Damage analysis in Fe–Cr–Ni centrifugally cast alloy tubes for reforming furnaces

Highlights

• Damage during service of HP grade tubes, modified with Nb and Ti addition.

• Laser-Optic Tube Inspection System (LOTIS) measurements performed in situ.

• Softening effects at high temperature are highlighted by tensile and creep tests.

• Tubes are put out of service when their deformation become not acceptable.

Abstract

Reformer furnaces tubes work under high temperature and pressure for a long time, which are very critical conditions for creep deformation and life of most common materials. Cast austenitic Fe–Cr–Ni alloys in the widely know HP grades are used for reformer tubes to allow a good service at temperatures that can be close to 1000 °C. This paper reports a study devoted to the damage analysis of reformer furnace tubes after more than 100,000 h of service. Tubes, made of a HP grade modified with Nb and Ti additions, were inspected in situ by a laser optic system to measure their internal diameter and evaluate creep deformation. With the aim of developing a criterion for deciding the substitution of components, samples of as cast material and samples, cut from the most deformed tubes put out of service, were considered to check changes of mechanical properties and metallurgical characteristics. Tensile and creep tests were carried out; moreover the metallographic observations included optical and scanning electron microscopy and energy dispersive X-ray microanalysis in order to measure locally the chemical composition.

Introduction

Reformer furnaces are widely used in the petrochemical industry to obtain hydrogen from hydrocarbon. This production takes place in radiant tubes, containing a catalyst, through an endothermic reaction between hydrocarbon (mostly methane) and water vapour.

Radiant tubes are the most critical components of a reformer furnace, as they are exposed to severe conditions for long time during service. These tubes, which have normally inside diameter of 60–200 mm, wall thickness in the range 10–25 mm and length of 10–15 m, are designed for a nominal life of 100,000 h in service, for temperatures up to 980 °C and internal pressures of 10–40 bar [1].

Due to the high service temperature, heat resistant alloy with good mechanical properties and corrosion resistance are required. On the basis of these considerations, during the last years new alloys and manufacturing processes have been developed to meet the severe requirements imposed on the radiant tubes [2],

Since reformer furnaces need to operate reliably and without unplanned shutdowns, tube materials have been improved over the last 50 years [3]. Formerly, in the 1960s and 1970s, the alloys mostly used was the 25Cr–20Ni–0.4C–Fe, designated as HK-40; during the following decades, the HP-40 grade (25Cr–35Ni–0.4C–Fe) has become common because of its better mechanical properties at high operating temperature [4]. The high Cr–Ni alloy required for reformer tubes are not easily drawn or extruded, thus cast structures are produced. Centrifugal casting is now generally used, giving a more even microstructure, with grains oriented in radial direction that provides greater strength and better creep resistance.

The additions of Cr and Ni are well known for improving corrosion resistance and creep strength at elevated temperature. Reformer tubes fail by creep through the formation of creep voids during service: many researchers showed that the occurrence of these voids may be related to the microstructural features [5]. In order to improve creep behaviour, the austenitic matrix of HP alloys is strengthened by a dispersion of hard deformation-resistant carbide particles. So high temperature strength of this alloy depends on the stability of Cr carbides which undergo thermal effects during long service time, changing their morphology and in general tending to particles coarsening in order to minimize surface energy [6].

Starting from the mid 1990s to the present, the HP-40 grade has been microalloyed with Nb and, more recently, with Ti and Y with the aim of enhancing creep strength. As pointed out in literature, many authors experimented that a fine dispersion of carbide particles can be stabilized by modifying the HP grade composition adding stabilizing elements, such as Nb around 1% [7], Ti up to approximately 0.8% [8] or Yttrium about 0.3% [9].

Creep properties depend on material microstructure and it is well known that tensile and creep strength decrease with the increase of tubes lifetime [10]; metallurgical features, as carbides type and distribution, change correspondingly [11]. Unfortunately investigations on these phenomena require extracting some tubes during planned maintenance shutdown of plant and then cutting samples from them. The cost of each tube is very high (some tens of thousands of Euros), moreover the condition of a sample tube may or may not be a representative of the total number of tubes in the furnace, so it is worth to develop methods that allow to perform measurements on a large number of tubes, possibly without extract them from plant.

As quoted in [12], conventional non-destructive testing techniques, such as eddy current and ultrasonic measurements, applied to reformer tubes detect creep damage in the form of internal cracking and are useful during the last stage of service in reformer furnaces; degradation phenomena of material, as carburization and oxidation, can be inspected observing ferromagnetic layers on tubes surfaces; measurements of magnetic permeability and low magnetic fields can give satisfactory information on effective tube wall thickness, but cannot indicate creep failures unless micro-cracks appear.

A comprehensive approach should include reformer tube inspection, remaining life assessment and identifying operational problems [13]: internal and external inspection methods assess periodically tube conditions at various stage of creep deformation and identify problems with burners or local hot spots due to gas flow conditions, evaluating also the effect of high-temperature corrosion.

However reliable criterion describing the amount of degradation in creep resistant reformer tubes during application has not been established so far [14]: experiments to predict residual lifetime of creep-resistant tubes and pipelines should be based on investigations performed during service, but up today they require the pipe decommissioning to perform metallographic examinations and mechanical properties tests.

It should be also underlined that the assessment of tube residual life is affected by several uncertainty factors:

• The actual value of operating temperature (temperature is measured by means of optical pyrometers with a measurement error of ±20 °C).

• Reliable values of material creep properties (long-term properties for such alloys are only known by manufacturer catalogues).

• The actual stress state acting on the tubes during service: for such components thermal stresses during start-up and shut-down may be significant and lead to a creep-fatigue interaction damage mechanism [15].

Therefore reliable means of inspection are required for a safety operation of such components.

In the last decade reforming plants have relied on the quality inspection data of tube diameter produced in situ by the Laser Optic Tube Inspection System (LOTIS) [16]. Diameter measurements of tubes deformation describe quantitatively creep damage. Information could be stored in a database to be utilized by a life-assessment program, which calculates the probability of crack initiation and tube failure. Nevertheless this technique is still in progress, due to the complexity of factors influencing the life of catalytic tubes, and currently various methods, based on other types of piping inspection, are experimented by plant operators in order to develop a procedure for failure forecast and prevention [17], [18].

On the basis of this introduction, our work is addressed to the inspection of reformer furnace tubes and evaluation of the most deformed tubes residual life, after long time service at 900–950 °C. Tubes are made of the modified HP steel grade with Nb (1.5% in weight). The tubes under investigation were inspected in situ by the LOTIS technique and put out of service as their deformation was judged not acceptable. With the aim of developing a criterion for deciding the substitution of components, samples for mechanical and metallurgical experiments were cut from the decommissioned tubes. Mechanical properties were evaluated by tensile and creep tests. Creep data are collected and analysed in the form of Larson–Miller diagram in order to predict the creep residual life of deformed tubes. Metallographic samples were observed by optical and by scanning electron microscopy; energy dispersive X-ray microanalysis were performed to check locally the chemical composition. The experimental results are put in comparison with the ones obtained on samples in the as cast conditions of the same HP grade.