IN-LINE INSPECTION-CONCLUSION ELF'S 20-YEAR EXPERIENCE CONFIRMS EFFECTIVENESS OF SMART PIGS

Marcel Roche, jean Pierre Samaran
Societe Nationale Elf Aquitaine (Production)
Pau, France

Societe Nationale Elf Aquitaine has acquired more than 20 years' experience with in-line inspection of pipelines, using tools from seven service companies.

MAGNETIC FLUX LEAKAGE

The first in-line inspection operation carried out by one of the companies affiliated with Elf Aquitaine occurred in 1971 and used a Linalog tool. This inspection was to evaluate the extent of external corrosion attacks at the bottom of a 24-in. OD, 69-km uncoated pipeline laid directly on the ground in a desert area.

Installed in 1969, this line looped the 24-in. OD, 778-km (483 mile) oil pipeline from Zarzatine (Algeria) to La Skhirra (Tunisia). The results given by inspection were verified by spot external measurements and considered good.

Later periodic inspections, performed in 1980 and 1987 on the main pipeline, detected important external corrosions in some salty areas. These attacks occurred beneath disbondments of the over the ditch-applied bituminous enamel coating.

Inspections also detected internal corrosion produced by accumulation of water at low points as a result of intermittent reductions of flow velocity and even to some flow stoppages.

Linalog tools were also used five times from 1974 to 1980 for in-line inspection of a hot fuel-oil 10-in. OD, 15-km pipeline between the Feyzin refinery and the Chasse-Sur-Rhone thermal power station in France.

Pushed by water at a velocity of 2-3 mps, the pig gave useful indications on external corrosions underneath the polyurethane foam heat-insulating coating directly applied on the steel surface. These operations permitted this pipe to remain in service after repairs which were limited to the most corroded areas.

The same kind of operation failed (unreliable results) when the Corrocontrole tool of Pipeline Service was used to inspect a 10-in. OD, 60-km, fuel-oil pipeline between Donges refinery and Vern-sur-Seiche, as well as a 14-in. OD, 2.8-km coastal pipeline at Port-La-Nouvelle.

The 30-in. OD, 203-km (126 mile) pipeline transporting oil from Piper to Flotta terminal in the North Sea (operated by Elf Enterprise Caledonia, formerly Occidental Petroleum) was inspected by Linalog tools in 1977 and 1979. These tools were replaced by the TransCanada tool in 1984 and by the British Gas tool in 1990.

This latest inspection revealed serious internal corrosion pits which led to a partial replacement of the pipe. British Gas' tool also replaced the Linalog from 1988 for inspecting the two 32-in. OD, 362-km (225 mile) pipelines transporting Frigg gas to St. Fergus (operated by Total Oil Marine).

British Gas' tools were considered more efficient, differentiating, for instance, between internal and external defects by the use of a secondary row of sensors situated outside the magnets. This secondary row gives information only on the internal side of the pipe wall. The cost of operation is, however, much higher than with the other tools.

A British Gas tool was also used in 1986 to inspect a 26-in. OD gas line in the Frigg field. No significant corrosion was found, but external Inconel 625 overlays were situated inside tunnels at the base of a concrete platform.

The Linalog tool was also used in 1982 to carry out a zero point check of the 20-in. OD, 65-km Yanga-Djeno pipeline offshore Congo before it was placed in service.

This kind of reference inspection is now considered unnecessary because the tools are permanently improved and give more and more quantitative results.

IN GABON, CONGO

Shore and offshore oil pipelines have been inspected in Gabon and Congo since 1988 with various H. Rosen Engineering pigs.

The first survey in Gabon concerned two 16-in. OD, 27-km parallel oil pipelines. One of them leaked because of internal corrosion damage.

After a general survey of the market, this inspection company was chosen because of its flexibility in adapting its tools to older installations containing 1.5D, 45, 400-mm bends.

The corrosion-detection pig is equipped with two rows of sensors, allowing differentiation between internal and external metal losses, as does the British Gas tool. An advantage is its compactness, which permits much shorter and lighter tools.

More than 1,000 internal pits situated in the lower part of the pipe have been located over 11 km of one of the pipes and sized from 20 to 100% depth.

Some external nondestructive testing (NDT) measurements, carried out after excavations, showed a good correlation with reported features, except for areas with general corrosion for which the reported penetration of attacks was higher than the actual value.

The electronic geometry pig (EGP) of the same company, working on a touchless electromagnetic sensoring device, is an excellent tool to detect and size unknown geometrical features such as short-radius bends, valves not fully opened, and dents in older installations.

This pig, however, failed to discover thick wall bends which damaged a special tool designed to negotiate 1.5D, 90, and 400 and 500-mm bends.

