Butane Splitter Fundamentals

1. Purpose of the Deisobutanizer (Butane Splitter)

The butane splitter, formally known as the deisobutanizer (DIB), is the final distillation column in the NGL fractionation train. It receives the mixed butane overhead product from the debutanizer and separates it into two distinct products: isobutane (iC₄) recovered overhead and normal butane (nC₄) taken as the bottoms product. Although both components are classified as butane, their physical properties and market values differ significantly, making the separation economically justified in many NGL facilities.

Isobutane is the more valuable of the two products. Its primary use is as alkylation unit feed, where it reacts with light olefins (butylene, propylene) in the presence of HF or H₂SO₄ catalyst to produce alkylate, a high-octane, low-sulfur gasoline blending component. Isobutane also serves as a refrigerant (designated R-600a), aerosol propellant, and petrochemical feedstock for propylene oxide and methyl methacrylate production.

Normal butane has a different set of end uses. It is blended directly into gasoline to adjust Reid Vapor Pressure (RVP), particularly during winter months when higher volatility is permissible. Normal butane is also a component of commercial LPG, and it can be sent to an isomerization unit where catalytic conversion transforms it into additional isobutane to supplement alkylation feed supply.

Butane Product Applications

The economic driver for installing a butane splitter depends on the price differential between isobutane and normal butane, which is typically $0.05–0.15/gallon. For large NGL plants processing 50,000+ BPD of mixed butane, even a small price differential can justify the capital investment. In refineries with alkylation units, the internal value of isobutane often exceeds the market price, further strengthening the economic case.

2. The Close-Boiling Separation Challenge

The butane splitter presents one of the most challenging separations in the NGL fractionation train. Isobutane has a normal boiling point of 10.9°F (−11.7°C) while normal butane boils at 31.1°F (−0.5°C), yielding a boiling point difference of only 20.2°F. This narrow spread results in a relative volatility (α) of approximately 1.10–1.30 depending on operating pressure, making the iC₄/nC₄ separation a classic close-boiling system that demands a large number of separation stages and high reflux ratios.

The relative volatility of the iC₄/nC₄ system is defined as:

Because both isomers have very similar molecular weights (58.12 g/mol), molecular structure, and intermolecular forces, their K-values at any given temperature and pressure are very close. The resulting low relative volatility means that each theoretical stage provides only a small incremental enrichment, requiring many stages stacked in series to achieve the desired product purities.

To appreciate how challenging the butane splitter separation is, it is useful to compare it against the other columns in the NGL fractionation train:

Comparison of NGL Column Separations

The impact of low relative volatility on column design is dramatic. Where the depropanizer achieves the desired separation in 30–45 trays with reflux ratios of 2–4, the butane splitter requires 60–120 actual trays with reflux ratios of 5–15. The high reflux ratio means that internal liquid and vapor traffic within the column is many times the net product flow, resulting in large column diameters and massive reboiler and condenser duties. Energy consumption per barrel of product separated is the highest of any column in the NGL train.

Operating pressure has a significant effect on relative volatility. Lower pressures increase α and reduce the required tray count and reflux ratio, but may require refrigeration for overhead condensation. The typical operating pressure range for butane splitters is 60–120 psig, selected to allow condensation of the isobutane overhead with air cooling or cooling water while maintaining a reasonable relative volatility.

3. Column Design — High Tray Count Considerations

The shortcut design methods (Fenske-Underwood-Gilliland) applied to the butane splitter yield very different results from other NGL columns due to the low relative volatility. Each step of the FUG method is heavily impacted by the narrow α value.

Fenske Equation — Minimum Stages

The Fenske equation at total reflux gives the absolute minimum theoretical stages:

For a typical butane splitter producing 96 mol% iC₄ overhead and 95 mol% nC₄ bottoms, with α_avg = 1.20, the Fenske equation yields N_min of approximately 35–50 theoretical stages. Compare this to a depropanizer where N_min is typically only 8–12 stages for similar product purities. The logarithm of a number close to 1.0 is very small, causing the stage requirement to escalate rapidly.

