High speed fluids

Don't know how to use a machine? Looking for efficient setups? Stuck in a mission?
the_potty_1
Long Handed Inserter
Long Handed Inserter
Posts: 62
Joined: Sun Aug 16, 2020 12:12 pm
Contact:

High speed fluids

Post by the_potty_1 »

Hi

In my experience, most fluids flow OK without doing anything special. Ten oil refineries running flat out don't typically back up, with only a scattering of pumps acting mostly as non-return valves. However, I'm running a modest nuclear plant, I built 4 reactors in a square, but found that I only need two running for my power requirements, so don't feed the other two at all. Providing water for this rig turned out to be problematic, I built it before I'd pushed my wall out, so it's quite far from the water pumps. I could rip it up and rebuild nearer to the pond, but that's a factorio solution not an engineering solution. IRL people don't move nuclear reactors.

I assume I'm on the new 1.0 release, and I hear there was a fluid overhaul. I was running the 0.18 beta, and I just checked, I'm now opted out of beta versions, and I see a new Wube circuit on the load screen. Congrats on the release :D

Back to the water. I laid down an overkill of offshore pumps, feeding a pair of storage tanks, then two pipe to ground, a right angle, another pump, another tank, and basically followed that pattern (without the right angle) to the reactor. My flow maxed out just over 1400. After the update my reactor was basically running dry, so I quickly slapped down a train to transport from the tanks by the pumps. Between the two, the reactor's running great, but both are required. Without the land pipe, the reactor would run dry in the time the train takes to fetch a load. I could add a second tank to the train, and more storage next to the reactor, but I really feel like I should be able to run water faster through pipes.

I redid the pipework to speed it up, I removed the tanks along the way, and mixed every pipe-to-ground jump with a pump, the entire way. I now get 1714 of a theoretical 12000 on every pump. Of the offshore pumps, one is running 1000 of 1200, another is running 400 of 1200, and the other four are idling. The pump from the offshores to the first tank runs 1712 unless the train docks for a refill, then it jumps to around 4700. The same for the pump between the first tank and the second. The train loads from the second tank. I have to assume that pipe to ground is the limiting factor. I will experiment with shorter stretches, parallel lines, and just normal pipe. Might as well try just pumps as well, or alternating pumps and tanks. God help me, I'm getting a flow over 10000, even if I have to turn on the other two reactors and lay down a gratuitous radar array to quadruple my power usage.
Serenity
Smart Inserter
Smart Inserter
Posts: 1017
Joined: Fri Apr 15, 2016 6:16 am
Contact:

Re: High speed fluids

Post by Serenity »

The proper fluid overhaul was scrapped because there were unexpected issues with it. There was a lot of optimization, but it still behaves as it always has
Yoyobuae
Filter Inserter
Filter Inserter
Posts: 511
Joined: Fri Nov 04, 2016 11:04 pm
Contact:

Re: High speed fluids

Post by Yoyobuae »

the_potty_1 wrote: Sun Aug 16, 2020 1:06 pm I redid the pipework to speed it up, I removed the tanks along the way, and mixed every pipe-to-ground jump with a pump, the entire way. I now get 1714 of a theoretical 12000 on every pump.
First, you cannot get 12000 fluid/s thru an underground pipe. An underground pipe counts as two pipe tiles and pipelines with two pipe tiles between pumps are limited to 3000 fluid/s max: https://wiki.factorio.com/Fluid_system#Pipelines

Second, every pipeline you make will always be limited to the weakest link. The fact that you are only getting 1714 fluid/s max instead of 3000 fluid/s means that somewhere along the way you've placed more than two pipe tiles in between pumps. It is EXTREMELY easy to make a mistake introduce a bottleneck in your flow. The problem can be anywhere between the producers and consumers. The producers and consumers themselves could be the problem too.

