Аутор: Paul C. Buddingh, P.Eng. Member, IEEE Universal Dynamics Limited 100 - 13700 International Place Richmond, BC V6V 2X8 Canada
Copyright Material IEEE – Paper No. PCIC 2002-11
Ја. УВОД
This case study describes an investigation by the author of harmonic filter failures at a Chemical plant in North America. The plant utilizes large static converters to take incoming High Voltage low current 60 Хз, AC power and rectify it into to Low Voltage, very high current DC power for operation of the electrochemical cells. Harmonic current generation is expected in this type of power system and harmonic filters are commonly used to limit harmonic levels and protect power system components.
A call from the plant indicated that they were experiencing what appeared to be overheating of a set of reactors used in a harmonic filter associated with one of the plant’s converter systems. The reactors in the 5ог harmonic branch of the filter had discolored, and dark bands were evident on the glassfiber surface of the reactors.
The filter was originally installed in 1988 and had a history of problems. The 5ог harmonic reactors had failed before, and a clear cause was never identified. As historical information was reviewed and measurement data collected, it became apparent that something unusual was occurring.
This paper outlines the power and harmonic filter systems at the plant, discusses how uncharacteristic harmonics are generated, analyzes the difficulty, identifies the cause and provides an action plan used to correct the problem.
II. POWER SYSTEM CONFIGURATION
The plant has two production lines, Lines A and B, each consisting of a series of electrochemical cells.
Line A consists of a 1978 vintage 6-pulse rectifier in a single-way ANSI 45 configuration with inter-phase transformer. The primary voltage is 13.8 кВ. Each of the 6 phases or “legs” has eight parallel thyristors. A phase lock loop (PLL) type control system using discrete analog electronics is implemented.
Смоква. 1: Line A Harmonic Filter
A three-branch harmonic filter is installed, consisting of branches tuned precisely to the 5ог, 7ог и 11ог harmonic with 6.9 effective MVAR of capacitors.
The Line B rectifier system is supplied directly at 66 кВ, in an ANSI 45/46 12-pulse configuration shifted an extra 15° apart to make a 24-pulse system. The rectifiers are equipped with a single branch harmonic filter, also at 66 кВ, tuned at the 4.7ог harmonic and rated at 15 MVAR effective.
III. BACKGROUND
It has been well known, since at least the 1930’s, that rectifiers produce harmonic currents as they convert electric power from AC to DC. A classic paper from the days of the mercury arc rectifiers, still relevant today, was written in 1945 by J. Ц. Читати. [1]. The proliferation of large thyristor rectifiers in the late 1960s and early 1970s created a resurgence and exacerbation of harmonics issues, largely a result of the increased size of the converters (in the 20 MW to 30 MW range). These new larger rectifiers typically required large capacitor banks for power factor correction, creating an ideal environment for parallel resonance disturbances. In response, a number of excellent papers were produced addressing this new twist on an old problem [2] [3].
Смоква. 2: Electrical Single Line Drawing Showing Main Power Distribution
This paper is not intended to be a primer or theoretical treatise on harmonics. There are many excellent works referenced that explain power system harmonics in detail. In particular, Ј. Arrillaga et al, “Power System Harmonics” [4], препоручује се. A few highlights pertinent to this case, међутим, will be summarized.
IV. SOME THEORY
Half-controlled converters consisting of a mix of diodes and thyristors will not contemplated in this paper. Half-controlled converters inherently produce even harmonics and are not used in high power applications.
As discussed in detail in the reference papers, a wellbalanced “ideal” static converter – који је, a converter with equal currents in each phase of the rectifier will produce harmonics on the AC side of the converter according to:
h = kp ± 1 (1)
где: h harmonic order k any integer (1, 2, 3,...) p pulse number of the circuit with a magnitude:
Јах = I1/х (2)
где: Ih harmonic current I1 fundamental current magnitude h harmonic order
У пракси, the commutating reactance and phase retard angle of the thyristors will somewhat reduce the amplitude of the current at each of the following characteristic harmonics:
Хармоничан 5 7 11 13 17 19 23 25
Цуррент 0.175 0.111 0.045 0.029 0.015 0.01 0.009 0.008 (Јединица)
These normal or “characteristic” harmonic frequencies starting at the 5ог и 7ог harmonics are expected from a 6pulse rectifier. Слично, a 12-pulse system will have characteristic harmonics starting at the 11ог и 13ог and a 24-pulse will have characteristic harmonics starting at the 23рд и 25ог секундарне фреквенције, итд. The rectifier acts as a harmonic current source, injecting these harmonic currents back into the AC system. If the AC system is reasonably symmetrical and timing of the rectifier firing is exact, the resulting harmonic currents will be equal in all three phases.
