• What is the typical particle size for HCGO Backwash filters? The frequency of the HCGO backwash filter has increased significantly when the wash oil amount is decreased in the unit. I have checked the data sheet which states that coke particle size>20 microns. Have you encountered any issues with HCGO backwash filters before? If the problem is not related to the filter size, what other factors might be causing it?



  • Michael Clements, FTC, mclements@ftc-houston.com

    Typical particle size will depend many factors such as feed blend, time in coker run, operating temperature, tank management, etc... However, you should expect coke fines < 20 micron and insolubles. It is critical to understand that the insolubles will bind smaller particles together to form large agglomerations that will be capture in the feed filter. This is commonly overlooked and has a major impact to feed filter operation as well as the reactor. The filter itself should be aligned with the catalyst manufacturers recommendation. The most commonly used is 25 micron @ 98.6% as per ASTM F795-88.

    I would not expect an impact to the backwash frequency based on wash oil amount, there should not be a direct correlation there. The freqeuncy is determine by the amount of contamination being sent to the feed filter. If an internal backwash system is being used, these can be problematic as the insoluble plug the media not allowing regeneration. Thus shortening frequency. Additionally, the internal backwash system is used set up to send the backwash upstream to the inlet side. This never allows the contaminates to be removed from the system and will eventually cause an upset to the feed filters. If an external backwash system is operated, it cab eith be the feed itself or the backwash source. The backwash source should be hot enough to dissolve the deformables and the cycle should be long enough to properly clean the media. As long as the backwash source is good and the system is operating properly, this would only leave the feed source. Sizing of the feed filter may play a role as well. The system should be design to operate at a 0.5 GPM/ft.2 or less. If it is not, you can expect increased frequencies based on the contamination level.

    Ultimately, it is best to confirm filter sizing, technology, operation, system and then feed quality. If everything is properly sized and operating correctly, it would indicate a change in feed quality.



  • Marcio Wagner da Silva, Petrobras, marciows@petrobras.com.br

    Unfortunately, issues related to HCGO (Heavy Coker Gas Oil) filters are not uncommon in the refineries. If the HCGO feed is coming cold from a storage tank, it's possible to face high pressure drop due to high viscosity, due to be a chemically unstable feed the the storage tank should rely with a nitrogen blanketing to avoid the olefins oxidation which will raise the pressure drop in the filters due to gums and polymers precipitation.

    Another key point is the operating performance of the main fractionator of the delayed coking unit. Disturbances in the delayed coking unit can cause coke fines dragging to the downstream processes, an alternative is to check if the wash oil flow rate of HCGO fractionating section is adequate, lower flow rates can cause coking deposition and worse quality of the stream.

    Other points to be checked are the backwash flow rate as well as a good characterization of the particles, especially to determine if the main issues are the organic or inorganic particles. If there was a significant reduction in the feed temperature to the HCGO filter system this also can lead to poor performance and the process needs to be checked to get around this issue.