Handout 1

ELEC9713: Industrial and Commercial Power Systems p. 1

ELEC9713 Industrial and Commercial Power Systems

ELECTRICAL SUPPLY SYSTEMS

1. Overview Industrial and commercial power systems represent a microcosm of almost the full gamut of electrical supply systems with, in addition, a significant range of other more specialised applications of electrical, electronic, communications and electrical energy utilization systems. The term “building” will be used in this course to include any industrial and commercial installations with substantial internal electrical distribution infrastructure. It will thus include both commercial high-rise buildings and factory sites with electrical supply at up to 11 kV and with the possibility of both 11 kV and extensive 415 volt distribution systems. In the current state of the art in the supply of electrical energy for such building services, there is an increasing need to make the overall electrical systems in large commercial buildings and in industrial sites - more energy efficient, with better energy management safer in all aspects (including personnel safety, fire and

equipment safety) of adequate power quality with regard to harmonics

and over-voltages


able to accommodate modern information technology systems

be compliant with the new EMC and EMI regulations for electrical systems

provide monitoring systems to assess the condition of the electrical installation.

Because of the potential dangers to personnel, safety precautions are stringently applied and there are a large number of codes and regulations that have to be complied with in the electrical system design and operation. Modern OHS requirements impose a Duty of Care on the infrastructure operator and make safety of paramount importance. The aim of these codes is to give safe operation for both personnel and equipment. A detailed knowledge of these regulations is necessary for the design and operation of building electrical systems. In Australia, the major overall requirement is compliance with the Australian Standard AS/NZS3000, the Electrical Wiring Rules: such compliance is a statutory requirement which is called up in the Regulations of the appropriate electrical supply system legislation of the various Electricity Acts in all of the Australian States and Territories. AS3000 is now based on international standards for such electrical installations. The new Wiring Rules, published in 2007, are based, to a considerable extent, on the IEC (International Electrotechnical Commission) Standard IEC60364, Electrical Installations, which is the basic international standard for low voltage electrical distribution and utilization installations.


In addition, there are many other Australian Codes and Standards which are required to be complied with in building electrical systems and in the various items of electrical equipment used in building services. Many of these are generally applicable to any electrical system, but many are quite specific to building services applications. These documents will be referred to from time to time during the course. 2. Power requirements The type of electrical supply system used in a commercial building or a factory depends primarily on the total power requirements of the various utilization activities that are taking place within the building or factory site. These power demands must be estimated accurately before the details of the electrical supply configuration can be determined and designed. They will determine the ultimate form (and voltage level) of the supply system to and within the building/site. The power requirements are obtained by an estimation of the maximum demand for electrical power. This will normally have three possible components: manufacturing equipment requirements general fixed wiring infrastructure requirements general free-standing mobile equipment supply (GPOs)


An estimate will require detailed loading and duty cycle estimates of each of these. In the first of these, this is easily done as the manufacturing load is easily calculated. It may include large motors, arc furnaces, power electronic equipment, welders, assembly lines, ovens, presses. Some of these loads may require special supplies, such as constant voltage, low harmonic, low noise systems. In the second area, the total of all permanently installed infrastructure equipment such as heating, lighting, air-conditioning, lifts and any similar motor drives etc must be determined. The Wiring Rules has tables which will allow the determination of typical levels of such loads for buildings. In the third component, the estimate takes account of the number of general purpose power outlets (GPOs) in the building. In this last component, the method of determination of maximum demand for domestic and commercial type loads is specified in the Wiring Rules (Australian Standard AS3000, or the equivalent in other countries). The general details of the method of determination are shown in the attached details. (Tables C1 and C2 from AS3000). When the exact load details are unknown, an estimate technique called ‘After Diversity Maximum Demand’ (ADMD) is often used. Based on results from similar installations, typical load density values (VA/m2) are derived for different types of floor area usage. Data for commercial (e.g. offices, shopping centres, hotels, theatres),


light industrial premises and other similar buildings can be obtained from distribution utilities. The following is from EnergyAustralia NS0112 Design Standards (which is now also included in AS3000, see Table C3):

In the estimation of the power demand requirements, it is necessary to make some allowance for future growth in power requirements, including possible additional switchboard circuits, in new buildings or in existing buildings. Typically:

residential premises, 10% fully air-conditioned offices, 15-20% commercial premises, 20-25% shopping centres, light/medium industrial, 25%


This allowance for expansion of power supply capacity must also be included in the initial consideration of the selection of transformer capacity and cable sizes for feeder circuits and reticulation of power. Once the power demand estimation is made and the required incoming supply is specified, the requirements can then be detailed for the incoming cable rating, the main transformer capacity, the substation size and capacity and the switchboard size and capacity. Only when this final design is completed can fault level calculations be performed to determine fuse and switchgear ratings and other protection needs. There are two requirements that must be considered in choosing cable sizes. These are: (a) current carrying capacity, and (b) voltage drop. These are not always consistent with each other’s requirements. For example an adequate current carrying capacity may give too high an impedance and thus too high a voltage drop. 3. Means and Requirements of Electrical Supply The requirements of the electrical supply system may include any or all of the following features: Specified voltage levels Limited harmonic content (quality of supply) Method of supply Safety of supply


Reliability of supply Maintenance Electrical protection Back-up supply (UPS) DC supply

3.1 Voltage Level The supply voltage levels which are available from the electricity supply utilities for use in commercial and industrial locations in urban areas are, typically:

High Voltage: 11 kV, 3-phase supply [by cable or overhead line]

Low Voltage: (a) 240/415 volts, 3-phase, 4-wire system

(b) Single-phase, 240 volts, 2-wire system [by underground cable, aerial cable or by overhead line].

Other supply voltages or configurations may be available in some specific locations, for example in rural areas and for some large industrial installations, but these would be relatively rare exceptions to the above options. Once the power demand and supply requirements are determined by the system designer (usually an electrical consulting engineer), it is necessary to then contact the electricity distributor. They will have a range of additional requirements which must be complied with before they will connect supply from their network to the


site. Note that the electricity provider may be different from the actual electricity distributor, in line with the new regulations which allow contestability among all electrical utilities (and others) in the supply of electricity to consumers. Each electricity distributor is licensed to operate its distribution system over a designated geographical area. In all, there are 3 electricity distributors covering the state of NSW: EnergyAustralia, Integral Energy, and Country Energy. Note that each Australian state sets its own requirements on electricity supply connection. In NSW in particular, these requirements are laid out in the Code of Practice “Service and installation rules of NSW: the electricity industry standard of best practice for customer connection services and installations” (2009 edition), published by the NSW Government, Department of Water and Energy. Also, each distributor may impose further requirements. For example, see “ES1 Customer Connection Information” (2007) by EnergyAustralia.

3.2 Quality of Supply This is now an important consideration in the operation of electrical supply systems in buildings, for two main reasons:

1. The increasing use of power electronics has introduced a higher harmonic level into the supply voltage and this


can have deleterious effects such as the increase of losses in equipment such as transformers and motors with iron cores.

2. Much of the equipment now in use (particularly IT items) is now more susceptible to voltage variation, transient overvoltages and harmonics.

The main features of quality of supply that must be considered in the design of building and industrial sites installations are:

Voltage regulation Frequency of AC supply Voltage waveform distortion interference with communications & control equipment Transient overvoltages

3.3 Method of Supply Up to a power demand level of about 200 kVA in electrical demand, the incoming power supply would commonly be by a low voltage three phase, 415/240V, supply obtained directly from the supply utility’s low voltage (LV) mains, by either overhead or underground connection. Note that LV is any voltage level up to 1000 V ac (or 1500 V dc); HV (high voltage) is any voltage level above 1000 V ac. Up to 3000 kVA, the supply method would be by utility-owned transformer(s) installed in a utility-owned and maintained substation, typically located on the consumer’s premises. The supply to the consumer’s electrical system


(at the point of attachment to the consumer’s terminals) would be at low voltage (415/240V) from the secondary of the transformer. The individual transformers used would be units of 11kV/415V and 750 – 1000 kVA in power rating. They would normally be oil-filled transformers if installed outdoors or possibly dry-type transformers if installed inside a building. Dry-type units are used to reduce fire hazards within buildings. At maximum power demands of greater than 3000 kVA, the supply to the consumer would be at high voltage, most likely 11 kV, with the consumer providing and owning the HV substation and switchgear installation (and handling the maintenance and switching operations associated with the substation). In this case the consumer would be required to employ electrical staff or contractors adequately trained in the maintenance and operation of high voltage equipment. In applying for connection to the utility system, the consumer must notify the supplier of any expected abnormal operating characteristics of operation. This may include, for example, current and voltage surges and non-linear loads which may be significant harmonic generators. The utility supplier will also require details of protection device settings in the consumer’s systems, so that appropriate discrimination with the utility protection operation can be achieved and an adequate earthing system installed.


3.4 Type of Supply Connection The supply system configuration and connection to the consumer may be by means of any of the following systems: (a) High or low voltage from the distribution supply, by

either: Aerial lines (with either bare, covered or

bundled conductors) Underground cables (either 3 single phase or 1

3-phase cable) (b) Low voltage supply from a utility on-site substation, by

means of any of: Low voltage aerial lines (bare, bundled or

covered conductors) Low voltage underground cables Low voltage busbar trunking system (greater

than 2000A per phase)

(c) Dedicated high voltage line from a utility HV substation. This may be by either overhead line or underground cable, depending on the location and the requirements.

[For example, UNSW is a high voltage customer and receives its electrical power at 11kV from two separate connection points to the EnergyAustralia 11kV distribution system. It (UNSW) then operates its own 11kV distribution supply system within the campus]


3.5 Safety Safety regulations require adherence to all relevant Codes and Standards for the applications of the installation. In general, all personnel safety and other general hazards such as fire precautions are covered by the Wiring Rules and its associated documents. [Other general safety aspects include lifts, emergency lighting, fire extinguishing systems etc.] However there are also likely to be other specific hazards which can arise depending on the installation and its loads. These may require isolation of machinery, hazardous areas, equi-potential areas, anti-static locations etc. 3.6 Reliability of Supply The level of reliability depends on the application. Process industry requires high reliability, while commercial operations are less dependent on supply, although computer systems need high reliability. Factors to be considered include:

Supply voltage level (HV supply at 11 kV is more reliable than LV)

Redundancy in circuits Proper protection design (discrimination, etc) Proper maintenance of equipment Choice of equipment.


3.7 Maintenance This is a major issue in current electrical systems. General requirements may include:

Moisture control Ventilation and cooling Corrosion Regular visual inspections Regular testing Regular monitoring and record keeping and analysis

The major issue relates to testing and monitoring and whether to use:

Regular routine testing Testing as required Reliability centered monitoring and testing

Currently the last of these is the favoured method in terms of the optimal compromise between cost and efficiency. 3.8 Back-up supply Many applications require some form of back-up electrical supply, whether it is just for basic systems or for maintenance of full supply. Un-interruptible supply systems (UPS) are becoming more common in building services. They may be via diesel generators or by battery operated power electronic inverter systems.


3.9 DC supply In some situations, mostly in industrial locations, there may be a need for a DC supply for adjustable speed motors, for electrolysis or other purposes. Generally the means of obtaining DC are by modern power electronic converters although older installations may still use rotating machine DC generators driven by AC motors or even mercury arc rectifier systems in very old installations. 4. In-house Distribution System Once the maximum demand and the form of supply are decided on, the internal electrical distribution system must be designed by (usually) the consulting engineers. The circuit supply layout design will be determined by the size of the load, the reliability required, the voltage level and the diversity of the load. Some typical internal supply configurations are shown on later pages. The configurations are generally determined according to requirements of reliability and demand. For example a ring main system may be required for allowing alternative supply to any location, or a primary selective radial system may be preferred. In simple small system a basic radial system is used: this can be increased to an expanded radial system for larger loads and to primary and secondary selective systems for more complex loads.


4.1 Supply system layout options (i) Simple radial system Used for small loads. One primary service and distribution supply transformer supplies all feeders. Simplest possible arrangement with no duplication or redundancy. Cheapest and least reliable option. Reliability is obtained by using quality components. Loss of a cable or the transformer will lose supply. Must be shut down for routine servicing

(ii) Expanded radial system For larger loads than the simple radial system. Same advantages and disadvantages as that system.


(iii) Primary selective system

Allows possibility of alternative supply from two sources on the primary side of transformer(s). Gives improved reliability. This method has two separate primary feeders which can be switched as required. When supply is lost to a load it can be transferred to the other supply source by automatic or manual switching. The sources can be paralleled. Maintenance is now possible without loss of supply. This method has higher cost than the radial systems due to the necessary duplication of components such as switchgear and cables.

(iv) Primary loop system

This system gives greater reliability in the case of failure of the primary cable. Each load can be supplied from either end of the cable in the event of a fault. Finding the cable fault may be difficult in some cases however. Requires several closures to find a fault, so may be dangerous. One section may be energized from either end. Primary selective system is a better option.


(v) Secondary selective system This is achieved when pairs of unit substations are connected through a normally open secondary tie circuit breaker. If the primary feeder or transformer fails, the main secondary feeder CB is opened and the tie CB closes. The general operation of the secondaries is as radial systems. Maintenance is possible. Good reliability. Requires consideration of loading should there be a sustained loss of one circuit.

It is possible to combine this system with a primary selective system to increase reliability if required.


(vi) Secondary spot network

In this configuration, large loads are supplied from one single common secondary busbar which has a number of parallel primary feeders connected to it. Uses special protectors in the form of circuit breakers to each secondary connection. If a primary feeder fails, the protector CB is designed to prevent reverse fault in-feed by opening in such an event. This is the most reliable system for large loads. Expensive.

Used extensively for low voltage, high load density applications such as large commercial buildings. Rarely used in industry however.

(vii) Ring bus system

This configuration will automatically isolate a fault. No interruption of supply for single faults. Note the number of switches and CBs used: the cost is quite high. Allows safe maintenance without loss of general supply.


The design layout will then require a number of specific component parts to make up the system. The components will not all be required in all cases and will depend on the supply voltage and whether the customer needs to have dedicated substations. The following gives the full range of requirements needed for such an installation: 4.2 Substation This will include the following items of equipment, within a general enclosure:

the general enclosure, the transformer(s), HV switchgear, the protection system, backup battery systems, monitoring equipment for energy, voltage, current,

power factor etc


4.3 Switchboards (switchgear assemblies) The switchboard will comprise primarily the output feeder cabling to the various loads (or floors of the building for example) together with their separate switchgear and protection and the main switchgear unit. The switchboard will also have an internal busbar system with rectangular busbars used to interconnect the various feeder circuits. It may be either high or low voltage and will include the following:

switchgear, protection means (relays, fuses, CTs), protection coordination, internal arcing detection, busbar connections to sub-circuits etc.

