Deethanizer Fundamentals

1. Purpose and Position in NGL Fractionation

The deethanizer is a distillation column that separates ethane (C₂) and lighter components from the propane-plus (C₃+) NGL stream. It is a core unit in the NGL fractionation train, positioned after cryogenic recovery (turboexpander or cold box) and before the depropanizer.

Feed to the deethanizer is typically the C₂+ bottom product from the demethanizer, or in some plant configurations, the raw NGL liquid directly from the cryogenic recovery section. The light key component is ethane (C₂H₆) and the heavy key component is propane (C₃H₈). The overhead product is an ethane-rich stream that may be sold as petrochemical feedstock or rejected to fuel gas, while the bottom product is the C₃+ NGL that feeds the depropanizer.

NGL Fractionation Train Sequence

The C₂/C₃ separation is one of the easier splits in the NGL train due to the relatively large difference in boiling points between ethane (−128°F) and propane (−44°F), resulting in favorable relative volatility. However, the column operates at elevated pressure and cryogenic overhead temperatures, which require refrigeration for the condenser.

2. Shortcut Column Design Method

The Fenske-Underwood-Gilliland (FUG) shortcut method provides a rapid preliminary estimate of minimum stages, minimum reflux, and actual stage count for a deethanizer column. While rigorous tray-by-tray simulation is required for final design, the FUG method is invaluable for screening studies and initial sizing.

Fenske Equation — Minimum Stages

The Fenske equation calculates the minimum number of theoretical stages at total reflux:

Where:

• N_min = minimum number of theoretical stages
• x_LK,D, x_HK,D = mole fractions of light key and heavy key in distillate
• x_LK,B, x_HK,B = mole fractions of light key and heavy key in bottoms
• α_avg = column average relative volatility of light key to heavy key, taken as the arithmetic average of the values at the two column ends per GPSA Eq 19-4

Relative Volatility

Relative volatility is the ratio of equilibrium K-values for the light key and heavy key:

For the C₂/C₃ separation, α typically ranges from 2.5 to 4.5 depending on pressure and temperature. The average is taken between the top and bottom of the column as an arithmetic mean, α_avg = (α_top + α_bottom) / 2, per GPSA Eq 19-4. GPSA notes square root averaging as an alternative but adopts the arithmetic average as the standard approach. Higher α means easier separation and fewer stages.

Choosing the K-Value Basis

Relative volatility drives the stage count, and it is only as good as the K-values behind it. This matters more on a deethanizer than on any other column in the train, because a deethanizer runs near 400 psig. In the Wilson correlation the factor P_c/P appears in both K-values of the ratio α = K_C2/K_C3, so operating pressure cancels exactly and Wilson's relative volatility is independent of pressure by construction. That approximation is reasonable at low pressure and untenable at deethanizer pressure, where it returns a much higher α and therefore a much lower stage count.

This calculator defaults to a Peng-Robinson equation-of-state flash, which is pressure-aware, and reports both bases side by side with each result so the difference is visible. On the default case the two differ by roughly a factor of 1.7 in α, which is worth ten or more actual trays — with Wilson on the optimistic side. Wilson is retained as a selectable option so a hand check can be reproduced, and carries a warning when used.

The DePriester chart is not offered on this column. Its correlation coefficients for ethane could not be verified against reference data, and ethane is the light key here, so the method was withheld rather than shipped on unverified numbers.

For detailed design, enter the relative volatility directly using the override field. That is the right place for a value from a process simulator, from the GPSA Section 25 convergence-pressure charts, or from plant data, and it bypasses the built-in correlations entirely.

Underwood Equation — Minimum Reflux

The Underwood equations determine the minimum reflux ratio (R_min) at infinite stages. The method involves solving for the root θ from:

Where q is the feed quality (q = 1 for saturated liquid, q = 0 for saturated vapor). Then the minimum reflux is obtained from:

Gilliland Correlation

The Gilliland correlation relates the actual number of stages (N) to the actual reflux ratio (R) using the dimensionless variables X and Y:

The Molokanov approximation of the Gilliland correlation gives:

Typical design practice sets the actual reflux ratio at 1.1–1.3 times R_min. The FUG method provides preliminary estimates; rigorous simulation (e.g., Aspen HYSYS, ProMax) is required for final design, particularly for multicomponent feeds.

