Reliability of Open Public Electric Vehicle Direct Current Fast Chargers

Published David Rempel1, Carleen Cullen1,2, Mary Matteson Bryan1, Gustavo Vianna Cezar3,
1Department of Bioengineering, University of California, Berkeley, CA, USA
2Cool the Earth, Kentfield, CA, USA
3SLAC National Accelerator Laboratory, GISMo Group, CA, USA

Abstract. In order to achieve a rapid transition to electric vehicle driving, a highly reliable and easy to use charging infrastructure is critical to building confidence as consumers shift from using familiar gas vehicles to unfamiliar electric vehicles (EV). This study evaluated the functionality of the charging system for 657 EVSE (electric vehicle service equipment) CCS connectors (combined charging system) on all 181 open, public DCFC (direct current fast chargers) charging stations in the Greater Bay Area. An EVSE was evaluated as functional if it charged an EV for 2 minutes or was charging an EV at the time the station was evaluated. Overall, 72.5% of the 657 EVSEs were functional. The cable was too short to reach the EV inlet for 4.9% of the EVSEs. Causes of 22.7% of EVSEs that were non-functioning were unresponsive or unavailable screens, payment system failures, charge initiation failures, network failures, or broken connectors. A random evaluation of 10% of the EVSEs, approximately 8 days after the first evaluation, demonstrated no overall change in functionality. This level of functionality appears to conflict with the 95 to 98% uptime reported by the EV service providers (EVSPs) who operate the EV charging stations. The findings suggest a need for shared, precise definitions of and calculations for reliability, uptime, downtime, and excluded time, as applied to open public DCFCs, with verification by third-party evaluation.

Keywords: electric vehicle charging infrastructure, performance, renewable energy, zero emission vehicles

Background

Reliable, functional, open, public Direct Current Fast Charge (DCFC) electric vehicle (EV) charging stations are critical as countries rapidly transition to EVs. A recent survey of EV drivers in California (N=1290) reported mixed experience with existing EV chargers (CARB, 2022a). They reported experiencing broken plugs (9%), unexpected shut off during charging (6%), charging station not functioning (22%), payment problems (18%), and the need to contact customer service via cell phone (53%). This experience appears to contradict a simultaneous survey of the EV service providers (EVSPs) who reported 95 to 98 percent uptime of their public chargers. An accurate assessment of the reliability, functionality, and uptime of the existing public EV chargers is needed to provide guidance for the successful buildout of the EV charging infrastructure.

Open EV charging stations are those open to all EVs (NREL, 2022). Closed systems, such as Tesla Superchargers, will not accommodate all EVs. Public charging stations are those that are open to the public 24 hours per day 7 days per week (AAI, 2022; NESCAUM, 2019). Examples of non-public charging stations are those in paid parking lots or those limited to customer and employee use. Open, public DCFC charging stations are designed to charge different models of EVs and, therefore, have multiple connector types, such as CCS (Combined Charging System; SAE, 2018), CHAdeMO, and Tesla connectors. Charging stations have one or more kiosks (also called posts), with each kiosk situated adjacent to one or two parking spaces. A kiosk may have one or more EVSEs (Electric Vehicle Supply Equipment) or ports (OCPI, 2020). An EVSE or port provides power to charge only one vehicle at a time even though it may have multiple cables with the same or different connector type (Figure 1). The EVSE provides information on charging and controls the delivery of electricity to the cable (DOE AFDC, 2022). Each kiosk typically includes a payment system that collects payment information from credit cards, debit cards, membership cards or smartphone applications; the transaction may be by tap, insert, swipe, or near field detection depending on the payment method. Another method of payment is Plug and Charge where the only action required is to plug in the EV and the EV is automatically identified and linked to a previously established payment method (ISO, 15118).

Figure 1. A model of an EV DCFC charging station with 2 kiosks or posts, 3 EVSE charge ports, and 4 connectors. Kiosks may have multiple connectors of the same or different types (e.g, CCS, CHAdeMO). [from DOE AFDC, 2022]

The National Renewable Energy Laboratory (NREL) Alternative Fuels Data Center (AFDC) maintains a national database/map of public EVSEs. The database includes charging station location and number of EVSEs (ports) and connection types at each station (NREL, 2022). The data is updated on a periodic basis by EV service providers (EVSPs); some states require updates at least monthly (CARB, 2022b). In addition, commercial smartphone, tablet, and desktop apps, such as PlugShare, provide EV users with information on the location of EV charging stations, the name of the EVSP, the number and types of connectors, the maximum power delivered, and other information.

There are different methods of measuring reliability of an electrical system, but essentially, it is the degree to which the performance of the system results in electricity being delivered to the customer in the amount desired (ORNL, 2004). The reliability of an EVSE, that is, the functional state, can be considered from the perspective of the EVSP or the EV driver. The EVSP may detect the state of an EVSE through its communication network, or as calls to a service number by EV drivers, as a measure of reliability. From the EV driver perspective, a reliable EVSE is one that charges the EV, for the expected duration, after using an appropriate payment method, at the expected rate (i.e., kW). The upper bound on charge rate is influenced by many factors including the EV’s state of charge, the maximum rate allowed by the EV, and the charging station nominal rate. The Alliance for Automotive Innovation (2022) defines a reliability standard as one specifying a minimum uptime requirement. States have different minimum uptime requirements for EVSEs that are paid for with public funds. For the Northeast States (NESCAUM, 2019) “Each connector on each public DC fast charging station pedestal shall be operational at least 99 percent of the time based on a 24 hour 7-day week (i.e., no more than 1.7 hours of cumulative downtime in a 7-day period).” For California, “The equipment must be operational at least 97 percent of the standard operating hours of the charging facility for a period of 5 years” (CEC, 2021).

However, the use of uptime as the reliability metric is controversial since there is no standard definition nor is there a standard calculation methodology. Given the complexity of the EVSE ecosystem and technology stack, from hardware to software, ensuring a high uptime and assigning “uptime ownership” of each EVSE may be difficult and may require standardization across different jurisdictions.

The EVSE ecosystem is composed of different stakeholders. For example, when an EVSE is installed, it is connected to the local utility electrical infrastructure that delivers power to EVSE. The EVSE is installed by a certified installer, operated by the charge point operator (CPO) and located at a site where it may be owned and managed by a site host or the EVSP. The EVSE is connected to an internet service provider (ISP) network and a payment system. Finally, the EVSEs may be serviced by an EV servicing company.

Depending on the jurisdiction, the overall responsibility for keeping the EVSE functioning, can be either with the local electric utility, the installer, the site host, the CPO, or the servicing company. These stakeholders may be independent or may be integrated, i.e., installer can also be the CPO, etc. These stakeholders will likely have different levels of visibility over the status of the system. For example, the site host might have information about the electrical infrastructure and outages and physical damage to kiosks but not information about the functional status of each kiosk, whereas the CPO may have continuous EVSE status information. This partial visibility of the EVSE operation poses a challenge in maintaining a high uptime from the EV driver perspective. Moreover, since these stations are in public locations, events such as road blockage due to construction, theft, or vandalism can occur, which are beyond the immediate control of the CPO. Therefore, the complex nature of the ecosystem and the lack of a clear definition and metrics describing EVSE uptime may interfere with stakeholders’ accountability.

For the purposes of this study, a functional EVSE is one that can charge for a minimum of 2 minutes, using an appropriate payment method, without the need to make a service call. An EVSE includes all the system components within a kiosk that are necessary for a successful charge, including the port, screen, network communication, payment system, power source, software, cable, and connector. If a kiosk has more than one cable with a CCS connector, the functionality of each connector is evaluated and reported as a separate EVSE.

The purpose of this study was to systematically evaluate whether open, public DCFC EV chargers with CCS connectors were functional in the 9 counties of the Greater Bay Area. California has the greatest density of public open DCFC chargers in the US (NREL, 2022) and within California the density is high in the Greater Bay Area.

Methods

All open, public DCFC EV charging stations with EVSEs with CCS connectors in the 9 counties of the Greater Bay Area were identified using the NREL NFDC database and the PlugShare.com website. Stations with CCS connectors with a charge rate >= 50kW were identified. The 9 counties were Alameda, Contra Costa, Marin, Napa, San Mateo, Santa Clara, San Francisco, Solano, and Sonoma. Non-open EV charging stations, e.g., Tesla, as well as non-public EV charging stations, e.g., stations in paid parking lots, private workplaces, or business sites with restricted access hours, were excluded.

The identified EV charging stations were visited by a driver with an EV with a CCS charge inlet. Each EVSE at the station was tested by plugging the CCS connector into the EV and attempting to initiate and sustain a charge for 2 minutes. If the charge was successful, the EVSE was classified as functional. The unique kiosk and CCS connector number or name were recorded. If the parking space was occupied by another EV and the EV was charging, the EVSE was classified as functional. If the parking space was occupied by a non-EV or by an EV and not charging, it was classified as not tested. If none of payment methods tested worked, or the EVSE was not functioning, or did not initiate or sustain a charge, the EVSE was classified as nonfunctional. If the cable was too short to reach the EV charge inlet, the EVSE was classified as a design failure.

The payment methods tested included 2 different functioning credit cards and the vendor mobile app or membership card. Payment methods were tested in the following order, credit card 1 insert, credit card 1 swipe, credit card 2 insert, credit card 2 swipe, then mobile app or membership card, until one of the payment methods was accepted. Each method, i.e, a swipe, was attempted twice before moving to the next payment method. The credit cards used for testing were Mastercard, Visa, and Amex. If any of the payment methods worked and led to a 2 minute charge, the EVSE was classified as functional. The EV drivers were instructed not to call the service number if the EVSE did not work; a functioning EVSE should not require a call to a service number.

Twenty volunteer EV drivers assisted in the testing of the EV charging stations. Only EVs with CCS charge inlets were used. The vehicles used for testing were the Chevy Bolt, Kia Niro, Hyundai Kona, Ford Mustang Mach E, and Porsche Taycan. The EV battery charge level was less than full at the time of testing. The volunteers were trained on the study methods and assigned EV charge stations to test. The survey was completed using a Qualtrics survey on a mobile device while the driver was at the charging station.

A random sample of 10% of the stations was tested at two points in time, approximately 1 week apart, to determine whether the functional state of the EVSEs changed over time.

Results

A total of 181 open public DCFC EV charging stations and 678 EVSEs with CCS connectors were identified in the 9 counties of the Greater Bay Area and visited between February 12, 2022 and March 7, 2022. Of these 678 EVSEs, in 21 instances, the adjacent parking space was occupied by a non-EV (7) or an EV that was not charging (14); therefore, these 21 EVSEs were excluded from the evaluation. The remaining 657 EVSEs that were evaluated are listed by EVSP in Table 1.

Table 1. Evaluated open public DCFC EV charging stations and EVSEs by EV Service Provider

.

1 An EVSE includes all the system components in a kiosk necessary to deliver a charge to a single connector.

Reliability of EVSEs

The functional states of the 657 EVSEs are summarized in Table 2. 72.5% of the EVSEs were functioning at the time of testing; 57.8% were tested and charged for 2 minutes and 15.4% were occupied by an EV that was charging. 22.7% of the EVSEs were not functioning. System electrical failures, e.g., screen blank or non-responsive, text on screen of “charger unavailable” or “connection error”; payment system failure; or charge initiation failure, were the most common causes of failure. A charge initiation failure occurred if the charge did not start after the payment was accepted or the charge started but was interrupted before 2 minutes of charging was completed. A payment system failure was recorded only after all payment methods were tested, each twice, and all failed. A broken connector, e.g., cracked or with bent pins, was recorded for 0.9% of EVSEs.

The cord was too short to reach the EV inlet for 4.9% (N=32) of EVSEs tested. This design failure was recorded at a ChargePoint station (1), EVgo stations (4), and Electrify America stations (27). The EVs tested were driven into the parking space either forward or backward during testing to position the EV inlet as close as possible to the charging kiosk. The EVs used, when it was recorded that the cord was too short, were all Chevy Bolts.

Table 2. Functional states of 657 CCS DCFC EVSEs.

.

1 Charger error, unavailable, under maintenance, etc.
2 Connection, network, communication error, etc.
3 12 of these were evaluated with 2 credit cards but not an app or membership card
4 Short session failure
5 At 3 EVSEs the space was too small to safely back into

Three EVSPs, ChargePoint, Electrify America, and EVgo accounted for 97.3% (639 of 657) of the EVSEs evaluated. The functional states of the EVSEs for the 3 EVSPs are summarized in Table 3. It should be noted that most of the Electrify America kiosks each had 2 CCS connectors that were each tested and reported as independent EVSEs. However, the 2 CCS connectors could not be used simultaneously. If each of these kiosks were considered as a single EVSE, with functionality determined if either just one or both connectors provided a successful charge, the percent of functional EVSEs for Electrify America would have increased from 73.9 to 77.1%.

Table 3. Functional State of EVSEs by the Top 3 EV Service Providers

.

Payment Methods For the 375 EVSEs that charged for 2 minutes, the payment methods that worked are summarized in Table 4. The payment methods were tested in the order presented in Table 4. For example, 50.4% of the successful charges occurred after just the first credit card was inserted. However, 24.5% of the successful charges required an app or membership card for payment, i.e., attempts to pay with 2 credit cards were not successful.

Table 4. Payment method that worked, in the order tested, for the 375 EVSEs that charged for 2 minutes.

.

