Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4659  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

Foreign Inclusions, Enclosed Cavities, Partial 
Discharge Models and Discharge Energy: A Short 

Review Regarding Solid Dielectrics and Composite 
Insulating Systems 

 

M. G. Danikas  
Democritus University of Thrace 

Department of Electrical and Computer Engineering 
Power Systems Laboratory 

Xanthi, Greece 
mdanikas@ee.duth.gr 

S. Morsalin 
University of New South Wales 

School of Electrical Engineering and Telecommunications 
High Voltage Laboratory 

Sydney, Australia 
sayidul.morsalin@unsw.edu.au

 

 

Abstract—Partial discharges in cavities are one of the main 
reasons of breakdown of insulating materials and insulating 
systems in general. The present paper aims at commenting some 
aspects of the relation between partial discharges and insulation 
damage. Some basic models explaining the workings and impacts 
of discharges in cavities are discussed. Discussions about the 
possibility of damage in the presence of multiple cavities as well 
as of the risks of minute discharges and/or charging phenomena 
below the so-called inception voltage are conducted. It is 
remarked that the question of the relationship between discharge 
parameters and insulation damage is still an important one, 
especially for composite insulating systems. 

Keywords-solid dielectrics; partial discharges; inception 
voltage; insulation damage; discharge magnitude; discharge 
energy; charging phenomena 

I. INTRODUCTION  
  Partial discharge is the name given to an electrical 

discharge, which involves only a portion of the dielectric 
without fully bridging the electrodes [1]. Partial discharges 
often happen in enclosed cavities of insulation and thus they 
are sometimes called internal partial discharges. They can be 
extremely harmful to solid insulating materials and they start 
appearing when a cavity is subjected to an AC stress greater 
than its breakdown value. They may occur regularly at each 
half period of an applied sinusoidal voltage [2]. Partial 
discharges in a dielectric develop in enclosed cavities (or 
because of inclusions) of lower dielectric strength than the 
dielectric strength of the surrounding material. The insulation 
in the inclusion breaks down at a stress lower than the 
dielectric strength of the surrounding material. The voltage at 
which discharges start occurring regularly in an enclosed cavity 
is called inception voltage [3]. When the voltage is reduced 
under a certain limit, discharges stop occurring. This limit is 
thus called extinction voltage [3] (all the above under the 
assumption that AC voltage is applied). Enclosed cavities may 
be found in extruded plastics, lapped impregnated paper and 

cast resins, they may be filled with oil (as in layers and in butt 
gaps of oil-impregnated paper insulation) or may consist – as 
inclusions - of various foreign particles (like textile fibers or 
dirt).  It is the aim of the present paper to give a review 
regarding the problems arising because of partial discharges 
and their injurious effects on the insulating materials. However, 
the present paper is by no means an exhaustive review of all of 
the aspects of partial discharges and their deleterious effects. 

II. PARTIAL DISCHARGE MODELS 
Partial discharges in enclosed cavities may be represented 

with the well known a-b-c model (or capacitor model) [4, 5]. 
This model is based on the assumption that the enclosed cavity 
may be represented as capacitance, while the neighboring to the 
cavity healthy insulation may be represented by another 
capacitance and yet another capacitance may represent the rest 
of the insulating material. In such a case, the apparent charge Q 
detected by a PD detector may be given as: 

Q = Cb·ΔVc       (1) 

where, Cb is the capacitance of the healthy insulation 
adjacent to the cavity and ΔVc the voltage change that is caused 
because of the discharge inside the cavity. This quantity, 
although not determined by the dimensions of the cavity alone, 
as Cb is affected by the insulation thickness, is considered 
approximately a right measure of a discharge [6]. Related to the 
abc model is also the energy of an individual discharge, which 
is given as: 

W = 0.7·Q·Vi       (2) 

where, Q is the apparent charge of an individual discharge 
and Vi the discharge inception voltage [6]. 

Such a model, although very useful for many decades, was 
subjected to criticism [7], where it was opined that the notion 
of capacitance is limited to metal electrodes and not to 
insulating surfaces facing each other. Thus, another model was 

Corresponding author: M. G. Danikas 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4660  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

proposed, named Pedersen’s model [7], based on 
electromagnetic theory: 

Q = k·Ω·εr·ε0·(Εi – Εl)∇λ0       (3) 
where, Q is the charge induced on the measuring electrode 

by the partial discharge in a cavity, k the cavity geometric 
factor, Ω the cavity volume,  Εi the inception electric field for 
streamer inception, Εl the limiting electric field for ionization,  
εr the relative permittivity of the surrounding solid insulating 
material,  ε0 the permittivity of the free space and  ∇λ0 the 
function which gives the ratio of the electric field at the 
position of the cavity (in the absence of the cavity) to the 
voltage between the electrodes. The quantity (Εi–Εl) is 
calculated in terms of the streamer criterion, i.e.:  

(Εi/p) = (1+B/√2·a·p) Εl/p       (4) 
where, B is a constant characteristic for the gas in the 

cavity, a the radius of the cavity and p the pressure in the 
cavity. As discussed in [8], the merit of (3) is that the charge Q 
is related to a variety of parameters. This model was referring 
to streamer type discharges.   Such a model was preceded by 
the model proposed in [9].  According to that model, the 
apparent charge Q is given from the following equation: 

Q=εr·ε0·E2·γ·V/la       (5) 

where εr and ε0 as in (3), V is the volume of the cavity, γ is 
the λ0, E2 is the partial discharge inception voltage according to 
Paschen’s law and la the geometry factor depending on the axis 
ratio of the cavity. It is evident from (5) that this model refers 
mainly to Townsend type of discharges (since it considers the 
partial discharge inception stress from Paschen’s law 
viewpoint). In actual fact, the models of [7] and [9] have a 
number of common parameters. 

Criticism of Pedersen’s model came from [8, 10], where it 
was pointed out that  Pedersen’s model is valid for initial 
experimental conditions and that (3), which is vital to the 
model, assumes a breakdown in a uniform gap (cavity 
breakdown, however, is not, strictly speaking breakdown, but it 
leads to redistribution of charges on the surface of the cavity). 
Subsequent criticism came from [11], where it was indicated 
that a direct correlation between the space charges and the 
change of capacitance exists and that the criticism against the 
change of the capacitance of the system in relation to space 
charges has been opposed by utilizing Pedersen’s equations. 
However, it must be mentioned that –as far as some 
experimental results with enclosed cavities in polymers are 
concerned– Pedersen’s model fared quite well (in comparison 
to the capacitor model) and gave closer theoretical discharge 
values to the experimental ones [12-14]. On the other hand, one 
should point out that in these experiments [12-14], only partial 
discharges under initial experimental conditions were 
investigated.  

III. CONSEQUENCES OF PARTIAL DISCHARGES 
Quite early, there were experiments with plane discs from 

clean polyethylene having small cylindrical cavities and a 
uniform electrode arrangement [15]. The author of [15] 
calculated that about 10-15 cm3 of polyethylene was eroded by 
each discharge having a magnitude of 10 pC. The deterioration 

was dramatically increased if the applied voltage was raised 
above the inception voltage. In yet another work [16], the same 
researcher pointed out that among the main factors affecting 
the rate of deterioration of an insulating material are the ratio of 
the applied voltage to the discharge inception voltage, the 
magnitude and energy of the discharges, the waveform and 
frequency of the applied voltage, the resistance of the material 
both to the discharge erosion and to chemical attack by the 
byproducts generated by the discharge and finally the electrical 
and chemical characteristics of the ambient medium. 
(resonance of the work – mentioned in [15, 16] - on the effect 
of the waveform of the applied voltage can be found in [17] 
and also in a more recent work [18], where there was an 
investigation on partial discharges in rotating machine 
insulation). The energy of discharges affects the local 
temperature rise at the point of impact of the discharges, and 
subsequently the value of the local dielectric strength is 
attained [19]. Differences in discharge characteristics and time 
to breakdown arise if the cavity is (a) sealed or ventilated, and 
(b) surrounded by the insulation or adjacent to a conductor. 
Moreover, conditions of pressure and temperature used during 
the processing of polymers (molding, curing) may greatly 
affect their “molecular packing” and consequently their 
resistance to discharge channels [20].  

Investigation of the behavior of discharges in air gaps 
facing solid insulation indicated that an increase of discharge 
magnitude with increasing voltage, which was explained in 
terms of a continuous recombination and neutralization of 
deposited surface charges which will reduce the shielding 
effect [21]. Such a phenomenon will keep the discharge going 
and will cause a fresh and greater generation of surface charge. 
Testing artificial and natural cavities in insulating materials 
(polyethylene and polyethylene terephthalate), the author of 
[22] found that the inception voltage of natural cavities formed 
in polyethylene was two to four times the calculated inception 
voltage, assuming that the electric field is uniform in the cavity. 
Oxalic acid crystals and water vapor, when present in the 
cavity, are factors of serious deterioration of the material. 
Confirmation of the above can be seen in [23], where it was 
stated that chemical byproducts contribute to charge deposition 
on the cavity inner walls and, thus, to the modification of 
discharge magnitudes in subsequent discharges. 

Internal cavities with discharges cause lesser damage 
compared to cavities undergoing discharges, which are 
adjacent to electrodes [24]. This has been observed also in [15]. 
Discharges in cavities with a large diameter/depth ratio cannot 
extinguish as in cavities with a small diameter/depth ratio. The 
former type of cavity is likely to have the more injurious effect 
on the insulation. In [23], it was reported that for very small 
cavities, the inception voltage decreases as the cavity depth 
increases, following the Paschen curve of breakdown, thus 
agreeing with [24]. The increase of the number of discharges 
per unit time with frequency, the increase of the number of 
discharges with increasing electric stress in the dielectric, the 
increase discharge magnitude with increasing depth of the 
cavity as well as with its area, all these were observed 
experimentally in [6]. Recent research indicated that, although 
an increase in cavity diameter causes more discharges of large 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4661  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

magnitude, an increase in cavity depth causes an increase in 
inception voltage [25]. This conclusion is at variance with [23]. 

The authors of [26], working with a non-uniform electrode 
system and polyethylene samples, observed that the main 
chemical change occurring in polyethylene when exposed to 
discharge activity are crosslinking, increase in unsaturation and 
hydrogen evolution. The rate of chemical change depends on 
the total discharge energy and concentration of end products. 
The total volume of hydrogen V (in ml) in a system at a time of 
t seconds is empirically given by: 

V=√2·K·t       (6) 
where, K is the hydrogen evolution coefficient which is 

directly proportional to the total discharge energy per cycle. 
Furthermore, in [26] it was pointed out that semiconducting 
surface films play a significant role in retarding the ultimate 
failure of the insulation, agreeing in this respect with [6]. 

The discharge rate rises almost linearly with the applied 
voltage irrespective of the gap setting and vapor pressure in a 
spark gap [27]. It was also reported that the discharge process 
in artificial cavities is more constant than that in natural 
cavities. The relation between discharge energy and insulation 
damage is a subject which was and is a vital one for the 
insulation community. In [28], it was thought preferable to 
measure continuously both discharge magnitudes and repetition 
rates in order to better correlate discharge damage with 
discharge activity rather than measure either the maximum 
discharge magnitude or the total discharge energy. Later 
research by the same author [29] confirmed the above 
conclusions with a variety of insulating materials, such as 
epoxy resin, polyethylene and mica under AC conditions. 

Research in internal discharges in artificial cavities in 
polyethylene [30], reported that discharges generally decrease 
in magnitude and repetition rate with time and that absorbed 
water can greatly influence the pattern of discharges. The 
decrease of repetition rate of discharges with time happens 
because of a voltage decrease on the side wall due to the 
decline of the side wall resistance in the cylindrical cavity. The 
apparent charge of maximum discharge (Qa) in a cylindrical 
cavity is given by 

Qa=ε0·ε
*· (VGO·S)/(D–d)        (7) 

where, ε*  is the specific dielectric constant of the material, 
D is the thickness of the whole specimen and d the cavity 
depth, S being the area of the cavity and VGO the inception 
voltage of the cavity. Qa decreases with time because the 
discharging area is narrowed by the development of low 
resistance to inner top and bottom surfaces. On the same line of 
research, it was reported that the impulse inception stress 
increases with decreasing cavity diameter at constant cavity 
depth as well as with decreasing depth at constant diameter 
[31]. The inner workings of discharges in enclosed cavities 
were the questions tackled in [32], where it was observed that 
the nature of internal discharges is greatly affected by the 
assembly of the electrode system and the adhesion of the films 
and that such discharges become unstable with time, resulting 
thus in extremely long lifetimes. Such observations were 
confirmed in [33], where it was reported that the internal 

surface of the cavity is subjected to physical damage or 
erosion, or a change in surface electrical resistivity. Pressure 
change may also occur, again affecting internal discharge 
activity. Such factors are responsible for the apparent 
characteristics of discharges in cavities, i.e. their intermittency, 
cessation and resumption. Furthermore, in [34] it was pointed 
out by the same authors that gas pressure plays a determining 
role for internal discharges and, moreover, the question of 
multiple cavities formed by thermal expansion in extruded 
cables was discussed. 

Experimenting with different arrangements and liquid 
nitrogen impregnated taped cable insulation, some researchers 
[35] postulating that discharges occur within bubbles and 
indicated that the logarithm of the discharge magnitude 
increases linearly with stress. They proposed that the applied 
voltage Va, the inception voltage Vi, M the discharge 
magnitude and K a constant, are related through the following 
equation: 

Va=Vi+K·lnM       (8) 

Performing similar experiments, the authors of [36] 
indicated that the discharge inception voltage depends on the 
product of butt gap depth and helium density. Discharges start 
from a point in a butt gap. The probability of the presence of 
such a point depends on the gap area. In yet another study with 
pressboard/oil insulation [37], it was shown that, whereas an 
increase in the magnitudes of discharges by increasing the 
applied voltage is possible, a decrease of discharge magnitudes 
is also possible owing to the volume of discharges taking place 
in gas bubbles in such an arrangement. Discharges start always 
in the electrically weaker medium, agreeing in this respect with 
[36]. Moreover, deterioration of the surface of the pressboard is 
in most cases a necessary condition in order that dangerous 
discharges will appear in its volume causing thus ultimate 
failure. 

The dependence of discharge inception voltage both on 
pressure inside the cavity and on cavity diameter was noted in 
[38]. In fact, it was noted that the discharge inception voltage 
decreased at low pressures and with increasing cavity diameter, 
whereas the extinction voltage depends on the conditions of 
charges trapped on the surface of the dielectric. Similar 
observations were made some years later [39]. In [39], it was 
also reported that, under the assumption that the insulation is 
subjected to combined electrical and thermal stresses, a large 
cavity gas pressure reduction rate is accompanied by a short 
time-to-breakdown, whereas a small pressure reduction rate is 
typical for long times-to-breakdown. In other words, the cavity 
gas pressure does reflect the ageing rate of polymer specimens. 
Some researchers noted also a dependence of both inception 
and extinction voltages on pressure [40]. Electrode design and 
material processing play also a deciding role according to [40], 
something also noticed in [32].  

A direct relationship between discharge intensity and the 
rate at which deterioration is taking place in an enclosed cavity 
was noted in [41]. This relationship was confirmed in yet 
another study by the same author [42], where it was also 
observed that discharges initially of large magnitudes decrease 
with time. Such conclusions were in agreement with [43], that 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4662  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

suggested that the area associated with each discharge 
decreases with time and this can happen if the discharges occur 
between sites where degradation products accumulate. As the 
degradation products spread over the surface of the cavity, 
smaller areas between them are available to discharges. Further 
confirmation of the above was given in [44], where it was 
reported that the cavity area affected by a discharge is usually 
much smaller than the area of the whole cavity in which it 
occurs, and consequently the discharge site is not uniformly 
discharged. 

A study of gaseous degradation products of corona exposed 
low density polyethylene showed that two degenerative  
processes were in action [45]: a) crosslinking of polyethylene 
which may be due to ultraviolet radiation and electrons, 
crosslinking being evidenced by hydrogen evolution, and b) 
erosion of polyethylene, which may be due to ion 
bombardment resulting in organic gases being evolved (CO, 
CO2, CH4). Higher temperatures evolve smaller amounts of 
organic gases since at higher temperatures molecules can 
dissipate more of the mechanical energy through segmental 
motion rather than by means of bond breaking processes (e.g. 
erosion). In [46], the effect of polarity for cylindrical cavities is 
remarkable only when the applied voltage is equal to the 
inception voltage, otherwise, as the applied voltage becomes 
three or four times larger than the inception voltage, the 
polarity effect almost disappears. This can be explained as at 
higher voltages, important secondary phenomena, such as 
photoionization and dissociation of negative ions formed by the 
preceding discharges, originate from the gas phase and not 
from the metal electrodes. 

The subject of the relation between discharge energy and 
insulation damage was treated in [47]. The authors found a 
relationship between the loss of weight and the dissipated 
power of discharges. They rejected the notion of the 
importance of isolated discharges and their relation with 
insulation damage. Their work was in agreement with [29, 48]. 
In [49], discharge magnitudes were shown to decrease with the 
thickness of cavity, whereas they tend to increase with the 
cavity diameter. The experimental data of [49] are in 
accordance with those published in [24] but they were at 
variance with those published in [6]. An approach similar to the 
one adopted in [47] was also adopted in [50, 51], where, 
investigating a complex insulation system (consisted of 
Uniaxially Orientated Polyethylene Tapes (UOPE) and 
Doecylbenzene oil, the total discharge energy was correlated 
quite well with the surface area damage of the UOPE tapes (in 
fact the correlation coefficient between the aforementioned two 
quantities was in the region of 0.71–0.97, which is considered 
to be quite satisfactory). It is true, however, that, in such 
complex insulating systems, besides the surface area damage in 
the aforementioned work, other deteriorating mechanisms may 
take place, such as evolution of gases [52]. Moreover, different 
types of discharges may cause increase or decrease of the 
gaseous evolution [52]. A more recent publication [53], is in 
accordance with [52] (albeit using a different electrode 
arrangement and different insulating materials), pointing out 
that three types of internal discharges are identified during 
ageing tests, namely, the first stage of discharges is linked to 
the streamer-type, the second stage to the Townsend-type and 

the third stage to the swarming pulsive micro-discharges 
(SPMD). 

The questions of sample thickness, cavity size and partial 
discharges were treated in [54]. It was shown that with larger 
specimens, the probability of having bigger cavities increases, 
and hence a longer free path for the ionization products. 
Thicker specimens have normally higher discharge activity 
than thinner ones, for the same value of applied electric field. 
Relevant to the above, and confirming [54], is work performed 
some years earlier, where it was reported that thicker insulation 
needs more sensitive discharge detection equipment. Moreover, 
the latter conclusion is valid independently of the sort of the 
enclosed cavity in the insulation, and independently of the 
discharge model used for the interpretation of the discharge 
events inside the cavity [55, 56]. Deteriorating results because 
of discharge activity in enclosed cavities were noted in [57], 
where it was reported that the rate of deterioration increases 
with an increase in concentration of discharges. Deep erosion 
may ensue with an instability in the dielectric-gas interface. 
Such a phenomenon may have as a result subsequent 
discharges to the same site. The observations of [57] were 
confirmed in a very recent paper [58]. Further confirmation of 
the above was offered in [59], where the authors carried out a 
micro-analysis of chemical byproducts regarding the local 
erosion in enclosed cavities. Localized spots are determined by 
their chemical byproducts. Oxidative degradation mechanisms 
are at work and as such they change the discharge patterns.  

Important observations regarding the creation of cavities 
from enclosed conductive contaminants were given in [60]. 
The authors postulated that high electric stresses develop at the 
tip of such contaminants and therefore high-energy electrons 
are injected into the bulk at the negative polarity and trapped 
electrons are extracted towards the tip at the positive polarity. 
Such a phenomenon, when repeated in each cycle of the AC 
voltage, causes a small degraded area at the tip of the 
contaminant and with the passage of time may render a cavity, 
which finally may lead to deterioration paths. It should be 
noted that the ideas about the injection and extraction of charge 
carriers in [60] could also be seen in earlier publications [30, 
33, 34]. Elaborating on the internal discharges, yet another 
researcher dealt with their two main mechanisms, namely 
Townsend-like discharges and streamer-like discharges [61]. 
His research indicated that active species from one half cycle 
can generate initiating electrons on the next, in accordance with 
[30, 33-34]. 

The deteriorating effects of discharges depend on the 
insulating materials, as shown in [62], where it was observed 
that discharge activity may lead to depolymerization of silicone 
rubbers. Indeed, discharge activity in RTV silicone rubber 
affected the propensity of macromolecular chains to crystallize 
at very low temperatures more than in HTV silicone rubber. A 
relatively recent paper [63] indicated the relationship between 
applied voltage and discharge intensity and also the 
relationship between discharge intensity and discharge energy. 
The discharge energy increases almost linearly with discharge 
intensity, with the authors noting that even a small discharge 
whose magnitude is 15 pC, at an inception voltage of 0.8 kV, 
will release energy of 8.48 x 10-9 joules. There was, however, 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4663  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

no relationship given between discharge energy and insulation 
damage. 

The relationship between cavity diameter and discharge 
magnitude for enclosed cavities was discussed in some recent 
publications [64, 65]. Both of them indicated an increase of 
partial discharge magnitude with cavity diameter and also a 
decrease of discharge inception voltage with cavity diameter. 
Such results, albeit with different insulating materials, were in 
agreement with previous work [24]. The series resistance in the 
HV circuit has as task to limit the current in the transformer 
secondary between the HV transformer and the test cell, 
especially when we are concerned with laboratory experiments. 
As was noted in [66], the resistance is usually calculated by 
considering the maximum voltage available and the maximum 
current allowed. This on one hand has as a result to limit the 
damage in case of breakdown, but on the other hand the choice 
of the series resistance also affects the partial discharge 
activity. The series resistance is an important parameter in 
determining the current flowing through the test cell as well as 
the waveforms induced on the test cell’s terminals. Since the 
discharge magnitude or current induced at the terminals is 
dependent on the current flowing through the HV side, it 
follows that a smaller current on the HV side will result in a 
decrease of the individual discharge current pulses. 

IV. PARTIAL DISCHARGES AND MULTIPLE CAVITIES 
From the above –non exhaustive- review, it is evident that 

there is a number of factors affecting the discharge behavior 
inside enclosed cavities and, consequently, the insulation 
damage. It is also evident that –especially for laboratory 
experiments– numerous researchers are confined in the study 
of only one well defined enclosed cavity. This is logical since 
the primary aim is to understand the discharge activity in one 
cavity before one proceeds to the study of more complex 
insulating systems. In real life, however, multiple cavities may 
be present. Earlier research in this direction revealed that in the 
case of adjacent enclosed cavities in an insulating sample, a 
triggering mechanism may be at work, i.e. an eventual 
breakdown in one cavity may trigger breakdown in a second 
one, and so on [67, 68]. As was noted earlier [69-71], although 
the waveform of the recorded discharge pulses may not differ 
greatly between an isolated cavity and two (or three) adjacent 
cavities, the apparent charge recorded is different. This 
precisely may suggest that a triggering mechanism may be at 
work. 

V. INFLUENCE OF VERY SMALL PARTIAL DISCHARGES  
Very small discharges may damage an insulation, as 

suggested earlier in [29] and also in [50, 51], where it was 
pointed out that such discharges may have a cumulative effect, 
which with time will produce more material degradation 
(polymer disruption, carbonization, release of gaseous 
byproducts etc.). Such views were contested however in [6]. 
Pioneering work done in [72] revealed that chemical changes 
observed in an antioxidant free high molecular weight 
polyethylene (AOFHMW-PE) in cavities above inception 
voltage were very similar to those obtained with values of 
applied voltage below inception. In fact, it was found that the 
presence of byproducts on the surface of similar nitrogen 

species only when corona and sub-corona took place indicates 
that corona aging and sub-corona aging result in similar aging 
products. Since corona currents and resulting species lead to 
failure of polymer insulations in relatively short times, their 
findings bore strong evidence that sub-corona aging could lead 
to serious damage as well, even though a conventional partial 
discharge detector could not detect any activity. Subsequent 
work [13-14, 73-74] tends to confirm [72]. Very small partial 
discharges and/or charging phenomena below inception voltage 
were observed not only in conventional polymers but also with 
polymer nanocomposites, albeit with lesser frequency [75, 76]. 
Earlier research pointed out the importance of small partial 
discharges and their influence on possible damaging effects on 
solid insulation [77-82]. Certainly, the question of the effect of 
very small partial discharges on insulation degradation is a 
complex one since it has to do with not only the means of 
discharge detection and their respective sensitivities but also 
with the exact terminologies of the various regimes of 
discharges. Moreover, there is always the question of the 
thickness of the investigated insulation, the size of minute 
enclosed cavities as well as the limitations of the discharge 
detector available, as was noted before [55-56, 83-84]. 

It is without any doubt that the points raised in [72] merits 
more attention and research. It must, however, pointed out that, 
although there are indications that minute discharges (possibly 
undetected from conventional discharge equipment) and 
charging phenomena below inception voltage may exist, more 
work has to be performed with a variety of insulating materials 
before any definite conclusions can be reached. 

