Controlling liquid consumption priority

[BeltRunner]

Let's say you are building an Advanced Oil Cracking facility to produce Petroleum Gas, but you also need Lubricant for some intermediate products.
If you don't have a lot of Crude Oil, you may face a problem when all your Heavy Oil is used by Heavy-to-Light cracking, and you get no Lubricant.

Here is a solution:

Example

How it works:

Diagram

Crucial part here is Storage-Tank-to-Pipe junction with combination of small pumps:
If a low amount of liquid (A) enters the system, liquid flow in Storage-Tank-to-Pipe part will be low (since A / Storage_Tank_Capacity is much lower then A / Pipe_capacity),
while the flow in Storage-Tang-to-small-pump is constant (~10/s).
This way all* liquid will flow into small pump (high priority output) direction unless its full.


* - only ~99.8% actually. In order to eliminate leakage and reach 100%, the second Storage-Tank-to-Pipe junction and back loop were added.

[Ackbarspiff]

Nice design. I will use this for my next factory.

[Salmelu]

I've been thinking about the same today, because of the same problem - making lubricant even though I need to crack the rest of the heavy oil.

I like your solution too, I'll try how it works

[sillyfly]

Could you explain why (if at all?) is the loop at the bottom necessary? (the one with tanks -> pump -> underground pipe -> pump -> tanks)

If you have already explained it I am sorry, but I have not understood it.
Thanks.

[BeltRunner]

Could you explain why (if at all?) is the loop at the bottom necessary?

First of all, if you are ok with small (~0.2%) leakage of liquid into low priority output, you can use basic design:


If you wand "perfect" 100% solution, then loop at the bottom is necessary to redirect leaked 0.2% of liquid back to the system input.

[sillyfly]

I see. To put it in my words - the loop makes sure there is always a "source" that wants to draw from the tanks, and thus makes anything without a pump lower-priority.

As for the solution with only one pump - it relies on the high-priority consumer being able to consume faster than the liquid is produced, which is not always the case.

Anyway - this is very helpful! thank you for this design.

[Nova]

Is it okay to remove the pipe between the two storage tanks?

[Rahjital]

That would break it because the low priority output would get no liquid.

[Nova]

Ah, no, I would like to put the two tanks closer together - removing the pipe and then connecting the two tanks directly.

[Rahjital]

Oh, that should work just fine, unless there is some quirk with liquid flowing from tank to tank. I think BeltRunner put the space there to fit that electric pole in.

[Saltan]

Hey guys,

I've made a few tests and it sounds like the system show by BeltRunner is not working anymore : I builded exaclty the same configuration show in the picture, and added two tank on each output pipes. I have then connect a water pump to the input pipe, and check the water flow : The tank on the "low pririty" output was fill more quickly than the other one. It seems like the small pump is limiting the maximum output of the flow.

Did you do the test ? Which effective solution can be found to control the liquid consumption priority ?

Thanks

[Zeblote]

Couldn't you just put the lower priority output behind a pump and set a combinator to only turn on the pump if the other tank is full?

[Saltan]

That may work and be very effective but that's a bit "space consuming".

I have found an other simple way to do it : the pump on the high priority output is limiting the flow. So if you add an other pump on the input pipe, you will limit the amount of liquid in the system and let your output pump manage all the liquid without being "overflooded"

A picture will explain is better :


In order to have a better result the high priority output pipe should be short, or have multiple pumps.

Thx for the help !

[Saltan]

Well that solution is not fully working in case of complex liquid managment system.

I am working on your solution Zeblote to found the best way to do it ("best way" here = Which one whould fullfill my needs).

[PiggyWhiskey]

Have you thought about changing the way you do it slightly.

Let's say you have a tank for each of the 5 liquids (Crude, Heavy, Light, Petroleum, Sulfuric Acid, Lubricant.)
And you want to crack down to petroleum, but don't want to leave yourself short on Lubricant (Or Light Oil->Solid Fuel)

It's simply a matter of laying out some small pumps connected to each tank with certain filters.



This quick setup is completely self sufficient and expandable.
Prioritising:
Petroleum (Sulfur/Plastic)
Heavy/Lubricant
Light Oil

With the LIght Oil it makes Solid Fuel only if the tank is full and potentially causing a backlog.
Although you could remove this entirely, your Lubricant may suffer if you have too much light oil and it can't crack down to Petroleum (Backlog in Oil Refinery)

I've marked on the image what settings to have for the pumps.
If you want to prioritise Sulfur over Plastic it's easy enough to put another Small Pump on the Petroleum Tank and add another condition and give the one you want more of a higher value. (Plastic P > 1.1k Sulfur P > 1.0k for a priority to Plastic)
That also keeps 1k in the Petroleum Tank. It is kinda useless, but I found it's easier to move fluids than if it was set to 0 or a really low value.

[Saltan]

Thanks for your reply PiggyWhiskey !

I have end up with a solution very similar. The only difference is that i don't connect all to a central system, but each pump to his associated tank. This is possible because i builded a very compact refinery, your solution is way better for a bigger or a more complex refinery.

My goal was to destroy the oil if i had too much. Here is what i build :


The full factory :

Factory

fixed image --daniel34

[PiggyWhiskey]

That last picture wouldn't load for me.

Rather than destroy anything, why not crack it down to what you need more of?

[V453000]

I personally just use 1 tank for each liquid type, and connect wires to small pumps.

With conditions:

If heavy oil > 2000, start cracking it into light oil. (by starting the pump which leads to cracking heavy->light)
If light oil >2000, start cracking it into petroleum gas. (by starting another pump which leads to cracking light->petroleum gas)

By connecting lubricant production directly to heavy oil tank Before the pump, the lubricant plants always take heavy oil if there is any. And if there is too much heavy oil, then cracking happens.

That way I do not really need anything special to be built, but it works flawlessly. The only thing that can occur is that petroleum gas tank gets full, and the refineries stop because of that.

In games where I have enough oil, I just connect petroleum gas -> solid fuel plants, with a pump condition of petroleum gas >2400 or such. In games where I want to save oil, I just ignore it as sooner or later petroleum gas consumption will be HUGE anyway.

Rather than destroy anything, why not crack it down to what you need more of?

I think this is useful when you do not have access to advanced oil processing and thus no cracking. Keeps stuff running and you do not have to build 1234297 tanks to store stuff in . Especially great if you plan on staying green science for a longer time (I like that when I want to have more time to get laser turrets, trains, electric furnaces and construction robots, instead of rushing for blue lab packs)

[PiggyWhiskey]

If you're going to make solid fuel, you should do it at the light oil level. IIRC you get 2 solid fuel from light oil, and only 1 from that same amount cracked down to petroleum. And 1.5 if you use heavy oil (cracked down into the same amount of light oil....if you catch my drift.)
That's why my design had the light oil/solid fuel setup.

Also, you are correct if you don't have advanced oil processing.

[V453000]

Yeah, if you care about the amount of fuel, that definitely is more efficient, and if you give it enough headroom, petroleum gas might never fill up either (: I just generally do not care about the amount of solid fuel as it is just a "mean to enable refinery production again" ... though reasons to that are probably little