted to heat
Heat - Heat Efficiency - how efficiently radiation is converted to heat
Mod. - Moderation - How well it slows down the neutrons
Cond.- Conductivity - The higher the number, the more RF you get from geneated heat.

During each update the reactor calculates radiation generation, simulates moderation and neutron absorption by coolant, calculating fuel heat and reactor heat.

Radiation has 2 types:
* slow, which can be absorbed by other fuel rods (to increase irradiation, thus decrease fuel usage) or by coolant (to generate RF)
* hard aka fast - does nothing unless is moderated by a block in a reactor

First step in calculating generated energy is calculating how much heat and radiation is generated in each fuel
rod.
Fuel temperature has following effects
- making the radiation harder (more fast neutrons, base level is 20%),
- making the fuel rods produce less total radiation - output takes a nosedive around 1000 degrees and levels off about 3000.
- making the fuel rods use more fuel (NOTE: While comments in code say that is the case, the fuel usage value is calculated without taking heat level into account)

Size of a reactor also has an effect on generated radiation - the more total fuel the reactor has, the more radiation it produces, while the amount of fuel rod columns reduces the production exponentially - higher reactors are better than flat ones.

Second step is simulating the radiation propagation - radiation travels 4 blocks in each cardinal direction, on every block the radiation is:
- absorbed (intensity * absorption coefficient * (1-hardness)
- moderated - hardness is divided by moderation coefficient
Absorbed radiation is then converted to heat (absorbed * heat efficiency) and added to total reactor heat.

Third step is transferring the heat from fuel rods to the reactor. Heat transferred is directly proportional to the sum of conductivity factors of blocks surrounding the fuel rod.

Fourth step is transferring heat from the reactor to the outside world - either to the ambient air (20 degrees C) or to the coolant (100 degrees C). Base heat transfer is equal to temperature difference between the reactor and outside world, multiplied by 0.6 (thermal conductivity of iron)
Here the power generation method splits between passively cooled and actively cooled reactors:
- Passively cooled reactors
Base heat transfer is multiplied by 0.2 - that's the amount of heat actually transferred, which is converted to RF with 0.5 efficiency.
The transferred heat is subtracted from the reactor heat
- Actively cooled reactor
All available heat is converted to steam (assuming sufficient amounts of cold coolant and space for steam in the internal tanks) and removed from the reactor heat. NOTE: Size of coolant tanks depends on the reactor size and is limited to 50 buckets - this makes it impossible to power more than 25 turbines from a single reactor.

Finally small amount of heat is lost to the ambient air, dependant on the reactor heat level and it's surface area.

As can be clearly seen, the best between-rod moderator is gelid cryotheum, followed by resonant ender and graphite (both equal), with the famed diamond taking a distant 6th place.

If the size of the reactor is limited to the 7x7 size moderation properties of the coolant are irrelevant, therefore only the product of absorption and heat efficiency coefficients needs to be considered. Resonant ender wins in this category by small margin over cryotheum (0.675 vs 0.627)

If multple layers of coolant are possible, each layer increases power output essentially for free (excluding initial setup cost). In those situations high absorption value of resonant ender is not required, even harmful. Calculating estimated radiation transfer shows that the best coolant for multiple layers is by gelid cryotheum with signifiant lead over blazing pyrotheum and enderium blocks.
Gelid Cryotheum has the disadvantage of being not enough absorbing - 4 layers will convert to heat only 98% of generated radiation. This is easily solvable by changing the layout to have 3 layers of gelid cryotheum and the outermost layer of resonant ender - this increases the radiation capture to about 99.5%. The increase in generated power is quite small and it makes the build considerably harder due to properties of the fluids involved, but it lets squeeze a bit more fuel efficiency.

The analysis above is based on reading the source code of the BigReactors mod available at https://github.com/erogenousbeef/BigReactors and doing a static radiation flow simulation in Excel. Results of limited in-game testing support the findings, with the 13x13x13 reactor with outer layer of resonant ender and gelid cryotheum everywhere else being the most efficient cube reactor tested, providing almost double fuel efficiency over the "reference" 7x7x7 with diamond and ender.
 
Last edited:
  • Like
Reactions: triblades and Pyure

triblades

New Member
Jul 29, 2019
80
0
0
Very nice find. :) Thanks for your hard work! (I would think :p)
I see some reference to the placement of fuel rods. I could easily miss your point, if made, but is there an optimal placement pattern according to your review?

One tiny thing:
Could you please put a version number on your post of which Big Reactors version you got the numbers from? In case something would change in the future ;)
 

MikrySoft

New Member
Jul 29, 2019
14
0
0
Mod version number added.