VI. FURTHER COMMENTS ON INSULATION DAMAGE 
As noted in [23], the observable effects of partial discharges 

can be an increase in the loss component of the charging 
current taken by the capacitive sample, an increased dielectric 
power loss, radiation of energy in the form of sound, light or 
ultraviolet radiation, changes in pressure and chemical changes 
in the dielectric material. Of these effects, the most important  
-regarding the relation of discharge energy to the material 
damage– is the last one, i.e. the chemical changes. However, 
such changes can manifest themselves in different ways, 
especially if the insulation is a composite one. If, for example, 
we have a composite insulation of solid/liquid, chemical 
changes can be gaseous byproducts, changes on the surface of 
the solid insulation, changes in the liquid itself. Which one of 
these changes can be related to the discharge energy, or for that 
matter to any discharge parameter? Any of them or all of them? 
Such questions are in need of an answer. Although there have 
been numerous publications on condition monitoring of 
insulating systems [85-88], there are relatively fewer papers on 
the actual relationship between discharge energy and material 
damage. This does not mean that such a relationship does not 
present some difficulties, especially with composite solid/liquid 
insulation, as was noted in older publications [89, 90]. In [89], 
for example, the authors failed to obtain a quantitative 
relationship between the partial discharge parameters and the 
gaseous byproducts at atmospheric pressure, explaining that in 
the course of a test, absorption and evolution of gases were not 
related to each other with a permanent relationship. It is, 
however, a better and more reasonable approach, than to try to 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4664  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

correlate insulation damage with individual discharges, no 
matter how large their magnitudes are. Recent efforts, however, 
indicated that for different equipment with defects of the same 
type, the percentage of gases has similar values. The 
development of defects in high voltage equipment is 
accompanied not only by the growth of gas concentrations but 
also stabilization of the percentage of contents of gases at 
strictly defined levels. This gives some hope regarding the use 
of the gas content of oils as an additional criterion for 
interpreting the results of analysis of gases dissolved in oil and 
thus to try to correlate them with discharge energy [91]. In this 
respect, important work performed relatively recently should be 
taken into account [92-94]. 

It is true that in the context of the present paper, particular 
attention was given to the total partial discharge energy. This is 
not only in agreement with earlier publications [29, 47] but also 
in accordance with much older ones [95], where it was 
suggested that the energy dissipation of the individual 
discharge in an enclosed cavity appears to be more important 
than the discharge magnitude and the significance of 
cumulative energy dissipation was stressed. 

VII.  CONCLUSIONS  
Partial discharges in enclosed cavities consist a major 

danger for insulating materials. Past research indicated that it is 
better to try to correlate discharge energy with insulation 
damage rather than individual discharge magnitudes with 
insulation damage. The role of sample thickness, the role of 
cavity dimensions, the role of series resistance in the HV 
circuit,  the role of multiple cavities in the same insulation as 
well as the possible damaging role of small discharges and/or 
charging phenomena below inception voltage are reviewed. 
However, it should be noted that the notion of insulation 
damage is still an elusive one, especially if we consider 
composite insulating systems.  

RREFERENCES 
[1] N. B. Timpe, “Partial discharge measurements in distributed parameter 

systems: cables”, Engineering Dielectrics, Vol. I: Corona Measurements 
and Interpretation, ASTM Special Technical Publication, pp. 134-175, 
1979 

[2] R. Bartnikas, “Corona discharge processes in voids”, Engineering 
Dielectrics, Vol. I: Corona Measurements andInterpretation, ASTM 
Special Technical Publication, pp. 22-67, 1979 

[3] A. Kelen, “Studies on partial discharges on solid dielectrics: A 
contribution to the discharge resistance testing of insulating materials”, 
Acta Polytechnica Scandinavica, No. EI 16, pp. 138, 1967 

[4] A. Gemant, “Die Verlustkurve Lufthaeltiger Isolierstoffe”, Zeitscchrift 
der Technischen Physik, Vol. 13, pp. 184-189, 1932 

[5] A. Gemant, W. von Philippoff, “Die Funkenstrecke mit Vorkodensator”, 
Zeitscchrift der Technischen Physik, Vol. 13, pp. 425-430, 1932 

[6] F. H. Kreuger, Discharge detection in high voltage equipment,  
Heywood, London, United Kingdom, 1964 

[7] G. C. Crichton, P. W. Karlsson, A. Pedersen, “Partial discharges in 
ellipsoidal and spheroidal voids”, IEEE Transactions on Electrical 
Insulation, Vol. 24, No. 2, pp. 335-342, 1989 

[8] M. G. Danikas, G. E. Vardakis, “The case of Pedersen’s theory to model 
partial discharges in cavities enclosed in solid insulation: A criticism of 
some of its aspects from an electrical engineers’ and from a physicists’ 
point of view”, Journal of Electrical Engineering, Vol. 52, No. 5-6, pp. 
166-170, 2001 

[9] H. Repp, K. W. Nissen, P. Roehl, “Partial discharges in voids: inception 
conditions and detection limits”, Siemens Forschung und Entwicklung 
Bericht, Vol. 12, pp. 101-106, 1983 

[10] M. G. Danikas, I. W. McAllister, G. C. Crichton, A. Pedersen, 
“Discussion: partial discharges in ellipsoidal and spheroidal voids”, 
IEEE Transactions on Electrical Insulation, Vol. 26, pp. 537-539, 1991 

[11] Z. Achillides, M. G. Danikas, E. Kyriakides, “Partial discharge modeling 
and induced charge concept: Comments and criticism of Pedersen’s 
model and associated measured transients”, IEEE Transactions on 
Dielectrics and Electrical Insulation, Vol. 24, No. 2, pp. 1118-1122, 
2017 

[12] M. G. Danikas, “Discharge studies in solid insulation voids”, CEIDP 
Conference on Electrical Insulation and Dielectric Phenomena, USA, 
October 28-31, 1990 

[13] M. G. Danikas, A. M. Bruning, “Comparison of several theoretical 
subcorona to corona transition relations with recent experimental 
results”, IEEE International Symposium on Electrical Insulation, USA, 
June 7-10, 1992  

[14] A. M. Bruning, M. G. Danikas, “Report on continuing work on parallel 
and non-parallel electric field chemical aging of polymer cavities”, 4th 
International Conference on Conduction and Breakdown in Solid 
Dielectrics, Italy, June 22-25, 1992  

[15] J. H. Mason, “The deterioration and breakdown of dielectrics resulting 
from internal discharges”, Proceedings of the IEE, Vol. 98, Pt. I, pp. 44-
59, 1951 

[16] J. H. Mason, “Breakdown of insulation by discharges”, Proceedings of 
the IEE, Vol. 100, Pt. IIA, pp. 149-158, 1953 

[17] W. Boenig, “Luftgehalt und Luftspaltverteilung geschichteter 
Dielektrika”, Archiv fuer Elektrotechnik, Vol. 48, pp. 85-96, 1963 

[18] A. Kelen, M. G. Danikas, “Evidence and presumption in PD 
diagnostics”, IEEE Transactions on Dielectrics and Electrical Insulation, 
Vol. 2, No. 5, pp. 780-795, 1995 

[19] J. H. Mason, “Breakdown of solid dielectrics in divergent fields”, 
Proceedings of the IEE, Vol. 102, Pt. C, pp. 254-263, 1955 

[20] J. H. Mason, “Discharges”, IEEE Transactions on Electrical Insulation, 
Vol. 13, No. 4, pp. 211-238, 1978 

[21] E. Friedlander, J. R. Reed, “Electrical discharges in air gaps facing solid        
insulation in high-voltage equipment”, Proceedings of the IEE, Vol. 100, 
Pt. IIA, pp. 121-131, 1953 

[22] S. I. Reynolds, “On the behavior of natural and artificial voids in 
insulation under internal discharge”, AIEE Transactions on Power 
Apparatus and Systems, Vol. PAS-77, Pt. III, pp. 1604-1608, 1959 

[23] M. S. Naidu, V. Kamaraju, High voltage engineering, Tata McGraw-
Hill, New Delhi, India, 2000 

[24] E. C. Rogers, “The self-extinction of gaseous discharges in cavities in 
dielectrics”, Proceedings of the IEE, Vol. 105A, pp. 621-630, 1958  

[25] A. A. Hossam-Eldin, S. S. Dessouky, S. M. El_Mekkaway, R. A. Abd 
El-Aal, “Internal discharges in cavities in solid dielectric materials”, 
Journal of Electrical Engineering, Vol. 9, No. 4, Art. No. 13, 2009 

[26] R. J. Meats, A. W. Stannett, “Degradation of insulation materials by 
electrical discharges”, IEEE Transactions on Power, Apparatus and 
Systems, Vol. PAS- 83, pp. 49-54, 1964 

[27] R. Bartnikas, G. L. d’ Ombrain, “A study of corona discharge rate and 
energy loss in spark gaps”, IEEE Transactions on Power, Apparatus and           
Systems, Vol. PAS-84, No. 9, pp. 770-778, 1965  

[28] I. Y. Megahed, B. C. Mansfield, R. E. Wootton, “Detection and 
measurement of discharges in cavities in solid dielectrics”, Proceedings 
of the IEE, Vol. 114, No. 11, pp. 1822-1824, 1967 

[29] I. Y. Megahed, “The discharge-repetition rate in cavities in epoxy resin, 
polyethylene, and mica under alternating voltage conditions”, IEEE 
Transactions on Electrical Insulation, Vol. 10, No. 2, pp. 69-74, 1975 

[30] T. Tanaka, Y. Ikeda, “Internal discharges in polyethylene with an 
artificial cavity”, IEEE Transactions on Power, Apparatus and Systems, 
Vol. PAS-90, No. 4-6, pp. 2692-2702, 1971 

[31] R. J. Densley, B. Salvage, “Partial discharges in gaseous cavities in solid         
dielectrics under impulse voltage conditions”, IEEE Transactions on 
Electrical Insulation, Vol. 6, No. 2, pp. 54-62, 1971 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4665  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

[32] H. Okamoto, M. Kanazashi, T. Tanaka, “Deterioration of insulating 
materials by internal discharge”, IEEE Transactions on Power, 
Apparatus and Systems, Vol. PAS-96, No. 1, pp. 166-177, 1977 

[33] T. Tanaka, A. Greenwood, Advanced power cable technology: basic 
Concepts and testing, CRC Press, USA, 1983 

[34] T. Tanaka, A. Greenwood, Advanced power cable technology: present 
and future, CRC Press, USA, 1983 

[35] S. J. Rigby, B. M. Weedy, “Partial discharges in liquid-nitrogen 
impregnated taped-cable insulation”, Proceedings of the IEE, Vol. 123, 
No. 2, pp. 165-170, 1976 

[36] K. Horii, A. J. Pearmain, A. J. McNerney, “Correlation of electrical 
breakdown of supercritical helium in short gaps with partial discharge in 
cable samples”, Crygenics, Vol. 23, No. 2, pp. 102-106, 1983 

[37] P. N. Nikolopoulos, D. N. Diamantopoulos, “Behavior of pressboard-oil 
insulation under alternating voltages of industrial frequency in a non- 
uniform electric field”, Proceedings of the IEE, Vol. 130, Pt. A, No. 3,          
pp. 145-151, 1983 

[38] S. Selvakumar, R. S. Nema, “Low pressure discharges in narrow          
cylindrical voids”, 3rd IEE International Conference on Dielectrics 
Materials, Measurements and Applications, United Kingdom, September 
10-13, 1979 

[39] A. C. Gjaerde, Multi-factor ageing of epoxy: the combined effect of          
temperature and partial discharges, Ph.D. Thesis, Norwegian University 
of Science and Technology, Faculty of Electrical Engineering and 
Computer Science, Department of Power Engineering, 1994 

[40] B. Stenerhag, A. Danemar, “Partial discharge characteristics of some 
liquid impregnated system”, 3rd IEE International Conference on 
Dielectrics Materials, Measurements and Applications, United Kingdom, 
September 10-13, 1979 

[41] J. P. Reynders, “Electrical detection of degradation caused by partial 
discharges in polythene”, IEE International Conference on Dielectric 
Materials, Measurements and Applications, United Kingdom, July 21-
25, 1975  

[42] J. P. Reynders, “Measurement of the effects of partial discharge activity 
on low density polyethylene”, 3rd IEE International Conference on 
Dielectrics Materials, Measurements and Applications, United Kingdom, 
September 10-13, 1979 

[43] R. R. Opoku, E. G. Robles, J. H. Mason, “Degradation and breakdown 
of polypropylene film by internal discharges”, IEE International 
Conference on Dielectric Materials, Measurements and Applications, 
United Kingdom, July 21-25, 1975  

[44] J. H. Mason, “Enhancing the significance of PD measurements”, IEEE  
Transactions on Dielectrics and Electrical Insulation, Vol. 2, No. 5, pp. 
876-888, 1995 