Jean Babtiste Fourier’s theory is used to mathematically explain the resulting harmonic spectrum. A 6-pulse rectifier is made up of two single-way, 3-pulse rectifiers, either connected in series in the form of a bridge configuration, or in parallel, as in this case. Fourier’s theory shows that for 3pulse systems the 3рд, 9ог, 15ог…harmonics are zero. A singleway, 3-pulse system is not half-wave symmetrical about the zero axis and produces even harmonics 2, 4, 6,…. The 180° configuration of the two parallel rectifiers creates a 6-pulse symmetrical system, што обично елиминише парне фреквенције.
Смоква. 3: Шема 6-импулсне двоструке Вај везе са међуфазним трансформаторима
У стварном свету, увек постоје неки заостали абнормални непарни и парни хармоници на страни напајања наизменичном струјом. Оне су класификоване као "некарактеристичне" хармонијске фреквенције.
Уобичајено, „некарактеристични“ хармоници су узроковани несавршеностима у систему напајања наизменичном струјом укључујући толеранције фазних углова намотаја трансформатора, комутационе реактансе и присуство улазних хармонских напона напајања наизменичном струјом. Ове несавршености на страни напајања наизменичном струјом утичу на тајминг паљења тиристора, пошто је његов синхронизациони сигнал узет са основне фреквенције наизменичне струје. Нормално, асиметрија је мала, резултујућа дисторзија је мала, а ефекти минимални.
It is assumed that phase-control timing or firing is identical for all semi-conductors on a phase, phase-to-phase timing is coordinated and that each phase group is precisely fired with respect to each other. For cancellation, we need precise and repeatable firing. This is another area where tolerances play a large part. Deviations in firing will also generate uncharacteristic harmonic currents. In a properly designed and operating rectifier, the “uncharacteristic” harmonics are normally minimal, and are not a concern.
Harmonic filters are designed, стога, based on accepted “theory”, only to treat the normal characteristic harmonics. For cost reasons, they are not normally designed to handle excessive “uncharacteristic” harmonic currents.
У. ANALYSIS
Постојао је низ препрека у истраживању и анализи система претварача постројења. Једна је била анализа проблема прегревања без могућности директног мерења 5ог струја гране хармонијског филтера. То је отежавало добијање потпуне слике о постојећим хармонским условима. Хармонички филтер се састоји од три гране прецизно подешене на 5ог, 7ог и 11ог хармоничан. Свака грана се састоји од реактора са ваздушним језгром са сетом кондензатора за фазе А, Б, и
Ц. Филтер се напаја једним металним „Тецк“ каблом преко прекидача опремљеног струјним трансформаторима. Једина практична тачка везе за мерење била је на струјном трансформатору који напаја све три гране филтера.
ТАБЛЕ 1 Измерене хармонске струје на улазу исправљача линије А
Хармоничне струје које производи исправљач биле су разумне са некарактеристичним компонентама вишим од идеалних, али не толико необично за исправљач из 1978. године. Било је запажено, међутим, да су мерења на улазу исправљача имала мањи износ од 4ог хармоничне струје него на улазу филтера. Ово је дало први наговештај да су некарактеристичне хармонске струје извор невоље у реактору.
Мерења на прекидачу линије А показала су да је систем напајања наизменичном струјом прихватљив и да није разлог за забринутост.
Када су извршена мерења на филтеру линије А, све је изгледало разумно. Измерене струје нису прелазиле оцену реактора, а температуре околине су биле унутар тестне оцене реактора од 30°Ц.
