EPANET AI Assistant

A free, browser-based AI chatbot for analyzing EPANET water distribution network models. Runs entirely as a static page — no server, no installation, no cost.

---

Overview

EPANET AI Assistant lets water engineers and researchers upload an EPANET .inp file and ask natural-language questions about the network. It is powered by Compound Beta via the Groq API and has built-in knowledge of:

• EPANET 2.3 — all .inp sections, PDA, leakage modeling, PCV valve, new API (ENgetnodevalues, ENgetlinkvalues, ENtimetonextevent…)
• EPyT — EPANET Python Toolkit, 500+ API methods with code examples (OpenWaterAnalytics/EPyT)
• WNTR — Water Network Tool for Resilience, pressure-driven simulation, resilience metrics (USEPA/WNTR)

---

Features

• Upload EPANET .inp files directly in the browser (drag & drop supported)
• AI answers about hydraulics, water quality, network components, and Python code
• Network summary panel (junctions, pipes, reservoirs, tanks, pumps, valves, all detected sections)
• Conversation history within the session
• 10 suggested questions — click to send instantly
• API key stored only in your browser (localStorage) — never sent anywhere else
• Mobile-friendly responsive layout

---

Live App

Deploy to GitHub Pages in a few clicks — see Deploy below.

---

Quick Start

1. Get a Free Groq API Key

No credit card required.

1. Sign up at console.groq.com
2. Go to API Keys → Create API Key
3. Copy the key (starts with gsk_)

Free tier: 14,400 requests/day · Compound Beta (free on Groq)

2. Open the App

Open index.html in any modern browser, or deploy it to GitHub Pages.

3. Start Chatting

1. Paste your Groq API key → Save
2. Upload an EPANET .inp file
3. Ask questions or click a suggested question

---

Deploy to GitHub Pages

1. Fork or push this repository to your GitHub account
2. Go to Settings → Pages
3. Under Source, select Deploy from a branch → branch main → folder / (root)
4. Click Save — your app will be live at https://<username>.github.io/<repo-name>/

Make sure index.html is at the root of the repository (or the Pages source folder).

---

Example Questions

Category	Question
Network overview	Summarize this network
Components	What are the pipe diameters and lengths?
Hydraulics	Are there any potential pressure problems?
Pumps	What pump curves are defined?
Quality	Explain the water quality settings
EPyT	How do I run this network with EPyT in Python?
WNTR	How do I analyze resilience with WNTR?
WNTR	How do I simulate a pipe leak using WNTR?

---

Example Network

The examples/ folder includes L-TOWN.inp — the benchmark network from the Battle of Water Networks (2020). Load it to test the assistant.

---

Files

File	Description
index.html	Complete app — HTML, CSS, and JavaScript in one file
examples/L-TOWN.inp	Benchmark water network for testing

---

Privacy

• Your Groq API key is stored only in your browser (localStorage) and sent only to api.groq.com
• Your .inp file is read locally and never uploaded to any server
• Conversation history exists only in memory and is cleared on page reload

---

Built With

• Groq API — ultra-fast LLM inference
• Groq Compound Beta — Groq's compound AI system with tool use and web search
• EPANET 2.3 — hydraulic and water quality modeling
• EPyT — EPANET Python Toolkit (KIOS Research Center, University of Cyprus)
• WNTR — Water Network Tool for Resilience (US EPA)

---

Citation

If you use this tool in your research, please also cite EPyT and WNTR:

EPyT:

Kyriakou et al. (2023). EPyT: An EPANET-Python Toolkit for Smart Water Network Simulations. Journal of Open Source Software, 8(92), 5947. https://doi.org/10.21105/joss.05947

WNTR:

Klise et al. (2023). Water Network Tool for Resilience (WNTR) User Manual Version 1.0. U.S. Environmental Protection Agency.

---

License

MIT — free to use, modify, and distribute.