May 2021

Special Focus: Maintenance and Reliability

Stability evaluation of hot spots in the RDS process reactor

During the end-of-run (EOR) period of the authors’ company’s residue hydrodesulfurization (RDS) process in early February 2019, some hot spots occurred in the catalysts of the reactors.

During the end-of-run (EOR) period of the authors’ company’s residue hydrodesulfurization (RDS) process in early February 2019, some hot spots occurred in the catalysts of the reactors. The normal operating temperatures of the catalytic bed of the reactors are about 370°C–400°C, and the temperature gradually rises as the severity of the process increases during the EOR period. However, the temperatures at that time reached 600°C–1,100°C. Fortunately, the planned shutdown of the process was in March, so personnel were able to adjust the operational conditions for the remaining month.

Impurities such as metal, sulfur and asphaltene accumulate in the catalysts, resulting in uneven feedstock flow. Areas that are not in contact with the feedstock undergo a rapid exothermic reaction inside, which is called a “hot spot” phenomenon. It occurs occasionally during the EOR period because a lot of impurities accumulate in the catalysts and the operational severity of the process is increased to fully use the remaining catalyst life. In addition, if the catalyst loading quality is poor, it rarely appears, even during normal operation.

When a hot spot occurs, the bed temperature rises rapidly in a short time due to a rapid heat reaction, and the range is expanded and continuously deteriorated. The inside of a hot spot mass comprises a solid chunk of coke and is made worse by not penetrating the feedstock coolant.

In this study, the authors modeled hot spot phenomena experienced to analyze the temperature effects of the catalysts, reactor shells and internal parts. This was completed through computational fluid dynamics (CFD) analysis and, with finite element analysis (FEA) techniques, these results were used to evaluate structural reliability. Also reviewed was the corrosion mechanism of internal structures exposed at high temperatures during this time.

Ultimately, the purpose of this study was to assess structural stability by estimating the exposed temperature of reactor internals and the shell in the event of similar hot spot phenomena in the future—and to predict corrosivity and devise measures during operation. In other words, if an abnormal temperature zone occurs due to a hot spot, then an operational guide could be provided to determine whether the process can be operated continuously or if feedstock supply must be halted and measures taken (such as nitrogen cooling).

Hot spots

Hot spot cases are very diverse, and their characteristics may differ depending on shape, temperature and location of occurrence. Hot spot sizing is not easy, but it can be estimated using thermocouples for temperature sensing of catalysts. It could be easier if thermocouples for multi-point sensing were installed. The actual size can be checked during the catalyst replacement work during a shutdown. The hot spot area rises rapidly in a short time and is observed as a hard coke mass.

The analysis was done in three stages. These include:

  1. The temperature profile—in axial and radial directions—was analyzed through CFD after modeling hot spots. The authors compared several items to find the best model, since there were a wide variety of cases.
  2. It was possible to know the exposure temperature of the surrounding structures through CFD of hot spot modeling. The mechanical stability of the structure was evaluated by reflecting this exposure temperature through the FEA method. It was the structure affected by the shell, interbed beam and outlet collector.
  3. The important item was the corrosion review. In general, Tp347 is a corrosion-resistant stainless steel in the typical operating envelope of a hydroprocessing reactor. However, at elevated temperatures above the design condition, Tp347 is known to be susceptible to high-temperature corrosion and/or metallurgical change associated with chromium carbide precipitation.

Temperature profile analysis of catalysts and structures by CFD modeling

The purpose of this test is to check the temperature profile in the downward and radial directions, according to the hot spot conditions. The modeling selected a reactor having an interbed beam. The assumed conditions for CFD modeling are shown in TABLE 1.

The structure of the reactor, the hot spot size and location are shown in FIG. 1. It consists of two catalyst beds and an interbed in the middle of the reactor. The bottom head has an outlet collector installed. The shell is made of 2.25 Cr-1 Mo steel with a Tp347 weld deposit, and the internal structure is made of Tp347 material.

FIG. 1. CFD modeling.
FIG. 1. CFD modeling.

The following are the results of the CFD modeling:

  • Temperature profile of the axial direction: The lower direction of the hot spot showed a long tail-type temperature profile (FIG. 2). Three models were found to be affected by temperatures up to about 7,000 mm in the lower direction. Higher temperatures will result in longer temperature distances, and larger sizes will also result in longer lengths. When a hot spot occurs, the lower structure (i.e., the interbed beam and outlet collector) is directly affected by the temperature. It is seen that this is related to the directional flow of the feedstock, and it can be estimated that a long tail-shaped temperature profile can be seen in the lower direction. However, it was only possible to confirm how far the influence would be through this modeling.
FIG. 2. Temperature profile for sizing hot spots.
FIG. 2. Temperature profile for sizing hot spots.

