Thoughts on ore processing

[septemberWaves]

Note: I'm using the phrase "imprecise sorting" to refer to sorting of the ores that spawn on the map (jivolite, saphirite, etc.) to a variety of pure ore outputs, and the phrase "catalytic sorting" to refer to the combination of several different types of those major ores with a catalyst (mineral or crystal) to produce a single desired type of pure ore.

When backreading through one of these discussion threads on Angel's mods I noticed some discussion regarding crystal catalysts and how geode production from flotation of ores is generally insufficient to produce enough crystal catalysts for ore production. It got me thinking about ways of doing ore processing and I've realized that catalytic sorting is probably best done as a secondary supplement to imprecise sorting of main ores, rather than as the only way to produce pure metal ores.

I'll have to do some testing in my survival game when I have time to figure out exactly what kind of system is needed, but what I have in mind would involve imprecise sorting of the main ores at various tiers of refinement (chosen depending on requirements of the higher-tier ores; probably will vary between factories but I would estimate that variation in production requirements will be predominantly dictated by secondary mods that aren't integral to Angel's and Bob's mods). Each pure ore would be combined together to a single belt (or a belt array, if the factory is large enough) to be sent to the smelting facility for that ore. For cases when one ore type is backlogging and causing the primary sorting machines to stop producing their other metal ore products, the output belt of any given metal ore will be measured across several belts (I don't know how many are optimal; it probably depends on the production speed of the system), and a localized circuit network will determine whether to enable the catalytic creation for that ore type. In case geode production still isn't enough (which I think is unlikely provided geodes and crystal catalysts are stored in bulk), a secondary geode input from geode washing will be used on a lower priority.

I'll update this post when I have both the time (irl) and the resources (in-game) to test this system; I think it'll be quite interesting regardless of the results.

[steinio]

You get endless geodes from washing plants if that is what you worry about crystal catalysts.

Mineral catalysts from crushed stone should be sufficent. Don't know.

[septemberWaves]

You get endless geodes from washing plants if that is what you worry about crystal catalysts.

Mineral catalysts from crushed stone should be sufficent. Don't know.

Regarding geodes from washing plants: "In case geode production still isn't enough (which I think is unlikely provided geodes and crystal catalysts are stored in bulk), a secondary geode input from geode washing will be used on a lower priority."

As for mineral catalysts, to be honest they'll probably be sufficient just from slag (and if not, most of my hydrogen production will probably come from dirt water electrolysis purely for the sake of having excess slag in case it is needed). But the point of the system I described is to minimize catalyst requirements by sorting the main ores individually for the majority of production.

[Arch666Angel]

The efficiency in refining comes from using all products and maybe use paths that dont require supporting processes. But as I always say it's up to you to find the best way that's suiting your playstyle and complexity you want to handle.

[Light]

When backreading through one of these discussion threads on Angel's mods I noticed some discussion regarding crystal catalysts and how geode production from flotation of ores is generally insufficient to produce enough crystal catalysts for ore production. It got me thinking about ways of doing ore processing and I've realized that catalytic sorting is probably best done as a secondary supplement to imprecise sorting of main ores, rather than as the only way to produce pure metal ores.

I've always started with imprecise sorting and moving into a mixed process like you're describing here. However, I'll reach a point where I'll make a sizable washing/filtering section right beside the pure ore sorting and phase out the imprecise sorting in favor of pure only.

I've played maps with impure only and mixed only which do work just fine, but as Angel said it really depends on your style. From my experience I've found that none of them are really right or wrong. Purity usually comes at a cost of more resources and/or effort than impure, so it's only natural the same applies here. With that said, mixed is probably the better fit if you don't want to dedicate a section just for major catalyst production and use only what you have.

[septemberWaves]

With that said, mixed is probably the better fit if you don't want to dedicate a section just for major catalyst production and use only what you have.

