• Asset Management
• Insulation Design & Materials
• Pollution Assessment & Remediation
• Lightning & Surge Protection
• Laboratory Testing
• Condition Monitoring

This time next year, T&D engineers and professionals from across the globe will gather in the beautiful Algarve region of southern Portugal for what promises to be the best INMR WORLD CONGRESS ever.

A curated team of experts in insulation design, surge protection, condition monitoring and laboratory testing will deliver more than 150 technical presentations covering practical topics and real-world solutions that help ensure reliable network performance.

Many events in the industry these days feature CEOs who are invited to give their views on shaping power grids of the future.

Our event takes a different approach. It focuses instead on the practical problems and challenges faced by power grids today. These include:

1. Optimizing insulation design of new overhead lines and substations;
2. Condition monitoring to prevent failures;
3. Improving pollution performance in harsh environments;
4. Reducing lightning-related outages;
5. Making systems more resilient in the face of climate change; and
6. Ensuring that critical network components are correctly manufactured and will perform as required.

The 2027 INMR WORLD CONGRESS will see engineers talking to other engineers to share knowledge, research findings and service experience. That, we believe, offers maximum benefits for attendees.

To support the many technical lectures delivered during the conference, there will also be a PRODUCT & TECHNOLOGY EXHIBITION featuring 80 world-leading suppliers. There, you will be able to see latest product designs, discuss your needs and build valuable contacts with experts who can one day help support you in your professional challenges.

We invite you to view the impressive array of speakers as well as their highly practical topics by visiting: www.inmrworldcongress.com.

There, you will also be able to learn more about this unique skills enrichment and networking event and register to attend.

Thank you and hoping to see you next year in Portugal!

Marvin Zimmerman
Congress Organizer & Chair

Florent Giraudet
Conference Director

The post Only One Year Until the 2027 INMR WORLD CONGRESS appeared first on INMR.

This edited contribution to INMR by Alfonso Ambrosone and Christopher O’Donnell at GE Vernova highlights the critical role of advanced online monitoring systems that enable real-time assessment of bushing health, facilitating early detection of emerging faults and supporting condition-based maintenance (CBM) strategies. Online bushing monitoring systems provide continuous, real-time diagnostics that significantly improve visibility into a transformer’s internal state.

By utilizing sensors that track essential parameters such as capacitance and power factor (tan delta), these systems offer early detection of insulation degradation, moisture ingress, and thermal anomalies issues that typically evolve gradually and evade detection through periodic assessments. This proactive monitoring approach results in enhanced reliability, reduced unplanned outages, and more accurate performance benchmarking across the transformer fleet.

Three case studies are also discussed which show how trending of online data helped trigger alarms and then lead to actionable offline test data for pro-active bushing replacement. This enhanced visibility enables better integration of operational information, empowering Operations Managers to make more informed decisions and significantly reduce the risk of major failures.

Hitachi Energy Transformer Components and Service

Switzerland

RHM International

USA

See more suppliers of Bushings

Key Failure Mechanisms of Condenser Bushings

Oil-Impregnated Paper (OIP) condenser-type bushings are widely used in high-voltage transformers due to their excellent dielectric properties and long service record. They are composed of a central conductor wound with alternating layers of paper insulation and conductive foil, forming condenser or capacitive layers. These layers are housed within a protective weather-resistant casing, typically porcelain, and impregnated with insulating oil to enhance dielectric strength. Within the bushing, two primary capacitances exist, designated as C1 and C2 (see Fig. 1).

C1 represents the total capacitance between the central conductor and the test tap, which is connected to the outermost capacitive layer and serves as the primary measurement point during diagnostic testing. C2 denotes the capacitance from the test tap to ground and is not part of the active circuit during normal bushing operation.

The capacitive layers are engineered to ensure uniform voltage distribution across the bushing by functioning as a voltage divider.

Despite their robustness, OIP bushings are susceptible to certain failure modes typically associated with moisture ingress, oil degradation and seal failure, which can compromise reliability and lead to catastrophic equipment failure if undetected:

• Moisture Ingress
Degraded seals or malfunctioning breathers can permit ambient moisture to enter the bushing, leading to increased dielectric losses and a rise in the dissipation factor (tan delta). Moisture ingress is one of the primary contributors to accelerated aging in OIP bushings, as it significantly diminishes the insulating properties of the oil-impregnated paper system.

• Oil Degradation
Over time, thermal aging and oxidative processes degrade the insulating oil, resulting in the formation of acids and sludge. These byproducts reduce the oil’s dielectric strength and can create conductive paths, increasing the risk of electrical breakdown.

• Partial Discharge (PD)
The presence of moisture, voids, or contaminants within the insulation system can initiate partial discharge activity. PD is characterized by localized electrical discharges that do not immediately bridge the insulation but gradually erode it, potentially leading to complete dielectric failure if left unaddressed.

• Thermal Runaway
Elevated dielectric losses—often triggered by moisture ingress—can lead to localized heating within the bushing. This internal heat accelerates insulation degradation in a self-reinforcing cycle that may ultimately result in thermal runaway and catastrophic failure.

• Mechanical Cracking
External mechanical stress, environmental exposure, or manufacturing defects can cause cracks in the porcelain housing. Such damage can result in oil leakage and expose the internal insulation to air and moisture, further compromising the bushing’s electrical and mechanical integrity.

Online Bushing Monitoring

Offline or time-based testing programs for bushings provide only periodic snapshots of their condition and thus offer no assurance that faults or catastrophic failures won’t occur between testing intervals. This inherent limitation calls for a re-evaluation of reliance on traditional time-based diagnostic testing as the sole method for assessing bushing health.

On-line bushing monitoring in power transformers based on continuous measurement and phasor analysis of test tap leakage current has been around for 50 years. In recent years due to an increased in substation monitoring online bushing monitoring is slowly increasing. This real-time approach is based on the principle of capturing and tracking high-resolution magnitude and phase data of leakage currents, enables real-time monitoring of C1% and Relative Power Factor changes. This methodology can be made more accurate by having an external voltage reference [3].

The types of bushing fault conditions that can be detected include but are not limited to grading layer failure (C1% increase), moisture ingress (PF% increase), and the results of internal discharge (affecting both C1% & PF%). These electrical characteristics serve as sensitive indicators of changes within the bushing’s dielectric system, enabling timely identification of deviations from nominal performance. By maintaining constant surveillance, online monitoring systems can effectively capture subtle shifts that may indicate the onset of degradation.

Measurements

The online bushing monitoring system involves the addition of high-resolution and high-precision ammeters at each bushing’s test tap, positioned between the pin and the earthed flange of the bushing. These ammeters measure currents that are directly proportional to the line voltage at the bushing, thereby providing crucial data on the condition of the insulation. The further chapters describe one of the standard methods of bushing monitoring and the measurements possible by using the total leakage current, measured at the test tap.

Fig. 1 shows a typical condenser bushing C1 electrical structure that has a physical layer for C2 (controlled C2) by design of the bushing manufacturer. In this design the bushing monitor would be connected to the voltage tap and the key electrical parameters monitored are capacitance (C1) and dissipation factor (tan δ), both of which provide insight into the condition of the insulation system.

Dissipation Factor – DP (Tan Delta)

Dissipation factor, also known as tan delta, is one of the key parameters used to assess the health of insulation. This dimensionless quantity represents the phase angle difference between the applied voltage and resulting current in an insulating material, offering insight into the dielectric losses. Ideally, in a perfect insulator, the current leads the voltage by 90°, implying no power loss. However, real-world insulators are not perfect and exhibit small amounts of leakage and dielectric loss, so that the capacitive current ID, resistive current IR, and total leakage currents can be represented in a phasor triangle.

When measuring the current at the bushing test tap, the total leakage current is acquired, which is the vector sum (hypotenuse) of the capacitive (displacement) and resistive leakage currents “(1)”.

where:
– I_R: resistive current (in phase with voltage)
– I_D: capacitive current (leads voltage by 90°)
– I_T: total leakage current – the vector (phasor) sum (measured at the test tap)

Extracting the resistive component (I_R) from the total measured IT in online bushing monitoring systems is technically challenging, since it typically constitutes less than 1% of the total leakage current. To measure the bushing’s absolute energy dissipation, the main lead voltage phase would be required, which is not measured in the approach described in this paper. However, the phase between two test tap currents can be measured between phases A and B and A and C.

The measured current inter-phase angles at the test taps provide information on the relative change of δ; if both δA and δB increase by the same amount, the relative change will not alter. The relative change of each δ (A, B & C) can be calculated from the measured angles in an arbitrary frame of reference.

In addition to the challenges stated above, it must be considered that ambient and operating temperatures will affect the readings by causing fluctuations, especially of the resistive component. Therefore, for accurate online tan delta monitoring of bushings, temperature compensation algorithms may be applied, referencing a baseline at 20°C, which helps ensure reliable analysis.

Change in Bushing Capacitance

Dielectric strength of a bushing is defined by the C1 capacitive layers, consisting of paper and foil meticulously wrapped around the central bolt conductor. Integrity of these layers is crucial for maintaining proper insulation properties. If a foil layer fails, the C1 capacitance would increase, leading to a corresponding rise in leakage current at the bushing test tap. This increase signals a potential degradation in the bushing’s dielectric properties.

Continuous bushing monitoring, based on the measurement of the bushing leakage current at the test tap, enables the detection of changes in the capacitive layers. The amplitude measured at the bushing test tap can be approximated to that of the displacement current amplitude. Thus, for a given nominal operational voltage in the main lead of the bushing, the Capacitance C1 can be calculated using the calculated formula “(2)”.

where:
– I^D₀: capacitive displacement current (ideal behaviour)
– f: nominal operational frequency
– C: nominal capacitance
– V₀: nominal grid operational voltage

Since the I_D is directly proportional to the line voltage, small fluctuations in system voltage will cause proportional changes in the leakage current. Additionally, the capacitance values might slightly change with temperature, especially in oil-impregnated paper bushings, which adds fluctuations to the measurement.

Alternatively, the expected test tap current amplitude can be calculated using the operational voltage, nominal frequency, and the bushing’s nominal C1 again using “(2)”. The percentual measured current amplitude difference relative to the expected value can then be easily calculated. For a given nominal voltage V₀ and constant frequency f, this is also the C1 percentage change relative to its nominal value, allowing for the relative change of bushing capacitance to be easily calculated “(3)”.

Polar Plot

The polar plot is a 2D representation of a 3D set of variables using a nonorthogonal basis defined by three normalized vectors at 0°, 120° and 240°; i.e. it is the vector sum of three vectors of length 1 pointing at 0° (A), 120° (B) and 240° (C) each multiplied by its associated parametric value.

Thus, variations which are common to the parameter set of all 3 bushing, like temperature fluctuations or power grid irregularities described in the previous chapters, are geometrically removed when plotted in the polar plot. If, however, one of the three parameters’ changes respect to the others, this will show in the Polar Plot and the angle of the vector sum will indicate which value changed respect to the others.

The polar plot approach is a highly effective tool for online bushing monitoring, offering a clear, relative, and intuitive view of the condition of 3-phase bushings by tracking phase angle and current vector changes over time. It is less sensitive to load and voltage changes, since it’s based on relative deviations, which are often more reliable, rather than looking at absolute values. At same time it provides clear visualization of imbalance, indicating the affected phase and provide good tracking as the affected bushing’s vector moves outward (increased magnitude) and shifts in phase clearly visible on the plot.

Field Experience / Case Studies

The following case studies illustrate continuous, dynamic assessment of bushing health while the transformer remains energized.

Case 1: Bushing Power Factor Change

This case illustrates progressive insulation degradation, accurately detected by the online bushing monitoring and clearly visualized using the Polar Plot.

This online bushing monitoring system was monitoring a 50MVA Generation Transformer at a Solar Power Plant. In early 2019, online monitoring data revealed within a period of one month an increasing trend in the relative dissipation factor (tan delta) for one of the bushings on a three-phase transformer. This significant change of >1000% (relative value!) was visualized through both a time-domain trend and a polar plot, signaling the need for immediate investigation.

The rising trend in relative dissipation factor for Phase B, LV-side indicated increasing dielectric losses, most likely due to insulation aging, contamination, or moisture ingress. The change was significant compared to the other phases, ruling out system-wide environmental factors and confirming a localized issue. The polar plot confirmed this diagnosis visually, with Phase B clearly deviating from the cluster, reinforcing that the issue was confined to that specific bushing as shown in Fig. 4.

The line chart shows the evolution of the relative dissipation factor over time for all three phases (A, B, and C). Phase B (blue line) exhibits a significant and steady increase in dissipation factor beginning mid-February 2019, with sharp spikes peaking in March as per Fig. 5.

This case demonstrates the effectiveness of online dissipation factor monitoring as a predictive maintenance tool. By detecting subtle but critical changes in insulation condition, the utility was able to take pre-emptive action—replacing the degraded bushing before failure and thereby avoiding unplanned downtime and potential equipment damage.

Case 2: Relative Change in Bushing Capacitance

The following shows a bushing ‘save’ detected by online bushing monitoring and clearly visualized on the Polar Plot. The online monitoring system detected a sudden change in C1 capacitance of one of the bushings in the primary side of a 220kV Generation Transformer.

Fig. 6 shows the time-series chart of the total leakage current IT with stable baseline values across all three phases until beginning on November, when the measured current of phase B (green line) showed significant increase.

Fig. 7 shows the clear indication of the phase B bushing capacitance change by more than 10% without further necessity for the operator to analyze trend lines. Subsequently a controlled shutdown was scheduled to inspect the transformer and the suspect bushing.

Offline testing at 10 kV confirmed abnormal capacitance levels for HV B phase bushing with a C1% increase of 17% from the FAT, respectively 9% change from the latest measurement during commissioning of the monitoring system, as shown on the Table 1.

Thanks to real-time monitoring and clear visualization of the capacitance trend, the utility avoided a potentially catastrophic failure. There was no unexpected service interruption, and the transformer remained in operation for an additional year before planned corrective actions, including a scheduled outage and bushing replacement program, were carried out.

Case 3: Bushing Power Factor Change

This case study examines the early stages of power factor degradation in a 364 kV substation transformer. The high voltage B phase bushing exhibits a noticeable change in relative dissipation factor (tan delta). Fig. 8 illustrates the increase in relative power factor from 300% to 1200%.

The dependence of temperature is a critical factor in assessing changes in the relative power factor of bushings during online analysis. As the insulation of bushings degrades from its original factory condition due to aging, moisture, or other influences, there is often a noticeable correlation between rising temperatures and an increase in relative power factor.

In Fig. 9 the trend lines of the measured inter-phase angles indicate that AB measured value exhibits a stronger temperature dependency in relation to the top oil temperature (transformer load). This is clear that AB inter-phase angle trend line shows increased separation from other at elevated temperatures, and this behavior has been linked to B phase.

GCC Electrical Testing Laboratory

Kingdom of Saudi Arabia

TECNALIA Electrical Labs

Spain

See more Laboratories

Conclusions

Bushing faults are inherently unpredictable unless subjected to regular monitoring, to mitigate this issue the implementation of online monitoring systems provides a proactive and continuous solution. These systems facilitate real-time observation of insulation performance without necessitating equipment shutdowns or manual interventions.

A notable advantage of online monitoring is its capability to identify slowly developing faults, which traditional periodic inspections often overlook. Issues such as gradual moisture ingress or localized overheating tend to evolve incrementally, potentially eluding detection until they reach advanced stages. Continuous monitoring enables the early identification of such trends, thereby supporting predictive maintenance strategies and minimizing the risk of catastrophic failures. As a result, online bushing monitoring signifies a critical advancement in enhancing the resilience and intelligence of asset management practices for high-voltage equipment.

Moreover, when integrated into substation SCADA or asset management platforms, online monitoring plays a pivotal role in digital substation and smart grid initiatives. It bolsters reliability centered maintenance (RCM) efforts and aids reducing unplanned outages.

Bibliography
[1] Kane, C ; Golubev, A (2005) On Line Monitoring of bushing on large power transformers
[2] Cigre W.G A2.43 (2019) Transformer bushing reliability
[3] Picher, P ; Rajotte, C (2007) Cigre Field Experience with on-line Bushing Diagnostics to improve Transformer
[4] W. A2.34, TB445, “Guide for transformer maintenance,” Cigre, 2011
[5] IEEE Std C57.143-2012, IEEE Guide for Application for Monitoring Equipment to Liquid-Immersed Transformers and Components, IEEE Power & Energy Society, 2012
[6] K. Elmer, J. Lapworth, and M. Ryder, “Transformer bushing monitoring: Implementation and benefits,” in Proceedings of the IEEE International Conference on Condition Monitoring and Diagnosis (CMD), 2008
[7] R. Hartingsveld and P. van der Wielen, “Transformer Bushing Monitoring by Power Factor and Capacitance,” in TechCon North America, 2009
[8] H. Borsi and E. Gockenbach, High Voltage Engineering: Testing and Diagnosis, Springer, 2010
[9] J. D. Belanger and J. Jalbert, “On-line monitoring of high-voltage equipment: Practical implementation and case studies,” IEEE Transactions on Power Delivery, vol. 20, no. 1, pp. 425–432, Jan. 2005
[10] A. Behrendt, B. Harnisch, and M. Krüger, “Modern techniques for continuous monitoring of bushings and tap changers,” in CIGRÉ Session, Paper A2-104, Paris, 2006
[12] R. Brown, Electric Power Distribution Reliability, CRC Press, 2008
[13] Dr. D. Robalino, “Accurate Temperature Correction of Dissipation Factor Data for Oil-Impregnated Paper Insulation Bushings: Field Experience,” 2011

The post Life Management for Transformer Bushings: Strategies & Benefits appeared first on INMR.

In this edited past contribution to INMR, Prof. Guan Zhicheng (retired) from Tsinghua University Campus in Shenzhen, China explains that lightning induced overvoltages fall into one of two broad categories based on the physical processes involved: those caused by direct strike to towers, shield wire or conductor; and those where lightning strikes the ground near lines, generating an induced overvoltage on conductors. Experience demonstrates that the former imposes a far greater hazard and that the latter most threatens lines of 35 kV or less.

Based on where a line is hit, overvoltages from direct lightning strike are of two types. In the first case, lightning strikes a point on the tower or shield wire and the resulting current causes its potential to ground to rise significantly. When this difference in potential between lightning strike point and conductor exceeds the lightning impulse discharge voltage of the line’s insulation, flashover or breakdown of the air gap occurs. By contrast, should the absolute value of the tower or shield wire potential at the moment of lightning strike be higher than that of the conductor, there is back flashover.

In the second case, overvoltages generated by lightning bypassing the shield wire and striking the conductor are shielding failures.

The performance of a system’s lightning protection can be measured using two indices – lightning withstand level and lightning related tripping rate. The former refers to the maximum amplitude of lightning current (in kA) that the line can withstand without leading to insulation flashover under direct lightning strike. The higher this is, the better will be a line’s lightning performance. Lightning related tripping rate refers to the number of trips on a line (per 100km/year) under standard conditions or after being converted to the equivalent of 40 lightning days per year. Lightning tripping rate is therefore a comprehensive index of a line’s lightning protection performance.

Since shield wire is often installed along the length of a line of 110 kV and higher, it has been widely assumed that lightning related tripping on such lines was due mainly to backflashover while incidence of shielding failure was lower. But data from several countries demonstrate the opposite.

For example, past lightning related tripping rates on Chinese power lines have shown that most were due to shielding failure and much less to backflashover or other causes. Moreover, tripping of 750 kV and ±500 kV lines were virtually all due to shielding failure. Indeed, comparative rates of shielding failure as a proportion of all lightning tripping on 66 kV, 110 kV, 229 kV, 330 kV and 500 kV lines were such that the higher the voltage, the higher the percentage of shielding failure related tripping.

Measures to increase the lightning withstand level of a transmission line and reduce lightning related hazards include: decreasing grounding resistance of towers; adjusting protection angle of the shield wire; using a lightning protection differentiation design for double circuit towers; installing additional lightning rods or line surge arresters (TLSAs) in areas with highest risk. Moreover, using lightning monitoring equipment helps in locating and replacing damaged insulators in a timely manner.

Dealing with backflashover has a distinctly different focus than for shielding failures. Decreasing ground resistance of towers is highly effective in preventing backflash related tripping; the main measures to prevent shielding failure are decreasing a wire’s protection angle or installing more lightning rods and line arresters.

The lightning withstand levels of transmission lines in the event of shielding failure are low. Calculated lightning shielding failure withstand levels for 110 kV, 220 kV, 330 kV and 500 kV transmission lines are only 7 kA, 12 kA, 16 kA and 27 kA respectively.

In places such as China, for example, the probability of lightning current over 20 kA is 59% and that over 50 kA is 27%. As such, occurrences of shielding failure will typically lead to line tripping and therefore demand that more attention be given to preventing them, especially on higher voltage lines.

Line arresters can significantly reduce lightning related tripping and their practical application has proven remarkably effective. If their cost can be further reduced, they would probably be used far more widely. At the same time, UHV AC lines have an additional shield wire installed above the center phase conductor on critical sections to increase withstand performance and reduce lightning shielding failures.

It is also important to increase the lightning withstand level of insulators. Since composite insulators offer superior withstand to pollution flashover, their insulation distances are usually shorter than for equivalent porcelain or glass strings. Moreover, a composite insulator’s grading ring’s lightning withstand level may be relatively low. Given that the lightning withstand level of 110 kV lines is already low, the net negative impact on lightning performance when composite insulators are specified can be significant.

Power companies have already noticed this and their specification of composite insulator lengths is no longer based solely on pollution flashover performance but also takes into account effect on lightning withstand level. Still, as long as the air gap between the grading rings of composite insulators is not less than that on porcelain or glass strings, a line’s lightning withstand level will not decrease.

Finally, any discussion of lightning related hazards to lines must consider not only lightning related tripping rate but also failure rate due to lightning. If reclosing is successful on a line struck by lightning, no service interruption will result. Given this, installing parallel-connected protection gaps on insulator strings helps keep the arc from a lightning flashover sufficiently far from the surface to avoid interruption due to damaged insulators.

The post Lightning & Power Systems appeared first on INMR.

It has long been known that corona can lead to insulation failure. However not all aspects of the problem are fully understood and are still being researched, including magnitude and duration of corona to initiate degradation, best detection methods and development of suitable tests to predict performance in its presence.

When it comes to composite insulators, corona activity can originate from hardware, voids within the material or from interfacial defects. Most of the light produced by such corona has a wavelength shorter than 400 nm and therefore falls in the UV range. By contrast, most solar radiation is in the 400-700 nm visible range – the shorter wavelengths filtered by the earth’s ozone layer. In fact, some peaks in the UV region of the corona spectrum match or exceed those in the solar spectrum.

Polymeric materials are more susceptible to degradation from UV produced by corona than from solar radiation, particularly if the corona takes place close to the material. Corona ruptures stable oxygen molecules (O2) to create radicals that combine with the molecules to form ozone (O3). The ozone then attacks double and triple bond sites in elastomeric materials such as silicone rubber or EPDM. The result is cracking. Even tiny amounts of ozone in the ppm range are sufficient to initiate cracks, however the time required for this depends on material formulation.

Although most modern elastomers are stabilized against this threat, some eventually succumb to ozone attack should its concentration become sufficiently high. Corona also produces oxalic and nitric acids in the presence of surface moisture from humidity, dew or fog. Depending on pH, this can also locally degrade polymers.

Corona can even ‘drill’ holes in a material, suggesting that degradation is not solely due to chemical attack by ozone. In fact, researchers have calculated the temperature at the tip of the discharge and shown it to be high enough to cause ‘evaporation’ of even inorganic materials. There is also suggestion of mechanical attack, like sandblasting, due to the impact of repeated discharges on a material. It is indeed rare in power engineering that any one physical phenomenon can trigger so many possible modes of degradation.

To illustrate the potential impact of corona, past research involved ground-based inspection of composite insulators on 115 kV, 230 kV and 500 kV lines in the U.S. southwest using a corona camera and binoculars. Several 230 kV towers were equipped with composite insulators with and without corona rings on adjacent phases. The 115 kV insulators did not have corona rings whereas the 500 kV insulators had rings on the line as well as on the ground end hardware. Given the dry conditions of the service territory, the utility operating these lines had decided not to replace the 230 kV insulators not equipped with corona rings. This allowed comparative study of the effects of corona on composite insulators since all other factors were the same (i.e. same design, material, manufacturing details, location and system voltage). The corona observed on these insulators, many of which had been in service for over 25 years, was sporadic and originated from hardware.

The 230 kV insulators being evaluated comprised three generations of composite insulator technology. Some were removed for further examination and it was found that the shed closest to the line end without a corona ring exhibited minor to serious changes in the form of hardening, cracking and discoloration. It was clear that these insulators were approaching their end of useful life and indeed would have probably already failed in locations having more precipitation. By contrast, none of the insulators equipped with corona rings showed such degradation.

This demonstrated that when it comes to composite insulators, design and need for a corona ring depend not only on voltage level but also service environment. For example, in the case of insulators in environments with heavy contamination, high altitude or frequent wetting, it is prudent to have a line end corona ring even at lower transmission voltages.

The post Corona: Deadly Enemy for Polymeric Insulation appeared first on INMR.

This edited past contribution to INMR by T&D expert, Dr. William Chisholm, discussed alternative mitigations, which include re-positioning dampers, inspecting more frequently or selecting other installation configurations. It also offers a case study that illustrates how these factors were considered during installation of TLSAs on a 230 kV line as alternative to continuous counterpoise.

Protektel sp. z o.o.

Poland

PFISTERER

Germany

See more suppliers of Arresters

The first successful lightning surge arresters for electric power systems used hollow ceramic insulators, filled with silicon carbide blocks to provide non-linear voltage limiting electrical characteristics. The silicon carbide arresters had internal components to quench and interrupt AC power system follow current that normally keeps flowing through a flashover arc. The hollow ceramic insulators were sealed with metal plates, providing over-pressure protection but also allowing long-term moisture ingress that shortened service life. These arresters were used mainly for protection of equipment. Two separate technologies converged to improve the reliability and service life of arresters:

• The non-linear electrical performance of metallic oxide varistor (MOV) materials was superior to silicon carbide. This technology was patented by Matsuoka, working at Matsushita Electric Company, in 1970 and had displaced silicon carbide in many applications by 1980.
• Polymer rubber materials, including ethylene propylene diene monomer (EPDM) and long chains of polydimethylsiloxane (silicone), protected the MOV elements against moisture while reducing weight and cost, compared to ceramic housings. The industry wide substitution of polymer for porcelain took place over the brief period from 1987 to 1995.

These improvements made it practical to use transmission line surge arresters (TLSAs) to improve lightning performance by applying them at some or all insulator positions, as well as at equipment or line terminals. Electrically, TLSAs using MOV materials have been successfully installed to limit transient overvoltages on line insulators for more than 30 years. For example, successful application of TLSAs on an unshielded 230 kV transmission line since 2001 used arresters with 60 mm block diameter. Exper¬ience with improving MOV technology allowed reduction in block diameter from the initial 61-76 mm to only about 40 mm for shielded lines, with overhead ground wires (OHGW) that limited their energy duty by diverting a large proportion of current to ground. Further reduction in cost and increase in performance has been achieved using a direct mold silicone construction that eliminates internal air spaces. Originally developed as components to be retrofitted to existing lines, TLSAs have now become an important design and mitigation alternative to investment in extensive grounding systems for new line designs. One case study set out the decision criteria leading to fitting of TLSA to all three phases of a 230 kV transmission line during construction, as an alternative to using continuous counterpoise in an area where rock resistivity was extremely high.

Mechanically, however, TLSAs have demonstrated in-service problems, the most common being rapid degradation of flexible leads and their terminals. Once a lead breaks away or pulls out from the tower or the bottom of an arrester, its uncontrolled motion in the wind can reduce electrical clearances. Rigid configurations, including the external gap EGLA type, do away with flexible leads and offer longer service life.

Unfortunately, many TLSA installations have been made in ways that affect or even completely defeat the mechanical protection given by Aeolian vibration dampers. Aeolian vibration is a low amplitude high frequency vertical vibration that accumulates millions of stress cycles. This is particularly a concern for lines that operate over a wide temperature range, e.g. -40 to +40°C. At low temperature, high conductor tension especially needs effective vibration control. Conductor damage has been associated with installation of components that have the same weight and size as TLSAs, notably aircraft marker balls, anti-galloping detuning pendulums and spacer dampers. Galloping is large amplitude low frequency oscillations involving complex vertical, longitudinal and torsional motions that accumulate thousands of stress cycles ad are common when there is an ice layer on conductors. Rocking motion along the line direction is a concern, especially if a resonant frequency of the span is synchronized to the pendulum frequency of the arrester itself. An arrester installation that causes galloping is theoretically possible in particular span lengths but has not been reported in practice – as of yet. There is considerable experience with spacer damper systems for bundle conductors. Lessons learned from these problems can be applied to applications of TLSAs, thereby achieving long service life – especially if arrester standards expand to include mechanical issues during the procurement and installation process. In this respect, IEC 60099-8 provides an example of leadership by setting out some vibration requirements for EGLAs.

A CIGRE Task Force consisting of mostly mechanical engineers evaluated one specific problem of non-gapped line arrester (NGLA) interactions with vibration dampers and formulated recommendations. This work set out the following recommendations:

1. On retrofit applications of TLSAs, Stockbridge type dampers should be moved along the conductor so that the distance from the TLSA clamp to the Stockbridge damper is at the optimum distance, as it presumably was from the suspension insulator clamp;
2. TLSA clamps on conductors should be mounted over reinforcing armor rods;
3. Loose connecting wires should be long enough to accommodate all motions possible of suspension insulator strings, without pulling them tight;
4. Alternatively, clamps on loose connecting wires should be applied over a reinforcing layer of armor rods at both ends;
5. There appear no standards that relate to mechanical reliability of surge arresters. A survey of industry service experience of failures leading to realistic test requirements seems necessary;
6. Support hardware should provide freedom of motion of the surge arresters across and longitudinal to the conductor.

EPRI conducted studies of the stress on flexible leads from lateral motion of conductors and insulators. An important side-benefit of this work was ongoing documentation of installation errors at various utilities, e.g.:

1. Arresters that are fixed between tower and conductor, making an inverted braced-post configuration with insufficient tension and compression strength ratings to restrain conductor motion from suspension insulators;
2. Arresters installed with insufficient clearance to phase, considering the arrester fails as a short circuit to ground and brings line potential to the normally grounded end;
3. Explosive disconnects installed at the bottom of the arrester, thus allowing a long, energized lead to swing freely from the conductor;
4. Flexible leads that break part way between arrester and phase conductor, leaving a short stub that becomes a source of corona and electromagnetic interference;
5. Rigid leads that reduce electrical clearance and transfer cantilever forces to the arrester, which may not be rated for these loads;
6. Combinations of connection lead and strain relief chains that bypass the electrical isolation function of the explosive disconnect.

Vibration Damper Placement

Many utilities have installed TLSA bodies or leads onto phase conductors, adding the new components somewhere between insulator clamps and vibration dampers. But few have followed Recommendation #1, i.e. to move the Stockbridge dampers at the same time as arresters are installed and thus preserve the original damping design. In cases where TLSAs have been placed close to dampers, as in Fig. 2, this advice to move the dampers into the spans, away from the TLSA, should be implemented immediately. It would be helpful to inspect some of the conductor beneath the TLSA to establish the extent of damage to inner conductor layers after a few years of service without proper damping.

In cases where TLSAs have been placed about halfway between dampers and insulator clamps, the CIGRE analysis provides some breathing space for developing an action plan. Fig. 3 shows that vibration damper placement is a compromise, absorbing some energy at wind speeds of 1 and 7 m/s and a larger fraction of energy from winds having 3-5 m/s velocity.

Mechanical engineers recommend a vibration damper placement at 80% of the loop length at 7 m/s as a good overall choice, considering the usual distribution of wind speed over a year. When a heavy TLSA is placed at 40% of the loop length, halfway between insulator clamp and an existing damper, Fig. 4 shows that there is reduction in damper effectiveness. However, the original design may have included some extra margin in damper power dissipation at vibration frequencies associated with certain wind speeds.

The placement of the TLSAs in Fig. 2 is approximately 20% of the original distance from damper to suspension clamp. Thus, the x-axis value in Fig. 4 is 0.2 and damper effectiveness has been reduced to 25-40% of the original value. This is too great a reduction to ignore. The dampers should be re-positioned. Since field tests have confirmed that TLSAs act as vibration nodes when suspended from phase conductors, the cumulative damage to the conductor under the arrester clamps should also be evaluated by periodic inspection. Electrical engineers have mastered concepts of resonance with RLC circuits and the recent IEEE introductory guide to Aeolian vibration can help guide their move into the mechanical domain.

Reliable Conductor Clamping Systems

CIGRE Technical Brochures provide historical consolidation of utility experience. TB 277, for example, describes the following design characteristics for clamping of spacer systems to phase conductors:

A spacer clamp should be capable of easy, reliable installation, preferably verifiable by ground based inspection and should provide a safe, reliable, non-damaging, long-term grip on the sub-conductor. High, localized clamping stresses should be avoided and the clamp conductor groove should be smooth and free from irregularities. The clamp should be smoothly profiled to minimize Corona and RIV discharges at specified voltages. Component parts should be secured and an energy storage mechanism is required to prevent loosening due to the effect of vibration, thermal cycling conductor and elastomer creep (in the case of elastomer lined clamps). All materials should be compatible with the conductor and avoid corrosion.

The clamps should:

• Maintain a suitable grip for the whole life of the line and in the whole range of service temperatures;
• Avoid conductor damage during installation and service;
• Be free of corona and RIV at maximum line voltage;
• Be installed without disassembling any parts;
• Be designed to minimize risk of improper installation;
• Be capable of being safely removed and re-installed on conductors;
• Allow installation survey from the ground;
• Ensure that individual components will not become loose during service.

Substitution of ‘arrester’ for the highlighted ‘spacer’ is an appropriate goal for users and suppliers of NGLAs. Experience with rigid, articulated, spacer-damper and flexible spacers can guide selection of suitable TLSA hardware. Spacer clamps are generally made of primary aluminum alloy. Rubber-lined clamps used in many spacer-dampers may not be appropriate for NGLAs since a separate electrical connection would need to be made to allow the flow of lightning surge currents through the arrester’s MOV elements. However, a range of interesting articulation systems is in use for spacers and spacer dampers. Fig. 5 shows typical spacer clamping systems and components that could also be used in reliable NGLA installations. Safety plates with tabs that bend after tightening, or breakaway nuts that confirm correct installation torque are also mentioned as ways to avoid the conductor abrasion or hammering damage in Fig. 6.

Alternate Mounting Configurations

A range of mounting configurations for EGLAs and NGLAs has been described. For example, flexible leads, as shown in Fig. 7 fail rather quickly in service and can benefit from re-design that separates electrical and mechanical components. This leads to designs with higher parts count, weight and cost that should be validated with the same limit-cycle tests (10 cm, 106 cycles) as spacer systems.

There are several ways to address the limited durability of flexible leads, including:

• Specifications that require limit cycle testing, much like spacer dampers;
• Specifications that would apply an appropriate number of cycles and amplitude, considering the clamp mounting location and damper let-through frequency;
• Installation practices that place the disconnect-lead clamp near or on the insulator clamp;
• Use of stress relieving loops;
• Frequent and specific inspection.

Lead failures are easy to locate from helicopter or ground inspection, although it is difficult to establish whether the explosive disconnect has operated. It is possible that drone-based inspection will become cost effective and bring a closer focus on areas of interest near ring tongues and collars of flexible leads. As the leads break at the arrester or tower and fall, they may reduce electrical clearances. Electrically, the greatest benefit of partial TLSA application is often gained with installations on the bottom phases. Fig. 8 shows a mounting configuration with flexible lead that presents minimal danger to electrical clearance if it fails as designed and disconnects from the phase. However, if the lead breaks away from the top of the NGLA, nuisance flashovers can occur under side winds that blow the flexible high voltage connection towards the tower while it is still connected to the phase.

Recently, SaskPower in Canada completed construction of a shielded 230 kV line running parallel to an existing, unshielded 138 kV line for 300 km. Most of the access was by helicopter and this formed at least 10% of the total line cost of CA$ 330M. The lines run across a geographic feature known as the Canadian Shield, which is granite of high electrical resistivity. With no overhead ground wires, every lightning stroke to the I2P 138 kV line in Fig. 9 caused a flashover and short circuit across its polymeric insulators

The I2P 138 kV transmission line was constructed with continuous counterpoise, although it had no shield wires to provide a suitable return path for fault current that would allow rapid operation of protective relays. Early in the project plan for the new I1K line, grounding investigations were carried out to establish whether continuous counterpoise would improve back flashover performance since the new design was fitted with an optical fiber groundwire (OPGW) for improved communication between terminals. A ‘Zed Method’ applied a voltage step between the base of existing I2P towers and a test lead, laid on the rock. The surge impedance of the test lead provided the transient connection to ground giving 400 mA pulse current. The tower base voltage rise was monitored in response to this impressed current. In a few cases, measured surge impedance of the tower base, with its counterpoise in place, was almost the same as the impedance of the insulated wire, draped on the ground. More typically, the guyed structure had an impedance of 140  in isolation that fell to a constant 77  when the counterpoise was reconnected. The tests concluded that continuous counterpoise, draped across the Canadian Shield rock surface, did not provide much of a grounding path. Each leg of counterpoise had a surge impedance of about 360 , meaning that more than 10 radial wires would be needed to achieve a footing impedance of 35 . At that stage, many options were considered and refined down to two choices for the new I1K line:

1. Moving the counterpoise up and suspending it from the tower, in an under-built location below the phases, where it could be inspected by helicopter; or
2. Installing surge arresters across all three phases

The selected arresters in Fig. 11 met the IEC 60099-4 Edition 2.2 Line Discharge Class 3, with 58 mm MOV block diameter. In Edition 3 of this standard, the new classification is SM with Q¬rs¬ of 2.0 Coulomb, corresponding to Energy Class E of IEEE C62.11/2012. The arresters weigh about 60 kg each with additional mass from the corona ring. With the choice of arresters over grounding made early in the project plan, it was possible to advise the manufacturer of the vibration dampers that their design, the twin pairs of dumbbell weights at the left of Fig. 11 would also need to protect the conductor under the arrester clamp. Close inspection shows that armor rod has been placed between conductor and NGLA saddle clamp, further improving mechanical durability and hopefully leading to long service life.

Another consideration in application of polymeric-housed TLSAs is shown in Fig. 12 and in a region prone to forest fire. One test along the I2K line near Island Falls showed that the ground temperature after a recent forest fire was hot enough to melt the connection from tower to counterpoise. In this case, the counterpoise was found to be otherwise intact, since it provided a continuous reduction in footing impedance when re-connected to the tower. This kind of damage is difficult to inspect from helicopter and forms an additional argument for selecting TLSAs (or underbuilt ground wires) as an alternative to grounding with continuous counterpoise.

Improvements in Arrester Mechanical Standards

There are many standards that describe mechanical vibration test methods for transmission line components. Some important definitions are found in Figs. 13, 14 & 15.

GCC Electrical Testing Laboratory

Kingdom of Saudi Arabia

CATU Test Laboratory

FRANCE

See more Laboratories

Bundle spacer systems are tested with mechanical oscillations that simulate in-service conditions. For example, a galloping test applies a 1 Hz oscillation frequency with 10-20 cm amplitude for 106 cycles. Components pass the test if there are no visible changes. This would seem to be a reasonable requirement for complete arrester/flexible lead systems. However, galloping tests are severe. An adaptation of the vibration test in Fig. 15 was set up in Fig. 16 to simulate galloping effects on a 44 kV NGLA located about 710 mm from a suspension clamp. Galloping amplitude of about 0.8 m was achieved on a 20 m span. The purpose was to simulate damage found in the field and Fig. 17 shows that this was achieved.

The damage in Fig. 17 was partially mitigated if the clamp was tight and flush to the top of the arrester. The stud fatigue damage shown in Fig. 18 occurred after 20k to 66k cycles. This is consistent with findings about galloping endurance of Stockbridge dampers on galloping test lines.

Conclusions

Transmission line surge arresters are large enough and heavy enough to affect the performance of vibration dampers. Field tests have confirmed that TLSAs act as vibration nodes when suspended from phase conductors. CIGRE guidance recommends that existing vibration dampers should be displaced along the conductor, treating any conductor-mounted arrester as a new clamp and node to be protected. Damper re-positioning should be an inspection and remediation priority for installations where arresters have been placed close to pre-existing dampers. Concerns with long-term durability of the conductor should include re-selection of appropriate dampers for the new span as well as improved arrester attachment using armor rods or other stress reduction clamps. Guidance, including standards for limit-cycle testing, should be obtained from clamping systems that have proven suitable for bundle spacers.

TLSA accessories, including flexible lead clamps, corona rings for NGLAs and arc receivers for EGLAs, should meet appropriate vibration and flexion criteria for an anticipated 20 to 50 year service life. Placement of light accessories close to existing insulator clamps is recommended to reduce level of vibration. Since some accessories are about the same weight as vibration dampers, placement near existing dampers should be reviewed with manufacturers and supported by bending amplitude verification testing.

The post Achieving Long Service Life from Transmission Line Arresters appeared first on INMR.

This edited past contribution to INMR by industry expert, Paul Leufkens, assessed developments in methodologies as well as relevant standards and technologies when testing power cables and accessories.

PFISTERER

Germany

DALEKOVOD OSO

Croatia

See more suppliers of Cable Accessories

Generic MV & HV Cables & Accessories: Past Developments at IEC & IEEE

Over the past decades, the volume of power cables being produced has been increasing by 8 to 10% each year. While most of these new cables are for distribution networks, annual growth in application of HV cables has recently been in the double digits. Key issues driving this trend are growth in the renewables sector and demand from within urban areas where, due to land limitations, cables are being installed in preference to overhead lines. Apart from volume, lengths of cable systems are also increasing and HV networks that cover hundreds of kilometers are no longer rare. Such longer lengths create challenges not only for the factory but also when it comes to on-site testing. Lengths of distribution networks are increasing as well, often covering 15 to 20 km.

In regard to applied insulation material, there is now a clear trend to use of only polymer insulated cables. Polymeric materials such as polypropylene (PP) have been developed and are being used for HVDC cables. New materials present new challenges, especially in terms of diagnostic assessment, such as dielectric losses.

Modern day production processes for power cables and their accessories are well-controlled and one of the main product philosophies is being ‘fit for purpose’. For example, insulation thickness is kept to the minimum necessary to reduce costs. But this approach comes with the downside that almost every workmanship error will eventually lead to a system defect.

