APIA-RSC Seminar Wednesday 11 th June, 2008 Hyatt Hotel, Canberra Field test voltages for Fusion- Bonded Epoxy (FBE) coated steel pipe Mehdi Behrouzvaziri, Lex Edmond, Wayne Neil, Nolene Byrne, Stephen Zhang, and Maria Forsyth Monash University Dept. of Materials Engineering
Objectives Determine the voltages needed to reliably detect 0.2, 0.5 and1.0 mm diameter intentional holidays. Determine the probe voltages that nominally defect-free coating can resist at ambient and elevated temperatures. Determine the effect of water immersion at ambient and elevated temperatures (60 0 C and 80 0 C) and the effect of subsequent moisture loss (drying out) on the coating/holiday detection after immersion. Examine any locations where defects are found on apparently defect-free coating, and attempt to identify the cause for these flaws. Examine the output waveforms of holiday detectors.
Outline Introduction Methodology Results Effect of water immersion Ambient and elevated temperatures Detection of artificial coating defects Ambient and elevated temperatures Examination of coating defects Investigation of waveforms DC and pulsed Conclusions Project extension
Introduction AS3894.1 V = 250 F T V = Applied potential (V) T = Coating thickness (µm) F = Coating factor (1 for FBE)
Methodology The dimensions of steel gas pipe used for testing were: - 168 mm outside diameter (OD) - 4.2 mm wall thickness - 450 µm average coating thickness (AS2331.1.4) Calculated test voltage: 5.3 kv Coated with Napgard 7-2501 fusion bonded epoxy Cut into 900 mm lengths
Methodology (cont) Elevated immersion testing components: - A heated stainless steel vessel - Immersion heater and controller - Thermocouple (calibrated) - An insulating thick walled (25mm) polyethylene pipe - Silicon tube - Copper fitting/pressure relief
Methodology (cont) Holiday detectors Unit Type Manufacturer Model A Continuous DC Monash Uni Purpose built B Continuous DC PCWI DC15 C Continuous DC Spy 780-1291 (1-5 kv) D Pulsed PCWI P20 E Pulsed Spy High voltage Little Jeeper
Methodology (cont) Monash detector components: 1- Power module from the mains power 1 2 3 (12 V AC output) 2- Electronic control box (signal processing and trigger switch) 3- High voltage module and sense circuit 4- Fine wire bristle brush probe 5 6 4 5- Dead man safety trigger 6- Earth lead which is attached to the pipe
The P20 pulsed commercial PCWI Detector (Unit D)
Methodology (cont) Oscilloscope - Tektronix TDS 320 - Two channel unit - 100 MHz bandwidth - 500 mega-samples per second High voltage was attenuated by a factor of 1000 by a Tektronix attenuator which was rated: - 3 pf - 100 MΩ - Max 20 kv - 40 kv peak
Effect of water immersion Long term (111 days) at ambient temperature - Without drying after immersion - Dried 57 days after immersion Short term (14 days) elevated temperature immersion - 60 ºC - 80 ºC
Long term (111day) ambient immersion without drying 8 No. of holidays 7 6 5 4 3 2 Number of new holidays detected at this voltage Total number of holidays detected 1 0 1.5 2 2.5 3 3.5 4 4.5 5 5.5 Test voltage (kv)
Dried 57 days after long term (111 days) ambient immersion Total no. of holidays 8 7 6 5 4 3 2 1 0 Immediately after immersion Dried 57 days 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Test voltage (kv)
Short term (14 days) immersion at 60 o C No. of holidays 20 18 16 14 12 10 8 6 4 2 0 Number of new holidays detected at this voltage Total number of holidays detected 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Test voltage (kv)
Short term (14 days) immersion at 80 o C 10 No. of holidays 8 6 4 2 0 Number of new holidays detect ed at this voltage Total number of holidays detected 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Test voltage (kv)
Summary of immersion tests Immersion condition Delay before testing after removal from water Pipe temp. during continuity testing Total number of defects detected at 5.5 kv 111 days / 15 C None Ambient 14 111 days / 15 C 57 days Ambient 8 14 days / 60 C None ~50 C 18 14 days / 80 C None ~70 C 20 No. of defects normalised to the area of one standard pipe (0.48 m 2 )
Detection of 0.5mm artificial coating defects at ambient temperature and humidity 100 100 80 % detected 60 40 30 55 20 0 10 10 5 0 0.4 0.6 1.2 1.5 2.0 2.5 3.0 Test voltage (kv)
Detection of artificial coating defects after Immersion at 80 o C (for 14 days) 100 Holidays detected (%) 80 60 40 20 0.2mm 0.5mm 1.0mm 0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Test voltage (kv) 6 5 >5.5 5.3 Detection Voltage (kv) 4 3 2 3.0 2.5 1 0 0.2 0.5 1.0 Holiday diameter (mm)
Examination of coating detects Defect found in FBE coating where a holiday was detected in a wetted coating (X 60). Void observed in the coating at location where a defect had been detected during continuity testing (X 55).
t Partial thickness hole located in the wetted coating. Detected at 4.5 kv during continuity testing. The specimen has been tilted at an angle so that the location of the hole within the coating thickness (t) can be viewed. The coating surface normally in contact with the steel is in the foreground.