This is pretty much what I have in mind. I don't want to go over the top with catalyst production and require catalyst production for ores when it can be avoided. On top of that, I'm using MadClown01's addons, and those add some ores to the world which aren't all used in catalytic sorting so some of them must be done the imprecise way in order to be used at all, meaning a mixed system is somewhat forced with these addons (although this isn't what led me to consider this idea, I noticed this after). I also definitely want to try using the ferrous and cupric ore mixes since they appear to be extremely efficient if refined fully (if my calculations are correct, you end up getting a total of three assorted pure ores per two basic ores, which with final tier processing means three ingots per ore). Those can't use catalytic sorting either, but the high efficiency means prioritizing them is probably a good idea.

[Light]

I'm using MadClown01's addons, and those add some ores to the world which aren't all used in catalytic sorting so some of them must be done the imprecise way in order to be used at all, meaning a mixed system is somewhat forced with these addons.

This mod is to be used on my next map when I find the time for it. Given mixed systems are forced I would be curious to know what MadClown ores you found worth refining, as they may be more worthwhile or cheaper than a pure Angel system depending on what they give. I do know some things given aren't very useful but act as sort of a logistics byproduct you can choose to use if you so desire. They may actually turn more useful if the system is to be run consistently over a long period though, which the mod seems to encourage.

I also definitely want to try using the ferrous and cupric ore mixes since they appear to be extremely efficient if refined fully (if my calculations are correct, you end up getting a total of three assorted pure ores per two basic ores, which with final tier processing means three ingots per ore). Those can't use catalytic sorting either, but the high efficiency means prioritizing them is probably a good idea.

Something I've also never used for more than a short period. If you're right then those processes are certainly very underappreciated and need more love.

You've convinced me to give mixed sorting another try given the updates that have come along since my last attempt. Given I won't have the time for it for a while, I'll be keeping an eye on this thread to see how your experience is doing and if it's as promising as it sounds.

[septemberWaves]

Given mixed systems are forced I would be curious to know what MadClown ores you found worth refining, as they may be more worthwhile or cheaper than a pure Angel system depending on what they give. I do know some things given aren't very useful but act as sort of a logistics byproduct you can choose to use if you so desire. They may actually turn more useful if the system is to be run consistently over a long period though, which the mod seems to encourage.

I'm actually quite early in my current playthrough. I haven't got to a point yet where I've mined anything beyond the starting area aside from saphirite, and I haven't even discovered all of the MadClown01 ores (I'm missing adamantite, elionagate, and sanguinite, which as far as I'm aware are the rarest). I am, however, in the process of assembling my first ore processing system which will use a mixture of methods only up to the refining level of chunks (which means titanium and tungsten will remain out of reach for now, but I also do not yet have a need for them) and without ferrous and cupric ores; this should provide a good enough testing ground to allow me to move towards the full-scale maximum-tier process and figure out the details of the system, as well as gaining a clearer idea of the required ratios of ores.

As for which MadClown01 ores are actually worth refining, I imagine it will depend both on the factory and the ores available on the map. It took me a while to find any infinite bobmonium so I initially thought I might have to initially rely on other ores for infinite tin (although now I do have access to some); in that case some of the new ores would probably be required in larger quantities and different proportions than on a map which, for example, had a shortage of jivolite. The rarest ones (adamantite and elionagate) both produce fluorite, so I think they're both worth refining in any case (and adamantite is the only source of osmium, which is useful in the late-game). Phosphorite is the only ore source of phosphorus (though it can also be obtained from crushed stone), although phosphorus has few applications at present so the ore will probably see minimal use (although it is required for maximum-tier tungsten refining, and I plan on making an addon later this summer which will (among other things) probably give some extra purpose to phosphoric acid) - actually though, it does produce uranium ore, so that's worth taking into account. Alluvial sands are particularly interesting; they're probably a good supplement to lower-tech factories for some ores, because it's one of the earlier sources of quite a few rarer ores (including being I think the only way to obtain chrome before blue science), and the sand is very quick to mine so large-scale sluicing likely requires few miners (though probably quite fast belts for a whole ore patch). Oil sand is another interesting one; I imagine it'll make an excellent supplement to oil production (perhaps even an alternative to multi-phase oil if there's not much oil on the map). One of the ones I've found most commonly is orichalcite, and that's probably going to end up being one of my more important copper sources (secondary to cupric ore mix) as well as producing several other useful byproducts in different ratios (and I think that might be one of the ones that isn't used in catalytic sorting, but I don't have the game open right now so I can't check). That's not all of them, just the ones I remember right now.
Essentially, I don't think any of them are pointless exactly; I think it's worth refining all of them, just the ratios in which they are used will vary depending on what you need (and probably to some degree on the complexity of your ore processing, as I imagine a system which ignores ferrous and cupric mixes will have significantly different ore input ratios than one which uses those mixes).