Switchboards are generally quite specific in design and need to be designed for each application as required. It is only in domestic and similar low power level situations where a standard switchboard design is normally used. 4.4 Cables, Busbars etc The type of cable or busbar choice is very important: a number of factors must be considered in choosing the inter-connecting conductor configurations:

current ratings, insulation ratings fire performance,


segregation of circuits, bundling of cables (effect on thermal rating), magnetic fields and any potential interference effects IP [Ingress Protection] requirements to prevent

contamination ingress.

4.5 Voltage regulation and power factor Voltage regulation and its associated determinant of voltage drop are very important factors in building service design (and in any electrical supply system, for that matter). An excessive voltage drop (leading to lower than rated voltage at the equipment) can cause overheating with some equipment (rotating machines for example) and lamp dimming and flicker in lighting for example. There are strict voltage level requirements in AS3000 and voltage regulation or power factor correction may be necessary to keep voltage drop within limits at maximum load. Industrial/commercial loads require significant reactive power (kVArs), e.g. motors, furnaces, electric discharge lighting. Customers must maintain p.f. not less than 0.9 lagging. There are no restrictions on where to install p.f. correction equipment in the circuit. For more details, refer to Section 6 of NSW SIR (Capacitor installations). The following diagram show the most general form of one-line diagram of a building distribution system. The incoming supply feeds a high voltage distribution switchboard which may supply a number of circuits in a ring main, for example. The switchgear will be normally


contained in withdrawable rackmount units. Each switchgear will then supply an 11000/415V transformer which will then provide supply to a main 415 volt switchboard (or “Switchgear and Controlgear Assembly” to give it its correct title) in the substation. The switchboard comprises a number of “Functional Units” which are essentially moulded case circuit breaker (MCCB) or similar units. Each functional unit will supply a 3-phase cable system to an area switchboard, which will then supply a number of local switchboards each of which may be a switchboard on a floor or wing of the building. This local switchboard will then supply the final sub-circuits to the various power outlets and to any appliances permanently wired to the supply system in the area. From such one-line diagrams, it is necessary to determine a number of characteristics of the electrical system. These include, in particular, the distribution line voltage drop between the point of supply from the utility and the final sub-circuit where the power is consumed. AS3000 states that the voltage drop between the consumer’s terminals and any point of the installation shall not, at maximum demand, exceed 5% of the nominal supply voltage at the consumers terminals. There is also a requirement that the voltage drop in the utility service line should not exceed 3% of the nominal voltage when at maximum demand. This is mainly a requirement applying to the utility and the requirements for their service cable impedance.



5. VOLTAGE DROP An accurate assessment of voltage drop is required for building systems in order to comply with the requirements in AS3000. This requires accurate knowledge of the conductor or cable or busbar resistance and series reactance, as well as any transformer impedances. The power factor of the load and line operation, the load patterns and maximum demand and any other transient conditions such as motor starts are also required information for a full and accurate assessment of voltage drop. Phase voltage and current balance or unbalance is also an important factor to incorporate in the design as even small unbalances in the phase voltage can have serious effects on equipment, such as three phase motor operation, for example. Some indication of voltage effects and phase in-balance are shown on the sheet over. 5.1 Voltage Drop Determinations There are two methods of obtaining voltage drop determinations:

1 By direct calculation using either manual methods or by software packages.


2 By use of the approximate figures in Tables included in AS3000 and its companion Standard, AS3008.1 [Cables for AC voltages up to and including 0.6/1 kV].

Manual Voltage Drop Calculations For the typical length of 50 Hz building system supply circuits, the shunt capacitive reactance of cables etc. is negligible and the voltage drop calculations can be done quite adequately and accurately by use of the short line approximation represented by the equivalent circuit shown below. Vs is supply voltage, Vr is load voltage, Z = R +jX is the series connecting impedance and cos is the supply operating power factor. Note that in the most general case, Z will include any transformer impedance that may be in the circuit. It may also be necessary in some cases to use either transformers or cables in parallel and this can cause some potential problems in design to achieve balance in loading. The requirement is to determine the voltage drop Vs – Vr . In general only the magnitude is required, but in some cases the phase may also be needed The effects of voltage variation and phase voltage unbalance on the operating characteristics of some items of electrical equipment are given below:



5.2 Voltage Drop determinations for supply 5.2.1 By Calculation

R jX

SV SV

I I

0V 0R RV Z R jX

We need to find S RV V

or more generally: S RV V is also acceptable

S RV V IZ RV IR jI X (i) For a lagging load lags by RI V

I

VR

I R

I X

SV

I Z

VR

I R

I Z

sinIR

I X

cosIX

sinIX

cosIR


We have: S RV V IR jI X

But: 0 and oR RV V I I

Thus: cos sinS RV V IR IX

cos sinj IX IR

and:

2 2cos sin cos sinS RV V IR IX IX IR

or just using the real part:

cos sinS RV V IR IX

and thus the voltage drop (regulation):

cos sinS RV V IR IX .

This formula is quite accurate enough for general use and is used for voltage drop calculations extensively. Note that we require the values of R and X which are constant, and I and which can vary. Here denotes absolute magnitude of the angle (unsigned). The regulation is load dependent. In many cases, we know VS but not VR, so then have to use:

R SV V IZ This is not so easy to solve as the equation:

S RV V IZ


(ii) For a leading load leads by RI V

I R

I X

I Z

I I

SV SV

0V 0R RV

Here: cos sinS RV V IR IX

sin cosj IR IX

Again, if we just take the real part:

cos sinS RV V IR IX

[However, it is less accurate now because the imaginary part is higher.] However, in general, we can use:

cos sinS RV V IR IX

for regulation with + for lagging loads and – for leading loads.


Note that the Z R jX can represent a transformer impedance and so it can be used to find regulation of transformers. 5.2.2 By Using Tables


5.3 Voltage Drop Determinations for Transformers

5.3.1 By calculation:

Transformers can be represented by an impedance Z R jX and so we can use the same formula:

cos sinV IR IX where R and X are the total R and total X L referred

to either the primary or secondary side. The transformer nameplate gives the value of Z expressed as a %

% 100

R R

ZZ

E I

where RE and RI are rated voltage and current. Z is

determined by the short-circuit test.

SC RV I Z


As a percent, Z% is the same for the primary and secondary reference: Z R jX

% % %Z R jX R can be measured and X can be calculated from:

2 2X Z R

2 2% % %X Z R Typically, when power >200 kVA, X R :

Z is in the range 3-8% ; R is in the range 1-2% 5.3.2 By use of known data for transformers using impedances


Figure 13

Approximate voltage drop curves for three-phase transformers, 225-10 000kVA, 5-25kV

5.4 Parallel operation of transformers and feeders The operation of feeders and transformers in parallel is not straightforward. Consider two parallel impedances:

1R 1jX

1 1 1Z R jX

2R 2jX

2 2 2Z R jX

1I1S

2I2S 1 2S=S S

V

1 2I=I I

I

The load power (complex S) is supplied with S1 from Z1 and S2 through Z2, with corresponding currents I1 and I2. Obviously:


1 1 2 2 1 2 and I Z I Z I I I

Hence: 2 11 2

1 2 1 2

and Z Z

I I I IZ Z Z Z

The complex powers are:

1 2S S S

* * *

1 2V I V I V I

Thus: * *

* *2 2

1 11 2 1 2

Z ZS V I V I S

Z Z Z Z

Similarly: *

12

1 2

ZS S

Z Z

Also: *

21

1 2

%

% %

ZS S

Z Z

; and S2 similarly

[Note that the % values must be referred to the same base value.] Thus, the supplied power divides according to the ratio:

* *

1 2 2

2 1 1

%

%

S Z Z

S Z Z


Thus, parallel lines or transformers divide loads in inverse proportion to their impedance. Thus, it is important for transformers to be matched in impedance when they are operated in parallel. If not matched, one may be overloaded when supplying a total power that is the numerical sum of their ratings. The transformer ratings must satisfy the requirement as stated in the equation above and they should operate at the same overall power factors to get maximum VA ratings. Ideally, Z1 and Z2 should have the same phase angle (so that

2 1Z Z is a real number). Also, the voltage ratios must be the same. Otherwise, circulating currents will occur. Even a very small voltage difference can cause substantial current that can increase losses and cause saturation of core material.


6. SUBSTATIONS The substation location should be chosen to be at the electrical load centre if possible, so as to keep voltage drop to a minimum and to assist in maintaining voltage regulation as much as possible within the requirements. In very high rise buildings and in extensive factory or similar sites with many buildings, a number of substations may be required to distribute power equitably. Normally these would be connected in a ring main system at high voltage (11 kV). The supply authority will normally determine the type and size of the substation requirements for a specific situation. 6.1 Types of Substations

Pole-mounted substation

The primary component of the substation is an un-enclosed 11,000/415V oil-insulated transformer, rated up to about 500 kVA for providing supply to the building at 415/240V. The substation will also be equipped with associated high voltage fuses of the expulsion or high rupturing capacity (HRC) type, surge arresters or arcing gaps, low voltage HRC fuses and an earthing system for connection of the transformer neutral. Pad-mounted substation

These are metal enclosed kiosk-type cubicles mounted on the ground. The main component is again an oil-filled transformer of similar or possibly higher rating than for the


pole mount type. The kiosk incorporates similar protection and control-gear to the pole-mounted substation. Voltages are also 11,000/415 and the maximum power levels are typically 1000kVA. Outdoor (fenced) enclosures

For higher power levels than 1000kVA, an outdoor fenced enclosure may be used. The transformer may be un-enclosed, with appropriate switches, protection and control-gear within the enclosure. The voltages are typically 11,000/415 V and power ratings are 1500 – 3000 kVA. In some industrial sites the voltages and the power levels may be much higher than the above, which are primarily applicable to building services only. Outdoor (building) enclosure

This consists of a dedicated small building containing transformers and the other usual items of equipment listed above. Voltages are 11,000/415V and power ratings are up to about 5000 kVA. Outdoor transformer with indoor control-gear

Consists of an open outdoor transformer with a small building for other equipment as outlined above. Voltages are 11000/415 V and power ratings are up to about 5000 kVA. Indoor basement or ground-floor substation

As above, but located indoors.


AS/NZS 3000:2007


ELEC9713 Industrical and Commercial Power Systems

SWITCHBOARDS

1. Introduction Depending on the size of the building or factory site and whether the supply is high voltage or low voltage, there may be requirements for both a main high voltage switchboard (SWB) and one or more low voltage SWBs or just a single low voltage SWB. The preferred name for the switchboard unit is a “Switchgear and Controlgear Assembly” (SCA). The basic aim of the SWB or SCA is to take the electrical power from the main supply source and then to feed or distribute power to the appropriate circuits within the building. The SWB has to perform this function in such a way that there is proper control of power flow and proper electrical protection against the damaging effects of faults. This protection is necessary to prevent personnel hazards and also equipment hazards and possible fires. It should be able to operate to isolate a faulty section in the minimum possible time consistent with the fault severity. The SWB should also be designed to present no danger of electric shock or injury to the operating personnel in the vicinity during normal or abnormal operation. Explosions in switchboards are a not infrequent occurrence which can cause significant injury to personnel. In many cases, work


is performed on the switchboard components while they are still live. 2. Component Parts of a Switchboard The major components of a switchboard are:

1 The Incoming Cables

These may be either high voltage (HV) or medium or low voltage (MV or LV). For high voltage, they will normally be either impregnated paper insulation (unlikely these days), cross linked polyethylene (XLPE) or ethylene propylene rubber (EPR) insulated cable. The last two types are the preferred types for new installations, with XLPE being the most common. EPR cables are more flexible and are preferred for specialized applications such as trailing leads in mines. For low voltages the cables may be XLPE or elastomer (EPR) type cable

2 Outgoing circuit conductors

These may be any of the following types:

Insulated cables, Insulated busbars, Busbar trunking systems Mineral insulated metal-sheathed (MIMS) cables Fire-resistant cables


Figure 1

Typical arrangement of switchgear in a switchboard


Figure 2: Switchgear enclosures and housings


Figure 3 Moulded case LV circuit breakers of varying ratings


3 Internal busbars These may be rigid copper (or aluminium) bars (insulated or uninsulated) in large SWBs or simply insulated single phase cables in small SWBs. Bare LV busbars are close together and are subject to high forces on short circuit and resonant force effects must be considered in determining supports.

4 Main isolating switch or section switches These allow segregation of the switchboard or its component parts to allow maintenance work on the SWB.

5 Circuit breakers These are HV or LV depending on the switchboard voltage level. For HV units the circuit breaker types used are oil, SF6 and vacuum units, contained in withdrawable rack-mounted carriers. Oil CBs are no longer used in new installations but are still very prominent. (In older switchboards there may also still be some high voltage air-break units with insulating splitter plates, but these are very rarely used now). For LV and MV (less than 1000 V) units, the circuit breakers are invariably of the air-break type using the


“de-ion” principle, with isolated metal splitter grids. Large MV CB units may be also rack-mounted but the modern SWB will have moulded-case circuit breakers (MCCBs) for the higher current ratings (more than about 100 Amps) and miniature circuit breakers (MCBs) for the lower rating levels (less than 100 Amps). MCBs would normally be used in the smaller sub-main and local SWBs in a building.

6 HRC fuses and CFS units These are also used in MV and LV switchboards for high level fault protection and, in many cases, there are combinations of HRC (high rupturing capacity) fuses and overload switches with limited interrupting capacity used (combined fuse-switch or CFS units) because of their economy. See Figure 1 for more detail of typical usage of switchgear and Figure 2 for some examples of rackmounted switchgear. Figure 3 shows examples of MCCBs.

7 Protection relays These are used for the higher voltages, together with their associated instrument transformers (current transformers (CTs) and voltage transformers (VTs)). Overcurrent protection units are used to activate timing relays so as to provide proper fault protection operation. At lower voltages, the circuit breakers will


normally have in-built fault detection sensing and thus no separate relaying is required.

8 Metering equipment The metering of a SWB will include: line and phase voltage, line current in each phase, total power, power factor metering.

9 Over-voltage surge protection Modern switchboards will also have some over-voltage surge protection designed into both the HV and LV sides to protect equipment against the effects of any over-voltage transients that may be generated within the system or conducted in from external sources.

3. Requirements Switchboards are usually quite specific to their particular requirements in a building and thus they tend to have a one-off or unique design, with little scope for standard design of large switchboards such as will be found in large buildings. However, in large building systems, the smaller sub-circuit distribution boards, located on each floor level for example, may be of a standard design. Similarly, domestic switchboards are standard designs and are uniform across Australia.


