Cemeg Action

A comparison of the CDU heavy diesel fraction with normal results showed a significant reduction in the width coincident with more paraffinic feeds.  Moderation of the move on cut points brought dosages back to acceptable levels.

Key Work:  Routine characterisation of refinery streams provided the basis for understanding the problem.

The bidding system for purchasing additives was to test for the best responding additive in a set of diesel samples at the start of winter blending. The refinery had several modes of operation following market demands, which significantly changed the diesel character demanding an additive effective in a wide range of blends, rather than a high response in a specific blend. 

Cemeg Project

The purchasing process was modified to select a product effective across all modes and grades.

Exchange partners were finding CFPP results several degrees above the -12 deg C specification and claiming compensation.  

Cemeg Investigation

Batch release results were all on grade at the time of release.  However, retained samples from a cool sample store were giving erratic results on retest.

Study of CFPP suction times showed uneven trends in the suction times with reducing temperature,  so called ‘hesitations’ or ‘aspiration time effects’.

The additive in use had been optimised to give maximum response in narrow and paraffinic  diesels and also not interfere with cold soak filtrations with 1.5micron GFA filters.  Similar to Case History 2, the additive was too soluble, lacking crystallinity and had to be changed for a more reliable product.

A refinery dosing diesel with Wax Anti-Settling Additive (WASA) to make wax float rather than sink when formed on cooling.  

Cemeg Recommendation

The refinery was importing widely different feeds and operating major cracking units, with complex and variable blend ratios for components of differing character.  The variability was found to be too high for consistent control of the densities of the wax and liquid diesel and separation inevitable in certain blends.  The refinery was advised to focus on the correct filter aid performance (EVA polymer) to achieve a good wax crystal form and diesel operability from a porous filter cake, rather than optimise additives and blends to control the wax density to be the same as the liquid diesel.

A vessel arriving at the destination port was retested for quality and failed CFPP with a result just below the cloud point, as if there was no additive in the fuel.

Cemeg Investigation

The shipment was quickly analysed and found to be broad and with a high final boiling point.  Going back to the refinery of origin, the batch had passed for CFPP but apparently lost 10 deg C of response in transport over several days. Retesting in different laboratories gave widely different CFPP values, some on grade but also well-out of specification.

Checks on the additive used showed it to be effective in current batches but apparently had been selected primarily because of a low additive pour point to help with on-site storage.  Further laboratory testing in blends of increasing end point and width indicated the additive was lacking crystallinity and too soluble to be reliable in high boiling blends.

As the winter season progressed, the FBT results started to increase above 1.1.

Cemeg Investigation

Laboratory hand blends were found to give both a good CFPP response and minimal increase in 1.5 micron GFA filtrations, even at high additive dosages.  A check on the blending system found the additive mixing temperature had fallen to a borderline level.  Increasing the diesel temperature for mixing returned the FBT results to acceptable low levels. Important monitoring and early correction, as undissolved crystals of EVA can block filters in the market.

A small number of batches stuck at CFPP values -18 to -20,  just out of specification and increasing dosage made results worse.

   Weakness of the CFPP test.

In the CFPP test, the sample soaking times between suctions increase as the sample cools down to -20. The cooling block then changes from -34 to -50 deg C and the soaking times between suctions is greatly reduced.  This creates a pressure point for fails in borderline samples at -18 to -20. 

The CFPP test is very useful down to 12 degrees below cloud in proving the EVA additive is present and modifying the wax crystallisation.  Going down further in temperature, suction times through the 45 micron mesh of the instrument are following the viscosity of the cooling mixture of modified crystals and viscous liquid diesel.  Low results are welcome, but important not to over optimise the additives to gain a low viscosity and ‘pretty’ CFPP result, at the expense of crystal form and a porous filter cake. The nature of the filter cake is an important factor in a diesel fuel filter cartridge on vehicle cold start.  The CFPP test results at greater than 12 deg C delta below cloud are thought by many chemists to over-estimate real vehicle operability.

Cemeg Support

If the diesel is too waxy, the longer soaking time can badly influence the results. Cemeg provide assistance to identify waxy feeds and adjust blends to dilute the wax.  Crude ranking from assay data and monitoring CDU gas oil n-paraffin contents by correlation. 

A refinery blending near to the minimum Cetane Index and boosting the quality with additive to achieve the cetane number as measured by IQT.

   Cemeg Consultancy

The benefit in cetane from 2-ethyl hexyl nitrate is generally reliable and predictable across blends.  In this case the Cetane Index calculation (CI) was found in certain blends to be overpredicting for cetane in not agreeing with cetane numbers for the base fuel as measured by IQT (CN).  Corrections for two of the blend components were established to help correct the CI calculation to reflect the CN. An on-line analyser was later installed to measure CN directly in the dosed blend header and help adjust dosages.

A refinery using the mono-acid chemistry to achieve the HFRR wear scar specification.

   Cemeg Support

The refinery was observed to be using 250 plus ppm when other plants on apparently similar blends were running below 200ppm.  Comparison of a number of split samples run on the refinery and other plants HFRR instruments indicated a small but significant delta with the refinery measuring higher on average.  Round robin studies had all been considered acceptable and for the samples circulated, near to the average. 

However, studying a number of samples, the statistics highlighted the dosage impact of being in mid-quartile 2 below, rather than 3 above the average. In certain blends, the response curve can be quite shallow above 200ppm and combined with the poor repeatability of the HFRR test, the higher dosages were needed to achieve acceptable first time pass rates.