A Review on Nanocomposite Based Electrical Insulations

Abstract

The potential of nanocomposites have been drawing the intention of the researchers from energy storage to electrical insulation applications. Nanocomposites are known to improve dielectric properties, such as the increase in dielectric breakdown strength, suppressing the partial discharge (PD) as well as space charge, and prolonging the treeing, etc. In this review, different theories have been established to explain the reactions at the interaction zone of polymer matrix and nanofiller; the characterization methods of nanocomposites are also presented. Furthermore, the remarkable findings in the fields of epoxy, cross-linked polyethylene (XLPE), polypropylene and polyvinyl chloride (PVC) nanocomposites are reviewed. In this study, it was observed that there is lack of comparison between results of lab scale specimens and actual field aged cables. Also, non-standardization of the preparation methods and processing parameters lead to changes in the polymer structure and its surface degradation. However, on the positive side, recent attempt of 250 kV XLPE nanocomposite HVDC cables in service may deliver a promising performance in the coming years. Moreover, materials such as self-healing polymer nanocomposites may emerge as substitutes to traditional insulations.

1. INTRODUCTION

Several studies carried out in last two decades has shown that nanotechnology is not limited to areas such as drug delivery, automobile, semiconductor etc. With the benefit of having a large surface area, nano-particles have proved its efficacy over microparticles.

The introduction to nanodielectrics by T. J. Lewis in 1994 has given a different perspective for the application of nanotechnology for electrical insulation development [1]. Nanoparticles have at least one of the dimensions in the nanometer range (1 nm= 10-9 m) [2]. Microcomposites improves one of the characteristics of the electrical insulation at the cost of another, whereas nanocomposites improve the overall characteristics (i.e. mechanical, electrical, thermal, etc.) due to enhanced interactions at interfaces [3]. Hence, these materials have emerged as suitable substitutes to microcomposites. Based on this, different articles have used the concept of nanocomposites for indoor and outdoor electrical insulation applications [4-6]. The major developments in the field of electrical insulation over the years are shown in Fig. 1. However, the studies revolve around polyethylene (PE), silicon rubber, epoxy resin and polyvinyl chlorine (PVC) based nanocomposites. From this context, our study critically reviews the findings from different researchers and also emphasizes on XLPE insulation which has attracted attention because of its encouraging characteristics.

Fig. 1.Major developments in the field of electrical insulation

This paper also discusses the different theories to explain the reactions at interface contact areas, preparation methods, and review of different nanocomposite insulations, opportunities and challenges ahead in the field of nanocomposite insulation. Also, noted characteristics such as surface degradation in epoxy resins and water trees in XLPE nanocomposites are highlighted in brief.

 

2. REACTIONS AT THE INTERACTION ZONE

The improved properties of nanocomposites are due to the increased reactions at interaction zone which surrounds the nanoparticle, as shown in Fig. 2 [7]. The surface area of nanoparticles is three times greater than microparticles, which in turn enhances the reactions at interaction zone. To explain the reactions at the interface, different models have been proposed. The Lewis [1] model is based on two layers, whereas the model proposed by Tanaka is based on a multi-layer approach [8] in which the innermost layer is of the least thickness, increasing subsequently from 1st to the 4th layer. It consists of bonded layer, bound layer, loose layer and double layer over the three layers mentioned. Tanaka has pointed out that defects and impurities are present in the outer double layer. By considering these impurities as traps, it has hypothesized that traps are accountable for change in different dielectric properties such as space charge, i.e. larger surface area of nanoparticles leads to change in density of traps which reduces the space charge. To explain this, the addition of nanoparticles leads to an increase in charge carrier mobility and decrease in average hopping distance relative to polymer matrix.

Fig. 2.Different components of polymer nanocomposites [7].

