A. Kittur, Saudi Aramco, Ras Tanura, Saudi Arabia
A mercury absorber vessel is designed to remove mercury from a light naphtha stream. A previous study showed that 17 psig was the maximum acceptable limit at the design flowrate conditions for the vessel, and the plant operated at the maximum allowable pressure drop (∆P). However, to meet the production requirement from a debottlenecking study, a higher-than-design flowrate would be required across the mercury recovery unit (MRU), which would require a higher permissible ∆P across the vessel without compromising the mechanical integrity of the equipment, as shown in FIG. 1.
As per the original equipment manufacturer (OEM), the mercury filtration bed could handle a 17-psig ∆P. The pressure differential transmitter measured the differential pressure over the vessels, and the ∆P trends were obtained, as illustrated in FIG. 2.
During normal operations, a steady increase in the ∆P is typically expected across the filtration bed structure. In the previous operational run, it was observed that this ∆P approached the set value of 17 psig much more quickly before the scheduled end-of-run (EOR) conditions. Operational adjustments were required to prevent this ∆P from exceeding a 17-psig set value before the scheduled EOR, which often compromised the efficient productivity of the MRU.
Based on the third-party study of the overall process assessment of debottlenecking and related ∆P issues of the mercury absorber vessels, it was recommended that the maximum allowable ∆P for the vessel was dependent on the design of the vessel’s bottom collector. Therefore, a detailed design review of the maximum ∆P that could be safely sustained by the bottom outlet strainer was required to establish if it was possible to increase this previously established maximum allowable limit.
Based on the known design and the geometric features of the vessel’s bottom collector, a conventional strength calculation cannot be performed for an accurate outcome. Therefore, a finite element method (FEM)-based analysis is required to establish the actual strength of the bottom outlet strainer based on operating stresses equal to 25% of yield strength, corresponding to a design factor (DF) of 4.0 (standard engineering practice). The finite element analysis (FEA) should be a linear, static, elastic, structural stress analysis based on the following mechanical, thermal and fluid loads witnessed by the bottom collector during the active service of the MRU vessels.
The existing condition of the structural plates of the bottom collector should be assumed to be partly corroded due to a corrosion allowance of 3 mm [safe life corrosion (SLC) of 30 yr] provided by the OEM manufacturer. The complex geometry associated with the strainer basket and the thermo-mechanical loading mandated a detailed stress analysis based on ASME Sec VIII Div 21 methodology. The accurate stress distribution was simulated using the FEM.
FEM. The full structural sizes according to the OEM’s drawings had been modeled, and there had been no reduction in thickness due to corrosion for the OEM’s new outlet strainer, as shown in FIG. 3. The appropriate materials properties were chosen for ASTM A-517 Gr-70 as the material of construction for the strainer.2 All existing fabrication fillet welds of external stiffeners are assumed to be 100% sound, resulting in the required frame strength, and the fabrication fillet has been modeled accordingly. For the old outlet strainer, which has already seen operational service as a conservative basis for calculating the strength of the bottom outlet strainer, we will assume the short-term corrosion rate (SCR) = 0.005779 in/yr. For the operating life of 10 yr (2009−2019), the resulting wall loss assumed on the bottom outlet strainer = 0.05779 in. Based on the manufacturer’s corrosion allowance of 3 mm, partly corroded (wall loss due to corrosion = 1.46 mm/0.05779 in.) structural plates were modeled. As the OEM’s drawing had no records of non-destructive examination, ultrasonic/radiographic testing done on the shop welds, the quality of the welds has not been confirmed, resulting in 80% joint efficiency.
The ¼-in. diameter support rods, wrapped and spot welded to a conical section and spot welded to a top plate, have not been modeled in the FEA. The wire mesh screen wire was a 1/16-in. (4 mesh) with a 3/16-in. opening. This screen material is a stainless-steel wire mesh that overlapped ½ in. with the opening. The vertical, top and bottom joint ring plates had also not been modeled. A higher load ∆P bearing capacity could have been simulated if the structural items were modeled in the FEA. The FEA was performed to establish the actual strength of the bottom outlet strainer and the consequential maximum permissible ∆P, corresponding to a DF of 4.0 (standard engineering practice). Therefore, the completed FEA analysis was conservative, enabling a higher maximum allowable ∆P without compromising the mechanical integrity of the bottom outlet strainer.
Loading and boundary conditions.3 The static dead weight load of the ceramic ball and other filtration materials were considered as follows (Eq. 1):
Total = 26,026.5 kg (57,378 lb), corresponding to Φ ID of vessel = 90 in.; Area = 6,358 in.2 (1)
This was based on the updated loading diagram and calculated weight for one vessel, as shown in TABLE 1. The static head of the process liquid in the 28-ft column at 0.72 specific gravity = 8.75 psi was applied on both the top and slant faces of the bottom strainer, respectively. The pressure dropped due to the flowing fluid. ∆P = 20 psi, applied on both the top and slant faces of the bottom strainer, resulting in:
The resultant operating load (∆P + head + dead wt%) on the bottom outlet strainer was calculated: 3,720 lb + 1,628 lb + 2,117 lb = 7,465 lb. The fabrication workshop performed a load test of the check design for a uniform vertical load of 10,000 lb. Based on the load test of 10,000 lb, the existing design seems to have an established design factor of 1.4. This is likely incorrect, as the bottom outlet strainer can withstand a far greater load. The base plate and the locating sleeve outside surfaces are constrained for the boundary conditions with a sliding (frictionless) support. This is to simulate the restraints on the bottom strainer during operation. The ¼-in. base plate has been ensured contact with the vessel bottom head’s circumference (FIG. 4).
Results—Elastic stress analysis. For the new (never installed in operation) outlet strainer:
For the old (experience installed operating service) outlet strainer:
Takeaway. The maximum allowable ∆P of 28 psid can be safely tolerated across the MRU vessel. This is based on an FEM analysis completed on a conservative basis, enabling a higher maximum allowable ∆P without compromising the mechanical integrity of the bottom outlet strainer. HP
REFERENCES
ASLAM KITTUR is a mechanical engineer with expertise in troubleshooting integrity issues related to refinery static equipment. Kittur is the Static Equipment Engineer for the Technical Services Division of Saudi Aramco. He has more than 23 yr of experience in the energy industry (upstream, offshore and downstream), specializing in performing stress analysis to understand the complex failure mechanisms of critical equipment. He focuses on establishing root causes for repetitive failures and developing solutions to ensure a reliable design for all operating conditions. Kittur earned an MS degree in solid mechanics and employs FEM-based structural and thermal simulations often required for a Level-III API-579 Fitness-For-Service Evaluation. Before joining Saudi Aramco, he served in a similar role in Canadian oil sands facilities for 10 yr.