Appendix N

 (a) make any warranty, express or implied, with the respect to the use of any information, apparatus, method, or process disclosed in this report or that such use may not infringe privately owned rights:

Introduction

Mr. John Lewis of Shakespeare Company contracted with Georgia Power Company Research Center / Southern Electric International to perform critical impulse flashover tests on a CCA pole and a fiberglass pole. In addition to these tests, 12 - 2" X 18" samples of fiberglass pole were tested for QUV exposure, resistive leakage, and critical impulse flashover values.

Pole Configuration Impulse Flashover Tests

Critical Impulse Flashover Tests were performed on a CCA and a fiberglass pole, Both poles were connected in the configuration shown in Appendix (N)A. In order to determine the critical impulse flashover values, impulses were applied to each test configuration using the up-down method described in ANSI/IEEE Std. 4 - 1978. The average value of these flashovers was computed and the voltages were corrected to standard atmospheric conditions using ANSI C29.1 voltage corrections. Each configuration was tested under wet and dry conditions. The water spray pattern used in the test was the IEEE Std. 4 - 1978 Revised Wet Test. The critical impulse flashover values are given in Table 1. The flashover values can be used to obtain Basic Impulse Level values for insulation materials. The BIL value for a particular insulation material is the level of voltage which the insulation material should be able to withstand without a disruptive discharge. For this reason, critical impulse flashover values are higher than stated BIL values for an insulation material. ANSI Standard C29 uses approximately 90% of the critical impulse flashover value as the BIL value for an insulator. This value could be used as the BIL value to choose for these poles. During the flashover tests both poles experienced the splitting of the fibers on the external surface of the poles. This splitting exposes the internal portions of the pole to weathering. After the third impulse application, the frayed fibers on the pole's surface caught on fire. The flames were extinguished in order to minimize damage to the pole's surface while determining the flashover values. Each consecutive flashover produced some burning of the fibers. The flames required extinguishing approximately 4 times. Photographs of the CCA and fiberglass poles are shown in Appendix (N)B.
 

Table 1 Critical Impulse Flashover Values
Impulse Polarity	Fiberglass Pole Flashover Values (kV)		CCA Pole Flashover Values (kV)
	Dry	Wet		Dry	Wet
+	1110	944		925	786
-	-950	-948		-1030	-724

Fiberglass Pole Section Tests

Twenty-four 2" X 18" pole samples were received for testing. These samples were divided into two groups. Twelve were placed in storage and twelve were placed in QUV and condensation aging. The samples were cycled between QUV @ 60 degrees C every 4 hours for 1036 hours. A data sheet describing this test is included in Appendix (N)C.

After the QUV/Condensation aging tests were completed, three samples from each test group were soaked in tap water for three days. Following this soak period, three more samples were taken from the two dry test groups for comparison against these soaked samples. Two comparison tests were performed.

The first of these comparison tests was an ac resistive leakage current test. This test measures the effect of moisture absorbed during the soak test. Any moisture absorption in the test samples will produce increased resistive leakage current. Because the test is an ac voltage test, even moisture which does not extend for end to end in the test samples can be detected. The results of these tests are shown in Table 2.
 

Table 2 93-732 C93748 Shakespeare Fiberglass Pole Samples High Voltage AC Resistive Current Measurements
Sample Description	Resistive Leakage (µA)	Watts Loss (W)
Dry Control #1	<1	.01
Dry Control #2	<1	.01
Dry Control #3	<1	.01
Dry QUV #1	<1	.01
Dry QUV #2	<1	.01
Dry QUV #3	<1	.01
Soaked Control #1	354	7.13
Soaked Control #2	45	.91
Soaked Control #3	216	4.41
Soaked QUV #1	1	.03
Soaked QUV #2	8	.16
Soaked QUV #3	3	.06

The second comparison test was an impulse flashover test. Each sample was subjected to a series of 1.2 µS X 50 µS lightning impulse waveforms. Lightning impulses were applied to each sample until enough pulses had been gathered to determine the critical impulse flashover voltage. The number of impulses used was not less than ten and no more than 20. As in the pole testing described earlier, the up/down method described in ANSI/IEEE Std. 4-1978 was used to determine the flashover value. The peak value of each of these pulses was recorded. The average of these peak voltages was recorded as the flashover voltage for the sample. Because these were comparative tests, the atmospheric conditions were only recorded. No corrections were applied to these test voltages. The results of each sample's flashover value are shown in Appendix (N)C. The flashover voltages for each of the three samples from each test condition (dry control, soaked control, etc.) were averaged. This average is recorded in Table 3 for each test condition.
 

Table 3 93-732 C93748 Shakespeare Fiberglass Pole Samples Critical Impulse Flashover on 2" X 18" Samples
Test Condition	Average Critical Flashover (kV)
Dry Control	-320.6
Dry QUV	-322.9
Soaked Control	-263.6
Soaked QUV	-285.5

Conclusions

The test data from the pole configuration impulse flashover test shows only slight variations between the flashover values for the fiberglass and CCA poles. However, under wet conditions, the CCA pole displayed a lower flashover value than the fiberglass for this configuration. Typical BIL values for distribution systems are much below the flashover values for these configurations. Therefore, it is likely the transformers, lightning arresters, and switches on the system may limit the voltages on the lines to less than these flashover values for induced lightning surges. During these flashover tests, some burning of the frayed fibers on the fiberglass pole was observed. For this reason, a comparison of the power arc flashover recovery of both poles may be desired.

The data from the fiberglass pole sections indicates those samples which were exposed to QUV aging have become more resistant to moisture pickup than the new samples. One possible explanation for the reaction could be the QUV light and heat in the test equipment has helped to cure the paint and resin materials. The malleability observed in the coatings during mounting supports this conclusion. The increased resistive leakage for the water soaked control samples indicates these samples absorbed more water than the QUV aged samples. The effect of this moisture pickup on the impulse flashover value is based on the pattern in which the moisture entered the samples. In most cases, moisture absorption decreases the flashover of the test sample. This decrease in flashover value is caused by an uneven moisture distribution which results in a non-uniform electric field. This non-uniform field distribution allows the plasma arc to initiate more easily, thus resulting in a low flashover value. The flashover value for those samples which picked up moisture in this test displayed these characteristics. The soaked control samples had the lowest flashover value, -263.6 kV lower than the soaked QUV samples. The flashover of the dry samples was almost 60 kV higher than the soaked control. Therefore, the moisture is definitely decreasing the withstand capability. The difference between the dry control and dry QUV samples was only 2.3 kV and should be considered insignificant.

---

Shakespeare Composite Structures

19845 Highway 76, Newberry, SC 29108 · 803.276.5504 Fax: 803.276.8940
www.skp-cs.com