In-service Duct Welding

Rodrigo CORREIA de Jesus

Welding Specialist in Pipeline and Piping Operation

Vitória - Brazil

rcj@rcjeng.com

Introduction

Successful welding on in‐service pipelines and piping systems requires that two separate vulnerabilities be considered; burning through the pipe wall during welding, and forming hydrogen cracks. Burning through (illustrated in Figure 1) occurs when the inside surface temperature of the pipe reaches temperatures at which the pipe wall has little remaining strength. This situation typically occurs when the pipe wall is relatively thin and the heat input exceeds the ability of the pipe wall and the pressurized fluid to quickly conduct the heat away from the weld.

Hydrogen cracking (illustrated in Figure 2) can occur either during the welding process or some time after welding is completed. It most often occurs in hard heat affected zones, but it can also occur in weld metal deposits. It requires the simultaneous presence of a susceptible microstructure (martensite), and sufficient amounts of both stress and hydrogen in the weld zone. Avoiding hydrogen cracking involves minimizing one or more of those three factors. For example a low hydrogen welding process is typically preferred over the use of cellulosic flux covered SMAW electrodes and heat inputs are selected to produce relatively slow weld cooling rates that prevent the formation of martensite.

Problems with welding on in‐service piping typically occur as a result of various myths and misconceptions about how to avoid burning through and hydrogen cracking. Some of those misconceptions are addressed below.

A Hardness of HV 350 is Not the Magic Number to Avoid Cracking

Traditionally, it has been common to specify that hardness of the heat affected zone not exceed HV 350 when qualifying welding procedures for use on in‐service piping. CSA Z662 states that for fillet welds and branch connections made onto piping containing flowing fluid, “…weld metal or HAZ hardness values in excess of 350 HV shall require an valuation of the welding procedure specifications to determine that they are suitable for the avoidance of hydrogen‐induced cracking”. However, it is not the hardness, but rather the presence of a crack susceptible microstructure that influences the likelihood of cracking. Hardness limits are merely used as indicators of detrimental amounts of martensite.

The fact that is seldom recognized is that the hardness associated with martensite is different for steels of various compositions that are commonly found in pipelines, and, the measured hardness will be different depending upon how much of the martensite is present. Furthermore, the amount of martensite that represents a significant susceptibility to ydrogen cracking is dependent upon the amount of hydrogen present in the weld zone. Weld zones with lower weld metal hydrogen contents can tolerate more martensite without cracking.

Another factor that affects crack susceptibility is the steel thickness. After a weld solidifies, it can be supersaturated with hydrogen. Hydrogen then diffuses to free surfaces, where it can escape. When the material being welded is thick, diffusion distances are greater, which results in more hydrogen remaining in the weld. Thickness also affects restraint levels with thicker materials developing higher levels of residual stress.

Therefore, the specified maximum hardness limit needs to consider the composition of the steel being welded, the thickness of the steel, as well as the likely amount of hydrogen in the weld zone [ref9]. As illustrated in Figure 3, a hardness of HV 350 can be non‐conservative for low carbon equivalent steels thicker than 0.375 in (9.5 mm). On the other hand, for relatively high carbon equivalent steel, for example higher than 0.42 CE IIW, a hardness higher than HV 350 can be tolerated when low hydrogen weld practices are used that produce hydrogen contents no higher than 4 ml/100 gm deposited weld metal. For thinner steel, the hardness limits can also be higher, as illustrated in Figure 4.

Conclusions

Welds of good structural integrity can be made safely onto in‐service pipelines and piping systems by selecting welding procedures that balance the susceptibility to hydrogen cracking with the susceptibility to burning through. Adjusting heat inputs to maintain acceptably low inside surface temperatures while still minimizing the likelihood of forming martensite and using low hydrogen welding practices are keys to success.

Table 1 Effect of Heat Input on Inside Surface Temperature and Critical Cooling Rate of Fillet Weld, Calculated Using Model from Ref.4

References

1. Howden, D. G., "Welding on Pressurized Pipeline," Loss Prevention, Vol. 9 (New York, NY: American Institute of Chemical Engineers, 1975), pp. 8‐10.

2. Wade, J. B., "Hot Tapping of Pipelines," Australian Welding Research Association Symposium, Paper No. 14 (Melbourne, Australia, 1973).

3. Cassie, B. A., "The Welding of Hot Tap Connections to High Pressure Gas Pipelines," J. W. Jones Memorial Lecture (Pipe Line Industries Guild, October 1974).

4. J. F. Kiefner, R. D. Fischer, and H. W. Mishler, "Development of Guidelines for Repair and Hot Tap Welding on Pressurized Pipelines," Final Report, Phase 1, to Repair and Hot Tap Welding Group, Battelle Columbus Division, Columbus, OH, September 1981.

5. Bruce, W. A., Li, V., Citterberg, R., Wang, Y.‐Y., and Chen, Y., "Improved Cooling Rate Model for Welding on In‐Service Pipelines," PRCI Contract No. PR‐185‐9633, EWI Project No. 42508CAP, Edison Welding Institute, Columbus, OH, July 2001.