<![CDATA[Dielectric spectroscopic measurements on transformer oil-paper insulation under controlled laboratory conditions]]>

A. Setayeshmehr et al.: Dielectric Spectroscopic Measurements on Transformer Oil-paper Insulation

1070-9878/08/$25.00 © 2008 IEEE

1100

Dielectric Spectroscopic Measurements on Transformer Oil-paper Insulation under Controlled

Laboratory Conditions

A. Setayeshmehr1, I. Fofana2, C. Eichler1, A. Akbari1, H. Borsi1 and E. Gockenbach1

1Institute of Electric Power Systems, High Voltage Engineering Section (Schering- Institute)

Leibniz Universität Hannover, Callinstr. 25 A, D-30167 Hanover, Germany 2Canada Research Chair, tier 2, on Insulating Liquids and Mixed Dielectrics for Electrotechnology,

University of Quebec at Chicoutimi, 555, Boulevard de l’Université, G7H 2B1, Chicoutimi, Qc, Canada

ABSTRACT For reliable operation of power transformers, the condition of the insulation system is essential. This paper reports on a detailed study of the effect of ageing, temperature and moisture on frequency and time domain spectroscopic measurements carried out on oil-impregnated pressboard samples as well as on a distribution transformer under controlled laboratory conditions. Because field measurements are generally performed after de-energizing the transformer, extreme care is required in interpreting the results due to inherent temperature instabilities. To avoid large thermal variations that may affect the results, a customized adiabatic room was built around the transformer for measurements above the ambient. Capacitance ratio and direct current conductivity deduced from the spectroscopic measurements, helped to interpret the data. Because, low frequency measurements techniques are time consuming, alternative to a transfer of time domain data into frequency domain data was investigated.

Index Terms — Dielectric spectroscopy, time domain, frequency domain, conductivity, Aging, moisture, temperature, oil-paper insulation.

1 INTRODUCTION

TRANSFORMERS are the “heart” of any electric power distribution and transmission systems and it is essential that they function properly for many years. In most of power transformers, the main insulation is a combination of mineral oil with cellulose materials. When this insulating paper is adequately impregnated with oil, it offers the user a material with insulating and mechanical properties of remarkable suppleness. The good dielectric properties combined with the wide availability and cost benefit of oil paper insulations have, therefore, made these materials, transformer insulation of choice for nearly a century [1]. However, this insulation system deteriorates in service due to service conditions. The reliable performance depends on its basic character which may affect the performance of the transformer. During service, a variety of processes occur, some of them being inter-related, which degrade the insulation. The degradation/aging of transformer insulation is recognized to be one of the major causes of transformer breakdown [2-5]. The cost of premature and unexpected failure of a power transformer can be several times its initial cost. There is not

only the refurbishment or replacement cost but also possible costs associated with clean-up, loss of revenue, and deterioration in quality of power delivery. With increasing age, there are potential risks of extremely high monetary losses due to unexpected failures and outages. Condition monitoring of the insulation of transformers has become an important issue since many transformers in electrical industries around the world are approaching the end of their design life. Indeed, condition monitoring can be utilized to attempt the prediction of the insulation condition and the remaining lifetime of a transformer. In this context, the adequacy of existing and the application of new diagnostic tools and monitoring techniques gain increasing importance. Increasing requirements for appropriate tools to diagnose power systems insulation non-destructively and reliably in the field drive the development of diagnostic tools based on changes of the dielectric properties of the insulation. Some of these modern diagnostic methods include the Recovery Voltage Measurement (RVM), Frequency Domain Spectroscopy (FDS) and Polarization and Depolarization Current Measurements (PDC) [4]. These two later became only recently available as user-friendly methods, and can be used to monitor, diagnose and check new insulating materials, qualification of insulating systems during/after production of power equipments non-destructively [6, 7]. Manuscript received on 19 October 2007, in final form 14 March 2008.

IEEE Transactions on Dielectrics and Electrical Insulation Vol. 15, No. 4; August 2008

1101

Field measurements are generally performed just after de-energising the transformer. Under such circumstances, large thermal variations may affect the results, since moisture distribution inside the insulation is not in complete equilibrium condition [8, 9]. Because, spectroscopic measurement results are highly operating conditions dependant, care are required to interpret them. Low frequency measurements being particularly long time consuming process, large temperature instability may affect interpretations. In this contribution, spectroscopic measurements (FDS and PDC) were performed on a distribution transformer, at temperature above the ambient, moisture content in oil acted as parameter. To reduce the influence of large thermal variations, a customized adiabatic room was built around the transformer. The obtained results are discussed in regard to the wetness and temperature influence on the insulating system. In order to analyse moisture content, insulation temperature and ageing, independent on spectroscopic measurements, investigations have been performed on oil-impregnated pressboard samples under controlled laboratory conditions. Measurement duration in time domain being shorter than frequency domain ones, PDC measurement were transformed in frequency domain using Fourier Transformation. The obtained results depict good agreement with measured one and indicate the feasibility of using low frequency data to separate moisture, ageing and temperature effects.

2 DIELECTRIC SPECTROSCOPY TECHNIQUES

Dielectric spectroscopy in time or frequency domain offers new opportunities for an off-line, insulation condition assessment of HV electric power equipment and its predictive maintenance non-destructively and reliably in the field. These techniques are global methods, i.e. each test object is regarded as a “black box” accessible only by its electric terminals. Therefore, only global changes of the insulation can be identified but not localized defects [6]. Because, results from these tests are highly operating conditions dependant, practical measurements issues should be considered [6-15]: • Inherent to all dielectric spectroscopy measurements in either

time or frequency domains is their “off-line” character, i.e. equipment in operation must be removed from service before performing measurements.

• Dielectric measurements require constant insulation temperatures during application for accurateness, as the polarization phenomena are temperature dependant. Adequate experience and extreme care are needed to interpret the results in the presence of temperature variations and thermal instability within the equipment.

• The charging time/period for PDC measurements should be long enough to allow all polarization processes to be completed. For good results, a minimum polarization time of 10,000 s should be used on large power transformers.

• Rain seems to generate leakage dc currents which superimpose the desired measurement currents. The magnitude depends on

the rain intensity and sometimes exceeds100 nA, what makes a measurement difficult to interpret.

• As measurement instruments used are usually quite sensitive to electromagnetic disturbances, the “electromagnetic compatibility” of such instruments must be guaranteed. Therefore, the test voltage levels of the instruments cannot be too low.

• The maximum polarization voltage is determined by the transformer’s geometry and the resulting electrical field strengths in the oil-paper-insulation. If field strength exceeds 10 V/mm, different effects (charge injection from the electrons and field-induced dissociation leading to apparent higher oil conductivity) can lead to non-linear effects. The polarization voltage should be kept as low as possible. A polarization voltage between 100 and 500 V has led to satisfying results on all measurements performed on various distributions and power transformers in power plants and switchyards.

