CO2 has a critical point at Tc = 31.1 °C, Pc = 7.38 MPa. Just above that point the fluid is a single supercritical phase, but its density is nowhere near an ideal gas — it behaves almost liquid-like, which is the entire reason this cycle exists.
Density is computed from the Peng–Robinson equation of state, solved for the compressibility factor Z:
P = RT/(V−b) − aα(T)/(V²+2bV−b²)
Z³ − (1−B)Z² + (A−3B²−2B)Z − (AB−B²−B³) = 0
ρ = M·P / (Z·R·T)
Near the critical point Z drops as low as ~0.3–0.5 (vs. Z≈1 for an ideal gas), so the same mass of CO2 occupies far less volume. Since compressor flow work scales with ∫V·dP, a lower Z means dramatically less compression work for the same pressure rise — the compressor-inlet temperature slider shows this directly: move it toward 32 °C (just above Tc) and the density readout jumps while compression work (and hence the whole cycle's efficiency) improves.
The rest of the loop uses a standard recuperated-Brayton model with isentropic turbomachinery efficiencies (turbine 90%, compressor 85%) and constant cp ≈ 1.2 kJ/kg·K, γ ≈ 1.28 for supercritical CO2 in the hot leg:
T2s = T1·PR^((γ−1)/γ) T2 = T1 + (T2s−T1)/η_c
T4s = T3·PR^−((γ−1)/γ) T4 = T3 − η_t·(T3−T4s)
Recuperator: T2' = T2 + ε(T4−T2), T4' = T4 − ε(T4−T2)
w_turbine = cp(T3−T4) w_compressor = Z₁·cp(T2−T1)
Q_in = cp(T3−T2') η = (w_turbine − w_compressor) / Q_in
This is a simplified, idealized cycle model (no pressure losses, constant properties outside the compressor-inlet density correction) — real sCO2 Brayton plants add a recompression split and precise NIST-grade property tables, but the mechanism you're controlling here — squeezing compression work by staying near the critical point — is the actual reason utilities and next-gen nuclear/concentrated-solar plants are building sCO2 turbines instead of steam Rankine cycles.