Тако, шта је изазвало прегревање? Додатни трагови су откривени док смо прегледали историју рада исправљача. Разговори са особљем за одржавање постројења указали су на то да је недавно завршена обимна реконструкција енергетског дела исправљача, са инсталираним превеликим уређајима. Ово је елиминисало поновљене кварове тиристора до којих је дошло пре реконструкције и био је јак показатељ да је проблем повезан са неправилностима у контроли.
Ако време паљења није идентично за паралелни сет тиристора, неравномерно оптерећење може резултирати стварањем појединачних кварова полупроводника и гравитирати ка каскадном низу кварова кроз систем јер све мање уређаја носи све више и више оптерећења. Уградњом уређаја већих димензија, биљка је елиминисала симптом.
Следеће, систем напајања је анализиран са посебним нагласком на прецизирању било каквих абнормалних услова хармонијске резонанције.
Проверене су различите конфигурације електроенергетског система које се користе током нормалног рада постројења. Занимљиво откриће је направљено када је линија А радила са искљученим исправљачем линије Б и филтером. The 5ог грана филтера у линији А (серија тачно подешена 300 Хз) утврђено је да показује снажну паралелну резонанцу са електроенергетским системом на 4ог хармоничан када је систем линије Б ван употребе. Када систем Линије Б ради, паралелна резонанца је и даље присутна, али ни приближно тако значајно.
Накнадна анализа је показала да ако је линија Б искључена и исправљач производи најмање 5% 4ог хармоника струја, појачава се и изазива а 40% струјно преоптерећење у 5ог грана филтера линије А.
This finding provided the theoretical basis for a growing suspicion that an even harmonic resonance was the source of the reactor overheating. One question remained: the plant normally operates at full capacity, 24 hours per day, all year long — could a short annual maintenance outage on Line B be sufficient to cause the overheating and resulting dark bands on the reactors?
Reactors have a normal maximum temperature rise of 60°C over a 30°C ambient temperature. The manufacturer reports that reactor insulation will not discolor until it reaches 130°C. To reach this temperature, the total current in the reactor would need to increase to 140% of the reactor rating. Since the reactors have little thermal mass, this temperature would occur in the order of minutes.
Armed with this data, теорија која је висока, некарактеристични хармоници који су узроковали прегревање могли би се тестирати. Други сет мерења је направљен на линији А да би се квантификовали прекиди, некарактеристичне хармонијске струје које долазе из исправљача и њихово појачање у 5ог хармонска грана филтера.
Пажљив протокол мерења потврдио је да се појачање у ствари дешава. Меасурементс оф 20% до 58% од 4ог хармоника струја (као проценат укупне струје филтера) забележени су у периоду од приближно 13 секунди на филтеру линије А. Утврђено је да је 5ог гранани филтер је црпео скоро половину укупне струје филтера, и 70% од 4ог хармоника струја. Као резултат, за кратке периоде ови реактори су оптерећени са више од 200% називне струје. Са спуштеном линијом Б, ефекат би вероватно био знатно гори.
Овај последњи податак је употпунио слику.
Смоква. 5: Сериес & Параллел Ресонанце оф 5ог Филтер & 13.8 кВ Бус
ВИ. MORE THEORY
As discussed above, even harmonics can be created in rectifier systems by firing timing irregularities. Galloway [7] describes harmonic instability as the abnormal operation of a converter system due to the harmonic voltage distortion of the power source caused by the harmonic currents itself. J.D. Ainsworth wrote a classic paper on this same topic 35 пре година [8].
Galloway [7] explains the various modes of timing irregularities. The irregularities are defined into three types.
Тип 1 — Pulse Deviation — One of the six pulses does not occur in the correct time or manner. This results in an “across the board” increase in harmonic currents, with poor cancellation of odd harmonics and production of even harmonic currents due to half-wave dissymmetry about zero.
Тип 2 — Phase Unbalance — Phase unbalance does not produce evens; it acts like a single-phase rectifier and produces the full spectrum of odd harmonics with modulation components of ± 2 of normal harmonic frequencies.
Тип 3 — Group Unbalance — Pulses 1, 3 и 5 are displaced an equal amount from 2, 4 и 6. This results in the generation of even harmonics, који је, multiples of 3 ±1.
