Factorio Forums

Coal is finite, but not oil

Here I want to talk about another infinite source of energy: Oil -> Solid Fuel -> Boilers -> Steam Engine cycle
Now we will calculate everything and find the perfect set up. Make your choice.

You might want to have a look at one of the similar posts we get regularly :
viewtopic.php?t=7052
viewtopic.php?f=5&t=32413

Pumpjacks : Oil Refineries : Chemical Plants
100000 : 5000 : ~20200 -> 1000 : 50 : 202 -> 500 : 25 : 101 -> 20 : 1 : ~4.

No, you can't. You are need at least 7 + 13 + 1 chemical plants, then you are need 5 oil refineries. Less then 1 chemical plant for cracking heavy oil is not effective.

No, you can't. You are need at least 7 + 13 + 1 chemical plants, then you are need 5 oil refineries. Less then 1 chemical plant for cracking heavy oil is not effective.

Notice that the format is Pumpjacks : Oil Refineries : Chemical Plants. You have to use 20 pumpjacks, 1 oil refinery and 4 chemical plants.
Well, after doing some tests in actual game I think you need to change this slightly: you can't use 4 chemical plants because you can't use two recipies at once. So after balancing everything out a bit, I had to use 1 chemical plant for heavy oil -> light oil, 3 light oil -> solid fuel and just one petroleum gas -> solid fuel to obtain 100 solid fuel / minute (which corresponds to needed 1+2/3 solid fuel a sec). I can calculate proper ratio with Chemical Plants used for cracking heavy oil, making solid fuel out of light oil and petroleum gas -> solid fuel, but the ratio is probably going to be unreasonably huge. If I could somehow make chemical plants to change their recipe depending on whether there is enough heavy oil, light oil or petroleum gas, but it's impossible.

I'll try it anyway.

In my experiments with oil-fired steam, I used 10 refineries with adv refining. I then had if I recall correctly, 2 heavy cracking 18 light -> fuel blocks and 14 gas -> fuel blocks.

Gotta load up that save again at some point and re-count the ratio. But it made neat little blueprint and go fuel factories that only needed to be connected to an abundance of oil and something to consume all the fuel blocks.

n my experiments with oil-fired steam, I used 10 refineries with adv refining. I then had if I recall correctly, 2 heavy cracking 18 light -> fuel blocks and 14 gas -> fuel blocks.

Gotta load up that save again at some point and re-count the ratio. But it made neat little blueprint and go fuel factories that only needed to be connected to an abundance of oil and something to consume all the fuel blocks.

I don't understand why you all need so many refineries. I am talking about 0.1 oil/s yield here with no speed modules whatsoever for reason stated in OP post. You simply won't have nearly enough crude oil to need so many.

You might want to have a look at one of the similar posts we get regularly :
viewtopic.php?t=7052
viewtopic.php?f=5&t=32413

The topic by first link was created in 2014. Also, it only showcases one possible design and doesn't really go much into detail about the whole idea.
Second one was created recently, but unfortunately

No modules

whereas I tried to tackle this thing with modules and prove that efficiency modules lead to the best efficiency possible.
And still don't go in much detail.


Also, I seem to have made an error: I didn't account for boiler efficiency of 50%.
Which is annoying because it halves the efficiency. I need to redo some calculations.


UPD: I updated my calculations. Now they should be accurate. I still have to deal with "4 chemical plants" problem though.

EditorRUS wrote:

n my experiments with oil-fired steam, I used 10 refineries with adv refining. I then had if I recall correctly, 2 heavy cracking 18 light -> fuel blocks and 14 gas -> fuel blocks.

Gotta load up that save again at some point and re-count the ratio. But it made neat little blueprint and go fuel factories that only needed to be connected to an abundance of oil and something to consume all the fuel blocks.

I don't understand why you all need so many refineries. I am talking about 0.1 oil/s yield here with no speed modules whatsoever for reason stated in OP post. You simply won't have nearly enough crude oil to need so many.

I'm not using a single well. I worked out the ratios for a map where coal was scarce, but oil was in tremendous abundance.

And I'm not afraid to lay down a bit of plumbing. 0.1/sec is nothing as far as pipe flow goes, its nothing more than a ho-hum matter to string together a whole bunch of wells. Or better, barrel up these isolated well sites and bring those to the refinery that feeds the power station. But I stumbled upon a nice compact all-to-fuel refinery string based on 10 refinery sets, that can be easily tiled and will support in the neighborhood of 400 steam engines for each set without issue.