The fundamental theories behind dielectric measurements are already well known [6] while dielectric phenomena are discussed in Jonscher’s publications [16, 17]. However, to facilitate interpreting the measurements presented in sections 3 and 4, there is a short review that includes theory behind time and frequency domain measurement techniques.

2.1 TIME DOMAIN SPECTROSCOPY

The measurement of polarization and depolarization currents (PDC) following a dc voltage step is one way in the time domain to investigate the slow polarization processes [4, 6-13]. The dielectric memory of the test object must be cleared before the PDC measurement. The voltage source should be free of any ripple and noise in order to record the small polarization current with sufficient accuracy. The procedure consists in applying a dc charging voltage of magnitude Uc to the test object for a long time (e.g., 10,000 s). During this time, the polarization current Ipol(t) through the test object is measured, arising from the activation of the polarization process with different time constants corresponding to different insulation materials and to the conductivity of the object, which has been previously carefully discharged. Then the polarization (or absorption, or charging) current Ipol(t) through the test object can be expressed by [6-11]:

( ) ( ) ( )⎥⎦

⎤⎢⎣

⎡++= ∞ tftUCtI

o

ocopol δε

εσ (1)

where: Co : geometrical capacitance of the test object, Uc : the step voltage (charging voltage), σo : the dc conductivity of the dielectric material, εo : 8.852 10-12 As/Vm is the vacuum permittivity, ε∞ : the high frequency component of the permittivity, δ(t): the delta function arising from the suddenly applied step voltage at t = t0. f(t) : the response function of the dielectric material. The geometric capacitance of a core type transformer can be estimated by the cylindrical capacitance (equation (2)), where h is the average winding height and ra and rb are, respectively, the inner and outer radius of the insulation between windings [8].

A. Setayeshmehr et al.: Dielectric Spectroscopic Measurements on Transformer Oil-paper Insulation 1102

⎟⎠⎞⎜

⎝⎛

=

a

br

rhC

log

200

πε (2)

If the design of the transformers insulation is available, the geometric capacitance can be calculated. Otherwise, geometric capacitance can be estimated by measuring the capacitance Cm between the two terminals of the insulation system under test (it can be measured with any capacitance measuring ac bridge at/around the power frequency) and dividing by the effective relative permittivity εr of the combination of the composite oil-paper insulation system (Co = Cm/εr) [8]. The effective permittivity of oil impregnated pressboard samples may be written in terms of paper and oil permittivity (εpaper and εoil respectively) as [18]:

( ) XX1.

oilpaper

oilpaperr ε+−ε

εε=ε (3)

where X is the relative amount of paper in the composite system. The range of X is typically 20% to 50% [18] for a transformer. The voltage is then removed and the object is short-circuited at t = tC, enabling the measurement of the depolarization current (or discharging, or de-sorption) Idpol(t) in the opposite direction, without contribution of the conductivity. The polarization current measurement can usually be stopped if the current becomes either stable or very low. According to the superposition principle the sudden reduction of the voltage UC to zero is regarded as a negative voltage step at time t = tc. Neglecting the second term in (1) we get for t = (t0 + TC) [6- 10]:

( ) ( ) ( )[ ]ccodepol TtftfUCtI +−−= (4)

where Tc is the charging time of the test object. Figure 1 shows the schematic diagram of the PDC measuring technique while Figure 2 shows the typical nature of these currents due to a step charging voltage UC [6].

Figure 1. Principle of test arrangement for the “PDC” measuring technique.

The insulation between windings is charged by the dc voltage step Uc. A long charging time is required (10,000 s) in order to assess the interfacial polarization and paper condition. The initial time dependence of the polarization and depolarisation currents (<100 s) is very sensitive to the conductivity of the oil while the moisture content of pressboard influences mainly the shape of the current at much longer time [12].

Figure 2. Principle of polarization and depolarization current.

The dc conductivity σo, of the test object can be estimated from the PDC measurements currents. If the test object is charged for a sufficiently long time so that f(t + tc) ≅ 0, the dielectric response function f(t) is proportional to the depolarisation current. Using (1) and (4), it yields :

0

0

0)(

εσ

−=C

pol

UCi

tf (5)

C

depol

UCi

tf0

)(−

≈ (6)

(1) and (2) can be combined to express σo as:

( ) ( )( )bdepolbpolC

o titiUC

−≈0

0εσ (7)

Even without performing direct conductivity measurements on a oil sample, the oil conductivity can then be calculated using equation (7) where the ipol(tb) and idepol(tb) are the initial values in polarization and depolarization currents [18]. In the same way, the conductivity of the paper can be estimated from the long term values of the polarisation and depolarisation currents, by replacing the tb to tm, where tm represents the largest value of time, for which the currents have been measured, that is tm = 10,000 s for our investigations.

2.2 FREQUENCY DOMAIN SPECTROSCOPY Dielectric response in the frequency domain is another alternative method to study the polarization phenomena. While the time domain response corresponds to the variation of a current in time and is therefore a real function f(t) of the real time variable, frequency domain response defines two components of amplitude variation in phase and in quadrature with respect to the driving harmonic signal and has to be defined as real and imaginary functions χ(ω) and −jχ(ω) of the real frequency variable, where ω = 2πν is the frequency, ν (Hz) is the circular frequency in hertz and 1−=j . Assuming the test object is a complex capacitance, the relationship between the applied voltage U(ω) and measured current I(ω) can be written as follow:

)()()( ωωωω UCjI = (8)

[ ] )()('')(')( ωωωωω UjCCjI −= (9)

where C’(ω) and C”(ω) are the real and imaginary components of the dielectric capacitance.

Uc

Ipol (t)

Idepol (t)

to TCtc t

Uc

Ipol Idepol

Electrometer

Test object


1103

Applying Fourier transform to (4) it yields [6-11]:

( ) ( )( )

( )

( )

( )ωωχωε

σωχεωω

ωεωε

UjCjIo

oo

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡

⎟⎟⎠

⎞⎜⎜⎝

⎛+−+= ∞

44 344 2143421

"'

"' (10)

where χ(ω) = χ’(ω) - jχ’’(ω) is the Fourier Transform (FT) of the dielectric response function f(t) and defined as the complex dielectric susceptibility. Given that ε(ω) = ε’(ω) - jε’’(ω), the loss factor tan δ in frequency domain can be defined as follows [5–9]:

( ) ( )( )

( )( )

( )

( )ωχε

ωχωε

σ

ωω

ωεωεωδ

'

"

'"

'"tan

+

+===

∞

o

o

CC (11)

The importance of FDS measurements lies in the fact, that they allow the frequency scan of test object electrical properties such as the real (ε’) and imaginary (ε”) parts of the dielectric permittivity [1] as well as the dissipation factor (tan δ), power factor (PF) and capacitance (C) [19].