[Light]

I've found a map to my liking and used the seed for my solo map and the family map. After using a cheat to check the map in a wide area to see what to expect, I notice there's no pro-galena or oil sands to be seen anywhere and most of the ores are incredibly far to the point I'll need to make them into proper outposts with their own power and have 8-12 wagon trains to properly supply the base due to the long trip.

It appears I'll be using mixed sorting of elionagate and phosphorite to obtain thorium ore. I'll need some heavy firepower and 30,000+ rails to break the biter horde and find a proper ore source at some point or another. Even other MadClown ores are just a single tiny patch with no secondary field to be seen. My RSO configuration was set to be sparse for Angel's ores last time, so the inclusion of MadClown's ore patches is quite the step up in rarity and distance to find what I need. This hurts even more knowing there's no infinite ores within some of these tiny patches either, so careful use must be planned before mining until the big fields are found.

The challenge begins.

[MadClown01]

It appears I'll be using mixed sorting of elionagate and phosphorite to obtain thorium ore. I'll need some heavy firepower and 30,000+ rails to break the biter horde and find a proper ore source at some point or another. Even other MadClown ores are just a single tiny patch with no secondary field to be seen. My RSO configuration was set to be sparse for Angel's ores last time, so the inclusion of MadClown's ore patches is quite the step up in rarity and distance to find what I need. This hurts even more knowing there's no infinite ores within some of these tiny patches either, so careful use must be planned before mining until the big fields are found.

The challenge begins.

Glad to see they work just as intended

[mrvn]

I use the following setup (and not just for ores):

Everything is connected with large power poles with wires. Green wire carries what items are stored in output buffer chests. Red wire carries demand for items. So all output buffer chests have a green wire attached. All input buffer chests subtract their content from the desired levels and add that to the red wire. Train stations add to the red wire if they are below one train load and add to the green wire the amount above a train load. Having one train load buffered is their neutral point.

The difference between red and green wires is the amount of each that should be produced. This is the price for an item and can be negative. For each recipe then add up the price of outputs minus the price of inputs. If the score of the recipe is positive then enable the input or output inserters or belts. This will for example stop using the recipes that produce 2 copper ore + 1 iron ore when the excess amount of copper ore has a larger effect than missing amount of iron ore. The 2 iron ore + 1 copper ore recipe on the other hand will keep working a bit longer since the effect of iron is double and the effect of copper half. This will automatically select recipes that are beneficial to the overall availability of items. More efficient recipes get better scores and work longer / earlier.

Modifications:
- Use x * abs(x) as price for items to get a steeper reaction to missing or excess amounts.
- Output items with 0 in the buffer chest get a bonus of 1000000 so "imprecise sorting" produces them even if other products are well in excess. Use this when you don't have "catalytic sorting" for something.
- For a set of recipes (like all ore sorting) take all positive scores and compute the average. Now enable only recipes >= average. This should stop inefficient recipes when a more efficient recipe is able to run even if both would be benefitial.

Now if you use this for ore sorting then I suggest using this for smelting too. Set up normal copper ore smelting and processed copper ore smelting. When copper ore and plates are low the processed copper way has a better score because it gains more plates for less ore. But when copper ore and plates are plenty then suddenly the plain copper smelting becomes better preventing excess copper ore from blocking imprecise ore sorting. Or the recipe for molten iron from just iron ingots or the mixed iron + manganese ingots. The smelter can automatically switch between the two to prevent both iron ore and manganese ore from building up or running low.

