James D. Weith
Practical Enginuity, Mission Viejo, Calif.
Critical path diagrams, which have evolved from logic diagrams, can help refiners optimize the timing of FCC start-ups. Those start-ups require that certain steps occur simultaneously before feed goes to the riser.
This required timing or coordination is why establishing a critical path is valuable; it helps remove some of the inefficiencies from the individual steps for each section, so that the ultimate timing between them can be better synchronized.
Logic diagrams illustrate the necessary sequences for starting an FCC. These logic diagrams then become the basis for the critical path diagram.
While critical path diagrams are valid for any refining process, a typical FCC unit will demonstrate the method; FCCs offer a good range of possibilities for examining the capabilities and merit of critical pathing due to the complexity of their start-ups.
Traditional FCC start-up
FCC start-ups have traditionally required a hot reactor section and a main column that is free of air, hot, and dried out before the blind between the two can be pulled.
To a certain degree, the refiner can modify or rearrange the sequences that lead up to this point, provided the end result in each section occurs at almost the same time.
A block valve that replaces the blind in the reactor vapor line is becoming more common; therefore, one can treat the start-up as three distinct events that occur simultaneously. The reactor and regenerator are considered separately instead of as a system.
With FCCs that use neither a blind nor a block valve, the boundary between the systems must change. In this case, the reactor is included with the main column.
Start-up with blind
Logic diagrams simplify the process of visualizing the requirements for starting up FCC units in each scenario.
null
Fig. 1 shows a logic diagram for a traditional start-up with a blind in the reactor vapor line. The problem with this type of start-up is that the refiner has to stop and restart the procedure to remove the blind. The refiner must stop catalyst circulation, drain oil from the main column, and remove any gas using steam.
After the blind is removed, catalyst circulation resumes, oil level in the main column is reestablished, and the column is reheated to dry it out. Depending on how long it takes to remove the blind, resuming the start-up can last almost as long as starting the procedure from the beginning.
It is also dangerous and hazardous to remove the blind in a hot, steamy environment.
Start-up with block valve
null
Fig. 2 shows a logic diagram that illustrates a much simpler start-up sequence that uses a block valve in the reactor vapor line.
Replacing the blind with a block valve eliminates all the hazards and time lost for blind removal. Block valves, however, create another set of problems.
They create the potential for an obstruction in the line from the reactor to the main column; this requires pressure safety valves on the reactor side of the block valve. These, however, are a mechanical nuisance and discharge vapors from the relief valves are 920° F. or hotter. Hot vapors may require an additional quench system for cooling, or a complete redesign of the relief header to accommodate any thermal expansion.
Even though a block valve simplifies the start-up considerably, both methods still require a vent on the reactor vapor line to release steam during the reactor-purging phase or for circulating catalyst.
The vent line also creates hazards. More than one refinery has released black oil when the main column level got too high. At least one facility has repiped the vent to an open pit at grade to prevent this from occurring.
A block valve does not completely eliminate the need for a blind. It is still required as an equipment isolation blind whenever personnel must enter the reactor or main column.
The reactor vapor line block valve is only useful immediately after a turnaround. Once an FCC unit goes on stream, coke will accumulate in the valve guides and seat. It is nearly impossible to close the valve until it is cleaned out.
A block valve does not provide a perfect seal between the reactor and main column. In the severe service that the valve encounters during many years of operation, it will probably still leak even after its seat has been thoroughly cleaned.
The block valve is only a convenience, not a substitute for a blind. With an appropriate start-up logic, the block valve will provide the convenience without imposing a safety risk. The refiner must assume that the block valve will leak and must ensure that the leak occurs in the correct direction.
When the main column has a fuel gas blanket and the reactor still contains air, it is safer to have fuel gas from the main column leak into a cold reactor than to have air from the reactor leak into a hot main column. This is especially true if the reactor vapor line vent is open and the reactor has a small flow of steam.
When the main column has a fuel gas blanket, its pressure should be 1-2 psi greater than that of the reactor. Fig. 3 shows that logic in the start-up procedure.
The reactor vapor line vent should sweep out any gas that leaks from the main column; there should be no gas buildup in the reactor. More steam will flow to the reactor when catalyst circulation starts to heat it up.
Even though only Fig. 3 shows the step in which the vent on the reactor vapor line opens, this step still occurs in Figs. 1 and 2.
Start-up without blind or block valve
Once the refiner accepts the logic in Fig. 3, the same logic can eliminate the blind or block valve in the reactor vapor line.
Once it realizes that the reactor vapor line valve will leak, the refiner should treat the reactor and main column as a unit instead of separately. That is exactly what happens after a power failure.
Refiners do not install a blind or close the valve before restarting an FCC once power is restored. The refiner assumes that the reactor and main column are air-free and under a gas blanket while the regenerator has hot air and catalyst in it. The only thing between the reactor and regenerator is the catalyst seal above both slide valves.
