Asymmetric switching duty for high-voltage circuit-breakers
Almost all engineers are familiar with the concept of short-circuit levels and their importance in correctly selecting fault level ratings for equipment; however, in SPE's experience we often find that many companies have only considered symmetrical break current fault levels, and have not adequately considered the possibility of the increased duty placed on the switchgear by the presence of large supply transformers, motor loads, or embedded generation.
Author: Stephen Sommerville BSc, MEng, CEng, MIET
Almost all senior engineers are familiar with short-circuit studies, and understand their importance in proving that an electrical system is safe for operation, yet in SPE’s experience, the author finds that many engineers do not fully appreciate how a system with a high X/ R ratio can place an increase duty on circuit breakers, and potentially cause unsafe operating conditions. This article details the importance of X/ R ratios in short-circuit systems, and explains the relationship with the DC component of a fault current and the network time constant. This article will show that in most systems the X/ R ratio is low and no special measures need to be taken, but in systems with large transformers, local generators, or a large amount of motors load, the network X/ R can be much higher. The increase of X/ R ratio leads to a slower decay in the DC component of the fault current, so that when the circuit breaker operates, the circuit breaker experiences a much higher percentage of DC current than usual. This can lead to circuit breaker fault levels being exceeded and potentially delayed current zeros, which in turn can pose a risk to the system security and personnel. This article is intended to help engineers understand why asymmetry occurs, what its effects are and to provide practical measures to calculate the asymmetry and deal with any problems that may arise.
Almost all engineers are familiar with the concept of short-circuit levels and their importance in correctly selecting fault level ratings for equipment; however, in SPE's experience we often find that many companies have only considered symmetrical break current fault levels, and have not adequately considered the possibility of the increased duty placed on the switchgear by the presence of large supply transformers, motor loads, or embedded generation. This article is intended to explain this phenomenon and give working engineers a practical insight into analysing, and resolving issues associated with circuit breakers asymmetric duty.
Problems relating to high asymmetrical breaking duties are most common in large distributed network operator (DNO) substations, heavy industrial plants, process plants, water processing/treatment, power generating plants (large or small) and combined heat & power (CHP) installations, as it is these sites that most often contain large transformers, generators or a large amount of motor load. These devices have a high X/ R ratio, which increases the overall system X/ R ratio and network time constant. This leads to a slower decay in the fault current, which in turn means that any circuit breakers experience a higher percentage DC component when they operate. This increased DC component offsets the symmetrical fault current, making it appear larger than it actually is, and if the circuit breakers are not suitably rated for the duty then the system and personnel can be placed at risk.
Three-phase fault calculations revisited
The IEC 60909  standard gives very detailed guidance on how to calculate fault levels for different scenarios, but in practice, many engineers tend not to refer to this standard, and instead rely on simple hand calculations, or use a computer simulation package to determine the fault levels. While this is suitable in many cases, it is not always appropriate because hand calculations and simple computer programs usually do not adequately consider the DC component of a fault current. In other cases, some computer analysis programs can calculate the asymmetrical switching duty, but do not present the results in a clear way. This can lead to engineers misinterpreting the results and believing the system design is adequate, when in fact there is an underlying problem. It is therefore always good practice to familiarise oneself with the detail of IEC 60909 [ 1 ] and the limitations of any computer analysis program in use. To help illustrate the phenomenon of asymmetric switching duty, it is worthwhile to revisit the well-known characteristic curve for a short circuit, as shown in Fig. 1, extracted from IEC 60909 .
Fig 1 Short circuit current of a near-to-generator short-circuit with decaying AC component
This well-known curve indicates the behaviour of a three-phase fault, and shows how the AC fault changes over time. The short-circuit current in the first few cycles is all of a similar magnitude, but it decays with time as the DC component drops off; as time progresses, the magnitude of the AC fault drops off further as the synchronous generators move from sub-transient, to transient and then steady state. The implication of the DC offset is twofold, first, the peak short-circuit current is increased and second, the root mean square break current, at the point of the circuit breaker contact separation is also increased.
