Effective Power Generation

[rlerner]

I have been doing some work on a map to determine exactly the best combination of power generation possible for most utilization. It has a 540 radar load cell attached to steam turbines which is powered by boilers, inserters, and drills on a separate network. The point of this setup was to see exactly what I can crank out of an engine.

What did I come up with?

Each Engine requires 1.5 boilers to operate at 100 degrees at full load.
Each pump can handle up to 10 steam turbines (the pipe goes dry outside of that range)
You can have a maximum of 26 engines on a single circuit before the performance degrades at engine 27, and finally drops to nothing at 28+

So, here is my approximate best power generation configuration:
16 Electric Drills providing to 39 Boilers, using 3 pumps, and 26 engines.

Yield:

23 Radars can be powered with "full electric demand satisfaction"
45 Radars can be attached before a lamp begins to indicate no power on the same circuit.
Over 540 Radars can be attached, and utilized, but they will run very slow (guessing close to 600 for them to stop due to lack of power, but got bored building them).

The power setup above uses two power plants, the 26 engine one mentioned, and an additional one with 9 boilers, one pump, and 6 engines to power all the inserters and drills. So, assume this loss when building.


Overall, building this configuration will allow the maximum yield out of a single pipeline of water at maximum utilization. Unless you're building something crazy, this configuration should power what you're building without an issue (minus occasional laser spray).

Any other interesting power metrics? I've heard there's no "water pressure", but I'm finding that isn't the case.

[rlerner]

Posting separately since I see people being interested in the save for this, you can download it from here, and unzip into your saves folder:

http://www.robert-lerner.com/factorio/energy_test.zip (1.75 MB)

[Aza-Industries]

Nice work, I was thinking of doing this eventually but never got around to it.

My last steam setup used 26 engines and 3 pumps, I just tweaked it to those levels during gameplay and seeing how it was going.
I don't remember how many boilers I had though.

This sort of information would be good to compile in a single location.
I'm thinking of making a excel sheets with resources construction amounts and speeds to help work out how many of X I need to produce for Y to be made at a steady rate, or something like that..

[rlerner]

Wouldn't be a bad idea, I thought of compiling a list, but didn't jump on it since it seems they still plan on doing a large "re-balance" of the game. At least this info should help with an optimal configuration until they change it.

Awesome how you stumbled upon a similar configuration as well

[MF-]

What about putting it up on factorio wiki?

[ssilk]

This sort of information would be good to compile in a single location.
I'm thinking of making a excel sheets with resources construction amounts and speeds to help work out how many of X I need to produce for Y to be made at a steady rate, or something like that..

I would go further: it needs some kind of program to calculate that. Because the real need depends on too many things to create a fixed table (ok, brings good estimations, but is not exact).

[FreeER]

This sort of information would be good to compile in a single location.
I'm thinking of making a excel sheets with resources construction amounts and speeds to help work out how many of X I need to produce for Y to be made at a steady rate, or something like that..

I would go further: it needs some kind of program to calculate that. Because the real need depends on too many things to create a fixed table (ok, brings good estimations, but is not exact).

Seems like a simple enough program, we would just need to know exactly how much power an optimal steam engine would provide (or a rough comparison to the solar panel. Assuming the energy_consumption is the energy required to run at full power (otherwise a single solar panel would be unable to power a radar even partway). It'd be relatively simple math.

Tested:
1 steam engine powers
84 radars enough to remove the electric symbol (presumably 1% each) --note inserters do not receive power if you have this many hooked up to 1 engine lol
thus 84% of 1 radar

thus a steam engine at max performance would produce 504 power/second (600*.84) or 8.4 power/tick
for comparison a solar panel produces 60 p/s or 1 p/t.

Thus for any machine, well let's walk through a mining drill:
An electric mining drill uses 1.5 p/t
thus 1 steam engine can fully power 5 mining drill (and 60% of a 6th, though I believe the deficient would be evenly spread to power 560/6=93% of each drill)
If I wanted to fully power 200 mining drills I would need 36 steam engines, since I can only put 26 in a circuit I'd need two circuits. One of 26 the other of 10 (and can add it it as needed later lol)

the code to calculate would be something like

Code: Select all

Given 200 (x) of mining drills (y=look up power use=1.5)
z = x*y*8.4 --thus 200*1.5/8.4=36
print("You will need " .. z .. " steam engines to power " .. x)
while z > 26
circuits=circuits+1
z=z-26
end
print("You will need " .. circuits .. " circuits of 26 steam engines. Each circuit needs 39 boilers and 3 pumps")
print("You also need an additional circuit of " .. z .. " steam engines. While each steam engine only requires 1.5 boilers you may as well create the full 39 and add steam engines to complete the circuit later")
print("Do not forget that you will need to power " .. math.ceil(circuits*39+(z*1.5)) .. " inserters for the boilers as well")

Admitedly this may not be 100% accurate but I think it's somewhat close

[ssilk]

This is what I mean. Depending on what you power and which kind of power you use, you can make very different calculations. This very interesting. The time-factor makes this absolutly unpredictable. This is also good, cause I have played too much predictable built-up games. The advanced player can make very good estimations. This is good.
The disadvantage is, that the new player doesn't know how to play it, cause there are so many options. For example, my first big factory had only one ammunition factory. That was much too low. For the next I build 10. And hundreds, no thousands of turrets. That was a little bit too much, cause my iron-ore gone out for that.
My first factory of blue potions produced in the end one blue potion per second. That was also too much.

So, what I mean is some help for calculating the right amounts of anything.

[FreeER]

Depending on what you power and which kind of power you use, you can make very different calculations. This very interesting. The time-factor makes this absolutly unpredictable.

Not really, I mean obviously if you are powering radars rather than mining drills it will make a huge difference (10 vs 1.5 p/t), especially if you let the belts backup so that the drills are not constantly running. But the only real difference is rather you use steam engines (with enough coal/water) or solar panels, which is simply a matter of if you want to work with setting things up or using more resources for something that you can just plop down and hook to the grid

The disadvantage is, that the new player doesn't know how to play it, cause there are so many options.

Which makes it more fun to replay and see what you can improve on. Of course I haven't played the tutorial in awhile. Maybe I should go back and see if there is anything I think should be improved upon. But it's easy enough for new players to hop on the forums and check out all the amazing designs on here, I know I've learned quite a bit

So, what I mean is some help for calculating the right amounts of anything.

Well if someone really felt like it they could go into the prototypes lookup the research speeds (and take an average), that would give you how many packs you'd need per cycle * number of labs (which can easily be figured out quickly, I personally use 10). Then it's a simple matter of looking up how much time it takes to craft each pack and creating enough assembling machines to do so. The rest is simply supplying resources. It's not difficult simply time consuming and for me decreases the fun I get out of working it out while playing The same can be done for everything else as well. With exact measurements, since the recipes do not change (99% of the time anyway). Timing can be worked out fairly easily as well. It is perfectly possible to work out a spreadsheet for how much ore and gears,electric circuits, etc etc. you'd need to make a factory, I just wouldn't want to be the one spending the time on it lol. Of course if I did make an excel work sheet I'd want to set up forumlas so that I wouldn't have to look through and work it out myself (hint hint Aza )