I recommend that you first test your flow between storage tanks. Setup one or more full storage tanks at the input to the pipeline and place a few or more empty storage tanks at the output. Remember to pull from source storage tanks using a pump and feed the empty storage tanks with a pump as well. This should give you the maximum throughput for the pipeline in question (temporarily).
the_potty_1 wrote: Sun Aug 16, 2020 1:06 pm God help me, I'm getting a flow over 10000, even if I have to turn on the other two reactors and lay down a gratuitous radar array to quadruple my power usage.
For testing the maximum power/fluid throughput you have to ensure that every step of the way is not limited. Offshore -> water pipeline -> heat exchanger -> steam pipeline -> steam turbine -> electric grid -> consumer. All of it needs to be checked. And also don't forget to check reactor -> heatpipe -> heat exchanger side too.

And for flows over 6000 you cannot use pipes (regular or underground). Flows of 6000~12000 can only flow thru:
  1. Pumps
  2. Storage tanks
  3. Fluid wagons
  4. Boilers (water side)
  5. Heat exchangers (water side)
  6. Steam engines (steam only)
  7. Steam turbines (steam only)
  8. Mining drills placed on uranium (sulfuric acid only)
Basically you can only push 6000~12000 flows thru entities which have at least 200 fluid storage capacity. Pipes only have 100 fluid capacity (underground pipes count as two individual pipe tiles with 100 fluid capacity each).
Last edited by Yoyobuae on Sun Aug 16, 2020 3:20 pm, edited 1 time in total.
astroshak
Filter Inserter
Filter Inserter
Posts: 642
Joined: Thu May 10, 2018 9:59 am
Contact:

Re: High speed fluids

Post by astroshak »

My own solution (since nuclear requires a LOT of water) is to landfill lakes. This allows me to put the offshore pumps exactly where I want them, minimizing pipe throttling of throughput.

A 2x2 reactor setup needs, per the factorio cheat sheet (factoriocheatsheet.com), 48 heat exchangers. Each one can use 103 water/sec, for a grand total of 4944 water/sec. This means, you need 5 offshore pumps. This ALSO means that you really can only expect to have 11 heat exchangers in a row, being fed by a single offshore pump, and that additional heat exchangers need another row and another offshore pump.

(As an aside, this is why I put my reactors on landfilled lakes. I use rows of 13 I think heat exchangers in my BP, but at both ends I’ve got an offshore pump feeding the row directly. This is overuse on the offshore pumps, but it eliminates the throttling of the pipes and allows the heat exchangers to operate eat 100%.)

Each reactor can produce 40 MW of heat. They get a neighbor bonus of 100% per operating reactor fully adjacent to them, a maximum of 400% bonus (or 200 MW in total) if they are in the middle of four other operating reactors. Sadly, there is no vanilla means of automating the fuel supply for that reactor! So I built a BP that extends infinitely in one direction, a 2xN reactor setup. That 2x2 setup you put down? That can produce 480 MW of power when built up properly. A 2xN setup produces 480 MW with the first 2x2, but then each additional 2x2 you extend it adds another 640 MW. Use that to decide how many reactors you need to build; but expect it to be a lot if you want to power a beaconed/moduled megabase.
User avatar
disentius
Filter Inserter
Filter Inserter
Posts: 694
Joined: Fri May 12, 2017 3:17 pm
Contact:

Re: High speed fluids

Post by disentius »

A 2x1 reactor, see OP (160KW) needs 16 heat exchangers (10Kw each) -> 1648 water per second.
User avatar
disentius
Filter Inserter
Filter Inserter
Posts: 694
Joined: Fri May 12, 2017 3:17 pm
Contact:

Re: High speed fluids

Post by disentius »

If you build it like this, you dont have to build over a lake, just use one side. here is a a 1.12 GW (x8) reactor with that concept.
1.12GW Nuclear Water sideloading 1.png
1.12GW Nuclear Water sideloading 1.png (3.93 MiB) Viewed 6040 times
the_potty_1
Long Handed Inserter
Long Handed Inserter
Posts: 62
Joined: Sun Aug 16, 2020 12:12 pm
Contact:

Re: High speed fluids

Post by the_potty_1 »

I've seen the plans which require the landfilling of a lake, but I'm afraid I can't do that. Although now that water shows pollution, I know which lake I'm choosing not to fill in, ew, that's nasty.

