A geothermal downhole heat exchanger pipes fluid past hot rock; suspending a small volume fraction of high-conductivity nanoparticles in that fluid ("nanofluid") raises its bulk thermal conductivity and its heat-transfer coefficient, so the same pipe pulls more heat per second.
Effective conductivity — Maxwell model:
k_nf / k_bf = [k_p + 2k_bf + 2φ(k_p − k_bf)] / [k_p + 2k_bf − φ(k_p − k_bf)]
Mixture density, heat capacity & viscosity:
ρ_nf = (1−φ)ρ_bf + φρ_p
cp_nf = [(1−φ)ρ_bf·cp_bf + φρ_p·cp_p] / ρ_nf
μ_nf = μ_bf · (1 + 2.5φ + 6.2φ²) (Brinkman-type correction)
Convective transfer — Dittus-Boelter correlation:
Re = ρ_nf·v·D / μ_nf, Pr = μ_nf·cp_nf / k_nf
Nu = 0.023 · Re^0.8 · Pr^0.4, h = Nu·k_nf / D
Heat exchanger effectiveness (NTU method) along a pipe of length L, area A = πDL:
T_out = T_rock − (T_rock − T_in)·exp(−hA / (ṁ·cp_nf))
Q = ṁ·cp_nf·(T_out − T_in)
Brownian diffusivity of the suspended particles (Stokes-Einstein), which sets how fast the microscopic nanoparticle jitter you see below happens:
D_B = k_B·T / (3π·μ_nf·d_p)
- Particle type — Al₂O₃, CuO and Cu have very different bulk conductivities k_p, so the same loading gives very different enhancement.
- Volume fraction φ — more nanoparticles raise k_nf and h, but also raise viscosity, which is why real nanofluids rarely exceed ~5-6% before pumping cost outweighs the gain.
- Particle diameter — only enters here through D_B: smaller particles diffuse faster and jitter more visibly, matching real Brownian-motion measurements.
- Flow velocity & rock temperature — set the convective driving conditions the correlation above runs on.
Real-world relevance: this is the same effective-medium + Dittus-Boelter + NTU pipeline used in nanofluid geothermal-drilling literature to justify 15-40% gains in downhole heat extraction from enhanced coaxial and U-tube exchangers.