Transmission lines with high and extra-high voltage that are untransposed, meaning the conductors are not positioned at the corners of an equilateral triangle and lack an overhead grounding wire (OGW), significantly contribute to the generation of negative-sequence voltages and, to a lesser extent, zero-sequence voltages. As the electrical load increases, the incidence of negative-sequence voltages also rises.

During the late 1970s and early 1980s, Eskom operated only two extensively long 400kV transmission lines that delivered electricity to the Western Cape. One of these lines was transposed, while the other remained untransposed.

At that time, the concept of Negative Phase Sequencing was not widely understood among engineers, nor were the factors that caused it.

Negative Phase Sequencing

Before the practice of live-line maintenance was established, it was necessary to periodically shut down each of the two transmission lines for upkeep.

Confusion ensued at Eskom Western Cape when the transposed line was deactivated for maintenance. Despite an uninterrupted power supply, a major customer’s equipment, which was equipped with highly sensitive control systems, experienced frequent shutdowns.

This prompted an investigation, leading to a consultation with a professor from UCT who provided insights into Negative Phase Sequencing. With this new understanding, I immediately directed the maintenance team to cease work on the transposed line and await further instructions.

We then coordinated with the customer to align our maintenance schedule with their plant’s downtime.

The subsequent introduction of a third Transmission Line to the Western Cape significantly reduced the recurrence of such issues.

To prevent similar incidents, we mandated that the untransposed line should not be the sole power source for the Western Cape.

Automatic Change-Over

Reflecting on past events, it later became clear that Negative Phase Sequencing was behind the numerous unexplained electrical disturbances I had been tasked with resolving.

A notable incident occurred years earlier when a colleague and I were summoned to determine why the standby generator at the undersea cable terminal was erratically activating and deactivating. Observing the voltmeter, we noticed a sudden spike in one phase voltage followed by a swift drop in another. At the time, the phenomenon was baffling, and we failed to link it to upstream occurrences.

Consequences

Click here to read more about a recent incident and an explanation of the consequences of these type of network faults.

In a balanced three-phase network, the magnitudes of the voltages and currents in all three phases are the same and these voltages and currents phase are shifted symmetrically by 120 degrees to each other. If the magnitudes are not the same and phase-shifts are different from the 120 degrees, the network is unbalanced.

Eskom strive to supply a balanced set of voltages at the supply points. Although Eskom may attempt to supply a balanced set of voltages, they do not. This is probably true for the point of supply at the generator terminals, but between the generator terminals and the end user are Transmission, Distribution and often Reticulation Networks that are not maintained or monitored with the same diligence that is required to supply balanced voltages. Often the network operators are oblivious to unbalances occurring on the network. Perhaps there are Negative Protection System installed on the Transmission Networks, but it is doubtful that such protective devices exist on Distribution and Reticulation Networks.

Low Voltage (LV) network unbalances are caused by unequal phase impedances, single-phase laterals and other structural asymmetries of LV networks, uneven allocation of single-phase consumers across the three phases on mostly rural reticulation networks, unbalanced three-phase loads and random variation of consumption over time. In addition, the presence of single- phase distributed generation, for example single-phase solar inverters also contributes to the phase imbalances occurring in LV networks. However, load asymmetry represents the main cause of voltage and current imbalances in LV networks. Asymmetric faults are another source of voltage or current imbalances, but these are usually transient events and are usually cleared quickly from the grid, so it could often be ignored.

In many cases operators and maintenance crews are unaware of the negative consequences that phase imbalances can have on LV networks and electrical equipment. Current imbalances would lead to a reduction in the serviceable loading capacity of LV cables and distribution transformers. Because of the imbalances, some of the phases could carry higher loads while the remaining phase of phases could be lightly loaded but, the limiting factor for the addition of three-phase loads is the current on the highest loaded phase or phases.

Current imbalances can also cause additional heat losses in distribution transformers and LV cables in both the phase and neutral conductors. These types of losses represent a significant part of the total losses occurring in LV networks.

When end users, such as municipalities, are supplied by unbalanced voltages, induction machines and power converters face adverse effects such as reduced efficiency, increased losses, potentially dangerous overheating and, in some situations, premature failures. At severe voltage or current imbalance levels, some types of protection relays could malfunction, leading to miscoordination, nuisance tripping and lack of selectivity.

