Log Reduction
“Our sanitizer kills 99% of bacteria!” Sounds impressive — until you realize 99% of a million still leaves ten thousand survivors. The log scale exists because percentages lie to human intuition. Log reduction is how microbiologists talk honestly about killing.
Definition
Log reduction measures microbial destruction on a base-10 logarithmic scale. Each log is a tenfold drop: 1-log kills 90%, 2-log kills 99%, 3-log kills 99.9% — each step destroying 90% of what survived the last one.
Key facts
- The scale matters because microbial populations live in orders of magnitude. “99% killed” versus “99.999% killed” differ by a factor of a thousand in survivors — but they sound almost the same in percentages.
- The regulatory targets you’ll meet: 5-log pathogen reduction in juice (FDA juice HACCP), 6.5 to 7-log Salmonella reduction in cooked meats (USDA), 12-log for C. botulinum in low-acid cans.
- Sanitizers and disinfectants earn their claims in log terms — e.g., a validated 5-log reduction of specified organisms within a defined contact time.
- The math: log reduction = log₁₀(starting count) − log₁₀(final count). And remember what it doesn’t promise — survivors can remain. Log reduction describes the process, not sterility.
- Validation — challenge tests, thermal-death data — proves the process delivers its claimed reduction under worst-case conditions, not just in the lab’s best run.
Why it matters
Misreading this scale is a classic and dangerous error. “99.9% effective” gets presented as “basically sterile” — it’s not; it’s 3-log, leaving one in a thousand alive. When you’re evaluating a sanitizer claim, a cooking validation, or a supplier’s kill-step data, think in logs: how many orders of magnitude, against which organism, under what conditions? That question cuts through marketing like nothing else.
Sources: Codex Alimentarius; FDA; WHO.
Quick reference
| Log reduction | % destroyed | Survivors of 1,000,000 |
| 1-log | 90% | 100,000 |
| 3-log | 99.9% | 1,000 |
| 5-log | 99.999% | 10 |
| 7-log | 99.99999% | ~0.1 |