To improve water quality in Probably Stolen, install Quality-boosting modules such as Quality Modules or Refinement Modules in your Water Purifier or Moisture Farm. Both machines use standard module bays, where the installed modules' Performance, Efficiency and Quality values are added together. Choose modules for the machine you are upgrading: higher Quality affects its output, while Performance and Efficiency govern speed and power use.
Water Purifier and Moisture Farm quality targets
A community water guide reports that a Water Purifier with Quality of at least 100% can reach a maximum water purity of 99.9%. The creator connects that result to the purifier's Quality and also describes a tooltip showing the maximum purity. Treat this as a demonstrated target for the setup shown, rather than an official universal formula for every water-processing situation.
The same guide gives different thresholds for the Moisture Farm: Quality 100 produces high-quality water, while Quality 150 or above produces pure water. These are reported output-grade thresholds for that machine, so the purifier's target should not be substituted for the farm's pure-water target. The claims concern water quality and do not establish a fixed production volume, which matters when deciding what an upgrade will actually solve.
Choose between Quality, speed and power savings
The standard Quality Module provides Quality +8% and is listed at 70 Credits, making it the straightforward option when you want that stat. The standard Refinement Module provides Quality +25%, but also Performance -8% and Efficiency -15%, at a listed price of 100 Credits. Refinement therefore comes with speed and power tradeoffs; its larger Quality contribution should be considered alongside the machine's other totals.
Performance determines speed through the formula base speed plus base speed multiplied by Performance and divided by 100. The standard Performance Module supplies Performance +16% for 70 Credits, whereas the Overclock Module supplies Performance +36% and Efficiency -44% for 100 Credits. The water guide specifically makes overnight completion of an entire jug dependent on machine Performance, so a quality-focused purifier is not automatically an overnight solution.
Efficiency determines power through base power minus base power multiplied by Efficiency and divided by 100, with ordinary power use at least 1. The standard Efficiency Module adds Efficiency +16% for 70 Credits, while the Eco Module gives Efficiency +32% and Performance -16% for 100 Credits. Negative Efficiency increases power use, so read that sign carefully on Refinement and Overclock modules; low power mode is a separate case that sets machine power use to 0.
Fit modules and expand the standard bay
The standard bay measures 7 x 5 cells, but its initial layout has only 26 usable cells because the corners and central plus shape are blocked. The next stage has 31 usable cells with the corners still blocked, and the final stage opens all 35 cells. A Module Bay Expansion Kit opens the next stage, and the water guide demonstrates using up to two kits to open the bay fully.
When arranging modules, the water guide shows E and Q rotating a selected module and right-click flipping it. Use those controls to try another orientation before assuming a module cannot fit around the bay's blocked spaces. If an existing module is stuck, the Wiki identifies the Module Extractor as the tool that frees it, which is a different issue from needing more open cells.
Check random shapes and trader stock
Random module shapes can change both stats and value, so a familiar module name does not guarantee the standard version's contribution. The Wiki gives C-shaped Performance, Efficiency and Quality modules a stat multiplier of x0.75, while S-shaped and H-shaped Overclock, Eco and Refinement modules receive x1.25. Compare the actual shape and stats when judging a water upgrade, especially when you are trying to reach a particular machine Quality target.
Dr. Jackson rolls random module stock on each visit, including Performance, Efficiency, Quality, Overclock, Eco, Refinement and Node modules. That stock system does not guarantee a particular purchase, so a plan requiring a Refinement Module still depends on what is available. From day 21, or from the start when playing without the intro, random modules can also roll 0 to 2 effects.
Read placement bonuses and effects before installing
Top Mount adds +20% of a module's stats while it touches the bay's top edge, and Side Mount adds +20% while it touches the left edge. Those conditions make the position of an affected module relevant alongside whether its shape fits. Premium multiplies base stats by x1.2 and costs more, while Inferior multiplies them by x0.8 and costs less, so check effects as well as the module type.
A Node (Medium) has no base stats and feeds adjacent Receiver modules, whose Receiver effect adds +10% of their stats per adjacent Node. Random nodes always have Support Node, which gains 20% of the stats of each adjacent module. A node's usefulness therefore depends on its neighbors and their effects; do not count it as a standalone Quality Module when planning the bay.
Overcharged doubles stats but has a 10% chance of ruining the module on each use, while Degrading doubles stats and lowers every positive stat by 1 per use. Learning Algorithm instead starts positive stats at 0 and raises them by 1 per use until they reach twice their original values. Because use-based effects count whenever the machine works, these modules can change your setup over time rather than providing a permanently fixed contribution.
Consider the Neural Core tradeoffs
The Neural Core Module (Capped) lists Performance +100%, Efficiency -100% and Quality +50%, while the Uncapped version lists Performance +200%, Efficiency -100% and Quality +100%. The Wiki classifies them as High-Level Contraband and Critical-Level Contraband respectively, making their classification part of the comparison alongside the attractive stats. Both carry negative Efficiency, so their Quality and speed gains also come with increased power use.
Test water and distinguish treatment tools
The official store description says water must be tested and filtered, and the Destructoid guide explains testing by dragging testing strips into the water source. That guide recommends testing when buying water bottles or containers, rather than selecting stock without checking it. During treatment, the community water guide describes using AquaScan periodically to see how clean the water has become.
The community guide distinguishes UV lamps, which target microbes, from water filters, which target organic waste, mineral and physical contaminants. It demonstrates UV treatment plus filtration for purchased basewater, but warns that self-produced basewater may respond differently depending on how it was made. Check the water itself when choosing treatment, because matching grade labels do not guarantee that the same combination will produce the same result.