Posts by Omicron

And here I thought to myself, "I could swear there was a 4-chamber design somewhere". And I searched around for 5-10 minutes, but my google-fu was weak

Unfortunately this design has an extremely high running cost due to the reflectors - if you paid the running cost with UU-matter, you would have to spend nearly 60% of all energy this reactor generates just to keep it running, resulting in a net efficiency of less than 3.

You should only use this if you have a huge overabundance of copper, tin and coal, and a scarcity of uranium.

And if you do use it, it can be built cheaper still with 5 chambers.

Yeah, breeding is definitely the way to maximize your uranium supplies. While you still play 1.4.7, you ought to take a look at Peppe's "Hybreeder", a efficiency 9.175, 367 EU/t hybrid reactor that can breed fuel for itself and 3-4 others as it goes It's probably the pinnacle of reactor design of the great (if short) era of hybrid reactors.

Full automation possible using GregTech advanced regulators, railcraft and applied energistics autocrafting. Not trivial to pull off though, it's a fun engineering challenge to figure out, but if you do it right you only need to insert uranium, coal dust, copper and tin and the system will take care of everything else for you.

Half right, half wrong.

For starters, you are correct in your assumption that the 4.93 figure posted by the reactor planner is wrong. The planner does not do it right because it doesn't take into account the "hybrid effect" (an unintended feature that lets certain fuel cell combinations output far more than conventionally possible, up to +75%). The correct way to determine efficiency is indeed to divide total EU output by fuel cells inserted. I dubbed it "cell value efficiency". Why bother giving it a name? Because Gregtech 3.x makes things hella complicated with double scaling plutonium, which will probably end up requiring hybrid reactors to post multiple individual efficiency figures by fuel types instead of using the overall "CV efficiency", so it might be useful to be able to tell them apart by name. More info here.

But I digress. You also made an error, namely in the way you calculated cell value. Good job on factoring the 2.5 plutonium cycles per thorium cycle in, that's the thing most commonly overlooked. But not all fuel cells are equal. You need to base your math off of the EU value that each fuel type will generate at efficiency 1 (i.e. single cell, no neighbours). In 1.4.7, for uranium and thorium that value is 1 million, but for plutonium it is 4 million. So for this reactor the calculation would look like this:

Thorium cell value: 3 (slots) x 4 (cells per slot) * 1 (cell per cycle) * 1 million (base value) = 12 million
Plutonium cell value: 3 (slots) x 1 (cells per slot) * 2.5 (cells per cycle) * 4 million (base value) = 30 million

The reactor has a cell value of 42 million EU.

353 million EU generated / 42 million EU cell value = 8.405 cell value efficiency.

It's not an 18, but it's still a pretty darn good value, considering the normal maximum is 7! Enjoy it while it lasts, Gregtech 3.x fixes the unintended hybrid scaling and nerfs both plutonium and thorium a bit on top of that. See this summary I made, with accompanying spreadsheet. (But keep in mind things can always change further, especially with a mod updating as fast as GregTech.)

Even more than in 1.4.7... how, though?

Plutonium has the same cell lifetime as before, but gives 50% less EU/t. That means over the entire lifetime of the cell, it will produce exactly half as much as a plutonium cell did in 1.4.7, given identical efficiency. On the flip side, if a plutonium cell in 1.5.1 runs with twice the efficiency as one in 1.4.7, it will produce the same EU/t and total EU.

So hypothetically, the doubling in efficiency scaling that plutonium received should completely equalize the two sides, right? Well no. Heat output scales linearly off EU/t, but exponentially off efficiency. That means that if the new plutonium needs twice the efficiency to achieve the same output, it will run much hotter even if its base heat output was half of what it was in 1.4.7.

Unfortunately the base heat is not half of what it was, it's higher. It went from a base 9 standard curve, to a custom spread that roughly matches base 8 at the lower efficiencies and approaches base 5.5 in the upper. Also, the new plutonium cannot match a flat doubling of efficiency values, because cells without neighbours start at 1/2/3 respectively instead of 1/3/5 or 2/4/6.

Therefore, in all possible scenarios, plutonium is either a.) producing less EU/t than in 1.4.7, or b.) running significantly hotter than in 1.4.7, or c.) both at the same time.

