ReactorCraft: Fission Reactor Designs (again)

[Demosthenex]

As a continuation to my previous thread on ReactorCraft fission reactor designs, I've reached the point in my game where I will be working on them again. Much has changed from the previous thread here:

http://forum.feed-the-beast.com/threads/reactorcraft-fission-reactor-designs.47199/

I'm on the final version for 1.6.x, 25z. I understand there are some changes already in place for 1.7.

In the last thread I experimented with a variety of reactor layouts and measured their output. This time I'd like to discuss the rules that govern reactors, and share with everyone so we can make better reactors!

The rules / pseudocode I have read so far, and will try to keep updated as we discuss:

• Boiler
  • Requires > 100 C heat to boil water, boiling can happen at 101 C
  • Each tick consumes 5 C of heat to convert 200mb water to 1 M^3 of steam
• Coolant Cell
  • Requires a reservoir of water on top
  • Each tick:
    • Reduces it's temperature by (temperature - ambient)/8 without
      requiring water (?)
    • With water it will
      • Remove (temperature - adjacenttemp)/2 heat from an adjacent block
      • 1 in 6 chance to consume the water
      • Doesn't refill until next tick so doesn't cool all adjacent blocks
• Fuel Core
  • 1 in 20 chance to fire neutron per tick (average 1 / sec)
  • Creates 20 C heat per neutron interaction with uranium fuel
  • Creates 30 C heat per neutron interaction with plutonium fuel
  • On neutron interaction always fire 3 new neutrons randomly
  • If air block next to it, lose ((temp-ambient)+1)/32 C heat per tick
  • Transfer (temp-adjacenttemp)/16 to adjacent blocks per tick
• Neutrons
  • Absorption chance when hitting a block
    • 90% chance on steel block
    • 60% chance on concrete block
    • 30% chance on water or flowing water
    • 20% chance on other opaque blocks, can cause side effects (lava, sickness, death)
    • 25% chance of neutron interaction with uranium core
    • 30% chance of neutron interaction with plutonium core
    • 25% chance of reflection against neutron reflector
      • If not reflected, in 25z always absorbed in 25z
      • 50% chance of absorption in 1.7 versions

EDIT: I should mention I'm reading Reika's public source code from

https://github.com/ReikaKalseki/ReactorCraft

To that effect, he has some incredibly clean code which makes excellent use of object inheritance to remove redundant code. I don't actually know Java, but I can interpret much of it.

To respect his efforts, I also made sure to ask as a courtesy if I could share a detailed discussion with everyone. He's been very open about his mods, and a responsive author!

 

[Pyure]

One thing I always tried to do was ensure each boiler in the reactor was always well above 100C. Energy wasted on "maintaining" 99C seemed wasted, although this may be in error. 200C "might" be a good target temperature if 5C is being subtracted every tick.

Because of this, it can be important to ensure you're not overdoing boilers. Boilers are coolants, and there's no point in keeping everything too cool when the point is to get steam.

 

[Pyure]

Also: if a boiler is at 101C, does it still subtract 5C and provide the regular amount of steam?

 

[Demosthenex]

Also: if a boiler is at 101C, does it still subtract 5C and provide the regular amount of steam?

Yes, as long as it's over 100.

 

[Demosthenex]

One thing I always tried to do was ensure each boiler in the reactor was always well above 100C. Energy wasted on "maintaining" 99C seemed wasted, although this may be in error. 200C "might" be a good target temperature if 5C is being subtracted every tick.

Because of this, it can be important to ensure you're not overdoing boilers. Boilers are coolants, and there's no point in keeping everything too cool when the point is to get steam.

This is actually a good discussion point, essentially, should reactors run hot or cold?

I take the opposite POV. If a reactor is building up heat, it isn't getting rid of it fast enough. Given the only way to get heat out of a properly insulated reactor is by making steam (not lost to air or using coolant cells), any heat leftover means you aren't making steam fast enough. That could even be dangerous.

 

[Pyure]

This is actually a good discussion point, essentially, should reactors run hot or cold?

I take the opposite POV. If a reactor is building up heat, it isn't getting rid of it fast enough. Given the only way to get heat out of a properly insulated reactor is by making steam (not lost to air or using coolant cells), any heat leftover means you aren't making steam fast enough. That could even be dangerous.

This is why I wonder if 200 *might* be a sweet spot. Every second, you basically have 20 chances to create steam: if you start at 150C, you get 10 steam conversions. If you start at 101C, you get just 1. If you start at 200, you should get the maximum number of conversions (20) per second.

At 500C you're probably wasting energy, but at 120C I suspect you're missing opportunities (and thus energy again).

All of this assumes you're "stabilizing" at 200C, inasmuch as ReC allows a reasonably stable heat level (we still haven't convinced Reika to implement this).

 

[Demosthenex]

This is why I wonder if 200 *might* be a sweet spot. Every second, you basically have 20 chances to create steam: if you start at 150C, you get 10 steam conversions. If you start at 101C, you get just 1. If you start at 200, you should get the maximum number of conversions (20) per second.

At 500C you're probably wasting energy, but at 120C I suspect you're missing opportunities (and thus energy again).

All of this assumes you're "stabilizing" at 200C, inasmuch as ReC allows a reasonably stable heat level (we still haven't convinced Reika to implement this).

Remember, energy isn't wasted unless you allow air cooling (no insulation) or use a coolant cell. If you had a reactor that ran at 100C, and instantly converted any heat into steam, it's very efficient.

 

[Pyure]

That can't be right. If so, you could heat a boiler to 99, insulate it, remove all heat sources, and it would stay at 99 forever. I'm pretty sure that's not happening.

Edit: fwiw, I'm looking at the same code and I don't see where temperature passively drops over time either :\ But it must be there somewhere

Edit2: It looks like we're inheriting a class which implements the interface updateTemperature(). I got java-confused and couldn't quickly figure out which implementations we're using, but the ones I checked are dropping temperature over time.

What that means for our debate is up in the air, I'd love to hear of some trial testing of a "boiling point" versus 200C reactor

 

[abculatter_2]

Why dont you use breeders?

 

[Demosthenex]

Ok, here's a diagram of several reactor layout's I've tried in v6e.

http://imgur.com/RTh8jUI

I've measured their output by using a steam pump where I break and replace the output pipe and measure the amount of steam output over a one minute period. Then I calculate the steam blocks per second.

The small 2x2 reactor in the upper right is a very simple and safe beginner design. If you make it three tall, you can power a steam turbine.

Another note I wanted to make regarding reactor design. I'm sure there's a better matrix math that we could use to describe the reactors and improve their design, but I've taken to counting how many cores are visible from any one core. As a result, I've found I should never have a core with more than five other cores visible in any direction. They heat up way too fast to cool safely.

 

[colt29]

Very interesting, I'm new around here, but am planning on getting much more involved with ReactorCraft, are you guys on any servers?

 

[Demosthenex]

Very interesting, I'm new around here, but am planning on getting much more involved with ReactorCraft, are you guys on any servers?

I host a mean one. Our group has a history of very difficulty survival packs, and we're the only ones using Reika's mods with finite water. It's whitelisted because it's friend and family in polite company, but you can PM me if you'd like to apply.