[45] K. D. Wolter, J. Tanaka, J. F. Johnson, “A study of gaseous degradation 
products of corona-exposed polyethylene”, IEEE Transactions on 
Electrical Insulation, Vol. 17, No. 3, pp. 248-252, 1982 

[46] J. Golinski, J. H. Calderwood, S. Zoledziowski, A. Sierota, “Partial 
discharges in a cylindrical void with a metal rod electrode”, IEEE 
Transactions on Electrical Insulation, Vol. 17, No. 6, pp. 560-569, 1982 

[47] G. Garcia, B. Fallou, “Equipment for the energy measurement of partial 
discharges”, 1st International Conference on Conduction and Breakdown 
in Solid Dielectrcs, France, July 4-8, 1983 

[48] F. Viale, J. Poitevin, B. Fallou, J. F. Morel, R. Buccianti, S. Yakov, S. 
Cesari, E. Serena, “Study of a correlation between energy of partial 
discharges and degradation of paper-oil insulation”, CIGRE, Report 15-
12, pp. 1-9, 1982 

[49] A. R. Nosseir, M. M. A. Salama, A. Soliman, M. Rizk, “Discharge 
detection and measurement in voids in solid dielectrics”, 1984 IEEE 
International Symposium on Electrical Insulation, Canada, June 11-13, 
1984 

[50] M. G. Danikas, A study of the behavior of a uniaxially orientated 
polyethylene tape/oil insulating system subjected to electrical and 
thermal stresses, Ph.D. Thesis, University of London, Queen Mary 
College, Department of Electrical and Electronic Engineering, 1985 

[51] A. J. Pearmain, M. G. Danikas, “A study of the behavior of a uniaxially 
oriented polyethylene/oil insulating system subjected to electrical and 

thermal stresses”, IEEE Transactions on Electrical Insulation, Vol. 22, 
No. 4, pp. 373-382, 1987 

[52] R. Liao, J. Yan, L. Yang, M. Zhu, B. Liu, “Study of the relationship 
between damage of oil-impregnated insulation paper and evolution of 
phase-resolved partial discharge patterns”, European Transactions on 
Electrical Power, Vol. 21, pp. 2112-2124, 2011 

[53] T. Tanmaneeprasert, P. L. Lewin, “The use of partial discharge analysis 
for the diagnosis of electrical ageing in polymeric insulation containing 
cavities”, Electrical Insulation Conference, Canada, June 19-22, 2016 

[54] K. Theodosiou, D. P. Agoris, I. Gialas, I. Vitellas, “Measurements of 
partial discharges and power losses in PET films in high voltage AC 
fields”, Journal of Electrical Engineering, Vol. 58, No. 5, pp. 250-256, 
2007 

[55] M. G. Danikas, “Some possible new applications of a partial discharge 
(PD) model and its relation to PD detection sensitivity”, European 
Transactions on Electrical Power, Vol. 6, No. 6, pp. 445-448, 1996 

[56] M. G. Danikas, G. E. Vassiliadis, “Models of partial discharges (PD) in 
enclosed cavities in solid dielectrics: A study of the relationship of PD 
magnitudes to the sensitivity of PD detectors and some further 
comments on insulation lifetime”, Journal of Electrical Engineering, 
Vol. 54, No. 5-6, pp. 132-135, 2003 

[57] A. A. Hossam-Eldin, “Analysis of discharge damage and breakdown in 
gaseous Cavities in insulating materials”, Journal of Physics D: Applied 
Physics, Vol. 17, pp. 421-428, 1984 

[58] E. Jahoda, J. Kudelcik, “Internal partial discharge in cavity of 
polyurethane”, Procedia Engineering, Vol. 192, pp. 365-369, 2017 

[59] T. Tanmaneeprasert, P. L. Lewin, “Electrical treeing and ageing 
characteristics in cavities of low density polyethylene dielectrics on 
partial discharge measurements”, IEEE Conference on Electrical 
Insulation and Dielectric Phenomena, Canada, October 16-19, 2016 

[60] T. Tanaka, T. Okamoto, K. Nakanishi, T. Miyamoto, “Aging and related         
Phenomena in modern electric power systems”, IEEE Transactions on 
Electrical Insulation, Vol. 28, No. 5, pp. 826-844, 1993 

[61] J. C. Devins, “The physics of partial discharges in solid dielectrics”, 
IEEE Transactions on Electrical Insulation, Vol. 19, No. 5, pp. 475-495, 
1984 

[62] J. P. Habas, J. M. Arrouy, F. Perrot, “Effects of electric partial 
discharges on the rheological and chemical properties of polymers used 
in HV composite insulators after railway service”, IEEE Transactions on 
Dielectrics and Electrical Insulation, Vol. 16, No. 5, pp. 1444-1454, 
2009 

[63] M. Kannan, P. Sreejaya, “Partial discharge detection in solid 
dielectrics”, International Journal of Scientific and Engineering 
Research, Vol. 4, No. 8, 2013  

[64] Z. Lei, J. Song, M. Tian, X. Cui, C. Li, M. Wen, “Partial discharges in 
cavities in ethylene propylene rubber”, IEEE Transactions on Dielectrics 
and Electrical Insulation, Vol. 21, No. 4, pp. 1647-1659, 2014 

[65] I. Kolcunova, M. Sipos, “Measurement of partial discharge in disc-
shaped voids”, 8th International Scientific Symposium on Electrical 
Power Engineering (Elektroenergetika 2015),  Slovakia, September 16-
18, 2015 

[66] M. G. Danikas, A. Kelen, “Diagnostic PD pulse techniques: a discussion 
on possibilities and limitations”, Nordic Insulation Symposium, Norway, 
June 10-12, 1996 

[67] D. P. Agoris, N. D. Hatziargyriou, “Two voids model for simulation of 
partial discharges in solid dielectrics”, Modelling, Simulation & Control, 
Vol. 42, No. 2, pp. 1-14, 1992 

[68] M. G. Danikas, “Partial discharge behavior of two (or more) adjacent 
cavities in polyethylene samples”, Journal of Electrical Engineering, 
Vol. 52, No. 01-02, pp. 36-39, 2001 

[69] D. P. Agoris, N. D. Hatziargyriou, “Approach to partial discharges 
development in closely coupled cavities embedded in solid dielectrics by 
the lumped capacitance model”, IEE Proceedings-A, Vol. 140, No. 2, 
pp. 131-134, 1993 

