pipes - "water flow resistance"?

[gr0mpel]

we now know the the steam engines reduce the water flow ("consume" water).

however i cant explain to myself why the exact same setups of 30 boiler + 20 engines + 3 pumps allways differ from each other in terms of water flow/energy production
the only thing that changes between 5 or 6 rows adjacent to each other is the amount of pipes from the pumps to the first boiler.
and if i connect all eninges at the end ,the water seems to flow to the row of enginges where the pumps are the furtherst away..
so im asking if theres some kind of water resistance or consumption in pipes like in steam engines?
did someone noticed the same effect/behaviour?

[Holy-Fire]

3 pumps? 2 is enough for 20 engines.
Also, I think steam engines are weaker in 0.6 (510W) and thus 20 engines now require only 26 boilers rather than 30.

Can you describe your observation in more detail? It sounds less like pipe resistance, and more like idiosyncracies of where water chooses to flows in branch outs.

Also there is a bug with water temperature when pipes are merged, instead of having a combined temperature sometimes the cooler income prevails. If you've noticed drops in performance this could be a cause.

[gr0mpel]

i will try to upload the savegame or something like that.
luckily you'r able to see in which direction the water flows in pipes.

yes i know that a pump can handle 10 engines. why i placed 3 for every 20 engines? well because even with 3 pumps the last 2-4 engines dont have their full or even no perfomance at all in some setups..

my experience is that the energy production isnt reliable atm.
i build 140 steam engines in a setup with 3 pump 39 boilers and 26 enginges..so i thought - well i have about 70kW to work with
but at 39kW the whole system got f* up and would not produce any more.
so i rebuilt it with 3 pumps 30 boilers and 20 enginges and its still not working 100% as it should


savegame: win 0.6.1
http://uploaded.net/file/qdm15zxq

[Holy-Fire]

savegame: win 0.6.1
http://uploaded.net/file/qdm15zxq

As much as I'd like to examine it you've chosen a horrible download service, try uploading to a reasonable one.

[SilverWarior]

I strongly disadvice using loops of pipe conections. Why? Since not all engines use equal amount of water per every game tick even if they are conected to same power network water presure begins to fluctuate on different parts of your pipe loop which in the end resutls in water on part of your pipe loop which is farthest away from water pipes to actually move back and forth. So the water cols down and therefore doesn't provide necessary heat for steam engines to work.
You can quickly see if this is the problem by simply removing one pipe near the steam engines that doesn't work and therefore breaking the pipe loop.

It is preferably to use branched pipe design instead of looped one.

[ssilk]

I guess: The pipe-system is very similar to the belt-system. And the rules are very easy: The speed is very, very high and the direction of the pipe-pieces are not fixed like on belts, they change depending on "pressure", so that the direction is always from high pressure to low pressure. Low pressure is defined as "hello, I'm a steamengine, I need hot water".

So imagine belts instead of pipes. Then you will see, that circled layouts will have problems.

[gr0mpel]

i kind of solved the problem..
experimented (or: randomly connected pumps and boilers and steam engines) a little bit and suddenly it worked. but only with one specific connection (among 20 others that i made). removing this one connection and it all goes to hell again

[Holy-Fire]

I guess: The pipe-system is very similar to the belt-system. And the rules are very easy: The speed is very, very high and the direction of the pipe-pieces are not fixed like on belts, they change depending on "pressure", so that the direction is always from high pressure to low pressure. Low pressure is defined as "hello, I'm a steamengine, I need hot water".

So imagine belts instead of pipes. Then you will see, that circled layouts will have problems.

If the water flow did not have bugs, there would be no problem with connecting pipes in loops.

[ssilk]

I would not say bug. The whole pipe system, steam etc. has its own physics, like the belts.
It's your job as a player to discover the rules.

I would bet, even the devs will be surprised, when players will find things, which never supposed to be in it.

