Weld Repair Of A Petrochem Component

Weld Repair of a Petrochem Component

OMMI (Vol. 1, Issue 3) December 2002

Weld Repair of a Petrochemical Component P. Veron, EQUIPOS NUCLEARES S.A. Maliaño (Cantabria), Spain

Dr.-Ing. Pedro Veron Head of Reliability and Corrosion Welding Metallurgist Professor of the University of Cantabria Spain.

Abstract In this paper we present an experience in the repair of two identical pressure vessels in a petrochemical factory with service induced flaws. First step was to determine the cause of the observed cracks. Next, Fracture Mechanics assessment proving that the vessel was able to work safely with the fissures. Service was soon restarted, avoiding a long outage and a rush repair. A successful weld repair was conducted several years later. No crack growth was observed in the intermediate outages. The vessels have been working for many years without any trouble.

2.

Component and problem description

In 1986 a refinery found several cracks during an in-service inspection in a scheduled shut down after first five years of service. The fissures were found at a nozzle location. The two affected components were two identical heat exchangers (stainless steel weld overlaid ASTM A182 F22). The cracks were linked to screwed holes and extended deep into the overlay (Fig. 1 to 3).

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

CRACK

Fig. 1 Nozzle design (left) and nozzle with crack (right)

Fig. 2 Dye check indication observed in ISI corresponding to1984 shut down

Fig. 3 Dye check indication close up.

2

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

3

The vessels designer had specified this 38mm thick weld overlay (Fig. 1) to secure an internal structure in screwed holes machined there (Fig. 4).

Fig. 4 Initial design: thick weld overlay (marked with an arrow) and internal structure secured in it. Changes: Scratched area of overlay removed by machining and structure held in other place (not shown) So, to get rid of the cracks, this inner structure was modified to fit on another area of the vessel and the thick weld overlay machined to a thinner (~9mm) one (Fig. 5). But, unfortunately, the cracks did not disappear after this machining (Figs. 6 and 7).

Fig. 5 Nozzle with machined weld overlay

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

4

Fig. 6 Dye check indications after overlay machining

Fig. 7 Dye check indication after overlay machining. Close up.

The ultrasonic examination on the machined surfaces showed that the cracks extended deep into the overlay (Fig. 8). As a consequence, a weld repair without post weld heat treatment (PWHT) was discarded because the low alloy steel of the nozzle would be affected.

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

5

Fig. 8 Ultrasonic examination after overlay machining

3.

Integrity assessment

At this stage the refinery contacted us, concerned by the safety issue of the component with cracks in the remaining weld overlay. In our opinion the cracking was due to low cycle fatigue caused by high thermal stresses induced by a 38 mm thick stainless steel weld overlay acting on stress risers (machined holes). A fracture mechanics assessment was performed by us following the procedure of reference [1]. Conservatively, the crack tips were supposed to be in the low alloy steel. The transients considered were start-up, steady state and shut down.

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

6

Fig. 9 Start-up transient. Cool down was the reverse one.

Fig. 9 Start-up transient (Cool down transient was the reverse one) Finally, the critical crack depth was found to be 16mm, larger than the remaining cracks depths (less than 9mm). In addition there were other considerations that reassured us about the capability of the components to withstand service safely without any weld repair. These were: •

The cracking mechanism was no longer present, and,

•

The remaining cracks were shallower and would be under lower stresses. On the other hand, a weld repair might introduce new flaws (reheat cracking, distortions), especially when performed under time pressing conditions.

So, under these considerations, the components restarted operation as it was soon afterwards in 1986, without any weld repair. 4.

Interim service experience

During the 1988 and 1991 shut downs the ISI revealed neither crack growth nor any new cracking. In 1991 shut down, because of the refinery strong will, a weld repair was performed successfully by EQUIPOS NUCLEARES S.A.The weld repair procedure, defined by ENSA, was based on the temper bead technique and

Weld repair of a petrochemical component OMMI (Vol.1, Issue 3), December 2002

7

included a local PWHT at 690ºC because of the working conditions of the vessels (temperature and partial pressure of hydrogen). For that purpose the remainder of the thick overlays was machined off. After that no crack was present, as revealed by dye check, confirming that they were confined to the stainless steel weld overlay. Thus the weld repair and PWHT were successfully performed. The component has been working since 1991 without any incidence.

5.

Summary

The aim of this paper is to present how an actual industrial case of damaged pressure vessel was successfully resolved.

References [1] Adams N J I. & Welland G G, “The Practical Application of Fracture Mechanics to Hydro-cracker Reactor Vessel”, 5th Int’l. Conference on P.V.T., San Francisco, USA, 1984,Volume II, p.p. 777—793.