Brief description of gas pipeline stress-corrosion

Analysis of accident statistics at main transportation pipelines in Russia shows that more than half of accidents are caused by stress-corrosion defects. This type of pipeline wall destruction is also called stress-corrosion cracking. Stress-corrosion cracking defects substantially differ from defects related to metal losses, such as corrosion pits and pitting corrosion. Metal-loss defects are characterized with relatively smooth edges, when the stress-corrosion cracking defect represents a colony of unexpanded cracks with the depth of 5-20% of pipe wall thickness.

Stress-corrosion cracking appears on pipe outer surface. The cracks are oriented perpendicular to the direction of maximum tensile stress, i.e. mostly along the pipe axis. At some instant the net of surface cracks unites into one main crack. After that, the main crack elongates very fast, which causes pipeline rupture.

Major factors influencing corrosion cracking process are: tension intensity; insulating coating condition; composition and structure of alloy the pipes are made of; cathodic protection condition at given pipeline section; forming technology and quality at pipes fabrication that influence the level of residual stress in weld-affected zone; type of surrounding soil. Stress-corrosion cracking is common to main transportation pipelines.

Stress-corrosion cracking in oil pipeline walls appears less often. Formation and rate of propagation of stress-corrosion cracks in colony is influenced with a combination of many factors. It is almost impossible to predict the accidents due to this process without in-line inspection.

Main purpose of in-line inspection is the determination of safe operation life of pipeline separate sections with all the defects existing at that section: constructional defects, metal loss defects, and stress-corrosion defects.

Literature analysis shows that during the last decade the number of accidents at main transportation pipelines stipulated with stress-corrosion exceeds the number of accidents due to metal loss defects. At the same time the number of records on stress-corrosion cracking defects in in-line inspection reports is hundred times less than the number of records on metal loss defects. This evidences for a very low degree of stress-corrosion defects detectability in the initial phase of development, when the local colony of surface cracks hasn’t yet transformed into a main crack.

Increase of reliability of detection of dangerous defects of stress-corrosion cracking and pipes external corrosion during regular in-line inspection leads to optimization (decrease) of scope of gas pipeline overhauls maintaining accident risk at the existing level or decreasing it.

Known methods of in-line inspection for stress-corrosion detection

At the present time three methods of stress-corrosion cracking defects detection during in-line inspection of steel pipelines are available:

1) TFI-method (transverse field inspection method) - method of registration of magnetic flux leakage at transverse magnetization of pipe section. At transverse magnetization magnetic flux leakage at longitudinal cracks is more evident. This method implements magnetic PIGs moving inside the pipe in gaseous or liquid medium.

2) CD-method. This method consists in total ultrasonic scanning of pipe wall at excitation of ultrasonic wave by means of piezoelectric transducers. In this method ultrasonic waves emitted by onboard piezoelectric transducers are transmitted into pipeline wall though the layer of fluid surrounding PIG at certain angle to pipeline wall internal surface. Operation due to this method requires an ultrasonic flaw detector moving in fluid. Herewith, for gas pipeline inspection the so-called floating water plug is created by means of additional batching pigs which located upstream and downstream of flaw detector.

3) EMAT-method. This method consists in excitation of ultrasonic waves directly in pipe wall by means of non-contact electro-magnetic acoustic transducers located at inline EMAT flaw meter. This flaw meter is able to move in gaseous medium, so this method is also known as dry ultrasound method. However, EMATs are also good for liquid medium.

Fundamental drawback of transverse field inspection method

First two methods are characterized with the following: TFI – transverse field inspection method does not reveal shallow (up to 15% of pipeline wall thickness) stress-corrosion cracks, because transducers sensible to change of magnetic field during in-line inspection are by definition inside the pipe, and stress-corrosion cracks are always formed on the pipeline outer surface.

If transducers moved along the surface covered with unexpanded cracks (like at inspection of dug-up pipe with outer scanners), there would be a substantial magnetic field intensity gradient at crack-like defects. However, at in-line inspection there is a metal layer (which thickness is almost equal to nominal thickness of pipeline wall) between transducers recording rejection of magnetic field in defect area and a defect in the form of superficial unexpanded shallow crack.

Under these conditions magnetic field variations recorded with transducers of in-line flaw meter have rather fuzzy nature. As far as magnetic signal amplitude drops proportionally to square of distance between transducer and crack bottom, the magnetic signal from external crack with the depth of 10-15% is comparable in amplitude with the noise level recorded with these transducers due to pipeline inner surface roughness. Herewith the magnitude of new pipeline inner surface roughness amounts about 0.1 mm and increases up to 0.3 mm and higher in pipelines being in operation for several years.