Reviewing the curve in Fig 1 , it should make it apparent that there is also another factor in play; and this is the time when the circuit breaker opens. If a circuit breaker is fast acting and operates at around 20 ms, it will be subjected to a much higher percentage of DC component than a slower circuit breaker that operates at around 100 ms, because the DC component will have had a chance to decay. In practice, most circuit breakers will usually operate within 50–100 ms, so under normal circumstance the DC component will have had chance to decrease by a significant portion, such that the fault current appears either with a small DC offset.
The area of interest for short circuit studies is primarily when the DC component decays more slowly than normal, or when the operating time of the circuit breaker is fast; as the fault level experienced by the circuit breaker appears to be much higher due to the higher percentage of DC component. In this case, the circuit breaker is placed under increased duty, and can potentially put people and the system at risk, as it may not be able to successfully break the fault current. The percentage DC component in the fault current is therefore the key variable that must be understood when undertaking short circuit studies.
In extreme cases the DC offset can be of sufficient magnitude that the AC fault current waveform does not pass through a current-zero for several cycles, and operation of the breaker must be delayed, or the circuit breaker will chop the current, or fail to extinguish the arc. Fortunately, delayed current zeros are unusual for smaller systems, and tend only to be encountered on large generating stations, but where a system has a very high X/ R ratio engineers should be alert to their possibility. A more detailed discussion of the implications of this phenomenon is beyond the scope of this article but when dealing with large power generation equipment engineers should be alert to this potential scenario.
It should therefore be clear that when undertaking a short-circuit study, there will almost always be a DC component in the fault current, and in most cases the value of this DC component is sufficiently low, that it does not affect fault current calculations. Where a fault current has a high DC component, this causes the symmetrical AC component of the fault current to be offset, making it appear larger. It is therefore important that any engineer undertaking short circuit studies understands the importance of the DC component and how it can potentially increase the duty on circuit breakers.
X/R ratios and the network time constant
The previous section explained how an increased DC component can lead to increased duty of circuit breakers, but the obvious question is – what causes an increased DC component? In simple terms, a high DC component is caused by a network that has a high X/ R ratio, which in turn is caused by an excess of system reactance (inductance) when compared to system resistance. There are detailed theoretical electrical engineering reasons for this, which relate to theoretical RC & RL networks and their constants. A detailed explanation of this theory is beyond the scope of this article and is discussed in many textbooks; but in simple terms, the inductance in the system stores energy during a fault, and it takes time for this stored energy to dissipate, which is the cause of the DC decrement curve. The more inductance in the system, the longer the time taken to dissipate the energy and the higher the DC component. For fault analysis purposes, the network X/ R ratio is usually calculated by an appropriate computer analysis program, as it becomes impractical to calculate by hand for any, but the smallest systems.
For most practical scenarios, a high X/ R ratio is caused by: (i) large supply transformers, (ii) large quantities of directly connected motors, or (iii) embedded generators. All of these components have a high reactance and X/ R ratio, which contribute to the overall system X/ R ratio. In most cases, it is therefore common to encounter this problem in generating plants, heavy industrial plants, or systems with large grid in-feeds. Typically, manufacturers, must be contacted to obtain the relevant parameters for determining system X and R values, but for preliminary studies the Alstom Network Protection Automation Guide [ 2 ], chapter 5 provides some useful tables which indicate typical X/ R ratios of various sizes of transformers. One of these tables is reproduced in Fig. 2 for ease of reference. Whilst the table is somewhat simplified, it can be seen that transformers rated above 25 MVA, tend to have large X/ Rratios, whereas smaller transformers do not. It is of course noted that transformers actual X/ R ratio will vary considerably, depending on rating and manufacturer, but the Alstom NPAG provides a useful starting point if no data is presently available.
Fig 2 Extract from the Alstom Network Protection Automation Guide 
For equipment manufacturers designing circuit breakers, X/ R values are not as useful as the actual value will vary depending on the frequency of the power system. Instead, switchgear designers, work with a value known as the system time constant τ (tau), which is used to calculate the time taken for the voltage across an inductor to fall to zero. Readers may be interested to note that this constant, is not specifically mentioned in the IEC 60909  standard, but it is referred to frequently in the IEC 62271-100  standard on HV circuit breaker design. Therefore, unless engineers have read both standards they may not be fully conversant with the implications of higher X/ R ratios.