In this case, only one row of sensors can be used. It is then possible to try to differentiate between internal and external corrosions using two runs, the second being carried out after modification of the magnetization system or of the sensors in order to detect only internal metal losses.

The inspection of a 10-in. OD offshore pipe in Gabon has been possible and successful after replacement of 1.5D bends and a slight modification of the tool to pass dents detected by the EGP. In-line inspections with Rosen tools have generally succeeded in fulfilling their objectives. The thickest pipe wall inspected was 15.8 mm for an offshore 16-in. OD pipe.

In addition to the difficult geometric problems posed by the age of some of the pipelines, wax deposits in pipes have also represented a major challenge (Fig. 1).

Because the oil pipelines must be kept in operation during in-line inspection, it is necessary to try to solve this problem only with cleaning pigs. Using bidirectional pigs built with polyurethane discs and equipped with jetting nozzles on the nose cap greatly aided wax scraping by creating a turbulent flow in front of the pig.

This failed to solve the problem, however, in the worst case: an 18-in. OD, 205-km (127 mile) land pipeline.

Inspection tool sensors were damaged after a few kilometers due to blocking of the supporting springs by wax that had accumulated between the guiding discs. This happened even though the last cleaning pig arrived with a very small amount of wax.

To complete this brief summary of performance with Rosen tools, it must be said that a few signals attributed to external corrosion pits were reported on buried pipelines in Gabon. Some of these signals in fact resulted from cathodic protection measurement cable connections.

For some inspections carried out in France on 8-in. OD and 12-in. OD pipelines reported external attack; were, in reality, internal attacks. This was explained by the wall thickness considered to be at the upper limit for this technique.

ULTRASONIC PIGS

Pipetronix Ultra Scan was used to inspect a 17-km, 10-in. OD production water pipeline on the Lacq gas field in France in 1988. At that time, this equipment was not fully operational.

Three operations have been necessary to achieve the fixed goal which was accurately to size the internal corrosion pits which had led to a leak. Because the tool was unable to cover the entire circumference of the pipe, several runs were required.

Despite the problems encountered, the final report was satisfactory, and some checks confirmed that the reported features were correct. These results were used successfully to make local repairs and keep the pipeline operating according to French regulations.

It should be noted that this inspection was made possible by intensive cleaning operations using solvents while the pipeline was out of service. Other inspections with newer 14-in. and 8-in. Ultra Scan tools, covering the whole circumference, were performed in 1991 on the same Lacq field with success.

A different approach has been followed by Rontgen Technishe Dienst (RTD; Fig. 2 ) to develop cable-operated inspection tools using ultrasonic sensors.' This system resolves the problems caused by the limited size of data processing and storage equipment which no longer resides inside the tool,

Using a cable limits the inspection to a certain length (2 km with a copper wire, 6 km or longer in the future with glass fiber technology) and requires the operating company to stop pipeline service during inspection.

This tool was used successfully in 1990 to inspect a 4.6-km, 26-in. OD pipeline in New Caledonia operated by the Societe Metallurgique le Nickel affiliated with Elf Aquitaine.

This discharge pipeline connects a storage facility to offshore single buoy mooring. Access to the line was made midway, on a peninsula, which permitted two runs of 2 km each. The tool used was self-propelled.

Significant internal corrosion was located and sized in some areas. No external corrosion was detected.

TEST LOOP TRIALS

A 10-in. OD test loop (Fig. 3) has been built by Elf Aquitaine's Centre d'Essais du Fourc in the South of France to test capabilities and limitations of the tools proposed by the market.

The length between launching and receiving pig traps is 120 m (396 ft), and the thickness of pipes is 11.13 mm, except for a 3 m spool which is 21 mm thick.

Seven spools containing calibrated internal and external artificial defects with various shapes and dimensions are introduced.

Water is used to transport the pigs at approximately 0.8 mps. Maximum pressure is 15 bar.

The following six tools have been tested in this loop:

Linalog and Rosen tools for the magnetic flux leakage technique
A prototype tool using measurement of the phase variations of an alternative electromagnetic field (Syminex, France)
Pipetronix Ultra Scan and TD Williamson Flaw-sonic for the ultrasonic technique adapted to pigs
An RTD cable-operated ultrasonic tool.

Each run in the test loop is performed under the responsibility of the tool owner after mutual agreement with the test loop operators concerning parameters. The inspection company issues a standard report which is studied by Elf Aquitaine and compared with existing defects.

A clause of confidentiality prevents public disclosure of results for the time being, but the general results underlie the following comments on tool use.