Measurements made in the plant seemed to indicate that a Type 1 problem was occurring due to random timing variations, as periods of elevated harmonics across the spectrum, including even harmonics were noted. With the older control electronics, међутим, the failure mode was hard to isolate, and a Type 3 problem, with inter-phase saturation, could be occurring.
The inter-phase transformers are typically designed to absorb only a small amount of imbalance between the rectifier halves and can quickly go into saturation. When the rectifier system is not well balanced, the output currents of the two 3-pulse groups flowing in opposite directions in the interphase produce significant dc magnetization of the core. As it goes into saturation and becomes ineffective, the rectifier operates as a two, separate, 3-pulse groups with the star points connected and semi-conductors only conducting over half of the normal 120°. The resulting 60° conduction angle results in about a 17% increase in semi-conductor power (вати) loss. This results in a substantial increase in heating of thyristors, fuses as well as the secondary of the transformers.
This unbalance also results in an effective DC current that the transformer secondary must carry. The transformer can go into saturation, increasing losses and creating large amounts of heat and a disproportionate amount of third harmonic current.
VII. FINDINGS
The pieces of the puzzle were starting to come together. More and more evidence pointed to an even harmonic resonance as the cause of the filter overheating.
The origin of the difficulties experienced is a thyristor firing circuit problem. The age of the control system and resulting electronic component “drift”, appears to have created a Type 1 timing irregularity.
Firing asymmetry was no longer directly affecting the operation of the rectifier with the oversize thyristors that had been recently installed, but was still affecting the harmonic filter under certain plant operating conditions.
The Line A rectifier is over 30 years old and, while well past its original design life, continued operation of these robust machines is common in the electrochemical industry. The Achilles heal of these units is typically the aging electronics of the control system. Electronic equipment has a bathtubshaped reliability curve and this equipment is likely on the upward slope of that curve. Укратко, control system problems are to be expected with older rectifiers.
Measurements demonstrated that with Line B operating, large amounts of 4ог harmonic current overloaded the Line A filter 5ог branch for short periods. The reactors have little thermal mass, and can reach extreme temperatures in the order of minutes. For at least 13-second periods, the reactors were exposed to a 200% оптерећење. If Line B is shut down under these conditions, the currents are likely to be significantly higher. A redeeming feature is that Line B is shut down infrequently for short intervals of maintenance. The cumulative effects of repeated overheating over time has stressed the reactors.
У 1992, one of the 5ог harmonic reactors was replaced. This explains why only two of the three existing reactors are showing signs of damage. The newer reactor has not been exposed to the same degree of repeated overheating as the two older fifth harmonic reactors.
A secondary concern is the DC offset effects on the interphase and secondary circuit of the transformer. While the transformer is in good condition, elevated DC currents can substantially increase heating and lead to long-term degradation. Tap changers, core clamps and other internal hardware can have localized heating effects with increased levels on harmonic currents [10], particularly with the uncharacteristic harmonic currents for which the machine was never designed.
VIII. ACTION PLAN
A physical inspection of the 5ог reactors on Line A was completed and although stressed, were not in immediate danger of failing, particularly if Line B is kept on-line.
The installation of a new rectifier control system is a substantial capital expenditure, and the plant is now considering this step. In the meantime, the following measures have been put into place.
Смоква. 6: Line A Rectifier Control System
Прво, the peak sensing protection relay is being replaced with a modern programmable relay that is sensitive to the low order harmonic frequencies experienced on this system. This will provide alarming and tripping of the filter bank if the reactors are in danger of overload. This relay also measures and records harmonic levels.
Други, redesigned 5ог harmonic filter reactors are being installed to move the parallel resonance between the filter and the power system to below the 4ог. The new design will greatly decrease the sensitivity to resonance. New reactors have been ordered and the replacement has been scheduled.
Коначно, the interval between transformer dissolved gas samples has been decreased to improve monitoring of the transformer condition. Dissolved gas analysis is a great tool to evaluate the condition of transformers, particularly when faced with uncertain harmonic stress. Corrective action can then be taken as required.
IX. CONCLUSIONS
A sustained rectifier 4ог harmonic level of 5% or more, at a time when Line B is off, has overloaded the reactors and caused them to run hot and discolor. Over the years, there has been a cumulative effect intensifying the condition. If nothing was done, plant operating history has established that failure would follow.
