Search

John Aikman (Grace Catalysts Technologies)

While there are several operational and mechanical factors that can influence a unit’s particulate emissions, the question asks specifically about the FCC catalyst; as such, the following discussion will address characteristics of fresh catalyst only.

There are four basic characteristics of FCC catalyst that can have direct effects on particulate emissions. These same characteristics will also affect particulate losses to the fractionator and slurry product. The first characteristic is simply the amount of fines content coming into the unit with the fresh catalyst due to the manufacturing process. Figure 1 is an example of a typical fresh catalyst particle size distribution, with a theoretical depiction of a cyclone’s ability to retain fresh catalyst particles. DPTh is the smallest particle diameter which can reliably be collected by a cyclone and is used to model cyclone performance. Particles below this size will be lost by the cyclone.

A review of the Grace Ecat database showed that none of the FCCU’s in North America can retain any 0-20 μ range particles. In addition, they only retain an average of approximately 4.0 wt% in the 0-40 μrange. Fresh catalyst typically ranges anywhere from 9 to 16% of 0-40 μ depending on the supplier andmanufacturing process. Some units require higher amounts of 0-40 μ range particles to help with circulation.

The next characteristic of fresh catalyst that must be considered is the particle density. he DPTh mentioned above will decrease with increased catalyst particle density, per Equation 1 below. This means that cyclones can retain smaller particle sizes as the particle density increases. This is due to the centrifugal force acting on a heavier particle.  However, particle density is not the same as apparent bulk density (ABD). Industry typically measures and reports ABD as part of the routine Ecat analysis, but this should not be mistaken for particle density for cyclone efficiency purposes. Since Al2O3 is denser than SiO2, catalysts with higher alumina content will have higher catalyst particle density.

The third characteristic is the inherent attrition resistance of the fresh catalyst. Industry measures the attrition resistance via a variety of tests, with the primary goal of providing a relative indication of catalyst attrition resistance. Grace utilizes the DI test or Davison Index. On the DI scale, a lower number is less likely to cause attrition and generate microfines. It is usually not valid to compare attrition resistance results obtained from different laboratories. Additionally, it is important to note that the energy applied to a catalyst sample during attrition testing is much more severe than commercial conditions.

As discussed above, the majority of the microfines created in the FCCU will leave the unit through either the reactor or regenerator cyclones, with the latter potentially contributing to increased particulate emissions at the stack.

'The attrition resistance of the catalyst is a function of the manufacturing process and the binder material utilized during the manufacturing process. Figure 2 is an example of how a refiner improved the FCCU stack opacity with catalyst formulation. The reduction was achieved changing to a Grace supplied catalyst with lower DI and lower 0-40  μ content in the fresh catalyst.

The final characteristic of fresh catalyst that affects particulate emissions is its morphology. Morphology can be defined as the study of the form and structure of a particle and its specific structural features. A catalyst particle that has a smoother exterior surface is less likely to generate microfines in an FCCU. Even catalysts with a low fresh DI measurement can cause increased particulate emissions if there are surface irregularities resulting from the manufacturing process. In order to demonstrate this visually, Figures 3 and 4 present SEM’s (scanning electron microscopy) of “bad” and “good” fresh catalyst morphology for a side-by-side comparison.

Figure 3 and 4 SEM’s of Fresh Catalyst (magnified X250)

“Bad Morphology” “Good Morphology” In conclusion, there are several characteristics of fresh catalyst that can be controlled to reduce particle losses and thereby reduce flue gas emissions. Specifically, to lower emissions the fresh FCCcatalyst should possess the following characteristics: a particle size distribution with an optimal range of 0-40 μ particles, higher catalyst particle density, lower DI, and superior morphology. Grace’s alumina-sol technology provides superior binding to the catalyst particle leading to best-in-industry attrition resistance. The versatility and performance of alumina-sol catalysts coupled with Grace’s manufacturing capability, have resulted in wide-market acceptance and as a result, Grace is the preferred FCC technology for loss sensitive units around the world.

