Posts by Omicron

Peppe, Requia, care to have a quick look and share your thoughts? https://docs.google.com/spread…VHhEZzUtX0dTZ3E2aVE#gid=3

I finished documenting the fuel type properties (as far as the Computercube can be believed) for the latest GregTech build. Things are starting to take shape; right now there are no more obvious bugs, but some things open for interpretation are left.

In summary, both thorium and plutonium got nerfed to half their 1.4.7 values. Thorium lost half its lifetime, while plutonium lost half its EU/t. As a compensation, plutonium scales up much faster in efficiency thanks to its double tick feature (but only up to 11 still, not 14 like Greg intended). You can see the results in the tables at the top of the relevant tab.

Also interesting is the effect on centrifuging re-enriched isotopes: if you do not have at least a dual neighbour situation (such as a 2x2 square of cells), you are getting more EU by not centrifuging but simply upgrading the isotopes to uranium cells. Also documented in the relevant tab.

The hybrid effect may or may not be gone. When combining uranium and plutonium, of which I know they work correctly, they deliver exactly the expected efficiency in all possible combinations, with no deviations up or down at all. However, thorium is another story.

I'm not yet sure that thorium works correctly, because although its neighbour scaling was fixed, the hybrid tests showed some really strange results. Back in 1.4.7, you had bad results when using two single cells, but things got better when upgrading to a dual thorium or even a quad thorium for maximum effect. In this build, however, the opposite seems true: you're getting a bonus when using two single cells, and moving away from that layout, no matter what combination, results in a smaller (and sometimes even negative) hybrid effect.

Note that this happens not only with plutonium as a partner, where the changes made to how plutonium ticks has a large effect on efficiency. It also happens with uranium. And the only thing that changed in the uranium/thorium pairing is the fact that thorium's burn duration got halved, nothing else. It still has the same EU/t, the same neighbour scaling. I don't understand why such a thing would completely reverse the trend - especially when the uranium/plutonium hybrid shows that there is no hybrid effect happening between them at all. If anything, it should either reinforce the trend, or also show no hybrid scaling whatsoever. On the other hand, thorium seems to behave exactly like it should in all the other tests aside from hybrid scaling...

What do you think? Do the figures make sense for you, and I'm just failing to see it? Is my math off somewhere, maybe? Or do you also think something is fishy with the way thorium behaves here?

Really? Then the NEI tooltip in my 1.4.7 testbed was wrong. Haven't looked at regulators with 1.5.1 yet.

The advanced regulator does not accept power from the front, i.e. its output face. Thus it is safe to attach it to the side of a reactor, which outputs power itself.

However, you have placed the advanced regulator in such a way that its upper side facing is flush with the reactor chamber above. Therefore, as soon as the regulator does something, uses energy and requests new energy, the nuclear reactor will attempt to fill that energy request. Since the regulator can only accept up to 32 EU/p, it explodes because your reactor is likely attempting to cram several hundred EU into it at once.

I was just citing the fact to explain where I am pulling the numbers in the math from.

Oh, and as Murphy's Law dictates, just after I linked that post to you I realized that I made a huge mistake in it Better read it again now, I fixed it.

@ Crispen: The reactor planner does the efficiency on most hybrid fuel designs wrong, sometimes with very large deviations. It's better to do it by hand.

First, count the number of fuel cells you have in your design: twelve thorium (3 quads) and 3 plutonium. Next, multiply each by the number of replacements they need over the course of a full cycle, which is dictated by the longest-burning fuel type. In this case, thorium is the longest-burning fuel, so you multiply it by 1. Plutonium needs to be replaced twice and then consumes another half charge, meaning you multiply it by 2.5. Finally, multiply each by the amount of EU a single cell of that type will produce at efficiency 1. For thorium and uranium, that's 1 million; for plutonium, that's 4 million.

12 * 1 * 1 million = 12 million
3 * 2.5 * 4 million = 30 million

You get a sum of 42 million. This number is called the reactor's "cell value". Divide the total EU yield of your reactor by the cell value, and you get efficiency:

353 million / 42 million = 8.405 <--- this is your real efficiency, even though the planner only states 4.93.

The same math works for any reactor in any configuration, but the reactor planner does it correctly as long as you only use a single fuel type, so you only really need to bother with it when you are mixing fuel types or you don't have the planner on hand.

