Skip to content

Repository files navigation

Water Chemistry Engine

A reusable, scientifically grounded Python engine for characterizing, blending, treating, comparing, and eventually optimizing water for brewing, fermentation, and other validated uses.

The engine is intentionally independent of web frameworks, databases, graphical interfaces, and product-specific persistence. End-user applications consume this package from separate projects.

Project status

The 0.3 supported-consumer-API milestone is complete. The engine can resolve reported source-water chemistry, blend multiple characterized sources, apply supported mineral additions, calculate the resulting water, compare it with target/reference criteria, and return auditable contribution, instruction, and notice data.

Python 3.11 is the project compatibility baseline, with CI coverage through Python 3.14. Public APIs remain pre-1.0 and may evolve as real consumer applications exercise the engine.

Version 0.3 establishes a supported package-root consumer facade covering both the deterministic forward-result graph and the complete source-reporting and provenance input graph. Its compatibility expectations are documented in the consumer API guide. Reported and target pH use FermUnits' semantic PHValue; calculated working-water pH remains explicitly deferred until a validated reusable model is ready.

Forward-calculator capabilities established in 0.2

  • source-water and target/reference profile models;
  • preservation of exact values, ranges, bounds, ND, reported statistics, reporting context, source-document metadata, and chlorine/chloramine data;
  • explicit source-resolution policy, including opt-in exact-range midpoints;
  • fixed multi-source blending with conservative unknown propagation;
  • supported mineral-addition stoichiometry and deterministic treatment;
  • source, blend, and final target/reference comparison;
  • combined source/treatment ion-contribution reporting;
  • structured preparation instructions;
  • structured notices for assumptions, unresolved inputs, model limitations, and deferred target-pH calculation.

Planned after 0.3

  • 0.4: curated target/reference profiles and practical treatment materials;
  • 0.5–0.6: automatic and ranked treatment optimization;
  • 0.7: reusable working-water pH if a defensible model is ready;
  • 0.8: BeerJSON/FermentationJSON interchange, conformance work, and 1.0 hardening.

AI-assisted document ingestion, accounts, persistence, browser UI, and native application code belong to separate consumer applications rather than this repository.

Repository structure

water-chemistry-engine/
├── src/
│   └── water_chemistry_engine/
├── tests/
├── docs/
├── reference-data/
├── schemas/
├── scripts/
└── test-vectors/

src/water_chemistry_engine

The importable Python engine package, using the standard src/ layout. It owns scientific calculations, domain models, validation, warnings/notices, optimization as it is added, and structured calculation results.

External applications such as web, automation, Mechani-Brew, and future native clients can consume the engine while providing their own interfaces and persistence.

Installation

Water Chemistry Engine requires Python 3.11 or newer. Install the published package with uv or pip:

uv add water-chemistry-engine==0.3.0

or:

python -m pip install water-chemistry-engine==0.3.0

Version 0.3 exposes a supported package-root facade. APIs remain pre-1.0 and may evolve under the consumer API compatibility policy.

Quickstart

This example resolves and blends equal volumes of two reported source waters. Calcium therefore blends from 40 mg/L and 80 mg/L to 60 mg/L:

from fermunits import Q_

from water_chemistry_engine import (
    ForwardWaterSource,
    Ion,
    IonConcentration,
    SourceResolutionPolicy,
    SourceWaterProfile,
    calculate_forward_water,
)

source_a = SourceWaterProfile(
    name="Source A",
    concentrations=(IonConcentration.mg_per_liter(Ion.CALCIUM, 40.0),),
)
source_b = SourceWaterProfile(
    name="Source B",
    concentrations=(IonConcentration.mg_per_liter(Ion.CALCIUM, 80.0),),
)

result = calculate_forward_water(
    (
        ForwardWaterSource(source_a, Q_(1.0, "liter")),
        ForwardWaterSource(source_b, Q_(1.0, "liter")),
    ),
    source_resolution_policy=SourceResolutionPolicy(allow_exact_range_midpoints=False),
)

calcium = result.final_state.concentration_for(Ion.CALCIUM)
assert calcium is not None
print(calcium)

Expected output:

60.0 milligram / liter

The explicit source-resolution policy prevents the example from silently choosing representative values for ranges. See the consumer API guide for the complete supported workflow, including treatments, targets, notices, and audit results.

Development stack

  • Python 3.11+ (CI tests 3.11–3.14; 3.11 is the compatibility baseline)
  • uv
  • FermUnits
  • pytest and Hypothesis
  • Ruff
  • mypy
  • GitHub Actions

Development

Install uv, then run from the repository root. The checked-in .python-version selects the Python 3.11 compatibility baseline:

uv sync --dev
uv run pytest

The full CI gate tests Python 3.11, 3.12, 3.13, and 3.14. It also checks the lockfile, formatting, linting, strict typing against Python 3.11 semantics, and builds the engine distribution.

Documentation

Primary project documents are stored under docs/:

  • WATER_CHEM_DESIGN.md — scientific and architectural design;
  • WATER_CHEM_REFERENCES.md — source and reference register;
  • ROADMAP.md — active engine release path;
  • PROJECT_STRUCTURE.md — repository/package boundaries;
  • CONSUMER_API.md — supported 0.3 package-root facade and integration guide;
  • reviews/ — point-in-time external review records.

Release history is summarized in CHANGELOG.md.

License

This project is licensed under the Mozilla Public License 2.0.

About

A reusable, scientifically grounded Python engine for characterizing, blending, treating, and comparing water chemistry for brewing, fermentation, beverages, food, and other uses.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages