IEEE Transactions on Dielectrics and Electrical Insulation Vol. 5 No.

1, February 1998 103

Space Charge Accumulation in Two Power

Cable Grades of XLPE
K. R. Bambery and R. J. Fleming
Department of Physics, Monash University, Clayton, Melbourne, Australia

ABSTRACT
The accumulation of space charge in XLPE (cross-linked polyethylene), cross-linked using DCP
(dicumyl peroxide) or a silane-based grafting process, was studied via the LIPP (laser-induced-
pressure-pulse) technique. Planar samples 0.5 mm thick were obtained from the XLPE insula-
tion of power distribution cables. dc fields to 10 kV/mm were applied at temperatures in the
range 20 to 90°C. Usually the DCP samples developed heterocharge, and the silane samples ho-
mocharge. However, the observed space charge densities were not very different. Nearly all
of the charge accumulated within 100 p m of each electrode, with very little in the remainder
of the volume. Both sample types showed a near-perfect inversion of the equilibrium space
charge profiles on reversing the applied voltage polarity. This observation is explained in terms
of charge injection at the electrodes, electron transfer between electrode and XLPE in either di-
rection involving the same narrow 'window' of combined donor and acceptor states in the in-
sulator, centered on the Fermi level.

1. INTRODUCTION
T HE accumulation of space charge in the polymeric insulation of
electric power distribution cables can increase the local electric
field beyond the breakdown strength of the insulation, and thus cause
cable failure [l].Such space charge may result from electrons or holes
injected from the cathode or anode, electrons and positive ions gener-
ated by electric field-assisted thermal ionization of impurities in the bulk
or from an inhomogeneous volume polarization [2-51. Although space
charge accumulation occurs only to a very limited extent, if at all, un-
der ac electric stress, significant charge densities can be generated dur-
ing dc testing of ac cables before installation, leading to premature fail-
ure when ac stress is applied subsequently In recent years cable man-
ufacturers have modified the composition and morphology of the insu-
lation so that the accumulation under dc stress at room temperature is
reduced. However, with a move towards dc transmission, particularly
for long underwater links, and with emergency operating temperatures
of 125°C or higher, further research on methods of eliminating or at least
further reducing space charge accumulation is required.
Today, XLPE (cross-linked polyethylene) is the polymeric material
most widely used as insulation in power cables. The two main chem-
ical methods of crosslinking low density PE to produce XLPE use ei-
ther a peroxide (frequently DCP) which decomposes to form free rad-
icals, or a silane-based grafting mechanism. In the former method the
peroxide is incorporated into the polymer at elevated temperature and
pressure, and decomposes to yield free radicals which extract hydro-
gen atoms from the PE chains. The resulting reactive sites then com-
bine to effect crosslinking. The silane method [6] involves grafting a
chemically-multifunctional silane-based molecule onto the PE chains,
using a method which also requires a peroxide but at much lower
concentration than in the DCP method. After grafting, crosslinking is
achieved through hydrolysis or condensation reactions.
The DCP method is the longer established. It is preferred by some ca-
ble manufacturers on the grounds that, since the crosslinking reactions
in the silane method commonly occur at 90°C in a water bath, the result-
ing insulation is more liable to failure through water tree formation. On
the other hand some power authorities do not differential e between the
two processes for the manufacture of medium voltage cables. Indeed it
has been claimed that the concentration of water remaining in the silane
insulation after manufacture is lower than that remaining in the DCP ma-
terial [6].
In this paper we present data comparing space charge accumulation
in planar samples of DCP cross-linked and silane cross-linked XLPE, dc-
stressed at temperatures in the range 20 to 90°C. The space charge pro-
files were obtained using the LIPP technique. We also present the results
of thermal step depolarization measurements on the same samples.
2. EXPERIMENTAL
2.1. SAMPLE PREPARATION
XLPE planar samples measuring 5 cm x 5 cm x 0.5 mni were peeled
from lengths of commercial power cable, using a wide blade on a metal
turning lathe. Self-fusing semiconducting polymeric tape was attached
-
to both surfaces of each sample to form circular electrodes, 3 cm in diam-
eter and 50 pm thick. The semiconducting tape ensured good acous-
tic impedance matching at the electrode/insulator inter faces, thereby
minimizing acoustic reflections. A 2 mm thick aluminurn disc was at-
tached to one of the electrodes and acted as the target for the laser beam.

1070-9878/98/ $3.00 0 1998 IEEE

104 Bambery et al.: Space Charge Accumulation in Two Power Cable Grades of XLPE

It was painted with a black marker pen to increase the absorption of
the laser energy and hence increase the signal strength, and was always
grounded,
2.2. LlPP APPARATUS
The LIPP apparatus was as already described [7],except that the sam- 10 kV/mm.
ple holder was modified so that the target electrode could be heated and
its temperature recorded by an adjacent thermocouple (Figure 1).Tem-
perature control was achieved using a Eurotherm 815 controller con-
nected to the heater. The samples were heated only on the ground side
because of the practical difficulties of attaching a heater to the HV elec-
trode. Consequently a temperature gradient existed across the thickness
of the samples during measurement. The temperature difference be-
tween the target and HV electrodes increased linearly from 0°C at 20°C
to a maximum of 13°C at 90°C. In the remainder of this paper, quoted
temperatures relate to the target electrode.
Sample Electrodes
hielding
-
- peaks earlier than it did in the calibration trace (no space charge).
FIGURE1. Temperature controlled sample holder schematic.
2.3. SPACE CHARGE DENSITY
CA LIBRATI0N
The spatial resolution of the measurements, taken as the product of
-
the rise time of the pressure pulse and the speed of sound waves in the
sample, was 20 pm.
The traces presented below were all obtained at zero applied voltage.
Except for the calibration measurements, the applied field strength was
3. RESULTS
3.1. ROOM TEMPERATURE
MEASUREMENTS
The space charge distribution was periodically determined during
poling of the samples at room temperature, until it reached equilibrium
(typically after 2 days). This equilibrium distribution changed very little
when both electrodes were shorted to ground for 2 days. The external
field was then re-applied with reversed polarity until the space charge
distribution again reached equilibrium. This equilibrium distribution
was also stable at room temperature under short circuit.
3.1. I . DCP SAMPLES
Figure 2(a) shows a typical uncorrected trace obtained for a DCP sam-
ple poled with the target electrode as the anode. The electrode charge
peaks in the low-field calibration traces were taken as indicating the po-
sitions of the electrodes; Figure 2(a) and all subsequent traces have been
marked accordingly. It will be seen that the electrode charges appear
outside the electrodes. This is a consequence of the finite width of the
pressure pulse; as the leading edge of the pressure pulse enters the in-
sulation and interacts with the space charge present there, its center and
trailing edge are interacting simultaneouslywith the charge of opposite
sign induced on the electrode by the space charge. As a result, the signal
Figure 2(b)was obtained by correcting Figure 2(a)approximately for
the finite width of the pressure pulse, and its attenuation and dispersion
in transit across the sample. Note that the negative peak, appearing at a
distance of 0.08 mm from the target electrode in Figure 2(a), and due
N

Initial data obtained for each sample consisted of a LIPP trace, recorded to 'ringing' effects in the pressure pulse profile, has been eliminated by
at zero applied voltage to confirm the absence of space charge, and a sec- the correction process. All other traces below have been corrected sim-
ond trace recorded at low electric field (1.0 kV/mm) to determine the ilarly. Figure 2(c) shows the equilibrium distribution established after
pressure pulse profile and provide the space charge density calibration poling with the reverse field polarity.
data. The calibration method is based on the theory of the LIPP tech- These samples developed heterocharge in the volume immediately
nique outlined by Gerhard-Multhaupt [8]. The necessity for calibration adjacent to both electrodes, and the space charge profile is nearly the
at low field strengths has been detailed by Alquie et al. [9]. same but with opposite polarity after field polarity reversal (Figures 2(b)
The LIPP traces were recorded on a digital storage oscilloscope and and (c)).This reversal pattern was observed in many samples. A similar
then corrected approximately for the effects of finite width, attenuation result has been reported for XLPE-insulated cable samples by Wang et al.
and dispersion of the pressure pulse, using a procedure which will be [5]. We shall call this the MIE (mirror image effect) from now on.
described in detail in a later paper. To facilitate direct comparison of
3.1.2. SILANE SAMPLES
traces recorded at different temperatures, the variation of the speed of
the pressure pulse with temperature must be taken into account [lo]. Figure 3(a) shows a typical equilibrium space charge profile for a
To do this, the speed of the pressure pulse was determned at all tem- silane sample after poling with the target electrode as the cathode. Fig-
peratures by noting the time interval between the signals from the two ure 3(b) shows the profile after poling with the field polarity reversed.
electrodes at low applied voltage. The space charge distributions were In this sample homocharge accumulated immediately adjacent to both
then calibrated in terms of distance through the sample measured from electrodes. Again a near perfect inversion of the profile on reversing the
the target electrode. applied voltage polarity (MIE) is seen.
 :Ijy,>d
IEEE Transactions on Dielectrics and Electrical Insulation Vol.5 No. 1, February 1998 105

(a) lTargetl
m ................:........i......... L ........,.........
.-
= ........

3
P
0
3
m
Y
3 0
g ........ .........I.........:................. ..:.................
(b)
...............................................

...........................,......... .........

!!
-0 ............. L...

................ .! ....... .................

0
...:.........L ........ :..................

....:.-~~
i ~ ~ . :
..... ..... ..................

I 4 2

0.0 0.1 0.2 0.3 0.4 0.5

Distance (mm)
FIGURE3. (a) Space charge profile for a 500 pm thick planar silane
XLPE sample after poling at 10 kV/mm with the target electrode as
the anode. (b) Space charge profile for the same sample after poling
at 10 kV/mm with the target electrode as the cathode.
I I I
0.0 0.1 0.20.3 0.4 0.5
Distance (mm)
FIGURE2. (a) LIPP trace for a 500 pm thick planar DCP XLPE sam-
ple after poling at 10 kV/mm with the target electrode as the anode.
(b) Space charge profile determined from the previous trace after
approximate correction for the finite width, dispersion and atten-
uation of the pressure pulse. (c) Corrected space charge profile for
the same sample after poling at 10 kV/mm with the target electrode
as the cathode.
3.1.3.ELECTRIC FIELD PROFILE
The ’self‘ field E ( a ) of the space charge may be calculated by in-
tegrating the space charge density across the thickness of the sample.
Thus, starting from Poisson’s equation
........
dEk) -
~- - P(2)
(1) 0
dz EoEr
we obtain
W
Distance (mm)
FIGURE 4. (a) Space charge profile for a 500 pm thick planar DCP
where the ground electrode is at z = 0, the counter electrode (also at XLPE sample after poling at 10 kV/mm with the target electrode as
ground potential) is at z = d, p(z) is the space charge density, and E, the anode (identical to Figure 2(b)). (b) Self field of the space charge
and er have their usual meanings. E (a)values are then obtained using shown in (a).
piecewise integration starting from x = d. The result is shown in Fig-
ure 4(b)for the corrected space charge distribution shown in Figure 4(a);
the latter is identical to Figure 2(b).
of 2.5,5,10 and 20 kV/mm (both polarities) were used. The density of
3.1.4. FIELD DEPENDENCE OF SPACE space charge certainly increased with increasing field in both samples,
CHARGE DENSITY
but at present the volume of data is insufficient to allow us to specify the
The dependence of the equilibrium space charge profile on the ap- dependence more accurately, e.g. linear or higher power law. The MIE
plied field strength was studied briefly in two samples. Field strengths was seen at all field strengths.
106 Bambery et al.: Space Charge Accumulation in Two Power Cable Grades of XLPE

3.1.5. DIRECT CURRENT

MEASUREMENTS -
During room temperature poling the charging currents through two $4 A

samples of each type of XLPE were measured with an electrometer con- .PI
Vlk
nected between the target electrode and ground. These samples were 8&
equipped with a grounded guard ring to prevent possible surface cur- a&<
7
rents contributing to the measured current. After the space charge pro- 3%
file had reached equilibrium (- 48 h), the steady state conduction cur- S 3?
rent at a field of 10 k V / m was < 10 PA. The magnitude of this current
was independent of the polarity of the applied voltage, within experi- I I I I
mental error, for a given sample. 0.0 0.1 0.2 0.3 0.4 0.5
Distance (mm)
FIGURE6. Space charge profiles for a 500 pm thick planar silane
3.2. TH ERMAL DEP0LA RlZATl0N XLPE sample in thermal equilibrium at various temperatures be-
MEASUREMENTS tween 30 and 90"C,following poling at 10 kV/mm with the tar-
get electrode as the cathode. The dashed horizontal line running
Thermal step discharge measurements were made on identically pre- through each trace is the zero for that trace.
pared DCP and silane samples. The experimental procedure was as fol- 4. DISCUSSION
lows.
4.1. POSSIBLE SPACE CHARGE
(1)The field was applied at 90°C for 1h, after which the sample was ACCUMULATION MECHANISMS
cooled to 30°C and maintained at that temperature for 1h with the field
still applied (to allow thermal equilibrium to be established). The principal characteristics of these data are as follows:

(2) The sample was shorted to ground and a LIPP trace obtained. 1. the space charge profile reached an equilibrium density within
48 h,
(3)Further measurements at zero applied field were made at 10°C in- 2. small steady conduction currents were measured in the samples
tervals to 90"C,the sample having stood at each temperature for 10 min 48 h after application of the field,
before the measurement was made. 3. the equilibrium space charge density close to the cathode was not
equal to that close to the cathode, for either applied field polarity,
Typical results for a DCP and a silane sample are presented in Fig- 4. nearly all the space charge was formed immediately adjacent to
ures 5 and 6 respectively. At 30°C heterocharge accumulated in the DCP the electrodes,
sample and homocharge in the silane sample. In both cases the charge 5. reversal of the applied field polarity led to a very similar equilib-
gradually disappeared when the samples were heated in short circuit. rium charge profile and density, but with opposite polarity (Fig-
It is of interest to note that, in Figures 5 and 6, the space charge den- ures 2 and 3), i.e. the MIE, and
sity accumulating adjacent to the HV electrode is equal to or larger than 6. the equilibrium charge density increased with increasing applied
that adjacent to the target electrode, contrary to the situation shown in field strength.
Figures 2 and 3.
Together, these characteristics severely limit the possible mechanisms
of space charge accumulation.
4.2. INHOMOGENEOUS
R2
h
MOLECULAR DIPOLE
ua
.I
CONCENTRATIONS.
ulti
E&
a m A model involving an inhomogeneous distribution of molecular dipoles
us through the sample thickness is initially appealing, i.e. one might expect
$3
A that alignment of the dipoles in the opposite direction when the field is
02 v reversed would lead to the MIE.
The dipole density distribution must be inhomogeneous, leading to a
0.0 0.1 0.2 0.3 0.4 0.5 polarization gradient. Then the total charge density p(x) in the sample
Distance (mm) volume is given by
FIGURE 5. Space charge profiles for a 500 pm thick planar DCP
XLPE sample in thermal equilibrium at various temperatures be- (3)
tween 30 and 90"C,following poling at 10 kV/mm with the tar- where ps is the usual (monopolar) space charge and P is the polariza-
get electrode as the cathode. The dashed horizontal line running tion due to dipole alignment. -dP/dz is an apparent volume charge
through each trace is the zero for that trace. density [11], originating in the electric field distribution generated by
 ~

IEEE Transactions on Dielectrics and Electrical Insulation Vol. 5 No. 1, February 1998 107

the dipoles. Following reversal of the external field polarity, the appar- Lewis [12]has shown that electron transfer by tunneling across a con-
ent space charge density profile will reverse polarity simply because the ductor/insulator interface, in either direction, will involve only a nar-
dipoles will orient in the opposite direction. Thus the MIE would be ex- row 'window' of combined donor and acceptor states in the insulator,
pected provided p8)(. was negligible. centered on the Fermi level, and the states within the same energy range
Apparent homocharge or heterocharge would be generated in the in the conductor. Since we observed approximately equal steady state
bulk, depending on the direction of the polarization gradient. How- conduction currents under opposite applied voltage polarities, it fol-
ever, the major problem with this model is that it is not consistent with lows that the equilibrium barrier heights and widths for electron extrac-
induced charge on the electrodes. Homogeneous and inhomogeneous tion and injection (via tunneling) at a given electrode must be equal.
polarization in the bulk both generate heterocharge on the insulator sur- This can be seen from the expression for the transfer ciirrent density
face, but these surface charge layers would almost certainly be indistin- based on Fermi's Golden Rule, Equation (17) in [12], which incorpo-
guishable from the equal and opposite induced charge on the electrodes, rates the quantum mechanical matrix element for tunneling. Since the
because their separation would be much less than the spatial resolution height and width of a given barrier in equilibrium will depend on the
limit of the LIPP technique. Thus no charge would be detected on the local electric field, and therefore on the local space charge density, we
electrodes, contrary to observation. expect equal amounts of negative and positive space charge adjacent to
a given electrode when the applied voltage polarity is reversed, i.e. the
4.3. CARRIER INJECTION AND space charge accumulation mechanism outlined above leads to the MIE.
IMPURITY IONIZATION
The discussion above relates specifically to the homocharge case. If a
As stated in the Introduction, space charge in XLPE is often consid- sample begins to develop heterocharge when the field is first applied, a
ered to be the result of (a) electron injection or extraction at the electrode corresponding sequence of events will occur, leading to similar equilib-
or insulation interfaces, or (b) ionization of impurities in the bulk fol- rium states, except that, when the field is reversed, electrons will have
lowed by electron or ion separation. Our preferred model leading to the to be transported across the sample before charge neutralization at the
MIE incorporates both of these mechanisms. Electron injection or extrac- electrodes can begin to occur.
tion at the electrodes is assumed to occur through quantum mechanical
tunneling, and the positive ions created by ionization of impurities are 4.3.2. HOMOCHARGE AND
assumed to have negligible mobility HETEROCHARGE
4.3.1. ORIGIN OF THE MIRROR IMAGE Whether a given sample develops homocharge or heterocharge, when
EFFECT the field is first applied, depends mainly on the mobility of electrons in
Consider an insulator without space charge. When a field is applied, the bulk. If this mobility is sufficiently large to ensure that electrons
electron injection from the cathode into the insulator, and ionization liberated by ionization of impurities close to the cathode, and those in-
of impurities within it, occur simultaneously. Suppose negative space jected into the insulator from the cathode, are fairly rapidly transported
charge begins to accumulate in the insulator adjacent to the cathode (ho- towards the anode, heterocharge will form close to the cathode. Het-
mocharge). erocharge will also form at the anode provided the electron extraction
It follows that the electron injection rate at the cathode ( z = 0) must rate is less than the arrival rate. This again implies a cerfain minimum
exceed the electron transport rate across a surface z = 6, where 6 is electron mobility If the mobility is too low, homocharge will form.
the minimum distance from the cathode at which the electron transport DCP XLPE contains significantly higher concentration; of crosslink-
properties are characteristic of the bulk of the insulator rather than its ing byproducts than silane XLPE [13]. These volatile byproducts include
surface. Simultaneously, positive space charge will begin to accumulate acetophenone, a-methyl styrene and cumyl alcohol. Li and Takada [14]
adjacent to the anode, due to impurity ionization and electron extraction found a change from homocharge to heterocharge accurnulation with
in that region. increasing impurity content in LDPE samples. Suh et al. [15] reported
The accumulation of negative space charge adjacent to the cathode similar findings for increasing cross-linking by-product concentrations
will increase the height of the barrier through which the electrons tun- in XLPE, and suggested that these impurities may become charged and
nel to enter the insulator, and so the injection rate will fall. Similarly, migrate to the counter electrode under dc stress. Aida et al. [16] observed
the barrier height for electron extraction at the anode will increase, and increased dc conduction currents in LDPE doped with 1%wt acetophe-
the extraction rate will fall. Eventually a steady state will be reached in none compared to undoped LDPE.Our observation that heterocharge
which the injection and extraction rates are equal to the transport rate formation occurs in DCP XLPE suggests that the higher concentrations of
in the bulk sample. This will correspond to a steady state conduction crosslinking byproducts in this material increase the electron mobility
current, and at this stage space charge accumulation will cease at both in the bulk. They may do so by providing additional donor or accep-
electrodes. tor sites for hopping conduction, or they may form intermolecular links
Suppose the applied voltage polarity is now reversed. The accumu- which enhance the connectivity of the electron transport network [17].
lated charge adjacent to the electrodes will begin to disappear and be
4.3.3. ELECTRON MOBILITY
replaced with charge of the opposite polarity Again a steady state will
eventually be reached, characterized by a constant conduction current Little or no charge was detected in the bulk of the samples, strongly
and constant space charge density adjacent to the electrodes. suggesting a high electron or ion geminate recombinaticn probability
108 Bambery et al.: Space Charge Accumulation in Two Power Cable Grades of XLPE
there. However, the fact that a steady state current was measured indi-
cates that electrons were being transported from cathode to anode. As-
suming that the minimum detectable charge density was 0.1 pC cm-3
(based on experimental s/n ratios), then a steady state current density
of 1 . 4 ~ 1 A0cmV2 ~ ~ at~ a field strength of 10 kV/mm implies a lower
bound of 1.4~10-~~ cm2V-'s- for electron mobility (including the
effects of trapping) in the bulk. Electron mobilities in XLPE do not ap-
pear to have been reported in the literature. Those reported for LDPE fall
into two groups, one around l o p 4 cm2V-'s-l [18,19]at room temper-
ature, and the other in the range to cm2V-ls-l [20,21]
at room temperature and at 80 to 95°C. The first group of authors [18,
191 appear to be reporting mobilities unaffected by trapping.
4.3.4. INTERFACES AND BARRIER
HEIGHTS
Modification of the effectivebarrier height for electron injection or
extraction at the electrodes is an important part of the space charge ac-
cumulation mechanism proposed above. The approximate magnitude
of the barrier height reduction (in the case of homocharge accumulation)
in the present samples, due to accumulation of space charge in the in-
sulator, is readily estimated. According to the Schottky injection model
[221,
J=AT2exp
[ I?;@
,,/=
--
(4)
D@=
47CE0ET
where J is the steady state conduction current density, @ the unper-
furbed barrier height, A@the reduction in the barrier height due to the
field E at the electrode/insulator interface, A = 1 . 2 ~ 1 Am-2K-2
0~
and the other symbols have their usual meanings. Taking E = 10 kV/"
and E~ = 2.3'(for XLPE) gives A@ = 0.079 e? If E isiow reduced to
9 kV/mm (typically) as a result of space charge accumulation adjacent
to an electrode, A@ = 0.075 ell A fall of 0.004 eV in A@(equivalent
to an increased barrier height) would lead to a reduction of 14%in
N
the steady state conduction current, at room temperature. It should be
remembered that this reduction is due exclusively to accumulation of
-
space charge adjacent to an electrode. In practice, much larger reduc-
tions are frequently observed during the first 3 h after application of
the voltage, but these are probably due to other effects, e.g. aging of the
electrode/insulator interface and dipole orientation.
The unperturbed barrier heights at the two electrode/insulator inter-
faces are unlikely to be equal, due to microscopic morphological differ-
ences across the contact area. Thus equal amounts of space charge at
both electrodes for a given field polarity would not be expected. Ac-
cordingly Figures 2 and 3 show higher space charge density adjacent to
the target electrode, Figure 5 shows approximately equal densities, and
Figure 6 shows higher density adjacent to the HV electrode. Integrating
over all the samples, there was no bias towards either electrode.
5. CONCLUSIONS
The following conclusions may be drawn from this work.
1. The crosslinking method appears to determine the polarity of
the space charge adjacent to the electrodes. Specifically, DCP
crosslinking favors heterocharge, and silane crosslinking favors
homocharge. However, the charge densities do not differ greatly.
2. A near-perfect inversion of the equilibrium space charge profiles,
observed on reversing the applied voltage polarity, is ascribed to
electron transfer by tunneling across the electrode/insulator in-
terfaces. Trus transfer, in either direction, involves the same nar-
row window of combined donor and acceptor states in the insu-
lator, centered on the Fermi level.
3. An estimated lower bound for the electron mobility in the bulk
of the samples is 1 . 4 ~ 1cm2VV1s-l.
0 ~ ~ ~
ACKNOWLEDGMENT
The authors wish to thank MM Cables, Sydney for providing lengths
of XLPE power cable from which the samples were cut. KRB acknowl-
edges the award of an Australian Postgraduate Research Award (Indus-
try), and RJF thanks the Australian Research COuncil for financial SUP-
port to purchase the laser used in the n ~ ~ ~ r e m e n t s .
REFERENCES
[l] T. Tanaka and A. Greenwood, Advanced Powev CabJe Technology, ch. 2. 6, CRC Press
Inc. Boca Raton, Florida, 1983.
[2] J. Lewiner, "Evolution of Experimental Techniques for the Study of the Electrical
Properties of Insulating Materials" IEEE Trans. on Electr. Jnsul., Vol. 21, pp. 351-360,
1986.
[3] T. Mizutani, "Space Charge Measurement Techniques and Space Charge in
Polyethylene", IEEE Trans. Diel. and Electr. Insul., Vol. 1, pp. 923-933,1994.
[4] G. M. Sessler, "Charge Distribution and Transport in Polymers", IEEE Trans. Diel.
and Electr. hSul., vol. 4, pp, 61&62&,1997.
[5] X. Wang, D. Tu, Y.Tanaka, T. Muronaka, T. Takada, C. Shinoda and T. Hashizumi,
"Space Charge in XLPE Power Cable under dc Electrical Stress and Heat Treatment",
IEEE Trans. Diel. and Electr. Insul., Vol. 2,pp. 467-474,1995.
[6] A. D. Nichols, "The Manufacture and Application of Cables by the Monosil Process",
1990Conference (Vol. 60),ElectricitySupply Engineers Association of New Zealand,
Wellington, pp. 100-116,1990.
[7] R. J. Fleming, M. Mdby Henriksen, M. Henriksen and J. T. Holball, "LIPP and
PEA Space-Charge Measurements in LDPE" Proc. 5th Int. Conf. on Conduction and
Breakdown in Solid Dielectrics, IEEE, Leicester UK, pp. 234238,1995.
[SI R. Gerhard-Multhaupt, "Analysis of Pressure-WaveMethods for the Nondestructive
Determination of Spatial Charge or Field Distributions in Dielectrics", Phys Rev B
Vol. 27, pp. 24942503,1983.
[9] C. Alquie, T. Ditchi and J, Lewiner, "Measurement of Charge Distributions in Insu-
lating Materials for Hv Cables", IEE Int. Conf. on Properties and Applications of
Dielectric Materials, Beijing, pp. 434-438,1988.
[lo] K. R. Bambery and R. J. Fleming, "The Temperature Dependence of Space Charge
Accumulation in Cross-Linked Polyethylene", J. Thermal Analyis, Vol. 50, pp. 19-
31,1997.
[ll] B. I. Bleaney and B. Bleaney, Electricity and Magnetism, Ch. 1, Oxford University
Press, 1963.
[12] T. J. Lewis, "Electrical Effects at Interfaces and Surfaces", IEEE Trans. on Electr. In-
Sul., Vol. 21, pp. 289-295,1986.
[13] A. W. Field and D. McAllister in Electric Cables Handbook, eds. E. W. G. Bungay and
D. McAllister, CRC Press, Inc. Boca Raton, Florida 2nd edition. Ch. 2,1990.
[14] Y.Li and T.Takada, "Space Charge Distribution in Multi-ply LDPE", Conf. on Elec-
trical Insulation and DielectricPhenomena (CEIDP),IEEE, Victoria Canada, pp. 397-
402,1992.
[15] K. S.Suh,S.J. Hwang, J. S. Noh and T. Takada, "Effects of Constituents of XLPE on
the Formation of Space Charge" IEEE Trans. Diel. and Electr. Insul., Vol. 1, pp. 1077-
1083,1994.
IEEE Transactions on Dielectrics and Electrical Insulation Vol. 5 No. 1, February 1998 109

[16] E Aida, G Tanimoto, M. Aihara and E. Hosokawa, ”Influenceof Curing by-products [21] E. H. Martin and J. Hirsch, “Charge Transport and Carrier Mobilities in Insulating
on Dielectric Loss in XLPE Insulation” Conf. on Electrical Insulation and Dielectric Polymers”, J. Non-Crystallie Solids, Vol. 4, pp. 133-137,1970.
Phenomena (CEIDP),IEEE, Pennsylvania, pp. 465473,1990. [22] L. A. Dissado and J. C. Fothergill, Electrical Degradation and Breakdown in Polymers,
[I71 T. J, Lewis, ”A Simple General Model for Charge Transfer inl’olymers”, Faraday Dis- Ch. 9, Peter Peregrinus Ltd., London, 1992.
cuss. Chem. Soc., Vol. 88, pp. 189-201,1989.
This manuscript is based on a paper given at the 9th International Symposium
[18] D. K. Das Gupta and T. Noon, ”Phototransient Spectral Shifts in Polyethylene with
on Electrets, Shanghai, China, 25-27Sepfember 1996.
High Fields”, J. Phys. D. Appl. Phys., Vol. 8, pp. 1333-1340,1975.
1191 T Tanaka and J, H, Caldewood, ”Transient Currents and Electron Mobility in khUSCT@tWas received On 18 October 1997, in revised form ;?I January 1998.
Polyethylene”,J. Phys. D.Appl. Phys., Vol. 7, pp. 1295-1302,1974.
[20] H. J. Wintle, ”Decay of Static Electrification by Conduction Processes in
Polyethylene”, J. Applied Phys. Vol. 41, pp. 4004-4007,1970.