Raetzke and Kindersberger [9] have developed the interface volume model to explain the role of interface content in enhancement of dielectric properties. It shows that for higher interface thickness (in nm), maximum value of interface volume content (in vol %) is observed. But, in case of lower interface thickness, values such maximum value of interface volume content is not observed. This is shown in Fig. 3(a). To support this, the highest value of dielectric properties is observed for highest interface content. For e.g. maximum arcing time observed in Fig. 3(b) for F1 silica nanofiller is at 5 wt% content of nanosilica. Assuming nanosilica diameter of 30 nm and interface thickness of 20 nm, maximum value of interface volume is observed around 5 wt% as shown in Fig. 3(a). Also, for F2 silica nanofiller, due to lower value of interface thickness (below 10 nm) no maximum value of interface volume is observed. However, the model assumes particles in spherical shape with same diameter and homogeneous dispersion of nanoparticles. Hence, this model works under ideal conditions, which is highly unattainable in real experiments. So, the simulation can be further carried out under real parameters.

Fig. 3.(a) Change in interface volume content with nanofiller content, (b) change in arcing time with nanofiller content [9].

Tsagaropoulos and Eisenberg model [10] assumes the two layers, but is different from the Lewis model. It states that the inner layer is highly mobilized, with the outer layer being immobile for charges. The model basically deals with two glass transition temperatures (Tg) for polymer nanocomposites. This is based on strong interactions at interaction zones for lower content and vice versa. Gist of Tsagaropoulos model is that it explains the behavior of nanocomposites with the help of ‘Tg’ whereas Tanaka’s model is the correlation of its multilayer model with different phenomenon, such as treeing, PD [11,12] etc. Also, the assumptions in Tanaka’s model contradicts the Tsagaropoulos model in terms of mobility of the outermost layer.

The above proposed models explaining the different reactions occurring at the interfaces have failed to explain the exact role of nanoparticles in enhancing the properties of nanocomposites. The common assumption in all proposed models is that structural properties in interface layer is changed, which results in structural and chain dynamic changes around the nanoparticles. This may stand as justification for change in surface structure of nanocomposite.

Attainment of expected parameters such as low permittivity, dielectric loss, space charge etc. are ascribed to reduced Maxwell-Wagner polarization at the interfaces of nanocomposites, which are dependent on the dispersion of nanofiller into polymer matrix. This suggests that due to nanometric dimension, the polarization is reduced [13]. However, the explanation about ‘Gouy-Chapman-Stern layer’ by Lewis seems satisfactory. This layer around the nanoparticle has high conductivity as compared to surrounding polymer matrix, which accelerates the charge movements at the interfaces [14]. To support the hypothesis of increased interactions at the interface, numerical analysis presented in [15], real and imaginary values interface permittivity are less than the polymer matrix which may be due to reduced polymer chain mobility, as shown by Todd and Shi model. However, in reference [15], it is assumed that all nanoparticles are of same size and diameter, which is an ideal condition. Generally, the dimensions of nanoparticle are average values, making the same technique inefficient in real cases. Hence, standard deviations in nanoparticle dimensions will play a major role in numerical analysis.

The Todd and Shi model is based on the single layer concept. This model calculates the permittivity of the composite using permittivity’s of the filler (not necessarily nanoparticle) component, the matrix component and the interface region as well as the volume fractions of each. This model can be applied to any uniform or randomly dispersed composites. This model explains the concept of overlapping of interaction zones using the concept of interface overlap probability function. However, the limitation of this model is that it considers all particles to be of spherical shape. Hence, it may not be reliable for single and two dimensional nanoparticles [16].

The possible contributions from different layers of nanoparticles to different dielectric properties of nanocomposites, is well tabulated by Tanaka. Regardless, other plausible explanations may further support the multi-core model. In this case, the interface thickness is the major aspect as described in [9]. Hence, combination of numerical analysis with different models developed may help in better understanding of role of interfaces. For e.g., numerical analysis presented in [15] with 3-D simulation and multi-core model developed by Tanaka in [8], together may deliver promising results with respect to possible contributions of different layers of nanoparticle. Also, it can be concluded that a single theory or model is not sufficient to explain the interactions at the interfaces.

 

3. PREPARATION

Generally, the dispersion of nanoparticles is tailored using well documented methods such as in-situ polymerization, melt intercalation, ultrasonication, sol-gel, etc. [17,18].

3.1 Role of dispersion in nanocomposite characteristics

Dispersion plays a vital role in nanocomposite characteristics. The homogeneous dispersion of nanofiller is desirable since it avoids the inconsistencies in the results. In order to achieve the homogeneous dispersion, surface functionalization of nanoparticles by coupling agents such as silane, is preferred. However, this cannot produce an ideal nanocomposite. The impact of the preparation method on dielectric properties of epoxy nanocomposites is shown in Fig. 4(a), 4(b), 4(c), and 4(d) [19]. Nanocomposite (NEP) is prepared using ultrasonication whereas ultrasonication is omitted in microcomposite (MEP) preparation. In case of NDNEP (not well dispersed nanocomposite), only one hour electrical mixing is used. Similar results are observed using the sol-gel method for silica nanocomposites [20], whereas alumina and MgO nanocomposites are prepared using in-situ polymerization.

Fig. 4.(a-d) Impact of method of preparation on different dielectric properties of epoxy nanocomposites [19].

It is reported that the addition of nanofiller may lead to an alteration in the structure of polymer matrix which may lead to cavity formation due to hygroscopic nature of nanofiller. Improper selection of processing parameters may result in surface degradation of the polymer matrix [21]. The use of ultrasonication to prepare homogeneous nanocomposites leads to reduction of the aspect ratio (L/D) [22].

From literature survey and Table 1 it has been observed that,

Table 1.Effect of nanofiller content on characteristics of different nanocomposites.

i. There is no fixed temperature range for preparation of polymer nanocomposites. ii. In case of XLPE nanocomposites, the temperature of degassing is 60~80℃. Also, the temperature of curing should not exceed 160~170℃. iii. Curing time for XLPE nanocomposites should not be too high, which may lead to pre-crosslinking, following which it would be tough to extract it from the twin screw extruder or kneader machine. iv. Drying and surface modification of the nanoparticle is desirable to obtain the homogeneous dispersion.

3.2 Threshold content of nanofiller

Specifically in nanostructures such as carbon nanotubes (CNT) and graphite platelets, the threshold content (sometimes referred as percolation threshold) has been an intensively researched parameter in recent years. In case of CNT nanocomposites, it is the value after which the electrical conductivity of the nanocomposite increases drastically. In general, it is the minimum amount of external filler to be added in polymer matrix (in terms of percentage volume fraction) after which there will be a sudden transition in measured values. For fixed number of CNT tubes, an increase in its length or decrease in anisotropy leads to decrease in percolation threshold value [22]. The nanofiller having highest aspect ratio will lower the percolation threshold [23]. The statistical studies used for quantification of percolation are as follows [24],

i. Monte-Carlo method ii. Quadrat based method iii. k-function method iv. Nearest neighbour indices (NNI) method

The Monte-Carlo method uses a large number of squares positioned randomly within the micrograph and collects the number of particles within each square. From the comparison between a histogram of collected numbers and ideal Poison distribution, it analyses the dispersion of the nanoparticles. But, alone it does not yield precision in quantification. In quadrat based method, the skewness of particles is measured (which ideally should be zero). Higher the skewness, poorer is the dispersion. In the NNI method, index value greater than 1 indicates good dispersion and less than 1 indicates agglomeration. The k-function method is the extension of NNI method in which multiscale information is available. The k-function value less than ϖh2 (‘h’ is the radius of particle) indicates good dispersion, and vice versa for agglomeration. To obtain good precision results the Monte-Carlo method along with any one of the remaining three methods is generally used [25].

The threshold content for different nanocomposites is tabulated in Table 1. Also behavior of different electrical, mechanical and thermal characteristics has been discussed with plausible explanations.

(Note: Hereafter, Abbreviations used in Table 1, Table 2, Table 3 and Table 4 are as follows,

M- Micro filler, N- Nano filler, M + N- Micro and Nanofiller, ST- Surface treated, NST- Non surface treated, phr- per hundred parts of resin, I- Increase, D- Decrease, # - One variable at the end signifies continuous increase or decrease respectively.)