Within the standards and guidelines there has not been a lot of change over the past decade and these continued to be based mainly on withstand testing only. Although diagnostics are recommended, no threshold values are mentioned. Within CIGRE, several working group activities have been undertaken both for MV and HV cables. A new brochure, TB841, has been published that discusses alternative methodologies for after-laying testing of HV and EHV cables. Although the brochure was not the original goal, it nevertheless provides a valuable overview of which methods can be applied. For MV networks, an active Working Group on MV cable asset management (B1.58) closed in 2022. This WG aimed to propose asset strategies to be applied on several type of cables (i.e. polymer, paper, mixed) for commissioning, after-repair and maintenance testing. All test and diagnostic methods, i.e. the complete toolbox, were to be described along with a pros and cons analysis of each, apart from a section dedicated to asset management.

MV Off-Line Site Testing, Diagnostics & Monitoring

Underground MV power cables have become an important component of modern distribution networks. Experience from network operators has shown that localized degradation, cable joints and terminations often constitute the ‘weak spots’ in these systems, apart from factors such as ageing, water ingress, inadequate workmanship and utilization of inappropriate or defective accessories. Hence the need to deploy testing and diagnostic techniques on both new and serviced-aged MV cable systems to ensure highest possible reliability, as outlined in guidelines and standards such as the IEEE 400 series and IEC 60502-2.

The standards today are written mainly for cable withstand testing. While there is some guidance on diagnostic methods for TanDelta and partial discharges (PD), this is either limited or outdated. Looking to the IEEE 400.2 TanDelta tables, for example, one can question: ‘will the cables from 2015 still have the same recommended trending limits as cables installed in the 1980s and 90s? Also, looking to PD diagnostics on paper-insulated cables, ‘can any threshold values be defined since this type of cable is not designed to be PD-free?’.

Withstand tests typically require application of continuous voltage at a prescribed voltage level (normally above nominal operating voltage) for a prescribed time. The cable system under test will either ‘Pass’ or ‘Fail’. If the cable system fails, i.e. breaks down during testing, it will either be repaired or replaced. A withstand test can be classified either as a simple withstand test or a monitored withstand test. The latter differs from a standard withstand test in that other parameters are also monitored e.g. dielectric losses, leakage current or partial discharges. Fig. 1 shows a generic schematic of a monitored withstand test (MWT).

As a basic principle, a simple or monitored withstand test is designed to consistently overstress a cable system to some acceptable risk level. As such, for a withstand test to be effective, it must include the following fundamental elements: defined voltage exposure, repeatable voltage exposure and a well-defined failure rate.

The most commonly used excitation voltages are VLF (very low frequency), either sinusoidal or cosine-rectangular, followed by DAC (damped AC). In some cases, power frequency or near power frequency is used and, in rare cases, a soak test is also performed. Use of DC voltage on extruded cable systems for withstand testing has been discouraged by IEEE 400.1, ICEA S-94-649-2004 and IEC 60502-2.

Withstand testing for medium voltage cable systems is performed mainly in the context of commissioning/acceptance/electrical testing after installation of new cable systems. The goal of a withstand test during commissioning is to ensure the cable system is safe to energize and for quality control purposes. The impact of cable failure during testing is considerably lower and repairs can be made relatively quickly and cost effectively compared to in-service failures and resulting unplanned (emergency) outages and repairs. Additionally, a withstand test can also be performed on aged cable systems.

It should be noted that in the case of aged cable systems monitored withstand testing is recommended above simple withstand testing. Moreover, some industry guidance/standards recommend reduced test voltages relative to cable commissioning contexts. The decision to conduct withstand testing should always weigh the cost of failure during testing versus the cost of in-service failure and subsequent unplanned outages and repairs. The main disadvantage for withstand testing of aged cable systems is that there is a risk of inducing failure during the test, especially for cables that are already degraded. This could cause additional complications in aged cable systems due to issues such as lack of spare components, difficult accessibility for repairs, outage availability, etc. Additionally, in cable systems that are already degraded, residual life could be reduced by applying withstand testing with a voltage higher than nominal voltage.

Generally, it can be stated that simple withstand testing will give a certain degree of confidence when putting a cable into operation. However, not all defects can be detected during a simple withstand, no matter what excitation voltage is used. Applying diagnostic methods as in a monitored withstand or diagnostic context helps ensure higher reliability of the cable system.

Looking to diagnostic methods on MV cables, these can differ in core and context specific methods. Core methods are VLF tan delta, TDR (time domain reflectometer), PD testing and jacket testing. These methods are suitable for almost every cable type. Despite lack of consensus on acceptance criteria for most of these, such methods are well known and documented. Modern devices are easy to operate and the core methods should be considered as the basic toolbox for cable testing and diagnostics, providing the user with a fast and reliable result.

The purposes of context specific methods are multiple and these can be recommended during a commissioning test for baselining purposes, or at higher level of diagnostics or troubleshooting to increase accuracy of the measurement. Typically, these context specific measurements are applied on critical system infrastructure such as in nuclear power plants. These methods should be considered an ‘added value’ to the core methods. Due to their specificity, some may require more experienced operators in order to obtain exploitable results. Context specific methods include FDR (frequency domain reflectometry), neutral resistance measurements, FDS and TDDS (frequency domain spectroscopy and time domain dielectric spectroscopy), insulation resistance measurements, PDC (polarization, depolarization current measurement) and conductor resistance measurements.

Two main core methods being widely applied are VLF TanDelta and PD testing. VLF TanDelta measurements were originally intended to detect water trees in degraded polymer insulated cables. However, as cables being installed have become more resistant to water tree growth, especially long bridging water trees, the purpose of VLF TanDelta measurement has been changing as well. Experience gathered over the past decades is moving from a global measurement to a local measurement. Comparing phases with one another makes it possible to identify local issues, however these are not localized. Still, this gives operators important information about the reliability of their networks and is extending application of VLF TanDelta measurements to mixed cables systems.

PD testing on MV cables is broadly performed (mainly off-line). A proper test strategy should be applied to:

• Maximize the probability of identifying PD inducing defects (reducing risk of false negative results);

• Discriminate such defects internally and from noise (reducing risk of false negative and false positive results);

• Localize identified defects (allowing for targeted repairs);

• Allow for condition assessment that can allow for pass/fail assessment or potentially deferred investment, ideally through relating results to in-service (operating) conditions.

A wide range of offerings exist among PD equipment vendors and service providers, ranging from ‘black-box’ type approaches with proprietary algorithms only used by service providers, to fully customizable and adaptable systems requiring high levels of expertise, to ‘push-button’ type boxes which in principle can carry out measurements for unskilled users.

To obtain a better understanding of how to deal with all such options and of their impact on accuracy and performance of field PD assessment, a good starting point is to understand the basic technical considerations that most influence the outcomes of a PD measurement.

The most commonly used systems for MV networks are single-ended PD measurement systems and use of either DAC, VLF CR or VLF sinusoidal excitation voltage. There is lack on guidance when looking to trending limits, especially for aged cable systems. Apart from user experience and recommendations, there is no standard. In addition, as for the differences between test voltage sources, there will likely never be general guidance on acceptance levels.

The numbers of monitored MV cable systems currently is still low. Demand is rising but, apart from the issue of cost, there are also some technological hurdles, as reviewed in Table 1. With on-line PD methods, the chance of false positives or false negatives is still high, which is not acceptable to most end-users.

Finally, with all measurements that can be performed for both withstand testing, diagnostic testing or monitoring, proper data handling is also of importance. The need for an asset management system/software is constantly increasing, not only in terms of handling measurement data but also in regard to such data as age, location, number of faults, etc. In the past, diagnostic measurement was limited and could be treated differently. Now, with the current broad use of diagnostics methods by utilities, only by applying proper asset management, guided by software, will it become possible to increase the availability and reliability of power cable networks.

HV Cables: On-Site Testing & Diagnostics

Since reliable energy transmission and distribution are fundamental for every economy, the various aspects of maintaining quality control for newly installed as well as in service power cables is of great importance. Key questions for power utilities in regard to maintaining/updating internal procedures for reliable cable network operation include:

1. How best to perform, in a sensitive yet non-destructive manner, detection of poor workmanship defects in newly installed HV cable circuits?

2.How best to perform non-destructive diagnostics of HV cable circuits in service to determine their actual condition?

Over the past 30 years of testing distribution and transmission power cables, different test methods have been introduced and are currently in use, including: continuous AC resonant (ACR), damped AC (DAC) and very low frequency (VLF). From the start of offshore wind farm installations, such onshore methods have also been applied for testing offshore power cables. Unfortunately, considering the greater need for quality control for offshore HV cables and for long onshore HV cables, all these methods can present challenges, as illustrated in Table 2. Different studies during the last 20 years have shown that regarding voltage stresses, partial discharge occurrence and dissipation factor measurement, differences can exist between the ACR and DAC methods. Nonetheless, when used by experts, both are a valuable tool for assessment. It has been documented in recent publications that when testing transmission power cables, DAC after-laying and diagnostic testing allows sensitive PD detection. Moreover, using DAC calibrated PD detection is possible on-site, with background noise below a few tens of pC.

IEC standards such as 60840, 62067 and 63026 place emphasis on manufacturing and therefore most of their content deals with testing aspects in the factory. An up-to-date description of methods for on-site testing of high voltage cables has long been necessary but is taking place only slowly, especially on alternative on-site test methods for ACR. Moreover, most standards provide no up-to-date guidelines regarding field experience in after-laying and maintenance testing of cable circuits. CIGRE’s TB 841 is not standard but rather a recommendation and provides recent methods and experience. Traditionally recommended over-voltage testing with a binary test outcome, i.e. ‘breakdown’ or ‘no breakdown’, will reveal major defects. But PD detection should be added when possible and a check should always be made for the risk of igniting faults that would not have occurred under operating voltages.

Some TSOs are in the process of developing or have already developed their own more dedicated procedures to test newly installed cables. These procedures are based on service experience and selected IEC documents as well as on existing international standards made in cooperation between power companies, cable manufacturers and testing organizations. These guidelines have come to represent the current state-of-the-art for non-destructive methods for both after-installation and maintenance testing and diagnostics.

Introduction of on-site PD monitored voltage withstand testing of power cables has opened the possibility of reproducible conditions for PD detection. Based on research and about 20 years of testing of power cables up to 230 kV at numerous different 3^rd parties offering after-laying testing, the following parameters could be recommended as general criteria for DAC and ACRT testing to determine if a cable has passed the test and is sound for operation (see Table 3).

Submarine Factory & On-Line Cable Testing

Submarine cable systems differ in several respects from underground cable systems. Among the more notable differences are manufacturing and delivery lengths being much longer as well as use of factory-made joints. Manufacturing lengths typically depend on the maximum time an extrusion line can run before it must be cleaned. This converts to maximum length through the relationship between volume that can be extruded per unit of time. Volume of insulation that can be extruded per unit of time relates mainly to voltage class and cable cross-section, i.e. it generally takes more time to extrude a kilometer of cable of high voltage class or large cross-section. That means that typical manufacturing lengths of the highest voltage classes can be 10-20 km, whereas much longer lengths can be extruded in one run when the voltage class is lower and/or cross-section is smaller. Mass-impregnated cables are produced batch wise, where length depends on size of the impregnation vessel. The focus below is on extruded cable technology.

