The recorder on his inspection system went crazy. Why was this single piece of drilling pipe producing such erratic amplitudes on the chart? The inspector removed the inspection head and moved to the next tube. This pipe joint produced normal indications. It can’t be the equipment, he thought. Visual inspection of the mystery joint coupled with a remotely controlled video camera revealed no anomalies, inside or outside. What would produce indications typical of massive corrosion pitting on this smooth steel? What could this be? We’ll come back to him in a few minutes.


BACKGROUND
Looking back in history one can now appreciate the recent evolution of non-destructive tubular inspection equipment and yet note that there is more to be accomplished. For 25 years the electronic inspection of Oil Country Tubular Goods (OCTG) centered around the flux leakage discipline. During this period of time, which some call the black box era, new and used ferrous OCTG were casually checked for flaws such as stress risers (notches and pits) using search coils. These coils, operating on the principle of electrical induction, when used in a strong DC electromagnetic field could often locate surface breaking cracks or pits on the OD of magnetizable steel pipe. They fall disastrously short in the ability to identify major defects occurring on the pipe’s I.D. In addition, search coils are insensitive to gradual changes in the induced magnetic flux path. These minor diversions of flux produced by shallow reductions in the tube’s wall rapidly become larger, often resulting in down-hole failure during the pipe’s next "trip" into the well.


EVALUATION: Old Technology...at work?
In the 80’s Exxon & Chevron spearheaded evaluations of all electromagnetic inspection (EMI) they received from pipe service companies. Their investigations supported pronouncements, through a variety of topical papers, that most if not all search coil based EMI systems were insensitive to detrimental flaws.

Said Exxon, "The EMI method is the dominant oilfield pipe inspection method used today. Most automated EMI units in the field today are wall-thickness limited due to a lack of sensitivity to internal flaws. From our experience, this situation becomes more acute when pipe wall thickness approaches one-half inch."

Additional work with EDM notches, replicating off-axis flaws, had shown Exxon that, "many EMI units (using search coils) are relatively insensitive to this flaw."

Chevron USA, Inc. provided additional information centering around the difficulties in using search coil based EMI equipment to locate small defects. Chevron reported, "As expected, (during testing) the EMI units could not pick up 5% I.D. flaws as evidenced by their...detection efficiency. They faired only slightly better on 10% I.D. flaws. Of particular concern to ...(Chevron) was the relatively poor performance on 10% O.D. flaws..."

Tubular inspection costs are a small amount of the typical well’s overall expense. This amount, if spent on quality inspection, could save larger amounts of money used in fishing out unexplained failures in the pipe’s body wall. "The cost of these tubulars (used in oil and gas wells) represents a major portion of the total drilling costs (about 20%), but a failure of one of these tubulars can cause normal costs to skyrocket and may result in abandonment of the well, or possible loss of the drilling rig or even human life."


A MYRIAD OF DEFICIENCIES
There are many reasons why the search coil has difficulty locating small, naturally occurring or man-made defects:

Speed sensitivity: Coils are sensitive to speed changes. They require a constant velocity to maintain the pre-set amplitude output.

Signal devaluation: Normalized signals can lose up to 75% or more of their amplitude on either side of the coil’s mid-line.

Non-linearity of electromotive force (EMF): The search coils will add the effect of all flux lines enclosed within it’s circumference. A primary signal is then masked, blended with non-rejectable indications and results in a poor signal-to-noise ratio.

Coil output EMF is low when encountering slight changes in the flux path as those normally produced by loss of metallic area (LMA) and shallow bottom pitting.

Non-specific coil voltages do not permit creative computer signal processing.

The search coil is not sensitive to off-axis (22.5º-45º) defects.


A BETTER MOUSETRAP
The emphasis on quality inspection promoted by oil companies has influenced a new development in flux leakage evaluation of OCTG greatly enhancing electromagnetic competency. Through the use of solid state Hall devices, defects can now be located more accurately, through greater depths and at increased speeds.

The Hall element is a semi-conductor which responds in a linear manner to increasing magnetic field intensity passing through it. "The Hall-effect, discovered in 1879 by E.H. Hall, results from the action of externally applied magnetic fields upon charge carriers in metals or semi-conductors." In short, an externally applied magnetic field can bend a constant DC control current in the semi-conductor Hall element. By bending the control current path two charged signal electrodes no longer have the same voltage potential. Instead, a voltage difference known as the Hall Voltage appears across these signal electrodes. The Hall voltage is proportional to the product of the control current and the normal component of the external magnetic field.

"The main advantage of the device is the small size of its active area. The second advantage of such devices is that they can be aligned to measure the normal magnetic field intensity or tangential magnetic field intensity of the flux leakage field from a discontinuity, with an amplitude that is not dependent on the speed of the sensor over the discontinuity."

In addition, when packaged as an IC, sensors may be computer tolerance matched facilitating parity across the inspection shoe’s surface as it contacts and moves along the pipe’s O.D. This provides the repeatability necessary to ensure accuracy in flaw detection. Standardization of signal output may now be accomplished. Further, the robust make-up of these sensors allows their reuse for many years.

