Bratticing Changes That Reshape Acoustic Interpretation
The Acoustic Consequences of Moving a Curtain
Research in Sound Propagation and Speech Transmission in a Branching Underground Tunnel found that low frequencies attenuate around intersections differently than high frequencies, with the geometry of the intersection itself shaping the acoustic response. Add a brattice curtain — a heavy fabric barrier used to direct airflow — and the acoustic response changes again. The Nature Scientific Reports paper on obstacle parameters and mid-high frequency noise propagation shows how obstacles alter mine-face noise propagation across frequency bands. Brattice is one of the most common obstacles in any active or recovering mine, and it gets moved repeatedly during a rescue to maintain fresh-air supply as the advance progresses.
Historic Bureau of Mines research documented in Use of Brattice Cloth in Coal Mines details how brattice placement affects airflow distribution at the face — placement as close as 10 feet has been shown in Replacing Brattice Cloth with Air Curtains to outperform 18-foot placement or air-curtain alternatives. NIOSH's Handbook for Dust Control in Mining documents brattice stopping lines, leakage impacts, and the tight tolerance between effective and ineffective placements. NIOSH Ventilation Planning Layouts for Large-Opening Mines reminds coordinators that stoppings closest to the fan are the most critical — move one of those and the whole circuit shifts.
Every one of those shifts also shifts the acoustic map. The quilt patches that were calibrated against the pre-move airflow and geometry no longer describe what the microphones are recording. If the command post ignores the change, the next few hours of patches will accumulate errors that compound.
The compounding is asymmetric in a way that catches command posts off guard. A small brattice move that redirects only 8 percent of the section airflow can change the acoustic propagation pattern by 15 to 20 percent across half a dozen patches downstream, because the airflow change interacts nonlinearly with the dust loading and the ventilation-noise floor. A captain who looks at the brattice change as a "minor" ventilation tweak will be surprised when the quilt's confidence band on a downstream pillar drops sharply within the next ten minutes. Conversely, a large brattice move that fully reverses an entry's airflow direction sometimes produces less acoustic confusion than a small move, because the new pattern is so different from the old that the classifier registers a clean transition rather than a contaminated overlap.
The takeaway for coordinators is to treat every brattice change as a mapping event and let the system decide its weight, rather than trying to filter out changes the captain considers minor.
Stitching Brattice Events Into the Quilt
EchoQuilt treats brattice relocation as a first-class map event. When a rescue team moves a curtain, the event is logged with a location, a direction of air redirection, and a timestamp. The quilt flags all downstream patches — patches whose acoustic signature depends on the affected airflow — as needing re-observation. The command post sees these patches turn amber, signaling that the previous acoustic data is historical and new data is needed before the patches are trusted again.
The stitching logic distinguishes three kinds of brattice-driven changes. First, direct acoustic changes: the curtain itself is an obstacle that reflects and absorbs sound, so any patch whose line-of-sight acoustic path passes through the brattice location must be re-measured. Second, airflow-driven changes: the new airflow carries ambient noise — tool vibration, ventilation hum, face fan drone — along a different path, so patches previously quiet may now carry background, and vice versa. Third, humidity changes: fresh air from an opened crosscut changes humidity levels, which affects high-frequency attenuation, so the quilt's calibration model for each affected patch must update.
The command post's brattice-event workflow is deliberately lightweight. When a captain radios or logs "brattice moved at crosscut 14 east, directing fresh air north," the briefing officer selects the brattice on the tablet and drags the new position. EchoQuilt cascades the re-observation flag automatically through the affected quilt region. The next rescuer passing through that region rebuilds the patches with fresh acoustic data. Total overhead for the command post: roughly 15 seconds of tablet interaction per brattice event.
Separating brattice-driven acoustic changes from collapse-driven acoustic changes is critical. A patch flagged because of a brattice move should not be interpreted as a sign of new collapse activity, and a patch flagged because of real collapse should not be explained away as a brattice effect. This separation uses the same signal source separation disciplines biologists apply when distinguishing bat echolocation from ambient stream noise — the underlying math is comparable, and the operational principle is identical: different signal sources must be tagged at ingest or they cannot be untangled later.
Advanced Tactics for Brattice-Aware Mapping
Three tactics separate brattice-aware mapping from ventilation-ignorant mapping. First, maintain a brattice inventory in the pre-incident map. Every permanent stopping, regulator, and check curtain should be tagged in the reference layer with its default position. Temporary brattice placed during normal mining gets tagged too, with a last-observed timestamp. When the rescue begins, the command post starts with a complete picture of where the ventilation controls are supposed to be — then tracks changes from there. This connects directly to ventilation signatures analysis after a roof fall, because a fall that destroys a brattice is simultaneously a geometry event and a ventilation event.
Second, coordinate with the ventilation engineer on the command-post staff. The mine's ventilation engineer should own the brattice layer on the EchoQuilt tablet. When rescue captains radio brattice moves, the ventilation engineer updates the layer and signs off on the new airflow model. This separation of duties keeps the geometry team and the ventilation team from stepping on each other's state, and it gives the IC a single authoritative voice on ventilation questions.
Third, expect brattice-driven acoustic noise to peak in the first 10 to 20 minutes after a move. The curtain itself flaps audibly as the airflow stabilizes; the fan load changes; rib dust gets stirred. EchoQuilt downweights acoustic data from this settling period. Rescuers passing through within the first 10 minutes of a move should know their observations will inform geometry more than structural-integrity assessments. Distinguishing this noise from actual settling interpretation after a collapse is one of the harder inference problems in the system and benefits from explicit brattice-event context.
A common mistake is to treat the brattice layer as a ventilation-only concern and exclude it from the rescue-team briefing. Rescuers should know the brattice state before they enter a section, because the curtain will affect their own acoustic footprint — their footsteps and breathing will generate different echoes on the other side of the curtain than on the near side.
A second common mistake is to treat brattice cloth itself as homogeneous when it is not. Heavy duct brattice, lightweight check curtains, mine-grade plastic, and emergency-deployed canvas all have distinct acoustic-absorption profiles, and the absorption profile changes the high-frequency attenuation downstream. Coordinators who maintain mixed brattice inventories should tag each curtain type in the reference layer so the quilt's calibration model picks the right absorption curve. Sites that standardize on a single curtain type — typically heavy duct brattice for permanent stoppings and a single check-curtain SKU for temporary placements — see lower acoustic-interpretation overhead than sites that allow individual sections to choose curtain materials. The standardization conversation often surfaces during the procurement review for any new mining technology and is worth the discomfort it sometimes generates among section foremen who prefer the curtain they grew up with.
Join the Waitlist for Mine Rescue Coordinators
Ventilation engineers and mine rescue coordinators responsible for active ventilation circuits during rescues can request early access to the brattice-aware build of EchoQuilt. We import your current ventilation plan, tag permanent stoppings and regulators on the reference layer, and train your staff on the brattice-event workflow. Priority goes to gassy coal mines operating under 30 CFR 75 ventilation plans and to metal/nonmetal operations with complex multi-fan circuits. Send us your current mine ventilation schematic and we will pre-build the brattice inventory for your pilot. The pilot also includes a curtain-type absorption library tuned to your standardized brattice inventory and a tabletop drill replaying a recorded brattice-cut event with full intake-return reversal logged against your AMS sensor net.