Gary Gianzon (Marathon Petroleum Company)

When one of MPC’s refineries starts processing heavy Canadian resid, they add 5 to 10 volume percent of slurry oil in the feed to mitigate making shot coke. The slurry also helps meet anode grade specifications on metals and sulfur. Processing slurry backs out resid processing which can impact unit economics.

FCCU slurry has a similar boiling range to heavy coker gasoil, so a large amount of slurry flashes out of the drum and ends up in the heavy coker gasoil product. The coke yield from slurry feed is around 2 to 3 x Concarbon (depending on coker unit operation) which is significantly higher than vacuum resid at 1.3 to 1.6 x Concarbon. If a high percentage (over 10 percent) of slurry is processed in the coker unit, the slurry can cycle up between the coker and FCCU unit. The amount of recycle built-up is somewhat self-correcting depending on operations in the coker and FCC and whether the HCGO is processed in a FCCU Feed Hydroteater.

Rajkumar Ghosh (Indian Oil Corporation)

We are adding approx. 3–4 wt% FCC Slurry oil in Coker feed in one of our Coker and about 10 wt% in another. We also had undertaken a study in the Delayed Coker pilot plant in our R&D centre. Our experiences with processing of FCC slurry oil in the Coker feed, based on field and pilot plant results, are as under:

a) Yield: The impact of slurry oil in Coker feed depends upon the quality of the base feedstock, CLO/slurry oil and also the pressure / temperature of the coke drums. If FCC slurry oil boiling point distribution and the coke drum pressure / temperature are such that most of the slurry oil vaporizes out of the coke drum, yield of coke and gas reduces with increase in distillate yield.

    In case of Fuel grade Coker, with CLO (with minimum overlap of LCO) below 10wt% in VR feed, coke yield by and large may be constant or may increase marginally depending on the relative quality of VR and CLO. Yields of total gas and liquid decrease marginally. Beyond 10 wt% (10-20 wt%) of CLO in VR feed, the coke yield may increase up to 4 wt%.

b) Capacity: The Coke produced with significant FCC slurry in Coker feed (>10 wt%) has a close-knit Coke matrix which ensures good porous structure to the Coke bed. This reduces the chances of hot spots and blowouts. But the negative impact of adding FCC slurry is pronounced where the coke drum is already limiting, as the porous structure results in lower coke bed bulk density and hence lesser vapor space in the Coke Drum. It may limit the Coker capacity.

c) Product quality: Tendency of formation of Shot coke significantly reduces with the addition of FCC slurry in the Coker feed, as it keeps asphaltenes in solution form. As per our experience at Panipat Coker, impact of slurry addition in the Coker feed is clearly visible on the Coke quality w.r.t. reduction in Shot coke formation. With increased FCC slurry in Coker feed, increase in Silica content in the green Coke would be a criterion to limit its wt% in the feed. This is significant for the Cokers producing Anode grade coke. Typical limit of Silica in Anode grade green Coke is 0.02 wt % max. Depending on the quality of the slurry oil and unit operating conditions, there may be a negative impact on the quality of the LCGO and HCGO. They will become more aromatics and heavier.

d) Energy: Slurry processing will require higher heater duty. High aromatic content in the slurry oil prevents the precipitation of Asphaltenes and thus increases the heater run length. Injection of slurry oil into the coke feed is limited by refinery configuration. In our Refineries with FCC and/or Hydrocracking units, we limit the slurry oil within 5 to 10 wt% on fresh feed to Coker. Increase in injection rate can lead to a massive recycle between the Coker and the FCC or will result in accelerated catalyst deactivation in the Hydrocracker unit.