Solid state sensors have several other advantages over search coils:

Linearity of output signals across all frequency components.

Off-axis defects (to 45° left or right hand) can be located. Small area sensors (0.060") provide the resolution and sensitivity to detect component segments of flaws which are not absolutely perpendicular to the magnetic field.

Flat semiconductor sensors may be mounted more closely to the inspected surface.

Because they can be used in large numbers, they produce excellent information on the condition of localized portions of the tube under inspection.


The following four cases show the advantage in using solid state sensors.

CASE # 1: WALL LOSS
This example shows how small area solid state sensors can point out a localized wall loss area missed by search coils and a conventional gamma ray wall thickness unit. The pipe in this example is a joint of 2-7/8" O.D., 7.90#, P105 production tube with PH-6 connections. It was in the inventory of a pipe rental company in the Gulf Coast area. One of many removed from an offshore well this joint was evaluated by a search coil inspection system and gamma unit. The first inspection company assessed the joint favorably showing no detectable injurious flaws. A second inspection company was called in because the drilling contractor wanted two evaluations before returning the tubing string down hole. This second EMI unit employed solid state sensors locating flaws and localized wall loss. A circumferential (1° taper) wall loss ring is machined at the center of a 5’ reference tube. This 5% wall reduced area establishes sensitivity to localized changes in the induced magnetic field. On the first pass, the wall loss sensors monitoring the DC flux density indicated a localized change. A large deflection appeared on the chart. Upon prove up, using a compression wave thickness gauge, an area approximately 1" in diameter revealed LMA in excess of 20% (0.058").


CASE # 2: SHORTENED UPSET AREAS
To provide extra thickness and strength, plain end drill pipe is upset prior to the attachment of the tool joints. These threaded connections allow tubulars to be made up or joined during the drilling procedure producing a drill string. This upset area provides the surface to which the once separate tool joint (pin or box) is attached through flash, friction or inertia welding. The pipe may fail in or around this attachment area due to improper welding or other external forces. In this instance a complete separation from the tube could occur (twistoff). Additionally, a tight fissure exposed to drilling fluid, under pressure, can widen producing a washout.

According to the American Petroleum Institute’s January 1, 1995 Recommended Practice 7G, Fifteenth Edition, "Recent industry statistics confirm that a major percentage of tube body in-service failures occur near the upset runout or within the slip area. Special attention to these critical failure areas should be performed during inspection to facilitate crack detection in drill strings which have been subjected to abnormally high stresses."

This formed and welded upset area must be properly accomplished for seamless drill pipe to survive the rigors of drilling. "Drill pipe is subjected to cyclic stresses in tension, compression, torsion and bending. Tension and bending are the most critical of these. Bending and rotation produce an alternation between states of tension and compression at localized points in the drill pipe-such as tool joints and areas near each upset." This is much like bending a wire back and forth until it breaks.


UNIQUE DISCOVERY
Recently a South Texas drill pipe inspection company made a unique discovery. Nearly new drill pipe (5", 19.50#, G105) washed out at the upset runout. This area appears as an angled ramp leading from the pipe’s nominal wall terminating at the thickened upset. Upon cutting the pipe open for visual investigation, runout lengths of only 5/8" to 7/8" in length were observed. According to this pipe service company, industry standards dictate that a proper runout should be between 2" and 4" in length. This length protects the pipe’s upset area from accumulated stress as it rotates through tension and compression during drilling, especially in deviated wells.

Subsequent electromagnetic inspection of this critical dimension on additional drill pipe has shown that the solid state sensor provides a distinctive chart signal not normally observed when the runout length is within tolerance. This abnormally rapid transition from upset area (.880") to nominal tube wall (.362") indicates that these foreshortened areas had forced the formation of a fatigue crack much too early in the tube’s useful life. Prove-up of this chart indication is now the basis for rejection of the tube.


CASE # 3: TIGHT LONGITUDINAL STRESS CONCENTRATORS
Another example of a defect not easily detected by search coils is the tight stress concentrator sometimes termed a cooling crack. This notch is produced during the quench and tempering process when modifying the yield strength of ferrous tubing. "Heat treating or quench cracks are typically forked, surface indications that are randomly placed in any direction on the test object." Longitudinally oriented defects of this type may be lengthy or extremely short. In any event, they are normally precursors to future failure. This flaw, in it’s emergence, is generally so tight that location with search coils is extremely difficult.

A pipe threading facility was having to credit some customers for premature in-service failures attributed to longitudinal defects on 2-3/8" O.D. tubing. The tube was easily magnetized but conventional search coil EMI units could not detect these extremely tight cracks. Due to customer returns, this manufacturer wanted to catch these flaws before shipping. Return costs are expensive and a quality inspection program is essential to the "bottom line".