Delivery lengths that are finally loaded onto the installation vessel are much larger than individual manufacturing lengths. For that reason, manufactured lengths are jointed together using so-called factory joints to form one large delivery length. Maximum delivery length is determined largely by factory storage capacity, i.e. size of turntables, and also by maximum load the vessel can handle and safely transport. Maximum delivery lengths are defined by the size and maximum weight capacity of the vessel’s turntables. The outer diameter and height of such turntables define the volume while the vessel’s properties define the maximum weight of the stack. Diameters of the turntables in most modern vessels can reach 20-25 m or more and the maximum weight the different vessels can carry sometimes surpass 10,000 tons. As such, maximum delivery lengths of power cables, while greatly dependent on cable and vessel type through their size and weight, nowadays easily reach well over 100 kms. Looking to the future, lengths of 150 km or more seem realistic. These different considerations about length can have profound implications on testing.

While cables are available in many different types, it is possible to categorize them according to certain hierarchies. Firstly, AC cables can be divided into 1- and 3-core types. The 3-core cables share the same armor package and the 3 cores can be installed in a single lay. By contrast, single core AC cables need 3 separate laying campaigns per phase. DC cables need only two poles (plus and minus) and, although often laid in a bundled fashion, the two poles are still two independent separately armored cables.

AC cables have a limitation in useful length between two connecting stations since they generate capacitive current. Depending on voltage class, such lengths are roughly maximized to between 50 and 150 km. By contrast, DC cables do not face this limitation and can attain route lengths of several hundred kilometers. This means that testing DC cables can face practical challenges due to the possibility of their significantly increased lengths.

AC cables (including submarine type) have been implemented at 500 kV whereas DC cables are commonly installed and projected at a level of 525 kV. Route lengths can be several hundreds of kilometers and it seems probable that the 1000 km barrier will eventually be surpassed using DC technology. It is clear that probability of a small defect in a cable system increases with its length. As such, testing, as one component of quality control, becomes very important. Nevertheless, it is also important to note that no testing can replace good workmanship and proper quality control of all steps – from sub-supply, to design, to installation. In this context, testing is rather an integral part of QA and QC.

Moreover, testing can also be categorized into three hierarchal classes: qualification testing, manufacturing and installation testing, and on-line testing & monitoring. Passing a pe-qualification test (PQ) or the addition of an extension of it (EQ) means that the manufacturer of a certain technology can supply such cable systems on a commercial basis, covered by such standards or recommendations. Holding a successful type test (TT) certificate means that the supplier can provide a certain technical solution on a commercial basis that meets the performance characteristics for an intended application, as covered by the standards or recommendations. These tests are conducted on a limited length of cable (tens to above 100 m) and only a small number of accessories. Such tests can be viewed as a student having passed a university exam but with no guarantee that they will do a good job every single day of the week in their upcoming career.

Manufacturing and installation tests cover every single meter of cable and every accessory. Such tests are routine tests, FAT (factory acceptance test) and possibly also a harbour test. The extent of these tests depends on cable system type as well as on the contract. Sample tests are performed on a sample of the projected cable length and number and serve to ‘measure’ the quality of design at certain points in time along the project line. The harbour test is intended to hand over responsibility for the project from manufacturer to installer in cases where these are different entities.

The challenge in testing long submarine cables comes with the long manufacturing lengths and very long delivery lengths. Typical extruded AC cables are routine tested at 2.0-2.5 U₀, the lower multiple representing the higher U₀ voltage class. The historical reason for this was that one did not want to exceed certain electric field thresholds above which the cable would age in a different way than from service. In regard to long cable lengths, there is another reason that practically limits the test factor. The higher the test factor (and with that the voltage), the larger the capacitive loading current. The test equipment must be able to deliver that. Originally, this problem was self-limiting because cables with the highest voltage classes often had the shortest delivery length due to the inherent issue of capacitive loading in service. This challenge has now returned with the introduction of extruded DC cables.

The recommendations state that the manufacturing lengths are to be tested with an AC voltage such that the AC stress is 20 or 23 kV/mm at the conductor screen. The challenge increases given the realization that the recommendation asks for an AC and PD screening test of the factory joints, if applicable. The term, “if applicable” was added because the challenge now becomes clear. Several manufacturing lengths are being factory-jointed together and, if possible, each joint shall be locally PD-measured. This means that the joint and the cable lengths before and after the joint shall be put on voltage, even though for short duration. The last joint, adding the last manufacturing lengths to the previous sections, make up the full delivery length.

HVDC cables having delivery lengths of 100 to 150 kms or more and tested at 23 kV/mm places severe challenges on the AC resonant test equipment, which must become very large. For example, the capacitive loading current during testing of a large cross-section 525 kV cable, tested in the interval 400-440 kVac, becomes around 1 kA if fed at one end and half that value if fed from both ends. These values are significant and care must be taken to ensure that the cable sheath can carry the current. Similarly, the heat generated in the cable stack also has to be considered.

One way of overcoming this challenge is to allow for other testing methods. TB 852, for example, offers a different way of addressing the QC of the joint using X-ray inspection. Another emerging technology is acoustic detection and in-line ultrasonic sensing of extruded core is a sometimes-utilized technology. Moreover, based on that technology where the cable is moving through a static head with ultrasonic sensors, another technology has now been developed as well, i.e. keeping the object fixed and static, while the ultrasonic sensing head is moving instead. Such technology could prove particularly useful in controlling quality of the factory joint insulation and its interfaces. Ideally, one could then limit size of the AC resonance test set by optimizing it only for testing manufactured cable lengths. Alternatively, accept the maximum length that the AC test resonance set can handle, utilizing AC and PD testing up to that length, and rely on acoustic sensing for the joints after that length.

Installation tests can be divided into two categories as well. One category summarizing the tests during installation and the second the classic after-installation test. The after-installation test for AC cable systems is again limited to the maximum route length specific to the voltage class. PD tests are then typically performed locally on the accessories. Obviously, there is a need for a mobile resonance or DAC set if the test is not performed by connecting to the grid.

Lately the question has arisen whether one should perform an AC after installation test on installed underground extruded DC cable systems (possibly including local PD tests on the prefabricated accessories). Considering the inherently long lengths of DC cable systems, the AC test set-ups become very large, as explained above in relation to factory testing. There is as yet no experience relating the effectiveness and therefore the benefit of such AC testing on performance of extruded DC cable systems. The complexity of this test must therefore be weighed against its perceived effectiveness. Clearly, without much experience on what is being gained, the choice becomes less technical and more economic or of a risk management nature.

Obviously, if one could measure partial discharges under DC voltage online, one would have the perfect monitoring tool. But the challenges here are significant. Electrical noise levels of the outdoor system are often high and the repetition rate of the discharges are inherently much lower under DC than under AC voltage. Nonetheless, research has been pushing progress in addressing that problem and, although not yet solved, this remains a distinct possibility in the future. CIGRE Working Group D1.63: Partial discharge detection under DC voltage stress will result in a Technical Brochure that summarizes the state-of-the art of the industrial practice of this phenomenon.

Cable systems these days are often equipped with an integrated fibre (FIMT) for temperature and acoustic sensing purposes. Although limited in sensing length, progress is being made to increase this. Acoustic sensing is mainly used to detect external damage or imminent such, whereas temperature sensing could preventively sense changes in the external thermal environment.

Finally, in regard to offshore windfarm cables, all the above remarks are valid for both interconnector cables and offshore wind export cables. They are also valid in some respects for array cables, but these are inherently shorter and have different design aspects.

Summary & Conclusions

The biggest challenge in cable system test methodologies is finding the right balance between efficacy and not damaging against a background of acceptable budget and realistically sized testing equipment.

When it comes to reliability of cables in general, there is much confidence in testing at the production location as being well-understood, controlled, effective and standardized. That is why users are looking for similar quality control for accessories and systems after installation.

Lengths of several hundreds of kilometers in HV networks are no longer rare.

Most commonly used systems for MV networks are single-ended PD measurement systems that make use of either damped AC or very low frequency excitation voltage. There is still lack on guidance when looking to trending limits of especially aged cable systems. Apart from user experience and recommendations, there is no standard and, as for the differences between the test voltage sources, there will most probably never be general consensus/guidance on acceptance levels. With all the measurements that can be performed for both withstand testing, diagnostic testing or monitoring, proper handling of data is also of utmost importance. There is a growing need for an asset-management system/software that not only handles the measurement data but also includes key statistics such as age, vintage, location, number of faults, etc.

Studies by different parties over the past 20 years have shown that there are some differences between ACR and DAC methods in regard to voltage stresses, partial discharge occurrence and dissipation factor measurement. Still, both methods, when used by experts, are a valuable assessment tool. Traditionally recommended over-voltage testing with a binary test outcome, i.e. ‘breakdown’ or ‘no breakdown’, may reveal major defects. But PD detection should be added where possible and a check should always be made of the risk to ignite faults that would not have occurred under operating voltages.

As concerns submarine factory and online cable testing, maximum delivery lengths of cable depend greatly on the cable and vessel type through their size and weight but can now easily reach well over 100 kms. This will likely rise to 150 km or more in the future. The recommendations state that the extruded DC cable manufacturing lengths are to be tested with an AC voltage such that the AC stress is 20 or 23 kV/mm at the conductor screen. The challenge increases when realizing that the recommendation asks for an AC and PD screening test of factory joints, if applicable. Testing of the last joint, adding the last manufacturing lengths to the previous ones making up the full delivery length, places severe demands on the AC resonant test equipment, which needs to be very large. Other means, such as acoustic screening methods, might be a tool for future development.

All insulation defects resulting from installation that might reduce the system quality under AC will produce PD and that is one way to identify them.