2.3 PDC OR FDS?

Analysis of the PDC measurements can provide reliable information about the condition of transformer insulation by PDC. This non-destructive method can provide the moisture content in the solid insulation material and the conductivities of the oil and paper. Other diagnostic quantities like tan δ, polarization index and polarization spectra can be calculated from PDC measurements directly. Dielectric frequency domain spectroscopy (FDS) enables measurements of the composite insulation capacitance, permittivity, conductivity (and resistivity) and loss factor in dependency of frequency. The real and imaginary part of the capacitance and permittivity can be separated. This non-destructive technique also provides the moisture content in the solid insulation material and C-ratio diagnostic quantity. A brief comparison reveals that FDS has better noise performance and separates the behaviour of polarizability (χ”) and losses (χ’) of a dielectric medium, while the dielectric response of an insulating system can be measured with the PDC method in shorter times and with a good accuracy [20]. Both methods appear to have their own strengths and weaknesses. There still exists a need for gathering more experiences before anyone can recommend one of them.

2.4 CONVERTING TIME DOMAIN MEASUREMENTS TO FREQUENCY DOMAIN

The dielectric response of an insulating system can be measured with the PDC method in shorter times and with a good accuracy. The FDS method separates the behaviour of polarizability (χ”) and losses (χ’) of a dielectric medium and may be preferred for the interpretation of the results. Therefore, the transformation of PDC data to frequency domain data is important. The duration to get time domain information up to 2,000 s lasts 1.2 h with PDC and 2 to 2.5 h with FDS. Up to more than 2,000 seconds is necessary to get a single value of the

insulation properties measurement at a frequency of 1 mHz. Changing the lower frequency to 0.1 mHz increases the measurement time from 1.1 h to 12 h, mostly improper for on site measurements [15]. Because, low frequency measurements techniques are time consuming processes, on going works enabling the transfer of time domain measurements (TDM) to frequency domain is gaining attention as alternative [20-22]. Under the assumption that the dielectric material is linear, PDC measurements performed at low field (field strength below 10 V/mm) can be transformed in frequency domain using Fourier Transformation. From our experiences with time domain to frequency domain transformations [20-22], the accuracy of the error in Fourier Transform was found to be less than 5% of the measured value. χ(ω) being the FT function of the dielectric response function f(t) and the complex dielectric susceptibility, it yields:

dtetfFjtjω

ωωχωχωχ−+∞

∫==−=0

)()()('')(')( (12)

( )∫+∞

−=−0

)sin()cos()()('')(' dttjttfj ωωωχωχ (13)

The real and imaginary parts of C(ω) can be yield as follows according to (10):

∞+∞

+= ∫ CdtttiU

C dpolC 0

)cos()(1)(' ωω (14)

∫+∞

=0

)sin()(1)(" dtttiU

C polC

ωω (15)

where C∞ is the high frequency capacitance of the test specimen. Measured polarization and depolarization currents can be fitted with several exponential functions using MATLAB Curve Fitting Tools Box, that is:

tn

tt neCeCeCti λλλ −−− +++= .....)( 2121 (16)

The real and imaginary part of the complex electric capacitance in frequency domain can be calculated as follows according to the polarization and depolarization exponential fitted currents.

∞=

++

= ∑ CCU

Cn

k k

kk

C 122

1)('ωλ

λω for depolarization (17)

∑= +

=n

m m

m

C

CU

C1

221)("

ωλω for polarization (18)

where 1/λk is the relaxation time constant and Ck is the coefficient related to charging voltage, charging duration and relevant relaxation branch parameters.

3 MEASUREMENTS ON A DISTRIBUTION TRANSFORMER

3.1 EXPERIMENTAL SETUP

A 220 V/35 kV, 100 kVA oil filled distribution transformer, which was in service for thirty years, has been removed from service since one year and stored in laboratory at room temperature. Under this condition the water equilibrium in cellulose/mineral oil systems can be used to evaluate the


moisture content in the paper. The water content in the oil was measured at 50 ppm at room temperature, before starting spectroscopic measurements. Using equilibrium curves of oil-paper system, the water content in the solid insulation is to be higher than 10% [23]. The transformer tank was directly earthed and excluded from the current measuring circuit, thus the test was not influenced by air humidity or surface leakage of any bushings. The polarisation current is therefore the combination of the absorption current due to polarization phenomena and the conduction current caused by conduction phenomena. Figure 3 shows the test arrangement used during the PDC measurements with a stabilised dc power source up to 4,000 V, a digital multi-meter for the current measurements with an uncertainty of 1% and a number of relays in order to operate the test setup by a computer.

Figure 3. Simplified diagram of dielectric response measurement (time domain) for transformer. The user-friendly interface of the developed software enables the operator to choose the voltage and time for charging and discharging. Once the operator sets the system into operation the measurement system becomes fully automated. For measurement purpose the switch S1 is closed, so that the polarization current flows in the transformer and decreases to nearly zero during charging time. After polarization duration, which lasted 10,000 s, switch S1 is opened and the transformer short circuited by switch S2. Similar to the polarization duration, a depolarization current in this stage flows, but in another direction. Both currents are stored for analysis in the computer. In order to perform the measurements at different temperatures, a customized adiabatic room has been build around the transformer (Figure 4). The transformer temperature was monitored using three Pt100 temperature sensors placed at different places on the tank. The values varied within a range of about ±3 °C. The mean temperature inside the transformer was then calculated as average value of the three sensors. The low voltage windings were heated through a short circuit transformer and this later, de-energized before the measurements. Since FDS measurements at very low frequencies increases the measurement time up to 12 h,

consequently, during the measurements, a decrease in the temperature up to about 10 °C was observed in some cases. The dielectric responses in the frequency domain performed with the insulation diagnostic Analyser IDA200 [19] gives the dielectric dissipation factor (DDF) and the capacitance C(ω) at discrete frequencies. The frequency range of the instrument was 10-4

to 103 Hz and the applied effective voltage 140 V,

while time domain measurements were performed with 200 V from 1 to 10,000 s.

Figure 4. Overview of the customized adiabatic room without the cover.

3.2 RESULTS

PDC measured at different temperatures on the transformer, plotted in a log-log scale, are presented in Figure 5.

BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

BBBBBBBBBBBBBBBBB

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJ

JJJJJJJJJJJJJJJJJJ

JJ

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHH

H

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFFFFFF

FFFFFFFFFFFFFFFF

3333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333

33333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333

333333333333333333333333333333333333333

3333333333333333

11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

1111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

1111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111

111111111111111111111111111111111111111111111111111111111111111111111111111

11111111111111111111111111111111111111111111

111111111111111111111111

111111111111111111

11

1E-10

1E-9

1E-8

1E-7

1E-6

1E-5

1E+0 1E+1 1E+2 1E+3 1E+4

Cur

rent

(A)

Time (s)

B Polarisation, T=50°C

J Depolarisation, T=50°C

H Polarisation, T=20°C

F Depolarisation, T=20°C

3 Polarisation, T=25°C

1 Depolarisation, T=25°C

Figure 5. PDC measured on the transformer at three different temperatures. The water content of the insulating oil measured at 20°C was 50 ppm.

From Figure 5, it can be seen that PDC current magnitudes increase with rising temperature. This is in agreement with results reported in the literature [8]. This phenomenon seems to be related to the loss of linearity at higher temperature. Spontaneous polarization is to be known as temperature-dependent. Any change in temperature causes a change of the dipole moments, measurable as a change of electric charges at both ends of the polar axis. Conductivity values for both oil and paper have been calculated from the measured polarisation/depolarisation

HV LV

Transformer

DC Source Controller

DC Voltage

Filter

pA

S1 (HV Relay)

Digitalmultimeter

R

S2 (HV Relay)

PC

s

A


1105

currents using equation (5) and are presented in Table 1. The conductivity of both paper and oil increases as temperature increases. This means that at higher temperatures, the condition of insulation worsens.

Table 1. Oil and paper conductivity as function of temperature.

Temperature (°C) σoil (pS/m) σpaper (pS/m)

20 0.77 0.5 25 1.2 0.55 50 26.5 5.7

Increasing temperature changes the DC conductivity of the dielectric materials as well as the polarisation phenomena. The magnitude of the data is to be related to the exceptional wetness of the transformer. The insulation conductivity is known to vary with temperature T (in Kelvin) according to the well-known Arhenius relationship [14]:

⎟⎠⎞

⎜⎝⎛−=σ

kTEexpA ac (19)

where Eac is the activation energy and Α a constant related to ions mobility in the insulation. Both oil and paper conductivities are found to increase exponentially with temperature. The validity of this equation in fitting oil and paper conductivity versus temperature has been demonstrated by Saha [8]. In order to access the effect of temperature on the PDC measurements when the insulation is less moistened, an on line continuous drying of the transformer via cooled cellulose filter has been performed. Details concerning this new technique, allowing a gentle, continuous desiccation without influencing the Dissolved Gas Analysis (DGA), developed at the Schering Institute at the "Leibniz Universität Hannover" can be found in [24]. During a continuous drying process, a mass transfer process of water results from the consequent imbalance, where moisture transfers from the paper to the oil via diffusion. Performing the PDC measurements just after the drying process will not reflect the true insulation condition, since complex dynamic processes occur as moisture diffuses. Consequently, reasonable time delay (about 2 weeks) has been given before commencing measurements, to attain stable temperature and moisture equilibrium. Measurement has therefore been performed weeks after water content in oil stabilised at an expected value. In Figure 6, the measured polarization and depolarization currents on the transformer at two different temperatures are plotted. Drying the transformer shows a significant reduction in the polarisation/depolarisation currents. The initial time dependence of PDC currents is well known to be very sensitive to oil conductivity while moisture content and conductivity of the pressboard influences mainly the shape of the currents at long time range (i.e. dc stationary current) [6, 7, 11, 12]. The high conduction of the insulation system, at higher temperature, can be easily identified by the large difference between the polarisation current and the depolarization current (Figure 6). Regardless the moisture content, it can be observed

that both the polarisation and depolarisation currents increase with temperature increase.

1E-10

1E-9

1E-8

1E-7

1E-6

1E-5

1E+0 1E+1 1E+2 1E+3 1E+4

Cur

rent

(A)

Time (s)

Polarisation T = 45°C

Polarisation T = 20°C

Depolarisation T = 20°C

Depolarisation T = 45°C

s

A

Figure 6. PDC Measurement result on the distribution transformer at two different temperatures, with a step charging voltage U0 = 200 V. The moisture content in oil measured at 20°C was 20 ppm.

Again, conductivity values for both oil and paper have been calculated from the measured polarisation/depolarisation currents using equation (5) and are presented in Table 2.


Temperature (°C) σoil (pS/m) σpaper (pS/m)

20 0.43 0.22 45 12.15 2.3

The conductivities of both paper and oil increased as temperature increased. The effect of moisture on PDC measurements are illustrated in Figure 7.

1E-10

1E-9

1E-8

1E-7

1E-6

1E-5

1E+0 1E+1 1E+2 1E+3 1E+4

Pol

aris

atio

n C

urre

nt (A

)

Time (s)

Polarisation, T = 50°C, m.c.=50 ppm

Depolarisation, T = 50°C, m.c.=50 ppm

Polarisation, T = 45°C, m.c.=20 ppm

Depolarisation, T = 45°C, m.c.=20 ppm

s

A

Cur

rent

Figure 7. PDC Measurement results at 20°C on the distribution transformer at two different moisture content, with a step charging voltage U0 = 200 V.

Figure 7 shows that the amplitude of long term dc polarisation/depolarisation current is very sensitive to the moisture content in oil and consequently in the insulation system. The capacitance and DDF of a dielectric is a complex function of at least two variables - frequency and temperature, although moisture and pressure may be other physical variables [16, 17]. A complete representation should therefore be "three-dimensional" plots, but these are cumbersome and are therefore seldom employed, although modern computer graphics enable one to plot third-angle projections in two dimensions. The prevailing method of representation consists


therefore in plotting the frequency dependence with temperature and/or moisture content as parameter. The frequency scan of the loss factor and capacitance are given in Figures 8 and 9. The dissipation factor is very high for the whole frequency range, which reflects the bad condition of oil and paper.

0

0,5

1

1,5

2

2,5

0,0001 0,001 0,01 0,1 1 10 100 1000Frequency

T = 45°C - W = 20 ppm

T = 45°C - W = 40 ppm

T = 20°C - W = 20 ppm

T = 20°C - W = 40 ppm

tan

δ

Hz

Figure 8. Dissipation factor measurement on the distribution transformer at two different temperatures and water content in oil.

0E+0

2E-8

4E-8

6E-8

8E-8

1E-7

1E-7

1E-7

0,0001 0,001 0,01 0,1 1 10 100 1000

Cap

acita

nce

Frequency

T = 45°C - W = 20 ppm

T = 45°C - W = 40 ppm

T = 20°C - W = 20 ppm

T = 20°C - W = 40 ppm

Hz

F

Figure 9. Capacitance measurement on the distribution transformer at two different temperatures and water content in oil.