Testing EV Charging Stations at Two Points in Time

Nineteen (19) randomly selected stations (88 EVSEs) were tested by 2 different EV drivers to determine if their functional state changed over time. The mean time between samplings was 8.0 days (SD=4.9). Eight of the EVSEs could not be compared between the time points because during one of the samplings the EVSE was occupied by a non-EV, an EV that was not charging, or the cord was too short. Of the remaining 80 EVSEs, 48 remained in a functional state, 14 remained in a non-functional state, and 18 (22.5%) changed state from functional to non-functional or a non-functional to functional (5 of these occurred with the same EV model). For the 14 EVSEs that remained in a non-functional state, the cause of failure was the same at both sampling times for 13 of them. The overall functional status changed little between the sampling times, i.e., 72.5% were functional at time 1 and 70.0% were functional at time 2.

Discussion

Of the 657 open public DCFC CCS EVSEs evaluated in this study, 72.5% were functional at the time of testing while 27.5% were either not functional or the cable was too short to reach the EV inlet. The most common cause of a nonfunctional EVSE was an electrical systems failure which included an unresponsive or unavailable screen, a payment system failure, a charge initiation failure, a connection failure, or a broken connector.

This is the first study we are aware of that systematically evaluated the functional state of open public EV chargers. The findings corroborate recent non-systematic surveys of EV owners. In a survey of 1290 EV owners, 34% reported that charging station operability issues were a barrier to using public charging stations (CARB, 2022a). In survey of 5500 EV owners, 25% of those who use public DCFCs reported a major difficulty with chargers being nonfunctional or broken (Plug In America, 2022). In the same survey, only 4% of Tesla owners reported a major difficulty with the Tesla closed DCFC system.

In the Greater Bay Area, 3 EVSPs, ChargePoint, Electrify America, and EVgo accounted for 97.3% of the 657 open public DCFC EVSEs evaluated. There were important functional and design differences between the stations installed by these EVSPs. ChargePoint had the highest percent of non-functional CCS EVSEs at 36.4% followed by EVgo (25.5%) and Electrify America (19.0%). The most critical design flaw was that 7.1% of the Electrify America cables were too short to reach the Chevy Bolt charger inlet, a problem that may be experienced by other EVs with the power inlet on the side of the vehicle. The cable length problem could be addressed with an industry standard on minimal cord length based on the kiosk location relative to the parking space.

The term reliability, when referencing an electrical system, typically refers to the percent of time, over a given time period, that the system is fully operational and able to deliver power at the intended level. This percent is also referred to as the uptime. For public EV charging stations, the definition from the Northeast States, is “the percent of time that a charging station must be functioning properly and available for use by EV drivers” and “Each connector on each public DC fast charging station pedestal shall be operational at least 99 percent of the time based on a 24 hour 7-day week (i.e., no more than 1.7 hours of cumulative downtime in a 7-day period)” (NESCAUM, 2019). New York, California, and the Federal Highways Administration require a minimum uptime of 97% (NYSERDA, 2021; CEC, 2021; FHWA, 2022).

The findings of this study suggest that the currently installed DCFC stations do not meet the 97 to 99% minimum uptime required by public funding agencies. The findings also appear to contradict the 95 to 98% national uptime levels reported by EVSPs (CARB, 2022a, p11). EVSPs do not report the details of how they define and calculate uptime. The EV charging infrastructure would greatly benefit from more data transparency and transparency on methodologies used by each EVSP in calculating uptime. For example, EVSPs could share data on the different subcomponent failure rates and whether the failure was localized, i.e., only affecting one EVSE due to a component failure, or systemic, i.e., affecting multiple EVSEs due to a communication or software problem. Such a reporting mechanism would benefit the entire industry by establishing an ongoing mechanism to identify the weak links in the ecosystem and developing a coordinated approach to addressing them.

While there are state reporting requirements for uptime; there are no precise state, national, or industry consensus definitions of nor calculation methods for uptime. A definition of uptime also requires a definition of the opposite, or downtime. Downtime is the total time that the EVSE is not operational. The clock on downtime should start when the EVSP has evidence that the system is unable to sustain a charge at the expected level. For example, recording downtime could start when there is (1) a system fault detected through the EVSP network where the fault results in the inability to charge, (2) a call to the service center by an EV driver to report nonfunctioning kiosk, (3) evidence of damage to physical components observed either in person or remotely, or (4) a nonfunctioning EVSE reported during a third-party evaluation of the station. If a failure is due to conditions outside of the control of the EVSP, e.g., upstream loss of power, cellular, or internet, it may be considered excluded time. If excluded time is used in calculating uptime, it should be subtracted from the reporting period time.

To improve the accuracy of reliability reporting, a third-party field audit of an EV charging station could be performed at the startup of the charging station and at periodic intervals thereafter. An audit of each EVSE should involve a standard methodology which could include an assessment of the allotted parking space, a measurement of the cable length, a test of payment methods and screen function, and a confirmation that power is delivered to the EV for a minimum period of time at the intended power level. A second type of third-party audit, following an Evaluation, Measurement and Verification (EM&V) process (DOE, 2022; CPUC, 2006), may also be useful to evaluate the EVSP system and data on uptime, downtime, and excluded time. Such audit findings should be made public.

To improve EV driver expectations and experience, accurate, real-time data on EVSE status should be made public. As mentioned before, the definition of reliability can be viewed from the perspective of the EV owner or the EVSE owner, and they are not necessarily the same. Acknowledging this difference, as the technology and regulatory framework matures and is better defined, is important to establish the correct expectations and prevent EV owners from giving up their EVs and returning to gas vehicles (Harding and Tal, 2021). Real-time data would allow EV owners to better understand the actual reliability of the EV infrastructure and adjust their expectations accordingly. Real-time data could be reported by EVSPs to the NREL Alternative Fuels Data Center (AFDC) and published on the National AFDC map and database. The data could also be made available for commercial applications that provide locations of EV charging stations and information on EVSE status to EV drivers.

Uptime may also be improved with standard maintenance and servicing agreements of EV charging stations. The Northeast State guidelines call for a 24-hour window for servicing an EVSE when the EVSE owner or operator is aware that an EVSE is not functioning (NESCAUM 2019). General maintenance may include the periodic checking of EVSE parts for damage; cleaning the EVSE kiosk, cables, and connectors; and removal of garbage and snow (NREL 2022).

Several limitations of the study should be noted.

First, the test of functionality required a 2 minute successful charge of the EV. A charging process may be interrupted for no apparent reason at any time during charging, so the 2 minute duration may be too brief a test period to fully evaluate functionality.

Second, the EV charging stations were evaluated at a single point in time, limiting conclusions about uptime. However, based on our reevaluation of 80 EVSEs, the functional state changed for 22.5% of the EVSEs, but the overall percent of functional EVSEs did not change.

Third, the test method used different payments methods, 2 credit cards and an app or membership card. A well-functioning system should work with just one payment method. However, if the test methodology had required successful charging with just one credit card, the percent of functional EVSEs would have dropped from 72.5 to 49.2%.

Fourth, the test methodology used did not include having the EV driver call a service number if they were unable to charge the EV. The need to call a service number for assistance might be considered by some a normally functioning system.

Fifth, classifying “occupied by an EV and charging” as functional may overstate the overall percent functional since it is unknown whether the EV owner called the service number to initiate charging.