It is fair to say that SWBs (particularly the smaller low voltage type) are not often designed by technically expert persons and, together with their unique one-off design, there is thus a need to perform detailed testing to prove a particular design. Even expert designers may have problems with some operational features (particularly thermal dissipation and temperature rise) where the overall operation may not be what would be expected from the characteristics of the individual components which make up the switchboard. For example, the thermal interactions of the component parts may limit the thermal ratings of components within the SWB to levels below their normal (isolated) ratings. De-rating factors may be required to be applied. This is particularly a problem for cable bundles entering a SWB. Similarly, the complex magnetic fields in a SWB may cause some variation in the calculated forces on busbars in the SWB or may cause some unexpected eddy current heating of any metal (particularly steel plate) that may be in the vicinity. The requirement for such extensive testing of switchboards means that the customer must be very specific in his required specifications when giving these to the SWB designer and constructor. The customer should also specify clearly what tests should be performed to prove the SWB operation and this should be agreed with the builder. In many cases these may be destructive tests and thus it will be necessary to count on multiple numbers for construction.


Switchgear and busbar requirements In general, the requirements for switchgear in new switchboard installations in buildings are:

A life of about 25-30 years at least A substantial (20-40%) spare capacity on new

installation to allow for expansion. Good quality and reliable switchgear in the various

outgoing functional units. Proper protection design, particularly in the area of

time discrimination with flexible variation of I-t characteristics possible..

Adequate interrupting capacity for future expansion Residual current (earth leakage) protection Adequate current carrying capacity Protection against ingress of contamination (dust,

moisture etc) Adequate compartmentalization to limit arc faults

The purchaser should specify, at the least, the following requirements for switchboards and switchgear: Voltage, power, current ratings. Specific rating for each circuit breaker and busbar

system The required fault level and the corresponding

protection operating time. Internal structure and segregation of compartments (if

required)


Ingress Protection (IP) numbers for protection against dust and moisture

Arc containment requirements Earthing requirements Electrodynamic forces and insulator mechanical

strength requirements. Thermal features-maximum temperature rises etc. Testing requirements (Type tests and Routine tests).

Australian Standards for design of SWBs The Australian Standard AS3439.1-2002 (Low Voltage Switchgear and Controlgear Assemblies – Part 1: Type-tested and partially type-tested assemblies) is the document which gives specific requirements for LV SWBs. There is a similar, though more diverse document, AS2067-1984 (Switchgear Assemblies and Ancillary Equipment for Alternating Voltages above 1kV). AS2067 also covers outdoor substations and specifies required clearances for bare HV conductors. Both documents also provide detailed guidelines for access prevention by un-authorised persons. Internal Segregation of circuits in the SWB With a number of separate circuits within the switchboard and with the knowledge that SWBs, with a multiplicity of internal components, are more susceptible than most items


to faults, the question of whether to segregate chambers or parts within the whole structure is an important feature of design. Segregation of chambers by metal walls will assist in containing faults and prevent them from spreading to involve other sections of the board. The major problem with SWBs particularly at low voltage is the arcing fault. The fault current in such SWBs can be very high and the arc that results will be a very high energy entity that can cause very significant damage by virtue of its high temperature, thermal radiation field and convective heat transfer. It can cause very significant damage to the board and to personnel. The problem is exacerbated by the fact that the arc impedance is significant and can reduce the current level in the fault and this can affect (slow) the response time of the overcurrent protection. High impedance arc faults are a major problem to the protection design engineer. Arc faults can be detected by various means such as optical sensors, pressure sensors, sensitive earth leakage protection etc. However in many cases there will be some requirement by purchasers to limit the effects of arc faults by detailed design of the SWB, involving segregating the various internal sections in some way to limit the spread of any fault arc within the board structure. The standard AS3439.1 defines four different forms of switchboard compartment segregation. These are designated as: Forms 1, 2a & 2b, 3a & 3b, 4a & 4b.


Note that Form 1 has no internal segregation of compartments within the SWB. The exact details of the design differences are shown in Figure 4. Arc Containment Internal arcing in switchboards is usually caused by some dielectric insulation failure within the SWB structure, caused for example by insulation ageing, by moisture, by solid particle contamination or even dropped tools while personnel are working live on the switchboard. Segregation of the internal parts can provide some limitation of the spread of the damage caused by arcing. Such damage can be very destructive. Figure 5 is an extract from AS3439.1 giving some details of how arc faults may be generated and how they can be prevented and contained. Residual current or earth leakage protection may be necessary to detect high impedance arcing faults. Ingress Protection (IP numbers) In common with many forms of electrical equipment, switchboards have to be protected against ingress of various contaminants (such as particles, dust and moisture) and there must also be some means of preventing access of personnel to live internal parts. The specific options and requirements are given by the use of IP numbers, such as


IP23, where the two numerals represent specific design requirements to prevent ingress. The first numeral relates to dust and particulate matter (and also to prevention of direct contact by personnel), while the second numeral relates to ingress of moisture. Thus IP00 would provide no protection whatsoever (completely open), while IP68 would be, effectively, a hermetically sealed enclosure. Because switchboards are normally located indoors and in locked and ventilated rooms with restricted access, the IP numbers are not particularly stringent in commercial building systems. IP21 may be a typical level of protection in a building SWB. However, in industrial manufacturing building SWBs, or for outdoor SWBs, the ingress protection level may need to be something like IP65. Figure 6 gives the specific design requirements for compliance with each IP numeral. More details are given in AS60529-2004 (Degrees of protection provided by enclosures)


Figure 4 Switchboard compartment forms of segregation

Form 1: no internal separation

Form 2a Form 2b terminals not separated from busbars terminals separated from busbars

Form 2: separation of busbars from functional units


Form 3a Form 3b terminals not separated from busbars terminals separated from busbars

Form 3: separation of busbars from functional units + separation of functional units from one another + separation of terminals from functional units

Form 4a Form 4b terminals in same compartment terminals not in same compartment as associated functional units as associated functional units

Form 4: separation of busbars from functional units and terminals + separation of functional units from one another + separation of terminals associated with a functional unit from those of another

functional unit


Figure 5 Requirements for arcing fault containment in SWB

enclosures


Figure 5 (cont.)


Figure 6 IP number classification system

AS 60529-2004


4. Design features In the design of switchboards, there are three major areas which must be addressed; 4.1 The insulation design This includes the 50Hz power frequency and BIL (Basic Insulation Level or lightning impulse voltage withstand level) insulation levels and also the appropriate creepage and clearance path design requirements. Because of the potential for accumulation of contaminants such as dust and moisture, the creepage distances over insulation surfaces are very important factors. Surface tracking (creepage) is a major hazard in SWBs. Figure 7 gives details of the insulation requirements of low voltage SWBs, as specified in AS3439.1. 4.2 The thermal design This is a very important consideration in the design and is complicated by the difficulties present in the theoretical calculation of temperature rise in such complex structures. This arises because of the enclosed nature of the SWB and the interactive heating effects between the many different components. Much experience and general “rules of thumb” and empirical procedures apply, but the only sure way of proving a thermal design is by performing temperature rise tests on prototype SWB designs.


Figure 8 gives typical temperature rise limits in SWB components. 4.3 The protection against electric shock This is particularly important in the open-type switchboards and it is necessary to have some means of protection against direct contact (with live parts) or indirect contact (with exposed conductive parts) during maintenance procedures on SWBs. Figure 9 details the requirements for attaining adequate protection against shock. An essential feature of most methods of protection against indirect contact is proper earthing of the switchboard. Fig 10 shows the various earthing schemes which have been designated by the International Electrotechnical Commission (IEC). For protection against direct contact, the IP number system outlined above can give adequate design requirements. 4.4 Testing of Switchboards In addition to the above design features, the testing of SWBs is of paramount importance in proving the design. Fig. 11 shows the various tests which need to be performed before a SWB should be accepted. There are both Type Tests (done only on one unit representative of the design)


and Routine Tests (done on every manufactured unit) listed. The full list of tests can be very expensive to perform as they can be very lengthy in their set up and instrumentation and test performance times (e.g. thermal tests may take many hours to achieve thermal equilibrium) and it may be necessary to sacrifice a SWB when carrying out short circuit and arcing tests. There are few test laboratories available with a full range of adequate facilities in Australia. The primary one is the Testing and Certification Australia (TCA) high current testing station at Lane Cove, Sydney. There is also another smaller and more limited facility at TestSafe Australia (associated with WorkCover NSW) at Londonderry, Sydney.


Figure 7 Insulation requirements for switchboard components and

structures

Dielectric Test Voltages

(a) for the main circuit

(b) for auxiliary circuits



Minimum creepage distances for different rated insulation voltages, pollution levels, and material groups are given in Table 16 of AS3439.1:2002.


Figure 8 Temperature rise limits for switchboards

(AS3439.1-2002)


Figure 9 Requirements for protection against electric shock from

switchboards

1. Protection against direct contact By insulation of live parts: which can only be removed

by destruction or by use of a tool. Insulation not inferior to cable insulation, i.e. 3.5kV withstand, PVC cable hardness, suitable for maximum temperature of 105oC.

By barriers or enclosures: protection against direct contact of at least IP2X.

2. Protection against indirect contact By using protective circuits By other measures: such as electrical separation of

circuits or total insulation


Figure 10: Earthing systems for switchboards

TN systems: one point directly earthed, exposed conductive parts connected to that point by protective conductor (PE). Three types:

TN-S system: separate neutral (N) and PE throughout TN-C system: N and PE combined into a single

conductor throughout. TN-C-S system: N and PE combined into a single

conductor in a part of the system.

TN-S

TN-C-S


TN-C TT system: one point directly earthed, exposed conductive parts connected to earth via separate earth electrode.

IT system: no direct connection between live parts and earth, exposed conductive parts connected to earth.


Figure 11 Details of required tests on switchboards

TTA = type-tested assemblies PTTA = partially type-tested assemblies



CABLES AND BUSBAR SYSTEMS

1. Introduction Electrical power distribution in buildings and industrial sites is achieved primarily by use of cables or busbars. Cables are used for the full range of current levels at low (400/230V) and high voltage (3.3, 6.6 and 11kV). Busbars are mostly used for the very high range of current carrying capacity, usually at medium/low voltage (400/230V). In addition to simple electrical energy supply, cables are also extensively used for control and communications functions in building and industrial sites. The choice of cable voltage for purely electrical energy supply depends on the size of the building or industrial site. Most mains installations will have cable systems operating at 400/230V and busbar systems will also operate at 400/230V. For extensive sites or for large loads the cables may be high voltage, at 11kV in modern sites or at 6.6 or 3.3kV in older sites (mainly industrial). Other types of cable may be used in some more sensitive areas: for example mineral insulated metal sheathed cable (MIMS - Pyrotenax) where fire resistance is required, or simply fire resistant cable which uses mica tape sheaths over the conductor and under the main insulation.


2. Cable structure The cables used may be either multicore or single core in structure (see Figures 1 and 2). At low voltage, single core cables consist of a single conductor (usually stranded copper or aluminium) with a main insulation layer of any of PVC (polyvinyl chloride), XLPE (cross linked polyethylene) or EPR (ethylene propylene rubber). In addition, there is usually an outer sheath of PVC or HDPE (high density polyethylene) or HEPR (hard ethylene propylene rubber) to provide some mechanical protection. As mentioned above or sites where a high level of fire resistance is required, copper clad mineral insulated metal sheathed (MIMS or Pyrotenax ) cable may be used or other fire resistant cables of the mica-glass taped type. At high voltage the cables may be oil or mass impregnated paper insulated if the installation is relatively old (more than about 20 years), but will be XLPE or EPR in modem installations, except in high voltage DC applications where paper insulation is still best. PVC is not used at voltages higher than 400V because of its high loss factor (DDF) which generates too much dielectric heat loss above 400V. XLPE and EPR, which do not soften as their temperature increases are examples of thermosetting synthetic insulating materials. They are vulcanized to improve their temperature withstand ability. XLPE is a very good high voltage insulant in that it has very low DDF losses and a high withstand voltage.


PVC is an example of the thermoplastic class of synthetic insulation. These materials soften as their temperature increases but regain their original form when they cool. Paper insulation is a natural cellulosic fibre which is operated in cables with the paper impregnated by a hydrocarbon oil or grease type compound. Although the paper in oil composite is a very good insulator, it has some drawbacks in that the cable joints and terminations are much more difficult to make and they must have a metal outer sheath to contain the oil. Multicore cables at low voltage usually have four conductors with three phases and a neutral. They may be either stranded copper or aluminium or, more often, solid section aluminium. The conductors are often sector shaped to give a better packing density in the overall structure. At high voltage the cables will be paper insulated if older style or XLPE or EPR in modern cables. EPR insulation is much more flexible than XLPE and is used more often in locations requiring high cable flexibility, such as for trailing cables in underground mines and similar applications. Figures 1 and 2 show some examples of typical cable designs and configurations.


Fig.1: single-core, multi-core.


Figure 2: Examples of LV and HV cables


3. Requirements of cable system designs There are some general installation requirements for power cable systems, in addition to the primary selection considerations of rating and voltage. These secondary installation requirements include: Separation of power and control/communication cabling

systems to prevent interference

Protection of cables from physical damage

Use of additional cable ducts to allow for future expansion

Care in the preparation of joints to prevent high contact resistance or low insulation levels

The major considerations to be applied in the selection of cables or busbar systems are: The current carrying capacity [determined by the

maximum permissible steady state temperature rise].

The voltage drop and regulation of the cable/busbar circuit at full load

The short circuit rating [determined by the maximum permissible transient temperature rise]

The insulation requirements and associated factors [jointing and termination].

The required level of fire resistance of the cable and busbar systems.


3.1 Current carrying capacity This is an important criterion in selection. The temperature of the insulation of the cable or busbar system must be kept below well-defined values to limit the ageing of the cable insulation. For this reason an accurate determination of the current capacity is necessary to ensure that the temperature will remain within allowable limits and thus not cause accelerated ageing. In general, a 10oC increase in the operating temperature would halve the insulation life. Figure 3 show the limiting temperatures for various types of insulated cables.

Fig. 3: Limiting temperatures for insulated cables

(Table 1 AS3008.1.1-1998)


This rating determination must be done for all cables and conductors. For some circuits there is no need to perform detailed calculations as there are specific requirements laid down by regulations for specified cable sizes in various situations. For example, the minimum rating for consumers mains cables is 32A and for sub-mains cables it is 25A. However, above this minimum level the ratings must be calculated with reasonable accuracy. The calculation is complicated by a number of variable parameters which may have a significant effect on the capacity. These parameters include the ambient temperature, the enclosure of the cable (if any), bundling of multiple cables etc. It is usual, and sufficiently accurate in most cases, to use the tables which are given in the Australian Standards AS3000 (The Wiring Rules) and in standard AS3008.1, an accompanying standard to AS3000, to determine the current rating capacity. Tables 3-21 of AS3008.1 give ratings for a variety of cable types and enclosures that cover most usual applications. Figures 4-9 are some typical examples. In the calculations the maximum load demand figures must be used as the basis for cable size determination and some allowance should be made for future expansion, de-rating from mutual heating of bundled cables, any effects of harmonics and frequency of current. Special attention must also be given to the current carrying capacity of the neutral conductor. This is often of smaller cross-section than phase conductors and this can sometimes lead to problems.


