Low-Cost Mine Ventilation Airflow Survey Instrument and Gallery Network Study

Low-Cost Mine Ventilation Airflow Survey Instrument and Gallery Network Study builds a compact, Arduino-based airflow measuring kit and then uses it the way a ventilation officer uses a full-size anemometer set: to run a traverse survey of a ventilation network, station by station, and turn the readings into air quantities, pressure drops and branch resistances.

Underground mines in India — the coal mines of Jharkhand and Telangana, the iron-ore mines of Ballari and Goa — depend on mechanical ventilation networks whose performance is checked by periodic airflow surveys. Commercial survey instruments (digital vane anemometers, manometers and smoke tubes) cost tens of thousands of rupees, so most college mining labs cannot give every batch hands-on practice. Our rig changes that: a pulse-output vane anemometer measures air velocity, an MPXV7002DP differential-pressure sensor reads the static-pressure difference across doors and regulators on a scale model, and an Arduino Uno time-stamps and logs every reading to a microSD card.

The survey target is a modelled gallery network — a plywood-and-acrylic single-entry system with a parallel split, an air door and a regulator, driven by a 12 V duct fan. The team calibrates the sensors against a reference anemometer and an inclined manometer, runs repeat traverses, computes air quantity (Q = v × A) and branch resistance (R = Δp/Q²) in Python, and checks how the model compares with a hand-worked Atkinson-equation calculation. The firmware is central to the build, so the source bundle includes the complete Arduino sketch and the Python reduction scripts.

Airflow surveys are the foundation of mine ventilation management, yet mining-engineering students in most Indian colleges practise them only as solved numerical problems. A survey-grade digital anemometer kit costs ₹40,000 or more, so departments cannot stock one per batch, and hands-on training is limited to a single demonstration unit. As a result, graduates join mines having never planned stations, held an instrument steady for a timed interval, or reconciled measured readings against a ventilation network calculation.

There is a need for a low-cost, loggable airflow survey instrument and a modelled gallery network on which a student team can run the complete survey workflow — station selection, velocity and pressure measurement, repeat readings, data reduction to quantity and resistance, and comparison with predicted values — at a total cost under ₹12,000, with firmware and data handling the students write and defend themselves.

Abstract

This project designs, calibrates and demonstrates a low-cost airflow survey instrument for mine ventilation teaching, built around an Arduino Uno, a pulse-output vane anemometer and an MPXV7002DP differential-pressure sensor. The instrument is validated in the laboratory and then used to conduct a formal airflow survey of a modelled mine gallery network, producing air quantities, pressure drops and resistances for each branch.

Introduction

Every underground mine is a network of connected airways whose behaviour is governed by pressure and resistance. The ventilation officer's basic tool is the airflow survey: measuring velocity and cross-sectional area at fixed stations, computing air quantity, and measuring the pressure differential across branches to derive their resistances. Comparing two surveys months apart reveals leaking doors, blocked airways or failing fans long before conditions become dangerous. Students usually see this only on the blackboard, because one survey-grade instrument set costs more than a department's annual lab budget.

Existing System and Its Limits

• Laboratory demonstration relies on a single borrowed anemometer; most students never take a reading themselves.
• Surveys are taught as formulas (Q = v × A, R = Δp/Q²) without instrument error, station selection or repeat-measurement discipline.
• There is no logged record — readings go into a notebook, so calibration drift and transcription errors cannot be checked.
• Limitation: no affordable kit exists that supports a full traverse on a physical model with automatic logging.

Proposed System

• A battery-powered Arduino Uno logger that samples the vane anemometer's pulse output at a fixed rate and the MPXV7002DP differential-pressure channel through its analog input, averages readings over a timed interval, and writes CSV rows (timestamp, pulses, pressure, temperature) to a microSD card.
• A calibration protocol: velocity calibration against a reference anemometer in the lab's wind section, pressure calibration against an inclined manometer, with a documented linear fit and repeatability check.
• A modelled gallery network of about 4 m of rectangular airway with a parallel split, one air door and one adjustable regulator, driven by a 12 V duct fan — a safe, room-sized stand-in for a mine split.
• A formal traverse survey: stations marked at the inlet, each branch and the return; repeat readings at each station; Python scripts that reduce the log to air quantity, air power and Atkinson resistance per branch.