Many refiners do not start up that way initially. Instead of assuming that the fuel gas blanket in the main column can leak into the reactor, the refiner should assume that fuel gas blanketing the reactor and main column could leak into the regenerator through the slide valves. The regenerator purges better than the reactor due to the large volume of air from the main air blower.
null
Fig. 4 shows the start-up procedure assuming the fuel gas leak. Pressure in the reactor and main column is still maintained 2 psi greater than that of the regenerator. This also eliminates the need for a vent on the reactor vapor line; this is the procedure that the refiner must use if there is no vent.
Fig. 4 only requires a logic diagram that includes a simultaneous reactor and main column steam out, and that establishes a fuel gas blanket before starting the main air blower.
Because of the large volume of air constantly flowing through the regenerator from the main air blower, the refiner can maintain a reactor pressure 2 psi greater than that of the regenerator until the catalyst has circulated long enough to establish catalyst seals above the two slide valves.
The refiner can then establish the correct DP for start-up.
The air-fuel gas interface therefore moves from the reactor-main column (Fig. 3) to the reactor-regenerator (Fig. 4).
In each scenario, however, the refiner must have good pressure control in the main column overhead accumulator. The wet gas compressor provides the best control, which Fig. 4 shows and Fig. 3 implies.
To implement this solution, the refiner must install a line from the fuel gas system to the wet gas compressor suction. The compressor can then circulate fuel gas from the fuel gas system, through the gas concentration unit, and back to the fuel gas system.
The line must be large enough so that it and the compressor spillbacks can accommodate the entire flow from the wet gas compressor.
This solution has two other advantages:
- It creates a purge system on the reactor and main column. If any air gets into the reactor during catalyst circulation, it will purge to the main column with lift steam in the riser. Once air reaches the main column, circulating fuel gas will take it back to the fuel gas system.
- The reactor-regenerator DP will remain stable during the start-up. As riser feed increases, the refiner only needs to close the line from the fuel gas system slowly to keep the wet gas compressor's spillback on control.
Once the reactor produces sufficient wet gas, the refiner can close the line from the fuel gas system. This eliminates flaring of wet gas and eliminates any pressure surges that occur when the wet gas compressor goes online after the start-up.
Surges always occur when the compressor governor and spillback controls seek their control points when the compressor is first brought up to speed.
Critical path diagram
Once a refiner is comfortable with a particular FCC start-up method, the logic diagram becomes the basis for the critical path diagram. Up to this point, arrows in the logic diagrams are all the same length because no time function is included.
The critical path diagram incorporates general time frames into the logic diagram. The critical path can be as simple and general, or as specific, as the refiner desires.
The initial critical path should be as simple as possible. When particular details are learned through experience, the refiner can add complexity or depth as desired.
The refiner can adjust time frames and arrows to correspond to the actual times required for various steps. The refiner can then expand the general steps to be more specific. For example, the steps that pertain to loading and heating the catalyst to 1,200° F. can expand to several steps (Fig. 5).
The bottom option in Fig. 5 allows for earlier regenerator torching if the proper conditions are satisfied. This is slightly more complex than the first alternative, which the refiner would learn through experience.
The physical layout of some FCCs allows a fast and easy catalyst loading. Other FCCs can be more difficult and slow to load, and the air heater can maintain a higher regenerator temperature. That would allow the use of torch oil earlier in the loading sequence once a sufficient regenerator catalyst level is available.
The critical path diagram is an evolving instrument. It should not be a permanent solution that the operators will eventually loathe and fear; it is a tool that should reflect the history of actual capabilities given reasonable conditions.
A critical path diagram should simplify the operator's workload, not intensify it. As more and more detail and complexity are added, the diagram is useful in determining the physical and personnel resources that are required for a start-up.
Most important, when management asks how long it will take to start up the FCC, the answer will be readily available. The answer will come directly from the critical path diagram, no matter how elementary or complex, because it is based on past history and experience for that specific FCC.
More detail in a critical path diagram will eliminate the awkward moments when everybody is standing around waiting for one particular event to occur before the start-up can continue.
For example, one refiner I know had to wait to inventory the condensate in all the steam drums used to heat the main column before oil was fed to the riser.
Because one of these drums is higher, it always took 1-1.5 hr to inventory it while the reactor was held at 940° F., waiting for feed. The standard joke in the refinery was that the clock in the FCC control room ran 90 min slow.
A critical path diagram highlighted the need to start inventorying these steam drums 90 min before the reactor would reach 940° F.
The author
James D. Weith (jweith@practicalengin uity.com) is the principal of Practical Enginuity, Mission Viejo, Calif., a company he formed in 2001. Weith has also worked as a process and project engineer at UOP LLC, a process-project engineer at Plateau Inc., a consulting refinery engineer at Unocal Inc., and principal engineer at Fluor Daniel Inc. He holds a BS in chemical engineering from the University of Colorado and an MBA and MSBA in management and organization from the University of Southern California. Weith is a member of AICHE and is a registered professional engineer in Colorado.
