Sei sulla pagina 1di 4

WATER ABSORPTION OF EXTRUDED CABLE COMPONENTS

S. PClissou
Hydro-QuCbec(IREQ)
1800 montCe Ste-Julie, Varennes
Quebec, Canada J3X 1S1

INTRODUCTION

The presence of water is detrimental to the electric performance of extruded medium voltage cables
since it leads to the formation of water trees. Hence, it is important to investigate the amount (of water
that can be absorb by the components of these cables. In the past, several studies [l-51 have been
performed to understand and describe these phenomena in a cable context, but, few have considered
full-size cables. Furthermore, the advent of new cable components deserves an updated investigation.
This paper is a continuation of previous work [6-71 on moisture absorption and diffusion in
conventional crosslinked polyethylene (XLPE) and water tree crosslinked polyethylene (WTRXLPE)
cable insulation. It presents data on XLPE, ethylene propylene rubber (EPR), three different
WTRXLPE, and also conventional and new technology insulation and conductor shields, using in all
cases only unaged cables. The influence of repeated water absorption and successive water content
measurements on the effectiveness of the three types of WTRXLPE insulation is also reported

EXPERIMENTAL

Table 1 gives the characteristics of the cables used in this work. Nearly all the cables are class 28 kV
with 750 kcm aluminum conductors and were made by three different manufacturers. One cablle, NT2,
has a typical size as per AEIC CS5 specification [8],and was made by another manufacturer. Three of
these cables, C1 to C3, were extruded with conventional XLPE and shield materials, four of them,
NT2 to NT6, with three different WTRXLPE and new technology insulation and conductor (except
NT2) shields. The EPR insulated cable, NT7, was made with compatible standard shield materials.
Except "I7 all cables were dry-cured.

Table 1. Cable characteristics.


NT7
28
750 kcm
AI
standard

EPR
standard
jtrippablc

-
D

To study water diffusion in these cables, chunks (4x5~15mm for 1/0, 7 x 7 ~ 1 5mt for 750 kcm) cut
from the insulation and pieces of shield (typically 1 x 5 ~ 1 0mm) were immersed in distilled water kept
at 80°C. Samples were taken at various times for water content measurements. Thie effectiveness of
WTRXLPE insulation was studied using the same type of chunks immersed in distilled water at 80°C
for exactly 30 days, several times in succession, followed by a water content measurement.

The measurements were performed with a Mitsubishi Moisturemeter, model CA-05, based on ithe Karl
Fischer titration method [9]. In this technique, the sample is placed in an oven where it is heated at

- 141 -
150°C and the diffusing moisture is carried by nitrogen gas to an electrolytic cell where its mass is
measured. To study the effectiveness of WTRXLPE insulation, the oven temperature was set
successively at 90°C and 150°C over constant periods of time. The first temperature simulated normal
operating conditions, the second was to remove all the remaining water. It should be noted that the
samples represented either the entire insulation or the shield thicknesses, in order to avoid misleading
local measurement [ 101.

RESULTS AND DISCUSSION

Conductor shields
Figure 1 shows a comparison of the water content between conventional and supersmooth conductor
shields as a function of the square root of time. After two years immersed in distilled water at 80°C
the conventional shields reached a water content of 1% and the supersmooth ones roughly 2.8%. It
took approximately 14 months to obtain a saturation. This difference in water content may be
explained by the type of carbon black used in these shields: furnace black for the conventional and
acethylene black for the supersmooth one. Regarding the influence of manufacturer, not much
difference was observed among datas. Obviously, new technology conductor shields absorb much
more water than conventional ones.

Figure 1 - Comparison of water content of Figure 2 - Comparison of water content of


supersmooth and conventional conductor cleaner and conventional insulation shields
shields after immersion in water at 80°C after immersion in water at 80°C

Insulation shields
Figure 2 presents a comparison of the water content of conventional and cleaner insulation shields.
The conventional shields, except C3, absorb approximately 18% of water, a value approaching
saturation, while the clearner ones and C3 attain the huge value of 28%. The different water content
of shield C3 may indicate that it is not a conventional material as originally expected. Apart from
component C3, there is some consistency among the three manufacturers. It is noteworthy that
saturation was not obtained for the cleaner insulation shield, even after 2 years; thus, much higher
values are to be expected at equilibrium. Such great water contents in comparison with the conductor
shields ones may be due to the additives contained in the formers to make them strippable. Of course,
this situation does not reflect field conditions, where the insulation shield is rarely exposed to
temperatures of 8O"C, but it permits relative comparisons. However, this is not the case of conductor
shields for which temperatures up to 90°C are possible. These results clearly show that new cable
components absorb much more water than conventional ones,

- 142 -
Insulatiom
The water content of the various insulation materials is depicted in Fig. 3. The insulation WTRXPE-
1 absorbs, at 80°C and saturation, 7500 ppm of water, a value which is 15 times greater than the one
obtain for the conventional XLPE. This large amount of water must be attributed to the waiter-tree
retardant (WTR) compound, since this cable was dry-cured, i.e. with almost no microvoids or water
reservoirs in its insulation. It is believed that this WTR is highly polar and therefore easily traps water
molecules. On the other hand, WTRXLPE-2 only reaches less than 2000 ppm of water, a result which
suggest a WTR additive less polar than the previous one. However, even after 2.5 years the water
content of this type of insulation did not saturated. This unexpected behaviour could be the result of a
chemical change (or degradation) of the WTR material into a more polar one, its efficiency is
therefore questionable. From the few data points obtained until now for WTRXLPE-3 and EPR1, it
can be estimated that their water content will reach respectively 1000 and 2000 ppm. Figure 2
indicates also that the influence of the cable manufacturer is also here not significaint, an information
which reveals a certain uniformity of these materials.