Feasibility

• Technical: every component is hobby-grade and documented; the firmware uses only standard Arduino libraries (interrupt counting, SD, I2C). The maths is undergraduate fluid mechanics.
• Economic: the complete bill of materials comes to roughly ₹7,000–12,000, well within a final-year project grant, and the instrument remains as a reusable lab asset.
• Operational: the model is tabletop-sized, needs no approvals since it uses a fan and not a real mine, and the DGMS-style survey sheet gives students practice with the documentation discipline expected in industry.

In scope

• One logger unit with two sensing channels (velocity via pulse counting, pressure via MPXV7002DP), SD logging and LCD display.
• Calibration of both channels in the college laboratory against reference instruments available there.
• One modelled gallery network (about 4 m of airway, 150 mm × 150 mm section) with one fan, one door and one adjustable regulator.
• Traverse survey practice, data reduction scripts and a comparison of measured vs calculated resistances.

Out of scope

• Intrinsically safe (IS) certification — the rig is for lab use only and never goes underground.
• Gas monitoring (CH₄/CO), dust sampling or automatic fan control.
• A real-mine visit-based survey; the study is deliberately confined to the model.
• Wireless telemetry — data is read from the SD card after each session.

The project follows a build → calibrate → survey → reconcile sequence organised as a two-phase VTU major project (Phase-I in Semester 6, Phase-II in Semester 7).

Testing approach: instrument-level checks (zero-offset, span, repeatability), survey-level checks (mass balance — inlet quantity should equal the sum of branch quantities within a stated tolerance), and end-to-end verification of the CSV log against manual stopwatch readings. Every error source (sensor resolution, area measurement, averaging time) is listed with its estimated contribution.

The instrument follows a simple sensor → microcontroller → storage pipeline, and the study follows a measure → reduce → reconcile pipeline; the two meet in the CSV log.

1. Sensing layer — the vane anemometer outputs one pulse per revolution; the Uno counts pulses on an interrupt pin over a timed gate. The MPXV7002DP spans ±2 kPa around 2.5 V at the analog input, giving the static-pressure difference across a door or regulator.
2. Controller layer — an Arduino Uno sketch (C/C++) gates each measurement window (default 60 s), averages readings, applies the calibration coefficients stored in EEPROM, and formats a CSV row.
3. Storage layer — a DS3231 RTC time-stamps rows and a microSD module persists them, so a survey survives a power loss and can be re-analysed later.
4. Reduction layer — offline Python scripts (NumPy, Matplotlib) convert pulses to velocity, velocity and area to quantity, and quantity and pressure to branch resistance, then reconcile the network.

flowchart TD
  AN["Vane anemometer (pulse output)"] -->|interrupt pulses| U["Arduino Uno logger"]
  PS["MPXV7002DP across door/regulator"] -->|0.5-4.5 V on A0| U
  RT["DS3231 RTC"] -->|timestamp via I2C| U
  U -->|CSV row| SD[("microSD card")]
  U -->|live values| LCD["16x2 LCD"]
  SD -->|copied after survey| PY["Python reduction scripts"]
  PY --> Q["Air quantity Q = v x A per station"]
  PY --> R["Branch resistance R = dp / Q^2"]
  Q --> RC{"Reconcile vs Atkinson calculation"}
  R --> RC
  RC --> REP["Results chapter + survey sheets"]

The reconciliation step closes the loop: where measured and calculated resistances disagree by more than the stated tolerance, the team re-examines joint leakage, area estimates and averaging time before accepting the value — exactly the judgement a survey report demands.

• Add a low-range differential-pressure sensor pair or a micromanometer stage to resolve very small branch Δp values more cleanly.
• Temperature and humidity compensation of air density using a BME280, so the rig also demonstrates density correction in ventilation calculations.
• A mesh-network version with ESP32 nodes logging several stations simultaneously, turning the traverse into a near-synchronous snapshot.
• Automated regulator control on the model with a servo, enabling a small closed-loop demonstration of quantity control.
• Scale-up study: replicate the survey sheet and reconciliation workflow on published mine ventilation data to test the same pipeline on real dimensions.