ABB and PEHLA Laboratories

Germany

CATU Test Laboratory

FRANCE

See more Laboratories

References

[1] E. Gulski, R. Jongen, J. Parciak, J. Sieminski, “Modern Methods for Quality- and Condition Assessment of High Voltage Cables” 2020, VDE VERLAG GMBH, Berlin, Offenbach, ISBN 978-3-8007-5353-6.
[2] Smit J, van Riet M, Staarink B, “Non-destructive after laying test with PD localization”, Jicable 2019, paper D3-4
[3] IEC 60840: Power cables with extruded insulation and the accessories for rated voltages above 30kV up to 150kV Test methods and requirements;
[4] IEC 62067: Power cables with extruded insulation and the accessories for rated voltages above 150kV,
[5] IEC 63026: Submarine power cables with extruded insulation and their accessories for rated voltages from 6 kV (Um = 7,2 kV) up to 60 kV (Um = 72,5 kV) – Test methods and requirements, 2019;
[6] IEEE 400-2012: Guide for Field Testing and Evaluation of the Insulation of Shielded Power Cable Systems Rated 5 kV and Above;
[7] IEEE 400.4-2015: Guide for Field-Testing of Shielded Power Cable Systems Rated 5 kV and Above with Damped Alternating Current Voltage (DAC)
[8] HD 632 S2 (CENELEC): Power cables with extruded insulation and their accessories for rated voltages above 36kV (Um=42kV) up to 150kV (Um= 170kV);
[9] UK Power Networks, ECP 11-0006, Engineering Commissioning Procedure, ECP 11-0006, HV insulation testing 2018;
[10] IEEE 400.3-2006: Guide for PD Testing of Shielded Power Cable Systems in a Field Environment;
[11] IEC 60270: Partial discharges measurements;
[12] IEC 60885-3: Test methods for partial discharges measurements on lengths of extruded power cable;
[13] IEC 60060-3: High Voltage test techniques Part 3: Definitions and requirements for on-site testing;
[14] Cigre TB 420: Generic Guidelines for Life Time Condition Assessment of HV Assets and Related Knowledge Rules, 2010;
[15] E. Gulski, R. Jongen, B. Quak, J. Parciak, A. Rakowska, Fifteen Years Damped AC Testing and Diagnosis of Transmission power cables, Jicable 2019, paper D3-1.
[16] Leufkens P, Gulski E, Jongen R, Damped AC for Commissioning & Diagnostic Testing of HV Cable Circuits, Proceedings of INMR World Congress (INMR 2019), 20-23 October 2019, Tucson, Arizona, USA, ISBN: 9781713803256;
[17] P. Cichecki, Testing and Diagnosis of High Voltage and Extra High Voltage Power Cables with Damped AC Voltages ISBN: 978-83-952726-0-8;
[18] F.J. Wester, Condition assessment of power cables using partial discharge diagnosis at damped AC voltages, ISBN 90-8559-019-1;
[19] CIGRE: Technical Brochure 841 After laying tests on AC and DC cable systems with new technologies, 2021
[20] PTPiREE, Frame instructions for exploitation of power cables up to 110 kV, 2011;
[21] PSE, Power cable lines 220 kV and 400 kV, 2020;
[22] M. Jeroense, Quality Control of HVDC Cables – The next industry challenge, JICABLE 2019, D8-3;
[23] TB 852, Recommendations for testing DC extruded cable systems for power transmission at a rated voltage up to and including 800 kV, CIGRE, 2021;
[24] S. Sutton, P. Willmott, “Determining Cables Metrics Using 3D Ultrasonic Scanning”, CIRED 2017, Paper 0418;
[25] IEEE 400.2-2013: IEEE Guide for Field Testing of Shielded Power Cable Systems Using Very Low Frequency (VLF) (less than 1 Hz).
[26] IEC 60502-2-2014:Power cables with extruded insulation and their accessories for rated voltages from 1 kV (Um = 1,2 kV) up to 30 kV (Um = 36 kV) – Part 2: Cables for rated voltages from 6 kV (Um = 7,2 kV) up to 30 kV (Um = 36 kV)
[27] Neetrac CDFI, Phase 2, Cable handbook 2016 – Chapter 9 Withstand testing
[28] ICEA S-94-649-2004: Concentric Neutral Cables Rated 5 Through 46 kV
[29] R.Probst, F.Petzold, H.Putter: Excitation Voltages for Partial Discharge Diagnostics on Medium Voltage Distribution Cables, IEEE T&D Denver 2018
[30] H.Putter, F.Petzold, P.Legler: Offline PD diagnostics using several excitation voltages, CIRED 2017 Glasgow, Paper 0049
[31] H.Putter, F.Petzold, D.Götz: Innovative Solutions for On-site Diagnosis of Distribution Power Cables, IEEE T&D Orlando 2021, Paper 0041
[32] A.Lathouwers, P. Jansen, E. de Meulemeester: TenneT’s giant leap to be able to replace 140 substations within next 10 year, while in service and coming from different lay-outs, CIGRE Paris 2020, B3-104

The post Cable Testing: Standards & Technologies appeared first on INMR.

This edited contribution to INMR by Jeff Butler of Hubbell Power Systems, explains how testing for these was defined to ensure performance would meet the demanding requirements.

EC Insulator Jiangxi Co., Ltd.

China

EB Rebosio SRL, A Gruppo Bonomi Company

Italy

See more suppliers of Fittings & Line Hardware

Spanning almost two decades of planning and development, the SunZia Wind & Transmission Project marks a landmark in development of the clean energy sector in the U.S. The project included construction of over 500,000 acres (810 km2) of generation consisting of over 900 windmills integrated to the grid with a 550 mile ±525kV HVDC transmission line. This line provides connection of generation assets in New Mexico to central Arizona load centers (see Fig. 1). With an investment exceeding $10 billion, the scope of the project includes servicing over 1 million American homes, making it the largest clean energy infrastructure project in the country’s history. Once fully operational in 2026, this Project will set a standard for other large-scale renewable energy projects.

One of the key elements of the Project is the ±525kV HVDC Transmission Line which will serve as a major artery for clean energy distribution across the southwest. A critical focus in project success was developing an effective test plan for key components and materials of the line, namely the transmission assemblies which included insulators and hardware.

In this regard, steps were taken to evaluate applicable standards from ANSI, IEEE and IEC that guide electrical and mechanical testing of products under high voltage transmission applications. However, since testing for HVDC is still not comprehensively defined, one of the challenges involved applying available standards to develop a test plan to verify performance.

SunZia Transmission Line Design

The SunZia Transmission Line Project consists of a 550 mile ±525kV high voltage direct current (HVDC) transmission line connecting central New Mexico to south-central Arizona. This line will have a capacity to transport up to 3,500 MW with the expectation to provide power to over 1 million American homes.

The transmission line design is a bi-pole design using pole conductors in combination with metallic return conductors. Structure designs included both tubular steel and lattice towers, depending on requirements and related assembly designs, as detailed in Table 1.

SunZia Transmission Line Testing

1. Evaluation of Relevant Industry Standards

Given the importance and scope of the Project, it was necessary to consider existing guidance and standards to evaluate performance of key components. However, as stated, there is still only limited guidance defined in available industry standards. Therefore, an investigation was performed to determine applicable current standards and how these could lead to a relevant test plan.

2. Results

An analysis of relevant IEC standards was performed (see Table 2). This uncovered a lack of clarity in terms of defining effective testing for polymer insulators for HVDC applications. Certain standards define they are not applicable for polymer/composite insulators but do not provide sufficient clarity as to whether this exclusion is intentional or accidental. There is guidance from IEC 61109-2008 “Insulators for overhead lines – Composite suspension and tension insulators for a.c. systems with a nominal voltage greater than 1 000 V – Definitions, test methods, and acceptance criteria”. However, despite the name of this standard referencing a.c. systems, the standard references procedures for electrical tests with the following:

“The electrical tests… shall be performed according to IEC 60383-2 to confirm the specified values.”

In referencing IEC 60383-2–1993, the standard states in “Section 2: Test procedures for electrical tests” that “[t]he lightning and switching impulse voltage and power-frequency voltage test methods shall be in accordance with IEC 60-1.” While impulse voltage is agnostic to whether it is for an AC or DC system, it was still important to apply available guidance from the standards in support of a proper test plan. Therefore, in an indirect but logical path, IEC 61109 states that it only applies to “a.c. systems” but then references IEC 60060-1 that does allow for direct voltage tests. Similar evaluations of relevant standards provided some guidance toward a proper test plan.

3. Transmission Assembly Material Testing

As applicable standards were evaluated, a test plan began to unfold. For glass bells, the standards provide clear guidance on HVDC testing. However, they do not provide such guidance for polymer/composite insulators. On the other hand, the standards relevant to transmission hardware testing are more direct with little distinction necessary for AC vs. DC since testing is largely mechanical or current based. Based on these reviews, the following testing outlines the particularities for insulators and assemblies specific to the challenges of HVDC and this project’s design.

A. Polymer Insulator Testing
Minor but important adjustments from the defined standards were needed to define the electrical testing needs for the polymer insulators. As the defined ratings for insulators were developed it was necessary to perform the testing to confirm the insulators provided would meet these ratings (see Table 3).

1. Electrical Testing: Polymer Insulators
The polymer insulator designs were selected with the expectation to meet the electrical requirements of the Project. However, it was necessary to use a software package to perform electric field analysis (EFA) to evaluate corona performance of the entire assemblies. Simulations showed that optimal protection would be provided by a non-standard corona ring due to the challenges of the applied HVDC voltage that the assemblies would need to endure. This analysis led to development of a patent-pending combination ring design used on the polymer insulators for the PC assemblies (see Fig. 2).

Moreover, evaluation of the design of the medium angle assembly resulted in optimal placement of the yoke plate mounted ring (see Fig. 3).

Insulator testing was conducted to verify the electrical ratings of the polymer insulators used on the project. The Metallic Return Conductor (MRC) polymer insulators were set-up as shown in Fig. 4 while the Pole Conductor (PC) polymer insulators were set-up as shown in Fig. 5. The PC polymer insulators were tested using the combination ring as determined from electric field simulation (see Fig. 2). The insulator assemblies were also set up as shown in Fig. 6 to evaluate corona performance. All insulators and assemblies met the ratings, as defined in Table 3.

The assemblies were also tested for corona/RIV values as defined per IEC 61284. The assemblies were tested without addition of the combination ring on the polymer insulators, as determined from e-field analysis, in order to evaluate the worst-case scenario for the assemblies. All assemblies met the requirements of the project for corona and RIV performance (see Table 4).

2. Mechanical Testing – Polymer Insulators & Assembly Testing
A test plan was developed to evaluate extreme wind loading on the tangent assemblies to determine if sufficient compressive load might be applied to the insulators that could compromise their mechanical integrity. The tangent assemblies for both 50k lbs. and 90k lbs. were mounted in the laboratory to apply a transverse load equivalent to the Project’s wind loading (see Fig. 7). The displacement, stress, and mechanical load of the insulator legs was recorded (see Table 4).

Finally, a tensile test to failure served to confirm that no impact was observed to the ultimate strength of the insulator in each leg of the assembly based on an estimated maximum stress for the fiberglass rod of ~100,000 psi (see Table 5).

Testing Silicone Rubber Compound

Testing defined in IEC 60587-2022 was used as the basis to develop a test method to evaluate performance of the polymer material relative to high voltage DC exposure. IEC 60587 defines an inclined plane test conducted on polymer plaque samples with applied voltage and contaminant flow to define a rate. However, since the standard does not differentiate between AC and DC applications, initial testing held the voltage and flow-rates constant for both AC and DC voltages (see Table 7).

