Langfields were delighted to be engaged by Essar Oil UK as the welding contractor for a project that involved the replacement and upgrade of the Fluid Catalytic Cracking Unit (FCCU) during a planned site shutdown. The article below goes into depth about all aspects of the project.
‘Worth Every Nickel’
Authors: Roy Rowlands C Eng CEWE M Weld I, Eur Ing Paul Jordinson C Eng CEWE F Weld I, Eur Ing Dave Godfrey C Eng CEWE F Weld I
As a part of project ‘’Tiger Cub’’ at the Stanlow refinery of Essar Oil UK Ltd, there was a requirement to replace and upgrade a key component part within the Fluid Catalytic Cracking Unit (FCCU) during the planned turnaround event in January 2018.
As a part of the replacement program it was necessary to join a new 1 ¼ Cr ½ Mo vessel to an existing Cr Mo vessel that had been in elevated temperature service for in excess of 30 years. The section thickness and diameter at the connection is 65mm x 1650mm.
The FCCU is one of the most important processes units at Stanlow. It converts the high-boiling, high-molecular weight long residue fraction from the upstream crude oil distillation unit into more valuable low boiling point, low-molecular weight fractions such as gas oils, gasolines and olefinic (double bonded) hydrocarbons such as ethylene, propylene, butylene as well as other valuable products.
A typical FCCU unit produces 11000 tons of product per day, with 78% of the process being converted to other products.
The FCCU is also critical to the refinery for producing very high pressure steam that is used to generate electrical power for the site.
The heart of the FCCU unit is the Reactor and Regenerator (R&R) and is where the cracking of the long residue takes place. This is achieved by contacting heated long residue feed (200 °C) with hot fluidised catalyst (720 °C) at the bottom in the unit’s reactor riser (R2150). The long residue feed is vapourised by the hot catalyst and is thermally cracked into the lower boiling point fractions in the riser. The hydrocarbon vapours and catalyst mixture (530 degree C) exits the top of the riser and pass into the primary cyclone via the transfer duct. The primary cyclone separates the catalyst from the hydrocarbon vapours, with the catalyst being collected and returned to the regenerator where it is re-activated by burning off entrained coke from the reaction process. The separated hot hydrocarbon vapours are fed into the FCCU’s fractionator column where they are de-superheated and drawn off as products based upon their boiling ranges. Un-condensed gases in the top of the fractionator column are compressed and sent to the gas separation unit for further distillation into high value product streams that are either further refined or converted in other downstream process units.
All images below supplied courtesy of Essar Oil UK Ltd.
Figure 1 is an illustration of the FCCU.
This paper discusses the complex welding engineering undertaken by the authors and the named welding contractor to complete the final on-site closure weld on the R2150 reactor riser using a nickel alloy filler metal deposited using manual GTAW and mechanized FCAW processes, to reduce the need for post weld heat treatment (PWHT).
As a part of ‘Project Tiger Cub’ the upgrade of the FCCU includes the replacement of the R2150 reactor riser for a more modern design with an increased diameter and length. In order to join the new reactor riser to the existing FCCU, it was necessary to undertake welding of old to new material. Essar identified the weld in the transfer duct as being complex in its nature due to restrictions caused by access and the location at the 56 metre level of the process unit, challenging metallurgical considerations including a desire to avoid in-situ PWHT and the shape of the transfer duct. The transfer duct installed has been exposed to a service temperature range of 400⁰C to 600⁰C whilst in service.
The existing transfer duct has a wall thickness of 65mm and is constructed from 1.25Cr-0.5Mo as per BS 1501-621-31B and ASTM 387 Gr 11. This duct had been in service for more than 30 years and its metallurgical condition was unknown but was expected to have deteriorated during service.
Given the age of the ASTM 387 Gr 11 material and a service life of 30 years + and the configuration and local restraint of the existing riser, it was critical to eliminate the need for PWHT on the final 65mm thick closure weld due the high probability of temper embrittlement being present within the materials substrate. A further consideration was the need to eliminate the risk of hydrogen cracking occurring in between completion of welding and application of PWHT. Due to site conditions and logistics it was not possible to guarantee that PWHT could be applied immediately after welding, giving rise to the possibility of hydrogen cracking occurring during this period. As the repair was carried out in open site conditions at height during February and March (during the ‘Beast from the East!’) this presented a serious risk.
