Did you intend for it to have only one heating cell? Right now, that design is both cooling-positive and very cold. Switching to a full 64-stack of heating cells makes for a significantly better breeder, although it's probably overkill for anyone running only a few reactors at once.
Posts by gattsuru
-
-
2.90h was not intended as a long-term use build, and that's the underlying issue. It wasn't used long, so designs on this thread aren't likely to work because no one bothered to design for it in the first place. Thorium has been significantly buffed since that release.
That said, 2.9h also has plutonium pretty good, so you may want to reconsider that part.
-
Is there something im doing wrong? I used your breeder design and left the reactor running for a day and a half, and i've only got 220 uranium cells.
What version are you using, and did you let the reactor heat up completely before putting in the heat sinks and isotopes? Thorium and Plutonium have been significantly altered in performance since a number of these designs were proposed, and depending on version you're running thorium may be completely useless (as in 2.90h), or simply most useful as a baseline load material (3.05+). -
The issue with cheap reactors is that they're often not terribly efficient, or hardly put out any EU/t. In fact, the three key factors of power output, efficiency and price are probably best described in a triangle. Your reactor can be anywhere inside that triangle, but as you move towards one of the ideals, you move away from both others... As such, you need to find an equilibrium somewhere, ideally one that matches your particular outside factors (like, if you only have a limited uranium income you need efficiency if you want to keep up the operation).
True, but there are certain places on that triangle that make more sense than others, and to a certain degree, cleverness lets you bend the rules a bit. That's especially the case given the high cost of the initial reactor, which heavily weighs EU/copper and EU/iron around. Finding the good spots is the tricky bit, though, and I'll admit not someplace I'm terribly skilled at.Quote(P.S.: cooling a 3x3 grid of quad thoriums in a 4-chamber reactor, good luck... I couldn't make it work with anything but 6, despite the lowered heat in 1.5.2. :p Though if you can manage a 5-chamber variant, I'm interested in your tricks...)
40 slots puts it just into the first space on a 6-chamber, I'm afraid : the best I can do with a 5-chamber is 8min up/2.5min down cycles before a component melts. OC Heat Vents are just too easy to fry. -
Yeah, the reduced costs of quad and especially double-cell options make a lot of designs much more reasonable from a perspective of running costs. You can do some cool stuff with even a 0-chamber reactor* using those options, but it requires a lot of extra infrastructure to automate a mixed-fuel reactor.
The hard part about high-efficiency spammable reactors is that the heat from even thorium quad cells can get out of control quick: that 3x3 block of quad thorium cells puts out 660 (GregTech 2.8) or 550 (GregTech 3.06) heat/tick. Even the newer, lower numbers take a theoretical minimum of 28 slots** of cooling, requiring an 4-chamber reactor and a crapload of copper startup cost. I don't think you can get a Mark I reactor higher than EFF4 without using (at least a dozen) OC Vents (EDIT: or reflectors, but that's even worse from a mass reactor spam viewpoint), and that's where a lot of the per-reactor cost comes from.
*pre-3.0 GregTech only -- it's only worth 70 eu/t in current, and needs another layer of vents to not explode.
** In practice, probably somewhere closer to 40.Of course, since we're by definition talking about large numbers of cheap reactors, very simple microcycles start to become appealing.
-
Can the rate of Uranium Cells turning into Near-Depleted Uranium cells within reactors be increased on worlds with B:8DepletedUranium recipe disabled? The ingot -> 8 cells recipe was really bad design space, since it pretty much mandated that your first reactor be a breeder, but the current settings make uranium loss rates pretty high, and mean you need to be running a lot of reactors a long time before a breeder reactor makes much sense.
-
Quite true. I'd actually love to see some 1 or 2 chamber reactor designs and see what we can do with it. I just took the 360 EU/t reactor and cut it in half as a one-chamber reactor which had 180 EU/t. Maybe someone can improve upon that.
This probably isn't any cheaper in practice, since you have to fill in the spare slots to avoid automation problems, but it's an easy 3.5 EFF 140 EU/t without too much extra costs, and the running expenses are significantly lower for a purely IC2 box.Copper costs get killer for thorium-plutonium boxes, I'm finding, and the 3.0 nerfs hurt EU/t/copper investment here a lot. Cutting down on OC heat vents and quads helps, but that hits performance pretty hard.
The problem with storage buses is that if all of your nuclear material runs out simultaneously, it won't be reloaded since it no longer has any nuclear material in the storage area. That's why I use import/export buses. It makes certain that it will always run smoothly.
The trick would be to not have a generic storage area, or any internal part of the reactor's AE network with stored uranium cells. By importing them from a second network through a chest, or by crafting them on demand, you always have fresh nuclear material entering the storage network just seconds after a reactor drains a cell, unless you run completely out of the materials on your central network. -
Much as it pains me to admit... this modular design pretty much kills CRCS, except in marginal cases where you want a higher Efficiency value (CRCS reactors tend to have a 5-6 efficiency rating, whereas this only has 3).
Efficiency does start to add up, though, at that scale, especially if you run in newer versions of GregTech without the uranium->8 deplete isotope recipe. 12-16 uranium ingots per ring is a good deal of equipment, and the full-size tower you've got there would run through a cell once every two minutes. For a couple cycles, it doesn't compare to the copper costs, true. But you need a lot of UU-Matter to rationally build a GregTech fusion reactor.On the other hand, if you really want to be a smart-ass, this general design can go apply to any single-fuel reactor system, and there are more than a few 1- and 2-chamber pure-thorium or thorium-plutonium reactors that can provide EFF4 or greater efficiency, albeit often at lower EU/t rates.