In the United States, voltage magnitude variations are limited to ±5% and in most European countries it is limited to ±10%. For the rest of the world, these limits are not much different.

The question is: is Eskom aware what is happening on the Distribution and Reticulation Networks? It is highly unlikely that they do. In a recent unrelated “survey”, I came across a 10-minute averaged voltage unbalance of 327% between Phase 2 and Phase 1.

The second question is: would this be restricted to just the one substation? It is highly unlikely. In another post I will talk about and incident where the Negative Phase Sequence occurred on the Transmission Network which affected the entire region.

In a balanced three-phase network, the magnitudes of the voltages and currents in all three phases are the same and these voltages and the phases voltages and currents are shifted symmetrically by 120 degrees to each other. If the magnitudes are not the same and phase-shifts are different from the 120 degrees, the network is unbalanced.

In the picture below, the Voltage Vector for the ideal network is shown with the magnitudes of the voltages being equal and the phases shifted symmetrically by 120 degrees to each other, while the solid lines showing the results recorded. The magnitudes of the actual voltages are not equal since the Yellow/White Phase 2 is almost negligible and the phases are not shifted symmetrically.

Eskom strive to supply a balanced set of voltages at the supply points. But they do not always do that. This is probably true for the point of supply at the generator terminals, but between the generator terminals and the end user are Transmission, Distribution and often Reticulation Networks that are not maintained or monitored with the same diligence to supply voltages with the same magnitudes and phase-shifts as mentioned above.

Often the network operators and maintenance crews are oblivious to unbalances occurring on the network. Perhaps there are Negative Protection System installed at the Generating Stations and on the Transmission Networks, but it is doubtful that such protective devices exist on Distribution and Reticulation Networks.

Low Voltage (LV) network unbalances are caused by unequal phase impedances, single-phase faults and other structural asymmetries of LV networks, uneven allocation of single-phase consumers across the three phases on mostly rural reticulation networks, unbalanced three-phase loads and random variations of consumption over time. In addition, the presence of single-phase distributed generation, for example, single-phase solar inverters also contribute to the phase imbalances occurring in LV networks. However, load asymmetry represents the main cause of voltage and current imbalances in LV networks. Asymmetric faults are another source of voltage or current imbalances, but these are usually transient events and are usually cleared quickly from the grid, so it could often be ignored.

In many cases operators and maintenance crews are unaware of the negative consequences that phase imbalances can have on LV networks and electrical equipment. Current imbalances would lead to a reduction in the serviceable loading capacity of LV cables and distribution transformers. Because of the imbalances, some of the phases could carry higher loads while the remaining phase of phases could be lightly loaded but, the limiting factor for the adding more three-phase loads is the current on the highest loaded phase or phases.

Current imbalances can also cause additional heat losses in distribution transformers and LV cables in both the phase and neutral conductors. These types of losses represent a significant part of the total losses occurring in LV networks.

When end users, such as municipalities, are supplied by unbalanced voltages, induction machines and power converters face adverse effects such as reduced efficiency, increased losses, potentially dangerous overheating and, in some situations, premature failures. At severe voltage or current imbalance levels, some types of protection relays could malfunction, leading to lack of coordination, nuisance tripping and lack of selectivity.

In the United States, the voltage magnitude variations are limited to ±5% and in most European countries it is limited to ±10%. For the rest of the world, these limits for voltage magnitude variations are quite similar.

The question is: is Eskom aware what is happening on the Transmission, Distribution and Reticulation Networks? It is highly unlikely that they do. When looking at Eskom’s Quality of Supply document, there is a lot being said about the consequences of unbalanced voltages, but I cannot find the voltage magnitude variations limits.

In a recent unrelated “survey”, I came across a voltage magnitude variation unbalance of 327% between Phase 2 and Phase 1. The instrument was set to record the values at 10-minute intervals and the 327% is the average over that period. See the picture below.