The higher the temperature and water content, the higher are the dissipation factor and capacitance at lower frequencies. Quantities as capacitance ratio (CR) which are important for generator and motor insulation diagnosis [17] can directly be "read" from the capacitance frequency scan. This is the ratio between C (at 0.1 mHz) and C (50 Hz) capacitances. For a high voltage transformer where the insulation system consists of the oil duct in series with the pressboard, as a “rule of thumb”, the ratio of the capacitance (CR) is about 3-5, for good insulation [12]. Table 3 summarises the C-ratio computed from values given in Figure 9. Table 3. C-ratio computed from values given in Figure 9.

Experimental conditions C-ratio T = 45°C and W = 20 ppm 41 T = 45°C and W = 40 ppm 57.26 T = 20°C and W = 20 ppm 33 T = 20°C and W = 40 ppm 51

Out of Table 3, it can be seen that, the higher the C-ratio, the wetter the insulation of the transformer. Moisture changes the dielectric constant of the equivalent capacitor so it also affects CR. The influence of temperature is also noticeable, since for

given moisture content, the CR increases with temperature rise. It should be noted that during dielectric response measurement at temperatures above the ambient temperature, temperature decreases up to about 10 °C in some cases, particularly during low frequencies measurements. This situation reflects onsite measurements performed just after de-energising the transformer. Even though at higher temperatures only a negligible part of water migrates out of cellulose into oil, the results may be affected to some extent. The authors will be engaged in investigations including the monitoring of the insulation temperature during dielectric measurements for temperature correction during unsteady temperature states in order to give a closer insight to the influence of temperature transient on polarisation phenomena. An alternative to time consuming low frequency measurements is the conversion of measuring results from TDM by PDC analysis [20-22]. Time domain (PDC) measurements were carried out with constant voltage 200 V from 1 to 10,000 s and the obtained results were converted into frequency domain as presented in section 2.3. The obtained results were compared to those measured in Frequency domain using the insulation diagnostic Analyzer IDA 200 [19] with an applied voltage 140 V from 0.0001 to 1000 Hz. The results shown in Figure 10 are in a good agreement.

1.0E-06

1.0E-04

1.0E-02

1.0E+00

1.0E+02

0.0001 0.001 0.01 0.1 1

Frequency

Tan

delta Measured Tan delta

Converted Tan delta

Measured Capacitance

Converted Capacitance

Hz

1.0E-01

1.0E-03

1.0E-05

1.0E-07

1.0E-09

F

Figure 10. Comparison with experimental data of the DDF and capacitance of the transformer obtained from FT of PDC measurements at a temperature about 20 °C The moisture content in oil was 20 ppm at 20 °C.

Additional results and methods concerning TDM to low frequency domain measurements can be found in the literature [20-22]. The proposed tools indicate good accuracies with FDS measurements and appear as promising alternative to time consuming low frequency measurements.

3.3 ON THE CONTEXT OF FIELD MEASUREMENTS

For on-site measurements, generally performed just after de-energising the transformer, whereby the temperature inside the transformer may experience large variations. For example, the transformer is switched off at an operating temperature of 65 °C, the dielectric measurement starts at 60 °C and in the end of the measurement the transformer is at 25 °C. Thus at Onsite measurements the water migration is commonly running, the transformer is in a non-equilibrium state. Saha et al [8, 25] reported PDC measurements on transformer under thermal transition (the temperature

Cap

acita

nce


1107

decreased during PDC measurements from 60 °C to 23 °C) and compared the results to those performed at ambient. According to their findings, the polarization and depolarization currents measured under transition were higher than the ambient ones. These authors therefore recommended allowing sufficient time for the transformer temperature to cool down to the ambient before starting measurements. Because at ambient temperature (20 °C) the water migration process takes a long time of 14 days per millimetre (water migration into oil-impregnated pressboard referring to [23]) that would be inappropriate for field measurements.

4 MEASUREMENTS ON OIL PAPER SAMPLES

In order to analyse moisture content, insulation temperature and ageing, independent on spectroscopic measurements, investigations have been performed on oil-paper insulation. The results are reported and discussed in the following section. Because dielectric responses are well-known to depend on the insulation structure/geometry in addition to the characteristics and parameters of the insulating oil and insulating paper at the condition when the measurement is performed, we will not compare magnitude but only shapes with previous data.

4.1 EXPERIMENTAL SETUP

Insulation life is normally determined by measuring the time to breakdown. Doing this in "real time" would have been rather exhausting, given that transformer insulation systems are expected to last several decades before failure occurs. It is therefore appropriate to perform accelerated ageing procedure, whereby the ageing process is accelerated in laboratory tests in order to greatly reduce the lifespan of liquid and/or solid insulation systems. An accelerated vessel ageing procedure is more rapid, less expensive, and provides samples with a controlled thermal history. It is also more amenable to the exploration of the materials and their condition during ageing [26]. Previous investigations in our laboratory have shown that 2000 h thermal accelerated aging worsens the electrical and dielectric properties of oil-paper insulation, than those taken out of a 30 years service-aged transformer [27]. The calendared pressboard samples (15 x 15 cm2 and 1 mm thick) were carefully dried under vacuum (<1 mbar, 48 h at 105 °C) before impregnation. Then, impregnation with degassed and dried commercial grade mineral oil (moisture content < 5 ppm) was performed. Finally, the samples were exposed to ambient air to reach the desired moisture level (as quantified by the moisture in oil), and stored—hermetically sealed—for one week before starting tests. Thermal aging was carried out on some samples at 135 °C in glass ageing vessels containing metallic catalysts (2.5 g/l copper and iron, and 0.5 g/l of zinc and aluminum of cuttings) with air inlet for duration of 500 h and 2000 h. The weight ratio of oil: pressboard: Cu: Fe: Al: Zn during the experiment was 100: 8: 0.3: 0.3: 0.06: 0.06.

Figure 11. Schematic of the test vessel used for the measurements.

Figure 11 shows the test vessel used for the measurements. The vessel is sealed to remain constant moisture equilibrium during the measurement periods. The measuring electrode diameters were respectively 8 and 13 cm. In order to perform the measurements at different temperatures, the test vessel was placed in an oven. In the literature equilibrium curves showing the relationship between moisture in oil and moisture in the solid insulation material can be found [23]. However, these curves are only valid if the moisture distribution inside the oil-paper insulation is in a complete equilibrium condition which is strongly dependent on temperature. As the time constant of moisture migration from oil to solid insulation and vice versa is about 333 h at 20 °C [23], samples were placed in the sealed vessel, at the constant measuring temperature, for at least 2 weeks to reach an equilibrium condition before performing measurements. The moisture content was controlled immediately after each measurement by Karl Fisher titration.

4.2 EFFECT OF TEMPERATURE ON MEASUREMENTS

The PDC measurements performed at three different temperatures are presented in Figure 12. The polarity of the depolarization current values has been changed to positive values to ease representation in the same figure.