A solid state sensor longitudinal inspection system was employed to learn if the small area solid state sensors could produce noticeable indications. The system induces an extremely strong active transverse DC field through two magnetic poles which rotate about the pipe’s surface. "In (their) miniaturized form these Hall generators permit the point measurement of magnetic d.-c. fields in fractions of 1 oersted up to the highest produceable field strengths." An array of differentially connected solid state detectors, spaced to eliminate the slowly changing magnetic field in this tubing, produced recognizable indications from these stress concentrators. Sensor size played a great roll in the ability to resolve these small defect areas. Solid State sensor responsiveness to extremely low levels of magnetic field leakage provided chart indications which alert the inspector to the defect’s location.


CASE # 4: FATIGUE?

BACK TO OUR INSPECTOR

Our inspector brought the strange joint of drill pipe to the drilling contractor’s attention. Not realizing the potential for disaster the tube was used on the next offshore well. This joint failed almost immediately. Retracing the inspector’s documentation showed that this was, in fact, the pipe which produced erratic signals. This scenario was repeated shortly thereafter with other such tubes failing in the well. The chart indications from these jointed tubulars are nearly "finger prints" of each other according to a Lafayette, Louisiana tubular inspection company. In the past year this company has rejected over 2,000 such tubes from one world-wide drilling contractor’s inventory. They have culled many other such tubulars for additional customers, using solid state sensors.

The example shown is of a thermal chart rendition depicting these erratic indications. This joint is a 4-1/2" O.D., 16.60#, X-95 drill pipe inspected on February 10, 1995. According to the inspection company, "Prove-up by use of the UT gauge could not find any significant pitting. Past joints of this type indication have washed out in the upset area. Test(s) by laboratories revealed that the pipe had heavy concentrations of sulfur and magnesium. We are continuing our testing at this time as we still believe the pipe is fatigued due to use or age."

The inspection company attributes the detection of this phenomenon to the sensitivity of solid state Hall sensors.


CONCLUSION
Historically the flux leakage method using search coils predominated. It is now known that analysis of anomalies in oil country tubular goods is greatly enhanced through nondestructive testing using solid state Hall sensors.

Groups of sensors can be standardized ensuring repeatability.

Unique flaws are being located.

EMI systems may now be logically tied to computers.

Finite defect location is possible through data generated by several hundred small area sensors. The safety of personnel, our environment, and expensive drilling equipment deserves no less than the best inspection possible on OCTG.


GLOSSARY
Flux leakage method: A technique for the detection and analysis of a discontinuity via the flux that leaves a magnetically saturated, or nearly saturated, specimen at the discontinuity.

Hall detector: A semiconductor element that produces an output electromotive force proportional to the product of the magnetic field intensity and a biasing current.

Tangential field strength: "The Hall generator method is especially well suited for the determination of the tangential field strength, that is, the true field strength during magnetization of a test part, because of the small dimension of the measurement probe.


ENDNOTES
1.  Moyer, M.C., Dale, B.A., "Methods for Evaluating the Quality of Oilfield Tubular Inspection", presented at the 1984 59th SPE Conference, Houston, Texas, September 16-19, 1984.
2.  Reynolds, B.W., Gill Jr., L.O., "A Practical Evaluation of Non-Destructive Inspection Units: Methods And Results", presented at an SPE meeting on May 19, 1987 in New Orleans, Louisiana.
3.  Moyer, M.C., Peterson, C.W., "The Importance of Quality Tubular Inspection", presented at the 1981 IADC Drilling Technology Conference, Calgary, Canada, March 10-12, 1981.
4.  McMaster, Robert C., "Electromagnetic Tests with Hall Effect Devices", Section 12, Part 1, in Nondestructive Testing Handbook : Electromagnetic Testing , 2nd Ed., Vol. 4, p. 317.
5.  Ibid
6.  Stanley, Roderic, Hiroshima, Tatsuo, Mester, Mike, "Introduction to Diverted Flux Applications", Section 22 in Nondestructive Testing Handbook : Electromagnetic Testing , 2nd Ed., Vol. 4, p. 639.
7.  American Petroleum Institute, Washington, D.C. 2005, Recommended Practice for Drill Stem Design and Operating Limits 7G, 15th edition, January 1, 1995.
8.  Petroleum Extension Service, edited by Jodie Leecraft, The Drill Stem, Unit 1, Lesson 3, Second Edition, 1981, p. 5.
9.  Atkins, David R., Urzendowski, Michael A., Warke, Robert W., "Discontinuities in Ferromagnetic Alloys", Section 4 in Nondestructive Testing Handbook : Magnetic Particle Testing, 2nd Ed., Vol. 6, p. 92.
10.  McMaster, Robert C., "Magnetic-Field Test Principles", Section 33 in Nondestructive Testing Handbook, Vol. 2, 33.12.
11.  Spinco, Inc., Lafayette, Louisiana.
12.  Ibid, 4 above, p. 656.
13.  Ibid, 4 above, p. 656.
14.  Ibid, 8 above.