Should you use Basic Refineries for stone?

Why you should not use the Basic Refinery if inefficiency is a thing that makes you uncomfortable

NOTE: Survival Kit numbers to be revisited

In a recent 1-1-1-1 playthrough I noticed that the survival kit seemed awfully much slower than what I recorded here, so I did a new test run because there was one important thing that didn't occur to me when I originally tested for, and wrote, this guide:

The Survival Kit is not a refinery, it's an assembler.

It functionally "assembles" ingots out of stone, which I have to assume was the only way Keen could manage to bodge both a refinery-style and an assembler-style conversion into the same block within the existing engine and its UI/UX limitations. (That's why you have to cue up those ingot bags, where a real refinery goes to work on its own.) As such, it is not affected by the refinery multiplier but the assembler multiplier instead. (See footnote on multipliers.) Meaning, as the original test was done with the default 3× assembler multiplier, the SK figures cannot really be blindly compared to the BR and R figures, and so I'll now need to rework and update the tables and charts.

Does that mean this entire guide is void?

No. The main purpose of this guide has always been to pit the performance (or arguably the lack thereof) of the Basic Refinery against the regular Refinery and debunk with hard numbers a long-standing misconception about the BR being The Better Choice (TM) for stone.

The numbers for BR and R are still correct and true, while the SK has always been included mostly for the sake of completeness.

I'll get around to updating the figures and charts eventually. For now, take the SK figures with the grain of salt described above.

Now back to the show. (BR still sucks. #Y4L)

Why this guide?

I kept seeing people rely on the Basic Refinery as their first, last, and eternal choice to deal with stone that accumulates during mining short of just sorting and ejecting it out of sight, and not just in the humble beginnings of the starter base but as a permanent solution.
And because this is wrong and bad, I've wielded the magic power of spreadsheets to let the world know exactly why and where and how, once and for all.

But mostly, in case my propensity for OVEREXUBERANTLY EXCESSIVE HYPERBOLES has clouded that purpose, it is to have a reliable listing of actual tests and test results, some hard data you can actually refer to instead of just vaguely knowing from hearsay or estimation.

Note: This guide is aimed at refining efficiency, not space efficiency, not refining effectiveness, not at circumstantial requirements such as progressing through the tech tree or bootstrapping a new base of operation. I realise and happily acknowledge people have individual playstyles and preferences, can and will encounter individual challenges for which one solution is more attractive or convenient than another, or just don't mind, but those individual cases are not scope of this guide. By all means, you build and play however you want.

Corollarily, if, in the presence of this guide, you insist on explaining why this guide is wrong (unless I made a factual error that needs correcting) or why the BR should be used after all, you willingly demonstrate and declare you have neglected or ignored to read any of this and that you thereby accept and invite any and all ridicule over said negligence or ignorance. You have been warned.

Quick facts

There are three blocks in SE that turn straw to gold stone to ingots. Each has its own immediately obvious (and relevant!) pros and cons:

• Survival Kit
  
  • Pro: You get a free one from most respawn options
    
  • Pro: Small
    
  • Pro: Small-grid variant available, with effectively only symbolic a difference in component cost
    
  • Con: Cannot refine what's needed to build another one
• Basic Refinery
  
  • Pro: Fairly cheap
    
  • Pro: Still fairly small
    
  • Pro: Can refine what's needed to build another one
• Refinery
  
  • Pro: Refines everything, including what's needed to build more (obviously)
    
  • Pro: Performance can be tuned with upgrades to suit situation
    (Anyone using anything else but Yield for only a second longer than needed to get the power supply in order is a fool and should be banned from playing any computer game for the rest of their existence.)
    
  • Con: XBOX HUEG
    
  • Con: Expensive

So, with the Basic Refinery apparently having no immediate cons, why the fuss?

SCIENCE!!

1. Observation

There are three significantly different blocks that are capable of the same job. They have different sizes, costs, performance properties, and names that hint at their specialisation, or lack thereof.

2. Question

Are they still equivalent in their performance at converting stone to ingots?

3. Hypothesis

Potentially not.

4. Prediction

They are potentially more efficient the more specialised they are. Then again, this is a KeenSWH game, so, technically, this prediction is as good as reading tea leaves.

5. Test

Here's our test setup.

• Space Engineers 1.199.025
  
• World's Refinery speed setting at Realistic.
  
• A fresh battery with known remaining capacity
  
• A cargo container, containing 100 tonnes of Stone
  
• The refining block in question, switched off
  
• Two timers
  
  • Timer 1 turns on the refining block and starts Timer 2.
    