I did not really go into rod placement, but my limited in-game tests show that I get better performance from 5 rods in X pattern than 8 rods in O pattern (for 3x3 reactor core). Eliminating the center rod might increase the fuel efficiency slightly, as this rod isn't irradiated by any others, but at the cost of lowering the power output.
I also had some success with 4x4 reactor core with 8 rods (4 in the middle, 4 corners), when Slowpoke101 requested a design for an even sized reactor.
I don't feel the need to use more rods as the 5 rods in 13x13x13 already saturate the reactors ability to output steam. If I wanted a bigger reactor (to throttle it down further and get better fuel efficiency) I would first go up to the size limit before adding more fuel rods as more control rods (for the same amount of fuel rod blocks) decrease the efficiency really fast, as I mentioned before.
 

Kuro-Maii

New Member
Jul 29, 2019
35
0
0
how many fuel rods did you test it with? in both cases 5 rods in an X pattern, or did you use some thing different for the 13^3 reactor?

I happened to be a bit late with y question so never mind
 

Pyure

Not Totally Useless
Aug 14, 2013
8,334
7,191
383
Waterloo, Ontario
Very much loved this walkthrough, thx Mikrysoft. I've had a big interest lately in attempting to design "ultimate efficiency reactors" and this has given me some better insight as to what's going on.

I'd love to see you make your own simulator one day. There's one out there already but it misses most usable blocks allowed and doesn't average out randomness (switching from a more-efficient moderator to less-efficient one might briefly show an improvement due to RNG. This should be averaged out so that you can immediately see an improvement.)
 

Skyqula

New Member
Jul 29, 2019
568
-1
0
Most of what you said is in line with my post here, I didnt read trough the code so there are a few interesting things here. Some of the calculations/values used where presented here and are more readable then digging trough the code. On to the next part, there are a few statements I dont agree with:
higher reactors are better than flat ones.

This is false for small reactors. The most important stat for fuel efficiency is fuel reactivity. For a small reactor this means a flat reactor so that rods can radiate eachother. Once a good design has been set with high fuel reactivity then building up starts to become interesting. For example: This 9x9x3 beats the "standard 7x7x7" quite easely in terms of efficiency. A proper 7x7x7 beats it though. But thats not strange considering we are talking about 21 fuel rods vs 40.
As can be clearly seen, the best between-rod moderator is gelid cryotheum, followed by resonant ender and graphite (both equal)

Where can we see this? Just the moderation stats? Because there is more going on. The best coolant between rods is either cryotheum or graphite. Wich one is better depends on the reactor temperature and fuel rod density. The reason resonant ender is not considered is because the unique aspect of graphite is that it lets nearly all radiation trough. Allowing higher fuel reactivity making up for the reduced thermal conductivity compared to cryotheum. That said, in big designs cryotheum wins just because of the thermal conductivity (assuming high rod density, as that provdes the highest fuel reactivity). A realy interesting option that requires more testing is Manyullyn (0.68, 0.88, 1.75, 4.5), wich may or may not actually be a contestant for first spot under the right conditions (possibly less coolant needed ==> more fuel rods).

If multple layers of coolant are possible, each layer increases power output essentially for free (excluding initial setup cost). In those situations high absorption value of resonant ender is not required, even harmful. Calculating estimated radiation transfer shows that the best coolant for multiple layers is by gelid cryotheum with signifiant lead over blazing pyrotheum and enderium blocks.
Gelid Cryotheum has the disadvantage of being not enough absorbing - 4 layers will convert to heat only 98% of generated radiation. This is easily solvable by changing the layout to have 3 layers of gelid cryotheum and the outermost layer of resonant ender - this increases the radiation capture to about 99.5%. The increase in generated power is quite small and it makes the build considerably harder due to properties of the fluids involved, but it lets squeeze a bit more fuel efficiency.

This is an interesting minegame for sure. But ultimately, given you can make anny component, a single layer of resonant ender and more fuel rods is more efficient then more layers of coolant. If you are limited to the amount of fuel rods you can make then multiple coolant layers would be a good idea.
 
Last edited:

MikrySoft

New Member
Jul 29, 2019
14
0
0
how many fuel rods did you test it with? in both cases 5 rods in an X pattern, or did you use some thing different for the 13^3 reactor?
All my reactors have 5 fuel rods in X pattern.

Skyqula: Very interesting simulator, I was not aware of it's existence.
The fact that my analysis is incomplete doesn't really surprise me - I concentrated on the coolant layout, ignoring rod layout almost completely. I also find that I need to repeat my efficiency tests for actively cooled reactors - with passively cooled reactors my 4 layers of coolant win over filling the reactor.
 

Skyqula

New Member
Jul 29, 2019
568
-1
0
Untill you start to use controll rod insertion, play around with it in the simulator. Every passive reactor has better efficiency when controlled down, with ofcourse the mayor breakpoint of a 1000 degrees being the most important one.
 
Last edited:

Darkone84

New Member
Jul 29, 2019
220
0
0
I have found with my own turbine testing that using the new Ludicrite Blocks for the coils are the best. In older versions Enderium Blocks were the best. You only need 32 Ludicrite Blocks compared to 37 Enderium Blocks.