Underwood Equation — Minimum Reflux

The Underwood equations for minimum reflux ratio yield R_min values of 5–12 for the iC₄/nC₄ system. The minimum reflux is determined from:

When α is close to 1.0, the denominator (α_i − θ) becomes very small for the key components, driving R_min to high values. The physical interpretation is that with little thermodynamic driving force for separation, the column must process large quantities of internal reflux liquid to achieve the desired purification.

Gilliland Correlation — Actual Stages

The Gilliland correlation relates actual reflux to actual stages via the Molokanov approximation. For the butane splitter, the operating reflux is typically set at 1.05–1.20 times R_min, which is much closer to minimum reflux than for other NGL columns. The reason is economic: because R_min is already so high, any significant multiplier would create prohibitively large energy costs. Operating near R_min increases the required number of actual stages but reduces the dominant cost driver, which is energy.

Mechanical Design Considerations for Tall Columns

With 60–120 actual trays at 24-inch tray spacing, butane splitter columns can reach heights of 120–250 feet (tangent-to-tangent). These extreme heights introduce mechanical design challenges that do not arise with shorter columns:

• Wind loading: Tall, slender columns are subject to significant wind-induced bending moments and vortex shedding. ASCE 7 wind loads must be evaluated, and vortex shedding analysis per ASME STS-1 may require helical strakes or spoilers on the upper shell sections.
• Seismic design: The natural period of vibration increases with height, potentially placing the column in an unfavorable region of the seismic response spectrum. Base shear and overturning moments per IBC/ASCE 7 govern skirt and anchor bolt design.
• Foundation design: The combination of high dead weight (filled with liquid during hydrotest), wind overturning moment, and seismic base shear requires octagonal or circular spread footings or drilled pier foundations. Soil bearing capacity must be carefully evaluated.
• Shell thickness: The lower shell courses require increased wall thickness to resist the combined hydrostatic head, internal pressure, and bending stresses from wind and seismic loads.

Split-Service (Two Columns in Series)

For very large duties or when a single column would exceed practical height limits (typically 200 feet), the butane splitter can be divided into two columns operating in series. The first column performs a partial separation, and the second column completes the purification. This approach reduces individual column heights, simplifies erection and maintenance, and provides intermediate heat recovery opportunities. The trade-off is additional plot space, piping, and instrumentation.

Typical Butane Splitter Design Parameters

4. Structured Packing as an Alternative to Trays

Given the extremely high tray count required for butane splitter service, structured packing is a compelling alternative to conventional trayed columns. Structured packing (Sulzer Mellapak, Koch-Glitsch Flexipac, Raschig Super-Pak, etc.) consists of corrugated metal sheets arranged in a regular geometry that provides a large surface area for vapor-liquid contact within a compact height.

The key advantage of structured packing in close-boiling separations is its low height equivalent to a theoretical plate (HETP). While conventional trays require 24 inches of spacing per tray and achieve 70–85% Murphree efficiency (resulting in 28–34 inches per theoretical stage), structured packing achieves HETP values of 12–18 inches for the iC₄/nC₄ system. This height reduction of 30–50% per theoretical stage translates directly to shorter columns, reduced shell weight, lower foundation costs, and less prominent visual impact.

Trays vs. Structured Packing Comparison

The lower pressure drop per theoretical stage offered by structured packing is particularly beneficial for butane splitters. With 60–120 stages, cumulative pressure drop across trayed columns can reach 6–18 psi, which creates a significant temperature gradient from top to bottom. This temperature gradient reduces the relative volatility on the lower trays (higher pressure = lower α), partially negating the benefit of additional stages. Structured packing, with its much lower pressure drop (1–6 psi total for the same stage count), maintains a more uniform relative volatility throughout the column.

Liquid Distribution and Redistribution

The primary design challenge with structured packing in tall columns is liquid maldistribution. Unlike trays, which inherently mix and redistribute liquid at every stage, packed columns rely on gravity-driven film flow down the corrugated sheets. Over extended packing heights, liquid tends to migrate toward the column wall (wall flow), reducing mass transfer efficiency. Industry practice requires liquid redistributors every 15–20 feet of packing height, or every 6–8 HETP. For a butane splitter with 80 theoretical stages, this means 6–10 redistributor levels, each adding cost and height.