3.3 Characterization

In earlier days, wide angle X-ray diffraction (in the range of 10~90°) (WAXD) was the most preferred method for characterization of multilayered nanostructures such as clays, graphite etc., due to its simplicity and availability. But, crystallinity and agglomeration of nanocomposites cannot be explained by WAXD. But, WAXD analysis can be applied only to multilayered nanostructures such as clays, graphite, etc. In addition, well resolved WAXD peaks (Braggs peaks) cannot be obtained [45]. Small angle X-ray scattering (in the range of 0~5°) (SAXS) is the substitute (in fact almost same) to WAXD. The difference between WAXD and SAXS is that the distance between sample and detector is less and thus diffraction maxima are detected at larger angles. SAXS determines the porosity, crystallinity, size, and shape, whereas TEM provides visuals regarding internal structure, defects and spatial distribution. However, TEM is time consuming and dependent on sample thickness. Hence, SEM and AFM are preferred over TEM. Frechette et al. [46] have attempted to explain the nanocomposite characterization using atomic force microscopy (AFM). But due to its complexity, it is in its primary stage.

For thermal characterization of nanocomposites, techniques such as differential scanning calorimeter (DSC), thermogravimetric analysis (TGA), fourier-transform infrared (FTIR), rheometer, dynamic mechanical analysis etc. are used.

From experiments, it has been observed that organic substances such as polymers are sensitive to electron beams produced from microscopic methods, which may lead to breakdown of chemical bonds, chain scission, mass loss etc. This can be mitigated by improving instrumentation and the operational efficiency of microscopy, or by using alternate methods such as chemical staining, surface etching, microtomy etc. [47].

 

4. POLYMER NANOCOMPOSITES

4.1 Epoxy nanocomposites

Epoxy resins have continuously undergone through advancements and have proved to be economical solutions as electrical insulation. Since its commercial introduction in 1950s, epoxy resins are used extensively for indoor, outdoor and switchgear apparatuses due to its excellent properties such as high chemical and heat resistance, high impact strength and hardness, high breakdown strength etc. The preferred applications are in the low and medium voltage range. Highly improved characteristics of insulation through the use of nanofillers have given a new perspective for epoxy nanocomposites. Remarkable findings such as prolongation of tree initiation and propagation [37], improved performance towards voltage endurance and partial discharge studies due to mitigation of polarization at interfaces by Maxwell-Wagner effect, low permittivity, tan delta and resistivity values, significant improvement in corona resistance, reduced space charge [48-53] are reported.

The inconsistent results with different percentages of nanofiller in the field of epoxy nanocomposites could be due to various reasons, such as:

4.1.1 Inconsistencies in breakdown strength

This decrease is attributed to poor dispersion, absence of surface modification and inconsistent processing parameters [54]. The surface modification may help to improve the breakdown strength as shown in Fig. 5(a) and 5(b) respectively. This is because most of the modifiers are antioxidants which will increase the breakdown strength. The use of modifiers may increase the absorption currents resulting from polarization. These ions work as electrical charge carriers which will increase the dielectric properties of epoxy nanocomposites [40].

Fig. 5.Effect of (a) non-modification of surface, (b) surface modification of nanofiller on breakdown strength of epoxy nanocomposites [54].

4.1.2 Increase in permittivity [55]

Generally, permittivity values of nanocomposites should be lower than that of its base polymer. However, increase in permittivity values has also been reported. This can be due to the higher permittivity of nanofiller than the base polymer, as well as due to overlapping of interaction zones. Also, the temperature and frequency dependence of permittivity plays a vital role in resulting permittivity values of nanocomposites.

4.1.3 Inconsistencies in ‘Tg’ [56]

Inconsistencies reported by [55] and [56] are shown in Fig. 6. The decrease in ‘Tg’ is due to presence of water nanolayer around the nanoparticle [57], and increase [56] is due to as nanofiller content increases the interparticle distance reduces which increases the overlapping of immobile polymer regions around nanoparticle which will in turn increase ‘Tg’ [13].

Fig. 6.Inconsistencies in glass transition temperature values of epoxy nanocomposites [55,56].