That configuration will be running at efficiency 9, which is definitely very good. Plutonium is a prime candidate for cooland cell reactors because you can make use of the scaling, which is something internal vents won't let you.

(Psst I think the first one's a wrong link :p)

New spreadsheet link: https://docs.google.com/spread…VHhEZzUtX0dTZ3E2aVE#gid=3

Greg's changes from 1.4.7 to 1.5.1 seem more or less final now, as he hasn't made any changes in the last 3 or so updates. Though he hasn't explicitly acknowledged it, so keep that in mind. Here's the current rundown as of 3.03a:

Plutonium:
* -50% power output
* pulses twice per reactor tick (doubled scaling, doubled breeding potential)
* reduced heat (custom flatter curve instead of standard scaling, roughly equivalent to a one third reduction on average)
* removed hybrid scaling (bugfix)

Thorium:
* -50% lifetime
* -20% heat
* removed hybrid scaling (bugfix)

Overall a resounding nerf, even without counting the hybrid bugfix. Plutonium now gives far less EU/t no matter the configuration, and thorium's total EU worth has been cut clean in half. The buffs in heat scaling and efficiency don't make up for it. And of course, the bugfix completely destroys hybrids as we know it. It was fun while it lasted, I suppose...

Regarding the design of cooling systems, thorium and uranium now produce the same amount of heat for the same EU/t (though thorium's slower speed obviously results in less total EU). That means thorium is slightly better off in this regard than in 1.4.7. Plutonium, however, is much worse off now compared to cooling uranium. Even single cells run searingly hot as soon as you make any attempt at utilizing the improved scaling.

Mind you, plutonium and thorium can still allow you to turn a net energy profit if you centrifuge re-enchriched isotopes for them. Provided you use the resultant cells in at least a 2-neighbour configuration, you'll get more total energy out than if you were using uranium (although it never goes above a few percent, compared to 1.4.7's greater than 55% bonus even without abusing the hybrid bug) while saving a few pieces of coal dust compared to making uranium out of the same re-enriched isotopes. But of course you need to bother with running a breeder for that.

The advantage of using hybrid reactors now boils down to being able to hit a specific heat profile better by mixing a few low-heat cells with your high heat ones. Basically the same thing you were doing when you were pairing quad uranium cells with single ones because you didn't have enough heat capacity left for a double. There is no more EU bonus to be had through hybrid configurations. Also, as mentioned above, plutonium hybrids can no longer be compared by overall efficiency scores, since the exact scaling depends on the ratio of plutonium to non-plutonium cells.

As a bottom line, you can say that GregTech reactors are now roughly comparable to standard IC2 uranium reactors. They don't give you tangible advantages, but rather they give you new options to achieve the same result. GregTech's advantage now lies solely in allowing you to silk touch uranium ore for significant bonus yield.

Are you sure you want to spend that much on 60 EU/t? I mean, playing GregTech these four thick reflectors take 39 emeralds (or 64 ender pearls), 128 copper, 64 tin and 64 coal. And you have to replace them every two cycles, which generate only 24 million EU during that time - only about 1.5 UU-matter.

It's very uranium efficient, that's true, but you're paying for that efficiency with murderous running costs in other resources.

By the way, I've discovered that I've been analyzing hybrid reactor performance wrong all this time. That's why I would like people to question and doublecheck my methodology >__>

While I go and redo all the tables, I discovered an unexpected side effect of the 1.5.1 version of plutonium and its doublepulse mechanics... basically, you can no longer use efficiency to compare plutonium-based hybrid reactors to each other. Here's an example:

Take a single plutonium cell surrounded by single thorium cells. The plutonium will be at efficiency 9 and the thorium at efficiency 2. The reactor generates 26.5 million total EU at 53 EU/t with a cell value of 4.5, for an overall efficiency of 5.889.

Now take a single thorium cell away and replace it with a dual thorium cell. If you do this in the reactor planner right now, the overall efficiency goes up - which is logical, as you are removing one efficiency 2 cell and adding two efficiency 3 cells, while all other components stay at the efficiency they were before. Clearly it is a more efficient reactor. However, after dividing the 28.5m output by the cell valoue of 5, you get efficiency 5.7... which is less than what the other reactor had. And this is not a hybrid scaling effect, mind you... there are no more hybrid scaling effects.