Eberhard Lucke (Commonwealth E&C)

In general, FCC Slurry has a similar effect as VGO in terms that it replaces residue in the feed and increases mainly the HCGO yield. The difference in this case is that FCC Slurry is a highly aromatic stream and is often used as additional Coker feed (up to 15wt% max. recommended) to reduce heater fouling and to push coke morphology to sponge coke (for anode grade coke). The heavy aromatics in the FCC Slurry help keeping asphaltenes in solution a lot longer and promote coke formation by poly-condensation, therefore increasing sponge type coke content in the coke bed (preferred for low sulfur, anode grade coke production). On the downside, FCC Slurry will contain entrained catalyst fines and – if too high in concentration – may have a negative impact on fouling rates in the charge heater(s). The fine catalyst particles can deposit inside the heater tubes, act as seeds for coking and may promote deposits of heavy oil and coke fines from the oil film inside the tubes.

ESTEBAN (Suncor Energy, Inc.)

We are going to skip the first slide.

The second slide shows a simple depiction of the layers of protection that we use at our different sites. In some cases, we have relief protection, basic process controls, and critical alarm systems on our feed drums to prevent a backflow scenario or the consequences of a backflow scenario. That being said, though, relief valves do not always provide an adequate level of protection for high pressure units. So obviously, take that with a grain of salt. Our primary layer of protection is provided by our trip valves which are activated by SIL-rated instruments. We do not have an SIL rating in all cases; but in some cases, it is required to get the level of protection we need. And then, of course, we also employ dual check valves of differing types downstream of our pumps. Those check valves will typically wind up on our critical check valve system as well.

The scenario is similar where you have backflow. It is not so much the concern of backflow of reactor contents through the furnace, but more just a loss of containment in the furnace itself. We do not treat these furnaces in our hydroprocessing units any differently than we do any of our other furnaces in the refinery. They all have an integrity operating window that we would like to stay in. That window defines at what burner pressures we need to operate and, of course, at what skin and overall box temperatures we can operate.

Our layers of protection are very similar here in that we have trip valves activated by SIL-rated instruments and which are only SIL-rated as required. And of course, we have basic process controls and critical alarm systems. In some, but not all, cases, we do have check valves downstream of our furnaces. That is not a standard at all our sites. However, on some sites, we are consistent about having check valves downstream of our furnaces.

KEVIN PROOPS (Solomon Associates)

I would like to comment on the heater part of the question. Reactor charge furnaces potentially have substantially higher consequences of failure than do most of the other furnaces in your refinery, so you need to be a lot more scared of them than you do of the other ones.

First, this is generally an exothermic process; so, the best case is probably that the furnace is not firing or is only minimally firing. Adequate feed-effluent heat exchange reduces firing and thus the risk of failure from flame impingement. Second is the inherent safety design. If you can go to a single-phase furnace instead of a two-phase furnace, then if it does rupture, your consequence will probably be a lot less. That also gets into the control system issues. Some refiners use hot oil utility instead of a fired heater in the hydrotreater. This is inherently safer.

Then you get into how to avoid a tube failure in the first place. There are a lot of ways to do that, but consider dry point in naphtha units, burner ring pressure controls and interlocks, and maintaining the cleanliness of the burners (fuel gas filtering). Adequate burner-to-tube spacing, feed filtration, tube monitoring (thermography), operator rounds frequency, and upgraded tube metallurgy can all add layers of protection.

Finally, it comes down to culture as well. You do not want to get into a situation of risk of a failure competing with profit to keep the unit maximizing at full throughput. Unfortunately, I have seen a case where that did happen: A furnace was experiencing flame impingement, and the operators did not reduce charge to the unit. After one shift, a tube failed, which led to the entire unit being consumed in a flash fire within a few seconds. We were very fortunate that there was no one outside at the time it happened; if there had been, we would have killed anyone in the unit.

ESTEBAN (Suncor Energy, Inc.)

We do treat them differently depending on their operating pressures and/or requirements. Certainly, from a design standpoint, the operating envelope for each individual piece of equipment changes how the furnace is designed overall. That being said, we evaluate all equipment using the same standard with a process hazard analysis to determine the appropriate layers of protection. Given the required layers of protection, we identify additional safeguards as required by LOPA. SIL-rated instruments, for example, are not required for all burner management systems. However, in some cases, they may be required because the consequences of failure are higher.