Tracking and erosion problem in epoxy nanocomposites are widely studied because they permanently damage the insulation. Tracking generally occurs due to discharge resulting from contaminants such as salt, dust, humidity and atmospheric chemical agents surface of insulation present on conductive carbon path. High heat energy results from the surface of insulation. It involves the formation of discharge and low oxygen content leads to formation of carbonized paths. The temperature resulting from discharge can rise up to 1,000℃ which completely paralyses the insulation

In case of epoxy nanocomposites, negative DC voltage is more susceptible to tracking than positive DC and AC voltage as shown in Fig. 7(a), 7(b), and 7(c) respectively. The nanoparticle acts as oxygen barrier which resists the tracking phenomenon and reduces the leakage current as shown in Fig. 7(d). The enhanced resistance of nanocomposites to surface degradation is due to strong bonding between epoxy matrix and nanoparticle. Hence, epoxy nanocomposites are considered as substitute to its microcomposites [58]. Other findings are explained in Table 2.

Table 2.Epoxy nanocomposites.

Fig. 7.(a) Leakage current under AC voltage, (b) leakage current under positive DC voltage, (c) leakage current under negative DC voltage, and (d) tracking time of epoxy nanocomposites under different voltages [58].

4.2 XLPE nanocomposites

Since its introduction in 1960s, use of PE and XLPE is widely increasing due to its favorable electrical and mechanical properties such as high mechanical strength, flexibility, low cost, easy processing and highly chemical resistant [66].

Through the crosslinking, its state changes from thermoplastics to thermoset and operating temperature increases from 70℃ to 90℃. However, the use of XLPE in overhead and underground transmission is limited due to treeing and space charge formation which decreases its service life. Since the last decade, use of nanofiller in XLPE insulation has been gaining much attention. The XLPE nanocomposites have shown some excellent characteristics such as space charge mitigation, improved breakdown strength and resistance against water as well as electrical trees, which makes XLPE nanocomposite a promising material for cable insulation [67-72]. The major developments carried out in the field of XLPE nanocomposites is reviewed in the following discussion.

The selection of nanofiller which would give the best results from XLPE nanocomposites has always been considered as the toughest task for researchers. The nanofiller which gives the best reactions at the interfaces will be considered as suitable. Polymers are polar in nature and surface modified nanofillers are non-polar in nature. This property helps to achieve better dispersion, minimizes the probability of agglomeration and increases the hydrophobicity of nanoparticles [73]. An innovative approach towards the selection of nanofiller has been made [74] which leads towards an improvement in the DC electrical breakdown strength. In this study, three different nanofillers were studied. The one which almost gives same characteristics as that of XLPE microcomposite was selected through the thermally stimulated current (TSC) data. As continuation to this research, the researchers have come with a practical application for 250 kV DC-XLPE nanocomposite cables in 2012 between Hokkaido Island and Main Island of Japan. The reported enhancements in electrical properties of XLPE nanocomposite cable are tabulated in Table 3. Line Commutated Converter (LCC) based HVDC has the limitation against the polarity reversal condition. Basically, polarity reversal condition occurs when there is change in power flow direction. In case of LCC HVDC system, during polarity reversal condition the insulation breakdown strength reduces and space change influence gets increased. XLPE nanocomposites cable has given positive results against the polarity reversal condition. This cable is an encouraging attempt towards practical realization of nanocomposite cable. In continuation to this, the same researchers have also reported the pre-qualification testing of 400 kV XLPE nanocomposite cable for HVDC application [68].

Table 3.Comparison of XLPE and XLPE nanocomposite cable.

4.2.1 Water tree

The hydrophobicity of the nanoparticles does not play a vital role against the water tree development. However, non-surface treated nanoparticles are better water tree retardant than the surface treated nanoparticles [75]. The improved resistance of nanocomposites against the water tree growth is due to scattering of water trees due to inclusion of nanofillers [76]. The scattering also helps in increasing the breakdown strength [77]. Contrary to this, [75] it has been reported and concluded that surface treated nanoparticles are more water tree resistant than nonsurface treated one [76]. This could be due to treating nanofillers with different surfactants. Also, it is worthwhile to mention that the method of crosslinking cannot cease the formation of water tree. To do so, some additives are added which helps to mitigate the water tree length [78-80]. The enhanced tree retardant-XLPE (TR-XLPE) [81] cable developed by Dow Chemical Company, USA has reported to have a 70% greater lifetime than TR-XLPE cable. MgO/XLPE nanocomposite is better water tree resistant than MgO/LDPE nanocomposite for lower nanofiller content (< 2%) [82]. The length of water tree also depends on the compatibilizer [83] used to obtain good dispersion for montmorillonite (MMT) nanofillers. Proper compatibilizer greatly decreases the water tree length [84,85].