From a welding engineering perspective, it was critical that the methology chosen would have to be robust and proven via a series of weld trials. The welding engineering team responsible for this work consisted of Roy Rowlands CEng MWeldI CEWE (Essar),-Eur Ing Paul Jordinson CEng FWeldI CEWE (Wood) and Eur Ing Dave Godfrey CEng. FWeldI CEWE (Godfrey Welding Engineering), who were supported during initial development by Marcello Consonni of TWI Ltd. The preferred contractor for the work was Langfields of Manchester. Langfields are very well known in industry as experts in exotic materials and, in particular, nickel alloy welding, where they have had good experience on numerous complex welding projects globally with the typical major oil and gas clients.
Selection of filler metals and welding processes, metallurgical considerations and welding procedure development and qualification
The first activity for a welding engineer is to review the influence of welding onto aged CrMo material (BS1501-2 Gr 621B) with typical chemical composition as shown in table 1 to new CrMo material re ASTM A387 Gr 11.
For in-service plant/components operating at temperatures within the range of 450°C to 650°C, 1.25Cr-0.5Mo steels are susceptible to reheat cracking and creep embrittlement. Temper embrittlement is another concern in welding of Cr-Mo steels and occurs within the same temperature range, but 1.25Cr-0.5Mo steels are less susceptible to temper embrittlement compared to 2.25Cr-1Mo and 3Cr-1Mo steels.
Reheat cracking and creep embrittlement typically occur in the heat affected zone (HAZ) of a weld, especially in a coarse grained HAZ, either during post weld heat treatment (PWHT) or during elevated temperature service.
The risk of these types of cracking is increased by the presence of peak stresses and the possibility of strain localization in the HAZ, presence of embrittling trace elements such as phosphorous, tin and antimony segregating to the grain boundaries, and a large grain size coupled with high hardness.
Based on the above, to avoid stresses from the welding operation impacting directly on any pre-existing HAZs, the location for the new site weld was located well away from any existing girth welds, which meant the new girth weld would be made in an area where the possibility of embrittlement would be lower. To assess the condition of the existing material, additional testing was performed on sections removed during the shutdown. The tests used to evaluate the material’s condition was elevated temperature creep testing, the results of which would not become known until the unit had been returned to service.
The conventional approach to making a butt weld on this material would be to apply a preheat of ≥220⁰C across an area extending for a minimum 75mm either side of the joint, make the weld using consumables of nominally matching composition and apply a Post Weld Heat Treatment (PWHT) of 670⁰C for a period of 1 hour per 25mm thickness, equating to 2.5 hours in this case. The potential difficulty of consistently maintaining the required temperature across a band >200mm in width, 65mm thick and 1650mm diameter also had to be taken into account. Due to the potential for errors to occur before or during PWHT consideration was given to how to eliminate or substantially reduce the need for PWHT.
In addition to the risk of temper embrittlement, the potential hazards associated with a conventional approach to this repair are summarised below:
- Difficulty in maintaining consistent preheat and risk of hydrogen cracking
- The required double sided weld would have required the welders to work inside the transfer duct with 220⁰C preheat applied, this was not considered feasible or safe
- Delay between welding and PWHT, risk of cracking as NDE would need to be performed prior to conducting any PWHT
- Unacceptable residual stresses and high hardnesses resulting from failure to maintain the required PWHT temperature across the specified band
Following this analysis it became clear that it was necessary to develop a procedure that gave increased security against these conditions. The options considered were:
- Use of matching consumables using an overlapping temper bead technique, no PWHT
- Use of high nickel consumables using an overlapping temper bead technique, no PWHT
- Buttering cut end of existing transfer duct with high nickel consumables, apply PWHT, make closing weld using high nickel consumables, no PWHT
Temper beading is an established method to avoid PWHT, but the effectiveness of the process is reliant on the welder maintaining the correct degree of overlay throughout deposition of the initial 2 layers. Failure to achieve sufficient tempering would lead to local hard zones that could then become crack initiation sites. Considering the size of the area that would need to be overlaid and the constantly changing orientation the welders would be encountering, it was concluded that this approach had a high chance of leaving untempered areas in the HAZ, so options 1 & 2 were discounted.