Since the medium-voltage supply comes directly from an Eskom substation which is probably about 20-metres away, one wonders what is going on at the Eskom substation since it is highly unlikely that the voltage unbalance is as a result faulty equipment at the municipal substation. It is as if one phase is completely missing. The same “missing” phase also show an abnormal high current. The neutral current which is supposed to be at or close to zero is also very high.

The second question is: would this be restricted to just the one substation? It is highly unlikely. In another post that is due to come out soon, I will talk about and incident where the Negative Phase Sequence occurred on the Transmission Network which affected the entire region.

Read more about the service we offer.

Newsletter

Join the mailing list and receive our latest newsletters by simply filling in your email address below and then click Subscribe.

When poorly maintained switchgear is continually operated, eventually something drastically is bound to happen. But when those extreme event goes unnoticed, the repercussion could have long-lasting outcomes.

Circuit-breakers must be opened when load-shedding starts and closed when ending. If not always, the contacts of these circuit-breakers are predominantly covered. So, when one contact does not make during the closing process, it would not be detected unless an analysis process is instituted, which is highly unlikely.

Negative Sequence Currents

With the one phase “missing”, the system becomes unbalanced, and this causes negative sequence currents. Such a situation can have harmful effects on power system equipment and operation. It can cause increased heating in motors, generators, and transformers, leading to potential damage or reduced lifespan. It can also create mechanical stresses in rotating machinery such as vibrations and torque pulsations. Additionally, negative sequence currents can lead to voltage dips, unbalanced voltages, and increased power losses in the system.

In the picture below, Ideal Voltage Vector is depicted with a Power Factor of 0.9 and at the rated circuit voltage (6.6kV).

However, the Voltage Vector obtained at 03:40:00 on 2023-09-05 was totally different as can be seen in the picture below.

The picture below shows the Ideal Voltage Vector with the recorded Voltage Vector superimposed on top.

The voltages recorded on Phases 1 & 3 were over 7.2kV whilst the recorded voltage on Phase 2 was almost negligible. The currents recorded clearly indicated the unbalanced system.

Consequences

Looking at the picture of the recorded Voltage Vector, one can image what the magnetic fields would look like in electric motors connected to this network. It would be an ellipse which is totally off-center and playing havoc on the bearings and causing a significant vibration.

Negative Phase Sequencing

It is uncertain that negative phase sequencing would ever be detected by operators and maintenance crews. Detecting and measuring negative sequence currents is nevertheless a critical part of maintaining the health of an electrical system. However, this can hardly be done without the use specialized equipment such as a Power Quality Monitor. There are other devices that can be installed on a permanent basis, such as Negative Sequence Current Protection, Differential Protection or Voltage-Based Protection Relays, but on medium voltage networks it is doubtful that this would be done.

Actual Incident

Recently, I came across yet another incident where this happened. After analyzing the results obtained by running my Power Quality Monitor for a week, I noticed rather large discrepancies in the current in the three phases. At low load conditions, the currents appear to have balanced out, but as the load increases, the current in the one phase increased significantly with an increase in the current in the neutral. This bothered me and I started analyzing the data and found that the percentage unbalance also shot up. After further investigation, I found that the voltage on one of the phases became almost negligible.

See how the voltages of two of the phases (Phase 1 & 3) shot up during a swell incident and how Phase 2 (Yellow) became almost negligible as shown in the picture below.

This is supposed to be a 6.6kV feeder with the grey dotted lines depicting the range. But during this incident that lasted 244ms, the voltages of Phase 1 and 3 increased to above 10kV while Phase 3 almost reached 0V. The increase of Phase 1 & 3 was over 51% increase above the rated voltage of the network.

This now requires further investigation to determine the cause of the voltage discrepancy.

Follow-Up Investigations

Whether an investigation will be done and when, is unpredictable.

This work was done for a consultant to determine whether additional load can be added without any upgrading of the network. It was done at a medium-voltage municipal substation with the knowledge and consent of the municipal engineering department. As a courtesy, the results were sent to the municipal engineering department by the consultant, and we must now wait for their reaction.

Agulhas Consulting Services

Kindly click on this page to learn more about Load Profiling specialist service we are offering.

Newsletter

Join the mailing list and receive our latest newsletters by simply filling in your email address below and then click Subscribe.