Evolution of dynamic rock support systems at the El Teniente mine
El Teniente — the largest underground copper mine in the world, operating since 1905 — has grown deeper and faced increasingly complex geomechanical environments. This paper traces how its dynamic rock support systems evolved through five decades of rockburst events, from single‑sector strategies to an integrated design logic built on one idea: energy dissipated at the tunnel boundary equals the energy absorbed by the support plus the energy consumed by rock rupture.
What "damage" has looked like, across 40 years
Fig. 1, Rojas, Muñoz & Landeros Córdova (2023) — (a) early primary exploitation, 1980; (b) Sub6 sector, 1990, the event that suspended operations there.
How a blast becomes a rockburst
Caving blasts the base of a rock column; gravity keeps the fracture process going upward. A seismic event is the radiated energy of a rock‑mass rupture — its character set by the geometry, geology and structure of the volume being mined.
Fig. 2, Rojas et al. (1993), as modified and reproduced in Rojas, Muñoz & Landeros Córdova (2023).
Two ways to manage the same risk
Mitigate at the source — push large‑magnitude events away from infrastructure with preconditioning (hydraulic fracturing) and by pacing disruptive mining (blasting, extraction, undercutting).
Control at the boundary — strengthen the rock support itself so that when a seismic event does arrive, observed damage stays within what the system can absorb.
A history read in rockbursts
- 1976First rockburst damage recorded in the mine's exploitation systems.
- 1990Fatal accidents in Sub6 — sector suspended.
- 1999Experimental‑mining framework (1992–93) enters design, planning and operations.
- 2005Mw2.5 event, Reservas Norte sector — planning impact at caving scale.
- 2013–15Mw2.0 events in the New Mine Level Project's main tunnels.
August 2005, Mw2.5
Fig. 3, Gonzalez et al. (2005), as reproduced in Rojas, Muñoz & Landeros Córdova (2023) — damage that pushed rockburst control into sector-level planning.
2013–2015, Mw2.0
Fig. 4, Rojas et al. (2015), as reproduced in Rojas, Muñoz & Landeros Córdova (2023) — deeper ground, and a new round of learning.
When the system works as designed
A seismic event during re‑entry deforms the TC Tunnel's excavation boundary. The loaded support is fully discharged, mechanised scaling clears the face, and new support is installed. Photogrammetry then shows something the geotechnical inspection missed:
Fig. 7, Rojas, Muñoz & Landeros Córdova (2023) — rock deterioration behind the surface reached deeper (blue) than the geotechnical estimate at the visible boundary.
Ground control catches it first
The yellow outlines mark where geotechnical inspection saw the excavation boundary deform, right after the September 2018 event — before anyone knew how deep the damage really went.
Fig. 5, Rojas, Muñoz & Landeros Córdova (2023) — loaded rock support system, geotechnical ground control record.
Scaling back to find out how far it went
Mechanised scaling during rehabilitation kept exposing more fractured rock than the geotechnical and stress estimates predicted — the gap that sent the team to photogrammetry.
Fig. 6, Rojas, Muñoz & Landeros Córdova (2023) — operational ground control records.
When capacity is exceeded
A severe rockburst hits the same tunnel with a much larger depth of failure. Boundary fracturing looks similar to a normal deformation zone — the difference only shows up at depth.
Fig. 8, Rojas, Muñoz & Landeros Córdova (2023) — (a) roof damage with evidence of loading; (b) deformation around the rockburst.
Where anchors held, blocks mechanically interlocked with intact elements — rock and grout fracturing dissipated energy the support system never had to absorb alone.
Six frames, one recovery
Zones with deformed support (top row) worked through the same mechanised sequence as the rest of the tunnel — scale, assess, re-support — despite carrying more damage than their neighbours.
Fig. 9, Castro & Bahamondes (2020), as reproduced in Rojas, Muñoz & Landeros Córdova (2023).
Where the >100 cm came from
The bulged mesh above held fractured rock displaced more than a metre outward before the central bolts broke — load redistributed to the surrounding elements instead of letting go.
Fig. 13, Rojas, Muñoz & Landeros Córdova (2023) — photographic evidence of rock mass displacement contained by the support elements.
Interlocking, up close
Rebars stay fully grouted while the blocks around them interlock mechanically — visible proof that rock and grout fracturing were doing some of the energy dissipation the support never had to.
Fig. 10, Muñoz (2021), as reproduced in Rojas, Muñoz & Landeros Córdova (2023).
Reinforcement and retention, working as one system
Bolts reinforce the rock closest to the excavation and anchor the mesh; a second, deeper layer of cables can anchor further into undisturbed ground. Loads must transfer, in order, without any element failing before the one behind it.
Fig. 11, Rojas, Muñoz & Landeros Córdova (2023) — diagram of rock support elements and stress distribution in the surrounding area.
The same bolt, before and after a dynamic load
Cable and rebar both hold the excavation boundary in place. Once loaded, the boundary displaces toward the tunnel, and the mesh — now carrying more tensile stress — has to pass the extra load on to whichever reinforcement elements haven't failed yet.
Fig. 12, Rojas, Muñoz & Landeros Córdova (2023) — schematic representation of rebars and cable, before and after being loaded.
Distributing excess load to neighbouring elements creates an arching effect — it stabilises fractured rock instead of letting it collapse.
— after Muñoz (2019), on retention‑system behaviour
Energy at the boundary = energy dissipated by the support + energy consumed in rock rupture.
The paper's central equation — and the reason retention capacity, steel toughness and failure‑depth estimates all have to be sized together, not separately.
Shotcrete becomes standard
External elements — plates and nuts — plus shotcrete application turn the bolt‑mesh‑shotcrete system into the productive‑level baseline.
Chainlink replaces welded mesh
Higher retention capacity, and better load transfer from mesh to external elements and on to the rockbolts (Rojas 1993).
Friction bolts at the face
Block stabilisation at the working face, better confinement‑wall anchoring, and the first operational QA/QC guidelines (Celis et al. 2006).
Layered systems, dynamic testing
Two meshes with dedicated anchors per layer allow progressive load transfer; dynamic testing and better steels refine plate, nut and bar geometry (Muñoz et al. 2014).
Higher‑capacity mesh, mechanised install
Mesh with greater energy‑dissipation capacity; mechanised installation becomes central to rockburst risk control, including face‑mesh for blast‑stage worker safety.
Micro‑alloyed steel bars
Higher system capacity without changing bar geometry or installation practice — capacity gained "for free" at the install stage.
More tonnes, and a support system that kept pace
Rockbursts didn't fall as production climbed — but each annotation marks the year a new support element entered the design, plotted directly against forty years of measured production and seismicity.
Fig. 14, modified and updated from Muñoz et al. (2016), as reproduced in Rojas, Muñoz & Landeros Córdova (2023).
The support system is the last link
Every prior stage — hazard mitigation, preconditioning, mine sequencing — exists to keep events small and far from infrastructure. If the support system ends up loaded, it means those earlier stages weren't enough on their own. Its job from there is narrow and specific: reduce rock rupture, limit displacement, and keep the tunnel functional — by design allowing displacement while consuming energy as it does.