The 3rd option allowed the use of a lower preheat of 180⁰C during the buttering operation while avoiding the possibility of hydrogen cracking occurring if, as happened in practice, a delay occurred between completion of welding and commencement of PWHT. It also meant that the area to be subjected to PWHT was a narrower band at the end of the cut transfer duct, enabling better access and more accurate placement of the heating elements. The full process was then to butter and PWHT the end of the new replacement duct (carried out off-site by Langfields at their Salford works), apply a layer of high nickel consumable to the cut end of the existing transfer duct and apply PWHT in position to a narrow band of parent material, align new to old sections, and make the closing weld using high nickel consumables, without the need for further PWHT. The use of nickel alloy also guarded against the risk of hydrogen cracking occurring before PWHT.
A number of possible welding procedures were debated, considered the best options being to qualify procedures for GTAW and FCAW, with a GTAW and SMAW procedure in place in case difficulties arose using the FCAW. The use of GTAW and SMAW processes are well established for welding Cr Mo materials using Ni Cr alloy filler wires, however the use of FCAW using a Ni Cr filler wires can be problematic in certain positions (typically between the 4 and 8 o’clock positions) on girth welds welded in position with defects such a micro porosity.
The eventual choice of welding consumables where:
- GTAW – Bohler UTP A068HH Classification AWS A5.14: ERNiCr-3
- SMAW- Bohler UTP 7015 Classification AWS A5.11: ENiCrFe-3
- FCAW- Bohler NIBAS 70/20 Mn-FD Classification AWS A5.34: ENiCr3To-4 used on the downhand buttering only
- FCAW- Universal Wire Works 82-T1 Classification AWS A5.34: ENiCr3T1-4 used on the girth butt weld in the 5G position only
To avoid applying a full back purge, the contractor, Langfields, proposed the use of double operator GTAW for the root pass, with the welder working on the exterior of the duct depositing the wire and a second welder inside the duct following the arc with a torch, but not depositing any filler wire. This alleviates any back grinding/gouging and, with the right skill in adding enough filler wire within the joint, prevents hot cracking typically seen in SAW welding when the weld depth/width ratio is insufficient. Langfields have previous experience of using this technique on the Stanlow site and it was acceptable to Essar use this application once again for the final golden weld.
Procedures were finally qualified to ASME Section IX for double operator GTAW/FCAW and double operator GTAW/SMAW. Due to time constraints the FCAW process was qualified in semi-automatic (manually operated) mode. Under the code it is permissible to use semi-automatic techniques to qualify fully automatic welding processes, which enabled the use of mechanized travel via 2 purpose-built Gullco track systems. Additional testing as per ASME IX and client requirements included hardness, charpy impact at -10⁰C in the cap of each welding process and mid thickness, and micro examination of the HAZ and fusion line of the buttering to the duct.
As the method of welding and the equipment was agreed, a program of training and testing was undertaken which involved the following activities:
- Training welders to deposit root passes simultaneously from both sides of the joint using the double operator technique
- Training welders to deposit buttering layers using SMAW in the vertical/overhead positions
- Training welders to use FCAW semi-auto welding process both in buttering and full penetration butt welding in the vertical/overhead positions
- Training welder operators to use FCAW on an automated system (Gullco Pipekat) in the 5G position and 2G positions, this would give us the full positional welding ranges in line with ASME 1X
- Qualifying the welders to ASME IX and the client specification using procedures qualified
Fabrication of new riser and preparation of welding end
Fabrication of the new riser was a complex build undertaken by a European vendor. The image below is of the new hot wall section of the R2150 riser in fabrication stage. The end of the connection to the existing duct has been overlaid using mechanised FCAW in flat position before the weld bevels were ground to shape.