The value of τ can be obtained through simple mathematical formulas as detailed below; but it is simply given by dividing the system inductance by resistance. The formulas for determining the network time constant can be seen in (1)–(3), which are obtained from IEC 62271-100 , section D1.1, and formula (4) gives the expression for determining the actual DC component percentage experienced by a circuit breaker at a given operating time, which is obtained from the same standard, section 4.101.2:
The values in the above formulas are all standard network parameters, and the only unusual values that readers may not be familiar with is the value of T op, which is the circuit breaker operating time and T r which is the value of a half cycle of rated frequency.
The IEC 62271-100  standard states that it is only when the network time constant τ, exceeds 45 ms (nominally a 20% DC component) that special considerations are needed in design and operation of a circuit breaker. If the value of 45 ms is entered into (3), it can be seen that it results in an X/ R ratio of 17 for a 60 Hz system and 14.1 for a 50 Hz system. Therefore, considering a typical system in the UK, if the X/ R ratio is above 14.1 (i.e. τ is greater than 45 ms), then the system must be examined more closely, as the circuit breakers may be under increased duty. The IEC 62271-100 standard also defines other special time constants of 60, 75 and 120 ms, which are used for testing specialist circuit breakers, such as generator circuit breakers which are specifically designed and tested to break high percentages of DC component.
In practice, the key to understanding if a problem exists or not, is to use a suitably sophisticated power system simulation package to analyse the system and to ensure the input parameters are carefully specified. For example, the electrical transients analysis program (ETAP) package can model all of the key parameters associated with a circuit breaker, including its break, and peak capabilities as well as its operating time, TRV capability and the network time constant the circuit breaker is rated for. This facilitates easy comparison between calculated fault levels and device ratings. It is important to note, however, that caution needs to be used when inputting this data and interpreting the results, as it is easy to make a mistake, and a good level of knowledge is often required from the studies engineer. The short-circuit study results that need to be analysed by the studies engineer include: (i) system X/ R ratio, (ii) percentage DC component, (iii) symmetrical three-phase break fault level, (iv) asymmetrical three-phase fault break level and (v) the peak fault level.
Where the analysis program is capable of directly comparing a circuit breakers asymmetric capability with the calculated asymmetric fault level, the results can be read simply from the output files. If this is not the case then the system X/ R ratio and/or DC percentage should be checked – if this is above the recommended limits of 17 for 60 Hz, 14.1 for 50 Hz or a percentage DC component is greater than 20% further action may be required. The short-circuit calculation results should be compared with the circuit breakers rated duty, and if the device capability is exceeded, or borderline, further analysis or design modification will be necessary.
The discussion in the previous sections may have seemed remote from many day-to-day design projects, but it is important to realise that these concerns can appear on what seem to be comparatively simple systems. To help demonstrate this point, a few typical scenarios have been modelled using the computer simulation package ETAP.
The first example, as shown in Fig. 3 , is a part model of a simple 33/11 kV substation and associated 11 kV distribution rings. This is a configuration that would be encountered by many DNOs, or large customers who have their own 11 kV distribution system. On first review, this system appears straightforward, as there are no major motor load, no embedded generation, and only a single large 33/11 kV, z = 10%, 40 MVA transformer, supplying an 11 kV, 20 kA switchboard, with some back-fed 11 kV fault contribution (the 11 kV distribution detail is omitted for clarity).
Fig 3 Simple substation fed by a large transformer modelled in ETAP
The test circuit shown in Fig. 3 is subjected to a normal IEC 60909 short-circuit study, with the results presented in Fig. 4 . It can be seen from Fig. 4 that the initial symmetrical short-circuit current ( I″ k) levels and break current ( I b sym) are high, but are within the equipment's rating by an acceptable margin. However, when a more detailed inspection of the output results is performed, it can be seen that the peak short circuit ( I p) current is nearly at the equipment limit, the asymmetrical break current ( I b asym) has been exceeded by over 1 kA and the DC current rating of the switchgear has also been exceeded.