No, what happens is that you are seeing a mixture of the normal scaling (of thorium) and the double scaling (of plutonium). And when you add more thorium cells, the mixture gets shifted towards the normal scaling side. The unfortunate result is that the efficiency of the first sample reactor and that of the second one are no longer directly comparable. The second one has better real efficiency, but the first one generates a higher number because that number is more strongly influenced by plutonium's double scaling.

I'm not quite sure how to handle this yet... hybrid reactors might potentially require an efficiency rating for each individual fuel type from now on.

I like the cooling system in your first one. The reactor stats are nothing special - you can get the same EU/t and efficiency with less cost and effort. But I've never seen that particular approach to keeping the edge vents from overloading. Pretty neat.

Regarding your breeder, that can be done a lot better. Check the second post in this thread for examples.

I like the first one. It's only a little slower but noticably simpler and cheaper than the offering listed on page one of this thread. Nice job, this thread rarely sees actually valid suggestions instead of junk spam!

A pair of quad thorium cells (or a 2x2 square of duals) will do 32 at efficiency 4. A 3x2 rectangle of quad thorium will do 128 at efficiency 5.33.

The second one currently takes 408 cooling in 1.4.7, which requires a fairly large reactor, but if Greg keeps the current values then 1.5.1 will bring that down to 326.4 which should let you cool it more cheaply.

Of course, using thorium means you'll likely also be generating plutonium on the side. Neither uranium nor plutonium will hit 32 or 128 exactly, since they scale of a base EU/t of 5 and 10 respectively. But you could use those for mass EU generation for your matter fab, where packet size doesn't matter.

P.S.: Wrong reactor planner link.

Well, if you consider "severe waste of resources" to be something bad, then yes, maybe

However in terms of running safety, when the reactor planner tells you "Generation time: full cycle" and calls it a "Mark I" in the heat bar up top, you can be sure it won't blow up on you. Provided you don't place components the wrong way. It can happen to anyone. For 100% safety, babysit the reactor during its first cycle.

Yes, it was designed for 1.4.7. And yes, you do have to replace cells multiple times in mid-run. Look through this thread for posts made by Peppe, he has descriptions and image galeries and even a world download for you to check out how the automation can be done.

Be prepared to learn a whole lot of complex stuff though, if you have never done something like that before...

A nerf is a nerf is a nerf, regardless of whether it was a bugfix or a balance change

By the way, thorium/plutonium and thorium/uranium pairs are still behaving a little funkily as of 3.02a. Each cell on its own is fine (nevermind, plutonium's heat scaling is out of whack, more details another day), and plutonium/uranium hybrids are fine, but most (but ot all) hybrid pairings involving thorium still deviate from their expected efficiency. Is that still a bug to be fixed, or are you done adjusting things?

(Just so I know whether I can start making new designs or not)

Quote from RandomLimes

I cant find any good 4 chamber reactor. So i would like to know if there are any good high (for a 4 chamber reactor)power that wont have a running cost except the uranium/plutonium/thorium? Im using greg tech also if anyone wants to know.

From the stickied reactor design thread: Peppe's 4-chamber thorium neutral hybrid. 367 EU/t, Efficiency 9.175. Running cost: slightly under 2 copper per hour.

Reminder: all current hybrid reactor designs cease working with GregTech 2.90 or later due to a massive nerf.

Try to keep things simple and practical while designing reactors. You might have a pleasing shape, but it's more expensive than necessary and has uneven condensator decay, which can cause problems in trying to automate it (depending on how you do it).

Try this: http://www.talonfiremage.pwp.b…6p7ttnw9oiebanl6m7a9z9m9s

Only 5 chambers, higher efficiency, higher output, cheaper building cost, even heat distribution (nevermind, due to the very nature of how things work, the edge cells generate less heat. Still, the rest of the advantages stand.)

Well, only the GregTech stuff is changing, standard IC2 is not. And the 1.4.7 designs are all finalized and listed in the first three posts... (okay, the 4-chamber thorium neutral hybrid is missing, I'll give you that.)

Negative hybrid effects aren't necessarily wrong. You had them in 1.4.7 as well 8check the first tab in the spreadsheet). What surprises me is that the effects I'm getting in 1.5.1 are so completely against the established trend.

At first I thought I calculated the expected efficiency wrong, but I doublechecked it. Also, if I had done it wrong, the trend should still be present, just shifted up or down a few tens. And not reversed.