Due to this fundamental drawback the TFI method (as any other method based on the effect of magnetic field rejection/expulsion in defect area) deceive expectations set on it in terms of monitoring of stress-corrosion cracking of gas main transportation pipelines.

TFI-method is suitable only for detection of deep longitudinal cracks (more than 20% of pipe wall thickness) and metal loss defects. However, when the local cracks in stress-corrosion cracking zone reach 20% of wall thickness, the process of local cracks net transformation into main crack can be very fast. Time interval to next in-line inspection may be insufficient for accident prevention.

At the same time too often in-line inspection may become an intolerable burden for organizations running gas pipelines. Thus, for stress-corrosion cracking monitoring it is required to detect reliably all the cracks (individual and in colonies) with the depth of 10% and even less.

Economical unprofitability of inspection in water plug

CD-method of ultrasonic examination in floating water plug turned out to be economically unprofitable due to losses related to taking a gas pipeline section out of service for a long period of time for drying after in-line inspection with this method, though the CD-method is sufficiently reliable in terms of detection of crack-like defects, including longitudinal shallow cracks in outer surface.

In gaseous medium this method is not applicable due to substantially large coefficient of ultrasonic wave reflection from gas-metal interface as compared to oil-metal or water-metal interface. This means that it is impossible to introduce an ultrasonic wave from piezoelectric transducer into gas pipeline metal wall and receive a signal reflected from external defect due to physical constraints related to laws of ultrasonic wave propagation and reflection. So, for inspection of main transportation pipelines with this method it is necessary to create a floating water plug and then to dry the gas pipeline.

EMAT - dry ultrasound method

Due to drawbacks of TFI and CD-methods the EMAT-method (dry ultrasound method) at its use for in-line flaw meters is the most promising direction of reliability increase in detection and grading of gas pipelines stress-corrosion cracking.

As a result of intensive development of electronics during last decades there were formed favourable conditions for this method application in in-line inspection. Presently the leading companies developing in-line EMAT-flaw meters (PII, ROSEN, Spetsneftegaz and some others) have already produced several generations of these in-line gauges.

It must be mentioned that EMAT-method is the general name for a variety of technologies and methods of excitation and recording of different types of ultrasonic waves. In other words, there exists a variety of EMAT designs, which, unlike piezoelectric transducers, are able to emit different types of ultrasonic waves (longitudinal, shear), as well as their combinations. Some of these waves are better for in-line inspection than the others.

For example, pipeline construction workmanship differs greatly in Russia and in Europe, especially when comparing pipelines built or repaired in the middle of 1990s. At that time Russia experienced lack of means for these works and quality of work execution dropped sharply. Generally accepted fact is that Russian gas pipelines have more construction defects and geometry imperfections (dents, edge mismatches, burr, electrode residues etc.).

These nuances are essential for choice of design of EMA-transducers for inspection of Russian gas pipelines, as far as they are exert decisive influence on total reliability and validity of EMAT-method. In order to understand why EMAT-method depends more than the other magnetic methods (TFI, MFL) on geometry defects we should take a closer look at features of electro-acoustic method of ultrasonic wave excitation in pipe wall and at examples of EMAT designs.

Electromagnetic acoustic transducers review

We turn our attention only to types of waves and EMA-transducers, which presently are most common in domestic and foreign practice of in-line inspection.

First of all it should be noticed that electromagnetic acoustic transducers can be of magnetostrictive or Lorentz type. In magnetostrictive transducers ultrasonic waves are excited at the expense of local compression or elongation deformations that occur in metals under cyclic variation of magnetic field intensity inside metal.

Herewith, static magnetic field lines (as opposed to cyclically varying magnetic field generated with surface eddy currents) in magnetostrictive transducers are always parallel to surface.

It is noteworthy that the amplitude of ultrasonic oscillations in metal caused by magnetostrictive effect depends heavily not only on intensity of magnetic field penetrating metal to certain depth, but also on metal electrical conductivity, as well as on its magnetic permeability. All these characteristics significantly vary from one steel grade to another and even from one end of given pipe 12 m long to its other end.

Besides, ultrasonic oscillations amplitude is affected with metal residual magnetization, which remains in pipeline walls for a long time after pigging with magnetic inspection or cleaning pigs. Thus, application of magnetostrictive EMATs in in-line inspection is presently limited.