Sixth, the test methodology did not determine whether the port was delivering power at the intended level; this should be included in future tests. Finally, the finding that the cable was too short to reach the EV inlet for 32 connectors is a major station design flaw. The identification of this problem was dependent on the EV model used for testing; testing with an EV that is not a Chevy Bolt may not identify this problem.

Conclusions and Recommendations

As more and more EVs are adopted nationally, the need for fully functional and reliable open public DCFCs will increase. Non-functional public chargers pose an important equity issue as residents in rented or multi-family dwellings usually charge at public charging stations. In addition, non-functional public chargers will have a significant impact on drivers on road trips. Furthermore, high rates of non-functional chargers may inhibit the adoption of EVs. The design of location and quantity of needed DCFC charging stations, for the build out of a national EV charge infrastructure, should not have to assume that a quarter of the EVSEs will be nonfunctional. The level of system failure observed indicates a poor quality of electrical design, components, or software plus the need for EVSPs to improve their identification of the EVSE functional status to trigger timely service. In addition, effective compliance measures are needed for EV charging stations that are part of a court settlement or paid for with public funds. Compliance measures require clear definitions of reliability, uptime, downtime, and excluded time. It may be useful to consider reliability metrics from other industries (e.g., data centers, cloud service providers, etc.), such as mean time to recovery or mean time between failures, etc. In addition, compliance measures may require third-party assessments of EVSEs, using a standard test methodology, at the time of initial operation and at regular intervals thereafter and an assessment of reliability data collected by the EVSPs.

Acknowledgements: We wish to thank the volunteer EV drivers who assisted in field data collection; these included Catherine Bohner, Suzanne Bryan, Lisa Chang, Ed Church, Jeff Cullen, Elena Engel, Ariane Erickson, CM Florkowski, Chris Gilbert, Howdy Goudey, Bill Hilton, Wiley Hodges, Linda Hutchins-Knowles, Douglas Mason, and Louie Roessler. Partial funding for the study was provided by Cool the Earth, a 501(c)3 nonprofit organization. The authors declare no financial conflict of interest

References

Alliance for Automotive Innovation. Planning for the electric future: Charging Station Attributes.
Accessed 3/3/2022: https://www.autosinnovate.org/about/advocacy/Recommended%20Attributes%20for%20EV%20Charging%20Stations%2009DEC2021.pdf
Avista Electric Vehicle Supply Pilot Final Report, Submitted to the Washington Utilities and Transportation Commission (UTC), October 18, 2019.
CARB 2022a. Electric Vehicle Supply Equipment Standards Technology Review. California Air Resources Board. February 2022.
CARB 2022b. Final Regulation Order, Attachment A. Accessed 3/3/2022: https://ww2.arb.ca.gov/sites/default/files/2020-06/evse_fro_ac.pdf
CEC (GFO-21-604- Clean Transportation Program Rural Electric Vehicle (REV) Charging. California Energy Commission, 2021.
CPUC 2006. California Energy Efficiency Evaluation Protocols: Technical, Methodological and
Reporting Requirements for Evaluation Professionals. April 2006. http://www.calmac.org/events/EvaluatorsProtocols_Final_AdoptedviaRuling_06-19-2006.pdf
DOE 2022. Evaluation, Measurement and Verification. Accessed 3/20/2022. https://www.energy.gov/sites/prod/files/2014/05/f16/what_is_emv.pdf
FHWA 2022. The National Electric Vehicle Infrastructure Formula Program Guidance, US DOT
Federal Highway Administration, February 10, 2022, p.22 https://www.fhwa.dot.gov/environment/alternative_fuel_corridors/nominations/90d_nevi_formula_program_guidance.pdf
Hardman, S., Tal, G. Understanding discontinuance among California’s electric vehicle owners. Nat Energy 2021; 6, 538–545. https://doi.org/10.1038/s41560-021-00814-9
ISO 15118-Parts 1 to 20. Road vehicles – Vehicle to grid communication interface. 2017-10-13.
NESCAUM. Building reliable EV charging networks: Model state grant and procurement contract provisions for public EV charging. Northeast States for Coordinated Air Use Management. Boston, MA. May 2019. https://www.nescaum.org/documents/model-contractprovisions-for-public-evse-5-24-19.pdf/
NREL AFDC Electric Vehicle Charging Station Locations. Accessed 3/3/2022: https://afdc.energy.gov/fuels/electricity_locations.html#/find/nearest?fuel=ELEC
NREL Charging Infrastructure Terminology. Accessed 3/3/2022:
https://afdc.energy.gov/fuels/electricity_locations.html#/find/nearest?fuel=ELEC;
https://afdc.energy.gov/fuels/electricity_infrastructure.html
NYSERDA 2021. DCFC Program, Program Opportunity Notice (PON) 4509. Accessed 3/3/2022: https://portal.nyserda.ny.gov/servlet/servlet.FileDownload?file=00Pt000000YDm19EAD
OCPI 2.2-RC1. Open Charge Point Interface, version 2.2-d2, 06-12-2020. Accessed 3/23/22. https://evroaming.org/app/uploads/2019/02/OCPI-2.2-RC1.pdf
ORNL. A toolkit of reliability measurement practices. Oak Ridge National Laboratory, June 2004.https://info.ornl.gov/sites/publications/Files/Pub57467.pdf
Plug In America. The Expanding EV Market. Observations in a Year of Growth. February 2022. Los Angeles.
SAE. SAE electric vehicle and plug in hybrid electric vehicle conductive charge coupler. J1772_201710. 2018-10-13. Accessed 3/28/2022. https://www.sae.org/standards/content/j1772_201710/

Authors: Address correspondence to David Rempel, Department of Bioengineering, University of California, Berkeley, 1301 S. 46th Street, UC Berkeley RFS Building 163, Richmond, CA 94804, USA; e-mail: david.rempel@ucsf.edu.

Source URL: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4077554

Hybrid Switch for Capacitors Bank used in Reactive Power Compensation

Published by Adam RUSZCZYK1, Krzysztof KÓSKA1, Konrad JANISZ2
ABB Corporate Research Centre, Poland (1), AGH University of Science and Technology, Poland (2)

Abstract. This paper describes new concept of a switch dedicated for control of three-phase capacitor bank used for low voltage, reactive power compensation. The proposed device utilizes fully controlled IGBT transistors that gives possibility to break capacitor current in moment that ensures zero voltage remaining on capacitor’s terminals. This minimize voltage stress for switch and capacitor as well as allows to turn-on again capacitors bank without unnecessary delay. This paper describes structure of hybrid switch and its operation with capacitor bank and detuning inductor.