Findings presented by CIGRE D1.27 evaluated effectiveness of the test method per IEC 60587 for AC and DC applications. In addition, the CIGRE D1.27 investigation was influenced by work in the Chinese Standard, DC/T 810-2002. One of the key outcomes of both efforts was to recommend that a flow rate of 0.2 ml/min be used for 4.5 kV and 5.5 kV DC voltages (see Table 7).

These test methods were evaluated with two proprietary silicone rubber polymer compounds offered. The samples passed testing for AC voltage, with visible erosion depth <1mm (see Table 8).

However, the same silicone rubber polymer compound samples failed the DC tests since erosion exceeded the maximum depth of 1 mm (see Table 9). These results aligned with findings from CIGRE D1.27 that a same flow rate for AC voltage is too aggressive for DC voltage testing. Therefore, based on these findings (and confirmed by CIGRE efforts and established Chinese standard), flow rate was reduced to 0.2 ml/min. Testing was then repeated on the two variants of silicone rubber compound at both 4.5 kV and 5.5 kV, this time with successful results (see Table 10).

These results confirmed the findings of CIGRE D1.27 that recommended flow rate for DC voltage for the IEC 60587 testing should be 0.2 ml/min to effectively evaluate performance of a silicone rubber compound, similar with the same test for AC voltage.

Conclusions

As this Project nears completion, with energization anticipated in 2026, its success has been due to the collaborative efforts of numerous stakeholders. This confirms the importance of early commitment to clear communication and disciplined focus toward a shared objective.

Despite challenges encountered, testing and evaluation of the transmission assemblies represented just one of several key milestones contributing to the overall achievement of this endeavor. As the industry continues to pursue HVDC projects of similar scale and complexity, it will be essential to establish clear, effective standards for testing to evaluate performance of critical components.

Bibliography
[1] SunZia Wind & Transmission – Fact Sheet [Internet]. [updated 2023 October; cited 2025 May]. Available from: https://patternenergy.com/wp-content/uploads/2023/10/20231005-SunZia-PROJECTS-Factsheet.pdf
[2] SunZia Wind & Transmission [Internet]. [cited 2025 May]. Available from: https://patternenergy.com/projects/sunzia/
[3] SunZia Transmission Project, US [Internet]. [updated 2024 July; cited 2025 May]. Available from: https://www.power-technology.com/projects/sunzia-transmission-project-us/?cf-view
[4] IEC 61109–2008
[5] IEC 60383–2–1993
[6] IEC 60060–1–2010
[7] IEC 61245–2015
[8] IEC 61325–1995
[9] Articulation and Bending Stress on V-String Long Rod Composite Assemblies. EPRI, Palo Alto, CA: 2024. EPRI Project 1-118633-01-01.
[10] IEC 60587–2022
[11] CIGRE D1.27-2015 – CIGRE Technical Brochure «Feasibility Study for a DC Tracking & Erosion Test.

The post Transmission Assembly Design & Testing for ±525 kV HVDC Project appeared first on INMR.

Temperature measurement on the cable surface and partial discharge sensors are examples of such monitoring systems and partial discharge measurements are established tools for condition assessment of cable accessories as well.

Most outdoor terminations these days are filled with an insulating fluid, typically silicone oil, and a diagnostic of this fluid therefore offers another means of overall assessment, apart from PD measurement. The key parameters in this regard are moisture content, dielectric properties and dissolved gases.

This edited past contribution to INMR by experts at Brugg Cables in Switzerland described methods to assess terminations used on cables with extruded insulation and which are the interfaces to other components of the power grid. These can be classified into indoor terminations for connection with transformers and gas-insulated switchgear or outdoor terminations for connection to an overhead line or switchyard.

DALEKOVOD OSO

Croatia

PFISTERER

Germany

See more suppliers of Cable Accessories

Design of Fluid-Filled Outdoor Terminations

A typical fluid filled outdoor cable termination is shown in Fig. 1. Much the same design is used for voltages Un = 60 kV to 500 kV. The main component of this accessory is the stress control element, which uses a semi-conductive electrode made of silicone rubber or EPDM for geometric field grading. The outer insulator can be porcelain or a composite housing, depending on customer preference.

As mentioned, silicone oil is the main insulating fluid in use these days. A certain chain length of silicone molecules is necessary to avoid migration into the silicone rubber of the stress cone. The consequence is a relatively high viscosity if the stress cone is made from silicone. But there are also terminations still under operation that are filled with mineral oil in spite of the fact that environmental impact of leaks has caused the shift in technology towards the silicone oil, which is also non-flammable.

Analysis of Dissolved Gases

Dissolved gas analysis (DGA) is a well-established technology for assessing condition of power transformers and also widely used for evaluation of old oil-filled paper (LPOF) type cables. In both cases, the insulating fluid is typically mineral oil, as also used in the past in outdoor terminations for extruded high voltage cable. The aim of a DGA is detection of partial discharges in the insulating medium so as to prevent failure of the termination. Hot spots that can sometimes occur in power transformers are highly unusual in a termination since any heat losses can easily be removed due to the large surface area relative to oil volume. The basic procedure in a dissolved gas analysis consists of the following:

• Taking sample from the termination;

• Extracting the gas from the oil;

• Analysing the gas sample by computer tomography;

• Interpreting findings according guidelines of the manufacturer.

Convective flow causes a homogenous mixture of the oil during operation and a sample taken from the lower base plate of the termination can therefore be regarded as representative. Sampling using a probe is theoretically possible during service but typically has to be done offline for reasons of safety. It is obviously important to avoid any contamination of the sample. There are a number of possible procedures for extracting gases from the oil and each procedure might yield somewhat different results. As a result, it is recommended to always use the same laboratory. Moreover, what is also important in such assessment, apart from absolute value, is the trend.

Since there has not been a standard that could be used for interpretation of results, the experience of the accessory manufacturer has been necessary to properly assess fluid condition based on gas content. The limits, given in Table 1 are based on such experience and are classified into A, B and C. As long as gas content stays below the values in A, no further maintenance action is required. Findings in the range of B, suggest that further tests or a repetition of testing after a shorter than normal interval are recommended. Reaching the limits given in C generally implies that immediate action must be taken.

The main gases for such evaluation are hydrogen (H₂) and acetylene ethyne (C₂H₂). H2 indicates that partial discharges of low energy are occurring inside the termination. If C₂H₂ is detected as well, it could be concluded that there have been partial discharges of higher energy or even some electric arcing taking place inside the fluid. The presence of methane and propane are further indicators of internal electrical activity in the insulating fluid, mainly in relation to the mineral oil in an oil-filled cable or transformer. However, evaluating the condition of a termination on a polymeric XLPE cable is not always possible due to the fact that both of these gases are generated during cable production as by-products of the cross-linking process. These gases could therefore remain in the insulation if the cable is not properly de-gassed after extrusion and migrate during operation into the insulating fluid of the termination.

While partial discharge measurement provides only a momentary view of electrical activity of a termination, DGA is a representation of the previous service history of the termination. Such measurement therefore provides an indication if and to what extent electrical activity has occurred inside the termination. In any case, analysis of dissolved gases allows no localization of failure inside the termination. As long as only discharges are occurring, the termination will not fail during a subsequent short time interval.

First testing of terminations could be done 12 to 24 months after system energization but assessment during operation might be deemed necessary if there any concerns due to certain events, such as failure of a termination on a neighbouring phase in the system. External events such as high pollution level accumulating on the surface of the termination is another reason for more immediate condition assessment. Manufacturers typically recommend repetition of DGA every 2 to 5 years, with the specific interval selected depending on service environment. Of course, a shorter time frame between successive tests will allow better understanding of any possible trends. It is also important is to define beforehand what remedial actions are to be taken if certain limits have been reached.

Evaluation of Moisture Level Inside Termination

Any moisture present in silicone oil will decrease its dielectric strength. Fig. 2, for example, shows this dependence for two different silicone oils.

From this chart, it is clear that dielectric strength decreases dramatically above a concentration of 100 ppm. Based on these results as well as service experience, manufacturers therefore have defined a limit of 90 ppm, which is significantly higher than for new oil.

Moisture ingress is typically due to improper sealing. Leaks could also be caused by damaged components or by errors during installation error and are one of the main reasons requiring maintenance intervention on a termination. Measurement of moisture content in the insulating oil of a termination is a relatively easy diagnostic tool since only small samples are needed. Assessment is then also faster for evaluation of the breakdown voltage of the sample.

Case Histories from Service

Two case histories are presented here to illustrate manufacturer experience in diagnostics of fluid-filled terminations. In both cases, moisture content was determined and a DGA performed. The first involved assessment of a 420 kV termination installed in gas-insulated switchgear (GIS), as shown in Fig. 3. These kinds of terminations are typically completely filled with silicone oil and also equipped with a vessel that compensates for expansion of the silicone oil.

Evaluation of the silicone oil was performed because the termination was flagged during a PD-test and results of the analysis are given in Table 2.

Moisture content was found to be higher than expected for a newly installed termination, even if below the limit. The ratio of methane and ethane formed during cross-linking of XLPE-insulation is approximately 10 to 100 but in this case the ratio was different. This suggested that the relatively high ethane content in combination with the hydrogen present was caused by discharges in the silicone oil. The assessment eventually led to the conclusion that the termination had to be visually inspected and the unit, with insulator removed, is shown in Fig. 4. A yellow jelly-like material was seen on the upper part of the termination and this residue of aged silicone oil was caused by electrical discharges in the oil. Since the cable core was affected only on the surface, it was possible to simply clean it as the necessary remedial action. This result allowed re-assembly of the original termination and an outage of the termination was thus avoided with only minimal testing.

In the second example, a termination failed during service. Based on this, the insulating medium of the two remaining terminations (Phase L1 and Phase L2) were evaluated to detect whether partial discharges had occurred. For this reason, dissolved gases and moisture content of the silicone oil were assessed, with results shown in Table 3.

These results showed that there was some moisture but below the limits provided in Table 1. While some methane and ethane were also detected, concentration of the two main diagnostic indicators – hydrogen and acetylene – was below their upper limit. Due to this, both terminations could be assumed to be free of partial discharges inside their oil.

Conclusions

Periodic evaluation of the condition of fluid filled terminations should be an essential part of any preventive maintenance strategy for cable systems. The reasonable time interval for this will depend greatly on the environmental conditions of the termination. There are proven diagnostic tools that allow both manufacturers and users of high voltage accessories to perform reliable assessments of their terminations. Determination of moisture content is a major element of evaluating condition of a fluid filled termination while an additional diagnostic tool is analysis of dissolved gases.

It has been shown that there are two key gases that allow reliable assessment of the operating condition of an outdoor termination. Hydrogen (H2) shows partial discharges of low energy but should acetylene (C2H2) also be found in the oil, it could be concluded that discharges of high energy are occurring. Visual inspection is recommended in such a case. Evaluations of terminations performed on site have confirmed that dissolved gas analysis delivers reliable results.

The post Assessing Condition of HV Cable Terminations appeared first on INMR.