In theory, proper usage of Storage Buses, level emitters, and either a crafting station and/or a second AE network can let you use a single ME Storage Bus per reactor instead of an export and import bus, which might make 0-chamber simple pocket reactors viable in this tower configuration. That gets very tricky, though, since the Storage Buses have counting problems with nuclear reactors to start with, and they'll only move items as they enter or leave the network. Cost reduction and space may be worth it, though.
-
Ok, then I think I'll use Sheenkeythelost's design. And what's CRCS?
CRCS stands for Continuously Re-applied Coolant System. The idea is to cycle coolant packs from a single active reactor out into multiple cooling 'reactors' that are filled with heat sinks instead of fuel. Because you use coolant packs, instead of condensors, they're reusable. By doing this, you can put together designs that would be impossible otherwise -- arrays of quad isotopes are doable, if tricky, and efficiency 6 becomes a lot more accessible. The costs of cooling towers, however, make for pretty significant startup costs, and you also need to have some good automation chops or the reactors will melt.If you want simple, reliable power and don't terribly care about uranium efficiency, a couple breeder reactors and forty of Sheenkeythelost's Pocket Reactors will do it.
-
Isn't the whole point of a fuzzy bus that it can tell if its damaged or not? Is this a bug thats being fixed? Or is it supposed to be this way?
According to the AE site, fuzzy buses should differentiate between damaged and undamaged types of the same item, so that seems to be a bug. It's one that also (in rv10.n) occurred with most coolant packs (as well as all chargable tools, which have an NBT value even when fully charged), but not with a Thorium Cell or Flint and Steel. Put it in their bug tracker, will see what the response is. Even if they worked properly for coolant packs, though, fuzzy buses can't tell the difference between a Flint and Steel:1 (nearly but not quite perfect condition) and a Flint and Steel:54 (only a couple uses from breaking). Add in the relatively slow action rate for these buses -- they only pull or place one item every five ticks -- and you're still going to want some outside timer triggering them.And probably turning off your power plant.
-
You can set up to eight versions of the same item on a Precision Export Bus, which can cover a somewhat wider range of damaged cell. Still not a terribly wide one, and this complicates the use of pretty much any AE device to deactivate your reactor when there's too many hot coolant cells around. I tried an almost-pure-Applied Energitics reactor circuit and it worked for three microcycles before going boom.
It's probably safer to just use ComputerCraft as a timer. Use fuzzy busses set to only act on redstone activation, then use one side of the computer to shut down the power plant, pull out every coolant cell whether it needs it or not, add the fresh coolant cells, and then turn the power plant back on. You can even use ME Level Emitters to check that you had enough fresh coolant cells before trying to load them in, then check that they were removed from the AE network before turning the computer back on. Note that all Level Emitters are Precision, as of AE-rv10.n, though, so they're not very good for telling how many hot coolant packs you have.
-
The heat vents in the corners are there to disperse heat created by the thorium cells. Cooling cells can only absorb heat from any fissionable material thats directly next to them, meaning the heat from the thorium would otherwise have nowhere to go other than into the reactor hull, with explosive results.
Ah, that'd be it, and an obvious reason. I guess I expected a more significant increase in rate of hull heat rise, since just four two-neighbor'd thorium worth of heat like that'd blow the reactor up in a couple minutes even outside of the plutonium. But the quad plutonium's on an entirely different scale of instant burn once the coolant cells run out.QuoteThe thorium isn't really there for extra power, its just acting as cheap neutron reflectors.
Yeah. I'd much rather spend a tin and a thorium dust over 8 copper and 4 tin of a neutron reflector, at least in a CRCS situation. Have to balance the extra heat issue outside of one. -
A) You don't have the lapis lazuli.
B) You don't have the redstone.
C) You can't sanely store enough redstone (400 stacks?!) for a full cycle even if you did find it.
D) Automation costs are nontrivial, and any serious latency turns your house into glowing crater.
E) Even if you did, those aren't terribly good uses of the materials. Removing those reflectors and squishing the entire array down into four rows on the first reactor, for example, leaves you with the exact same performance profile as you started with and saves a lot of resources. -
I was thinking: what if you made a hybrid system, running the plutonium in a CRCS setup (for better efficiency), while using some of your thorium as cheap neutron reflectors? Something like this. It has the advantage of avoiding the extortionate costs of iridium neutron reflectors, and it shouldn't overheat, but i'm not 100%. I can't try it on the computer cube since i'm not on 1.5.2 yet, could someone who is test it?
You get a little over five and a half minutes (342.95 seconds) at 2412 EU/t before the coolant cells pop in the Computer Cube (running 3.06d Gregtech over 1.115.336-lf IC2), which is... pretty nice. By comparison, Thick Neutron Reflectors return 2400EU/t for 343 seconds. Thorium definitely makes more sense, both from a cost and performance perspective. Not sure if the heat vents make much sense on the corners, or if there's anything that can go in those slots meaningfully -- I'm seeing less than a seconds of difference between heat vents and empty space.
