How to avoid aquatuner blocked pipes?

[Cipupec2]

I keep getting blocked pipes on my aquatuner, any idea of what cause it?  here is a simplified layout of what I use:

 

Spoiler

Spoiler

   

[Soulwind]

Well unless you're running some kind of switching mod, you've got your inputs and outputs reversed between the bridge, the shutoff, and the tuner.

[0xFADE]

Easy, there were several threads with people dialing in on the best solution.  I can't claim I did it right but...

The sensor controls the tuner.  Bridge priority means liquid won't transfer through the bridge if there is already liquid (from the aquatuner) going through.

I also have some bridge priority on that refinery.  It keeps the loop going when the refinery is off and stops the loop so the refinery can unload/refill.

[ghkbrew]

1 hour ago, Cipupec2 said:

any idea of what cause it?

The loop through the aqua tuner is a long run of pipe with multiple inputs and outputs.  ONI may be deciding the flow in certain segments should be in the "wrong" direction.  I would try adding bridges between the various connections to force flow in the right direction.

Is there a reason you're using this setup (loop packets through AT until they're cold enough) instead of something like @0xFADE showed (packets bypass tuner if they're cold enough already)?  In my experience it's almost always better to keep the loop flowing rather than stopping it to cool one hot packet.

[Cipupec2]

5 hours ago, Soulwind said:

Well unless you're running some kind of switching mod, you've got your inputs and outputs reversed between the bridge, the shutoff, and the tuner.

I don't believe I am. To clarify, my setup works, I have 6 aquatuner that worked for over 100 cycles, the problem is that every ones in a while one of them would get blocked for no apparent reason (which cause my production facility to get hot) and I have to use extract pipe content on the output. Which annoys as I can't figure out why would that happen.

4 hours ago, ghkbrew said:

The loop through the aqua tuner is a long run of pipe with multiple inputs and outputs.  ONI may be deciding the flow in certain segments should be in the "wrong" direction.  I would try adding bridges between the various connections to force flow in the right direction.

I don't think this is the case. The problematic segment has bridges on all pipes leading in/out  

Spoiler

 

4 hours ago, ghkbrew said:

Is there a reason you're using this setup (loop packets through AT until they're cold enough) instead of something like @0xFADE showed (packets bypass tuner if they're cold enough already)?  In my experience it's almost always better to keep the loop flowing rather than stopping it to cool one hot packet.

I have a priority bridge leading to the aquatuner so packets keep flowing in the cooling loop regardless. So the only difference would be in that my "temp filter" nearer the aquatuners (which works nice for me because I have 6 of them there and it makes management easier) 

 

@0xFADE, its a little bit hard to see but you don't seem to be using the thermo sensor / gas valve solution I seek.

 

 

[0xFADE]

If you wanted to set up something that is looping continuously on a couple packets then spits them out when they are cold I guess you can use those otherwise you are just using extra electricity.  I would have to see an extremely exact example of why you would want to do it that way rather than loop the entire thing till it gets down to the temperature you want.  If you were cooling down liquid for food it would probably be more efficient to use it as a radiator and cool down a pool that you start pulling from when it gets down to temperature.  When large enough it should become pretty stable with the incoming/outgoing liquid based on how much cooling it has.

You can chain it as much as you want.  Each could bet set with the same temp so that when you rapidly cool down to your target they all stop or half stop since you hit your target midway etc.  The image where the aquatuner is plugged in is cooling down all the oxygen generation with a long run of polluted water.  It stops running when the polluted water gets cold enough but the loop continues to loop.  Sometimes if they get backed up for some reason you need to pull a few packets out.

[Cipupec2]

29 minutes ago, 0xFADE said:

  Sometimes if they get backed up for some reason you need to pull a few packets out.

So you have the same problem, no? Is there any longer term solution to this, so I can focus on other parts of my base?

Otherwise you probably right, I could improve the design for efficiency, but that would have to wait for space or next game. right now I was going for simple block design I could easily expand, I used similar layout to what I tested for my base air cooling  and didn't used radiators because in my last game it always had issues with spikes so i wanted to try to come up with something more precise.

[impyre]

Here's what I use:

Plumbing:

Spoiler

Automation:

 

Spoiler

Basically, the input pipe for the aquatuner is always full because it gets priority. The temperature sensor disables the aquatuner if the next packet is too cold. The bypass gets priority on the output side, so the cold packet will bypass the aquatuner's input pipe and continue out. When warm packets are detected again, the aquatuner is enabled, the output clears, and normal flow resumes. I'm using the filter gate to smooth out operation so it doesn't toggle too much... because since the bypass is shorter it takes a couple of extra ticks to cut back on.

[0xFADE]

55 minutes ago, Cipupec2 said:

So you have the same problem, no? Is there any longer term solution to this, so I can focus on other parts of my base?

It mostly happens if you are feeding a loop the wrong way.  If you bridge in to the loop it should be fine because it is only adding to the loop if there is a spot.  But if you try and directly cram more liquid in it will back up.  It can also happen if the bypass on the tuner would put that loop in a state where it becomes backed up.  It seems like each bridge on a loop leaves one bubble worth of backup.  Like when I take out a bridge somewhere on a loop the loop will move forward like traffic coming to a stop.

That might be why I tend toward making closed loops.  Since it makes dealing with any backup problem as easy as taking a little bit back out.  Using the loop to cool something indirectly.  Yeah, that is my go to.  I only have some loops that are fed by the output of steam gens but I can probably make those indirect as well.

Since liquid reservoirs have signals now you could use that to buffer any loop and keep it from overfilling.  That is only a "recent" thing so I have not used it enough.

54 minutes ago, impyre said:

 

  Hide contents

 

That looks pretty good, and something you could insert pretty much anywhere you needed.

[ghkbrew]

@Cipupec2, I now suspect you're overfilling the aqua tuner loop.  My mental model of how pipes flow is:

Every 5th tick (once a second) all the pipes are updated (the flowing animation is a lie).  All the packets in a segment of pipe will move one forward one segment if there is an empty segment at the end.  At the end of this tick buildings update their fluid input/output ports.  Specifically bridges will teleport a packet from their input port to their output port (if there is an empty space under their output port after fluid flow).

So, in a continously flowing loop the input port of the bridges will always be empty.  If you stop flow at it's output port a new packet will move to the segment under it's input port and then sit there.  This is what happens when you overfill a loop.  All the pipe segments are full so nothing can move.

This is basically only a problem when you are both 1) filling a loop and 2) intermitantly stopping the loop.  I think what's happening here is the Aqua tuner is acting as an intermittently blocked bridge and making your loop over fill and then stop.

[Cipupec2]

@ghkbrew, yes, you right. At first I didn't realized which loop you meant, but thanks for your explanation i now know what to do to avoid playing plumber for the rest of the game.

[0xFADE]

Better described than what I did for sure.

[TheMule]

Useful post:

 

[Cipupec2]

interesting, I am going to try this double bridge solution