For some additional insight on hybrid reactor designs, I wrote up something here that might interest you.

I'm aware, but I compounded the info in one sentence because a hybrid reactor needs both thorium and plutonium, and because I am lazy

Well, if you have a TE pulverizer with GregTech's overloaded recipes (TE config setting required in newer versions) or an industrial grinder and use them to process silktouched uranium and coal ore, you'll get plutonium and thorium dust as side products. So depending on how a player goes about their mining, they might have some spare plutonium and thorium even without a breeder + centrifuge.

And then there's also the 16 uranium dust = 16 uranium cells + 4 thorium cells + 1 plutonium cell centrifuge recipe, for those who don't want to mess with breeding. Using that you could run a mixture of uranium based and hybrid based reactors.

Peppe, the planner is showing 234 copper cost for that design, not 186?

Also, here's a potential alternative, *IF* you have thorium and plutonium available to you (for example as side products from pulverizing/industrial grinding coal and uranium, or from a breeder reactor). It adds a minor copper running cost in order to pull significantly more energy and EU/t per uranium ingot invested and reduce maintenance overhead. Whether that's worth it to you of course depends on your situation and supplies.

http://www.talonfiremage.pwp.b…jvwb1laxo925kajba5w2rwzr4

Eu/tick: 116
Efficiency: 6.44*
Cost: Iron 100, Copper 223*, Tin 35, Gold 14
Hybrid balance: +2 thorium positive
Maintenance advantage: it runs twice as long before you need to replace a fuel cell
Running cost: 10 copper per 14 hours

* reactor planner calculates hybrid reactor efficiency and copper cost of multicells wrong.

Oh, while I was at it, I made this today.

What is this? It's the "hybrid effect", that's what it is. Put into numbers. It works with all possible kinds of fuel combinations, with varying results. Some cell combinations score above the expected efficiency for some reason, while others fall below.

Disclaimer: valid only for 1.4.7 until further notice.

It's an okay start, but it could be better.

For starters, the heat plating is unnecessary. You need those only if you want to run a breeder reactor that you keep permanently at a high heat level. For safety, use containment plating. But even those are unnecessary when you have a heat-stable full cycle reactor.

Second, your efficiency is not very good for a hybrid system. Your improved variant has a cell value of 66 million (12.5 plutonium cells and 16 thorium cells per cycle), and a total output of 397 million, for an efficiency of 397 / 66 = 6.015.

This is because you are using too many plutonium cells. Hybrid reactors function on the basis of the single plutonium -> quad thorium pairing. You want to maximize the number of cell pairings while creating the least possible amount of heat. Since plutonium cells run far hotter than even quad thorium cells, that means the ideal setup has as little plutonium as possible and as much quad thorium as possible. This is the highest efficiency cell layout you can build. Cell value of 20, combined with 222 million energy yield, results in an efficiency of 222 / 20 = 11.1. If you surround the whole thing with reflectors, you can you even push this up to 270 / 20 = 13.5. DERP! I miscalculated that. Why doesn't anyone call me out on it? What I linked has a cell value of 26, not 20. The ideal efficiency setup is of course an infinite lattice of thorium and plutonium, though that is in no way practical. Still, maximizing the amount of plutonium/quad thorium pairings is the ultimate goal. Minimum cell value for maximum total yield.

However, there's another thing to to keep in mind: when producing thorium and plutonium, you almost always get both products from the same recipe, and almost always in a 4 thorium : 1 plutonium ratio. That means that the ideal setup above has a problem: it consumes 67% more thorium to keep running than you would get while producing the plutonium. 6 extra thorium cells are required per reactor cycle that you need to procude somewhere... or on the other hand, if you produce it the same way you produce all your other thorium, you'll end up with 1.5 plutonium cells per cycle that are not used and just pile up in storage. What a waste of energy!

Thus it's generally recommended to design reactors to be "thorium neutral", i.e. to consume 4 thorium cells for every plutonium cell (unless you run a whole group of reactors, some of which are thorium positive and some negative, to make the whole group neutral). And you must keep in mind that plutonium cells have a much shorter lifetime than thorium. The ideal thorium neutral cell setup is pictured here.

If you read through this thread you'll see it in several places; there's a variant that can be cooled with only 4 chambers, and another variant that breeds its own isotopes while running.

Yeah, because the condensator actually consumes all the heat it can get. And not just the leftover heat the internal vents can't handle.