Fig 4 Simple substation fed by a large transformer ETAP short-circuit results
Normal hand calculations, or simple computer analysis programs of this system would have potentially missed these results, and an engineer giving only a cursory review of the system would not have identified the position. This could then have led to an unsafe operating condition, posing risk to both personnel and the system. Looking in more detail at Fig. 4 , the high asymmetric switching duty is primarily due to the transformer having a high X/ R ratio of 40, this leads to a long network time constant and slow DC decay. This simple example highlights a very common scenario at larger substations, where the simple presence of a large transformer can cause problems. In practice many substations of this nature, could have the problem exacerbated by the presence of embedded generators, or motor load downstream on the 11 kV network, which would further increase the fault level.
A second example is shown in Fig. 5, which considers a small industrial plant with a main 6.6 kV, 25 kA switchboard, containing several large MV motors, a small embedded generator, several LV MCCs and a utility infeed from a 33/6.6 kV, 16 MVA, z = 8% transformer. This a more complex plant, than the previous example, but is representative of the type of plant that industrial companies may operate.
Fig 5 Typical industrial plant modelled in ETAP
The test circuit shown in Fig. 5 is subjected to a normal IEC 60909 short-circuit study, with the results presented in Fig. 6. It can be seen from Fig. 6 that the initial symmetrical short-circuit current ( I″ k) levels, break current ( I b sym) and asymmetrical break current ( I b asym) are high, but are within the equipment's rating by an acceptable margin. However, it can be seen that the peak short circuit ( I p) current has been exceeded by nearly 2 kA and the DC current rating of the switchgear has also been exceeded
Fig 6 Typical industrial plant modelled in ETAP short-circuit results
As with the previous example, normal hand calculations, or simple computer analysis programs would have potentially not picked up this condition. Looking in more detail in Fig. 5 , the high peak current is due to the combination of transformer, MV motors and generator, with some contribution from the LV system. This is a more complex example, but shows how it is important to consider the peak fault current as well as the break current.
A third example, as shown in Fig. 7 , considers a small 15 MW peak matching generation station, containing an 11 kV, 20 kA switchboard and six small 2.5 MW diesel generators, which are connected to the DNO network via a 33/11 kV, z = 10%, 25 MVA transformer. The presence of a large amount of directly connected generators should be an immediate flag for any system designers and an indication that the fault levels could be problematic.
Fig 7 Simple peak matching DG plant modelled in ETAP
The test circuit shown in Fig 7 is subjected to a normal IEC 60909 short-circuit study, with the results presented in Fig 8. The results are similar to the first example and it can be seen from Fig. 8 that the initial symmetrical short-circuit current (I k) levels and break current (I b sym) are high, but are within the equipment's rating of 20 kA by an acceptable margin. When a more detailed inspection of the output results is performed, it can be seen that the peak short circuit (I p) current is nearly at the equipment limit, the asymmetrical break current (I b asym) has been exceeded by over 1 kA and the DC current rating of the switchgear has also been exceeded.
Fig 8 Simple peak matching DG Plant ETAP short-circuit results
This scenario is much less common, but as with the other it is also a deceptive case, as hand calculations, or a simple computer analysis programs would have potentially not picked up this problem, and led to an unsafe operating condition, posing risk to both personnel and the system. The asymmetric switching duty is caused by a combination of the transformers and generators both having high X/ R ratios, which combine to cause an overall X/ R ratio on the 11 kV busbar to become high, leading to a longer network time constant and a slow DC component decay.
The first step in any solution requires a detailed understanding of the problem, and for asymmetrical switching, the only realistic approach requires the use of sophisticated power system analysis software that has the ability to calculate the asymmetric fault currents and compare them with the capabilities of the selected breaker. When the severity of the problem is understood, it is then possible to begin analysing potential solutions to the problem. The key to resolving any potential issue depends on the magnitude of the problem, as a circuit breaker faced with a calculated DC component of 25% is a far less troublesome problem, than the same device facing a DC component of 50%.