  • Temperature profile of the radial direction: Conversely, the temperature effect in the radial direction was very limited. If the hot spot was only 150 mm away from the reactor wall, the wall temperature could be reduced below the design temperature (FIG. 3). This is closely related to the directional flow of the feedstock, which seems to rapidly reduce the range of high-temperature areas as feedstock flows from the top to bottom. This means that, even if the hot spot location is close to the wall, the temperature increase of the wall does not have to be a concern. In fact, in the authors’ experience, an increase in wall temperature was not detected. In the structural analysis of the reactor wall, the hot spot case at 50 mm was considered. A peculiar item is that the point where the maximum temperature appears is not the area of the hot spot occurrence. Since the maximum temperature area appeared below approximately 1.8 m, it is necessary to take this into account when checking the actual shell temperature. This can be confirmed in the CFD temperature profile on the right side of FIG. 3.
FIG. 3. Inside wall temperature.
FIG. 3. Inside wall temperature.

To summarize the CFD results, the temperature effect is much larger in the direction below the hot spot, while the effect is very small in the radial direction. This seems to be closely related to the directional flow of the feedstock.

Based on this, the master curve—which can be used for various hot spot sizes—is prepared to estimate the exposure temperature of the lower structure, by distance, according to the hot spot size and temperature (FIG. 4). The exposure temperature of the lower structure is 50°C–160°C lower than the hot spot indicator temperature. For example, with a hot spot of 900°C, the exposure temperature of the lower structure can be predicted to be as high as 750°C–840°C.

FIG. 4. Maximum surface temperature of the internal structure according to hot spot size and temperature.
FIG. 4. Maximum surface temperature of the internal structure according to hot spot size and temperature.

Evaluation of the shell and interior structure stability through FEA

A structural stability evaluation was performed on the reactor wall, interbed beam and outlet collector. Only 800°C and 500-mm diameters were considered for the shell and outlet collector, while 500 mm, 1,000 mm and 1,500 mm were considered for the interbed beam. The standard ASME VIII, Div. 2 Part 5 specification was applied to the analysis.

The shell’s material comprises 2.25Cr-Mo steel with thermal conductivity of 35 W/mK, and the internal structures are composed of Tp347 steel with thermal conductivity of 25 W/mK.

Evaluation results. The following are the results of the evaluation:

  • Reactor wall: Consider when a hot spot mass with a temperature of 800°C and a 500-mm diameter exists 50 mm away from the wall. As a result of the analysis, the maximum stress caused by the hot spot was 322 MPa—and there were no problems, as it was below the acceptance criteria of 331 MPa. It was analyzed that the maximum stress was applied inside of SCL-A-2-L1; whereas, it was found that lower stress was applied at the maximum temperature point (SCL-A-2-L2). The previous CFD analysis indicated that the maximum surface temperature of the wall was located at the lower part of a certain distance, not at the point of the hot spot occurrence. The stress analysis results were similar (FIG. 5).
  • Interbed beam: The top of the interbeam’s center showed the highest temperature (750°C), with a stress of 48.6 MPa (less than the allowable stress of 230 MPa). Conversely, the maximum stress at the bottom of the beam’s center was 172 MPa (less than the allowable stress of 315 MPa), and the temperature was 638°C. The maximum temperature was at the top, but the maximum stress was at the bottom. Although the surface temperature of the upper part of the beam was at its maximum due to the hot spot, it was in line with the expectation that the maximum bending stress would be the bottom point of the beam (FIG. 6).
  • Outlet collector: As a result of the evaluation of the outlet collector, the maximum temperature area was 551.6°C with the structure’s top, but the calculated stress value was 82 MPa (less than the allowable stress of 354 MPa), which was enough (FIG. 7). Conversely, the bottom area had a maximum calculated stress of 337.2 MPa (less than the allowable stress of 396 MPa). This area was where the maximum bending stress was applied, as shown in the inter-beam analysis. In other words, the surface temperature at the top of the structure exceeded the design conditions due to the hot spot, but it was not the area where the maximum stress was applied, so it could be assessed that there was no significant impact on the stability of the structure.
FIG. 5. Stress at the reactor wall.
FIG. 5. Stress at the reactor wall.
FIG. 6. Stress at the interbed beam.
FIG. 6. Stress at the interbed beam.
FIG. 7. Stress at the outlet collector.
FIG. 7. Stress at the outlet collector.

Summarizing the structural stability evaluation results, the surface temperature of the shell wall, interbed beam and outlet collector structure exceeded the design temperature by the hot spot. It was decided that there were no structural stability problems, even if the temperature of some of the shell and internal structures exceeded the design standard due to the hot spot effect. The cause seemed to be that the maximum temperature point and the maximum stress point did not coincide in the structure. In other words, due to the hot spot, the temperature exceeded the design conditions locally, but there was no problem with the stability of the structure.