1.0E-13

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

1 10 100 1000 10000

Time

Pol_

Dep

olar

izat

ion

Cur

rent

pol 17°C pol 39°C pol 58°C

dpol 17°C dpol 39°C dpol 59°C

S

A

Figure 12. PDC measurements on pressboard specimens at three different temperatures; the water content of the oil sample measured at 17 °C was 10.5 ppm.

Oil Level

HV-Electrode

Guard Measuring Electrode

Pressboard

Isolators


These results show that polarization and depolarization current increases with temperature increase as observed in section 3. Conductivity values for both oil and paper have been calculated from the measured polarisation/depolarisation currents using equation (5) and are presented in Table 4.


Temperature (°C) σoil (fS/m) σpaper (fS/m) 20 13.7 3.6 40 54.74 8.27 60 234.62 58.36

The conductivities of both paper and oil increased as temperature increased. This means that at higher temperatures, the condition of insulation worsens. The polarization and depolarization currents measured on both pressboard specimens at the different temperatures were transformed in frequency domain using equations (14) and (15). Each polarization and depolarization current was fitted with five exponential functions and the constants of each exponential function were calculated by fitting program. The Dielectric Dissipation Factor (DDF) and capacitance (C) of the complex dielectric capacitance are plotted in Figure 13. The DDF increases with temperature increase, and shifts the tan δ curve into high frequency. The capacitance also increases with temperature in the low frequency range.

0.0001

0.001

0.01

0.1

1

10

0.0001 0.001 0.01 0.1 1

Frequency

Tan

Del

ta

Tan T=17°C Tan T=39°C Tan T=58°C

C T=17°C C T=39°C C T=58°C

Hz

1E-6

1E-8

1E-9

1E-10

1E-11

F

Figure 13. Frequency scan of the DDF and C of the pressboard specimens at three different temperatures computed from FT of PDC measurements obtained in Figure 11.

4.3 EFFECT OF AGEING ON MEASUREMENTS

Ageing effects on the dielectric response measurements have been investigated on 500 h and 2000 h aged pressboard samples, under similar condition of oil, temperature, geometry and water content. As shown in Figure 14, the aging of pressboard causes an increase in polarization and depolarization currents. Conductivity values for both oil and paper computed from the measured polarisation/depolarisation currents using equation (5) are presented in Table 5. The conductivities of both paper and oil increased with accelerated ageing duration.

Table 5. Oil and paper conductivity as a function of ageing time.

Ageing duration (h) σoil (fS/m) σpaper (fS/m)

500 234.6 58.4 2000 723.4 78.7

1,0E-12

1,0E-11

1,0E-10

1,0E-09

1,0E-08

1 10 100 1000 10000

Time

Pol_

Dep

olar

izat

ion

Cur

rent

Polarization Current 2000 h

Polarization Current 500 h

Depolarization Current 2000 h

Depolarization Current 500 h

A

S

Figure 14. Polarization and depolarization currents measured on 500 h and 2000 h aged pressboard samples, under the same condition of oil, temperature (20 °C), geometry and water content.

The conductivity of an insulating material is a property, which can be related to the moisture content and different ageing byproducts present in it. Transformers ageing byproducts are mostly polar in nature and will affect conductivity as well as permittivity and the capacitance. Thus, knowledge about the conductivity of the oil and the solid insulation material can be used as an important basis for the assessment of the condition of the oil-paper insulation.

0.001

0.01

0.1

1

10

0.0001 0.001 0.01 0.1 1

Frequency

Tan

Del

ta

Tan 2000 h Tan 500 h

C 2000 h C 500 h

Hz

1E-7

1E-11

1E-9

1E-10

F

Figure 15. Frequency scan of the DDF and C of the pressboard specimens at three different temperatures computed from FT of PDC measurements obtained in Figure 14. The Fourier transformed PDC measurements in figure 13 were calculated using equations (14) and (15); and the DDF and capacitance of the pressboard samples plotted in Figure 15. The DDF as well as C tend to increase with ageing duration. Higher capacitance implies higher permittivity, and hence worse condition of the insulation. Therefore, at the lowest frequencies, different aging cases can be detected.

4.4 EFFECT OF MOISTURE

Figures 16 and 17 show DDF and capacitance values of new and 450 h aged oil-paper insulation samples with different moisture contents. It can be seen from Figure 16 that, moisture affects the DDF frequency scan of oil impregnated paper samples especially at frequencies between 10-2 to 102 Hz, regardless the ageing effect. This is in agreement with the common scheme for FDS reported by CIGRE Task Force 15.01.09 [14].

Cap

acita

nce C

apac

itanc

e


1109

BBBBB

BB

B

B

B

B

B

B

BB

BBBBB

BB

JJJJJJ

JJ

J

J

J

J

J

JJ

JJJJ

JJ

J

1E-3

1E-2

1E-1

1E+0

1E+1

2E+1

1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3

DD

F

Frequancy (F)

B in Virgin oil (20 ppm)

J in Virgin oil (11 ppm)

in Virgin oil with (20 ppm)

in Virgin oil (6 ppm)

Figure 16. Frequency scan of the DDF of pressboard samples in virgin oil at different water contents in oil. Dashed line represents new pressboard samples while solid line depicts 450 hours aged pressboard.

BBBBBBBBBBBBBB

B

B

B

B

B

B

B

B

JJJJJJJJJJJJJJJ

J

J

J

J

J

J

J

6E-11

1E-10

1E-9

1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3

Cap

acita

nce

(F)

Frequency (Hz)

B in Virgin oil (11 ppm)

J in Virgin oil (20 ppm)

in Virgin oil (6 ppm)

in Virgin oil with (20 ppm)

Figure 17. Frequency scan of the capacitance C of pressboard samples in virgin oil at different water contents in oil. Dashed line represents new pressboard samples while solid line depicts 450 hours aged pressboard. Concerning the frequencies scan of the capacitance, moisture influences oil impregnated paper samples at frequency below 1 Hz, irrespective of the paper ageing. Ageing influences the DDF and C of oil impregnated pressboard at all frequencies (from 0.0001 to 1 kHz) but moisture influences are restricted to limited frequency ranges. Therefore, the effect of moisture content and aging can be separated. Table 6 summarises the C-ratio computed from values given in Figure 16.

Table 6. C-ratio computed from values given in Figure 16.

Experimental conditions C-ratio New pressboard in virgin oil, m.c. = 11 ppm 6.9 New pressboard in virgin oil, m.c. = 20 ppm 8.7 450 h aged pressboard in virgin oil, m.c. = 6 ppm 8.1 450 h aged pressboard in virgin oil, m.c. = 20 ppm 9.4

Out of Table 6, it can be seen that, the higher the C ratio, the wetter the insulation is. The influence of ageing is also noticeable, since for given moisture content, the CR increases with ageing.