Lorentz EMATs

In Lorentz EMATs ultrasonic waves are excited at the expense of Lorentz force, which influences free electrons in metals, when surface eddy currents are within the static magnetic field. Moreover, magnetic field must be perpendicular to eddy current direction. Thus, the direction of surface eddy current, magnetic field lines and Lorentz force are mutually perpendicular and form a right-hand triple. It means that Lorentz force direction can always be determined according to right-hand thumb rule.

Eddy currents in metal superficial layer are excited at current cyclic variation in conductor of specific shape (so called RF-coil) located directly over the metal surface. Herewith the current in RF-coil and the surface eddy current are always oppositely directed.

Frequencies of EMATs are within the range from 100 kHz to several MHz. At frequency of 200 kHz an ultrasonic wave length in steel is about 10 mm.

Lorentz transducers are designed in such a way that eddy currents flow within the area of strong static magnetic field. Moreover, in in-line inspection a static magnetic field is generated with permanent magnets. In other industrial applications of EMAT-method without these stiff restrictions on electric power consumption strong static magnetic field, as a rule, is generated by means of electromagnets.

Major advantage of Lorentz EMATs is in their lesser dependence on magnetic permeability and electrical conductivity of pipe metal as against magnetostrictive EMATs. Besides, in Lorentz EMAT the amplitude of excited ultrasonic wave is almost independent on metal surface condition and on residual magnetization of given pipeline section.

So, presently in-line inspection generally implements Lorentz EMATs, because their work is more predictable in pipelines with variable thickness, constructed of pipes made of different steel grades. It should be noted that the magnetostrictive effects are present in Lorentz EMATs as well. Usually these effects are the parasitic ones.

However, e.g. ROSEN Company utilizes EMA-transducers in which ultrasonic waves with different direction and polarization are generated simultaneously at the expense both of Lorentz effects and magnetostrictive effects.

Horizontally and vertically polarized shear waves

It is also noteworthy that shear ultrasonic waves of two different polarizations are used in EMAT-method of in-line inspection. Shear waves can propagate only in solids, e.g. in metals. In shear waves crystal lattice atom oscillations are perpendicular to direction of wave propagation. This is the difference between shear waves and longitudinal compression/decompression waves, which may propagate not only in solids, but also in gases and liquids.

In longitudinal waves the oscillations of substance atoms or molecules occur in the direction of wave propagation. At propagation of shear wave in a plate or pipe the atoms may oscillate in the direction perpendicular to surface – this is the case of vertical polarization. When atoms oscillate parallel to surface shear wave is polarized vertically.

In in-line inspection an ultrasonic wave with vertical polarization is denoted as SV (Shear Vertical), and a wave with horizontal polarization is denoted with SH (Shear Horizontal). Denominations "vertical" and "horizontal" polarizations are used since the times when shear waves investigation was performed in labs on metal plates lying on the table. In that case the metal atoms actually moved vertically at propagation of shear wave with vertical polarization, and in horizontally polarized wave atom oscillations were horizontal.

These denominations lost their original sense for the cases when wave propagation in pipeline walls is considered. For pipes it would make more sense to denote vertically polarized shear wave as normal (or perpendicular) shear wave, and denote horizontally polarized wave as shear wave with circumferential (longitudinal) polarization. However, let’s adhere to usual terminology and use denominations SV and SH.

Advantages and drawbacks of SH and SV shear waves

It should be noted that for detection of longitudinal stress-corrosion cracks the SH and SV waves should propagate in circumferential (not in axial) direction along the pipe wall. At the present time both vertically (SV) and horizontally (SH) polarized ultrasonic waves are used in in-line inspections. Notably, there are no crucial advantages for one polarization over another. So, the number of devotees of SH and SV waves is almost the same.

ROSEN Company and Russian company Spetsneftegaz traditionally use SV-wave, and PII Company (General Electric division) and Russian company TRISTAN use SH-waves.

SH-waves have smaller dispersion, i.e. this wave front is not diffused with offset or after reflection from the defect. Velocity of SH wave propagation is almost independent on pipe wall thickness. Due to this the algorithm of signal processing for detection and sizing of defects in pipeline wall using SH-waves is very simple, which increases reliability of detection and recognition of defects.

Main drawback of EMATs design with horizontal polarization is the necessity of placing the RF-coil (through which pass pulse currents up to 100 Ampere) in a narrow clearance between powerful permanent magnet and pipeline wall. Moreover, this clearance should be maintained with relatively high accuracy and must not exceed several millimetres.

It is very difficult to maintain this clearance at in-line flaw meter velocity of about 2 m/s and magnet attraction to pipeline wall with magnitude of thousands of Newtons. RF-coil may be easily damaged when the magnet passes over circumferential weld or over geometry defect of pipeline inner surface (dents, edge mismatches, burr, electrode residues etc.).