Water content (ppm)

...........................

................................

........ .........

0 2000 4000 6000 e000 10000 12 U30


reduced tlme fsM
[+Cl e C 2 +C3 f3NT3 +NT4 +NT5 f N T 8 +Nl7 *NT2/

Figure 3 - Comparison of water content of the various insulations


after immersion in water at 80°C

The capability of the WTRXLPE materials to retain or trap water also provides an indirect indication
of the chemical stability or effectiveness of the WTR additive. In order to test this effectiveness
several water immersions and successive water content measurements at two temperatures were
repeated on the same samples. The results, given in Fig. 4, show that the water content of WTFXLPE-
1, measured either at 90°C or 150"C, is more or less stable up to the eighth run where a change is
observed. At this run the amount of water released at 90°C has significantly increased while at 150°C
it is much reduced, but the total amount is rather similar to the previous runs. These differences in
water release may indicate a chemical change in WTRXLPE-1 due to aging, but obviously this
deserves more work. On the other hand, WTRXLPE-2 reveals a relative increase. in water content
after each run. In terms of percentage, this increase is closer to 500% after the sixth run. Thiis WTR
material clearly undergoes an irreversible change. The successive thermal treatments cannot be
responsible for this change, since a similar one, or the same, occurs at only 8O"C, as indicated
previously [7]. Thus, there is some doubt about the effectiveness of WTRXLPE-2 to operate in
normal conditions, i.e. at 90°C [SI. For WTRXLPE-3,after the only two runs performed no change
was observed and the amount of water released is consistent with the one in Fig. 3. In conclusion,
these three WTRXLPE insulations have different behaviour as regarding the trapping of water,

- 143 -
Wafer conlent (ppm) water Conkent (ppm)
5000
WTRXLPE-1
. . . . . . . . . . . . . . . . . . .
nI
-I I - 1 .......................

3000 .......................

2000 ....................

1000

0
1s1 2nd 3rd 41h 51h 8th 7th 81h

5000 Figure 4 - Water content at two successive


WTRXLPE-3
4000.. . . . . . . . . . . . . . . . . . . . . . . . temperatures of three WTRXLPE insulations
after immersion several times in water at 80°C
.......................
3000.. for 30 days. The number over each histogram
2000 . . . . . . . . . . . . . . . . . . . . . . . . .
represents the total amount of water released
for both temperature.
1000. ........................
e40 W S

0-w I

The great capability of these new components to retain water could influence cable life, since water
trees grow with water. Even at temperatures lower than 80"C, the shields still can absorb a great
amount of water and, as water reservoirs, they can supply water to the insulation from both sides. It is
believed that the thermal gradient and thermal cycling would facilitate the transport of water in cable
insulation. Of course, the presence of greater amount of water in WTFCUPE insulation, as compared
to XLPE, indicates that they are effective in trapping water, preventing it, hopefully, to contribute to
water tree growth.

CONCLUSION

The water absorption of various kinds of cable insulation and shields has been studied. The three
WTRXLPE insulations absorb a large amount of water compared to ordinary XLPE, probably
because of the polar character of the WTR material, One WTRXLPE seems to lose its effectiveness at
80°C. EPR has also a great water content capability but less than one of the WTRXLPE. It is clear
from this work that new cable insulation and shields absorb much more water than conventional ones.

REFERENCES
[I] H. Matsuba, E. Kawai, K. Sato, 1976 IEEE Intern. Symp. Electr. Insul., July 14-16, 1976. pp.
224-7.
[2] E. Ildstad, H. Faremo, B.E. Knutsen, Nordic Insulation Symposium (Copenhagen) Nord-IS90,
paper W.4 (1984).
[3] J. C. Chan, S.M. Jaczek, IEEE Trans. Electr. Insul. Vol EI-13, No. 3, pp. 194-7 (1978).
[4] G.J. Le Poidevin, IEEE Trans. Electr. Insul. Vol EI-15, No 4 pp. 362-5 (1980).
[ 5 ] E. Ildstad, Nordic Insulation Symposium (Copenhagen) Nord-IS84, paper #7 (1984).
[6] S. PClissou, H.J. Wintle, IEEE Intern. Symp. Electr. Insul., June 7-10, 1992, pp. 165-8.
[7] S. PClissou, 1994 IEEE Intern. Symp. Electr. Insul., June 5-8, 1994, pp. 185-7.
[8] AEIC CS5-87, "Specifications for Thermoplastic and Crosslinked Polyethylene Insulated
Shielded Power Cables Rated 5 through 35 kV" (1987).
[91 J.H. Groeger, D. H. Damon, S. Pelissou, J.-P. Crine, H. St-Onge, 3rd Int. Conf. on Polym. Insul.
Power Cables (Jicable), pp. 276-82 (1987).
[lo] S. PClissou, J.-P. Cnne, J. Castonguay, S. Haridoss, T.K. Bose, M. Merabet, R. TobazBon,
"Nature et distribution de I'eau dans les clbles rBticulCs i la vapeur", 3rd Int. Conf. on Polym.
Insul. Power Cables (Jicable), pp. 270-5 (1987).

- 144 -

Potrebbero piacerti anche