4.2.2 Other dielectric properties

The surface treated nanofillers are more resistant to moisture uptake than untreated nanofillers for shorter period. However, this is not true for longer periods. Other than the positive effects such as space charge mitigation and increased resistance to water tree, a humid environment results in an increase in dielectric loss and decrease in AC breakdown strength under the influence of moisture, which are highly undesirable [86,87]. This decrease in breakdown strength can be due to poor bonding between nanoparticles and host polymer.

The crosslinking by-products such as methane, cumylalcohol, acetophenone and -methylstyrene also affect the electrical characteristics (e.g. it decreases volume resistivity and it increases the dielectric loss). It also leads to formation of heterocharge and absence of it leads to homocharge or no-charge [88]. The concentration of by-products depends on the temperature used for crosslinking. Hence, researchers have adopted degassing and annealing processes to remove the by-products [88-92]. Additionally degassing reduces the space charge accumulation [88]. Apart from degassing temperature conditions, the space charge formation is also dependent on the thickness of insulation under temperature gradient condition [93]. Other important findings in the field of PE nanocomposites and other nanocomposites are tabulated in Table 4 and Table 5 respectively.

Table 4.PE nanocomposites.

Table 5.Other nanocomposites.

 

5. DISCUSSION

The models hypothesized individually by Lewis, Tanaka, Tsagaropoulos and Eisenberg, Todd and Shi are not supported with experimental results w.r.t. electrical characteristics such as tree propagation, partial discharge, etc. They are unable to clarify the mechanisms which lead to the overall enhancement of properties through the use of nanocomposite insulations, which requires detailed experimentation and analysis. This can be supported by use of different techniques such as SEM, TEM and AFM. Tanaka has explained the effect of different phenomenon on different layers of multi-layer model.

Quantitative (finite element analysis) and experimental approaches [15] are very good attempts in understanding interface reactions. The permittivity of nanocomposite can be calculated if permittivity of polymer, interface and nanofiller are known. This model is based on a 2-D approach, which can be carried forward for 3-D analysis for better understanding of interface reactions.

From the above tabulated analysis in Table 1, Table 2, Table 4, and Table 5, it is rather difficult to have benchmark standards for the nanocomposite systems. The variation of permittivity, dielectric loss, and breakdown strength are frequency dependent. Also, their variation according to nanofiller content, are unpredictable. This prediction totally relies on homogeneous dispersion. The surface modification of nanofiller helps to achieve a better dispersion but that has an adverse effect on other properties as mentioned. Several researchers have analyzed the MNC which shows some promising results. In anticipation, the selection and content of microfiller content is again the crucial issue. Better results can be obtained for MNCs through the combination of micro and nanofiller content which have yielded best results individually [42].

In case of epoxy nanocomposites, several inconsistent results in terms of dielectric breakdown strength are observed [59,60]. The same nanofiller with same epoxy matrix gives different results which may be due to poor dispersion of nanofiller and method of preparation as in Fig. 4. The micro-nanocomposite performs better than nanocomposite. Also, surface treated nanoparticles aids in improvement in dielectric properties than non-surface treated nanocomposites. Epoxy/silica nanocomposite perform better than epoxy/alumina nanocomposite. This may be due to permittivity difference between silica and alumina.

XLPE has shown promising results. The main problem in XLPE is the electrical and water treeing phenomenon’s. The crosslinking mechanism also plays a vital role in XLPE characteristics. The by-products produced due to peroxide based crosslinking affect the electrical characteristics. Hence, silane grafting can be considered as an alternative over peroxide based XLPE nanocomposites. Radiation based crosslinking has stronger matrix bond strength, less curing time, and temperature. Hence, it can also be preferred over first two methods but it requires high investment.