Selection of special wear resistant protective films for RF-coil and creation of mechanical systems of clearance automatic maintenance are basic directions of improvement of Lorentz EMATs with horizontally polarized SH-wave.

PII Company (incorporated in GE) made some progress in inspection of gas pipelines in the USA and Europe by improving this design Lorentz EMATs with SH wave since 2003. Operational capability of this design for typical Russian gas pipelines may require verification with tests on running gas pipelines.

In Lorentz EMATs with vertically polarized shear SV-wave the magnetic field lines are parallel to surface. Thus, RF-coils inLorentz EMAT-flaw meters are placed between poles of permanent magnets or between metal brushes though which the magnetic field is introduced in pipeline wall. So the emitting RF-coil in these EMATs is not jammed in a narrow clearance between the magnet and pipeline wall, it is instead pressed down to inner surface with some minimum force.

This radically increases flaw meter lifespan and reliability of operation. Besides, general arrangement of such EMAT-flaw meter with ultrasonic SV wave is very similar to arrangement of magnetic flaw meter with longitudinal or transverse magnetization of wall (to MFL or TFI-flaw meter). The difference is that RF-coils emitting and receiving ultrasonic waves propagating in pipeline wall are placed between the magnet poles instead of Hall magnetic sensors or magnetoresistive sensors.

Pipe wall magnetization is performed by means of permanent magnets. Wall magnetization to saturation in EMAT-method is not required.

ROSEN Company for magnetization implements the magnetic pieces with floating suspension, and Spetsneftegaz uses magnets with metal brushes. During last fifteen years of magnetic flaw meters improvement both designs showed good results at Russian and foreign gas pipelines and oil pipelines in terms of operation reliability.

Basic drawback of EMAT-flaw meters with vertically polarized shear wave is the dependence of SV wave propagation velocity on pipeline wall thickness (ultrasonic wave dispersion). This complicates signal processing algorithm and interpretation of signals reflected from defects.

Monitoring of stress-corrosion cracking in gas pipelines

At the present time at straight runs of gas pipelines in-line EMAT-flaw meters detect unexpanded longitudinal cracks with lateral area of 40-100 mm2 at crack depth of 1-2 mm. This sensitivity is already sufficient for monitoring of stress-corrosion cracking of main transportation gas pipelines in the initial phase of development of stress-corrosion cracking defects.

Moreover, horizontally polarized SH waves are slightly less sensitive to shallow cracks (up to 15%-20% of wall thickness) as compared to vertically polarized SV1 waves also known as symmetrical Lamb waves.

Fast introduction of EMAT-flaw meters for regular inspection of gas pipelines is inhibited with their higher cost as compared to magnetic-field flaw detectors, as well as by insufficient range (reserve) of offline operation.

Opportunities of EMAT-method

The fact is that the power consumed by EMA-transducers is at least by an order of magnitude greater than the power consumed by piezoelectric transducers in ultrasonic CD or WM-flaw meters for the same amplitude of ultrasonic wave in pipeline wall. In comparison with magnetic TFI or MFL-flaw meters general power consumption by EMAT-flaw meter may be greater by two orders of magnitude.

Thus, to provide sufficient stroke reserve and inspection of linear runs of main transportation gas pipelines with the length up to 150 - 200 km it is necessary either to increase batteries number in EMAT-flaw meters (which is related to gain in weight and, consequently, decrease of piggability and reliability), or to decrease the number of onboard EMATs (which leads to inspection quality drop).

In this connection for further improvement of EMAT-flaw meters a task of power supply directly onboard the PIG is vital, because gas is supplied through main transportation pipeline by means of centrifugal-flow blowers with drive power of tens of megawatts. This means that in any gas pipeline the potential energy flux density of compressed gas is very high.

It is suffice to learn how to transform merely a basis point of this energy flux into electric energy onboard the flaw meter and the stroke reserve problem will be resolved without any loss of pipeline walls scanning quality.

Development of such technique with due account for compliance with all requirements to explosion safety and reliability is distinctively possible considering ultra-modern power electronics and electric drives. However, it will require certain capital investments.

Thus, specific programs and project financing on improvement of MAT-flaw meters in order to improve validity of detection of stress-corrosion cracking in gas pipelines of Russia UGSS net will be required. This will lead to improvement of in-line inspection quality and, as a consequence, to more rational planning of overhauls of gas pipeline sections along with reduction of general expenditures and gradual decline of accident risk.