5 CONCLUSIONS

In this contribution, the influence of temperature, ageing and water content on the dielectric response of distribution transformer as well as on oil-pressboard insulation has been investigated. From the obtained PDC and FDS results, that clearly indicate a strong dependency with these parameters, the following specific conclusions may be drawn:

o From the PDC measurements, it was found that polarisation and depolarisation currents increase with temperature increase. Also, the shape of polarisation current changes as temperature increases.

o Drying of the transformer shows a significant reduction of the polarisation/depolarisation currents.

o For FDS measurements, temperature, ageing and water content caused a higher increase of dissipation factor and capacitance at lower frequencies.

o From FDS (resp. PDC) data, it was shown that the C-ratio (resp. oil and paper dc conductivity) can be used to accurately monitor insulation condition. Both parameters were found to increase with temperature and/or moisture and/or ageing.

o Because FDS measurements techniques at low frequencies are time consuming processes (the duration of PDC measurements and Fourier Transformation of measured data is about half of that of FDS measurements duration), Fourier Transformation of dielectric response from PDC measurements were used with a good accuracy to diagnose temperature and aging effects on transformer insulation system. This alternative would clearly helps saving the expenditure of transformer outage time from service for off-line monitoring and diagnosis.

The strong dependency of the measurements with temperature, clearly suggest that, for on-site measurements, which are generally performed just after de-energising the transformer, whereby the temperature inside the transformer may experience large variations (up to about 20 °C in some cases), the transformer insulation must be in complete equilibrium condition. Indeed, this temperature dependency clearly underscores the requirement of constant insulation temperatures during application for accurateness otherwise extreme cares are needed to interpret the results in the presence of temperature variations and thermal instability within the equipment.

6 REFERENCES [1] T.O. Rouse "Mineral oil in transformers", IEEE Elec. Insul. Mag., Vol.

14, No. 3, pp. 6 – 16, 1998. [2] D. Woodcock Risk-Based Reinvestment - Trends in Upgrading the Aged T&D

System.[Online].Available: http://www.energypulse.net/centers/article/article_display.cfm?a_id=63.

[3] R. Fournié, Les isolants en électrotechnique : Essais, Mécanismes de dégradation, Applications industrielles, Collection de la Direction des Études et Recherches d’électricité de France, Eyrolles, Paris, 1990.

[4] T. K. Saha, "Review of Modern Diagnostic Techniques for Assessing Insulation Condition in Aged Transformers", IEEE Trans. Dielectr. Electr. Insul. Vol. 10, pp. 903-917, 2003.

[5] M. de Nigris, R. Passaglia, R. Berti, L. Bergonzi and R. Maggi, "Application of modern techniques for the condition assessment of

Hz

Hz


power transformers", CIGRE Session 2004, Paris, France, Paper A2-207, 2004.

[6] W.S. Zaengl, "Dielectric Spectroscopy in Time and Frequency Domain for HV Power Equipment, Part I: Theoretical Considerations", IEEE Elec. Insul. Mag., Vol. 19, No. 5, pp. 5-19, 2003.

[7] W.S. Zaengl, "Application of Dielectric Spectroscopy in Time and Frequency Domain for HV Power Equipment", IEEE Elec. Insul. Mag., Vol. 19, No. 6, pp. 9-22, 2003.

[8] T. K. Saha and P. Purkait, "Effects of Temperature on Time-Domain Dielectric Diagnostics of Transformers", Proc. of the 2003 Australasian Universities Power Engineering Conf. (AUPEC), Christchurch, New Zealand, September 28 – October 1st 2003.

[9] A. Seytashmehr, I. Fofana, A. Akbari, H. Borsi and E. Gockenbach, "Effects on temperature on the dielectric response of transformer", Proc. of the 15th Int. Symp. High Voltage Eng. (ISH 2007), paper T8-537, Ljubljana, Slovenia, August 27-31 2007.

[10] CIGRE Task force 15.01.09. "Dielectric Response Methods for Diagnostics of Power Transformers", Electra, Vol. 202, pp. 25-36, 2002.

[11] B. Breitenbauch, A. Küchler, T. Leibfried and W. S. Zängl, "Insulation diagnosis by polarisation and depolarisation current measurements", Proc. of the 13th Int. Symp. High Voltage Eng. (ISH 2003), paper #90, Delft (The Netherlands), August 25-29 2003.

[12] A. B. Supatra, "The Latest On-Site Non-Destructive Technique for Insulation Analysis of Electrical Power Apparatus", Weidmann-ACTI Annual Technical Conf., Sacramento, USA, pp. KEA.1-4, 2004. Available online at: http://www.weidmann-acti.com/u/library/bhumiwatpapernov2004.1.pdf.

[13] A. Shayegani, E. Gockenbach, H. Borsi and H. Mohseni, "The Influence of Aging on Results of Dielectric Spectroscopy on Impregnated Pressboard", Proc. of the 14th Int. Symp. High Voltage Eng., Tsinghua University, Beijing, China, paper H-01, August 25-29 2005.

[14] Cigre Task Force 15.01.09: “Dielectric Response Methods for Diagnostics of Power Transformers”, Electra No. 202, pp. 25 – 33, 2002.

[15] M. Koch and K. Feser, "Reliability and influences on dielectric methods to evaluate the ageing state of oil-paper insulations”, Proc. of the 2nd Inter. Conf. on Advances in Processing, Testing and Application of Dielectric Materials (APTADM), pp. 95-101, September 15-17, Wroclaw (Poland) 2004.

[16] A. K Jonscher, "Dielectric relaxation in solids", J. Phys. D: Appl. Phys. Vol. 32, pp. R57–R70, 1999.

[17] A. K. Jonscher, Universal Relaxation Law, Chelsea Dielectrics Press, London, 1996.

[18] T. K. Saha and P. Purkait, "Investigation of Polarization and Depolarization Current Measurements for the Assessment of Oil-paper Insulation of Aged Transformers", IEEE Trans. Dielectr. Electr. Insul. Vol. 11, pp. 144-154, 2004.

[19] Insulation Diagnostics Spectrometer IDA, Programma Electric AB, Eldarv. 4, SE-187 75 Täby, Sweden.

[20] A. A. Shayegani, E. Gockenbach, H. Borsi and H. Mohseni, “Investigation on the transformation of time domain spectroscopy data to frequency domain data for impregnated pressboard to reduce measurement time”, Electrical Engineering, Springer Verlag, Vol. 89, No 1, pp. 11-20, 2006.

[21] A. Setayeshmehr, C. Eichler, A. Akbari, H. Borsi and E. Gockenbach, "Condition Evaluation of Oil-Pressboard Insulation by Fourier Transform of Time Domain Dielectric Response”, Proc. of Nordic Insulation Symposium (Nord-IS 07), pp. 169-172, Lyngby, Denmark, June 11-13, 2007.