The second factor to consider is the magnitude of the symmetrical and asymmetrical fault compared to the symmetrical circuit breaker capability. For example, if an 11 kV circuit breaker has a fault level of 25 kA, and the calculated symmetrical fault level is 23 kA at 35% DC, then it is likely the circuit breakers rating will be exceeded and the designer must consider alternative strategies; but if the symmetrical fault level is only 10 kA at 35% DC, the circuit breaker is likely to be able to cope with the increased duty. Annex I of the IEC 62271-100 [ 3 ] standard provides some clear and practical guidance on how to consider the scenarios when the network time constant is above 45 ms. For network voltages up to 52 kV it recommends:
‘In cases where the short-circuit current requirement is lower than the rated short-circuit breaking current of the circuit-breaker by at least one class within the R10 series, a circuit-breaker tested with a time constant of 45 ms may fulfil the requirements. For example, a 50 kA circuit-breaker tested with a time constant of 45 ms may be adjudged to be adequate for a 40 kA location with a higher time constant.’
In simple English, if the network X/ R ratio is above 14.1 on a 50 Hz system, and the fault level is near the selected maximum switchgear rating, then it is suggested that the next standard R10 size of switchgear fault rating is chosen. For example, a maximum symmetrical three-phase fault level of 18 kA is calculated at 11 kV and the designer decides to select a 20 kA switchboard – the calculated fault level is within the equipment rating, and initially everything appears ok. The designer then uses a computer analysis package and realises that percentage DC component is 35% and will exceed the switchgears rated capacity. The designer then follows the guidance detailed above and selects one R10 switchgear size larger than is necessary at 25 kA, in order to meet the increased duty.
While the general guidance provided above, is suitable for many cases, it needs to be taken with a degree of caution, as its suitability will depend on the severity of the DC component percentage. In the case discussed in the previous page, if the percentage DC component is 30%, using a 25 kA would be adequate for meeting the duty, but if the DC component is 60% then even a 25 kA circuit breaker might not be adequate. This approach is clearly easy to implement when analysis is being performed on a new system, and the designer can specify a higher rating, but where there is an existing switchboard, or the designer is constrained to a maximum fault level due to switchgear manufacturers limits, the solution can be less obvious.
When faced with particularly difficult conditions, such as switchgear manufacturer limits, or very high X/ R ratios, it will be necessary to seek specialist advice from a consultant and/or liaise directly with the switchgear manufacturer. In some cases, it is possible to check the switchgear manufacturers test certificates, and identify the actual percentage of DC component the circuit breaker was tested against, as it is frequently higher than the minimum 20%; and so suitability can be obtained by inspection of the actual test results and certificates. If this is not possible, then more drastic measures may be required to reconfigure the network, replace existing switchgear with specialist circuit breakers or use current limiting reactors, or Is-limiters.
This article has given an overview of how the presence of large transformers, generators or a large amount of motor load can lead to an increased network X/ R ratio. This increased network X/ R ratio, leads to a slower decay of the DC component in the fault current, which results in circuit breakers having to break a higher percentage of DC component at the point of contact separation. This scenario is known as an asymmetrical switching duty, and can lead to excess duty on circuit breakers and unsafe operating condition. It was identified that undertaking asymmetrical analysis of fault levels is difficult by hand, and also that some power system analysis programs may not accurately calculate or present the results clearly, which can mislead a design engineer; consequently, it is necessary to use sophisticated programs for determining the asymmetrical fault levels and device capability.
For practical purposes, it was identified that IEC 62271-100 [ 3] states that where a percentage DC component is above 20%, (equivalent to an X/ R ratio above 14.1 for 50 Hz systems, or above 17 for 60 Hz) then a circuit breakers rating could be exceeded. It was shown that in this scenario, provided the DC component was not too high, the guidance given in IEC 62271-100 , suggests that the next R10 size series circuit breaker, above the calculated fault duty, can meet the increased duty. Finally, it was noted that where the percentage DC component is very high, or it is not possible to increase the switchgear fault rating, a more detailed solution is necessary that should involve input from a specialist consultant and, where necessary, the switchgear manufacturer.
- IEC 60909: ‘Short-circuit currents in three-phase a.c. systems’, 2016.
- Alstom: ‘.
- IEC 62271-100: ‘High voltage switchgear and control gear – Part 100: high-voltage alternating-current circuit-breakers’, 2001.