When an anchor rod fails due to corrosion, in addition to compromising the strength of the tower, it is also possible that the guy wire might touch the conductor and cause the line to ground fault. Therefore, this is an issue that involves both safety and reliability implications. In the last eight years, New York Power Authority (NYPA) has analyzed the condition of its anchor rods to prevent these failures. NYPA has approximately 6000 anchor rods supporting the transmission structures at voltages from 115 kV to 345 kV. The lines using most of the anchor rods are the Moses Adirondacks lines, built in 1942 and 1958, and Moses Willis lines, built in 1958 and 1978.

NYPA has been using nondestructive evaluation (NDE) techniques to inspect anchor rods for corrosion for the past six years. EDM International Inc. and Southwest Research Institute developed two nondestructive techniques for this inspection with funding from Electric Power Research Institute (EPRI) and several utilities. EDM International Inc. has the exclusive license from EPRI to use these technologies for inspecting anchor rods.

CYLINDRICAL GUIDED WAVE TECHNIQUE

From 2000 to 2001, NYPA tested anchor rods using the Cylindrical Guided Wave Technique (CGWT). This is an ultrasonic technique used to inspect long cylindrical objects such as anchor rods. The ultrasonic wave traveling inside the rod is guided by the shaft and is reflected by the rod end.

Before inspecting an anchor rod, the rod must be prepped to create a small flat surface in its eye. The flat surface must be perpendicular to the longitudinal axis of the rod. A small ultrasonic probe, which acts as both a transmitter and receiver, is coupled to the anchor rod with gel. An ultrasonic pulse is introduced in the rod through the probe. The pulse travels inside the rod and is reflected by the bottom end of the rod, if certain conditions are met. For the ultrasonic signal to be reflected back, the bottom of the rod must be relatively flat. After completing an inspection, the small flat surface created in the eye of the anchor rod is regalvanized with a zinc cold-coat.

NYPA found that the CGWT technique was initially time consuming without an automated grinding system. The technique was sensitive to the presence of surface corrosion along the entire length of the rod. It identified excessively corroded rods well, but many anchor rods with only surface corrosion were also identified as excessively corroded. In 2000 to 2001, NYPA tested approximately 400 anchor rods, and CGWT found most of them excessively corroded. However, excavation showed that most of these rods were still good. They did not lose more than 10% of cross-section due to corrosion and had considerable amounts of surface corrosion.

The CGWT technique works well in certain soils for rods with localized corrosion and also can be used to inspect rods in concrete.

MAGNETOSTRICTIVE SENSOR TECHNIQUE

To overcome the limitations of the CGWT, the Magnetostrictive Sensor (MsS) technique was developed in 2002. This guided wave technique uses the magnetostrictive effect present in ferromagnetic materials, such as carbon and alloy steel, to generate and detect guided waves. MsS technology allows elastic-guided waves at ultrasonic frequencies to be injected into materials from the outer surface without using a liquid couplant. MsS uses a dc-bias magnetic field and a probe coil through which a short pulse of electric current is applied. The alternating magnetic fields produced expand and contract the material via the magnetostrictive effect, thus generating guided waves in the test specimen. The reflections from the boundaries of the material to be inspected are detected using the inverse-magnetostrictive effect.

Depending on the diameter of the rod, three to four permanent magnets are placed on the rod and a ribbon coil is wrapped around it. The technique requires access to the top 6 inches to 8 inches (152 mm to 203 mm) of the anchor rod and does not require any surface preparation.

The measurement takes very little time if no digging is needed to expose the top of the rod. The wave reflects from the end of the rod, and a skilled technician is required to analyze the reflected wave to determine the level of corrosion. This NDE technique classifies the inspected rods into three categories that have been calibrated based on the loss of cross-section:

• Good — 0% to 15% loss of cross-section

• Moderate — 15% to 30% loss of cross-section

• Excessive — greater than 30% loss of cross-section.

NYPA used the MsS technique from 2002 to 2005 to test all of its anchor rod sites.

NYPA'S EXPERIENCE

Based on its experience, NYPA decided that it was cost-effective to have half-cell measurements and soil analysis performed on all untested sites. Half-cell measurements indicate current corrosion activity by comparing the corrosion potential of steel with respect to a standard copper/copper sulphate electrode. Sites that indicate large or moderate corrosion activity based on these measurements undergo further tests using the MsS method. This method then permits each tested rod to be classified into one of the three categories based on loss of cross-section.