As a first step, the filter protection relay was modified to detect a 4ог harmonic current overload and to alarm and trip as required.
The resonant effects of tuning the Line A harmonic filter exactly to each harmonic frequency to be treated was not considered in the original design. Tuning each of the 5ог, 7ог и 11ог branches to a frequency 2% до 10% below the target frequency would have alleviated the parallel resonance.
A redesign of the damaged filter reactors is complete and the new reactors are scheduled for installation.
The new larger thyristors, which were recently replaced, are able to withstand control system irregularities to a much greater degree, with a resulting improvement in reliability. Control system irregularities that earlier caused rectifier problems, међутим, still affect the AC power system.
Even harmonics will also cause the rectifier transformer to run hotter by saturation of the magnetic circuit. Metal clamps, fixtures and other components can overheat inside the transformer, creating localized hot spots. This can significantly reduce transformer life.
Steps have been taken to mitigate the immediate issues as noted and a replacement control system is under consideration by the plant.
Кс. ACKNOWLEDGEMENTS
I would like to thank John Kirichenko and the Plant Staff for the opportunity to work on this very interesting challenge and my colleague Bernd Schmidtke, P.Eng. for his outstanding work and insight on this project.
XI. REFERENCES
[1] J.C. Читати, “The Calculation of Rectifier and Inverter Performance Characteristics”, Proceedings of the lEEE, Лет. 92, Део 2, Не. 29, Октобар 1945, ПП. 495-509. [2] A.P. Jacobs and G.W. Walsh, “Application considerations for SCR dc power systems,” IEEE Trans. IGA-4, July/Aug 1968. [3] D.E. Steeper and R.P. Стратфорд, “Reactive Power Compensation and Harmonic Suppression for Industrial Power Systems using Thyristor Converters,” IEEE Trans. IA-12, 5/76 ПП. 235-255. [4] Ј. Arrillaga et al, “Power System Harmonics”, Вилеи & Синови, ИСБН 0471906409, 1985. [5] Power Converter Handbook, Canadian General Electric Co. Ltd., 1976. [6] ИЕЕЕ 519-1992 “IEEE Recommended Practices and Requirements for Harmonic Control in Electrical Power Systems”. [7] J.H. Galloway, “Harmonic Instability in Phase Controlled Rectifiers,” IEEE PCIC conf. record 1999, ПП. 171-175. [8] J.D. Ainsworth, “Harmonic Instability between Controlled Static Converters and AC Networks,” Proc. IEE, No.7 pp.949-957 July 1967. [9] Ј. Arillaga et al, “Power System Harmonic Analysis,” John Wiley, ИСБН 0471975486, 1998. [10] S.P. Kennedy, “Design and Application of Semiconductor Rectifier Transformers,” IEEE PCIC conf. 2001 record pp. 153-159. [11] Ј. Shaefer “Rectifier Circuits – Теорија & Design,” John Wiley & Синови, 1965. [12] B.M. Bird et al, “An introduction to Power Electronics,” John Wiley & Синови, ИСБН 10430 2 1983. [13] A. Kloss, “A basic guide to Power Electronics,” John Wiley, ИСБН 0471904325 1985. [14] P.C. Buddingh & Ј. Св. Mars “New Life for Old Thyristor Power Rectifiers using Contemporary Digital Control,” IEEE IAS transactions Sep/Oct.2000, ПП. 1449-1454.XII. VITA
Paul C. Buddingh graduated from Lakehead University in Thunder Bay, Ontario, Canada with a degree in Electrical Engineering. Upon graduation, he spent several years working out of Toronto, Canada as an electrical consulting engineer working in heavy industry. У 1991, he co-founded a company that developed a new magnetic approach to solving zero sequence harmonic problems in low voltage systems. У 1997, he moved to Vancouver, Canada and joined Universal Dynamics. He has been designing and installing harmonic filters for 15 године. His work is centered on designing high reliability power systems for difficult loads, power converter issues and resolving power system problems for a number of industrial customers across the Americas. He is a registered Engineer in the provinces of Ontario, Manitoba and British Columbia and an author of several IEEE papers.