Water Towers and Moving Water Uphill in Satisfactory

Why your water stops partway up a cliff, where the first pump actually goes, and whether a water tower is worth building at all.

PATCH 1.2.3.1 UPDATED 2026-08-10 INTERMEDIATE

Water is the fluid you will move the furthest and lift the highest, because the places you want it are rarely next to a lake. Most water problems are one misunderstanding about head lift.

The numbers you need

ThingValue
Water Extractor output120 m³/min at 100% clock
Water Extractor power20 MW
Water Extractor built-in head lift10 m
Pipeline Pump Mk.1+20 m lift, 4 MW
Pipeline Pump Mk.2+50 m lift, 8 MW
Pipeline Mk.1 / Mk.2 throughput300 / 600 m³/min

That third row is the one people miss. The extractor itself provides 10 m of lift, so a modest rise off the shoreline needs no pump at all. Your first pump belongs where the extractor’s own 10 m runs out, not immediately at the water’s edge.

Also worth remembering: a perfectly level pipe still needs about 1.3 m of lift to fill. Level is not free either.

Why water stops partway up

Head lift is not a suggestion. When a pipe climbs beyond the available lift, flow simply stops at that height. The pipe below is full, the pipe above is empty, and nothing indicates which segment is the problem.

Three rules that resolve almost every case:

Head lift does not stack. Two pumps side by side give you one pump’s worth of lift, not two. They must be spaced along the climb so each takes over where the previous one’s lift is exhausted.

Measure the vertical rise, not the pipe length. A pipe running two kilometres across flat ground needs no pumps. A pipe running fifty metres straight up needs three.

Pumps must be on a section that already has fluid. A pump in an empty pipe has nothing to push. Build upward from the source rather than downward from the destination.

Placing pumps on a long climb

Working from the extractor upward:

  1. First 10 m: covered by the extractor.
  2. Place a Pump Mk.1 at that point: covers to 30 m.
  3. Next pump at 30 m: covers to 50 m.
  4. And so on.

With Mk.2 pumps at 50 m each, the spacing is much wider and the pump count drops sharply. On any serious climb, Mk.2 pumps pay for themselves in reduced clutter alone, even ignoring the doubled power efficiency per metre lifted.

Do you need a water tower?

Usually not, and this is worth saying plainly because “water tower” is a phrase people arrive with from other games.

There is no water pressure system that a tower would feed. A tall tank does not give downstream buildings better flow. What a tower actually gives you is stored volume at height, which is only useful if your consumers are also at height and you want a buffer between them and the pumps.

The building people mean is a Fluid Buffer (400 m³) or an Industrial Fluid Buffer (2400 m³). Those are genuinely useful, and the reasons are:

  • Absorbing sloshing. Pipe flow fluctuates by design; a buffer smooths it.
  • Covering burst demand. A consumer that draws unevenly gets a steady supply.
  • Priming a system. A full buffer means the line downstream fills immediately after a restart rather than slowly refilling from the extractor.

Stacking similar buffers is supported, which is how people build tank farms compactly.

So build buffers where they help. Build a tower only if you like the look of one.

Sizing extractors to consumers

One Water Extractor at 120 m³/min feeds a specific number of machines depending on recipe. Rather than guessing, count backwards from what the consumers want and underclock the extractor to match.

Underclocking is cheaper than it looks: power scales by an exponent, so an extractor at 50% draws considerably less than half the power. If your consumers want 90 m³/min, running an extractor at 75% costs less power than running one at 100% and dumping the surplus.

The wiki’s own extractor table shows this directly: 30 m³/min costs 3.2 MW while 120 m³/min costs 20 MW. That is four times the water for over six times the power.

Alternatives to lifting water at all

Before building a pump ladder up a cliff, check whether you need to.

Move the factory to the water. Coal generators and refineries both need water in volume. Putting them near the shore and belting the solid inputs in is frequently simpler than piping water up.

Use a closer source. Water is abundant. A smaller lake nearer your factory usually beats a large one far away, and the extractor needs the water to be deep enough rather than large.

Send it by rail. Fluid freight platforms move water without caring about elevation at all. For a genuinely large climb this is worth pricing against a pump ladder.

Water for coal power specifically

Coal generators are where most players first need water at volume, and the ratio is worth having to hand rather than discovering by trial.

Each Coal-Powered Generator consumes water continuously alongside its coal. A row of generators therefore has a fixed water demand that scales linearly with how many you run, and one Water Extractor at 120 m³/min supports a specific number of them.

The pattern that works: put the generator row near the shore, feed water in over a short pipe with no lift, and belt the coal in from wherever the node is. Coal travels well on belts; water does not travel well uphill. Moving the harder thing the shorter distance is nearly always right.

If your coal is far from any water, that is still usually the right arrangement — a long coal belt is cheaper and more reliable than a long water pipe with a pump ladder.

Diagnosing a dead water line

When water is not arriving and you cannot see why, check in this order.

Is the extractor powered and on deep enough water? Shallow edges do not qualify; the extractor needs proper depth, which is why some apparently large lakes will not take one at the shoreline.

Is any part of the run climbing more than the available lift? Walk the pipe and watch where it changes from full to empty. That transition point is your answer.

Is a pump sitting in an empty section? It cannot push what has not reached it.

Is the far end simply consuming everything? A line that is full at the start and empty at the end with no climb in between is a demand problem, not a lift problem.

Common mistakes

Putting the first pump at the extractor. The extractor already gives 10 m. Place the first pump where that runs out.

Stacking pumps together. Lift does not add. Space them along the climb.

Building down from the destination. Pumps need fluid already in the pipe. Build upward from the source.

Building a tower expecting pressure. There is no pressure system. Use a buffer for storage, and place it wherever is convenient.

Running the extractor at 100% out of habit. Underclocking to match demand saves disproportionate power.

See also

Sources

  • Satisfactory community wiki, Water Extractor and Pipeline Pump pages — extraction rate, built-in head lift, pump lift values and power draw
  • Game data for 1.2.3.1, build 23855724, stable branch — building power figures and unlock tiers