The XLPE nanocomposites are generally prepared using twin screw extruder or lab scale kneader machine. The crosslinking temperature and time should be maintained within permissible limits since XLPE is difficult to extract it from the machine once it gets cross-linked. This would lead to an increase in the maintenance cost. Ultrasonication, which is well suited for LDPE, can be preferred to obtain good dispersion in XLPE nanocomposites.

The above discussed factors should be closely monitored for obtaining better nanocomposite characteristics. Briefly, the material prepared should be easily degradable. The bio degradability is also one of severe issue in polyethylene. This can also stand as future scope for PE based nanocomposites. The degradability issue can also be solved through the combination of XLPE and Polyhydroxybutyrate (PHB), which is an easily degradable plastic.

Challenges and opportunities:

The determination of nanofiller content which can give an optimal performance towards electrical, thermal and mechanical characteristics of nanocomposites, has always been a crucial task. Surface modification of nanofiller has not always met the expected outcomes. The surface degradation observed due to addition of nanofiller is the major problem in nanocomposites.

Amidst all advantages from the nanoparticles to electrical insulation development, it poses potential safety and health risk to workers. This involves the hazard to personnel due to inhalation exposure of dry nano powder and nanoparticles suspended in liquid during transportation. Since nanoparticles are attached to polymer matrix in nanocomposites, it may not pose a severe risk. The assessment of effect of toxicity of nanoparticle exposure has been done on animals, as reported in the report published by National Institute for Occupational Safety and Health (NIOSH), which can be further carried out for nanocomposites [111].

In coming days, the use self-healing polymer nanocomposites with the addition of healing agent will play vital role in electrical insulations. These types of nanocomposites show some excellent properties such as automatic healing, ability to heal the damage for multiple times, reduced maintenance cost, and possess almost the same performance as traditional materials [112-114].

These materials are prepared by adding liquid healing agents such as monomers, dyes, catalysts etc. If cracks occur, then the liquid agents are incorporated through capillary force and it heals the crack.

Self-healing agents are generally designed using following methods [115]:

a. Release of healing agents b. Reversible cross-links c. Miscellaneous technologies

The main difference between method (a) and (b) is that method (a) is autonomous, i.e. self-repairing, and method (b) is nonautonomous. In reversible crosslinking, an external trigger such as thermal, photo or chemical excitation is required.

It is well known that the water tree gets initiated after the formation of cracks on the surface of polymer due to its chain scission. Once the water tree is formed, the insulation is permanently damaged. This would pose a serious threat for system under operation. Recently through the use of nanocomposites, it has been observed that nanofillers are able to impede the growth of tree and they are unable to resist the formation of water tree fully.

Hence, for such cases to avoid permanent damage to cable insulation, self-healing polymer nanocomposites can be used. The function of self-healing agents is shown by Fig. 8. As seen in Fig. 8(b), the healing agent gets poured into a crack, leading to its closure (Fig. 8(c)). Use of self-healing polymer nanocomposites will be a new perspective for researchers in the field of dielectrics and insulation.

Fig. 8.Role of self-healing agent in crack embedment [115].

As previously discussed, surface degradation due to partial discharge phenomenon is the main concern in epoxy nanocomposites. Hence, use of self-healing agents with epoxy nanocomposites may serve the purpose, since self-healing materials improve the mechanical strength of the composite.

 

6. CONCLUSIONS

In this paper, different aspects of nanocomposites documented in recent literature are reviewed. There are several issues discussed by researchers such as theories to explain the interface reactions, and different methods to obtain the better dispersion of nanofiller. The earlier study shows that the practical application and commercial production of nanocomposites has not been fully developed for electrical insulation. The survey shows that in spite of inconsistencies in results, researchers have been extensively working to explore the potential of nanotechnology to optimize the performance of nanocomposites towards electrical, mechanical and thermal characteristics of insulation.

The use of nanocomposite in electrical insulation will increase due to its highly encouraging advantages through small quantities of nanofiller (0~10 wt%). Also, use of self-healing nanocomposites may acquire the considerable attention in coming days. Furthermore, the interdisciplinary studies in the field of electrical engineering, computer science, physics and chemistry may accelerate the research to discover new areas of applications. However, the cost reduction and safety issues will also play major roles in its production and implementation.