Optimising Ethanol Gasoline Blendstocks Through Addition of Ethanol to Blending Analyser Sample Lines Presented by ICON Scientific at PEFTEC 2015

This Paper, presented at PEFTEC 2015,  provides an overview of the problems associated with controlling and certifying the final product quality of bio-gasolines when ethanol is added to meet biofuels mandates. Ethanol is currently nominally added at 5%, 7%, 10% or 15% depending on country, location or marketing preference. The paper describes a method and equipment to allow the continued optimum use of on-line process analysers to control and where applicable certify the gasoline product at the refinery although the ethanol is added remotely at the marketing terminals. 

The ethanol is added at the marketing terminals to avoid water pick-up during transport to the marketing terminals due to the highly hygroscopic nature of ethanol. The ethanol has an impact (property boost or depression) on a number of the critical gasoline product quality specifications; most importantly octane, vapour pressure and distillation. Therefore refineries need to produce what is commonly known as a BOB (Blendstock for Oxygenates Blending) with product quality properties such that it produces a final on-specification gasoline when the requisite amount of ethanol is added at the marketing terminal. This boost effect is sample matrix specific i.e. depends on the specific component compositions and their proportions used to blend the BOB. The boost also depends on the base property value of the BOB and can be non-linear. This means the BOB's have to be conservative in respect to final product quality property give-away to cope with the worst case property boost or depression to ensure the final gasoline is on specification after the addition of the ethanol. 

For conventional gasoline blending, there are large incentives to use on-line analysers for on-line control and certification of the blend. However the incentives for using on-line analysers for BOB blending are more limited as they can only measure the neat BOB properties. When blending BOB's, the analysers can therefore only control to the conservative BOB property targets not the final gasoline product quality after the ethanol is added. They also can no longer be used to certify the final product with the loss of the significant benefits for sites that use on-line certification for conventional gasoline. 

Icon Scientific has developed and demonstrated a unique patented system that uses precision manufactured piston cylinders to mix a combined BOB/ethanol sample flow to the analysers in the required ratio with a very high precision (better than ± 0.05%). This minimises any effects from the uncertainty of the mixing. The system will eliminate the uncertainty of the boost models for BOB blending control and release and enable the recapture of the full economic benefits of on-line analyser control and analyser on-line certification of the final product. The estimated benefits range is $1M-$15M for analyser control of the blend and $2M-$25M+ for online analyser certification depending on the size of the refinery and the gasoline volumes blended. Other lower precision ethanol mixing systems will not be able to capture all of these benefits. 

The precision of the Icon Scientific mixer system is independent of the flow rate and the flow rate is automatically set by the actual instantaneous demand from the analysers so it only continuously mixes the small final sample flow required by the analysers. This minimises the ethanol use to only exactly the amount required by the individual analyser sample flows which reduces problems with the handling and management of ethanol at the blending site location. 

Possible alternatives to the Icon Scientific piston cylinder approach are either to use coriolis or turbine flow meters and a small flow control valve(s) or metering pumps to continuously control the ethanol addition to the sample. In both cases the fundamental precision will be in the range of ± 0.2% similar to a terminal tanker loading system which is not as good as the fixed volume precision piston cylinder approach. Therefore it is an added uncertainty for the blend target setting increasing the final product quality giveaway. Also these systems will not be able to mix just the small volume flow required by the analysers due to turndown and hysteresis effects on varying demand flows from e.g. cyclic analysers potentially making the precision even worse. This means they will need to mix the constant full analyser fast loop or a bypass loop flow so they will use a significantly greater volume of ethanol.