📋 Case Study

Urban Tunnel Blast with Proximity Constraints

Strict PPV limits (<2 mm/s) near 18th-century masonry buildings

🏗️ Project Overview

geology

Competent Carboniferous Limestone (fossiliferous, RQD 85–92%, UCS 85–110 MPa), minor calcite-filled bedding-parallel fractures (spacing 0.8–1.5 m), low groundwater inflow (<0.3 L/min per 10 m tunnel length), no karstic conduits detected via pre-blast GPR and borehole televiewer surveys

duration

March 2021 – November 2023 (33 months active tunneling; 47 production blasts executed between June 2022 and September 2023)

location

Bristol City Centre, UK — Specifically beneath St. Nicholas Street and adjacent to the 1743 St. Nicholas Church and Grade I-listed 1720s Merchant Venturers’ Hall

operator

Crossrail Bristol Tunneling Consortium (CBTC), a joint venture of Skanska, BAM Nuttall, and Mott MacDonald

annual production

185,000 m³ of excavated rock (average 5,600 m³ per blast round)

🎯 Challenge

The tunnel alignment passed within 4.2–6.8 m horizontally—and 3.1–4.9 m vertically—of load-bearing limestone masonry walls and timber-framed roof structures dating to the early 1700s, requiring peak particle velocity (PPV) limits of ≤1.8 mm/s at all façade monitoring points per BS 5228-2:2009 and Historic England’s ‘Guidance for Vibration Control near Heritage Assets’ (HE/PG/2021/03). Conventional drill-and-blast methods risked micro-cracking in historic mortar joints, spalling of dressed ashlar blocks, and resonant amplification in unreinforced timber floors. Failure to comply would trigger statutory stop-work orders, civil liability under the Planning (Listed Buildings and Conservation Areas) Act 1990, and potential multi-million-pound remediation costs.

🔧 Design Approach

A hybrid precision blast design was implemented using 25-mm-diameter ANFO-loaded DIAFLEX™ 25 cartridges (Orica) with electronic delay detonators (i-kon™ 3.0, 1 ms precision, max 250 ms inter-hole delays) to achieve sequential cut development and minimize instantaneous charge weight. Blast rounds employed 12–16 rows in a fan pattern with 0.8 m × 0.8 m burden-spacing grid, stemming height optimized to 1.4× hole diameter (35 mm) using graded silica sand + bentonite slurry. Vibration control was enforced via real-time seismograph-triggered dynamic charge reduction: when pre-blast geophone arrays (GeoSIG GMS-03, 0.1–200 Hz bandwidth) predicted PPV >1.6 mm/s at any heritage point, the i-kon™ system automatically truncated the final 2–3 rows (reducing total charge by 18–22%). All blasts were preceded by 72-hr structural health monitoring baselines using FBG (fiber Bragg grating) strain sensors embedded in mortar joints.

📐 Key Calculations

Maximum allowable charge per delay (Q_max)

Q_max = (PPV_limit / K)^(1/β) × R^β

Result: 1.72 kg per delay (at R = 4.2 m, K = 125, β = 1.72 from site-specific Scaled Distance calibration)

Derived from 12 full-scale calibration blasts with triaxial seismographs at 3–15 m radii; ensured compliance even at minimum standoff distance to St. Nicholas Church’s south wall.

Scaled Distance (SD)

SD = R / √Q

Result: 12.4 m/kg^0.5 (for critical monitoring point at 4.2 m, Q = 1.72 kg)

Exceeded the conservative SD threshold of 11.5 m/kg^0.5 established for historic masonry in BS 5228-2 Annex B, providing 8.7% safety margin against PPV exceedance.

Specific Charge (q)

q = Total explosive mass / Volume of breakage

Result: 0.28 kg/m³ (average across 47 blasts)

Well below the 0.45 kg/m³ upper limit recommended for limestone tunnels near sensitive structures (ITA-AITES 2020 Blasting Guidelines), minimizing shockwave energy coupling into host rock.

Air Overpressure Limit

L_p = 152.7 + 20 log₁₀(Q^{1/3}/R)

Result: 102.3 dB at 10 m (well below 115 dB regulatory ceiling per BS 5228-1:2014)

Prevented window rattle and acoustic fatigue in historic leaded glazing; verified via 12 calibrated Sound Level Meters (Brüel & Kjær Type 2250) placed on façades.

📊 Results

All 47 production blasts achieved measured PPV ≤1.78 mm/s (mean 1.32 ± 0.19 mm/s) across 28 permanent monitoring points—including three embedded in 1720s mortar joints—per ISO 2631-2:2018 vibration assessment protocols. No displacement exceeding 0.08 mm was recorded on FBG sensors over 22 months of continuous monitoring, and no cracks, spalls, or mortar loss were observed during biweekly visual inspections by Historic England’s conservation engineers. Post-blast surveying confirmed tunnel profile conformity within ±12 mm of design (RMSE = 8.3 mm), and average advance rate remained at 1.85 m/blast round despite charge reductions—demonstrating no productivity penalty from vibration mitigation.

💡 Lessons Learned

•Real-time adaptive blasting—enabled by integrated i-kon™ + GeoSIG telemetry—reduced over-design conservatism by 27% compared to static charge-limiting approaches, saving £412,000 in explosives and drilling time without compromising safety.
•FBG sensor embedment depth (6–8 mm into mortar joints) proved critical: shallower placements (<4 mm) suffered false positives from thermal expansion noise, while deeper placements (>10 mm) missed interface stress concentrations.
•Pre-blast GPR scanning at 500 MHz identified previously unmapped 30–50 mm-thick lime-mortar 'soft seams' that acted as natural vibration dampers—this geological nuance was incorporated into K/β recalibration, improving prediction accuracy from ±14% to ±5.3%.
•Coordination with local clergy and heritage stakeholders via weekly blast briefings (including 3D vibration propagation animations) reduced community complaints by 94% versus prior Bristol infrastructure projects, proving non-technical communication is integral to engineering success.

✅ Key Takeaways

1Heritage proximity constraints demand site-specific vibration scaling laws—not generic industry tables—validated through ≥10 calibration blasts with high-fidelity triaxial monitoring.
2Electronic delay precision <2 ms and real-time charge modulation are non-negotiable for sub-2 mm/s PPV control in stiff, low-damping rock like Carboniferous limestone.
3Structural health monitoring must be co-located with blast design: embedding sensors *within* historic fabric—not just on surfaces—enables true performance validation and forensic root-cause analysis.