You are here

External Weld Repair of Coke Drums


Delayed Coking Units allow oil refineries to process “heavier” oil. The main function of these units is to boil the residual oil feed to the thermal cracking temperature in order to produce gas oil and petroleum coke. The coke drums, where this operation takes place, are very large vertical pressure vessels, with size ranging from 4 to 9 m in diameter and 25 to 40 m in height. Such systems have been fabricated through the years using a range of materials from carbon steel (A516-70) to low carbon alloy steels (A204 C, A387-11, A387-12) with Cr (1-2.25) and Mo (0.5-1.0) additions, with internal cladding made of 13% (405) or 12% (410S) Cr steels. The max. shell thickness ranges between 14 and 42 mm. During operation, these vessels are subjected to severe thermo-mechanical cycling, which causes accumulative damage by creep-fatigue. Therefore, repair is not uncommon in such equipment. However, external repair technology faces several challenges and is still under development. OSU has recently engaged jointly with Shell in a preliminary study to address external repair technology and this project is a follow up on such endeavor.

This project will evaluate several procedures and filler materials combinations that could be used for external repair of coke drums. The proposed study combines a fundamental metallurgical approach with actual welding techniques, creep-fatigue testing and advanced characterization to address and understand the actual performance and failure characteristics of such complex repair joints under simulated coke drum operation conditions.


1.   Evaluate diverse welding techniques and filler metals combinations for external cokedrum repair.
2.   Develop an accelerated Gleeble-based test to evaluate coke drum material under complex thermo-mechanical cycling.
3.   Compare the coke drum shell material and welded repair joints performance under conventional isothermal low-cycle fatigue and Gleeble-based testing.
4.   Optimize external repair welding procedure using the developed test, advanced characterization and metallurgical (thermodynamic and kinetic) modeling.
5.   Obtain a better understanding of the coke drum repair joints microstructural evolution and failure mechanism. Finally, based on it provide recommendations for further fabrication and repair techniques/materials improvement.


Identification and implementation of most suitable repair techniques for external coke drum repair for the preparation of mockups. Some joints already produced during preliminary work at OSU will be used for preliminary work on this task. However, additional joints using additional base and filler materials will be produced. The special order production of filler metals will be considered (Special Metals, a member of MA2JIC will be invited to participate on this development/production stage).

Development of Gleeble-based accelerated test to evaluate coke drum material under thermo-mechanical cycling. The Gleeble allows precise, yet complex, thermal and mechanical cycles to be applied in order to achieve very well controlled and reproducible microstructures and thermomechanical histories. It is anticipated that such test will allow faster and lower cost materials and repair techniques optimization. The test development involves extensive direct comparison with regular isothermal low-cycle fatigue testing.

Thermo-mechanical testing of repair mockups reproducing coke-drum service conditions. Both Gleeble-based and isothermal low-cycle fatigue techniques (to be performed at EWI or other company) will be used. It is anticipated that there will be iterations between this task and the repair technique development (Task 1) to improve/optimize the repair technique.

As welded samples extracted from the repair mockups and subjected to controlled thermo-mechanical conditions during task 3 will be characterized using conventional (optical microscopy, microhardness mapping) and advanced characterization techniques. (SEM, TEM, XEDS, EBSD, TKD, nano-identation) to better understand the welded joint structure, its evolution during service and failure. If provided by any partner or supporting company, actual used coke drum material will also be evaluated. Iterations between this characterization stage and Task 1 and Task 3 is also expected. Metallurgical (thermodynamic and kinetic) modeling will be used to support the material evolution understanding and its effect on performance.


Industry Sponsors: Shell, AZZ, Stress Engineering Services

Faculty: Antonio Ramirez (OSU)

Graduate Student: Sebastian Romo Arango, Guilherme Abreu Faria 

Industry Contacts: Jorge Penso, Darren Barborak, Julian Bedoya