FRI (Fractionation Research Inc.) data confirm that properly designed structured packing installations with adequate redistribution can match or exceed tray performance for the iC₄/nC₄ system. However, poor initial liquid distribution or inadequate redistribution can reduce effective stage count by 20–40%, potentially requiring column height increases that negate the original packing advantage.

5. Energy Optimization and Heat Integration

The butane splitter is the most energy-intensive column per barrel of product in the NGL fractionation train. The combination of high reflux ratio (5–15) and large internal traffic means that reboiler and condenser duties are 2–5 times greater per BPD of feed than the depropanizer or debutanizer. For a 10,000 BPD butane splitter, reboiler duties of 30–80 MMBTU/hr are common, representing a substantial portion of the total plant energy budget. This makes energy optimization critical to economic viability.

Optimal Reflux Ratio Selection

The relationship between reflux ratio and total annualized cost is particularly steep for the butane splitter. Operating at R/R_min = 1.05–1.20 is standard practice, compared to 1.10–1.30 for other NGL columns. The economic optimum lies very close to R_min because energy cost dominates the total cost. Each 1% increase in reflux ratio above R_min increases energy consumption by approximately 1% but reduces the required stage count by a progressively smaller amount due to the asymptotic nature of the Gilliland correlation in this region.

Energy Optimization Techniques

Heat Pump (Vapor Recompression) Distillation

Vapor recompression is the most impactful energy-saving technology for butane splitters and is widely applied in commercial practice. The overhead isobutane vapor is compressed (typically by 15–30 psi) to raise its condensation temperature above the reboiler temperature. The compressed vapor then condenses in the reboiler shell, providing the heat of vaporization. The condensed liquid is flashed back to column pressure and returned as reflux.

The close boiling points of iC₄ and nC₄ mean that the temperature difference between the column overhead and bottoms is only 30–60°F, making the compression ratio and compressor power modest relative to the energy recovered. The coefficient of performance (COP) for heat pump butane splitters typically ranges from 5 to 10, meaning that each BTU of compressor work delivers 5–10 BTU of reboiler heat. This compares to a COP of 1.0 for conventional steam-heated reboilers (where 1 BTU of fuel produces approximately 1 BTU of useful heat after boiler and distribution losses).

Double-Effect Distillation

In a double-effect arrangement, the separation is performed in two parallel columns operating at different pressures. The higher-pressure column has a higher overhead condensation temperature, which is used to supply the reboiler duty of the lower-pressure column. The net effect is that only one column requires external reboiler energy (or a significantly smaller total external duty). This approach works well for the butane splitter because the moderate operating pressures allow a practical pressure differential without exceeding material or equipment limitations.

Feed Preheating

The simplest energy recovery measure is a feed/bottoms heat exchanger. The hot nC₄ bottoms product (140–200°F) preheats the incoming mixed butane feed, reducing the sensible heat load on the reboiler. With a 20°F minimum approach temperature, feed preheat typically recovers 5–15% of the reboiler duty. This measure has a rapid payback (typically less than one year) and should be included in all butane splitter designs regardless of whether more advanced heat integration is also employed.

Detailed process simulation with rigorous thermodynamic models (Peng-Robinson or SRK equation of state) is essential for final design of any butane splitter, particularly when heat integration or vapor recompression is employed. The sensitivity of stage count and energy duty to small changes in relative volatility, feed composition, and product specifications demands careful optimization that cannot be adequately addressed by shortcut methods alone.

References

1. GPSA, Chapter 16 — Hydrocarbon Recovery
2. GPA Standard 2140 — Liquefied Petroleum Gas Specifications
3. FRI (Fractionation Research Inc.) — Design Practice Reports for Structured Packing and Tray Performance
4. Kister, H.Z. — Distillation Design, McGraw-Hill
5. Humphrey, J.L. and Keller, G.E. — Separation Process Technology, McGraw-Hill
6. ASME Boiler and Pressure Vessel Code, Section VIII, Division 1