EditorRUS wrote:

No modules

whereas I tried to tackle this thing with modules and prove that efficiency modules lead to the best efficiency possible.
And still don't go in much detail.

When talking about efficiency, you must remember to state what it is with respect to.
Given that we're dealing with power generation and limited oilwells, I would expect the design goal to be highest net energy off a limited set of oil wells.
If this is the case, then speed modules (in the pumpjacks) and productivity modules (everywhere else) really are the optimal choice, as they vastly increase the output of solid fuel.

In the OP there's one critical statement, which you seemingly haven't considered too carefully.

EditorRUS wrote:Let's first calculate the energy cost for a unit of crude oil. Fortunately this is easy. Regular pumpjack consumes 90 kW a second and produces 0.1 units of oil a second, so 900 kJ of energy is spent to produce one unit of crude oil.
Adding 1 speed module 1 to the pumpjack increases oil production to 0.12 oil a second and energy consumption to 135 kW, which translates to 1125 kJ of energy. Speed Modules 1 are useless here.
1 speed module 2 makes it ~1108 kJ/unit of crude oil, still not good.
1 speed modules 3 gives 1020 kJ/unit of crude oil.
So the most efficient approach is actually having pumpacks with no speed modules.

The power loss from processing and production is minuscule compared to the gain from solid fuel, and the pumpjacks themselves only make up a fraction of this cost.
Now if I'm reading your calculations right:
A depleted pumpjack with a -80% power modifier produces oil worth an average 8 solid fuel every 100 seconds, costing 383.55 kJ per solid fuel, at a net of 96.93 MJ - or 0.969 MW worth of power production per pumpjack.
A depleted pumpjack with 1 level 3 speed modules produces worth an average 12 solid fuel every 100 seconds, costing 1433.5 kJ per solid fuel, at a net of 132.79 MJ - or 1.32 MW worth of power production per pumpjack.

Your conclusion (that efficiency modules in pumpjacks are most efficient) only makes sense if you want to optimize power per unit of crude oil, which in itself is a backwards design goal.

EditorRUS wrote:Coal is finite, but not oil
Let's first calculate the energy cost for a unit of crude oil. Fortunately this is easy. Regular pumpjack consumes 90 kW a second and produces 0.1 units of oil a second, so 900 kJ of energy is spent to produce one unit of crude oil.
Adding 1 speed module 1 to the pumpjack increases oil production to 0.12 oil a second and energy consumption to 135 kW, which translates to 1125 kJ of energy. Speed Modules 1 are useless here.
1 speed module 2 makes it ~1108 kJ/unit of crude oil, still not good.
1 speed modules 3 gives 1020 kJ/unit of crude oil.
So the most efficient approach is actually having pumpacks with no speed modules.

Using efficiency modules is a must to increase our, well, efficiency.
The limit is 20% of initial consumption. If we are to use only efficiency modules in the pumpjack, we will get 180 kJ/unit of crude oil, which is almost ridiculous.
If we try to use 1 speed module 1 and 1 efficiency module 3, then we get a flat increase of 20% in oil production for the same price of 900 kJ of energy, but that's still only 750 kJ of energy per unit of crude oil. Not efficient.
Therefore we conclude that we have to use 1 pumpjack with two efficiency modules 2 or 1 efficiency module 3 and 1 efficiency module 1. That gives us 180 kJ / unit of crude oil.

There is one huge misstake you make in your assumptions here:
The relevant value is not how much energy do you get per invested energy, but how much net power you get, which is net energy per timeunit, since as you already are saying: Oil is infinite!
So the speed modules are actually better than you think!
Even if you reduce the net energy per Unit of oil, you still get more enery out of it than you put in there:
Speed module 1:
For every 5 Units at 900kJ you produce 1 Unit effectively at 6.75MJ which even after postprocessing is a lot less than the 12.5MJ you gain from burning it.

Efficiency/Productivity modules actually hurt you energy production wise I'd say, since even though you might effectively pay less energy per block of fuel, you get less of them per tick after energy costs. And that is the only reasonable metric here. On later stages in the the refinement that does not necessarily hold true though, since you easily can build a second refinery for example doubling throughput and energy cost, which you can not do for the pumpjack (at least for this kind of analysis).

So your whole analysis falls down.
Stm