Streszczenie. Artykuł opisuje nową koncepcję łącznika dedykowanego do załączania banku kondensatorów używanych do kompensacji mocy biernej w sieciach niskiego napięcia. Zaproponowany łącznik wykorzystuje tranzystory IGBT, co umożliwia przerywanie prądu kondensatora w chwili, gdy napięcie na zaciskach kondensatora jest równe zero. Zmniejsza to stres napięciowy łącznika i kondensatora oraz pozwala na ponowne załączenie banku kondensatorów bez zbędnego opóźnienia. Artykuł opisuje strukturę łącznika hybrydowego oraz jego pracę z bankiem kondensatorów i dławikiem wygładzającym. (Łącznik półprzewodnikowy dla kondensatorów używanych w kompensatorach mocy biernej).

Keywords: Solid-state switch, Active clamping protection circuit, Reactive power control.
Słowa kluczowe: Łącznik półprzewodnikowy, Układ aktywnego ograniczenia napięcia Sterowanie mocą bierną.

Introduction

The circulation of the reactive power in the system causes different types of power quality effects therefore the generation of reactive power by consumers is restricted by the electricity supplier. The inductive power generation in the most cases is compensated by attaching capacitor banks. It is very simple, inexpensive and effective way of compensation. The drawback of the method is that the reactive power compensator (RPC) cannot compensate reactive power in continuous way due to a discrete value of the capacitors. The typical approach to solve this inconvenience is to split total capacitance of reactive power compensation (RPC) system into smaller blocks that can be simply connected or disconnected to the grid in order to match the needed capacitive power [1], [2].

The most common RPCs use electromagnetic relays to switch capacitor banks. One of the advantages of relays is low contact resistance, and thus low conduction power losses. On the other hand electromagnetic relays have undetermined operational delay and there is no possibility to synchronize them with zero crossing of the voltage. As a result large inrush currents occurs during connection of capacitor to the grid. Typical approach to this problem is utilization of auxiliary contacts with initial pre-charge resistors which are bypassed by main contacts for normal operation. This simple solution damps current to smaller values, but it is far from ideal because inrush current still exists [3].

A much more elegant solution uses solid-state switches (thyristors) as a replacement for relays. Firstly, the moment of turn-on of a semiconductor switch can be precisely controlled at the moment of the zero-voltage crossing [4]. Secondly, thyristors are characterized by relatively low forward voltage drop (approximately 1,5V) and, in consequence, low conduction losses in comparison to the other semiconductor components that are able to withstand voltage higher than 1,4kV. One of the thyristor’s drawback is its lack of turn-off capability. Silicon Controlled Rectifier (SCR) commutates when conducted current drops below, so called, holding current. In AC grid, a capacitive load is turned off during zero crossing of the current and hence peak value of the voltage. Because of voltage remaining across the capacitor, in the following grid voltage period, stress across the thyristor can reach approximately twice phase-to-phase peak voltage.

In this paper the new hybrid, IGBT based, three-phase switch is presented. Presented switch can break current of capacitor in its peak value that corresponds to zero voltage remaining at the capacitor.

Fig.1. Diagram of the three-phase switch for capacitor bank composed of two physical switches

Switching device is dedicated for three phase compensator bank. It comprises two physical switches (Fig.1) which break currents in two phases. This is enough to disconnect or connect three-phase delta-connected capacitor bank. Both switches comprises bidirectional solid-state switches and electromechanical relays connected in parallel (Fig. 2). Similar hybrid solutions are commonly utilized in applications where high conduction losses are not acceptable [5], [9].

Fig.2. Diagram of hybrid switch structure as a combination of electromagnetic relay and IGBT transistor

Hybrid solution combines benefits of a relay (low conduction losses) and semiconductor (synchronized turn- on and -off, as well as arc-less operation). The drawback is slightly increased cost and complexity of the device. Three phase switch, presented in Fig.1 and Fig.2, enables a connection of the capacitor bank without inrush current. The same switch is able to disconnect capacitors bank in a manner that afterwards all three capacitors are completely discharged (Fig.6). This result can be achieved by use of fully controlled switches S1 and S2 that can break capacitors’ currents in specific moments.

Significant problem appears when compensating capacitors are connected to a distorted grid. In that case the capacitor creates low impedance path for high-order harmonic current flow. It’s dangerous phenomenon which may cause serious consequences in the power quality. To avoid such situation the detuning inductors are connected in series with capacitor bank (more details are described in the Section – Filtering of high-order harmonics).

The overvoltage spikes are generated during the disconnection process of the capacitor by interruption of the current in the circuit with additional detuning inductance. It has destructive influence on the semiconductor switch with IGBT component. In order to prevent break-over of the transistor, active clamping circuit is proposed and described in Section – Overvoltage inducted during current interruption.

Thyristors based solid-state switch for capacitor bank

The result of operation of the three-phase switch based on thyristors components is presented in Fig. 3. Capacitor switch is composed of two independent physical switches (Fig.1). Both physical switches are made of two antiparallel connected thyristors in order to form a bidirectional valve. S1 and S2 work independently and are turning-on when the voltage seen across the switch is crossing zero value. The consequence of this fact is that the physical switches works in sequence. This ensures that the capacitor bank can be connected without inrush current. Moreover, transient states observed in phase currents are greatly reduced in comparison with electromechanical relay solution.

The commutation delay between S1 and S2 equals 90 deg. (5ms) for both turn-on and turn-off operation. It may look odd that phase delay between two switches in three-phase circuit is exactly 90 deg. (5ms) neither 60 nor 120 degrees. The reason of a such behavior is explained on voltage graphs shown in Fig. 4

Fig.3. Waveforms of voltages across capacitors (top); line currents (mid) and control signal (bottom) during turn-on and turn-off process. Capacitor bank are controlled by thyristor based switch.

Let’s assume that both switches S1 and S2 are in blocking condition of phase-to-phase voltages UL12, UL32 respectively. All capacitors are discharged. Let’s start with S1 closing it at first the zero crossing of the voltage UL12. Then capacitors C1, C2 and C3 are charged up. In series connected capacitors C2 and C3 create parallel branch to C1 and form a voltage divider for UL12 voltage. During this period switch S2 is connected between phase UL3 and midpoint of UL12 voltage. This voltage is in phase to UL3, but its amplitude is √3/2 higher according to height of an equilateral triangle created by phase-to-phase voltages (Fig.4).

Fig.4. Voltage vectors at capacitors during conduction and turn-off process

Switch-off process is analogical to switch-on, but the difference is that non-zero voltages remain at capacitors terminals afterwards. In the following grid period (20ms) switch has to withstand sum of grid amplitude voltage and capacitor’s voltage. In consequence of it switches and capacitors have to be rated for higher voltage that makes a practical system more expensive. This problem is unsolvable with use of thyristors components for solid-state switch because they break capacitors’ current near zero current condition, so maximum voltage. This situation is presented in Fig.3.

IGBT based solid-state switch for capacitor bank

Proposed solution uses IGBT instead of thyristors. To achieve bidirectional operation transistor is connected with single phase diode bridge (see Fig. 2 and Fig.12). IGBT is turned-on exactly like thyristor at zero crossing of the voltage. This prevents an occurrence of inrush current. But unlike thyristor, the

IGBT is turned-off in the moment when voltage, which is measured at selected capacitor, is close to zero. The switch, which operation is based on the described principles has been designed and tested. It proves that operation of two physical switches is enough to achieve zero voltage across capacitors after turnoff. This device is presented in Fig.5. In parallel to IGBT based switches electrometrical relays are connected.

