From Geology to Indoor Radon

Indoor radon concentration is the product of multiple independent factors, not their sum. The chain runs: uranium-238 (half-life 4.47 billion years) decays through thorium-234, protactinium-234, uranium-234, and thorium-230 to radium-226 (half-life 1600 years), which decays to radon-222 (half-life 3.82 days).

Each subsequent stage filters or modulates the signal: emanation releases only 3–70% of generated radon from grains into pore spaces (typically 13% for rock, 20% for soil). Soil transport through diffusion and advection depends on permeability (spanning 12 orders of magnitude from 10⁻⁷ m² for gravel to 10⁻¹⁹ m² for glacial till) and soil-gas concentration (typically 5–100 kBq/m³, extremes in tens of thousands).

Building entry driven by tiny pressure differences (~1–10 Pa from stack effect and wind) pulls soil gas through cracks and joints. Ventilation dilutes indoor air (median ~0.45 ACH in US homes). Progeny equilibrium factor F ≈ 0.4 determines actual alpha-emitting activity. Because these factors multiply, the outcome is far more sensitive to transport and building characteristics than to uranium content alone.

Bulk silicate Earth contains about 18–23 ppb uranium. Because uranium is highly incompatible in mantle minerals, it was extracted into the continental crust during planetary differentiation. The upper crust averages ~2.7 ± 0.6 ppm, the lower crust ~0.2 ppm, and oceanic crust ~0.1 ppm.

USGS states most rocks contain 1–3 ppm uranium. Granites, light-colored volcanics, black shales, and phosphatic rocks may contain up to 100 ppm. Kansas black shales contain 15–100 ppm, with phosphate nodules reaching 350 ppm. In the Sierra Nevada batholith, samples range from 0.17 ppm (gabbro) to 6.6 ppm (quartz monzonite).

This creates a three-order-of-magnitude range in source strength, but radium-uranium disequilibrium often decouples the two. Swiss Jura karst soils show radium-226 and thorium-230 enriched up to a factor of 20 relative to uranium parents because of differential mobility in water.

When radium-226 decays, the newly formed radon-222 atom recoils with about 86 keV energy, traveling 20–70 nm in mineral crystals before stopping. Only atoms born within one recoil range of a grain surface, and recoiling outward, escape into pore spaces.

Representative emanation fractions: 0.03 for intact minerals, 0.13 for rocks, 0.20 for soils. Pore water raises emanation by up to a factor of 5 because water stops recoiling atoms more effectively than air, preventing them from re-entering adjacent grains. However, recoil range in water (~77 nm) is actually shorter than in air (~53 µm) — the enhancement comes from preventing backscatter.

Most radon never leaves its birth grain. This is the first major filter: even in uranium-rich rock, if grain size is large and porosity low, most generated radon remains trapped.

Soil permeability ranges over 12 orders of magnitude. Soil-gas radon is typically 5–100 kBq/m³, with extremes in tens of thousands. Diffusion length in dry soil is ~0.5–2 m, limited by radon's 3.82-day half-life. Wet soil reduces diffusion coefficient from ~10⁻⁶ m²/s (dry) to ~10⁻⁹ m²/s (saturated).

Advective flow becomes significant above permeability ~10⁻¹² to 10⁻¹¹ m². Faults, karst conduits, and backfilled utility trenches act as fast pathways. This stage introduces the largest physical variability: neighboring sites can have soil-gas radon differing by a factor of 250 within 30 feet despite similar bedrock.

Large UK and French datasets show geology explains only about 10–25% of the variance in log indoor radon. House factors explain less than 10%. The remainder is soil transport variability, building entry characteristics, and occupant behavior — all multiplicative factors that dominate the geological signal.

Low-uranium karst limestones produce some of Europe's and Kentucky's highest indoor radon because of high permeability fracture networks and radium enrichment in residual soils. Uranium-rich bedrock under low-permeability clay can produce low indoor values because radon cannot migrate efficiently.

The European Atlas of Natural Radiation explicitly frames the pathway as a network rather than a line, cautioning that "simple regression models often perform badly, because they can hardly capture a high-dimensional space of convoluted covariates."

European pooled case-control studies: lung cancer risk rises ~16% (95% CI: 5–31%) per 100 Bq/m³ of long-term average radon concentration. WHO estimates radon causes 3–14% of all lung cancers in a country, depending on national radon levels and smoking prevalence.

Smokers are estimated to be 25 times more at risk from radon than never-smokers because of multiplicative interaction. Under ICRP 137, 300 Bq/m³ in a home corresponds to approximately 14 mSv/year effective dose.

The complete chain from uranium to lung dose means intervention can target different links: geogenic radon potential mapping (source), barriers and membranes (transport), sub-slab depressurization (entry, 50–99% reduction), and ventilation (dilution). Long-term measurement remains the only reliable way to quantify the actual endpoint concentration.