3. Operating Pressure Selection

The operating pressure of the deethanizer determines the overhead temperature, which in turn dictates the type and cost of the condenser cooling system. Selecting the optimal pressure involves balancing separation efficiency against capital and operating costs.

Higher pressure raises the overhead temperature, making condensation easier and potentially allowing the use of less expensive cooling. However, higher pressure reduces the relative volatility between ethane and propane, requiring more trays and higher reflux.

Lower pressure improves relative volatility and separation efficiency (fewer trays, lower reflux), but results in colder overhead temperatures that demand propane refrigeration or other cryogenic cooling for the condenser.

Pressure vs. Overhead Temperature

The economic tradeoff is significant: operating at lower pressure requires fewer trays (lower tower cost) but more expensive refrigeration (higher condenser cost and energy consumption). Most deethanizers in NGL plants operate in the 350–450 psig range, with propane refrigeration providing the condenser cooling. The optimal pressure is typically determined through a total-cost optimization study that considers column cost, condenser duty, refrigeration compressor power, and utility costs.

4. Tray Hydraulics and Column Diameter

Column diameter is determined by the vapor handling capacity, which is limited by the flooding velocity. The Souders-Brown correlation is the standard method for estimating the maximum allowable vapor velocity.

Souders-Brown Flooding Velocity

Where:

• U_flood = flooding vapor velocity (ft/s)
• C_SB = Souders-Brown coefficient (ft/s), depends on tray spacing
• ρ_L = liquid density (lb/ft³)
• ρ_V = vapor density (lb/ft³)

C_SB Factor by Tray Spacing

The design vapor velocity is typically set at 75–85% of the flooding velocity to provide adequate operating margin. The required column cross-sectional area is then:

Where V̇ is the volumetric vapor flow rate (ft³/s) at column conditions. The column inside diameter is then D = (4A/π)^0.5, rounded up to the next standard vessel diameter.

Tray Types

• Valve trays — Most common in deethanizer service; good efficiency over a wide operating range with turndown ratios of 3:1 to 5:1
• Sieve trays — Lower cost but limited turndown (2:1); susceptible to weeping at low vapor rates
• Bubble-cap trays — Excellent turndown but highest cost and pressure drop; used where zero liquid leakage is critical

Turndown is an important consideration for deethanizers because feed rates can vary significantly with upstream plant throughput changes. Valve trays are the most common choice for new deethanizer installations, offering the best balance of efficiency, cost, and flexibility.

5. Reboiler and Condenser Design

Condenser

The deethanizer condenser is typically a partial condenser, producing a vapor overhead product (ethane) and a liquid reflux stream. In some configurations, a total condenser is used when liquid ethane product is required (e.g., for pipeline shipment to an ethylene cracker).

The condenser duty is calculated from:

Where V is the total overhead vapor rate (lb/hr) and λ_avg is the average latent heat of condensation (BTU/lb). Because the overhead temperature is typically −30 to +20°F depending on operating pressure, propane refrigeration is the most common cooling medium, with the refrigerant boiling at −40 to −20°F.

Reboiler

The reboiler duty for the deethanizer is approximately equal to the condenser duty plus feed preheat and heat losses (typically 10–15% additional):

The reboiler temperature depends on the bottoms composition and column pressure, typically ranging from 200 to 350°F. This is well within the range of hot oil, steam, or process-to-process heat exchange. The most common reboiler types for deethanizer service are:

• Kettle reboiler — Most common for deethanizer service; provides good separation of vapor from liquid and handles varying heat loads well
• Thermosiphon reboiler — Used when lower residence time is desired or when plot space is limited; requires adequate static head for natural circulation

For detailed reboiler sizing methodology, see the Reboiler Sizing Fundamentals guide.

Typical Duties by Feed Rate

These values are approximate and depend strongly on feed composition, operating pressure, and product purity specifications. Detailed simulation is required for final engineering.

References

1. GPSA, Section 19 — Fractionation & Absorption
2. GPA Standard 2140 — Liquefied Petroleum Gas Specifications
3. Kister, H.Z. — Distillation Design, McGraw-Hill
4. ASME Boiler and Pressure Vessel Code, Section VIII, Division 1