Fig.4: Current-carrying capacity (Table 3 AS3008.1)







3.2 External influences on cable capacity The current-carrying capacity of a cable can be affected by external influences such as:

Grouping of cables Ambient temperature Depth of laying Thermal resistivity of soil Varying load conditions Effect of thermal insulation Effect of direct sunlight

Thus capacity values given in Tables 3-21 of AS3008.1 should be corrected by applying an appropriate rating factor. A common situation is the grouping of cables in close proximity such that they are not independently cooled by the ambient air or ground. The appropriate derating factors are given in Tables 22 to 26 of the Standard.

Fig.10: Cable grouping derating (Table 22 AS3008.1)


Fig.11: Cable grouping derating (Table 23 AS3008.1)

The current-carrying capacity specified in Tables 3 to 21 of AS3008.1 is based on the ambient air temperature of 40oC and soil temperature of 25oC. If the ambient temperature varies from this standard, the cable capacity needs to be


adjusted using the rating factor given in Table 27 of the Standard.

Fig.12: Rating factors for variation in ambient temperature

(Table 27 AS3008.1.1-1998) For different operating current or ambient temperature, the conductor temperature can be calculated using:

2

o o A

R R A

I

I

where: Io = operating current IR = rated current given in Tables 3-21 o = operating temperature of cable R = rated operating temperature of cable A = ambient air or soil temperature


3.3 Voltage drop and regulation The Wiring Rules have quite specific requirements for minimum voltage drop levels in consumers' circuits. They allow for a maximum of only 5% voltage drop between the consumers terminals (where supply is taken from the utility) and the load end of the longest sub-circuit in the building. Thus, if the supply is at the nominal value of 230V, the minimum permissible voltage anywhere in the consumer's circuit is 218.5V. This is roughly the general lower limit at which most 240V equipment will still operate satisfactorily. There is also a general upper limit of voltage rise of 5% above the supply level. This may occur if some highly capacitive loads are in use. 3.3.1 Allocation of voltage drop in consumer's circuits For a building distribution system, the typical allocation of voltage drop would be as follows:

Consumers mains wiring connections: 0.5 - 1 % Consumers sub-mains wiring connections: 0.75 -1.5% Consumers final sub-circuit connections: 2 - 3 %

The designer must choose and design for allocations from the bands such as to give a total of no more than 5% voltage drop at maximum demand loading. The actual voltage drop (Vd) can be calculated from:

1000

cd

L I VV

volts


where: L = cable length (m) I = current carried by cable (A) Vc = unit value of voltage drop (mV/A.m) For a single-phase, two-wire supply system, if using same conductor type for active and neutral:

1

2

1000c

d

L I VV

For a balanced three-phase supply system, no current flows in the neutral. The voltage drop per phase to neutral is voltage drop in one conductor and the voltage drop between phases is therefore:

3

3

1000c

d

L I VV

Thus: 1 voltage drop = 1.155 x 3 voltage drop 3 voltage drop = 0.866 x 1 voltage drop Tables 40-50 of AS3008.1 provide 3 voltage drop values for various cable configurations, types, conductor size and operating temperatures. If we know cable impedance (Tables 30-39 of AS3008.1) and the load power factor, the voltage drop can be estimated accurately from:

Single-phase: 1 2 cos sind c cV IL R X

Three-phase: 3 3 cos sind c cV IL R X


Note that the voltage drop will depend on the following factors:

power factor, current level, cable or conductor resistance, inductive reactance, length.

Fig.13: three-phase voltage drop (Table 41 AS3008.1)


Fig.14: three-phase voltage drop (Table 42 AS3008.1)

For unbalanced 3 circuits, voltage drop calculations can be performed on a 1 basis by geometrically summing the voltage drop in the heaviest loaded phase and the voltage drop in the neutral conductor. Another approach would be to assume balanced 3 load conditions and perform calculations using the current flowing in the heaviest loaded phase.


Fig.15: three-phase voltage drop (Table 47 AS3008.1) 3.3.2 Power factor correction In some cases of low factor loads (less than about 0.9) it may be necessary to install power factor correction equipment to limit current and hence voltage drop. It will also decrease the cost of the supply utility charge for power in most cases.


As most low power factors are lagging (inductive) loads, power factor correction is achieved by installation of compensating power capacitors used to generate leading reactive power. The capacitors may be either series connected or, most commonly in buildings, shunt connected. Installation of power factor correction equipment must be done carefully to avoid the potential effects of harmonic currents which may be amplified by the decrease in net impedance caused by the reactive power reduction in the circuit. It may be necessary to install filters to avoid such effects. 3.4 Short Circuit Temperature Rise Depending on the insulation material in use with the cables or busbar systems, the permissible temperatures of cables and busbars are different and thus different cables have different continuous ratings and also different short circuit ratings. The maximum allowable short circuit temperatures of insulation of cables are quite different to those used for determining steady or cyclic ratings. The permissible temperatures are higher for short circuits on the basis that the duration of the short circuit is typically only about 1 - 3 seconds before protection operates. For this period of time there will be little deterioration of the cable material. The choice of maximum short circuit temperature is more likely to be based on softening or mechanical effects rather than pure insulation damage.


The maximum permissible short-circuit temperatures (for duration up to 5 seconds) are detailed in Tables 52-54 of AS3008.1.

Fig.16: Maximum permissible short-circuit temperatures.


The short circuit heating formula is derived from the more general equation which covers adiabatic heating of the cable during the short duration of the transient short circuit. This equation incorporates the temperatures of the ambient medium and the insulation temperature of the cable at the instant that the short circuit occurs. It also includes constants which incorporate properties of the conductor materials. The equation that is used is given below and simply provides a maximum 2I t value relating to the cable type and temperature conditions and limitations. It includes the non-linear effect of resistance increase of the conductor as the temperature rises:

2 2 2I t K S where: I = short-circuit current (rms over duration), in amps t = duration of short circuit, in secs S = cross-section area of conductor, in mm2

20 1 01 20 ln 1 1K c

c = specific heat of the conductor material 20 = electrical resistivity of conductor at 20oC

= conductor material density 0 , 1 = initial and final temperatures

= temperature coefficient of resistance Values for the constant K can also be readily obtained from Table 51 of AS3008.1.


Fig.17: Table 51 AS3008.1.1-1998

The maximum phase conductor temperature on short circuit is often taken as 120oC. However it may also be that other conductors are also limiting factors. Thus, copper wire or tape screens have a limiting temperature of 350oC, lead sheath of 200oC and aluminium sheath of 200oC. And in some cases these may be the limiting factor if the return fault current passes through the sheath, say, which may have a higher resistance that the main conductor and thus generate more heating. For some multicore cables, it may be electrodynamic forces between the phase conductors rather than thermal constraints that determine the short current rating of such a cable. In such cases, a current limitation of the fault level by HRC fuse rather than an I2t limitation is required.


3.5 Insulation Effects The particular type and thickness of insulation of cables and conductors will depend on the voltage of operation and also on the application. For modern high voltage distribution systems, XLPE is almost exclusively used. For low voltage applications, there are many insulation types in use. These include:

PVC XLPE Elastomeric (ethylene propylene rubber [EPR]) Mineral insulated (MIMS)

Other factors that need to be considered in the insulation choice are flexibility, hardness, resistance to mechanical effects, effects of moisture and contamination etc. A major factor in the choice of insulation used is the effects of fire on the cable insulation. 3.6 Fire behaviour There are many problems generated by the potential damage to cables which are involved in fires. The fires may be either self-generated, by the cable, or externally- generated but causing significant damage to the cable insulation. The problems that arise are:


1. The cables may provide substantial flammable material in the chemical structure of their insulation

2. They generate significant smoke and soot when they burn and this can cause considerable damage to equipment otherwise unaffected by the fire heat.

3. Many of the fire products from insulation combustion are toxic and thus represent a significant health hazard.

4. Many products of the insulation combustion are corrosive and can substantially damage electronic equipment for example.

5. The cables should ideally be able to operate after significant damage by a fire: however this is not possible if the polymeric insulation burns away: there may be total loss of insulation integrity in this case.

The last problem is a particularly important consideration in buildings where power may be required during and after a fire to operate emergency equipment. Thus there is a requirement for fire-resistant cables in buildings and many circuits within a building, particularly high rise buildings, will be designated to have such cabling installed. Item 4 above is an increasingly important aspect because of the large penetration of computers and general electronic and communications equipment in modem commercial buildings. The major problem in this situation is with PVC which generates very large amounts of hydrogen chloride gas when it burns and this can cause havoc with metal in


electrical and electronic circuits, particularly when moisture resulting from fire hose and extinguisher use is present. The HCl is also very injurious to health. The response to this problem has been the development of a range of cables able to withstand the effects of fires or able to maintain insulation integrity after a fire. These will be outlined in the next section. In some cases, PVC has been banned for use in buildings and similar areas for general distribution. 4 Cable Types in Use The cable types in use in the overall building supply context are: 4.1 Aerial cables or lines These are used between buildings in site complexes or as service lines to buildings. The bare lines with no insulation covering use aluminium or copper or (less likely for buildings) ACSR (Aluminium conductor with steel reinforcement). The insulated overhead cables (which may be either aerial bundled cables (ABC) or simply covered conductors) are installed as either an aerial service line type of application or in some case are mounted on the side fascia of buildings. The insulation may be PVC or some other material which has been treated to be able to withstand the effects of UV


radiation. UV radiation from sunlight causes degradation of insulation and is the primary problem with polymeric insulation when used outdoors. ABC can be high voltage (up to 11kV) but the simple covered conductors are always low voltage cables. 4.2 Indoor distribution cables For main circuits these may be either single core or 3-core or 4-core. The insulation may be paper (unlikely in new installations) or XLPE, PVC, elastomer (EPR) etc. at low and medium voltage. The multi-core conductor types may be stranded cores, sector cores, with or without neutral or armour. Typically they will have PVC or PE outer sheathing for protection. 4.3 Fire Resistant cables Fire resistant cables are required in specific situations where emergency supply needs to be maintained even in the event of severe fire interaction with the cable. Such cables need to be able to supply power up to two hours after being engulfed in conflagrations. Some common types of such fire resistant cables are described below: 4.3.1 MIMS

Mineral insulated metal sheathed. Single or multi core. The insulant is magnesium oxide powder. There is an outer copper sheath. The powder is not as good an electrical


insulant as the normal cable polymeric materials but is able to withstand extreme fire temperatures (700 – 800oC) without any damage. Sealing of the metal sheath must be done carefully to prevent moisture ingress as this will destroy the insulation integrity of the powder. 4.3.2 Radox

This has a glass-mica tape layer wound over the insulation. The insulation is a polyolefin material and is halogen-free to prevent the production of chemically active and toxic halogen based by-products from the fire. 4.3.3. Firestop

This has a mica glass tape wound over the conductor. The insulation is XLPE. Firestop cable also uses halogen-free insulation. Fire resistant cables must be tested to resist the effects of fires. This is particularly the case with the last two types where there is substantial loss of polymer in the fire, leaving only the mica glass tape as the insulant. The test involves exposure to fire at about 1000oC for three hours with the cores live. No short circuits can occur. This is followed by removal from the oven and subjecting the burnt cables to a water spray test to simulate the effect of fire extinguishers. Again, no short circuit can occur if the cable is to pass the test. 4.4 Other cable types


There are many other cables in use in buildings, primarily of the communications type. These will be discussed at a later lecture. As stated earlier it is necessary to segregate the communications/control cables from power cables to prevent EM coupling effects which may lead to incorrect information transfer. 4.5 Magnetic interference from cables A recurring and increasing problem with power cables is the effect of the magnetic field generated by the power frequency current in such cables. There are two major areas of concerns in buildings: interference with IT equipment (e.g. computer terminals) and the less tangible one of the potential hazard to personnel. The end result is that some care must be exercised in location and arrangement of cables to minimise these effects. This aspect will be covered later in the course. 5 Cable Enclosures and Conduits These include the following types of enclosure systems: (they should all be halogen-free, if made of insulating material, for fire damage minimisation) Conduits: steel, rigid PVC, corrugated flexible PVC Tubes and pipes: metal, non-metal plastic, earthenware

etc Ducts: metal or non-metal


Cable support systems: cable trays, cable ladders troughs etc

Trunking systems: mainly for busbars. In the case of metal enclosures or support systems, the potential for eddy current heating must be taken into account. Insulation of sections may be necessary to prevent this effect. Also the enclosures may need to be earthed to prevent problems of electric shock by indirect contact if a fault occurs by insulation failure to the metal enclosure. The conduits and other enclosures must have appropriate IP numbers to prevent ingress of moisture and dust to susceptible areas. Some care must also be given to a determination of the impact of the enclosure on the thermal dissipation from the cables and the effect that this will have on the cable rating. It will often be necessary to de-rate the cable current carrying capacity because of thermal limitations caused by enclosures. 6 Busbars and busbar trunking systems (BTS) etc 6.1 Busbars For very high current capacity systems it is common to use rigid busbar systems of either aluminium or copper rectangular section [See Figure 13]. These are normally segmented (laminated) for very high ratings. This 'lamination' achieves two purposes: one is to improve heat


loss by providing greater surface area for thermal dissipation and the other is to limit increase of resistance due to skin effect. Both skin effect and proximity effect are significant problems that must be considered for high current AC busbar systems. The segmentation or lamination of the individual phase conductors limits the eddy current generation level which is the basis of the skin and proximity effects in AC conductors. DC systems are not affected in this way and thus only require such segmentation to assist with improving thermal dissipation. Note that the orientation of the busbar sections can have a very significant effect on the thermal dissipation by virtue of the impact on natural convection flows set up in the ambient air and this will affect the current carrying capacity of the bars. Similarly, a matt (non-shiny) surface with a high radiative emissivity will also enhance the current rating substantially through increasing the radiative heat loss. Busbars are prone to potential resonance effects caused by the electrodynamic forces between the conductors. Because of the very small spacing at low voltages and the high currents, the forces on busbars even at normal operating conditions can be significant. With the busbar support being normally at spaced intervals of perhaps a metre or two, the busbar acts as a fixed beam under the lateral electrodynamic force and it will thus have a natural resonant frequency, the value of which will depend on the spacing of the supports, the size of the conductor and the elasticity coefficient of the material.


Fig.18: Current-carrying capacity of busbars.


