Presentation 02.1

OMICRON 2013 International Protection Testing Symposium

Impedance Measurement in Madrids Underground Power Grid

Jose Maria Madrid / Ivan Lozano / Jose Manuel Roca, Gas Natural Fenosa, Spain

Abstract

In this paper we want to show our experience in GNF on impedance measurement of underground cables and the influence that the soil of a big city like Madrid has on zero sequence impedance.

This parameter, influenced by the return paths that can be found on the ground and the type of connection of the screens, found in a city like Madrid with unexpected return paths through the soil (pipelines, railways, building structures, etc. ...) does vary significantly from the theoretical values in zero sequence impedance.

For this study we tested all the underground cables that we have in 220kV, 132kV and some 45kV, finding notable differences between the theoretical and measured values.

GNF Grid

Gas Natural Group is one of the leading multinational companies in the gas and electricity sector, operating in 25 countries and with more than 20 million customers.

Following the acquisition of the electricity company, Unin Fenosa, third in the Spanish market, Gas Natural Fenosa has achieved its objective of integrating the gas and electricity business in a single company with extensive experience in the energy sector, capable of competing efficiently in energy markets subject to a process of increasing integration, globalisation and levels of competition.

In Union Fenosa we have over 40 years measuring impedance cables with equipment manufactured for us (see Figure 1), but in 2005, when we acquired an OMICRON CPC 100 we intensified this field work.

Nowadays, all the electric distribution circuits in the city of Madrid which are the property of Gas Natural Fenosa, are underground, with voltage levels of 220, 132, 45 and 15kV. All 220, 132 and 45kV lines are fitted with a distance relay as main or redundant protection. This makes it very important for us to measure the reliability of the impedance data lines in terms of protection settings calculations.

Since 2006 we have been able to measure all the 220kV circuits which are for us the most critical, and any position of 45kV that we believe needs to be measured for its unique features.

Fig. 1 Autotransformer

Fig. 2 CPC 100 in Azca Substation

Based on our experience, we consider the impedance measurement of underground power cables very important because we have found substantial differences between the values indicated by the manufacturer or calculated for us, and those actually measured by us. In addition, it allows us to detect implementation errors in the laying of cables related to the connection of the

Presentation 02.2

OMICRON 2013 International Protection Testing Symposium

sheaths in cable joints, errors in the grounding systems of sheaths or errors in the surge arresters.

Fig. 3 Cable Gallery

In this paper we will show some data in which it can be seen that in some cases the differences are quite significant and could give rise to over-or under-reach.

Fig. 4 Damage detected in a Link box

Cable sheath bonding in GNF

In GNF it is possible to normally find three types of bonding depending mainly on the voltage level, cable characteristic or other design purposes.

1. Single Point or Single End:

In a single end system screens are connected to the ground only at one end of the line, which do not offer a path for current to flow through them.

In the remaining points of the circuit, we will find a voltage between the screen and the ground and also between adjacent screens so that the maximum is the far end. This voltage is dependent on the length of the line and the circuit load. To avoid damaging surges, surge arresters are equipped at the none-grounded end of the line.

Single-point bonded cable installations need a parallel ground conductor, grounded at both ends

of the cable route and installed very close to the cable conductors, to carry the fault current during ground faults and to limit the voltage rise of the sheath during ground faults to an acceptable level. As the main advantage is their simplicity, the screens circulating insignificant current and magnetic fields between conductors are broadly balanced.

It is widely used in lines where it is necessary to maximize the allowed amp capacity in the conductor without the limitations that cause the current screen.

Fig. 5 Single End connection

2. Solid Bonding:

In this case, sheaths are bonded to earth grids at both ends (via link box). This is the common connection on our 45kV grid.

This is generally the most common method and does not require surge arresters, but on the other hand it reduces the transport capacity because of heating effects in the cable screen.

The cable grounded at both ends, makes the use of surge arresters unnecessary and overvoltage will not occur on the screens. It also eliminates the need for the parallel continuity conductor used in single bonding systems.

On the other hand current will flow through them, thus the transport capacity will be limited by the heating effects that occur in the screens by the aforementioned current flow.

Fig. 6 Solid Bonding

Presentation 02.3

OMICRON 2013 International Protection Testing Symposium

3. Cross Bonding:

A system is cross-bonded if the arrangements are such that the circuit provides electrically continuous sheath runs from earthed termination to earthed termination but with the sheaths so sectionalized and cross-connected in order to reduce the sheath circulating currents.

In this type of connection voltage will be induced between screen and earth, but no significant current will flow.

Therefore, examining the total line length and number of joints is required, so that the number of sections in which the line is divided is three or a multiple of three.

