Search

MICHAEL KIMBRELL (Becht Engineering) 

My preference is to reprocess cracked stocks through the Delayed Coker. This keeps the cracked products separate from the straight run products. If the site does not have a Delayed Coker, then reprocessing cracked stocks through the FCC fractionator is the next option. Again, this keeps the cracked materials separate from the straight run products.  

CHRIS CLAESEN (NALCO) 

Cracked stocks can lead to fouling due to polymerization reactions, especially lighter and diolefinic material. The cracked stocks can best be processed in units that can limit polymerization such as hydrotreaters and hydrocrackers. On the crude unit the fouling can partially be controlled by using polymer inhibitors. 

SAM LORDO (Consultant) 

The typical issues I have seen with the injection of cracked stocks into the crude unit is fouling in the preheat train downstream of the desalter and in the crude tower and associated sidestream circuits. Mitigation can range from commercial chemical additives (antifoulants) to relocation of the stream to another unit that is more suited to handle the crack stock, like, fluidized catalytic cracker (FCC), or Delayed Coker. When determining alternate dispositions, the nature of the expected foulant needs to be considered, e.g., inorganic, ammonia salts, organic polymeric materials, etc. 

BILL CATES (Hunt Refining) 

For our facility, we limit the TAN of the crude to control the individual yield streams TAN content. 

Understanding the mechanism of naphthenic acid corrosion, we model the hot circuits looking at the potential TAN of the stream, metallurgy and fluid velocity within the circuit.  Once this has been done, we evaluate the options, if the corrosion potential is high enough. 

The easiest option to be utilized is chemical injection since this can be installed underway.  Changing line size to slow velocity or changing metallurgy typically can only be achieved with a unit outage.  In each case, we will evaluate the risk of waiting until the outage to help make the decision.  Economics around changing line size or metallurgy will need to be evaluated at each facility site as compared to the operational cost of installing and using a chemical injection system. 

Our Crude Unit is several decades old so that very little piping is of enough metallurgy to resist naphthenic acid attack.  For this reason, we limit the TAN of the crude to a level that does not aggressively attack the metal.  We monitor the hot circuits using a variety of inspection techniques to ensure that no abnormal amount of corrosion is occurring.  This imposes limitations at to what crudes can be purchased since we have limited blending capability in the upstream supply system.  

In our Delayed Coker, we only do atmospheric distillation with the atmospheric resid being sent to the coke heaters.  The vapors from the coke drums are returned to the flash zone of the atmospheric distillation tower.   Since naphthenic acids tends to cycle up in the heavier streams, we allow crudes to be purchased and processed that contain a higher TAN.  Since naphthenic acids will be destroyed in the coke heater, the yields off the atmospheric tower tend to have a lower TAN than a vacuum gas oil would have.   

As with the Crude Unit, we model and monitor the hot circuits for TAN content and potential corrosion.  Again, our metallurgy is predominantly carbon steel, so we have installed chemical injection to utilize a filmer to protect the equipment on an as needed basis. 

SAM LORDO (Consultant) 

The answer is complex as it depends not only on the TAN but the naphthenic acid content as well as the sulfur species and content, composition and distribution. Hence, each crude blend needs to be assessed individually for processing on the any crude unit, considering the existing metallurgy, circuit configuration, stream temperature and flow characteristics. 

If you plan to process naphthenic acid crude you should first, consider: 

1. Developing a comprehensive assessment of the existing facilities with respect to processing higher Nap Acid crude.  

2. Determine where and at what concentration the Nap Acids will go when they enter the crude tower.  For example, a “safe” whole crude TAN may not be compatible with downstream facilities if the low TAN blend stock is stripped off in the first tower.   

3. Identify the areas of greatest risk. 

4. Determine a robust monitoring protocol, locations, type and responsibilities for its maintenance. 

DENNIS HAYNES (Nalco Champion)  

Care must be taken in determining an absolute limit for TAN in whole crude blend.  The limit can be impacted by the sulfur level in the crude, naphthenic acid distillation ranges, naphthenic acid type corrosivity as well as downstream temperatures and flow/turbulence relative to the metallurgies in place.  Therefore, there is not an absolute TAN limit that may be applied to all systems for all crude feed blend combinations.  Each would need to be reviewed on a case-by-case basis. 

MICHAEL KIMBRELL (Becht Engineering) 

For a properly designed vacuum heater, it should be possible for the heater to last the entire run between cycle ending turnarounds without having to decoke the heater. This requires the heater to have been designed with the appropriate heat flux and mass flux rates. Velocity steam is typically used to maintain adequate velocity as the oil is heated and begins to vaporize.  

Heater coking is very dependent on process film temperatures and residence time at temperature as well as coke precursor concentration. Heater coking is exponential with temperature and linear with time. Over firing the heater will accelerate heater coking and cause the heater to be decoked early. Modifying an existing unit to be a deep cut unit without modifying the heater to manage the higher duty will result in accelerated heater fouling. A heavier crude slate has a higher coke precursor concentration, so the heater fouling rate will be higher just by changing the crude slate. With a higher heat flux the process film temperature will be higher which will also increase heater fouling.  

I have not heard of anyone performing an on-line spall of a vacuum unit heater. If that were done, the coke from the heater coils would enter the vacuum tower and would exit out the vacuum bottoms stream. These solids would have the opportunity to foul the stripping trays in the vacuum tower or to get stuck in the vacuum bottoms pumps. The reason on-line spalling works for Delayed Cokers is that the coke that is spelled out of the heater passes is routed to the coke drum where coke is collected and then removed from the process before the coke gets to the fractionator.