It is a necessary requirement to calculate this natural frequency of resonance and if it is close to the 100 Hz frequency of the electrodynamic forces generated, then the support spacing must be altered to change the resonant frequency to move it further away from the electrodynamic force frequency. The bars may be insulated or uninsulated, although the modern trend in large buildings is for insulated bus bars, often in a trunking system (busbar trunking system[BTS]). 6.2 Busbar Trunking systems The trunking system consists of a tightly packaged sandwich of the three busbars and the neutral, with an insulating foil (melinex typically) between layers. It may require external fins for cooling because of the densely packed nature of the sandwich. They will also require some protection against ingress of moisture as this can be taken by capillary action to the live conductor and create a creepage path. Trunking systems are now being used with structures which are able to be joined with relative ease to connect plug-in systems. Feeder BTSs can have ratings from a few hundred amps to 10,000 amps. Plug-in systems can have ratings from about 100 amps up to 1500 amps. The outer metal enclosure must be earthed and the same precautions for eddy currents must be taken as in metal bus


bar enclosures. They must be type tested and routinely tested before installation. Tests must look at, in particular, the following aspects:

Temperature rise at rated load Short circuit temperature rise Clearances Creepage paths IP ratings Mechanical strength etc.

6.3 Flexible copper straps In addition to the rigid trunking systems, flexible insulated copper straps are now being used. These have the advantage of being able to be easily shaped on site to match the installation requirements. They can be used in small switchboards for example where space is limited.


Example 1:

Assume: 3 voltage drop in consumer mains: Vd = 3V balanced loads, i.e. disregard current in neutral Subcircuits wired with multi-core V-75 insulated and sheathed copper conductors, installed in single circuit configuration, unenclosed in air, clipped to a wall. Choose conductor size to satisfy voltage drop requirement. Three-phase circuit: 30A load. From Table 12 column 4, conductor size of 6mm2 has current-carrying capacity of 37A.

Total maximum permissible voltage drop: 5% 400 = 20V

Hence, voltage drop allowed in 3 final subcircuit: 20 3 17V

Maximum unit value of voltage drop:

1000 1000 17

6.3mV/A.m90 30

dc

VV

L I


Select Table 42 for multicore cables with copper conductors. Select 75oC column for normal operating temperature of V75 cables. The nearest lower unit value is 3.86mV/A.m. From column 1, this corresponds to cable size of 10mm2. Single-phase circuit: 30A load. From Table 9 column 4, conductor size of 4mm2 has current-carrying capacity of 34A.

Total maximum permissible voltage drop: 5% 230 = 11.5V

Hence, voltage drop allowed in final subcircuit:

11.5 3 3 9.77V

Maximum unit value of voltage drop:

1000 1000 9.77

3.62mV/A.m90 30

dc

VV

L I

Convert to three-phase value: 3 1 0.866 3.62 0.866 3.14mV/A.mc cV V

Select Table 42 for multicore cables with copper conductors. Select 75oC column for normal operating temperature of V75 cables. The nearest lower unit value is 2.43mV/A.m. From column 1, this corresponds to cable size of 16mm2. Note: in both final subcircuits, cable size had to be increased to comply with voltage drop requirements.


Example 2: Underground 1500A 3 supply is constructed using parallel circuits of 400mm2 V-75 single-core insulated and sheathed Cu cables. Determine the minimum number of active conductors required. (a) All cables in one conduit

Capacity of one 400mm2 cable= 510A (Table 6, col. 16) Refer to Table 22 for derating factor:

510 5 0.6 1530A

(b) Groups of conduits

Refer to Table 26(2) for derating factor:

510 4 0.79 1612A




Example 3: 3- circuit is to supply a load of 125A per phase. Use two V-75 four-core Cu conductor, insulated and sheathed cables bunched together on a surface in a confined ceiling space where ambient temperature is 50oC. Determine minimum conductor size and maximum route length if allowable voltage drop is 3%. Derating factor for cable bunching = 0.8 (Table 22, col.5) Derating factor for 50oC ambient = 0.82 (Table 27.1, col.9) Required minimum current-carrying capacity:

1 1

125 190.5A0.8 0.82

for two parallel cables

or 95.25A per cable. From Table 12, column 4, the conductor minimum size is 35mm2. 3- permissible voltage drop: 400 0.03 12VdV Unit voltage drop (Table 42): 1.11 mV/A.mcV Maximum route length:

1000 1000 12

173m62.5 1.11

d

c

VL

I V


Example 4: 1- circuit composed of two 16mm2 Cu single-core sheathed cable, V-75 insulation, installed unenclosed on a wall. Circuit is to supply 55A resistive load. Determine 1- voltage drop value when ambient air temperature is 40oC and 25oC. Operating current 55AoI

From Table 1, rated operating temperature o75 CR From Table 3, current-carrying capacity is 72ARI . (a) Ambient temperature o40 CA

2

o o A

R R A

I

I

operating temperature o60.4 Co

Use o60 Co then from Table 41:

3 voltage drop = 2.32 mV/A.m thus 1 voltage drop 1.155 2.32 2.68 mV/A.m (b) Ambient temperature o25 CA From Table 27, correction factor = 1.21

2

2555

72 1.21 75 25o

o44.9 Co

Use o45 Co then from Table 41:

3 voltage drop = 2.20 mV/A.m thus 1 voltage drop 1.155 2.20 2.54 mV/A.m

ELEC9713: Distribution Transformers p. 1


DISTRIBUTION TRANSFORMERS

for use in industry and commercial buildings

PART 1: General Overview

The transformers used in industry and in commercial buildings are generally less than about 1500 kVA in rating, although some may be up to 2500 kVA. However size and space limitations keep them to typically the 1000 kVA level. For interior use in buildings they are all naturally cooled apart from, in some cases, some rudimentary additional fan-cooling systems which may be installed some time after installation as an adjunct, perhaps to try to increase the maximum short–time rating of older transformers. Table 1 gives details of the range of distribution substation transformer ratings as used by Energy Australia. Note that the substations listed are not necessarily restricted to indoor types. Because of the potential danger from fire in enclosed spaces, such as substations in buildings, the type of insulation of transformers in such installations is often somewhat different to that used in outdoor applications. Such outdoor transformers use oil-impregnated paper as the basic dielectric.


Transformer type Current rating (A/phase @415V)

Power rating (3-) in kVA

Pole mount Pad mount Outdoor enclosure 1x1000kVA unit 2x1000kVA units Building substation 1x1000kVA unit 2x1000kVA units 2x1500kVA unit 3x1500kVA units

140-700A (5-25 @11kV)

1000A

(40A @11kV)

1500A 2000A

(60-75A@11kV)

1400 2000 3000 5500

(50-200A@11kV)

100-500 kVA

750 kVA

1000 kVA 2000 kVA

1000 kVA 2000 kVA 3000 kVA 4500 kVA

Table 1: Range of standard substation transformers.

In the standard oil-impregnated paper transformer (Figure 1) the primary insulation is Kraft paper wound around the transformer winding conductors. This paper is then impregnated with a liquid dielectric to exclude any air bubbles and to provide good thermal circulation for heat dissipation. For outdoor applications the main liquid insulation in transformers is mineral oil, with the windings being encased in a tank of such oil. The problem is that such oil is highly flammable and, if used in a building in an enclosed location, could represent a serious fire hazard. If


mineral oil insulated transformers are used in buildings then the room must be fitted with automatic fire extinguishers and there must be a bund structure at the base of the transformer to contain all of the oil should it be discharged from the transformer. Moisture is a major problem with oil transformers and they need to have precautions against ingress of moisture, such as silica gel dryers. Liquid insulated transformers are generally limited to the Class A materials temperature rise limits of about 60-70oC. Higher temperature rises can cause more rapid deterioration of the cellulose (paper) insulation on the windings and also deterioration of the oil itself. The lack of forced cooling and the relatively low insulation temperature limits put some constraints on the thermal ratings of such transformers for use in buildings. Because of the potential for deterioration of the transformer insulation it needs to be tested at regular intervals to determine the insulation efficacy and whether any deterioration has occurred. This is generally done in mineral oil transformers by dissolved gas analysis to monitor both oil and paper insulation and by dielectric dissipation factor testing. These diagnostics will be covered in more detail later in the course. The expense of fire precautions and the potential risk of fire in buildings has generated a current trend to use what are effectively non-flammable liquid transformers. There have been two basic types of non-flammable liquid insulated transformers developed:


the liquid insulated type, using synthetic oil and the dry insulation type using solid resin insulation to

replace the oil. 1 Synthetic liquid insulation materials 1.1 Askarel The initial move away from mineral oil (in the 1930s) was to the use of askarel liquid insulation. Askarel, which is essentially a polychlorinated biphenyl (PCB), is an artificial insulating oil which is almost non-flammable. It was used for many years in a large number of electrical applications, but it has some toxic effects, particularly if heated or burnt. It is now banned from use in most countries. It is quite likely, however, that there are still some PCB insulated transformers in service. 1.2 Silicone Oil Silicone insulating fluid, which is tetrachloro-benzyl toluene with about 40% trichlorobenzene, is essentially non-flammable and has no toxicity problems. It is the most favoured synthetic transformer insulating oil. It has a higher viscosity than mineral oil and askarel, and so its convective heat dissipation coefficient is not so good but its other electrical properties are very similar to those of mineral oil, although it is quite a bit more expensive.


Figure 1: Oil insulated distribution transformer ONAN cooled type (Oil Natural Air Natural)

2 Dry Type insulation Because of the cost of silicone oil and the need to provide expensive bund structures to hold the oil in case of a leak, there has been a very substantial move to dry-type transformers for use in buildings in recent years. These have no fluid impregnation, as the name implies. Some also have a much higher temperature rating, being typically about class C (150oC). Such high temperature rating transformers may use higher temperature-withstand synthetic-paper insulation such as Nomex, rather than


natural cellulose paper, on the winding conductors. This type of material is self-extinguishing if subjected to a flame from a fire. There are two different generic forms of dry-type transformers:

the open winding type and the epoxy cast-resin type.

2.1 Open Winding type This type is the true dry-type transformer and uses the simple structure of the (paper-insulated or nomex-insulated) winding in open air and in construction simply has many layers of insulating varnish coating applied to the windings. (see Fig 2b). There is a potential problem with moisture absorption and penetration into the varnish and insulation if these transformers are left un-energized for long periods of time. The paper on the copper windings will absorb moisture readily if the varnish layer is not absolutely impermeable. The moisture ingress will increase dielectric losses in the insulation (dielectric dissipation factor) and will also reduce the insulation strength. They are form-wound windings as can be seen from the structure. 2.2 Cast-resin type These have the windings in a cast solid epoxy resin structure as shown in Fig. 2a. They are much less susceptible to moisture ingress and absorption. The


application of the casting must be done very carefully however to ensure that the expansion coefficients of both the resin and the metal windings and core are the same. Any differential expansion or contraction will cause cracking of the casting. The windings are often sheet layers on the LV side. They are more costly than the open structure dry-type transformer and are often more expensive than silicone oil transformers.

(a) (b) Fig.2: Dry-type transformers (a) cast resin (b) open winding 2.3 Gas insulated transformers The other form of non-flammable transformers that are being used increasingly in buildings and in high-density areas are SF6 insulated transformers (Figure 3). They are very expensive but very reliable. SF6 is a non-toxic gas with


very good electrical insulation properties and with good thermal transfer properties. SF6 transformers typically operate at about 2 atmospheres of gas pressure where the dielectric strength is similar to oil. The greenhouse problems of SF6 may eventually force the pure SF6 dielectric currently used to be replaced by an SF6-N2 mixture. The dielectric strength of a mixture of 20%SF6 with 80%N2 is about 80% that of pure SF6.

Fig.3: SF6 gas insulated transformer

3 Comparison of characteristics of different transformer types

3.1 Cost


Fig 4a shows a cost comparison of the various transformer types. The SF6 and the cast resin are the most expensive and the mineral oil is the least expensive type. The cast resin type also has higher losses because of its more difficult thermal dissipation problems with thermal conduction being the only means possible.

Figure 4a: Comparison of costs of transformer types

In its most general application the cost of the transformer must include capital cost of installation and the cost of total losses amortised over the predicted life of the transformer. This aspect will be discussed later. 3.2 Losses Fig.4b shows a comparison of total losses (including the variable copper and the fixed iron losses) in cast-resin and silicone-insulated transformers at various loadings. At a high load factor (80%) there is essentially no difference in


total loss: at 50% loading the silicone oil transformer has lower losses (because of its inherently lower no-load losses). [Load factor is the ratio of average load to full rated load of the transformer]. The generation, measurement and determination of losses will be covered later in the course.

Figure 4b: Comparative losses of cast-resin dry type and silicone-oil transformers

Figure 4c: Comparison of losses of different transformers types [For 1000kVA, 11kV/415V]


Fig.4c shows similar comparisons with the other types of transformers. The liquid insulated units are seen to be generally better than the dry-type units particularly at high load factors. 3.3 Reduction of Insulation Life During operation, the loading of the transformers is of particular importance as it determines the total losses and these, in turn, determine the operating temperature of, and hence any deterioration of, the transformer insulation. If the insulation temperature rises to too high a level for its class, then the transformer lifetime may be reduced. The increased temperature causes increased chemical reactions in the insulation and these lead to the deterioration by changing the insulation composition. As a general rule of thumb the 10oC rule is often used, whereby an increase of continuous operating temperature by 10oC causes a reduction of insulation life time by about 50%. The loss of life versus temperature of operation details are given in typical loading guides for transformers which are published as Standards in most countries. In Australia they are in the Australian Standards AS 2374.7-1997 (oil-filled transformers) and AS 3953-1996 (Dry-type transformers). See Fig. 5 for some typical data that can be used to determine insulation deterioration (or lifetime reduction) if loading details and temperature rises are known.


Figure 5: Rate of deterioration of transformer insulation with change in operating temperature.

[The overall average operating temperature can be determined from a mean load factor K, which is derived from cyclic loading details of the transformers.] 3.4 Transformer Impedance & Short Circuits As the transformer impedance will be a major part of the impedance of any short circuit path, the effect of the transformer impedance on prospective fault current in a power system is very substantial and thus accurate transformer impedance data is needed to allow such calculations to be performed to determine protection


requirements. In some cases where accuracy is not paramount, general impedance data such as that shown in Fig 6 can be used for fault current determinations involving transformers. As can be seen, an average value of 5% or .05 per unit for transformer impedance is a good and often used approximation. Usually, the leakage inductance component is the major contribution to the impedance, particularly for high voltage transformers with their larger winding spacings. For precise fault calculations the impedance must be known accurately from data given on the nameplate and the resistance and inductance components must be known.