Great lengths of line where it is difficult to get the number of sections as a multiple of three uses Cross Bonding combined with one or two end sections with Single Point configuration.

An advantage of this system is that for a conductor arrangement in a triangle, the induced voltage in a steady state in three consecutive sections of screens is zero because it is the sum of three equal voltages outdated 120, because the mutual inductances between the conductors and screens are equal in all three phases. As a result there is no current flow through the screens.

Another advantage is that does not need a conductor parallel ground return, as the screens flows continuously from end to end of the line and are grounded at both ends, so that, the fault current can flow through them.

Moreover, due to the transposition of screens, the voltage induced in the screen during a fault is lower compared with Single End configuration.

The induced voltage on screens is highest in the transposition intermediate joints, and should not exceed 150V under nominal conditions of service and the maximum current for the conductor, taking into account the longest stretch. Induced voltages in a steady state and short circuits are calculated for each project.

In the transposition screen points must be installed a junction box provided with a screen surge arrester.

Fig. 7 Cross Bonding

Obtaining parameters before measurement

Usually, the first step is to receive data from the study of our online database (BDI), which is composed on one side with the data provided by the manufacturer and on the other with a calculation tool, and both sets of data are compared.

For our part we try to measure the impedance of all underground cables that make it possible for their construction work. There are occasions when it is not possible to have a cable access point.

After both parameters have been obtained, both are compared, contrasted and evaluated and the differences examined.

In most cases the calculated parameters are very close to the real ones.

Fig. 8 Theoretical parameters

The impedance calculating is mainly based on two factors:

a. Characteristics of the conductor (diameter of the screens, materials and type of insulation).

b. Physical geometry of the conductors and the ground wire.

Measurements experience

As mentioned before, our experience in cable impedance measuring means we consider this test to be very important, a useful tool to contrast theoretical settings with those obtained after measure.

Observe below some line-measured data, in which we can observe that the differences are slightly bigger in some cases and might lead to an over or under reach in case of faults.

Fig. 9 Comparative

Presentation 02.4

OMICRON 2013 International Protection Testing Symposium

The table shows some results reported, and we can check, for instance, in the circuit connecting the substations Azca-Norte a difference of 29.5% in positive sequence resistance and 70% in zero sequence resistance. Another striking case is the circuit connecting the substations Puente Princesa and Cerro de la Plata, which are measured after a gap of "98%" in the zero sequence resistance.

In previous cases we referred to XLPE insulation cables, but we also have examples of paper-insulated cables, such as circuit Prosperidad Hortaleza 220kV (O.F. technology). In this case, the calculation of their characteristics may seem easy because the layout and grounding of the sheaths is done in a more conventional way (Single End with surge arresters at the far isolated), but we still had a calculation deviation of 51% in positive sequence resistance and 52.8% in case of zero sequence resistance.

Due to the discrepancies found, errors have been detected (connection of screens, execution a Link Box). In the same way we have found damage and problems in the line and its earthing system after faults and shortcircuits in the surrounding grid.

The errors mentioned above have been one of the causes of the discrepancies, but on the other hand due to uncontrollable factors such as ground conditions, the current season, or the influence of other factors external to the line such as pipes, parallel lines, railway infrastructure, and more items that cannot be evaluated because significant changes can occur in a city in a very short period of time that directly influence these parameters and these can only be examined if a field test is performed.

Conclusion

After having carried out the impedance measurement of 220kV cables and a number of 132kV and 45kV cables, a conclusion can be reached.

Due to soil characteristics in a city like Madrid, the impedance of the lines needs to be measured because large differences can be found, especially in the parameter R0, which being dependent on return paths, is strongly influenced by the structural conditions that affect this type of circuit, whether medium voltage lines flowing through the same gallery, water or gas pipelines, railways, steel structures of buildings, the current season or characteristics of the soil. Furthermore, impedance line measurement helps us to check that the connection of the screens has been successful, according to current hypotheses.

In the end, many factors make it very difficult to estimate a theoretical calculation compared with the true scenario.

Fig. 10 Cable gallery

Literature

[1] Alstom Grid: Network Protection & Automation Guide. May 2011

[2] S. Kaiser 2004,Different Representation of the earth Impedance Matching in distance Protection Relays. OMICRON User Meeting 2004.

[3] Roeper, Richard, Short-Circuit Currents in Three-Phase Networks. Siemens Aktiengesellschaft, 1972

[4] William D. Stevenson & John J. Grainger, "Power system Analysis", McGraw-Hill, Jan 1, 1994.

About the Author

Jose Maria Madrid (9 October 1973) is a Systems Engineer from the University of Catalonia. He has performing coordination work commissioning of electrical installations since 2001 for different Spanish utilities having developed his main activity for Union Fenosa, both in generation and in distribution and transportation.