[SilverWarior]

I guess: The pipe-system is very similar to the belt-system. And the rules are very easy: The speed is very, very high and the direction of the pipe-pieces are not fixed like on belts, they change depending on "pressure", so that the direction is always from high pressure to low pressure. Low pressure is defined as "hello, I'm a steamengine, I need hot water".

So imagine belts instead of pipes. Then you will see, that circled layouts will have problems.

No pipes are not like belts. But yes water always flows from HighPresure point to LowPresure point. and that is what is causing problem with loped designs becouse you have water input on one side of the circle so the water is going to two haves of the circle but on the farther end of this circle water can flow back and forth becouse of constant fluctuations of water presure between the two halves of the circle.

[ssilk]

Ok, the comparison to the belts is a stupid idea. I had this picture of fast running belts which change direction in mind, but high and low pressure is of course more than enough.
Hm. The problem is eventually solved if every piece of pipe etc. would have a resistance (pressure decreases). The fluctuations should stabilize with it, like in reality. But this means, that the steam engines far away from the boilers need a bit more power to run. (?)

[cube]

Actually pipes are more complex than using just pressure difference, (water inertia has to be used as well), and they are currently a little broken (at least I consider them broken, you should be able to make a double pipe without problems )
Unfortunately this breakage doesn't have high enough priority, so it will probably stay for a while.

[Arakasi]

Unfortunately this breakage doesn't have high enough priority, so it will probably stay for a while.

That's a pity. Every time I start new game I forget about that and I am getting angry why it does not work as I expect.

[ssilk]

I'm sure you will fix it in the moment, when for example the oil-industry comes into game. I think that's no problem, because for making steam the current pipes are more than enough.

[gr0mpel]

Unfortunately this breakage doesn't have high enough priority, so it will probably stay for a while.

That's a pity. Every time I start new game I forget about that and I am getting angry why it does not work as I expect.

thats exactly what i feel too
its like: god damn it why you wont ELECRTICITY!!!!

[SilverWarior]

Actually pipes are more complex than using just pressure difference, (water inertia has to be used as well), and they are currently a little broken (at least I consider them broken,

They are not broken. In real life example you would face similar dificulties or shuld I say even more dificulties.
Right now the only way for water to cool down in Factorios is by it being used up by Steam Engines. In real life water would be cooling down even when just standing in pipes.

I just got idea of how you can quickly fix the current water pipes problem. Make another water pump which will be placed betwen two pipes and will make water to flow from one pipe to another. This way you can force water to constantly circle in circular pipe conections. This will prevent water flow fluctuations from even ocuring.

[ssilk]

Don't understand, can you show, what you mean?

[Holy-Fire]

Actually pipes are more complex than using just pressure difference, (water inertia has to be used as well), and they are currently a little broken (at least I consider them broken,

They are not broken. In real life example you would face similar dificulties or shuld I say even more dificulties.

I'm sorry, but they are broken. Of course there are challenges in real life but the behavior of water in Factorio does not remotely resemble the principles of how fluids work in reality, resulting in counterintuitive and difficult to work with phenomena.

[zer0t3ch]

Actually pipes are more complex than using just pressure difference, (water inertia has to be used as well), and they are currently a little broken (at least I consider them broken,

They are not broken. In real life example you would face similar dificulties or shuld I say even more dificulties.

I'm sorry, but they are broken. Of course there are challenges in real life but the behavior of water in Factorio does not remotely resemble the principles of how fluids work in reality, resulting in counterintuitive and difficult to work with phenomena.

That's what makes it so fun!

[n9103]

Don't understand, can you show, what you mean?

He's talking about a forced circulation pump. One that would be inline with the intended flow and one-way.
With water's inertia, the below setup would only flow clockwise after a very short startup time.

Code: Select all

/-E->-E-\
|       |
|       |
\-E-E-E-+-B-P

E Engine
B Boiler
P Pump
> Inline/Circulation pump -or- Check valve

A simple IRL fix is a check valve, but considering how the water currently works, that might not be so simple to implement properly.