Hot geothermal brine never boils the working fluid directly — it gives up heat across a shell-and-tube evaporator to a low-boiling-point organic fluid (isobutane here), which drives its own closed Rankine cycle. That separation is why it's called a binary cycle: two fluid loops, one machine.
1. Evaporator (real effectiveness-NTU heat exchange). Brine is the heat-capacity-rate minimum stream (the boiling working fluid absorbs latent heat at near-constant temperature, so its effective heat capacity is enormous). Heat actually transferred is capped below the thermodynamic maximum by the exchanger's effectiveness ε:
Q̇ = ε · (ṁ_brine·cp_brine) · (T_brine,in − T_cond)
2. Turbine (real isentropic expansion, derated). Saturated isobutane vapor leaving the evaporator at T_evap expands to the condenser pressure. Saturation pressures come from the Antoine equation for isobutane; the ideal (isentropic) temperature drop is derated by the turbine's isentropic efficiency η_t, exactly like a real turbine:
P_sat(T) = 10^(4.328 − 1132.1/(T[K] − 0.918)) bar
T2s = T1·(P_low/P_high)^((γ−1)/γ), T2 = T1 − η_t·(T1 − T2s)
w_turbine = cp,vapor·(T1 − T2)
3. Condenser. The expanded vapor rejects its remaining heat to the ambient-cooled loop and fully condenses back to saturated liquid at T_cond = T_ambient + cooling approach.
4. Pump (real feed-pump work). The saturated liquid is pumped back up to evaporator pressure; because liquids are nearly incompressible, this costs only a small fraction of what the turbine produces:
w_pump = v_liquid·(P_high − P_low) / η_pump
The working-fluid mass flow is sized so the evaporator's actual duty Q̇ is fully absorbed as sensible preheat plus latent heat of vaporization (a Watson-correlation estimate that correctly shrinks toward zero as T_evap approaches isobutane's critical point, 134.7 °C). Net power is turbine work minus pump work, times that mass flow. The displayed cycle efficiency η = w_net/q_in is checked live against the Carnot ceiling η_Carnot = 1 − T_cond/T_evap (absolute temperatures) — a real Rankine cycle with finite-effectiveness heat transfer and irreversible turbomachinery always sits measurably below that ceiling, never above it.
- Brine inlet temperature — hotter resource water raises T_evap and widens the Carnot window, but is clamped so isobutane stays comfortably subcritical.
- Brine mass flow — more geothermal flow raises the heat-transfer-rate ceiling (ṁ·cp), scaling total power roughly linearly without changing efficiency much.
- Heat exchanger effectiveness ε — how close the evaporator gets to the thermodynamic maximum heat transfer; real ORC exchangers run 0.7–0.9.
- Turbine isentropic efficiency — how close the real expansion gets to the ideal (isentropic) one; the gap is lost work that shows up as extra reject heat, not extra power.
- Ambient / cooling temperature — sets the condenser floor T_cond; a cooler sink widens the Carnot window from both ends, which is why ORC plants prefer cold climates or wet cooling towers.
Real-world relevance: this is the actual architecture of moderate-temperature binary geothermal plants (roughly 100–180 °C resources, common across most of the world's geothermal fields) that can't flash steam directly, and it's the same cycle used in ORC waste-heat-recovery units on ships, cement kilns and biomass plants.