IS 13365 (Part 2) : 2000Quantitative classification systems of rock mass-Guidelines, Part 2: Rock mass quality for prediction of support pressure in underground openings

IS 13365 Part 2:2000 gives the quantitative classification system of rock mass — guidelines for Rock Mass Quality (Q-system) for predicting engineering properties. It is the Barton Q-system in IS form — the second of the two routine rock-mass classifications used for tunnels, caverns and deep rock excavations in India, read alongside its sibling IS 13365 Part 1 (RMR / Bieniawski).

The Q-system is built to be support-class-directly-readable — it maps a rock-mass score to a recommended rock-support category for a given tunnel span / wall height ratio (the equivalent dimension, *De*).

It is read with the broader rock-engineering stack: - IS 13365 Part 1 — RMR (the routine cross-check classification) - IS 13365 Part 3 — quality classes for jointed rock-mass shear strength - IS 8764 — point load strength index of rocks (a feeder input) - IS 4464 — presentation of drilling/core information - IS 5878 (parts) — tunnel construction - IS 1893 — seismic input for underground/rock structures

The Q-value is the product of three pairs:

Q = (RQD / Jn) × (Jr / Ja) × (Jw / SRF)

• RQD — Rock Quality Designation (% of sound core ≥ 100 mm)
• Jn — Joint set number (1 → 20+; rising with the number of intersecting joint sets)
• Jr — Joint roughness (0.5 → 4; smooth slickensided → very rough)
• Ja — Joint alteration / filling (0.75 → 20+; clean fresh → thick clay infill)
• Jw — Joint water reduction (1.0 → 0.05; dry → catastrophic inflow)
• SRF — Stress Reduction Factor (1 → 400; competent rock → squeezing or burst-prone)

The three ratios capture three independent things: - (RQD / Jn) — block size - (Jr / Ja) — inter-block shear strength - (Jw / SRF) — active stress state (water and in-situ stress)

Range: Q runs from ~0.001 (exceptionally poor, squeezing rock) to ~1000 (exceptionally good, massive intact rock). Log scale — small Q changes hide large support changes.

Class buckets: A (exceptionally good, Q > 400) → G (exceptionally poor, Q < 0.01). Most working tunnels in Indian hydropower, metro, road and rail sit in classes C–E (Q ≈ 1 to 40).

Equivalent dimension De = (span or wall height in m) / ESR where ESR is the Excavation Support Ratio (lower for railway/road tunnels; higher for temporary mine openings). De vs Q on the support chart reads off the support class directly.

Data from a mapped reach of a 9 m span × 7 m height road tunnel in moderately-jointed quartzite:

• RQD = 70%
• Joint sets = 2 + random (Jn = 6)
• Joints rough, slightly weathered (Jr = 3)
• Joint walls clean, slightly oxidised (Ja = 1)
• Damp, no flowing water (Jw = 1)
• Moderate stress, no squeezing (SRF = 1)

Step 1 — compute Q: Q = (70 / 6) × (3 / 1) × (1 / 1) = 11.7 × 3 × 1 = ~35

Step 2 — class: Q ≈ 35 → Class C (good rock).

Step 3 — equivalent dimension. For a road tunnel, ESR ≈ 1.0. De = 9 / 1.0 = 9.

Step 4 — support category. Plot De = 9 against Q = 35 on the Q-system support chart → recommended support is systematic rock bolts at ~2.5 m spacing + 50-90 mm un-reinforced shotcrete on the crown, no steel sets.

Step 5 — cross-check vs RMR. The same reach in IS 13365 Part 1 would map to RMR ≈ 65-70 (Class II — good rock), recommending broadly the same support (bolts + shotcrete). The two classifications agreeing is what gives confidence in the support design.

Step 6 — re-rate every reach. Q is computed reach by reach; a single Q for the whole tunnel hides the dangerous poorer zones.

1. Using Q in isolation. Q and RMR are designed to cross-check each other. Designs using only one classification have repeatedly under-designed support in mixed-quality rock. Always rate both Q and RMR for the same reach.

2. Wrong SRF. SRF is the trickiest parameter — for the same rock, SRF in moderate stress is 1, but in squeezing rock conditions can be 5-20, dropping Q by an order of magnitude. Getting SRF wrong by one parameter step usually changes the recommended support class.

3. Treating Jw as static. Water inflow during excavation can drop Jw from 1.0 to 0.1, slashing Q by 10×. Design for the construction-stage water condition.

4. Ignoring random joints. Two joint sets + random is Jn = 6, not Jn = 4. Random sets count, and missing them over-rates the rock.

5. Using ESR = 1 for everything. ESR is the safety lever — 0.5 for nuclear caverns, 1.0 for road/rail tunnels, 1.6+ for temporary mine openings. Default ESR = 1 in a railway tunnel is fine; default ESR = 1 in a temporary access drift is conservative-and-expensive.

• IS 13365 Part 1 — RMR / Bieniawski rock-mass rating (cross-check)
• IS 13365 Part 3 — quality classes for jointed rock-mass shear strength (uses Q as input)
• IS 8764 — point load strength index of rocks (an alternative to UCS for the rating)
• IS 4464 — presentation of drilling/core information (the field-logging basis)
• IS 5878 (parts) — construction of tunnels (where the support classes are deployed)
• IS 11315 — rock-mass description methods (the field-mapping protocol)
• IS 13063 / IS 14448 — rock-slope and rock-anchor references
• IS 1893 Part 1 — seismic input for underground/rock structures

The Q-system is the support-design workhorse for Indian underground construction. Hydropower headrace tunnels (NHPC, NTPC, SJVN), metro running tunnels (DMRC, BMRCL, MMRC), railway tunnels (KRCL, IR new lines) and many highway tunnels (BRO, NHAI, NHIDCL) use Q for first-cut support categorisation, then refine with numerical modelling and instrumented observation.

The practitioner essentials: - Map reach by reach — typical reach length 25-50 m, shorter through transitions. - Always cross-check with RMR (IS 13365 Part 1) — disagreement between the two classifications is your signal to re-log and re-think, not pick the more favourable. - Re-rate after exposure — what the borehole said and what the heading actually exposes can diverge sharply. The observational method (re-log → re-rate → adjust support pattern) is built into the Q approach. - SRF deserves separate thought — it is the only parameter that depends on the engineering context (size, depth, stress state), not just the rock. A specialist call on SRF in deep or squeezing rock is worth its cost many times over.

Q and RMR remain *empirical correlation tools*, not analyses — their value is converting consistent field logging into a defensible first-cut support class and a strength estimate. For major works (large caverns, deep tunnels in difficult ground), the empirical class is confirmed by numerical modelling (FLAC, Phase2/RS2) and, decisively, by instrumented observation during excavation. The rock that is exposed always overrules the rock that was logged from a borehole.