Corrosion analysis

A review of the corrosion environment is essential, even if structural stability assessment results are satisfied. The internal structure is composed of a part supporting the load, a beam, a supporting bar and a mesh screen that simply prevents catalysts from passing through. A thermowell pipe is installed to measure the temperature. If these parts are severely damaged by corrosion or cracks, then catalyst leaks or structure collapse may occur during operation, and repair or replacement may be long and expensive. In particular, the thin mesh screen is susceptible to damage and, if the gap is widened and the catalyst escapes, normal operation is not possible.

In normal operating conditions, corrosion of the internal parts (Tp347 steel) rarely occurs, so no special care is taken. However, the problem changes at high temperatures above 500°C. If this temperature is exceeded, high-temperature sulfidation (HTS) corrosion and/or polythionic acid stress corrosion cracking (PASCC), induced by sensitization, may occur. While HTS is a well-known corrosion phenomenon within the typical operating range of a hydroprocessing unit, corrosion data above the design temperature of the reactor (454°C) is limited, especially for austenitic stainless steel. The corrosion rate of Tp347 stainless steel at 500°C–600°C has a maximum of 60 mils/yr. However, this value has a low confidence level because it is calculated by extrapolation.

In the authors’ case, the process ran for about a month after the hot spot occurred. As a result of the inspection of the internal structure during shutdown, a severe corrosion phenomenon was found on part of the mesh screen of the outlet collector and on several thermowell pipes on the wall. Even with the austenitic stainless-steel material, it was found that corrosion may occur within a short period of time under hot spot conditions where the temperature rises rapidly. As shown in FIG. 8, the mesh screen had a wider gap, and some of the mesh was completely corroded. If exposed for a longer period, the catalyst above may have been lost. Although the thermowell pipe was 3-mm thick, it was completely corroded within a short period of time, and high-temperature corrosion was accelerated by carburization and metal dusting phenomena. Based on the authors’ inspection data, the amount of corrosion was acceptable when the exposed metal temperatures were below 700°C, and the duration of high-temperature exposure was less than 1 mos.

FIG. 8. Corrosion at the internal parts.
FIG. 8. Corrosion at the internal parts.

In the case of PASCC, it is known that sensitization occurs first and contact with polythionic acid is required. Sensitization occurs more frequently with higher temperatures. According to literature, austenitic 347 stainless steel can be exposed at 495°C for 10,000 hr, but at 530°C for only 1,700 hr. In the authors’ case, the microstructure of the bolting bar—which is a part of the bottom collector assembly—was inspected using the filed replication technique. It was found that there were no signs of sensitization, even though corrosion was spotted at the nearby mesh. Note: The microstructural review was conducted after replacing corroded/damaged internal parts. Therefore, the authors believe that the remaining part of the reactor internal, which was not replaced, still maintains enough PASCC resistance because the temperature and duration of the hot spot were not enough to cause sensitization.

In summary, the possibility of HTS corrosion should be kept in mind, even if the hot spot period is a short duration. Furthermore, high-temperature exposure may cause sensitization of austenitic stainless steel, but the level of damage should be carefully reviewed by metallography.

Considering the exposure temperature, time and area, the authors decided to maintain a temperature of 650°C. Corrosion testing under the same conditions—high temperatures with high sulfur—was not easy. Only a hardness comparison was made through simple heat treatment experiments, and some of the 650°C limit was reflected.

Takeaway

CFD modeling of hot spot cases in the RDS reactor catalysts was performed to analyze temperature profiles for catalysts, reactor walls and internal structures. Accordingly, structural stability was evaluated through FEA for structures where temperature increases locally. Corrosion under this condition was also identified, and the following conclusions were drawn:

  • The larger the hot spot size or the higher the internal temperature, the larger the temperature area of the axial direction—and the temperature of the lower internal structure increases locally. In the downward direction, the influence of temperature was confirmed to about 7,000 mm, as it was affected by the directional flow of feedstock.
  • Conversely, the effect of temperature in the radial direction was relatively small. Even if it is only about 150 mm away, it has been cooled rapidly below the design temperature. This means that, even if an internal hot spot occurs, the temperature of the reactor wall does not rise significantly.
  • It was analyzed that the temperature of the internal structure increased locally due to the hot spot. Compared to the 500-mm distance, the structure was evaluated to be 50°C–150°C lower than the hot spot temperature.
  • As a result of evaluating the structural stability of the wall and internal structure of the reactor reflecting CFD, there was no indication that the structural stability was unsatisfactory, even if the design temperature was exceeded locally.
  • Two types of corrosion must be considered: HTS and PASCC. In this case, corrosion by HTS was observed on the mesh screen and thermowell pipe. For HTS, the exposure temperature is important, and the authors suggested a short-term allowable excursion temperature of 650°C. No PASCC was found on the remaining internal parts by field metallography because the temperature and duration were not enough to cause sensitization.

This study is the result of reviewing the hot spot case of the RDS reactor. It is difficult to apply these results to all hot spots; however, when similar hot spots occur, it is believed that reasonable decisions can be made by referring to them. HP

The Authors

Related Articles

From the Archive

Comments

Comments

{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}