Life Assessment of Reformer Tubes from Strain Measurement

The life of reformer tubes is controlled by creep. Creep is a deformation process which results in an increase in tube diameter, driven principally by the internal pressure. Inspection techniques that measure diameter, directly follow the progress of creep damage. However, the rate at which creep damage accumulates is not linear and is dependent on a number of complex factors including the large changes in creep properties as the material ages, the variable stress distribution through the wall of the tube and the uncertainties associated with actual metal operating temperature. This paper describes the development of a life assessment methodology to calculate the level of damage that has occurred up to the time of the inspection and  to predict remaining life under any chosen future operating regime directly from tube diameter i.e. strain data.

At the heart of the methodology is a specifically developed model that describes the rate at which creep damage accumulates and relates that damage to strain and strain rate. There are a number of life assessment methodologies based on assessment of strain data. The Omega technique has found recent widespread acceptance for creep life assessment and is based on a description of the strain behaviour of the material under consideration. This technique was an obvious starting point for modelling reformer tube creep data. However it was found that the creep curves, particularly for as-cast tube material, were not well described by the Omega technique. An alternative model relating strain and strain rate to the development of creep voids has been developed. A comprehensive database of tube material creep properties has been developed for input to the model. This data base has been created from reformer tube material that has been pre-aged to a range of metallurgical conditions. Parameters specific to the material and its metallurgical condition are obtained by fitting the creep curves obtained from this test program to the model.

The methodology is completed by detailed finite element analysis that takes into account the temperature and hence strain through the wall of the tube and how this temperature gradient varies along the length. The influence of plant cycling caused by start up- shut down and trips is taken into account. Of particular note is the fact that the model has little dependence on measured tube metal temperature. In fact, temperature is an output from the model calculated from the knowledge of strain (inspection data) and creep behaviour.