(a) (b) Figure 6: (a) Typical impedance and corresponding typical fault levels for various ratings (b) Effect of impedance on

fault levels of 11kV/415V transformers

3.5 Tappings on windings


Distribution transformers used in buildings do not normally have on-load tap changing (OLTC) facilities to adjust voltage level. However they do have permanent taps which can be altered to allow about a +/-10% variation in voltage output level, usually in about 1% steps. The taps must be manually changed while the transformer is de-energised and isolated. The tapping points are normally on the high voltage windings only in normal transformers. Figure 7 shows a distribution transformer with on-load tap changing facilities. The tap changer is shown at left of the three winding structures.

Figure 7: Transformer with on-load tap changing

[rating about 5000kVA]


3.6 Connections In general, building distribution transformers are star-connected on the low voltage side to eliminate circulating triplen harmonics (3rd, 9th, 15th etc.) in a delta-connected three-phase winding. The high voltage side is almost always delta connected. There are many possible variations of winding connections of transformers. These variations can affect the magnitude of the voltages but more particularly they change the phase shift between the primary and secondary windings. There are about 20 different connections possible if the use of zig-zag earths on transformers with delta windings is included, but the most common winding connections for standard distribution transformers are:

DY11 DY1 DY5 DY7

DY11 is the most commonly used connection for distribution transformers: it gives a 30o phase shift between primary and secondary. See Fig. 8 for the corresponding vector diagrams and an explanation of phase shift symbols and phase shift values.


Figure 8: Different connections of windings in transformers

and associated vector diagrams.


3.7 Cable terminations Cable terminations in transformers are generally achieved by means of a cable box on the transformer at the high voltage and low voltage sides for both types of transformer with paper insulated cable. Figure 9 shows examples of medium voltage (11 kV) terminations. Note the use of skirts in some cases to increase creepage path lengths.

Figure 9: Examples of medium voltage cable terminations

The LV cable box is usually air insulated, but the high voltage cable box is compound-filled with petroleum grease or some similar viscous insulant. Good sealing of the box is necessary to keep the moisture out. The modern preference for use of XLPE cable use makes terminations in the cable box a little simpler in that moulded heat shrink terminations are able to be used: these can be relatively easily applied for both high and low voltages. Paper insulated joints and terminations at high voltage are much more difficult to


produce and require considerable expertise to achieve good results (that is a joint without insulation problems). 3.8 Parallel Operation of transformers When transformers are used in parallel it is necessary to ensure that they satisfy the following requirements:

Have the same voltage ratios The same tapping points in use (that is the same voltage) Have the same vector diagram (same phase shift between

primary and secondary) The same impedance angle (this is preferable but not

imperative). If these conditions are not designed for, problems will occur. For example:

(1) There will be unequal loading of the transformers if the impedance angles vary. This can lead to overloading of one transformer and a lighter loading of the other. The same principles that apply to the sharing of load in parallel-connected feeders also apply to transformers.

(2) If the voltage ratio or the tapping points are not the same there will be circulating currents set up in the two transformers which will lead to possible overheating of the transformers and possible change in operation points on the magnetization curves.

(3) If the vector diagrams are different then the line and phase voltages will be intermixed and insulation stress will be stressed.


DISTRIBUTION TRANSFORMERS

PART 2 Operational Characteristics and Efficiency

1 Construction The basic power transformer comprises paper-insulated copper windings wound around a laminated magnetic steel core. For 50 – 60 Hz operation the core is laminated grain-oriented silicon-steel (or more rarely an amorphous magnetic metal core to reduce eddy current loss). For high frequency operation, where eddy current losses in the core become too high, even if it is laminated or amorphous, transformers may use a ferrite core which has very high electrical resistance and will thus significantly reduce eddy currents. For instrument transformers (current and voltage transformers [CTs and VTs]) where the core losses contribute to the measurement uncertainty, high permeability, low loss, materials such as Mu-metal are used in the core to give higher accuracy in metering applications. Usually the two windings (primary and secondary) are wound on the same limb of the core to reduce leakage flux. The high voltage winding is normally the outer winding, as shown in Figure 1, because of the higher electric field associated with it. Note that there are two standard core configuration forms in general use. These are:

The Core Form The Shell Form


Figures 1 and 2 show the two typical core arrangements for both single and three phase units. Note the different magnetic circuits which result from the two configurations of cores. The Core Form is the most prevalent type in use in Australia and in most other countries. The shell form was manufactured primarily in the USA

Figure 1: Core and winding structure of a single phase

transformer: (a) Core type, (b) Shell type

Figure 2: Core construction of 3-phase transformer

(a) Core type (b) Shell type or 5 limb core The windings are normally composed of paper-insulated copper strip or wire. The insulation on the copper is either lapped paper (with oil impregnation in oil-filled transformers) or enamel or possibly Nomex tape in the case of dry-type transformers.


The windings of main power transformers are all form-wound as opposed to being random-wound structures. There are two general winding configurations in use: Concentric layer winding: these windings are helical

layers wound axially along the core and the HV and LV windings are laid concentrically.

Sandwich windings: these windings are of the pancake or

disc type with radial rather than axial extension on the core. The HV and LV pancake winding discs are then sandwiched together as shown in Figure 3.

Figure 3: Types of transformer winding

(a) Concentric, (b) Sandwich – made up of disc sections. 2 Equivalent Circuits of Transformers Figure 4a shows an ideal transformer with perfect flux coupling (no leakage outside the magnetic circuit) between the primary and secondary windings. The equivalent circuit for this is shown in Figure 4b. Only the winding resistance is needed in the equivalent circuit. The transformer (the


mutual coupling section) is taken as ideal, with no losses and no saturation effects. Thus there is no magnetisation current needed and no distortion. With perfect coupling and no magnetisation current the mutual inductance of the windings shown will be infinite as the assumptions will, in effect, make the core material of infinite permeability.

Fig. 4a Ideal transformer Fig. 4b Full coupling of flux Equivalent Circuit However, in a practical transformer, the coupling of flux is not perfect and there will be some flux leakage, as shown in Figure 5a. There will be leakage at both HV and LV windings and this leakage flux will appear as a leakage self-inductance (mostly in air) in each winding. This represents the main internal impedance in a high voltage transformer. The equivalent circuit must now include this leakage inductance, as shown in Figure 5b.


Figure 5a: Leakage Flux

Primary Flux 1P , Secondary Flux 1S , Common flux 1P and 1S are leakage fluxes, typically 1 1-6% of

Figure 5b: Equivalent circuit with leakage inductance

For high voltage transformers the inductive reactances Xp and Xs are generally much higher than the winding resistances Rp and Rs and the leakage reactance thus represents the main short circuit impedance of the transformer. The total impedance is generally about Z = 3 – 8 % (the impedance voltage). Note that the percentage is based on the rated or base impedance of the transformer, ZB:


RatedB

Rated

VZ

I

where Vrated and Irated are the nameplate voltage and current for the transformer. Note that for 5%Z ,

0.05 0.05 RB

R

VZ Z

I

and for a terminal short circuit 200.05

R RF R

B

V VI I

Z Z

or 20 per unit. 3 Excitation Requirements (a) Magnetising current Im Because the core is not infinitely permeable it requires some finite level of ampere turns (magnetic potential - N1Im) in the windings to establish the flux in the core, even with no load connected. Because the core is not of zero reluctance, there will be some finite inductance of the core and the winding used to magnetise the core. This magnetising inductance (Lm) will be defined by:

m mL I where is the core flux and Im is the magnetising current.


Note that by using the magnetising inductance in the equivalent circuit we are taking the mutual coupling section of the transformer to be an ideal transformer, as below.

(b) Core power loss and Ic Core losses are quite significant in large power transformers and they must be included in the equivalent circuit representation. (In instrument transformers these core losses will be the major source of error in the measurement of current and voltage). The loss is included by insertion of a notional core loss resistance Rc connected in shunt as shown in Figure 6a. The value of the notional resistance is determined by use of the equivalent core loss current Ic, such that c p cI V R

2 2

c c c p cP I R V R

where Vp is the primary voltage and Pc is the core loss.

Ideal transformer


Thus: 2

Pc

c

VR

P

(c) Total exciting Current Io This is Io = Ic + jIm Thus the full equivalent circuit is as shown in Figure 6a.

Figure 6a: Full equivalent circuit

We can exclude the ideal transformer part of the equivalent circuit by rescaling the voltages and impedances on the secondary side using the usual turns ratio squared (a2) calculation. Note that a is that ratio of primary turns to secondary turns. We then have Figure 6b below where all quantities are referred to the primary side, or we can have Figure 6c below where all quantities are referred to the secondary side.


Figure 6b: Equivalent circuit referred to the primary

Figure 6c: Equivalent circuit referred to the secondary

Normally the primary and secondary quantities are all lumped together to give the most general equivalent circuit as shown in Figure 6d below where: 2

p sR R a R ; 2p sX X a X and Z R jX

cR mjXcI mI

'2V

0IR jX 1I '2I

1V

Figure 6d: Lumped equivalent circuit


Note that Im will lag V1 by 90o and that Ic will be in phase with V1 (barring small differences caused by R and X). The magnetic flux will be in phase with Im and will lag V1 by 90o. Thus, the phasor diagram of the final equivalent circuit is as shown below (using V2 as the phase reference).

1I

'2V

1I R

1I X

1V

1I Z

'2I

IC

Im0I

0I

Note that when the secondary is short-circuited the impedance Z determines the fault current. This is the normal method of determination of the transformer impedance Z (the Short Circuit Test). The secondary is shorted and the primary volts are raised until rated current I2 flows in the secondary circuit. Then:

21 testV Z I

Because I2 is the rated current value, [1 per unit or 100% in percentage terms],

1 test%Z V


The transformer nameplate will give a percentage value for Z, which is termed the “Impedance Voltage” rather than the impedance because of the above relationship. Note that

2 2Z R X

and X is thus able to be determined from:

2 2X Z R

R is the total winding resistance and is able to be measured with a resistance meter. Note however that R will be very temperature dependent and thus it should be measured at or near normal operating temperature of the transformer. The difference between R at ambient temperature and operating temperature may be as much as 30%. Problem Example: Consider a 4000/400V 10kVA transformer which has the following characteristics:

Primary winding resistance: 13pR

Secondary winding resistance: 0.15sR Total leakage reactance referred to primary: 45pX

Magnetising reactance referred to primary: 6kmX Core loss resistance referred to primary: 12kcR


Determine:

(i) total R and Z for transformer, referred to the primary. (ii) R and Z referred to the secondary and also all

impedances referred to the secondary. (iii) the input current when:

a) the secondary is open circuit b) 2 25AI at 0.8 lagging PF

(iv) Im, Ic, Io and the total core loss and the winding (load) loss in case (b).

(v) the total transformer power loss at full load and the full load efficiency at unity PF.

Solution:

Ratio 1

2

400010

400

Na

N

Thus 2

1eq 13 0.15 10 28R

1eq 28 45 Z j

2eq 1eq 2

1 280.28

100R R

a

2eq

145 0.45

100X

m2

600060

100X

c2

12000120

100R


4000

0.33A12000cI (in phase with V1)

4000

0.67A6000mI (lag V1 by 90o)

Hence: 0.33 0.67 0.745 63.5 Ao

oI j When 2 225A, 25 36.9oI I Thus, referred to primary: 21 2.5 36.9 2.0 1.5 AoI j Hence: 1 21 oI I I

2.0 1.5 0.33 0.67j j

2.33 2.167 3.18 43 Aoj Core loss 2 212000 0.33 1333 Wc cR I Copper loss 2 2

1eq 1 28 3.18 283 W (at load)R I

Total loss = 1616 W

Efficiency 10000

86.1%10000 1616

4 Transformer losses


Losses in transformers are composed of two separate components:

a) Load (copper) loss in the resistance of the windings b) Core (iron) loss in the core material (Hysteresis and

Eddy current loss) The copper loss (I2R) is load-dependent and scales as the square of load current IL. The loss will also be temperature-dependent through the resistance variation with temperature of the winding. The core loss is constant whenever the transformer is energised. Core loss is thus independent of the load. [There is also another loss component, which is caused by eddy current loss in the steel tank and in any other metal which is coupled by the AC magnetic field of the transformer. However this loss component is usually neglected.] Total copper loss is thus: (neglecting stray losses) 2 2

Cu 1 1 2 2P I R I R

21 1eq wattsI R

Core loss is: (i) Hysteresis loss n

h mH k f B W/m3

where f is frequency;


Bm is peak flux density; kh is a constant of the material and of the core configuration; The exponent n is material dependent and is in the range 1.5 < n < 2.5 (often use n = 2).

(ii) Eddy current loss 2 2

e mE k f B W/m3

where ke is a constant of the material and configuration. It is also temperature dependent as it includes the material resistivity.

Total Core Loss:

totalP H E

2h e mk k f fB (for n =2) W/m3

Note the effect of frequency: there is a significant effect even between 50 and 60 Hz and this can be an important difference. 50Hz 60Hz Hysteresis 1.0 1.2 20% increase Eddy Current 1.0 1.44 44% increase 5 Transformer Efficiency 5.1 Power efficiency


When supplying a load, the transformer power efficiency is given by:

Efficiency =

Power Out Power Out

Power In Power Out Losses

= 1 -

Losses

Power Input

For a load with voltage V2, current I2 and power factor cos,

2 22

2 2 2 2eq core

V I cos

V I cos I R P

Transformers are very efficient items of power equipment, with efficiencies normally in the range of 95 – 99%, but the efficiency is obviously dependent on the load and on the load power factor. It can be shown easily that maximum power efficiency of the transformer occurs when the load is such that

2core 2 2eqP I R

That is, when, core loss = load loss For typical transformers at full load,

load loss ≈ 3 – 4 times the core loss Thus, the load for maximum power efficiency is about 50 – 60% of rated load. [the transformer in the previous problem


example was not a typical transformer as the maximum efficiency occurred at 2.15 per unit load!]

5.2 Transformer Energy Efficiency is the instantaneous power efficiency of the transformer. However, because the load will vary (usually) in a cyclic manner, a more useful quantity to give efficiency of operation is the energy efficiency. This takes into account the duty cycle of the transformer operation and will take account of the core loss during no-load and light load situations. For example, for a daily load duty cycle of the following:

8 hours at full load, 0.8 lagging 6 hours at 0.6 per unit, 0.8 lagging 6 hours at 0.4 per unit, unity PF 4 hours at no load (but energised)

The energy efficiency is given by the following equation:

24hr energy supply

efficiency=24hr energy supply + 24hr energy loss

Thus:

2 2

2 2

[8x0.8 6x0.6x0.8 6x0.4x1.0 0]

[8x0.8 6x0.6x0.8 6x0.4] 24 total ]ec Cu

V I

V I P E

6 Transformer Tests


There is a need to monitor transformer load to make the most effective use of the transformer. It is also necessary to know the core and load losses. These will normally be given on the nameplate, but the hysteresis losses and eddy current losses will not be separated and this may cause some problems if there are harmonics to contend with. Thus some tests may need to be done. 6.1 Open Circuit Test [for core loss determination] This test requires normal operating flux in the core and hence needs rated voltage to be applied. There is no load connected so there is no load loss contribution in the measured power, which is thus only the (constant) core loss Po. The test requires measurement of supply voltage (rated value) V1, exciting current I0 and exciting real power P0. It should be noted that I0 is not sinusoidal because of the non-linearity of the B-H magnetising curve of the core material. This needs to be considered in choosing the ammeter and the wattmeter types. Open circuit test provides P0, I0 and also Rc and Xm for the equivalent circuit.