Fig.5. Hybrid switch for three-phase capacitor bank.

Presented hybrid switch is equipped with a simple control logic implemented on CPLD which is responsible for synchronization with the grid and generation of control signals for IGBTs and relays.

The turn on-off operation of the switch is presented in Fig.6. It presents capacitor voltages. One can observe that capacitors are discharged before and after operation of the switch. That definitely reduces voltage stress at capacitors and switches after disconnection. Moreover, it is not necessarily to wait until capacitors will be discharged before next operation. User is able to turn on the capacitor bank again without time restrictions. Dynamics of the presented hybrid solution is comparable with thyristor-based switch.

Fig.6. Capacitors’ voltages (top) and CPLD synchronization signal (bottom)

The consequence of capacitor turn-off with zero voltage condition is interruption of non-zero current. Capacitor’s current waveforms during disconnection are presented in Fig.7. In ideal condition, when current is interrupted in a circuit that has only capacitive character the current can be interrupted immediately. The current interruption in a circuit where even the smallest inductance exists a voltage spikes will be induced.

Fig.7. Line currents waveforms during disconnection at zero voltage condition
Filtering of high-order harmonics

The known problem with capacitor attached to the distorted network is a generation of high-order harmonics of current. This problem exists no matter of type of used switch technology: electromechanical, thyristor or IGBT. Line current of single-phase of R

load with purely capacitive compensator is presented in Fig.8a. The reactive power compensation reduces reactive power flow for fundamental harmonic but at the same time increase high-order current harmonics. Because of existence of the line impedance the distorted current causes the additional voltage drop that increases voltage distortions.

The solution is to use of detuning inductor installed in series with capacitor bank. Properly selected detuning inductor causes smoothing the capacitor current. This effect can be observed in Fig.8b.

Fig.8. The grid voltage (Ch1) and line current (Ch2) of the load with purely capacitive compensator (a) and compensator with detuning reactor (b) of reactive power

This justifies the necessity of use of detuning inductor. Detuning inductance L connected in series with compensating capacitor C creates series resonant circuit with a resonant frequency below the 5th (or 3rd) order harmonic, which is the most common in a harmonic-rich environment. In Europe, detuning by a factor of 3.78 (7%) times the line frequency is most common, whereas in other parts of the world, in particular in Asia, a factor of 4.08 (6%) is more often selected. For high demanding systems 2,83 (12,5%) or even 2,67 (14%) factor is used.

Fig.9. Frequency response of capacitive filter with different values of detuning inductor (XL is given in % of capacitor reactance XC)

To fulfill resonance frequency requirements each capacitor bank in RPC must be equipped with separated detuning reactor with properly selected detuning inductance. As it is shown in Fig.9, the LC filter operating below resonant frequency is in capacitive mode and above it in inductive mode.

Overvoltage inducted during current interruption

The detuning inductor is an effective solution for high order harmonics rejection. However, existence of additional inductance in series with compensating capacitor creates serious problem for IGBT-based switch. During interruption of the current an overvoltage is inducted which can break over the structure of semiconductor. Every physical circuit has small inductance introduced by connection wires, so even a lack of detuning inductor do not allow to neglect this problem. Fig.10 shows two examples of line current without (Fig.10a) and with detuning inductor (Fig.10b).

In Fig.10b the effectiveness of higher harmonics filtration can be observed when detuning inductor is used. In Fig.10a short voltage spikes of few microseconds duration are visible even with lack of detuning inductor. In both cases the overvoltage spikes were limited by the surge arresting circuit which protects IGBTs.

Fig.10. Line current without (a) and with (b) detuning inductor during S1 and S2 switch off with clearly seen inducted overvoltage ; Ch1 – switch voltage, Ch2 – line voltage and Ch3 – line current

The most common overvoltage protection device is metal oxide varistors (MOV’s). Although MOV seems to be good solution in many applications, the utilization in IGBT based switch is far from ideal.

MOV are dedicated for incidental operation as a surge arrester. The structure of metal-oxide degrades with every action cause degradation in the metal oxide material, which eventually leads to component failure. Theoretically, according to [10] low energy pulses can be suppressed infinitive number (Fig.11). But it is hard to ensure that the current magnitude and time duration will remain unchanged when the impedance of compensating branch may vary due to a capacitance and inductance change. Therefore the operating point for MOV can be moved into limited lifespan region (Fig.11).

Because varistors only dissipate a relatively small amount of average power they are not suitable for repetitive applications that involve substantial amounts of average power dissipation.

Additional limiting factor is ambient temperature that forces derating of surge power. To ensure long time of trouble less operation for solid-state switch the size of MOV has to be carefully selected. For the most cases it means the MOV has to be oversized.

Fig.11. Repetitive Surge Capability for 20mm Parts – Littlefuse [10]

In solid-state switch the overvoltage is present during every turn-off of IGBT, so after limited number of cycles MOV may fail. In this paper it is proposed to use transistor active clamping circuit, in which IGBT tries to protect itself by reducing di/dt of interrupted current in order to limit induced voltage to the safe level. This system is described in next chapter.

Transistor Active clamping circuit

Break of the load current in inductive circuit generates voltage equal UL=–L(diL/dt). It means that derivative of the current has to be limited to keep induced voltage below maximal. It is achieved by additional circuit composed of in series connected high voltage Zener diodes (Fig.12) connected between IGBT’s emitter and gate terminals.

Fig.12. Active clamping overvoltage circuit made of in series connected Zener diodes

This kind of protection circuit is commonly used with high power IGBT transistors [7], [8]. When IGBT is turning off an inductive load and inducted voltage UL exceeds the voltage threshold set by Zener diodes and the IGBT is driven back into conductive state by current injected into the gate. In fact IGBT remains in active state during current interruption and can be interpreted as variable resistance. The main drawback of this method is that all energy stored in inductance has to be intercepted by internal IGBT silicon structure.

In the laboratory setup with three capacitors 62μF in delta connection (Qc = 10 kvar @ ULL= 400Vac, f = 50Hz) to achieve 7% detuning reactance three phase choke has been used. Nominal inductance of this choke is 3,84mH per phase. Laboratory verification was made by interruption of instantaneous current of 27A. Worse switching condition has switch S2 that has to interrupt current flowing through in series connected inductances in phase L2 and L3. The energy stored in both inductances can be calculated as:

.

where: i – interrupted current, L – detuning inductor.

While the energy absorbed by silicon is an integral of a product of collector current IC and transistor voltage UCE in period of 220μs read from Fig.13

.

where: IC – collector current of transistor, UCE – voltage cross CE junction, Δt – current interruption period.

Fig.13. UCE voltage, IC current and IG current registered for IGBT transistor operation during turn-off process

All energy stored in circuit inductances has to be intercepted by transistor (EL≈ET). Therefore a special type of transistors should be selected. According to the datasheets [11,12] for two type of investigated IGBT’s the maximal acceptable energy is calculated in Table 1.

