Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ruttiger]

I'm guessing the old ways are dead by now, I really haven't done it since .38, but you used to be able to build pumps feeding other pumps, and the calculations would produce extra water, if slowly.  I have some water feature plans on a map without aquifer and not producing enough rain to even consider, in fact half the ponds in my map have dried up, and the rest don't seem to be emptying or refilling. 

So is there any way to create water?  I mean, worst case I create a water source in one of the ponds using dfhack, but I'd prefer not to do that if I don't have to, and another ridiculous boondoggle project on the pile is always a plus. 

[Mohreb el Yasim]

If your map is freezing, and if I recall correctly, you can always make a 3 deep pond freeze => unfreeze to 7 deep.
If you want to do it more than once a year. I think Magma beneath the water layer (not touching) unfreezes it.

[PatrikLundell]

The freezing method mentioned is the only one I know of.

I'd try three things:
1. Set up drains from the ponds for the very infrequent rains that may happen (with suitable anti intrusion measures).
2. Check the caverns for any cavern lakes.
3. Unless you already know, check if the embark has more than one biome, as different biomes may have different aquifer conditions.

Unfortunately, there is no way to import water, nor any (sacrilegious) way to turn wine into water.

[Quietust]

I'm guessing the old ways are dead by now, I really haven't done it since .38, but you used to be able to build pumps feeding other pumps, and the calculations would produce extra water, if slowly.

By ".38", I assume you mean "0.27.173.38a" and "0.27.176.38a/b/c" - evidently, there were some bugs back in version 0.27.169.33g and earlier where water would duplicate itself when going through U-bends, but that was fixed over 13 years ago.

[Moeteru]

I believe there's also a way to duplicate water using minecarts.
I remember reading somewhere that if you submerge a minecart in 6/7 water then it will absorb 1/7 and become full. If that minecart then passes over a dumping track stop it will create 2/7 water for a net gain of 1/7.

I haven't tested it but I think the following would work:

• Dig out a large-ish 1z deep reservoir
• Build a track stop somewhere inside the reservoir, set to dump straight back into the reservoir.
• Place a minecart on that track stop.
• Fill the reservoir up to mostly 5/7, with a few tiles of 6/7 water.
• The minecart should duplicate water until the level eventually reaches 6/7 in most tiles.

The only problem I can foresee is the minecart getting moved by fluid flow. Maybe put it in an alcove behind a fortification to reduce that risk.

[PatrikLundell]

Assuming dump generation works (I don't know), build track stops everywhere so it doesn't matter if the cart gets moved to another tile.

[Ruttiger]

Hmm, that could work.  Not generating quickly, and I suck ass at setting up minecarts and tracks (Just never really bothered producing a power system that would work for them), but I need water and I have enough to make that work. 

And I could automate that?  Hmm....

[PatrikLundell]

A track stop is just a simple construction (remember to designate a dump direction away from the wall, as dumping into the wall is the magic behind deployable drains), and your "route" just consists of the route itself and a single stop at the top of the track stop, with all conditions removed. Of course, you'd have to do that for every tile of the room AND it hinges on water duplication actually happening. You'd be able to automate it (if it works) using a pressure plate set to trigger on 7/7 water only, releasing water through e.g. a trap door. I'd use bars in front of the door to block carts from getting pushed out with the water.

[Moeteru]

You might be able to automate it by using a floor hatch linked to a pressure plate, both placed at the bottom of your water-farm chamber.
The pressure plate is configured to open the hatch when the water level gets to 6/7 or above, and the hatch should lead down into a large storage tank.
A hatch should be used instead of any other switchable barrier because it changes state instantly and doesn't destroy any water when opening or closing.

Unfortunately the pressure plate imposes an unavoidable 99 tick delay between opening and closing, so you're always going to move quite a lot of water on every activation. You can reduce this by locating the hatch and pressure plate at the end of a long tunnel, but I'd be surprised if you could get below 20 or 30 units. Pressure-reducing diagonals don't reduce the flow rate of unpressurised water.
If the average water level drops to 5/7 or below then the farm will stop working, so you need to make sure that the total volume of water in your farm is large compared to the amount drained each time the hatch opens. Even with an arbitrarily large tank you'll still lose water on average if you have only one minecart because the hatch will activate just as often as the minecart (both are triggered by random movements of 6/7 water) and more water is drained with each activation than is created. Therefore, to get a net gain of water in your farm you'll need to have at least as many trackstops and minecarts as the number of units (1/7) of water drained each time the hatch opens. It's a random process so you should aim for several times more than this minimum bound in order to make the system reliable. Also note that you probably can't have one minecart dumping into another minecart, otherwise the second minecart will simply move the 2/7 water which was just created to its output side for zero net gain.