0

0 01 0

0 0 0 0

total core loss

: m c

Pcos P

V I

I I sin I I cos

2

1 1 1 1

0 0 0

c mc m

V V V VR X

I P I I sin

6.2 Short Circuit test [for load loss determination] In this case, the I2 (and I1) is the rated current, but the applied voltage V1 is the impedance voltage level, only about 5%. Thus the core flux density� , is only about 5% and thus core loss is negligible, but full rated currents flow in the windings so that the measured power Psc is the copper loss in the winding resistances only. The test requires measurement of Psc, V1, I1 and I2. The test results give the copper loss and also Zeq and total winding resistance Req and leakage reactance Xeq.


1 1

scsc

Pcos

V I

1 1

11 1

' '1 2 1 2

+

=( ) ( )

eq sc sc

V VZ cos j sin

I I

R R j X X

2 '

1 1 2( ) total copper losssc scP I R R P 7 Effects of Harmonics on Transformer Operation The increasing level of harmonic content in the general power supply waveform is causing some potential problems for all electrical equipment items with magnetic core materials. This includes in particular the transformer. In addition, there is also an increase in the number of non-linear loads that are being used, particularly from the increasing use of power electronic controllers. Such devices are a problem because the losses which are generated in the transformer are frequency dependent and,


on a relative basis, the heating by the harmonic components scales with frequency. If there is harmonic content in the supply voltage, the core losses scale with the square of frequency. If the harmonic content of the load current is high, such as may occur with power electronic devices, there is a frequency dependent increase in the copper loss due to eddy currents (skin effect) and the transformer may need to be de-rated so that it does not overheat with a high harmonic load. In the following, the considerations of increased losses in transformers due to harmonics are those due only to current in the load and thus the increase is in the load loss due to eddy current generation by harmonic currents in the primary and secondary windings. The harmonic currents may also cause some harmonic distortion of the exciting voltage by virtue of the effect of the distorted current on the voltage drop in the leakage reactance of the equivalent circuit. This voltage harmonic distortion (from an assumed pure sinusoid) could then also lead to increased losses in the core material in addition to the windings. However it is found that the effect on the core loss of load current harmonics is not generally significant and thus it is usually neglected and only the load loss increase is considered when de-rating calculations are performed. This approach thus assumes a pure sinusoidal supply voltage. There are two approaches to estimating the de-rating required for the transformer:


the CBEMA (Computer and Business Equipment Manufacturer’s Association) Crest Factor Method

the IEEE K-Factor Method 7.1 CBEMA Crest Factor The Crest Factor is defined as the ratio of the peak value of the current (amps) of the distorted waveform to the true RMS value (amps) of the distorted waveform.

Peak current (amps)Crest Factor = C.F. =

True RMS current (amps) The de-rating factor is then determined by the ratio of 1.414

2 to the calculated crest factor.

1.414De-rating Factor =

C.F. Thus, for a pure sine wave, the C.F. is 1.414 and the de-rating factor is then unity (1.0). For a crest factor of 2.0, the de-rating factor will then be 0.707. Thus in the latter case a 100kVA transformer would need to be de-rated to about 70kVA to avoid overheating. While the above approach is a simple method it has some faults:

(i) the peak value of current is not necessarily truly representative of the harmonic content. Two different waveforms with the same level of total harmonic


distortion (THD) may have quite different values of peak current.

(ii) the measurement of the current data needed to calculate the crest factor requires an oscilloscope and a true RMS current meter. In particular the need for an oscilloscope is unwieldy for testing.

For these reasons the CBEMA crest factor method is not widely used and the more quantitatively accurate K-factor method is preferred. 7.2 The K-Factor Method The total harmonic distortion (THD) of a current waveform is defined as:

2

2

1

( )

THD = n

n

I

I

where n is the harmonic number and n=1 is the fundamental (50 Hz) component. The K–factor is defined as:

2 2

1

2

1

.K =

nn

nn

I n

I

where In is the nth harmonic current in amps


7.2.1 Application of the K- Factor

Because ( )n

n puRMS

II

I

where In(pu) is the per unit value of the nth harmonic current and IRMS is the true RMS current, we can express the K-factor as:

2 2( )

1

2( )

1

.n pun

n pun

I nK

I

Because the eddy current losses scale as the square of frequency, the K-factor provides a useful indicator of the increased heating due to the harmonic content. It further gives a quantitative means of calculating the de-rating factor for transformers. Typically, K may vary up to 20 or more for badly distorted current waveforms. 7.2.2 Calculation method for transformer de-rating factor The total losses (PLL) are defined as: 2

LL ECP I R P where: I2R = total winding loss at pure 50Hz operation.


PEC = eddy current loss in the windings We also define PEC(R) to be the eddy current loss at rated current (IR) at 50 Hz.

Thus

2

2EC EC(R)

1

2

2ECEC(pu)

1EC(R)

2 2EC(pu) EC(R)pu ( )

1

. watts

.

.

n

n R

n

n R

n pun

IP P n

I

P IP n

P I

P P I n

We also have: LLpu ECpu1P P

And for rated load LL R pu EC R pu1P P

Using

2

1

2( )

1

2 2 2( ) ( ) ( ) ( )

1 1

( )

.

nn

pu n pun

LL pu n pu EC R pu n pun n

I I

I I

P I P I n

The maximum permissible current is given by:

LL(R)pu EC(R)pumax( )

EC(R)pu EC(R)pu

1

1 . 1 .pu

P PI

K P K P


Example:

100kW of personal computers are supplied from a transformer rated at 150kVA. Given the harmonic current levels caused by the computers are as below, and that

EC R 10%P , calculate the K-factor and the required de-

rating factor of the transformer.

Harmonic no.

% of fundamental

Harmonic no.

% of fundamental

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

100 0.2 66 0.4 38 0.4 13 0.3 4.5 0

5.3 0.1 2.5 0.1 1.9

17 19 21 23 25 27 29 31 33

1.8 1.1 0.6 0.8 0.4 0.2 0.2 0.2 0.2

[In this case 1 rated current = 1.0 puI ]


We can neglect even harmonics in this case and just use odd harmonics. In calculating the harmonic contribution to the K-factor value, we stop when the contribution of a high harmonic becomes negligible. In this case this occurs after the 25th harmonic. We construct the table as follows:

n Current(pu)

Freq. (Hz)

2n puI 2n

2 2n puI n

1 3 5 7 9 11 13 15 17 19 21 23 25

1.00 0.66 0.38 0.13 0.045 0.053 0.025 0.019 0.018 0.011 0.006 0.008 0.004

50 150 250 350 450 550 650 750 850 950 1050 1150 1250

1.0000 0.4356 0.1444 0.0169 0.0020 0.0028 0.0006 0.0004 0.0003 0.0001 3.6E-5 6.4E-5 1.6E-5

1 9 25 49 81

121 169 225 289 361 441 529 625

1.0000 3.9204 3.6100 0.8281 0.1640 0.3399 0.1056 0.0081 0.0936 0.0437 0.0159 0.0339 0.0100

Totals: 1.6031 10.1732

2( )

1n pu

n

I

2 2( )

1

.n pun

I n

Thus:


2 2( )

1

2( )

1

.10.1732

= = 6.351.6031

n pun

n pun

I nK

I

For the de-rating calculation:

EC(R)pumax (pu)

EC(R)pu

1

1 .

PI

K P

The value of EC R puP will normally be available from the

manufacturer. We take a typical value of 0.1 in this example: Thus

max( )

1 0.1

1 6.35 x 0.1puI

pu

= 0.82 pu Thus, the de-rated permissible loading of the 150 kVA transformer is:

0.82 x 150 = 125 kVA Note:

Performing the same calculation as above for the same

EC R 0.1P , but for different values of K, we find the

following de-rating factors:


K max puI kVA

2 10 20 30

0.96 0.74 0.61 0.52

144 111 91 78

7.3 Comparison of the Crest factor values and the IEEE

values for the same loads The table below shows the variance of the two methods for a wide range of building loads with various harmonic levels. A negative value means the CBEMA value is lower and positive that the IEEE value is lower. Reference value is the IEEE value.


CBEMA kVA - IEEE kVA

x 100IEEE kVA


8 K-Factor Transformers K-factor transformers are designed to be able to be used for loads with harmonic distortion without the necessity of de-rating. If a K-factor transformer is to be used in an application it is necessary to know the load characteristics and the harmonic content over the whole load cycle and then to calculate the K-factor and specify a transformer with the required K-factor value. For most general applications a K-factor rating of 15 or less is adequate. Because they must be designed to reduce the level of eddy current generation in the windings, or to allow better dissipation of losses, K-factor transformers are:

more expensive (about two times) heavier (about 15-20 % more) larger

when compared to standard power transformer designs of the same kVA rating. They may have a shield between the two windings to limit harmonic induction: the basic conductor section size making up the transposed windings (particularly the low voltage) are made smaller to limit eddy currents (skin effect) while the overall conductors may be made larger to reduce ohmic heating by the power frequency current. Neutral conductors may be made larger to limit the heating effects of triplen harmonics.


The core is often made of better quality magnetic steel with lower hysteresis loss and perhaps thinner laminations to reduce core eddy current losses. The overall core size may be made larger to reduce operating flux density and hence eddy current and hysteresis losses. However the core losses are only a minor part in this aspect if the supply is free of harmonics and the only harmonics are in the load current. Cooling is also enhanced in K-factor transformer design. Overall the fundamental property of K-factor transformers is that they have lower losses than standard transformers of the same rating for the same level of harmonic distortion. In general, only the transformer winding loss is used in the K-factor calculation. The core loss is not important in this determination as at full load the load loss is the much higher level. However some designs do have lower core loss as discussed above. Dry-type transformers are more susceptible to harmonic effects because of their lower heat dissipation coefficients. This lower heat dissipation results from the lack of oil. Oil provides a much more efficient convection loss method (whether natural or forced) compared to the mainly conduction loss mechanism which dominates the thermal dissipation process in dry-type units. K-factor transformers will have a lower impedance than the equivalent rating standard transformer design.



Fault Calculation Methods

There are two major problems that can occur in electrical systems: these are open circuits and short circuits. Of the two, the latter is the most dangerous because it can lead to very high fault currents and these currents can have very substantial effects (thermal heating and electromechanical forces) on equipment that may require replacement of equipment and may even cause fires and other similar ensuing effects in the electrical system. Building systems are particularly at risk. To prevent problems from short circuits, it is necessary to design electrical protection systems that will be able to detect abnormal fault currents that may occur and then take remedial action to isolate the faulty section of the system in as short a time as is consistent with the magnitude of the short circuit fault current level. This requires that the fault current be predicted for a fault in any particular location of the circuit system. We thus need to establish methods of fault calculation. Fault calculation is not simple for a number of reasons:

There are many different types of fault in three phase systems


The impedance characteristics of all electrical items in the system must be known

The fault impedance itself may be non-zero and difficult to estimate

There may be substantial fault current contribution from rotating machines etc.

The initial cycles of fault current may be asymmetric with substantial DC offset

The earth impedance in earth faults can be difficult to estimate accurately

DC system faults also include inductance effects in fault current growth

For example, the possible fault types that may occur in a three-phase system are:

Three phase (symmetrical) faults (the most severe in terms of current)

Phase to phase fault

Single phase to earth fault

Three phase to earth fault

Phase to phase to earth fault Each of these fault types will have different fault current when they occur at the same location and the electrical protection system will need to take this into account when operating time is determined.


In the very simplified coverage of fault calculations that follows we will look only at symmetrical three phase faults. We do not cover any asymmetrical faults (phase to phase and single phase) except for some general comments on their behaviour. In general, three phase symmetrical faults will give the maximum fault current level at any location and thus such calculations represent worst case situations in general. Because they have low impedance systems, low voltage electrical systems, such as those in buildings, generate very high levels of fault currents. The prospective short-circuit current and the fault level (power) are important parameters that the designer of an electrical installation needs to know and can be obtained from the electricity distributor. The prospective short-circuit current is defined as the current which would flow as a result of a bolted 3-phase fault. Typical value at the point of supply for 230/400V NSW distribution systems:

Suburban residential areas: 10 kA Commercial and industrial areas: 25 kA

Knowing the fault level, the impedance of the upstream circuit and devices (e.g. transformers, conductors) can be derived. The prospective fault current varies at different points in the supply:


At the supply transformer terminals, it is limited by the impedance of the distribution transformer and conductors.

At the main switchboard, the fault current is further reduced because of the additional cable impedance of the consumers mains.

At the distribution board, the fault current is further decreased by the cable impedance of the submains.

Y

MSB DBSupply

Transformer

Consumersmains cables

Submainscables

Utilitycables / lines

YY

MSB DBSupply

Transformer

Consumersmains cables

Submainscables

Utilitycables / lines

Example: An 11kV to 400/230V transformer has a prospective fault current of 32kA at the secondary terminals. The consumers mains circuit has a route length of 25m, using single-core 120mm2 active conductors and 70mm2 neutral. The submains circuit has a route length of 35m, using 16mm2 multi-core cables. We want to determine the prospective fault current at the main switchboard and the distribution board (for the purpose of selecting appropriately rated protection devices and switchgear).


Transformer impedance:

230

0.00718 32000TX

VZ

I

Assume cable temperature is 45oC, from Table 34 of AS3008.1, impedance of consumers mains (1 phase):

0.170

25 0.00425 1000CMZ

Prospective fault current at main switchboard:

230

20.2 kA0.00718 0.00425SCI

Assume cable temperature is 45oC, from Table 35 of AS3008.1, impedance of the submains (1 phase):

1.26

35 0.0441 1000SMZ


Prospective fault current at the distribution board:

230

4.158 kA0.00718 0.00425 0.0441SCI

Note that the above calculations are for a short-circuit fault across three phases. A short-circuit from a single phase to neutral will produce a lower fault current. Here, we need to include the impedance of the neutral cable. 1. The Per Unit System Fault calculations are simplified very substantially if they are performed using the per-unit system and normalising all electrical quantities relating to the fault in per unit


values for the fault analysis. This allows the removal of the complexity of transformer ratios in the fault calculations. The transformer can be included as a simple impedance. In the per unit system we express voltage, current, kVA and impedance as per unit values of selected base values of those quantities. Thus

puB

VV

V VB is the voltage base

puB

II

I IB is the current base

puB

SS

S SB is the kVA base

puB

ZZ

Z ZB is the impedance base

It is usual to specify the two base values VB and SB and then the other two base values IB and ZB are able to be determined from the specified VB and SB values by normal (Ohm’s Law) electrical relationships:

BB

B

SI

V


2

B BB

B B

V VZ

I S

Normally, the voltage base VB is taken as the rated system voltage and SB is arbitrarily specified (often 100, 10 or 1 MVA is chosen), although a common method is to use the rating of a major element in the system such as a transformer or generator as the base SB. For balanced symmetrical three phase faults the fault calculation is able to be done on a single phase basis using the per unit phase impedances in the one-line diagram of the fault circuit. Some care must be taken to use the proper phase kVA and voltage levels in the single-phase circuit to calculate the appropriate base values of current and impedance.