So yes, the consequences are significantly greater on a high-pressure furnace. The assessed risk ranking would define how those layers of protection will appear. In some cases, you will see a simpler system on a furnace; and in others, much more complex layers of protection will be applied because of the potential consequences of equipment failure for that furnace. So, to re-phrase my response, I will say that we evaluate every piece of equipment using the same processes.

ESTEBAN (Suncor Energy, Inc.)

In order to reduce the risk of potentially catastrophic consequences related to backflow in high pressure hydroprocessing services Suncor Energy, Inc. uses several independent layers of protection at operating pressure boundaries. One common boundary is for hydroprocessing units, is between the unit feed drum and the reactor charge pump. A typical hydroprocessing unit will have relatively low design pressure equipment upstream of the reactor charge pump which boosts the operating pressure of the feed stream to the much higher reactor operating pressure. As such preventing back flow in the event of the loss of a feed charge pump is critical to prevent equipment failure in upstream equipment with catastrophic consequences. In this application Suncor Energy, Inc. applies the use the following layers of protection:

1. Primary protection is typically a Safety Instrumented System (SIS) that monitors the run status of the feed charge pump via multiple direct and indirect instrumented signals and activates quick acting trip valves and in some cases closes the feed charge control valves in the event of a shutdown. In some cases, depending on the unit specific hazard analysis these systems may be SIL-rated to ensure reliable operation when activated. In addition, these systems are often designed to be activated by any one of several different instruments used to sense a potential backflow scenario, i.e., low-low flow shutdowns and low-low feed controller differential pressure shutdowns.

2. In some cases, pressure relief valves are used as layers of protection for overpressure due to backflow, but caution must be applied when relying on a relief valve as protection for vessels, such as feed drums, since these valves are not always sized for backflow scenarios.

3. Mechanical safety systems are also employed depending on unit design. While these systems are often not credited in a process hazard analysis of a unit they can provide additional layers of protection. Typical installations include dual check valves of different design which are often deemed critical check valves that require routine maintenance.

4. Provided the design of the system and equipment in some cases basic process controls and/or critical alarms with operator response are employed as additional layers of protection.

In addition to backflow prevention and protection as it relates to pressure boundaries, furnace tube ruptures can result in backflow from multiple large high-pressure vessels to atmosphere with catastrophic consequences. In order to address the release of reactor and high-pressure circuit equipment, layers of protection must be applied to the feed furnaces that prevent operating windows that have the potential to create damage resulting in tube rupture. The layers of protection employed for this scenario do not differ from those on other furnaces in Suncor’s refineries, as all furnaces are evaluated for tube rupture scenarios. However, in this application Suncor Energy, Inc. applies the use of the following layers of protection:

1. Primary protection is typically a SIS that monitors furnace flows, temperatures, and fuel and box pressures via multiple direct and indirect instrumented signals and activates quick acting trip valves on fuel supply and in some cases closes the fuel supply control valves in the event of operation outside a preset operating window. These SISs often activate related SISs to stop process flows. In some cases, depending on the unit specific hazard analysis these systems may be SIL-rated to ensure reliable operation when activated. In addition, these systems are often designed to be activated by any one of several different instruments used to sense operation outside of the specified window, i.e., low-low flow shutdowns and high-high burner pressure shutdowns.

2. Mechanical safety systems are also employed depending on unit design. While these systems are often not credited in a process hazard analysis of a unit, they can provide additional layers of protection. Typical installations include check valves downstream of furnaces to prevent the backflow of reactor contents. In general, these check valves are not relied upon as fail-safe devices and are not considered critical check valves.

3. Provided the design of the system and equipment in some cases basic process controls and/or critical alarms with operator response are employed as additional layers of protection.