Sawmill Recovery Rate: Benchmarks and How to Improve Yours
Lumber recovery factor benchmarks — around 38% in EU hardwood mills, 48% in softwood, 50% global average — how overrun differs, and the levers that actually raise recovery.
What recovery rate means
Recovery (lumber recovery factor) is the sawn lumber volume out divided by the roundwood volume in, times 100: a mill cutting 100 m³ of logs into 48 m³ of lumber runs at 48% recovery. Everything else leaves the saw hall as slabs, edgings, sawdust and chips. The FAO global average sits near 50%; European softwood mills typically run 46–50% and hardwood mills 36–40%.
Recovery vs overrun — don't mix them up
Overrun compares the lumber you actually cut against a log rule's estimate (Doyle, Scribner, International); recovery compares volume out against volume in. A mill can show a +30% Doyle overrun and still run a mediocre 45% recovery — overrun measures the rule's error, recovery measures the mill's efficiency. Use the rule when buying logs and recovery when judging the saw line; Doyle, Scribner and International Log Rules Compared explains the rules, this page the efficiency side.
Benchmark table
Published ranges worth calibrating against:
| Mill type | Typical recovery |
|---|---|
| EU hardwood sawmills | 36–40% |
| EU softwood sawmills | 46–50% |
| FAO global average | 42–60% (mid ≈ 51%) |
| Tropical meranti, Malaysia | 55–57% |
| SE-Asia small mills, band saw | 50–54% |
| SE-Asia small mills, chainsaw | 37–41% |
| Vietnam tropical mills, typical | 35–55% |
Where the other half goes
A typical softwood mill splits the log roughly: lumber 48%, chips 30%, sawdust 15%, slabs and edgings 7%. The saw kerf alone — 2 to 5 mm per cut — removes 8–12% of log volume in a common softwood pattern. Whether the remaining half costs money or pays depends on residue markets: chips feed pulp and MDF, sawdust feeds pellets and bedding, and slabs sell as firewood or biomass.
Five levers that raise recovery
In order of payback: (1) sharpen more often and cut a thinner kerf; (2) match the sawing pattern to each log class — cant sawing beats random sawing on small logs; (3) optimise placement with scanning, even at a basic level; (4) cut for grade, not for volume — one clear board beats three knotty ones at the same volume; (5) measure recovery per shift so drift is caught the day it starts. Sawmill recovery computes recovery from volume in vs out and flags the benchmark band your mill falls into.