Feb-2020

High octane isomerates from light naphtha

Dividing wall column distillation is an energy efficient, cost effective alternative to conventional distillation for delivering high octane products from light naphtha.

DAVID KOCKLER
Consultant

Article Summary

Dividing wall columns have gained increasing acceptance over the past 30 years as an energy efficient means to separate ternary mixtures into three products by distillation. Dividing wall columns are the functional equivalent of thermally integrated two column prefractionator/main column arrangements, which are also known as Petlyuk columns. They are divided into two sections by an internal dividing wall. A feed consisting of a mixture of A/B/C (where A is the lightest column product and C is the heaviest column product) enters the dividing wall column on the prefractionation side of the dividing wall. In the prefractionation zone, the A/B/C feed is separated into two cuts. The lighter cut, consisting of A and B with only trace amounts of C, travels into the main column section above the dividing wall, and the heavier cut, consisting of B and C with only trace amounts of A, travels into the main column section below the dividing wall. In the main column section of the dividing wall column, the A/B and B/C cuts are separated into A, B, and C products, which are withdrawn as overhead, side draw, and bottoms products, respectively.

The amount of prefractionation that takes place in a dividing wall column is determined by the liquid and vapour traffic that takes place on the feed (prefractionation) side of the dividing wall. The vapour split that takes place between the fractionation zones on either side of the dividing wall is determined by the column hydraulics. This means that the placement of the dividing wall in the column effectively determines the vapour split and the amount of prefractionation that takes place in a dividing wall  column.

Since the dividing wall cannot be moved once a dividing wall column is installed, the operator cannot adjust the vapour split and the amount of prefractionation that takes place in the column. This can be a drawback in applications with large composition swings in the feed to the column, since a suboptimal amount of prefractionation may reduce energy efficiency.

Dividing wall columns which process isomerisation reactor effluent as the column feed have a unique advantage over other applications of the technology. The composition of the isomerisation reactor effluent is very consistent, even when the composition of the fresh feed to the isomerisation unit changes. The composition is very consistent because the reactions that take place in the isomerisation reactor approach an equilibrium between high octane molecules and low octane molecules. Using chlorinated alumina catalyst, for example, the ratio of isopentane to total C5 paraffins at the outlet of the isomerisation reactor will be very close to 0.76 and the ratio of 2,2 dimethyl butane to total C6 paraffins will be very close to 0.34. Different reaction equilibria exist at the outlet of isomerisation reactors for other combinations of compounds. As a result of the reaction equilibria that is maintained at the outlet of the isomerisation reactors, reactor effluent streams have very consistent compositions.

Significant energy savings and reduction in plot space and capex can be obtained by using a dividing wall column in place of a two column sequential distillation configuration. Reboiler energy savings of approximately 30% can typically be obtained in applications which are well suited to dividing wall columns. Capital cost savings can also be obtained by using dividing wall columns in place of two column designs. The cost of internals is somewhat higher than the comparable cost for conventional distillation column internals, but in general the total installed cost will be less than the total installed cost for a two column design.

Light naphtha isomerisation
Light naphtha isomerisation has been used extensively in refineries to produce high octane isomerate products which meet current gasoline specifications. When reformulated gasoline specifications with lower allowed benzene concentrations were implemented in the 1990s, many refiners elected to use isomerisation units to reduce the benzene content in the refinery gasoline pool. Naphtha splitters were revamped so that the majority of benzene and benzene precursors (cyclohexane and methylcyclopentane) ended up in the light naphtha fraction instead of the heavy naphtha fraction. The light naphtha fraction produced in naphtha splitters is sent to an isomerisation unit where benzene is saturated to form cyclohexane. The practice of sending benzene and benzene precursors to an isomerisation unit rather than to a catalytic reformer has significantly reduced the amount of benzene in the gasoline pool without the use of benzene extraction units.

The most important parameters in light naphtha isomerisation are the liquid product yields and the isomerate product research octane number (RON). The liquid product yield is determined principally by the extent of hydrocracking that takes place in the isomerisation reactors. Hydrocracking is an undesirable side reaction which converts light naphtha into light hydrocarbon gas molecules instead of more valuable gasoline blending components. Larger molecules (C7+) are more prone to hydrocracking than C5 or C6 molecules, so it is preferable to limit the amount of C7 in the light naphtha feed as much as possible. However, C7 molecules are generally present in low concentrations in light naphtha feeds as a result of including benzene and benzene precursors in the light naphtha cut.

Isomerisation increases the octane value of light naphtha by increasing the degree of branching of paraffin molecules. Normal pentane (nC5), which has a RON of 61.7, is isomerised to isopentane (iC5) with a RON of 93.5. Normal hexane (nC6), which has a RON of 31, is isomerised to 2-methylpentane (2MP) with a RON of 74.4 and 3-methylpentane (3MP) with a RON of 75.5. The 2MP and 3MP molecules are subsequently isomerised to 2,2 dimethylbutane (22DMB) with a RON of 94 and 2,3 dimethylbutane (23DMB) with a RON of 105. Isomerisation reactions are equilibrium limited. When unbranched paraffins are exposed to isomerisation catalysts in the presence of hydrogen, a fraction of the unbranched paraffins is converted to branched paraffins, but 100% conversion of low octane molecules is not achievable.

The RON target for the combined isomerate product in large part determines the specific energy consumption for an isomerisation process. As a RON target is increased, the reboiler energy required for the distillation columns downstream of the isomerisation reactors increases for two reasons. Higher isomerate product RON is achieved in isomerisation units by recycling low octane molecules from the product fractionation section back to the isomerisation reactors. A high RON target requires a sharp separation between low octane molecules and high octane molecules, which requires increased column reflux and a larger reboiler duty. Isomerate product RON can also be increased by increasing the rate of recycle of low octane components to convert a higher percentage of low octane molecules to higher octane molecules. Higher rates of recycle in turn increase the feed rate to the reactors and to the downstream distillation columns, which results in higher column reboiler duties.

The RON target for an isomerate product may also affect the Reid vapour pressure (RVP) of the refinery gasoline pool. RVP is an important property for gasoline blending components because of seasonal environmental limits that are imposed on gasoline fuel products. A strong correlation exists between the RVP and the RON of isomerate products. Highly branched isomers have higher vapour pressures than single branched paraffins, and single branched paraffins have higher vapour pressures than unbranched paraffins. Consequently, increasing the RON of an isomerate product results in a proportionately higher isomerate product RVP.

Isomerisation unit configurations
Various isomerisation unit configurations have been developed to achieve a wide range of isomerate product RON. The most basic isomerisation units are known as ‘once-through’ units. Fresh feed is introduced to a feed pretreatment section and then passed into a series of isomerisation reactors where the vaporised feed is contacted with isomerisation catalyst in the presence of hydrogen. The reactor effluent is sent to a stabiliser column where hydrogen and light hydrocarbon byproducts from hydrocracking reactions are removed from the top of the stabiliser. The stabilised reactor effluent is the isomerate product in a once-through process. Once-through isomerisation processes using currently available catalysts are generally capable of achieving an isomerate product RON of up to 85.