[70] D. P. Agoris, N. D. Hatziargyriou, “A modelistic consideration of the 
interaction between discharges in cavities of solid dielectrics”, 8th 



Engineering, Technology & Applied Science Research Vol. 9, No. 5, 2019, 4659-4666 4666  
  

www.etasr.com Danikas and Morsalin:  Foreign Inclusions, Enclosed Cavities, Partial Discharge Models and … 
 

International Symposium on High Voltage Engineering, Japan, August 
23-27, 1993 

[71] D. P. Agoris, N. D. Hatziargyriou, “Application of the EMTP program 
in the study of discharges in two cavities enclosed in a solid dielectric 
during its stressing with AC or impulse voltage”, CIGRE Greek National 
Committee, Greece, February 10-11, 1994 (in Greek) 

[72] A. M. Bruning, D. G. Kasture, F. J. Campbell, N. H. Turner, “Effect of 
cavity sub-corona current on polymer insulation life”, IEEE 
Transactions on Electrical Insulation, Vol. 26, No. 4, pp. 826-836, 1991 

[73] M. G. Danikas, “On the damage of insulating materials below inception 
voltage”, Journal of Electrical Engineering, Vol. 52, No. 11-12, pp. 367-
371, 2001 

[74] A. Harlin, M. G. Danikas, P. Hyvonen, “Polyolefin insulation 
degradation in electrical field below critical inception voltages”, Journal 
of Electrical Engineering, Vol. 56, No. 5-6, pp. 135-140, 2005 

[75] Y. Zhang, M. G. Danikas, X. Zhao, Y. Cheng, “Preliminary 
experimental work on nanocomposite polymers: Small partial discharges 
at inception voltage, the existence of possible charging mechanisms 
below inception voltage and the problem of definitions“, Journal of 
Electrical Engineering, Vol. 63, No. 2, pp. 109-114, 2012 

[76] Y. Zhang, M. G. Danikas, X. Zhao, Y. Cheng, “Charging phenomena 
below the inception voltage: Effects of nanofillers on epoxy”, Malaysian 
Polymer Journal, Vol. 7, No. 2, pp. 68-73, 2012 

[77] Y. Shibuya, S. Zoledziowski, J. H. Calderwood, “Void formation and 
electrical breakdown in epoxy resin”, IEEE Transactions on Power, 
Apparatus and Systems, Vol. 96, pp. 196-206, 1977 

[78] T. Tanaka, “Internal partial discharge and material degradation”, IEEE         
Transactions on Electrical Insulation, Vol. EI-21, No. 6, pp. 899-905, 
1986 

[79] R. Bartnikas, J. P. Novak, “On the character of different forms of partial         
discharge and their related terminologies”, IEEE Transactions on 
Electrical Insulation, Vol. 28, No. 6, pp. 956-968, 1993 

[80] C. Hudon, R. Bartnikas, M. R. Wertheimer, “Chemical and physical 
degradation effects on epoxy surfaces exposed to partial discharges”, 4th 
International Conference on Properties and Applications of Dielectric 
Materials, Australia, July 3-8, 1994 

[81] P. H. F. Morshuis, “Assessment of dielectric degradation by ultrawide-
band PD Detection”, IEEE Transactions on Dielectrics and Electrical         
Insualtion, Vol. 2, No. 5, pp. 744-760, 1995 

[82] M. G. Danikas, “Small partial discharges and their role in insulation 
deterioration”, IEEE Transactions on Dielectrics and Electrical         
Insulation, Vol. 4, No. 6, pp. 863-867, 1997 

[83] J. Sletbak, “Partial discharge measurements: possibilities and 
limitations”, 4th International Conference on Conduction and 
Breakdown in Solid Dielectrics, Italy, June 22-25, 1992 

[84] M. G. Danikas, "Some new relationships and a scaling law regarding 
partial discharges in spherical cavities enclosed in solid insulation", Acta 
Electrotehnica Napocensis, Vol. 39, No. 1, pp. 5-9, 1998 

[85] R. E. James, B. T. Phung, “Development of computer-based 
measurements and their applications to PD pattern analysis”, IEEE 
Transactions on Dielectrics and Electrical Insulation, Vol. 2, No. 5, pp. 
838-856, 1995 

[86] M. G. Danikas, N. Gao, M. Aro, “Partial discharge recognition using 
neural networks: A review”, Archiv fuer Elektrotechnik (Electrical 
Engineering), Vol. 85, pp. 87-93, 2003 

[87] G. C. Stone, “A perspective on online partial discharge monitoring for         
assessment of the condition of rotating machine stator winding 
insulation”, IEEE Electrical Insulation Magazine, Vol. 28, No. 5, pp. 8-
13, 2012 

[88] M. Wu, H. Cao, J. Cao, H. L. Nguyen, J. B. Gomes, S. P. 
Krishnaswamy, “An overview of state-of-the-art partial discharge 
analysis techniques for condition monitoring”, IEEE Electrical 
Insulation Magazine, Vol. 31, No. 6, pp. 22-35, 2015 

[89] D. Moraru, C. Popesco, M. Stoica, F. Tanasesco, “Detection of 
degradation produced by partial discharges in paper-oil insulation”, 
CIGRE Report, Paper 15-05, pp. 1-8, 1970 

[90] H. Gsodam, “Investigations of the energy of internal partial discharges 
by AC stress”, 4th International Symposium on High Voltage 
Engineering, Athens, Greece, September 5-9, 1983 

[91] O. Shutenko, “Faults diagnostics of high-voltage equipment based on the           
analysis of the dynamics of changing of the content of gases”, 
Energetika, Vol. 64, No. 1, pp. 11-22, 2018 

[92] M. Duval, “A review of faults detectable by gas-in-oil analysis in 
transfmormers”, IEEE Electrical Insulation Magazine, Vol. 18, No. 3, 
pp. 8-17, 2002 

[93] M. Duval, “New techniques for dissolved gas-in-oil analysis”, IEEE 
Electrical Insulation Magazine, Vol. 19, No. 2, pp. 6-15, 2003 

[94] K. Sahitya Yadav, Understanding the impact of thermal ageing on 
dielectric properties of transformer insulation, Ph.D. Thesis, Indian 
Institute of Technology, Department of Electrical Engineering, Chennai, 
India, 2017 

[95] G. Mole, “Improved methods of test for the insulation of electrical    
equipment”, Proceedings of the IEE, Vol. 100 (IIA), pp. 276-283, 1953