Table 1. The IGBT parameters comparison

.

The discharge of energy of inductance in the IGBT transistor takes about 220μs. It is definitely too short to transfer any heat outside. The process can be treated as adiabatic. The thermal image (Fig.14) confirm that there is no visible increase of transistor’s temperature.

The HGTG27N120BN has approximately 0,036g of silicon [13]. Specific heat of the silicon equals 0,7 J/(goC). Thus, the temperature rise of the silicon equals approximately 110oC. Fortunately, heat transfers to the copper lead frame of the transistor with relatively short time constant. Copper weights 4,0g [13]. That for specific heat capacitance of copper equal 0,386 J/(goC) gives 1,57 J/oC thermal capacitance of transistor in TO-247 package. In other words a single portion of 2,8 J of energy from detuning inductor would cause increase temperature of transistor about 1,78oC. Even the turn on/off cycle realized every second cannot increase significantly temperature of IGBT enclose.

Fig.14. Thermal image registered for IGBT transistor operation during turn-off process. Surface temperature – 28.2oC
Conclusion

Paper presents a new concept of hybrid switch which is dedicated for capacitive reactive power compensator. Single device is made of two physical switches installed in two phase lines. It has been experimentally proven that proposed switch is able to disconnect a delta connected capacitor bank in manner that afterwards all three capacitors are completely discharged. Moreover it has been showed that conduction losses can be reduced by introducing hybrid solution with parallel electromechanical relay.

Finally the IGBT overvoltage protection allows to use the presented switch in RPC systems with detuning inductors.

REFERENCES

[1] Gos z towt W.: “Gospodarka elektroenergetyczna w przemyśle”. Warszawa WNT, 1973
[2] Nar tows ki Z. Baterie kondensatorów do kompensacji mocy biernej. Warszawa WNT, 1967
[3] Application Guide: Contactors for capacitor switching, 1SBC101140C0203 2009 ABB
[4] Olivier G., Mougharbel I., Dobson-Mack G.: “Minimal transient switching of capacitors”, IEEE Trans. on Power Delivery, vol. 8, no. 4, 1993, pp. 1988-1994.
[5] Bachman P.: “Crydom RHP Series – 3 Phase Hybrid Solid State Contactor”, White Paper CRYDOM Inc. 2009
[6] Ironcore – Reactors Catalogue http://www.mangoldt.com/pdf/ HvM_Ironcore_Reactors_Catalogue_2011_ENG.pdf
[7] Garcia O. , Thalheim J . , Meili N. : “Safe Driving of Multi-Level Converters Using Sophisticated Gate Driver Technology”, PCIM Asia, June 2013.
[8] Bur khard B. : “Switching IGBTs in parallel connection or with enlarged commutation inductance”, PhD thesis, Bochum 2005
[9] Shukla A., Demetriades G. D.: “A Survey on Hybrid Circuit-Breaker Topologies”, IEEE Trans. on Power Delivery, Vol. 30, No. 2, April 2015, pp. 627-641
[10] Metal-Oxide Varistors (MOVs) – UltraMOVTM Varistor Series – © 2015 Littelfuse, Inc. – Specifications Revised: 08/20/15
[11] IRG4PH40KD – Insulated gate bipolar transistor with ultrafast soft recovery diode – datasheet
[12] HGTG27N120BN – 72A, 1200V. NPT Series N-Chanel IGBT
datasheet – obsolete product.
[13] AN-7516 – Safe Operating Area Testing Without A Heat Sink

Authors: dr inż. Adam Ruszczyk, ABB Corporate Research Center, ul. Starowiślna 13A, 31-038 Kraków, Poland, E-mail: adam.ruszczyk@pl.abb.com; mgr inż. Krzysztof Kóska, ABB Corporate Research Center, ul. Starowiślna 13A, 31-038 Kraków, Poland, E-mail: krzysztof.koska@pl.abb.com; inż. Konrad Janisz, Akademia Górniczo-Hutnicza,

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 91 NR 12/2015. doi:10.15199/48.2015.12.03

Power Quality – IEEE 519-2022

Published by Comsys AB, website: comsys.se, ADF Technology: Power quality – IEEE 519-2022

A commonly used and very important standard is IEEE 519-2022 (previous versions IEEE 519-1992 and IEEE 519-2014). The standard, among other things, puts two requirements on harmonics; and absolute maximum THDU level, and a variable maximum TDD level. All limits are applied to the Point of Common Coupling (PCC), which is the interface between utility (sometimes called operator) and consumer. The PCC can be located at any voltage level. In some cases, the PCC is considered to be an internal point in a system of particular interest; this is not in line with the original intention of the IEEE 519, which considered only the connection point between operator and user (consumer). These concepts are illustrated in the figure below:


IEEE 519 standard. Image by Comsys

Below is Table 1, from IEEE 519 (2022), p17, “Voltage distortion limits”, as outlined below:

Table 1. (IEEE 519-2022, pg.17) voltage distortion limits

Bus voltage V at PCCIndividual harmonic (%)Total harmonic distortion THD (%)
V ≤ 1.0 kV5.08.0
1 kV < V ≤ 69 kV3.05.0
69 kV < V ≤ 161 kV1.52.5
161 kV < V1.01.5*
.

*High-voltage system are allowed up to 2.0% THD where the cause is an HVDC terminal whose effects will have been attenuated at points in the network where future users may be connected.

Note that these levels are absolute, and not depending on the size of the operator/utility or the consumer. Also note that the resulting distortion level is the result of the combination of the background distortion and the load distortion created by the consumer.

Following below is an excerpt from Table 2 (IEEE 519-2022, pg. 19, replacing table 10.3, p78, “Current Distortion Limits for General Distribution Systems” in IEEE 519-1992). This table is of importance as it defines target levels to be achieved depending on the short circuit ratio ISC/IL. ISC is the rated short circuit current at PCC, and IL is the maximum demand load current at PCC.

Table 2. (IEEE 519-2022, pg.19) current distortion limits for systems rated 120 V through 69 kV

ISC/ILHarmonic limits a,b
2 ≤ h < 11		Harmonic limits a,b
11 ≤ h < 17		Harmonic limits a,b
17 ≤ h < 23		Harmonic limits a,b
23 ≤ h < 35		Harmonic limits a,b
35 ≤ h ≤ 50		TDD Required
<20c4.02.01.50.60.35.0
20<507.03.52.51.00.58.0
50<10010.04.54.01.50.712.0
100<100012.05.55.02.01.015.0
>100015.07.06.02.51.420.0
.

a For h≤ 6, even harmonics are limited to 50% of the harmonic limits shown in the table.
b Current distortions that result in a dc offset, e.g., half-wave converters, are not allowed
c Power generation facilities are limited to these values of current distortion, regardless of actual Isc/IL unless covered by other standards with applicable scope.  Where: 

ISC = maximum short-circuit current at PCC 

IL= maximum demand load current at PCC under normal load operating conditions

Note the major difference in how harmonics are limited at the current and at the voltage. For voltage harmonics, all requirements are absolute. For current harmonics, the authors of IEEE 519 chose to limit the current harmonics depending on how strong the voltage source is. This is reasonable and understandable; a strong grid will be able to suppress current harmonics to a much larger degree without the voltage being influenced than a weak grid. In very weak grids, voltage distortion and current distortion may have similar values. Hence, it can be argued that current emission requirements must be stricter in weaker grids.