Waveforms High voltage holiday detectors investigated Unit Type Manufacturer Model A Continuous DC Monash University Purpose built B Continuous DC PCWI DC15 C Continuous DC Spy 780-1291 (1-5 kv) D Pulsed PCWI P20 E Pulsed Spy High voltage Little Jeeper Continuity test unit requirements of AS 3894.1-2002, which were examined in this report. Requirement Constant voltage Detector category Pulsed Output voltage DC DC- substantially unidirectional Voltage accuracy ±5% of true value ±5% of true value Voltage precision ( ripple content ) 5% Not applicable Pulse frequency Not applicable 30Hz
Constant DC detectors Waveform from the Monash purpose built continuity tester set at 8.0 kv (with the Crest meter). One small dotted square represents 500 ms on the X axis and 2 kv on the Y axis. The zero voltage position is indicated by the square green marker on the left hand vertical (Y) axis. Waveform from a PCWI DC15 unit set at 5.0 kv. Peak voltage ~5.4 kv. One small dotted square represents 500 ms on the X axis and 1 kv on the Y axis. Waveform from the Spy 780-1291 (1-5 kv) constant DC unit set at 3.0 kv. One small dotted square represents 25 ms on the X axis and 0.5 kv on the Y axis. The time scale has been set to show the frequency (~80 Hz) of the voltage ripple (2.6 to 3.0 kv).
Pulsed detectors Actual and schematic waveform from the Spy Little Jeeper pulsed Unit set at 10 kv. Maximum zero to peak voltage is ~5.6 kv. Average pulse frequency ~24 Hz. One small dotted square represents 25 ms on the X axis and 2 kv on the Y axis. Waveform from the PCWI P20 (pulsed) unit. Set at 5.0 kv. One small dotted square represents 250 ms on the X axis and 1 kv on the Y axis. Peak voltage ~+3.8 kv. Average pulse frequency ~16 Hz.
Summary of waveform results Characteristic Monash Constant DC Detectors PCWI DC15 Spy 780 (1-5kV) Voltage output Constant DC Constant DC Constant DC Accuracy of voltage readout Ripple component (precision) N/A. No direct readout of voltage ~5% PCWI P20 Significant AC component Pulsed detectors Spy Little Jeeper Significant AC component ~7% ~2% ~24% ~44% ~6% (at 5.0 kv) ~12% N/A N/A DC or substantially unidirectional for pulsed units Measured value within ±5% of the display voltage Ripple voltage 5% of the peak voltage Pulse frequency N/A N/A N/A ~16 Hz ~24 Hz 30 Hz Comment Meets requirements Fails accuracy & precision criteria Fails precision criterion Fails all listed applicable criteria Fails all listed applicable criteria AS 3894.1 requirement
Conclusions The voltages required to completely detect intentional holidays in 450 µm FBE coating after having been exposed to the most extreme environment (80 o c immersion for 14 days) were: - 2.5 kv for 1.0 mm diameter holidays - 3.0 kv for 0.5 mm diameter holidays - >5.5 kv for 0.2 mm diameter holidays This compares with a specified testing voltage in AS 3894.1 of 5.3 kv. The voltage output and frequency of the pulsed holiday detectors examined are highly variable. Therefore pulsed detectors are not recommended for testing FBE coating. None of the commercial DC output detectors investigated met the AS 3894.1 requirement of 5% for ripple voltage component.
Conclusions (Cont) The use of Oscilloscope should be considered for checking the waveforms of detectors in the field. Voltages sufficient to reliably detect small through-penetration defects are likely to also identify abnormalities in the coating such as voids and porosity, once the coating had been wetted.
Project extension Select up to eight different commercially available holiday detectors Suitable for FBE Stated to be compliant with AS 3894.1 Select appropriate Crest meters and multi-meter type oscilloscopes with peak measuring capabilities Review and establish/confirm methods to accurately determine the output voltage and waveforms Test/calibrate each holiday detector wrt AS 3894.1
Project extension Choose up to four detectors for further work Investigate the ability of the selected holiday testers to detect intentional holidays for a range of agreed voltages Determine the breakdown voltages on apparently defect-free coatings
Project extension Investigate Detection efficiency of rolling ring and brush electrodes Effect of speed of electrode travel Stability of the detector output voltages Under varying load conditions At different speeds of electrode travel With different moisture levels on the pipe surface Methods and equipment required for field checking and testing the voltage output characteristics of the selected holiday testers
Acknowledgements The APIA Bredero Shaw McConnell Dow One Steel The Pipeline Trust Tyco Water PCWI
Acknowledgements Geoff Cope Leigh Fletcher (RSC Chair) Ankie Larsson (APIA Research Manager) Alan Bryson (Pipeline Trust) Peter Mayes (Bredero Shaw) Ashley Fletcher (Tyco Water) Ian Haddow,
Acknowledgements Dept. of Materials Engineering Mechanical Workshop Electrical Workshop Maria Forsyth Researchers ACES (ARC Centre of excellence for Electromaterials Science)
APIA-RSC Seminar Wednesday 11 th June, 2008 Hyatt Hotel, Canberra Thank you! Mehdi Behrouzvaziri Dept. of Materials Engineering Monash University