Liquid Fraction Calculator from Mole Balances
Estimate vapor fraction and liquid fraction in an isothermal flash using the Rachford-Rice mole balance relation and component K-values.
Flash Input Data
Component 1
Component 2
Component 3
If z-values do not sum to 1.0, the calculator automatically normalizes them before solving.
Calculation Results
How to Calculate Liquid Fraction from Mole Balances: An Expert Engineering Guide
Calculating liquid fraction from mole balances is one of the most common tasks in vapor-liquid equilibrium, distillation design, and flash drum analysis. In practical process engineering, you often know feed composition, temperature and pressure, and phase equilibrium constants. What you need is a fast, physically correct way to determine how much of the feed becomes vapor and how much remains liquid. That split controls separator size, compressor duty, condenser load, and downstream product quality.
At the center of this problem is a simple but powerful concept: total and component mole balances must both be satisfied at the same time. The phase equilibrium relation links them. When these equations are solved together, you obtain vapor fraction and liquid fraction. This page focuses on liquid fraction, usually written as L/F, where L is liquid moles and F is feed moles. Vapor fraction is V/F, and because total moles are conserved for a nonreactive flash, L/F = 1 – V/F.
1) Problem Setup for a Flash Calculation
Consider a feed stream entering a flash vessel. The feed has total moles F and component feed mole fractions zi. At equilibrium, two outlet phases may exist:
- Vapor phase moles: V with composition yi
- Liquid phase moles: L with composition xi
Total balance: F = V + L
Component balance: F zi = V yi + L xi
Equilibrium relation: yi = Ki xi
The K-value expresses phase preference. If Ki > 1, component i tends to favor vapor. If Ki < 1, it tends to stay in liquid. These constants depend strongly on temperature and pressure.
2) Rachford-Rice Equation and Liquid Fraction
Defining vapor fraction as β = V/F, we can rearrange the mole balances into the classic Rachford-Rice equation:
Σ zi(Ki – 1) / (1 + β(Ki – 1)) = 0
Solve this equation for β in the range 0 to 1 for a two-phase flash. Once β is known:
- V = βF
- L = (1 – β)F
- Liquid fraction = L/F = 1 – β
- xi = zi / (1 + β(Ki – 1))
- yi = Kixi
This is exactly what the calculator above does numerically using a robust bisection method. Bisection is slower than Newton methods per iteration, but it is stable and reliable for engineering calculators where invalid user input is common.
3) Why Mole Balance Based Liquid Fraction Matters in Industry
Liquid fraction prediction directly affects equipment and process economics:
- Separator sizing: residence time and disengagement area depend on liquid and vapor flow rates.
- Energy integration: compressor and condenser loads are tied to vapor amount.
- Control strategy: level and pressure loops rely on expected phase split behavior.
- Safety analysis: overprediction of liquid fraction can underestimate vapor handling requirements.
In refinery and gas processing services, even a small error in flash split can shift downstream tray loading, absorber efficiency, and product recoveries. For this reason, engineers treat phase split calculations as foundational design work, not a secondary check.
4) Data Quality: K-values and Feed Composition
A correct method still gives bad results if input data are poor. K-values are the biggest source of uncertainty when quick screening calculations are performed. Good practice includes:
- Use K-values from a temperature and pressure pair as close as possible to the actual flash condition.
- Verify feed composition closure. If z-values do not sum to 1.0, normalize before solving.
- For wide-boiling mixtures, check whether a simple K-value model is still valid.
- For highly nonideal systems, use activity-coefficient or equation-of-state methods.
5) Reference Property Statistics for Common Light Hydrocarbons
The physical volatility trend below explains why K-values for lighter components are usually larger at a fixed pressure and moderate temperature. These values are widely reported in thermodynamic references such as the NIST Chemistry WebBook.
| Component | Normal Boiling Point (K) | Critical Temperature (K) | Volatility Tendency in Typical Flash Service |
|---|---|---|---|
| Methane | 111.7 | 190.6 | Very high vapor preference |
| Ethane | 184.5 | 305.3 | High vapor preference |
| Propane | 231.0 | 369.8 | Moderate to high vapor preference |
| n-Butane | 272.7 | 425.1 | Moderate vapor preference |
Values shown are standard reference values commonly published in thermodynamic databases and are provided for trend comparison in mole balance interpretation.
6) Worked Computational Scenarios Using Mole Balance Logic
To make liquid fraction behavior intuitive, the table below compares different feed and K-value environments. All outputs are obtained by solving the Rachford-Rice equation and converting to liquid fraction by L/F = 1 – V/F.
| Scenario | Feed z-vector | K-vector | Computed V/F | Computed L/F |
|---|---|---|---|---|
| A: Mixed volatility | [0.50, 0.30, 0.20] | [1.8, 0.9, 0.4] | 0.429 | 0.571 |
| B: More volatile feed behavior | [0.55, 0.25, 0.20] | [2.2, 1.1, 0.5] | 0.604 | 0.396 |
| C: Less volatile condition | [0.45, 0.35, 0.20] | [1.3, 0.8, 0.35] | 0.256 | 0.744 |
Notice the pattern: as weighted volatility increases, vapor fraction increases and liquid fraction decreases. This is consistent with equilibrium theory and observed plant behavior.
7) Step by Step Manual Method
- Gather feed flow F, feed composition zi, and K-values at flash conditions.
- Confirm all z-values are nonnegative and normalize if needed.
- Write Rachford-Rice equation in terms of β.
- Check phase feasibility:
- If function value at β = 0 is less than 0, mixture tends to all liquid.
- If function value at β = 1 is greater than 0, mixture tends to all vapor.
- If signs differ across 0 and 1, a two-phase root exists.
- Solve for β numerically.
- Compute liquid fraction L/F = 1 – β.
- Back-calculate x and y compositions and verify closure:
- Σxi should be approximately 1
- Σyi should be approximately 1
8) Common Mistakes and How to Avoid Them
- Using inconsistent K-values: if K-values come from a different pressure than your flash drum, the split can be very wrong.
- Ignoring normalization: composition vectors that sum to 0.97 or 1.03 create hidden mass-balance bias.
- Forgetting single-phase outcomes: not every case has a root between 0 and 1.
- Rounding too early: keep full precision during iteration, then round only final reported values.
- Treating K-values as constants over large condition changes: in real design, recalculate K-values when T and P change.
9) Validation and Benchmarking
For engineering confidence, compare your calculator result with a process simulator on a few benchmark cases. When EOS models are used in simulation and constant K-values are used in a hand tool, some difference is expected. Still, trends should match:
- Increase pressure at constant temperature for many hydrocarbon systems and liquid fraction usually rises.
- Increase temperature at constant pressure and vapor fraction usually rises.
- Increase light-end concentration and vapor fraction typically rises.
If your result violates all three trends, inspect units, compositions, and K-value sources before trusting the output.
10) Authoritative Learning and Data Sources
For rigorous thermodynamic background and property validation, consult these authoritative references:
- NIST Chemistry WebBook (.gov) for thermophysical property data and volatility context.
- U.S. Department of Energy (.gov) for process and energy systems references where phase behavior impacts design and efficiency.
- MIT OpenCourseWare (.edu) for chemical engineering thermodynamics and phase equilibrium coursework.
Final Takeaway
The most reliable way to calculate liquid fraction from mole balances is to combine component balances, total balance, and equilibrium relations into the Rachford-Rice framework. Solve for vapor fraction first, then convert directly to liquid fraction. When inputs are consistent and physically meaningful, this method is fast, robust, and highly transferable across refinery, petrochemical, and gas processing applications.