[Frost]

I'm using MadClown01's addons, and those add some ores to the world which aren't all used in catalytic sorting so some of them must be done the imprecise way in order to be used at all, meaning a mixed system is somewhat forced with these addons.

This mod is to be used on my next map when I find the time for it. Given mixed systems are forced I would be curious to know what MadClown ores you found worth refining, as they may be more worthwhile or cheaper than a pure Angel system depending on what they give. I do know some things given aren't very useful but act as sort of a logistics byproduct you can choose to use if you so desire. They may actually turn more useful if the system is to be run consistently over a long period though, which the mod seems to encourage.

Figured I add my 2 cents. I am "only" 130h in and I am using the following:
Sanguinate as it can be sorted into pure manganese and nicely supplements the ferrous sorting. It is a rail base and priorities are set by the stations.
Phosphorite: all results of the sorting (slag, phosphorus and lime) are useful, however, by now it is blocked by too much phosphorus.
Oil sand: 130 h in and still haven't set up oil refining. I also produce fuel oil via farming. Oil is next though because I am running short on naphta. Science cost tweaker is brutal.
Alluvium: Saved my ass with some initial gold for blue science. Otherwise I would have needed to build some other mixed sorting.

Next step is nuclear, however, I am not quite sure what needs doing with the madclown rework. To be seen. Alot of tech is gated behind yellow and purple, and those are a bitch with SCT.

[Light]

I'm using MadClown01's addons, and those add some ores to the world which aren't all used in catalytic sorting so some of them must be done the imprecise way in order to be used at all, meaning a mixed system is somewhat forced with these addons.

This mod is to be used on my next map when I find the time for it. Given mixed systems are forced I would be curious to know what MadClown ores you found worth refining, as they may be more worthwhile or cheaper than a pure Angel system depending on what they give. I do know some things given aren't very useful but act as sort of a logistics byproduct you can choose to use if you so desire. They may actually turn more useful if the system is to be run consistently over a long period though, which the mod seems to encourage.

Figured I add my 2 cents. I am "only" 130h in and I am using the following:
Sanguinate as it can be sorted into pure manganese and nicely supplements the ferrous sorting. It is a rail base and priorities are set by the stations.
Phosphorite: all results of the sorting (slag, phosphorus and lime) are useful, however, by now it is blocked by too much phosphorus.
Oil sand: 130 h in and still haven't set up oil refining. I also produce fuel oil via farming. Oil is next though because I am running short on naphta. Science cost tweaker is brutal.
Alluvium: Saved my ass with some initial gold for blue science. Otherwise I would have needed to build some other mixed sorting.

Next step is nuclear, however, I am not quite sure what needs doing with the madclown rework. To be seen. Alot of tech is gated behind yellow and purple, and those are a bitch with SCT.

I've been playing the map with the family for just about the same amount of time and with exponentially increasing science costs as well. However, I've not even touched MadClown's ores since they're deep in enemy territory and my biters hit incredibly hard to be worth the chaos of trying to mine them. Due to this, I've been sorting only the Angel's ores to get everything needed with a massive pure ore sorting facility, so by the time I break the enemy lines the MadClown ores won't do much outside of supplementing what I already ended up building. It's a shame I couldn't try using them early on due to my settings, but maybe in the next year or two I'll be ready to try again since the games last so long and we burn out.

With that said, I have tapped into oil sands as my primary source of oils and alluvium for the gold ores as well, as these two were fortunately nearby. It's incredible how much of a bitch it is to keep steam going strong enough to fuel the miners, but it's much more reliable and faster than using the oil wells which are small and distant. It's to the point I'd rather hunt for the next sands instead of tapping into wells, at least until enough wells can be found to eventually surpass the sands due to their unlimited yields, which speed modules and no steam requirement makes them the top choice late-game.