So by shifting the setup you might be able to throttle the heat intake a bit.

Well, we basically ended up with the single LZH or cooling cell design. None of us could manage to make it fully coolable on just internal vents... I even got it down to 0 excess heat but still couldn't keep the components from melting over time. It's just not happening with the fuel cells consuming so much room, sadly.

Cheers Greg, thanks for your tireless works and for maintaining one of the fastest update cycles in the modding community.

This also shed some new light on the supposed death of hybrid reactors... in fact I already have an idea I want to try out once the fix goes out.

Okay, I called thorium cells in 1.5.1 "hideously useless" yesterday.

I must apologize: I was wrong.

Thorium cells are not hideously useless. They are bugged. Bugged to high heavens :p

That kind of explains the catastrophic numbers I've been getting in attempts at recreating hybrid setups...

Yeah. While it's a buff to pure plutonium reactors, it still can't compete in any way, shape or form with 1.4.7 designs. Or even non-nuclear EU sources for that matter.

Still, it was fun to throw together

Thorium is hideously useless now, by the way.

A quad uranium cell next to a single uranium cell has efficiency 3.6.
A quad thorium cell next to a single plutonium cell had efficiency 4.5 in 1.4.7.
A quad thorium cell next to a single plutonium cell has efficiency 2.667 in 1.5.1.

No such thing as a hybrid effect - in fact it's an almost 30% reduction compared to vanilla IC2.

What the hell are we going to be doing with all this junk?

Update: got the Efficiency 10 setup running stable. Upsides: awesome efficiency, only half the thorium overhead of previous attempt, less copper consumption due to using a dual cell, cooling system can likely be made cheaper by someone who's better at this than I am. Downsides: quad reflectors necessary. Still has thorium overhead. Only 200 instead of 280 EU/t.

The math says 11: https://dl.dropbox.com/u/44754370/plutoniumefficiency.png

Maybe with the doublepulse system, you can get above 11 if you pair plutonium cells with other plutonium cells? I haven't gotten around to test that. With reflects alone, though, 11 is all you're going to get.

Besides, even if you could get to 14, I doubt it would be coolable in a practical manner... This stuff is crazy enough at 7 already.

Currently toying around in a test install. Uranium cells are bugged in the GregTech computercube, unsure whether it's IC2's or Greg's fault. Also not important.

Much like Peppe reported, thorium is nerfed to half and plutonium is running insanely hot.

Currently examining plutonium efficiency. It actually scales quite nicely - if you place 1 reflector next to it, it acts as if there were 2 reflectors, due to the double pulse. As a result, maximum theoretical efficiency is 11 rather than 7 for a quad plutonium cell surrounded by reflectors.

I say 'theoretical' because that setup was impossible to cool in 1.4.7 and in 1.5.1 it is so much worse. I can't even get anywhere near a triple reflector setup, the reactor just liquifuies the moment I touch the simulation start button.

For starters, this is the best I could come up with - in the 1.4.7 planner it looks overcooled, but in 1.5.1 it just barely stays stable (we're talking somewhere around 670-680 heat here). Under the new plutonium rules, that's an efficiency 7 reactor (because it treats the 2 reflectors as if they were four). This is actually not half bad. Efficiency 7 plutonium was, as mentioned before, uncoolable in 1.4.7, but now it actually runs with just internal vents! And it only takes two of those mind numbingly expensive iridium neutron reflectors instead of 4. (Disclaimer: the realworld implementation reactor hasn't finished its cycle yet, but it looks okay so far...)

I'll freely concede that this is a clear buff, although one that does not yet reach the sheer convenience of hybrid reactors... after all, for every quad plutonium cell you're going to have four quad thorium cells piling up in your storage. And getting those consumed at a similar efficiency is going to be difficult because you'll need at least twice as many reflectors. And even then you're going to spend 5 out of 6 reactor cycles on slow, low-yield thorium...

More later, I got a spreadsheet to fill up, and then I'm going to test if I can get the "hybrid effect" to work in any shape or form.

If temperature fluctuations cause your reactor to peak above melting temperature now and then, the chamber might simply be melting off. Take out a heating cell for a bigger safety margin and see if that changes anything.

The iridium neutron reflectors are mind-numbingly expensive in legit play, though... unless you're into abusing villager spawning mechanics.

(and even then they're stil lextremely expensive.)