I tried alternating pumps and tanks, but it looks horrendous, and appears to run at 5k instead of 12k. Although I didn't bother to disconnect my solar array and add the power load to the nuclear plant, because the solution is obvious. Just pipe using pumps. If you need to corner use a storage tank. Oh yes, I replaced that corner pipe with a tank, and my speeds on the original piping went up to 3k immediately, and for my current usage, that's all I needed.
Yoyobuae
Filter Inserter
Filter Inserter
Posts: 511
Joined: Fri Nov 04, 2016 11:04 pm
Contact:

Re: High speed fluids

Post by Yoyobuae »

the_potty_1 wrote: Sun Aug 16, 2020 6:15 pm I tried alternating pumps and tanks, but it looks horrendous, and appears to run at 5k instead of 12k. Although I didn't bother to disconnect my solar array and add the power load to the nuclear plant, because the solution is obvious. Just pipe using pumps.
Screenshot_2020-08-12_11-09-50.png
Screenshot_2020-08-12_11-09-50.png (1.25 MiB) Viewed 6016 times
Alternating pumps and tanks will give you 12k fluid/s, assuming you don't bottleneck the flow elsewhere. I think the only application for something like this would be loading trains, since you kinda need to use storage tanks anyway.
the_potty_1 wrote: Sun Aug 16, 2020 6:15 pm If you need to corner use a storage tank. Oh yes, I replaced that corner pipe with a tank, and my speeds on the original piping went up to 3k immediately, and for my current usage, that's all I needed.
You don't need to use storage tanks for corners. But the fluid system sometimes behaves weirdly and just picking up and replacing a pipe can change the flow rate. Normally, you can have a corner with a pump before and after the corner, with up to two pipe tiles in between them.
the_potty_1
Long Handed Inserter
Long Handed Inserter
Posts: 62
Joined: Sun Aug 16, 2020 12:12 pm
Contact:

Re: High speed fluids

Post by the_potty_1 »

Well I didn't deconstruct the four reactors, but I was compelled by my OCD to rebuild the rest of the power plant :D I originally had 3 lines of heat pipe, surrounded by 6 lines of heat exchangers, and 6 lines of steam pipe running to feed the turbines on the other side of a row of steam storage tanks. I tore it down, placed a double line of heat pipe, connected to 2 rows of 24 heat exchangers, those directly feeding 2 turbines each, a total of 96. I then disconnected the solar arrays, and slapped down plans for 1600 radar. The water's flowing beautifully, and I only seem to need 4200/s (It can do 10k/s if required). But the plant was only producing 413 MW, it's supposed to produce 480. I looked, and 12 of the turbines furthest from the reactors were not running, even though the heat pipe was at 503 degrees at that end. I checked the ideal component ratios for 4 reactors, and it is 48 heat exchangers, but only 83 turbines. I cut it down to 42 exchangers with 84 turbines, and the power jumped up to the expected 480 MW. Note I had to replace the exchangers with 4 pumps and a single pipe, but each line at that point is carrying 2100/s water, and the pipe caps at 3000/s, so not a limiting factor. Fortunately I'm not so OCD that I can't live with this single pipe, or I would .. have .. to .. rip .. up .. the .. entire .. pump .. line .. and .. redo .. it .. Damn.
Yoyobuae
Filter Inserter
Filter Inserter
Posts: 511
Joined: Fri Nov 04, 2016 11:04 pm
Contact:

Re: High speed fluids

Post by Yoyobuae »

the_potty_1 wrote: Sun Aug 16, 2020 8:49 pm Well I didn't deconstruct the four reactors, but I was compelled by my OCD to rebuild the rest of the power plant :D I originally had 3 lines of heat pipe, surrounded by 6 lines of heat exchangers, and 6 lines of steam pipe running to feed the turbines on the other side of a row of steam storage tanks. I tore it down, placed a double line of heat pipe, connected to 2 rows of 24 heat exchangers, those directly feeding 2 turbines each, a total of 96. I then disconnected the solar arrays, and slapped down plans for 1600 radar. The water's flowing beautifully, and I only seem to need 4200/s (It can do 10k/s if required). But the plant was only producing 413 MW, it's supposed to produce 480. I looked, and 12 of the turbines furthest from the reactors were not running, even though the heat pipe was at 503 degrees at that end. I checked the ideal component ratios for 4 reactors, and it is 48 heat exchangers, but only 83 turbines. I cut it down to 42 exchangers with 84 turbines, and the power jumped up to the expected 480 MW. Note I had to replace the exchangers with 4 pumps and a single pipe, but each line at that point is carrying 2100/s water, and the pipe caps at 3000/s, so not a limiting factor. Fortunately I'm not so OCD that I can't live with this single pipe, or I would .. have .. to .. rip .. up .. the .. entire .. pump .. line .. and .. redo .. it .. Damn.
You cannot run two lines of 24 heat exchangers with just a double line of heat pipes (you need a triple one).

My reasearch on heatpipe and max length of heat exchanger rows:
  • Single heatpipe line -> two rows of 14 heat exchangers
  • Double heatpipe line -> two rows of 20 heat exchangers
  • Triple heatpipe line -> two rows of 24 heat exchangers
  • Quad heatpipe line -> two rows of 28 heat exchangers
  • Quint heatpipe line -> two rows of 31 heat exchangers
  • Sextuple heatpipe line -> two rows of 33 heat exchangers
the_potty_1
Long Handed Inserter
Long Handed Inserter
Posts: 62
Joined: Sun Aug 16, 2020 12:12 pm
Contact:

Re: High speed fluids

Post by the_potty_1 »

Huh. Well I just loaded my save and I'm again getting 413 MW, so I assume you're right. I will play around and see what reliably gives me 480.

EDIT in the screenshot above, they use a double line even though there are only 14 heat exchangers, is that for aesthetics, or does the gap between the reactors and the first heat exchanger count as if there were theoretical heat exchangers filling that space?
User avatar
disentius
Filter Inserter
Filter Inserter
Posts: 694
Joined: Fri May 12, 2017 3:17 pm
Contact:

Re: High speed fluids

Post by disentius »

Just empirical testing. One row was not enough to deliver heat to outer HE's.
Koub
Global Moderator
Global Moderator
Posts: 7955
Joined: Fri May 30, 2014 8:54 am
Contact:

Re: High speed fluids

Post by Koub »

The most confusing thing is that pipe building order matters for maximum throughput, but also flow direction.
See here : viewtopic.php?p=506022#p506022
Koub - Please consider English is not my native language.
Yoyobuae
Filter Inserter
Filter Inserter
Posts: 511
Joined: Fri Nov 04, 2016 11:04 pm
Contact:

Re: High speed fluids

Post by Yoyobuae »

the_potty_1 wrote: Tue Aug 18, 2020 8:15 am EDIT in the screenshot above, they use a double line even though there are only 14 heat exchangers, is that for aesthetics, or does the gap between the reactors and the first heat exchanger count as if there were theoretical heat exchangers filling that space?
Empty spaces on the sides of the heatpipe line don't quite count as filled by heat exchangers but heatpipes do loose a few degrees temperature per tile. So you naturally want to minimize the distance between reactors and heat exchangers.

When I tested heatpipes I only left the required two tile space (for fuel cell feeding) between the reactor and the start of the heatexchanger lines.
Post Reply

Return to “Gameplay Help”