Why do liquids move between storage tanks so slowly?

[rollie42]

I had assumed storage tanks abide by the same rules as pipes (high pressure -> low pressure); is this not the case for storage tanks? For example, part of my setup looks like:

Tank1: 22.5k
Tank2: 21k
Tank3: 19k
Tank4: 17.5k
Tank5: 16k
Tank6: 14k
...
Tank14: 3k
Tank15: 1.5k
Tank16: 6.0

Is this expected? I would have thought they would quickly equalize...

[Serenity]

Might have something to do with the numbers. Just placing one or two additional tanks to an existing setup makes them equalize very quickly

[rollie42]

The numbers? Not sure I follow; is there a good solution?

[ratchetfreak]

The numbers? Not sure I follow; is there a good solution?

Instead of a single row make it a double row

[sparr]

I don't know how it works behind the scenes, but I would not be surprised if each tank counts as 5 tiles of pipe. How fast does fluid equalize across an 80-tile pipe without a pump?

PS: pumps

[Ranakastrasz]

I don't know exactly how it works, but tanks work exactly the same as pipes. However, the capacity of a pipe has some effect on liquid flow, and the more capacity, the slower it flows.

if you use a pump, you can very easily drain or fill a tank.

[Shokubai]

Couple of issues.

1) Resistance - things slow down going linearly like this. Even with a flush input pipe its going to slow down by the last tank
2) Bandwidth - You have 1 input feeding all those tanks. Its never going to fill faster than that 1 input. Make 3-4 rows and bring in 3-4 lines from your source.

You could simply put 1 pump between ever tank or even 2nd/3rd tank but you still have the bandwidth bottleneck.

As far as why it's not balanced...you have the same problem with your output as with your input.

If you dont want to change your layout...
Create a horizontal pipe along the top for input fed by pump(s). This will act as a manifold.
Do the same at the south side for output but do it AFTER a pump on each (or every other) tank)

[Xeorm]

viewtopic.php?f=18&t=19851 Describes how liquid pressures work with math.

The relevant bit for your question is that entities share liquid based on pressure, but tanks do not get increased pressure, just more capacity. So a full tank transfers as much liquid in an absolute sense as a pipe does, assuming equal percentages of liquid, but takes much longer to fill/empty because of it.

[rollie42]

@Xeorm, thanks for the link - I get the idea. I guess I was confused because thinking about it, the physics of the storage tank should increase pressure as the contents increase, but I guess it isn't implemented that way (or my mental model of this is off)

[DerivePi]

As it stands, a fluidbox (pipe) with a base area of 1.0 will incorrectly have far less friction than a storage tank with a much larger base area (something like a factor of 36). A pipe with a baseArea around 0.98 conceivably has 0 friction.

The fluid box base area is not implemented properly and can easily be fixed (from my couch ). An easy fix is to divide the kinetic energy (flowEnergy) by the minimum BaseArea of the connected fluidboxes after transferring the fluids. Currently we have:
Q = ( amount1 / baseArea1 - amount2/baseArea2) * 0.4 + 0.59 * flowEnergy.
with the proposed change:
Q = ( amount1 / baseArea1 - amount2/baseArea2) * 0.4 + 0.59 * flowEnergy / baseAreaMin. (Note that this can't be simplified since the pressure difference should account for the individual fluidbox baseAreas)

The minimum BaseArea can be determined and stored when the Connection struct is constructed.