Underground Pillar Design

What the pillars are carrying, what they can carry, and the ratio between the two — plus what you give up in recovery to get it.
Underground room-and-pillar layouts only — a regular array of pillars on one horizon. It is not a barrier pillar or panel design, it does not model abutment loading from an adjacent gob, it is not a numerical stress model, and it says nothing about the back between the pillars. Bolts hold the back up; this page is about what holds the mine up. A competent person signs the layout.

A square rock pillar standing between two mined rooms in an underground hard-rock mine

Step 1 — The layout

Pillar width against room span decides both the load and the recovery.

Step 2 — Depth and stress

In-situ vertical stress is just the weight of what is above you.

Step 3 — Pillar strength

Eight published formulas, and they do not agree. Pick deliberately.

Step 4 — What the pillars look like

A number is a prediction. This is an observation, and it outranks the number.

The formulas were fitted to coal. The handbook is explicit that the statistical verification behind these constants concerns coal, and that there has been no similar verification for hard rock or even for softer minerals like trona, salt, oil shale or borates. Applying them outside coal is what the industry does, and it is still an extrapolation. The observed condition above is the check that costs nothing.
pillar stress
pillar strength
factor of safety
StepValueWhere it comes from

The layout, and what each pillar carries

Factor of safety against pillar width

The eight published strength formulas
Sourceabα βFitted to

SME Mining Engineering Handbook, Table 13.1-1, p. 1329, after Zipf 2001. Two general forms are printed with it: σP = σS(a + b·W/H) (Eq. 13.1-12) and σP = K·Wα/Hβ (Eq. 13.1-13). Obert-Duvall and Bieniawski follow the first; Salamon-Munro and Holland follow the second.

The daggers are not decoration. Three of the power-law formulas are printed “use with English units only” and one “use with metric units only”. A power law is not dimensionless, so running one in the wrong system does not scale — it gives a different answer with no sign that anything happened. This page refuses rather than converts.

A conflict worth knowing about. Rock Mechanics for Underground Mining p. 379 gives Salamon and Munro as α = −0.66, β = 0.46 — the same two numbers as the table above, with the letters swapped and a sign changed, because it writes the formula in a different form. Take α from one book and β from the other and the exponents land on the wrong variables. This page uses the SME convention throughout.

Rating a pillar by looking at it
RatingCondition

SME Mining Engineering Handbook, p. 1331, after Carmack et al. 2001. Hourglassing is the shape a pillar takes when the walls have spalled away and the load has migrated to a smaller core, which is a late stage and not a warning.

Published pillar widths by depth and mining height
Depth of cover, m

SME Mining Engineering Handbook, p. 1028, adapted from Vandergrift et al. 2004. Read down for depth of cover and across for mining height; the cell is a pillar width in feet. Mining heights run 0.9 to 2.7 m, so this is a COAL table and it is here as a sanity check on the shape of an answer rather than as a design for a hard-rock heading. The blanks are blank on the printed page: not every depth and height combination is covered, and this page does not interpolate into them.

What this tool does not know. The tributary area method assumes vertical loading only, that each pillar carries the full column above its share of the area, and that the load is spread evenly across the pillar. The handbook notes that none of the last one is true: stress concentrates at the pillar corners. It knows nothing about horizontal stress, about an abutment load from an adjacent mined-out panel, about a pillar sitting on a weak floor, or about time — pillars that stand for a decade can fail in the eleventh year. It does not design barrier pillars, and it does not check the back. Every published value has been checked against the printed page.

Spending more time on spreadsheets than on the job?

We build the reporting and admin automation for mining contractors, so the numbers keep themselves.

Book a 20 min call