In fact, now that I think about it you might be able to make the water movement slightly more deterministic by carefully choosing where to place your track stops and what direction to have them dumping in. I'm imagining some kind of loop of alternating track stops and floor tiles which would tend to move water around the loop while spilling excess out to the sides.

Anyway, this is all strictly in the realm of speculation until one of us verifies that minecart-based water creation works in the first place.

You'd be able to automate it (if it works) using a pressure plate set to trigger on 7/7 water only

If my theory of how minecart-water interactions work is correct (see table below) then the water level will never reach 7/7 except momentarily on the output tile of each minecart. The water creation rate drops to zero when the average level is slightly below 6/7 and actually goes negative (ie. water is destroyed faster than it's created) at levels above that.

Water level at input  |	Water level at output  |	Net units of water created  |	Notes
5/7 or less	any	0	A minimum level of 6/7 is required.
6/7	5/7 or less	1	Input is reduced to 5/7 and two units are created at output.
6/7	6/7	0	There's only space to add one unit of water at the output.
6/7	7/7	-1	Input is absorbed but the output side is already full.
7/7	5/7	0	This is what happens when one minecart feeds into another.
7/7	6/7	-1
7/7	7/7	-2	Again, the output is full so the input is just destroyed.

The negative values are based on the fact that minecarts seem to not care whether the tile they're dumping to can actually accept liquid. They try to put 2/7 liquid there, but if it fails then they still become empty, hence why portable drains work. Of course this still all needs testing.

[PatrikLundell]

More speculation: If the water generation chamber is 2 Z levels high then the water output into a full output tile ought to stack on top of the 7/7, in which cases you'd get no losses as long as you don't dump into walls. Note speculation, though, as I haven't tested it.

[mightymushroom]

More speculation: If the water generation chamber is 2 Z levels high then the water output into a full output tile ought to stack on top of the 7/7, in which cases you'd get no losses as long as you don't dump into walls. Note speculation, though, as I haven't tested it.

If that were true I would expect copious reports of overflowing from minecart-driven water reactors pouring from 6/7 into 6/7.


I tracked down a post where Fleeting Frames discusses this trick in the context of magma generation; should work for water, too. There's more scattered about in that thread including some illustrations. Although I must admit that I have never truly understood the exact setup used.

The new plan calls for an ex nihilo magma generator. 3z, exploiting that a minecart on top of minecart will take only 1/7 magma to fill up, and generating up to 60 magma per day. My theoretical max with an unit around this size is 200, but requires extensive signalling setup that I don't care to waste the mechanisms for.

(The biggest slowdown is that I don't have the exact colours I want - it should be a gradient of purple to red there. I try to include a way to stop, so I could replace them later...It must make for an acceptable approximation for now.)

[Bumber]

A track stop is just a simple construction (remember to designate a dump direction away from the wall, as dumping into the wall is the magic behind deployable drains), and your "route" just consists of the route itself and a single stop at the top of the track stop, with all conditions removed. Of course, you'd have to do that for every tile of the room AND it hinges on water duplication actually happening. You'd be able to automate it (if it works) using a pressure plate set to trigger on 7/7 water only, releasing water through e.g. a trap door. I'd use bars in front of the door to block carts from getting pushed out with the water.

You don't need a route if the minecart is just going to sit on the track stop. Just forbid it to prevent it from being hauled.

[PatrikLundell]

@Bumber:
Hm, wasn't aware of that. However, you'd still need to get the mine cart hauled there, which is probably easiest done by assigning it to a route stop.

Edit: Corrected "was" to the intended "wasn't" above...

[Fleeting Frames]

@Moetaru: You can indeed generate water (or magma) with a minecart, but you got it slightly wrong.

Specifically, a floating / flying cart will only take on fluids from 7/7 tile and will use only 1 unit of fluid to fill fully.

Spoiler: I have one running in Solstice Chronicle (click to show/hide)