As for nuclear, just be sure you can find Pro-Galena and start stockpiling so when you score the higher tech you're ready to maximize your nuclear fuel and weapons. (Sadly, the nearest Pro-Galena on my map would require a 3 minute helicopter ride. Which is a shame because I worked with MadClown to expand his tech which does include thorium for better effect.)

[BlueTemplar]

Hi, is it normal that sorting and basic smelting of crushed stiratite/saphirite is worse than their direct smelting ?
Feels like a trap...
Sure, you get extra 12 slag instead of each 5 iron/copper plates you lose, but this slag is only really going to be useful much later !

[xfir01]

Hi, is it normal that sorting and basic smelting of crushed stiratite/saphirite is worse than their direct smelting ?
Feels like a trap...
Sure, you get extra 12 slag instead of each 5 iron/copper plates you lose, but this slag is only really going to be useful much later !

Kind of. If you're putting sorted ore into a furnace to make plates you'll see a drop in efficiency, but if you put it into a blast furnace to make ingots -> casting route, you'll come out ahead.

[anorganicbear]

Hi, is it normal that sorting and basic smelting of crushed stiratite/saphirite is worse than their direct smelting ?
Feels like a trap...
Sure, you get extra 12 slag instead of each 5 iron/copper plates you lose, but this slag is only really going to be useful much later !

Kind of. If you're putting sorted ore into a furnace to make plates you'll see a drop in efficiency, but if you put it into a blast furnace to make ingots -> casting route, you'll come out ahead.

Actually this isn't complete. When you add in the inefficiency of sorting you end up with 0.5 plates per original ore vs 0.67 from direct smelting.

However you also get 1 additional ore from sorting (or 0.25 ore per original ore), which if you actually need that ore, it raises the efficiency to 0.75. But only if you need the secondary ore. It's still less efficient for smelting plates of a single type. This is kind of pedantic, since you can probably make use of the secondary ores pretty easily, at least early game. And later you have better and more efficient methods to smelt what you want. It mostly comes into play for more late-game setups when you have multiple processes going at once and sometimes you want to turn on/off secondary ore production when your buffers start to get full.

You actually don't break even until you reach ferrous ore processing and can start using manganese to help your efficiency (at least in the case of iron, I haven't finished checking out the other ore types yet).

[BlueTemplar]

I specifically mentioned basic smelting (with stone furnaces - before the advanced one with blast furnace).

[Termak]

I specifically mentioned basic smelting (with stone furnaces - before the advanced one with blast furnace).

Since i have discussed about this often on Discord too.

Is there any reason why smelting crushed saphirite and stiratite for iron and copper give 2 times as many plates as bobmonium/rubyte for tin/lead on normal mode. You wont get better yields until t2 blastfurnace recipes or something. On expensive mode this "feature" doesnt exist, which is good.

[BlueTemplar]

The first advanced smelting is already good enough :

2018-09-11_08-04-27.png (74.69 KiB) Viewed 8525 times

One issue that I don't deal with in that screenshot is the need for lots of iron for steel, you can easily go over the 2:1 (B.2) iron:copper ratio if you stay too long on basic steel smelting. That's why I tend to rush advanced steel smelting ASAP (which basically halves the needed iron) - handcrafting those 50 GreenSci (it' like 20 wood = 5 cut trees IIRC ?).

[mrvn]

The first advanced smelting is already good enough :
2018-09-11_08-04-27.png
One issue that I don't deal with in that screenshot is the need for lots of iron for steel, you can easily go over the 2:1 (B.2) iron:copper ratio if you stay too long on basic steel smelting. That's why I tend to rush advanced steel smelting ASAP (which basically halves the needed iron) - handcrafting those 50 GreenSci (it' like 20 wood = 5 cut trees IIRC ?).

From that screenshot I figure:

1) molten-cast 1 smelting is barely worth it: 16 vs. 18 plates total.
2) use molten-cast 1 smelting for your copper needs
3) use the molten-cast 1 iron for steel
4) suplement the remaining iron needs with direct smelting
5) rush towards manganese + iron smelting