  • Timer 2 turns off the battery after 600 seconds, shutting down the entire rig.
    (That's two-hands-worth-of-fingers minutes for you weirdos still counting in body parts and cereal grains.)

Technically, 5.5 times 'cuz that-them-tharr-y'all's pile are twenty US short tons, not metric ones.

Test results: Energy usage

	Energy started with	Energy remaining	Energy used	Power drawn
Unit	kWh	kWh	kWh	kW
Survival Kit	900	866.58	33.42	200.52
Basic Refinery	900	844.85	55.15	330.90
Refinery, 0 Yield modules	900	805.88	94.12	564.72
Refinery, 1 Yield module	900	806.40	93.60	561.60
Refinery, 2 Yield modules	900	806.66	93.34	560.04
Refinery, 3 Yield modules	900	806.40	93.60	561.60
Refinery, 4 Yield modules	900	806.40	93.60	561.60

So, the Basic Refinery sits between the Survival Kit and the standard Refinery. But without seeing what it actually did with that energy in those ten minutes, that's not really saying anything yet.

Also, there is some measurement noise apparent in the consumption for the Refinery.
Power draw was calculated from energy draw, assuming averaged continuous operation, not read from terminal.

Bonus observation: Speaking of the Refinery, using Yield modules doesn't actually increase energy consumption. Bug? Lazy "Efficient" Keen implementation? Intention? Who knows. At least those naysayers who use anything other than Yield modules on Refineries now have one excuse fewer.

Test results: Stone consumption

	Stone started with	Stone remaining	Stone consumed	Consumption rate
Unit	kg	kg	kg	kg/s
Survival Kit	100 000.0	46 000.0	54 000.0	90.00
Basic Refinery	100 000.0	62 560.0	37 440.0	62.40
Refinery, 0 Yield Modules	100 000.0	24 371.2	75 628.8	126.05
Refinery, 1 Yield Module	100 000.0	25 120.0	74 880.0	124.80
Refinery, 2 Yield Modules	100 000.0	25 120.0	74 880.0	124.80
Refinery, 3 Yield Modules	100 000.0	25 120.0	74 880.0	124.80
Refinery, 4 Yield Modules	100 000.0	25 120.0	74 880.0	124.80

Well, well, well, will you look at that. It was pretty plausible from the start that the Basic Refinery would be slower than the dedicated big brother, but that it managed to be slower than the Survival Kit still, and by a massive margin at that, is a surprise. One would think a block at least notionally more specialised in refining ought to deliver at least somewhat more attractive performance figures, but I guess thinking is not the thing to do when it comes to SE.

That the Refinery with zero upgrades scores a smidgen faster is likely just measurement noise again.

Test results: Efficiency A

Note: Values listed using Wh, not kWh, to keep the numbers short.

	Energy used	Stone consumed	Energy efficiency	Energy efficiency
Unit	kWh	kg	Wh/kg	kg/Wh
Survival Kit	33.42	54,000.0	0.619	1.616
Basic Refinery	55.15	37,440.0	1.473	0.679
Refinery, 0 Yield modules	94.12	75,628.8	1.244	0.804
Refinery, 1 Yield module	93.60	74,880.0	1.250	0.800
Refinery, 2 Yield modules	93.34	74,880.0	1.247	0.802
Refinery, 3 Yield modules	93.60	74,880.0	1.250	0.800
Refinery, 4 Yield modules	93.60	74,880.0	1.250	0.800

This is looking worse and worse. Not only is the BR slower than any of the alternatives, it also uses more energy for what it does manage to refine.

What a piece of rubbish.

Test results: Ingots produced

	Gravel	Iron	Nickel	Silicon	All ingots
Unit	kg	kg	kg	kg	kg
Survival Kit	756.00	1,620.00	129.60	216.00	2,721.60
Basic Refinery	366.91	786.24	62.90	104.83	1,320.88
Refinery, 0 Yield modules	1,058.80	2,268.86	181.51	302.52	3,811.69
Refinery, 1 Yield module	1,246.67	2,671.44	213.71	356.19	4,488.01
Refinery, 2 Yield modules	1,482.55	3,176.89	254.15	423.59	5,337.18
Refinery, 3 Yield modules	1,763.06	3,777.98	302.24	503.73	6,347.01
Refinery, 4 Yield modules	2,096.64	4,492.80	359.42	599.04	7,547.90

No surprise here. Less stone consumed, less ingots produced. But less by how much?