Starting in 2002, anchor rods that were originally classified as “excessive” using the CGWT technique were retested using the MsS technique. The anchor rods on the Moses-Adirondack 1 and Moses-Adirondack 2 lines had a ground clamp attached to them that was buried approximately 12 inches (30 cm) below ground. The ground clamps had to be loosened prior to testing these rods with the MsS technique. The process of digging and loosening the ground clamp was time consuming, so only three to four rods were tested at each site on the Moses-Adirondack lines to determine the overall condition of the anchor rods. For lines that did not have buried ground clamps, all the anchors from the selected test sites were inspected using the MsS technique because it takes little time to implement once at the site. The time-consuming part is clearing the site of vegetation to gain access to the anchors.

In the three years of inspection since 2002, more than 1000 anchor rods have been tested using the MsS technique. Approximately 65% of the rods were predicted to be “good” and another 9% of the rods were predicted to have “moderate” corrosion. Approximately 24% of the rods were predicted to have “excessive” corrosion; 4% of those are likely rock anchors. Knowledge of the length and type of rods is very important to obtain accurate NDE predictions.

NDE predictions were visually confirmed by excavating several inspected rods over the years. Verification results on 11 out of the 12 excavated rods were positive.

COST-BENEFIT ANALYSIS

The NYPA transmission system in the St. Lawrence region has approximately 6000 anchor rods. On average, each transmission structure has 16 anchor rods. Tests show that 256 anchor rods need replacing and 92 anchor rods are predicted to have moderate corrosion, which means they will be tested six years from now to see if they have to be replaced. Even if all anchor rods in the excessive and moderate categories have to be replaced, 5652 rods will be unaffected. This represents a significant benefit in avoided cost over the possible alternative of just replacing all the anchors.

LEARNING FROM EXPERIENCE

In the last six years, NYPA has learned the following:

• Each anchor rod site is different.

• The condition of the anchor rods at the same site is generally similar. Testing three to four individual anchors out of 16 anchor rods provides a good indication of the level of corrosion at the site on the lines that had buried ground clamps. It was assumed that the condition of the remaining anchor rods was similar.

• Agricultural use of the right-of-way increased the corrosion rate for the anchor rods as a result of fertilizers.

• Sandy soil in the mountains typically indicated no corrosion degradation.

NYPA's experience confirms that it is worthwhile to test the anchor rods using a combination of soil analysis and NDE techniques. Many times during this process, NYPA's management asked why not simply replace all the anchor rods. It was proven that this was not necessary, and that the NDE measurements lead to savings in avoided cost by not replacing good anchor rods.

George Stranovsky has worked in the electric utility industry since earning his BSEE degree from New York University in 1970. He has worked with AEP and GPU, and for the last 22 years with the New York Power Authority as a senior engineer in the research and technology development group, concentrating on transmission issues. george.stranovsky@nypa.gov

Fred LaChance began his career at New York Power Authority (NYPA) as a laborer before working his way to lineman. He has a BSCE degree from Clarkson University and is currently a transmission engineer responsible for the maintenance of NYPA's transmission assets in the northern region of New York. fred.lachance@nypa.gov

Dr. Arun K. Pandey received his MS and PhD degrees in civil engineering (structures) from Duke University. He joined EDM International Inc. in 1994 and is currently a project manager. Pandey is a member of the American Society of Civil Engineers, American Society of Nondestructive Testing and Sigma Xi. apandey@edmlink.com

The table is a summary of NDE predictions for the inspected rods from NYPA's system.

Line	Number of Rods	Good		Moderate		Excessive		Likely Rock
		Number	Percent	Number	Percent	Number	Percent	Number	Percent
MA1	167	84	50.3%	8	4.8%	65	38.9%	27	16.2%
MA2	141	76	53.9%	11	7.8%	54	38.3%	8	5.7%
WP1	137	90	65.7%	27	19.7%	20	14.6%
WP2	158	99	62.7%	15	9.5%	44	27.8%
MW1	133	114	85.7%	6	4.5%	13	9.8%
MW2	131	114	87.0%	14	10.7%	3	2.3%
PS1	68	64	94.1%	1	1.5%	2	2.9%
MRG1	27	13	48.1%	3	11.1%	10	37.0%
MRG2	26	1	3.8%	1	3.8%	23	88.5%
MRG3	26	22	84.6%	0	0.0%	3	11.5%	1	3.8%
PV20	35	10	28.6%	6	17.1%	19	54.3%	11	31.4%
Total	1049	687	65.5%	92	8.8%	256	24.4%	47	4.5%