Sandustry Guide and Wiki

getting started

25 Sandustry Tips I Wish I Knew Earlier

Reddit's rising posts show continuous demand for tips and best strategy advice. This guide distills current community wisdom — validated by official documentation where possible — into a concise checklist that helps new and intermediate players avoid the most expensive mistakes.

Early Access · checked 2026-08-29 · Checked 2026-08-29

1. Build small, then scale

Your first factory is a learning exercise, not a final design. Build one stable Sand→Gold loop. Observe where it clogs. Fix that bottleneck. Only then add a second loop. A small working factory teaches you more than a large broken one.

2. Rebuild without shame

Community advice consistently says: do not try to fix a messy factory. Isolate the working parts, delete the rest, and rebuild one module at a time. Rebuilding is normal progression, not failure. You will rebuild at least three times before late game.

3. Use gravity for water

Pumps break, pipes clog, and priming is tedious. Whenever possible, place your water source above your factory and let gravity do the work. A raised tank with a shutoff valve is more reliable than a pumped loop.

4. Separate every output

Gold, Slag, residue, and byproduct need independent routes. If they share a destination, one full collector stops everything. Label every route and test the full state before you walk away.

5. Overflow is safety

Every production line needs an overflow path to visible quarantine. Overflow is not waste; it is a signal that something downstream is full. Without overflow, a single full buffer silently stops your entire factory.

6. Test reload before trusting

A factory that works after building may break after save and reload. Test every critical line after reload before scaling. Falling-sand physics can settle differently on load.

7. Do not copy community blueprints blindly

Blueprints hide assumptions about version, startup state, and surrounding geometry. Inspect the footprint, test one clean material, and verify every interface before connecting to production.

8. Research the bottleneck you have

Do not follow a universal research order. If you cannot carry enough, unlock belts. If sorting fails, unlock Filters. If you are stuck at a puzzle, check Research for the next tool. Match research to your current constraint.

9. Keep maintenance access

Every machine needs a walkway, a shutoff, and a way to clear jams without demolishing walls. If you cannot reach a clog, the factory design is wrong, not the machine.

10. Measure, do not guess

Do not trust nominal machine rates. Time actual output at the collector. Use the slower sustained rate in your planner. A short burst of fast output hides a bottleneck that will appear later.

11. Planter Boxes leak

Flowers and Amethelis can escape Planter Boxes if growth exceeds capacity or if the box is damaged. Build farm enclosures with containment walls and a catchment below. Check the official Planter Box page for permeability behavior.

12. Steam does not pump

Steam is a gas. Pumps move liquid. If you try to pump Steam, it will not flow. Give Steam a vertical escape path and let buoyancy do the work.

13. Lava melts everything

Lava destroys ordinary containers, melts Pipes, and converts Water to Steam. Route Lava in fire-resistant containment with emergency shutoffs. Never let Lava intersect your water lines.

14. Catch zones need margins

Launcher handoffs need generous landing areas. Start wider than you think. Record overshoot and undershoot. Narrow the catch zone only after repeated clean batches.

15. One turbine per steam source

Steam is consumed by the first turbine it touches. A second row of turbines on the same line produces no extra power. Add separate steam chambers with separate turbines instead.

16. Gravity filters are coarse

Gravity-based sorting is fast for bulk separation but less precise than standard Filters. Use gravity for pre-sorting and Filters for final cleanup.

17. Do not automate until you understand

Automating a process you do not fully understand creates hidden failures. Run the loop manually until you can predict every failure mode. Then automate one step at a time.

18. Backup before experiments

Save a clean backup before testing blueprints, Lava handling, or new machinery. Cloud sync is not a backup. A failed experiment can overwrite your only good save.

19. Use the pause menu unstuck command

If you are physically trapped in a puzzle or dungeon, the official Beginner's Guide documents an unstuck command in the pause menu. Use it before digging through unknown boundaries.

20. Read the objective panel literally

When progression stalls, open the Research tab and read the exact requirement. Do not guess. The game shows you the next step. If it does not unlock, check for a full storage or unclaimed reward.

21. Dirt is not waste

Excavated dirt and stone can be used for containment, landing pads, and temporary walls. Before deleting it, consider whether it can solve a nearby layout problem.

22. Watch the official patch notes

Early Access mechanics change. A strategy from last month may be obsolete. Check the current patch notes before following an older guide or blueprint.

23. Label everything

Name your routes, buffers, and overflow paths. A factory you understand today will be confusing in a week. Labels make debugging faster and rebuilding cheaper.

24. Do not scale before testing failure

Before doubling production, test the full state: full collector, blocked destination, stopped machine, and reload. If any of these break the line, fix the failure mode before adding capacity.

25. Have fun with the physics

Sandustry is a falling-sand factory sim. Part of the joy is watching materials flow, settle, and interact. Do not optimize away every observation window. Leave a glass wall somewhere.

Sources and evidence notes

Frequently Asked Questions