(Before you ask: All Gravel:Iron:Nickel:Silicon ratios are equal across all conversions. It's always 35:75:6:10.)

Test results: Efficiency B

	Stone consumed	Ingots produced	Conversion ratio	Conversion ratio Iron ingots
Unit	kg	kg	%	%
Survival Kit	54,000.0	2,721.60	5.04	3.00
Basic Refinery	37,440.0	1,320.88	3.53	2.10
Refinery, 0 Yield modules	75,628.8	3,811.69	5.04	3.00
Refinery, 1 Yield module	74,880.0	4,488.01	5.99	3.57
Refinery, 2 Yield modules	74,880.0	5,337.18	7.13	4.24
Refinery, 3 Yield modules	74,880.0	6,347.01	8.48	5.05
Refinery, 4 Yield modules	74,880.0	7,547.90	10.08	6.00

I have no words.

Also, interesting to see that the Survival Kit and the Refinery have the same conversion efficiency.

Conclusion

So, should you use the Basic Refinery for stone?

6. NO!!

(Unless you want to. Who am I to tell you otherwise? I really don't care. I just write spreadsheets and snarky guides.)

The only excuses this sorry piece of pixel pile in the G-menu has for existing next to the proper Refinery are its advantages in size, initial cost, and raw power draw (330 kW vs 500 kW), but as soon as that Refinery can be afforded, this pathetic sack of junk should be erased from anyone's consideration. Once superseded, it is literally and demonstrably a waste of time, energy, and material.

♥♥♥♥ this thing.

Links, acknowledgements, notes

• Spreadsheet: https://docs.google.com/spreadsheets/d/1YaP-QNtsi1x9MH7xDFz6V_uycfXCo-OoVIDqK3yWvBM
  
• Basic Refinery 3D model: Keen SWH s.r.o.
  
• Gravel pile: Some guy getting a truckload of gravel delivered, https://www.youtube.com/watch?v=9oZU_s3qVu0
  
• Granny punt GIF: no foggy idea; it's literally older than some of the people who are going to be reading this
  
• Regarding multipliers: Unlike the refinery multiplier, which is a speed multiplier, the assembler multiplier is actually a divisor; it reduces the amount of ingredients needed to assemble a unit of a product as a fraction of the "uncut" original amount that was specified in the assembler "recipe" for that product. (See blueprints.sbc.) Evidently, yes, as long there is a "recipe" or "blueprint" defined appropriately, there is nothing that prohibitively segregates ore, ingots, and components, at least as far as conversion between them is concerned.

[Lama]

Great argument, i usually replace them myself too

[warrat106]

did you have fun with this experiment? it seems like it'd be dull watching stone refine into ore. all sarcasm aside, nice job.

[orion_blackstorm]

people use basic refineries for stone? I've always used survival kits

[Demon47G2327]

This is hilarious and helpful Thank You.

[HybridGrizzly]

I disagree on the complete superiority of yield modules, I have a drilling rig set up that makes stone effectively free, so I have a couple of refineries that can only take stone with speed modules

[andersenman]

Rounding; see spreadsheet for full precision.

[Mediterranean Barbarian]

Total stone consumed for Refinery should not have a decimal because the numbers multiplied are even numbered, being 126.05x600=75,630.0 I don't know why you have an 75 628.8.

[AutoMcDonough]

you're in an area where there's no right or wrong answers, if your priority is maximizing the ore you've mined then go with all yield modules. If you are swimming in ore and would rather just blast thru it asap then speed. I like to use efficiency with uranium just because the point is making more energy but it still positive energy with other modules.

[condottiere]

Googling my own topic shows Fe, at least, caps out at basically 2 speed modules. So 3-4 yield modules and 0-1 speed for low to very low yield ores, versus inverted modules for basic ores. Yes? No? :D Thanks for the fun and quantified guide plus provoking my line of inquiry!

[condottiere]

I enjoyed the snark, but as a filthy SE casual who's only now starting to look up guides for best practices and neat ideas... <raises hand>

Beyond scope of this guide, why are 4 yield modules overwhelmingly superior to 4 speed modules on a refinery? I totally understand for the, well, yield-per-mined-ore slash labor time. One full yield refinery outputs approx 2x the ingot output for same ore input quantity.

OTOH one full speed refinery = 5 refineries for PCU cost of 1 refinery. You'd need 2x the ore input for 2x the ingot output, but you'd get it done in 40% of the time. Minimizing mining labor effort could make this a useful time-vs-efficiency trade-off.

Unless speed modules ADDITIONALLY reduce efficiency?