[22] A. Akbari, A. Setayeshmehr, M. Farahani, H. Borsi and E. Gockenbach, "A Software Technique for Transforming Dielectric Data from Time Domain to Frequency Domain for Insulation Diagnosis of Power Transformers", Proc. of the 15th Int. Symp. on High Voltage Eng (ISH), paper T8-482, Ljubljana, Slovenia, August 27-31 2007.

[23] Y. Du, M. Zahn, B.C. Lesieutre and A.V. Mamishev and S.R. Lindgren "Moisture Equilibrium in Transformer paper-oil systems", IEEE Elec. Insul. Mag., Vol. 15, No. 1, pp. 11 – 20, 1999.

[24] H. Borsi, E. Gockenbach and V. Wasserberg, "Life Extension of the Transformer Insulation with an Innovative Online Drying System", Proc. of the 18th Intern. Power System Conference (PSC), Teheran, Iran, Vol. 5, pp. 41 – 48, 2003.

[25] T. K. Saha and P. Purkait, “Investigations on some important parameters of the PDC measurement technique for the insulation condition assessment of power transformers”, Proc. of the 6th Int. Power Engineering Conference (IPEC 2003), pp. 381-386, November 27-29, Singapore, 2003.

[26] J. Scheirs, G. Camino, W. Tumiatti and M. Avidano, "Study of the mechanism of thermal degradation of cellulosic paper insulation in electrical transformer oil", Die Angewandte Makromolekulare Chemie, Vol. 259, Issue 1, pp. 19-25, 1998.

[27] I. Fofana, H. Borsi and E. Gockenbach, "Oil Filled Transformer Retrofilled with ester Liquid – Facts and Arguments", Proc. of the 15th Intern. Symp. High-Voltage Engineering (ISH), paper T8-453, Ljubljana, Slovenia, August 27-31, 2007.

ACKNOWLEDGMENT The authors would like to thank Programma Electronic AB for letting them to use the IDA 200 Insulation Diagnostic System for the frequency domain measurements. Their thanks are also extended to the Alexander von Humboldt foundation for providing financial support to Prof. Fofana during his research stay in Germany (June 15th to August 15th 2006).

BIOGRAPHIES

Alireza Setayeshmehr was born in 1969 in Iran. He received the B.Sc. degree in 1993 from Ferdosi University and the M.Sc. degree in 1996 from the Tarbiat Modarres University, in electrical engineering, both in Iran. Since 1996, he has worked as a member of the academic staff of Chamran University, Ahvaz, Iran. Since 2003, he worked as a Ph.D. student at the Schering Institute of High Voltage Techniques and Engineering at the University of Hannover, Germany. His main research interests are monitoring and diagnostic of high-voltage power transformers.

Issouf Fofana (M´05) received the Electro-mechanical Engineering degree in 1991 from the University of Abidjan (Côte d’Ivoire), the Master and Ph.D. degrees, respectively in 1993 and 1996, from the Ecole Centrale de Lyon, France, where he has been a postdoctoral researcher in 1997. From 1998 to 2000, he was researcher (Fellow of the Alexander von Humboldt Stiftung from November 1997 to August 1999) at the Schering Institute of High Voltage Engineering Techniques (University of Hanover, Germany). He joined the Université du Québec à Chicoutimi (UQAC), Canada, as an

Associate Researcher in 2000. He is currently Associate Professor at the UQAC, Quebec, Canada. Since 2005, Dr Fofana is Chair holder of the Canada Research Chair, tier 2, on Insulating Liquids and mixed dielectrics for Electrotechnology (ISOLIME). He is registered as professional engineer in the province of Quebec, currently appointed to the Technical Committee on Energy and Power Systems of the International Association of Science and Technology for Development (IASTED), member of the IEEE Task Force on Atmospheric Icing performance of line insulators and member of the ASTM Committee D27 on Electrical Insulating Liquids and Gases. He has authored and co-authored over 100 scientific publications and has 3 patents.


1111

Christian Eichler was born in 1971 in Germany. He received the Dipl.-Ing. degree in electrical engineering in 2006 from the University of Hanover/Germany. Since September 2006 he worked as a Ph.D. student at the Schering-Institute of High Voltage Technique and Engineering at the University of Hanover/Germany. His main research interests are investigations about the influence of manufacturing techniques on the electrical properties of synthetic resins as electrical insulating materials.

Alireza Akbari was born in 1973 in Iran. He received the B.Sc. degree in 1997, in computer engineering from Azad University of Najafabad (Esfahan), Iran and the M.Sc. degree in 2004, in distributed computing system from Brunel University (UK). Currently he is a Ph.D. student at the Schering Institute at the University of Hannover, Germany. His main research interests are monitoring and diagnostic software system for high-voltage apparatus, Intelligent and agent based systems and computer applications in power systems.

Hossein Borsi was born in 1946. He received the Dipl.-Ing. degree in electrical engineering in 1972, Dr.-Ing. degree in 1976 and habilitation with “Venia legendi“ for “Hochspannungs-meßtechnik” in 1979 from the university of Hanover/Germany. From 1979 to 1985 he was Professor of Power Engineering at the University of Mashad, Iran, from 1980 to 1982 Dean of the faculty of engineering and from 1981 to 1985 scientific adviser at the ministry of energy in Iran. He is one of the four founding members of the current transformer factory “Reza Transwerke” in Iran and was from 1982

to 1985 its technical director. Since 1986 he is a lecturer of high voltage measuring techniques and Academic Director at the Schering Institute of the University of Hanover. He is a member of VDE, different CIGRE Task Forces and national Working Groups for standardization. He is author and co- author of more than 250 scientific publications and has more than 20 patents in the field of high voltage technology.

Ernst Gockenbach (M´83-SM´88-F’01) received the Diplom degree in 1974 and the Ph.D. degree in 1979 from the Technical University of Darmstadt. From 1979 to 1982 he worked at the High Voltage Test Laboratory of the Switchgear Factory Siemens AG, Berlin, and was responsible for the High Voltage Outdoor Test Field. From 1982 to 1990 he worked with E. Haefely AG in Basel, Switzerland, as chief engineer for high voltage test equipment. Currently he is professor of high voltage engineering and director of the Schering-

Institute of High Voltage Technique and Engineering at the University of Hanover. He is member of VDE and CIGRE, chairman of the CIGRE Study Committee D1 Materials and Emerging technologies, member of CIGRE Working Group D1-33 High Voltage Test and Measuring Technique and member of national and international Working Groups for Standardization of High Voltage Test and Measuring Procedures.

Documents

<![CDATA[Dielectric spectroscopic measurements on transformer oil-paper insulation under controlled laboratory conditions]]>