3

BB

B

SI

V

2

BB

B

VZ

S

where VB and SB are the line voltage and three phase kVA values. In the fault calculation the impedances in the fault circuit must include all significant components and all of these must have their impedance expressed in per unit terms using the appropriate base value. This requires changes in


some per unit values if they are already expressed (for example on the name plate) using different base values. This may commonly occur with transformer impedances. To change per unit impedances from one base value to another we have to use the following equation as the basis for change:

ohmspu ohms 2

B

B B

Z SZ Z

Z V

Thus: (i) For change of kVA base (SB), the new Zpu is given by:

new

pu new pu oldold

B

B

SZ Z

S

(ii) For change of voltage base (VB)

2old

pu new pu old 2new

B

B

VZ Z

V

(iii) For change of both kVA and voltage bases at the same

time:

2new old

pu new pu old 2old new

B B

B B

S VZ Z

S V


In most cases the impedances of items such as transformers, generators, motors etc, will be given on name plates in per unit or percentage terms based on the equipment’s rated voltage and power levels. These given values must be adjusted to the base values chosen for fault calculations if these are different from the nameplate values. For cables, overhead lines, busbars, etc, the impedances will most likely be given or obtained in ohmic values. These must then be used with the appropriate base values to get their per unit values referred to the common bases. Thus the appropriate operating voltage and chosen SB must be used to get ZB and IB. The base impedances and currents for a 1 MVA (1000 kVA) base and typical common voltage levels are shown below, using VB equal to rated voltage: [1000 kVA is the 3-phase base value]

Line Voltage

(V)

Phase Voltage

(V)

Base Current

(A)

Base Impedance

() 415

3300 6600 11000 33000 66000

240 1905 3810 6351 19053 38105

1391 175 87.5 52.5 17.5 8.75

0.1722 10.89 43.56 121.0 1089 4356


Example: A 3-phase radial transmission system is shown below. Calculate terminal voltage of the generator. Use a base of 100MVA for all circuits.

50MW0.8pflagging

30kV

132kV 33kV11kV 132kV

50MVAX=10%

50MVAX=12%

j100

Line

VS VS

50MW0.8pflagging

30kV

132kV 33kV11kV 132kV

50MVAX=10%

50MVAX=12%

j100

Line

VS VS

Base impedance of the line:

232

6

132 10174

100 10B

BB

VZ

S

Per-unit reactance of the line:

100

0.575174B

Z jj

Z

Per-unit reactance of sending-end transformer:

(new)pu(old)

(old)

1000.1 0.2

50B

B

SZ j j

S

Per-unit reactance of receiving-end transformer:

(new)pu(old)

(old)

1000.12 0.24

50B

B

SZ j j

S

Load current (using formula 3 cosL LP V I ):

6

3

50 101203

3 30 10 0.8

A


Base current for the 33kV line:

6

3

100 101750

3 3 33 10B

B

S

V

A

Hence, per-unit load current is:

1203

0.6871750B

I

I pu

Per-unit voltage of load busbar:

30

0.9133B

V

V pu

The equivalent circuit is shown below:

j0.2pu j0.575pu j0.24pu

0.687pu0.8pflagging

ESVS VR=

0.91pu Load

j0.2pu j0.575pu j0.24pu

0.687pu0.8pflagging

ESVS VR=

0.91pu Load

Hence,

0.687 0.8 0.6 0.2 0.575 0.24 0.91 0.0SV j j j j j

1.328 0.558SV j pu

1.44SV pu or 1.44 11 kV 15.84 kV


2. Fault Calculation Effects and Requirements Fault levels in a power system are required to be determined at the design stage to allow determination of the following parameters:

(i) overcurrent protection requirements (ii) peak electromagnetic forces (iii) thermal heating effects (iv) the maximum fault current (and the minimum fault

current) (v) the (time) discrimination requirements of protection

operation (vi) the touch voltages on earthed objects (personnel

safety) 2.1 Sources of fault currents In a complex electrical system, there are a number of potential sources of fault current when a short circuit occurs in the system. These are:

(i) the electrical utility supply grid system (ii) any in-house generation systems operating at the

time of the fault (iii) any motors operating within the system at the time

of the fault (iv) any electrical storage elements in the system (e.g

capacitors)


Static equipment such as power electronic inverters and converters, transformers, induction heaters are not sources of fault current. Capacitors in power factor correction systems and battery operated uninterruptible power supplies may be fault current sources however, although generally the contribution of fault current is low and of very short duration. The supply utility contribution to the fault provides a constant fault current, as will the in-house synchronous generation for a short period, but motors will provide decaying fault current contributions as their magnetic excitation fields collapse. Synchronous motors will sustain their fault current level much longer than induction motors. 2.2 Fault impedance variation In calculating fault currents, all components, including the source impedances, must be represented in the one line diagram by an effective impedance in per unit value. For the utility supply this is constant (a stiff source) but for the motors there is a time-varying impedance depending on the time after the short circuit. Depending on when the fault current needs to be calculated, any of three impedances may need to be used:

(i) sub-transient reactance (Xd”) (ii) transient reactance (Xd’) (iii) synchronous reactance (Xs)


We must use the sub-transient reactance for the fault current during the first few cycles, the transient reactance for the fault current up to a fraction of a second and the synchronous reactance for very long duration faults (usually synchronous reactance is not necessary as the protection should operate before it comes into effect). For synchronous motors only the sub-transient and transient reactance are normally used before the exciting field dies away and the fault current contribution is then effectively reduced to zero. For induction motors, only the sub-transient reactance is used before the fault current contribution dies to zero. 2.3 DC Offset This must be included in fault calculations, particularly in low voltage systems as the offset can increase the initial current levels substantially. The magnitude of the DC offset level is governed primarily by the X/R ratio of the faulted circuit. [The offset magnitude is also dependent on the angle on the voltage waveform at which the fault occurs. However the worst-case situation is always assumed in the fault calculation]. 2.4 Types of AC faults

The classes of faults that can occur in AC power systems are:

Three phase fault ELEC9713: Industrial and Commercial Power Systems p. 16

Three phase to earth fault Phase to phase fault Phase to phase to earth fault Single phase to earth fault.

The first of these gives the highest fault current and is the one which will be used in the following examples. However the most common fault is the last type, the single phase to earth fault and at low voltages the fault impedance becomes an important factor in that type, particularly at low voltages. The estimation of fault impedance in such cases is very difficult. The following diagrams show some of the above effects of fault currents.


Example of generator S/C current waveform


Fault type Magnitude 3-phase (most severe)

Line-to-line Line-to-ground

(usually least severe)

(E/Z) x multiplier About 0.87 x 3-phase fault

Depends on system grounding


3. Fault Calculation Methods For the simple fault calculations that we will cover here, we assume the following: (i) The fault is balanced 3-phase symmetrical.

(ii) All significant component impedances are included.

(iii) The fault itself has zero impedance [that is, it is a “bolted” short circuit].

(iv) Earth circuit impedance is neglected because of the balanced 3-phase nature which eliminates the earth impedance.

(v) The appropriate rated voltage is used as the voltage base value.

(vi) For LV systems where resistance is important, we

use the impedance determined by 2 2Z R X .

(vii) Record X R for all equipment, if necessary, to calculate the level of the DC offset multiplier after the symmetrical fault current has been calculated. It is necessary to know R and L separately.

The first step in the process is to convert all impedances to per unit values and to then use these to draw a single line diagram of the fault circuit, including all possible sources modelled as an ideal voltage source with their appropriate source impedance value connected. Then, by a process of circuit simplification the impedance diagram is reduced to a


single per unit impedance ZF connected to true earth and to an ideal voltage source.

puFZ

pu 1V

puFZ

pu 1V

Then the fault current and fault power in per unit value are:

pu

pu(pu)

FF

VI

Z and

2pu

(pu)

pu

F

F

VS

Z

Thus:

(pu) (pu)

pu

1F F

F

I SZ

when we define pu 1V

The actual fault current is puF BFI I I amps and fault

power is pu VAF BFS S S .

The advantage of using pu 1V is evident from the above.

4. Faults in DC Systems DC systems are becoming increasingly common with the use of power electronics and the calculation of fault


currents in such systems is also necessary to consider in modern commercial and industrial systems. In DC systems the impedance elements which determine the steady state fault current level are only resistance elements. However in most cases the system inductance will also have a significant effect in that it will determine the rate of increase of the fault current level in DC system faults. The L/R time constants of such systems are usually long enough that the steady state fault current will not be reached before protection operates and the protection will thus be interrupting current when that current is still rising. Thus DC fault calculations are not necessarily simple to perform. The sources of DC fault currents are, typically, any of the following:

DC generators Synchronous converters DC motors Rectifier systems Battery banks UPS systems

Another factor that must be considered in the design of the protection system is that DC arc currents are more difficult to interrupt than AC arc currents. An AC circuit breaker has 100 current zeroes per second to interrupt the fault current, while a DC breaker has none. Thus the arc interruption is much more difficult for DC than for AC. In a DC breaker


the arc voltage developed is an important factor in the protection design and in determining fault current levels. As a result of the difference between AC and DC faults, either specialised DC breakers or fuses must be used or, more commonly, if AC breakers are used they must be de-rated for use on a DC system. The fault calculation procedure must involve the determination of the time constant and thus the initial exponential rate of rise of current as it is most likely that interruption will occur during this period. A DC fault is modelled by a DC supply in series with a fixed circuit resistance, a fixed circuit inductance and a variable resistance in the form of the circuit breaker arc when its contact open (see figure over page). The governing equation during the initial transient is:

S R a

dIV V V L

dt

or S R a

dIL V V V

dt

Initially, when Va is small or zero, S R aV V V and dI dt

is positive and current increases, but later as the arc develops and lengthens, S R aV V V and dI dt is

negative and current decreases. The typical behaviour is shown below.


+

_

+

_

I

aVSVLR

C.B.

+ +RV

dIL

dt

DC fault circuit and C.B.

+

_

+

_

I

aVSVLR

C.B.

+ +RV

dIL

dt

DC fault circuit and C.B.

5. Fault calculation data and calculation example


The following tables give details relating to various parameters required for fault calculations and an example of a typical fault calculation procedure.



5.1 Example of a Simple Fault Current Calculation


(1) Utilitysupply

(2) Transformers

(3) Generator

(4) Cable

(4’)

(7)

(8)

(9)

(10)

(6) (5)

MotorCable

Cable

Cable

Cable

Power transformer

Current transformer

(1) Utilitysupply

(2) Transformers

(3) Generator

(4) Cable

(4’)

(7)

(8)

(9)

(10)

(6) (5)

MotorCable

Cable

Cable

Cable

Power transformer

Current transformer

Impedance circuit

We are required to find fault current at location A: voltage is 480V. Use base of 20 MVA for p.u calculation, i.e. 20 MVABS At 4.8 kV:

pu 1V 2

1.152 BB

B

VZ

S

2406 A3

BB

B

SI

V

At 480 V:

pu 1V 0.01152 BZ 24056 ABI


(1) Source impedance:

500 MVA fault level 25 p.u pu

10.04

25Z p.u

6X R pu 0.0066 0.0395Z j

(2) Transformer:

3000 kVA, 6% pu

200.06 0.4 p.u

3Z

8X R pu 0.05 0.4Z j

(3) Generator:

1000 kVA, 15% pu

200.15 3

1Z p.u.

10X R (negligible R) pu 3.0Z j

(4) Cable (4.8kV):

0.06 0.03j 1.152 BZ

pu

0.06 0.030.052 0.026

1.152

jZ j

p.u.

(4’) Cable (4.8kV):

0.05 0.02j 1.152 BZ

pu

0.05 0.020.043 0.017

1.152

jZ j

p.u.

(5) Motor:


200 kVA, 10% pu

200.1 10

0.2Z p.u.

10X R (negligible R) pu 10Z j

(6) Cable (4.8 kV):

s pu 0.052 0.026Z j p.u. (7) Cable (4.8 kV):

0.1 0.04 j 1.152 BZ

pu

0.1 0.040.087 0.035

1.152

jZ j

p.u

(8) Power transformer:

1000 kVA, 4% pu

200.04 0.8

1Z p.u.

4X R pu 0.194 0.776Z j p.u.

(9) Current transformer (480V):

0.0001 0.0005Z j 0.01152 BZ

pu

0.0001 0.00050.0087 0.0043

0.01152

jZ j

p.u.

(10) Cable (480V):

0.002 0.002j 0.01152 BZ pu 0.174 0.174Z j p.u.


(1)+(2)

(4’)

(7)+(8)+(9)+(10)

(5)+(6)

(3)+(4)

F

(1)+(2)

(4’)

(7)+(8)+(9)+(10)

(5)+(6)

(3)+(4)

F 1 2 0.0566 0.4395 0.443 p.uj

3 4 0.052 3.03 3.03 p.uj

5 6 0.052 10.03 10.03 p.uj

4' 0.043 0.017 0.046 p.uj

7 8 9 10 0.4637 0.9893 1.093 p.uj

Approximation :

1 2 3 4 0.443 3.03 0.384

0.384 4' 0.430

0.430 5 6 0.430 10.03 0.412

0.412 7 8 9 10 0.412 1.093 1.505


Total pu 1.505Z p.u.

Fault pupu

10.664I

Z

At 480V: 0.664 0.664 24056 15970 AF BI I

At 4.8 kV: 0.664 0.664 2406 1597 AF BI I Alternatively :

1 2 3 4 0.0440 0.3843j

0.0440 0.3843 4' 0.0870 0.4013j j

0.0870 0.4013 5 6 0.0805 0.3865j j

0.0805 0.3865 7 8 9 10j 0.5442 3758j

Total pu 0.5442 1.3758 1.4795Z j p.u.

Fault pupu

10.676I

Z

At 480V: 0.676 0.676 24056 16262 AF BI I

At 4.8 kV: 0.676 0.676 2406 1626 AF BI I


Handout 1

Documents

Transcript of Handout 1