The choice of the right tool must be made for each individual problem because the two possible technologies-magnetic flux leakage and ultrasonics-have different characteristics inherent to their principles.

Moreover, manufacturers make their own design choices for a given technology. Last, but not least, direct practical experience of the pipeline operating company is essential to optimize its choice.

Cost is also an important factor because, generally speaking, these operations are expensive. However, it is possible to provide some guidelines which are generally accepted or which come from Elf Aquitaine's experience.

EFFLUENT TYPE, VELOCITY

MFL tools can be used with gas as well as with liquids. Current operational ultrasonic tools, however, need a liquid between the transducers and the pipe wall. This means that these tools can be used only if a liquid slug (glycol for instance) or a gel is run between two conventional pigs for inspecting gas pipelines while keeping them in operation.

This process is not simple but can be done. Research intended to develop dry coupling for ultrasonic sensors is currently under way.

Another problem for ultrasonic tools is that they need an homogeneous liquid to prevent misinterpretation of measurements. This is a problem for oil pipelines containing gas bubbles, water drops, or wax in suspension.

The problem has been solved for tools used in wells by installation of a supplementary sensor and a reflector in order to measure the ultrasonic velocity which depends on the characteristics of the medium. Water slugs can also be used in this case.

MFL tools can be used at least up to 4 mps, but a minimum velocity of 0.3 mps is necessary to achieve a satisfactory detection level.

On the other hand, ultrasonic tools give better results when the velocity is low because the fixed impulse frequency determines the distance between readings, hence the minimum diameter of defects that can be detected.

All intelligent tools use electronics which can resist temperatures of 40-65 C.

The maximum allowable pressure varies from one tool to another but is generally lower than 100 bar (14,500 psi). This means that these parameters may have to be adjusted during inspection (reduction of pressure, cooling by means of a water slug, etc.).

PIPELINE DIMENSIONS, FITTINGS

Small-diameter pipelines represent a technological barrier for most intelligent pigs. Compacting of equipment can be compensated for by an increase in the length of the tool and the number of segmented modules.

In practice, the minimum inspectable diameter is 4 in. (Linalog). Others, like Rosen or Pipetronix, propose tools from 6 in. and larger.

Wall thickness is another parameter to be considered. MFL tools are limited to a certain value because complete magnetization of the thickness is necessary to achieve a good detection level. The wall-thickness inspection capability of these tools is about 10-16 mm for diameters from 6 in. to 16 in. and can reach 16-25 mm W. T. from 20 in. to 32 in. Development of newer, more efficient magnets will allow inspection of heavier wall pipe in the near future.

U.S. tools are currently more suited to heavy wall pipe.

For most tools, the minimum permissible 90 bend radius is 3D (even 5D for the smallest diameters). Some tools can negotiate 1.5D when diameters are greater than 16 in. (e.g., Rosen and Pipetronix).

This point, as well as the minimum ID acceptable in straight pipes and in bends, must be carefully checked before any decision is taken to use a given tool.

Of course, the pipelines must be equipped with full-opening valves. Suitable pig launching and receiving facilities are also required.

Some tools have a substantial length (Pipetronix and British Gas, for instance) which causes problems, especially on offshore installations. Temporary facilities can be used when possible.

Guiding bars are recommended for tees if the diameter ratio is greater than 50%. Complete cleaning is necessary before any in-line inspection operation. Bi-directional pigs using polyurethane discs are particularly useful for achieving this goal.

For oil pipelines, experience shows that sometimes as much as 10-15 cleaning runs are necessary to yield acceptable cleanliness. Considerable problems are encountered when wax deposits occur, as explained earlier. When ultrasonic pigs are used, cleanliness is an even more stringent requirement because deposits can interfere directly with the measurements. Solvent cleaning is sometimes necessary, which necessitates shutdown of the pipeline.

DEFECT CHARACTERIZATION

The more recent MFL tools (British Gas, Rosen) generally have two rows of sensors, the second row being outside the magnetizing system and giving information limited to the internal wall surface.

Comparison with the primary recording of data can give an indication of the location of defects related to the inside of the pipe wall.

U.S. tools differentiate internal and external defects as long as sensors are in a stand-off position.

MFL tools generally give better results concerning detection of small defects because they operate on a continuous recording basis. On the other hand, ultrasonic tools showed a partial detection of defects because measurements are made according to a certain impulse frequency. With ultrasonic tools, the sizing of defects is quantitative. With MFL tools, the recorded data must be interpreted to evaluate the dimensions of defects.

Because of the measurement principle, localized attacks are detected more readily than those having a gradually changing profile. Calibration of defects by external NDT carried out at certain locations is necessary to obtain a correct idea of the size of the corrosion damage detected.