Ore Processing Concepts

From GT New Horizons
Revision as of 02:29, 2 July 2022 by OrderedSet (talk | contribs) (Add medium sized EIO line example)

How To Do "Good" Ore Processing?

Although there are many implementations of doing ore processing, the desired features of a "good" solution remain the same. These desirable characteristics are:

  1. Flexible
  2. Expandable
  3. Unbreakable
  4. High Throughput
  5. Lag Friendly
  6. Minimal Human Intervention

[Flexible]

Ores in GT: New Horizons have multiple "routes" you can take which give different outputs. (See the bauxite chart to the right.) For bauxite you can take three routes:

  1. Macerate -> Macerate -> Simple Wash/Centrifuge (Grossular byproduct)
  2. Macerate -> Thermal Centrifuge -> Macerate (Rutile + Gallium byproduct)
  3. Macerate -> Simple/Normal Wash -> Macerate -> Simple Wash/Centrifuge (Rutile + Grossular (if normal wash) byproduct)

While for most bases, Rutile + Gallium Route 2 is superior, perhaps you already have automated gallium from essence berries. Then you might decide to do Route 3 instead, as it gives 2x more Rutile and therefore 2x more Titanium.

Every ore in the game will have a different "optimal" route for your run and base needs. This may also change over time. As a result, you want ore processing systems to be flexible - allowing different ores to go on different routes.

You also want to be able to automatically pull out things at any step of the process. For example, Rare Earths from GT++ require much higher power load, which would cause your existing macerator line to choke without modifications. You'll want to filter these out and process them in specially dedicated macerators.

[Expandable]

Depending on what batch you are processing, you may find a major bottleneck in your current system, or you simply want to replace machines with multis. This means your system must be able to expand without causing catastrophic infrastructure redesigns. Ideally you only want to change stuff in the current step you're dealing with (eg thermal centrifuging), not the whole line.

[Unbreakable]

No matter what weird combination of ores or even non-ores you feed in, your system shouldn't break.

[High Throughput]

Your choice of logistics system should not be a bottleneck for speed. For example, PRT (described in Pre-AE2 Autocrafting) has very low pipe speed (beyond not being TPS friendly at scale), so it would not be appropriate for a late game ore processing system.

[Lag Friendly]

Your choice of logistics system should not bring the server to its knees. This matters especially for late game players.

[Minimal Human Intervention]

You want to be able to throw everything into an input chest, and receive finished clean dusts and/or ingots in your output with zero intervention. You should only touch your ore processing system for upgrades or debugging issues.

Overview of Designs

placeholder

Filter Options

placeholder

[MV+] EnderIO Conduits - The "Central Line"

This shows the simplest version of a "central line" ore processing system. It has one route (priority -2) and is currently useless for anything practical. Gold chest = input and copper chest = output.
This shows the second most basic EIO ore processing setup - a central line with an attached macerator round robin module.

EnderIO conduits are your first and cheapest option for creating a truly flexible ore processing system. This system is built around a "central conduit line" which all ores go through, and connected "modules" route ores into machines. Because of its modular nature, it is easy to add new routes or expand old ones. This is made possible due to the EIO conduit "priority" system, which allows us to make override routes.

The priority system for the "central line" design goes as follows:

Priority 0: Override routes. For example, specific ores, rare earths, etc.

Priority -1: Type filter routes. These are your "default" paths that are taken if nothing overrides them.

Priority -2: Output routes. This goes to the final output chest.

The simplest version of this is an input chest set to extract always, some conduits inbetween, and an output chest at -2 priority. Then everything you input (ores, random tinker tools by accident, w/e) will go to the output chest. But obviously this isn't very useful.

So you start adding "modules" to your inbetween conduits. These provide routes for ores to go. Let's start with a macerator module.

[The Macerator Module]

For every module, you must choose two things: the filters and the machines. For a macerator, we want to automatically send anything that fits these characteristics:

  1. ore
  2. crushed ore
  3. crushed purified ore
  4. crushed thermal centrifuged ore

GT type filters are perfect for this, so let's add four of those and configure appropriately.

Now I will add a buffer chest - this will hold items routed by EIO, but not yet processed by machines. If the buffer chest gets full, items will go to the next available route - in this case, the output chest with priority -2. This can lead to undesirable behavior with very large inputs - for example, if the Rare Earth macerator buffer chest gets overfilled, it might accidentally get routed to the normal macerators instead (which is bad). Unfortunately, there are no easy fixes for this besides adding bigger buffers.

Now I round robin extract from the buffer chest into my macerators. EIO lets you easily select this option by just hitting a toggle in the menu. For the connections to the macerators, I set input on green (from my buffer chest) and output on red. This programming can get a bit tedious, so I recommend making a conduit probe whenever you can - this lets you copy conduit settings.

A medium sized EIO ore processing system. The system can be further expanded by increasing the length of the central line (horizontally, vertically - however you want).

Finally, I add a "transfer chest." This is just so that my macerator round robin logic doesn't accidentally get sent to the rest of the central line.

[The Rest of the Line]

Every module you add will look almost exactly like the macerator module. The only things you will swap out will be different filters and machines in place of the macerators.

For specific item overrides (eg. I want crushed bauxite specifically to go to thermal centrifuging), remember to use priority 0, as this will override items going to the normal macerator route we defined above.

On the right you can see an example of a fully fledged EIO ore processing line.

[How Desirable is the System?]

  1. Flexible: S Rank. It is very easy to add new modules, and they can be added in any configuration (before or after existing modules) without breaking things. Filters are very explicit and can be examined visually with NEI-like images.
  2. Expandable: S Rank. All that you need is a space for item I/O, and EIO lets you do input and output from the same side. If you need fluids later, like for rare earths, you can bundle an ender fluid conduit inside the item conduit.
  3. Unbreakable: A Rank. While this system by default is very robust to random items thrown in, like backpacks and tinker tools (these will automatically go to the output chest since they have no routes), the ability for logic to break down if buffer chests fill is a major weakness. While you can always add more chests, it is never able to be "truly" solved - a large enough input can still break your system.
  4. High Throughput: B Rank. Item conduits are by default quite slow (4 items every 0.8 seconds or so). You can upgrade them up to 64 per 0.8 seconds, but this is still much too slow for endgame. You can use conveyors into conduits to get a sufficiently high throughput, but using conveyors everywhere is relatively expensive and annoying compared to just using AE interfaces.
  5. Lag Friendly: A Rank. EIO conduits are fine unless you're stargate tier. They are pretty optimized, just not to the level of AE.
  6. Minimal Human Intervention: S Rank. EIO Conduits are perfect for no intervention ore processing. You only touch the system when adding new modules or doing new upgrades. If you accidentally drop your tinker tool inside, you can simply collect it from the output chest instead of having to figure out what multi input bus it's inside.

[Other Stuff to Remember]

  1. Although I made a "linear" looking central line in the images above, you can actually put the output chest anywhere. In practice I tend to put the output chest right next to the input chest to make interfacing with other logistics (eg. AE) easier. This is because the priority system doesn't care where an inventory is, only that it follows the priority rules.

[EV+] Applied Energistics 2

placeholder

[LV+] Limited Flexibility GT Pipes

placeholder