The joint configuration below was developed to allow full access to the root region while minimising volumes of weld metal, Figure 2.
The transfer duct was laser survey checked prior to the cut and beveled by waterjet cut using a specially made track. A hydrogen release treatment had been considered, but was discounted as the duct had not been in hydrogen service and had also been shut down in a controlled manner. The weld overlay application for the buttering runs were applied by the GTAW and FCAW welding processes using PQR numbers 817 and 820. Two purpose-made Gullco mechanised travel units were used for the FCAW, regular adjustment of the torch being needed to follow the varying joint geometry. An overlay of 20mm thickness was deposited, sufficient to manually grind to a finished thickness of between 12mm – 15mm, so that welding of the butt would not affect the heat treated duct end. This was followed by both volumetric and surface NDE then PWHT. Further NDT checks after PWHT where acceptable to ASME B31.3 section 341.3.2. The boundary between nickel alloy and A387 Gr 11 is clearly seen in Fig 2. with the thicker layer being the new riser end and shop deposited overlay.
The replacement riser was then safely lifted into position, installed and trial aligned. Further grinding was needed to bring the root faces into rough alignment, but no further overlay welding was necessary. Final root alignment was not perfect but judged to be manageable. As the root pass was now made between 2 high nickel clad surfaces there was no requirement for preheat enabling the welder to work comfortably inside the duct, and the small volume of shielding gas emanating from the torch was quickly dispersed. On completion, the root pass was DPI checked on both surfaces and minor areas of lack of fusion were found, which were locally excavated and repaired. The inner surface of the joint was then welded to completion using GTAW with 3.2mm diameter filler wire, which overall equaled the joint completion time of SMAW process, largely due to the lack of deslagging and interpass grinding required by SMAW.
The initial passes of FCAW on the outer side of the joint were deposited using a manual torch operation, with 2 welders working on the lower 2/3 and 1 welder working on the top 1/3. Once the more restricted part of the joint had been filled this was changed to mechanised travel using the Gullco tracks, with the torch reconfigured for butt welding. Occasional running repairs were necessary and interruptions were frequent but, considering the conditions, the welding operation went close to plan.
On completion of the weld, NDT was conducted to the requirements of ASME B31.3. The Phased Array (PAUT) method was used, as radiography would need a source such as Cobalt 60 for the 65mm material thickness, and the exposure time and duration would be unsafe and impractical in a shutdown environment working 24hours.
Further NDT checks after PWHT, PAUT, and DPI inspection of the completed joint were acceptable to ASME B31.3 section 341.3.2. With no weld repairs or rework needed.
From conducting research, material appraisal and successful welding trials, a welding methodology was agreed which avoided both the risk of hydrogen cracking in the HAZ and PWHT of the final girth butt weld. This approach gave a significant safety factor to a difficult and highly critical operation and resulted in a useful time saving. Site welding was carried out by 2 rotating 12 hour shifts over 6 days per week, with each shift supervised by the authors. The thorough development process and close adherence to procedure during site welding meant that welding problems were few and far between, with most stoppages being due to extreme weather conditions, site logistics, and restricted and difficult access.
This project was crucial to the success of the refurbishment and the successful conclusion shows the value of a thorough technical approach. It would have been possible to cut corners during development, but this could have led to interruptions on site due to unforeseen problems occurring. At worst, using the ‘conventional’ approach could have led to hydrogen cracking, delaying the shutdown schedule, critical start up, and presenting a real problem in the repair of a susceptible HAZ.
The careful approach was ‘’Worth every Nickel’’
Acknowledgments for the Project
Mr. Gary Bush IEng MWeldI, Essar Inspection Team Leader
Mr. Wayne Griffiths IEng MWeldI, Langfields Operations Director and Lead Welding Engineer
Mr. Marcello Consonni CEng SenMWeldI of TWI Ltd
Mr. Martin Eagle AWeldI, Gullco Managing Director