Total demand distortion

Total Demand Distortion (TDD) is defined as the ratio of the root mean square of the harmonic content, considering harmonic components up to the 50th order and specifically excluding interharmonics, expressed as a percent of the maximum demand current. Harmonic components of order greater than 50 may be included when necessary.

THDI uses the instantaneous fundamental current as reference. TDD uses the maximum demand current (maximum current) as reference. This means, at 100% load THDI = TDD. The difference between THD and TDD can be quite dramatic, as illustrated below.

TDD, not THDI. Be sure to make it clear if the requirement from the customer is TDD or THDI before specifying your ADF size. Preferably only specify THDI at 100% load or use TDD instead!

Source URL: https://comsys.se/our-adf-technology/power-quality-ieee-519-2022/

Tropical Climate Effects on Corona Power Losses on 275 kV Transmission Lines in the South Sulawesi System

Published by Ikhlas KITTA, Salama MANJANG, Ida RACHMANIAR, Faris MARICAR
Electrical Engineering Department, Hasanuddin University, Indonesia

Abstract. Electricity losses are very dependent on electric current and loss of the corona phenomenon. It is very clear that losses depend on network parameters, load behaviors and climatic factors. South Sulawesi is a tropical climate. This sector must be resistant to exposure to various factors of high tropical climate such as temperatures of 23.4⁰C – 33.3⁰C, sun irradiation that occurs more than 12 hours per day, relative humidity close to 100%, and average rainfall between 440-1322 mm . This climate factors will simultaneously transmit the transmission. This text aims to connect climate factors to achieve power losses on 275 kV transmission line in South Sulawesi. The results of this study found that temperature and duration of sun irradiation affected corona power losses on 275 kV transmission line in South Sulawesi.

Streszczenie. Straty elektryczne są bardzo zależne od prądu elektrycznego i zjawiska korony. Oczywiste jest, że straty zależą od parametrów sieci, obciążenia i czynników klimatycznych. South Sulawesi to tropikalny klimat. Sektor ten musi być odporny na działanie różnych czynników tropikalnych, takich jak temperatura 23C – 33C, promieniowanie słoneczne, które występuje więcej niż 12 godzin dziennie, wilgotność względna bliska 100%, a średnie opady wynoszą 440-1322 mm. Niniejszy artykułma na celu analizę czynników klimatycznych w celu uzyskania strat mocy na liniach przesyłowych 275 kV w Południowym Sulawesi. Wyniki adania pokazały, że temperatura i czas trwania promieniowania słonecznego wpłynęły na straty energii z wyładowań koronowych. Wpływ klimatu tropikalnego na straty mocy koronowej na liniach przesyłowych 275 kV w South Sulawesi.

Keywords: tropical climate, corona loss, transmission lines, South Sulawesi
Słowa kluczowe: klimat tropikalny, strata koronowa, linie transmisyjne, South Sulawesi

Introduction

The power system consists of generating units, transmission lines and distribution networks. The transmission network is considered as the backbone of the electric power system that connecting the power plant center with the load center. In general, the transmission line carries an electric current that reach hundreds of kilometers. The entire transmission system is interrelated due to economic reasons, security and reliability which is a transmission line requirement based on system planning [1]. Every time the generating unit is added to the system, there is a need for a transmission line to transfer power from the generating station to the load center. But the longer the transmission line is used, the greater the electrical power losses in the transmission line so that the electrical power that reaches the destination has been reduced which causes the transmission line efficiency to be low and the transmission line voltage regulation becomes high. To avoid this, the option is to increase the voltage on the transmission line from high voltage level to extra high voltage.

Therefore, one of the efforts to reduce electric power losses and improve the quality of stress in the province of South Sulawesi, one of the provinces in the country of Indonesia, or often called the South Sulawesi system, is insertion of 275 kV transmission network. In addition to these reasons, the application of the 275 kV transmission line is to connect power plants in the area of renewable electricity to the load center in South Sulawesi, namely the City of Makassar (the capital of South Sulawesi province). South Sulawesi has many primary energy sources, especially in the form of hydropower which can be developed into hydroelectric power plant. Hydropower potential that can be developed into around 1996 MW [2].

Indonesian, especially South Sulawesi, is on the equator line having a tropical climate, precisely the wet tropical climate. This is also influenced by the shape of the Indonesia condition which is an archipelago. Most of the land in Indonesia is surrounded by oceans. That is why Indonesia has a climate of sea that is moist and has a lot of rains.

The geographical location of South Sulawesi makes this area vulnerable to natural and environmental disasters, which drastically affect the transmission network in the area. The northern part of the area is hilly and mountainous, while the southern part is low land which has a tropical climate with high humidity and rainfall. The consequence is the occurrence of corona power losses (Pc) in the transmission network conductor [3], especially when applied extra high voltage on the network that is 275 kV voltage.

Fig.1. The map with the location of 275 kV transmission line in South Sulawesi [2]

The conditions that affect corona power losses are air movement, air temperature and humidity [4]. Atmospheric influences affect greatly the corona losses [5].The ionization process will stop if the electric field decreases. The effects of corona power losses are noise interference, frequency interference, interference with electronic equipment performance [6].

South Sulawesi is a tropical region, so the use of transmission networks for the distribution of electrical energy must be resistant to exposure to various factors of high-intensity tropical climates such as ultraviolet radiation from the sun about 12 hours during the day, air temperatures between 16-35 °C, relative humidity approaching 100% between early morning and early morning and high annual rainfall between 40-1000 mm. These factors will simultaneously hit the 275 kV transmission line, so that through this paper an explanation of the influence of tropical climate climatological parameters on corona power losses on 275 kV transmission line in South Sulawesi is explained.

Object Analysis

The case that we analyzed in this study is the South Sulawesi electricity system which is devoted to 275 kV transmission line with ACSR type (Gannet) for about 195 kmc. The location of the 275 kV transmission line is shown in Fig.1, where the base of the transmission line is in Palopo city, and the end point is in Makassar city.

Climatology Conditions of South Sulawesi

The transmission line in the South Sulawesi power system has a voltage of 150 kV and 275 kV which already exceeds 1000 kmc. This South Sulawesi system is strongly influenced by Indonesian tropical climate conditions. The northern part is a mountainous area that contains tropical rain forest which is a source of water that flows through a network of rivers and creeks. This northern part is a potential place for renewable energy for hydro power plant. In the central part there is a vast plain for agriculture that has strong wind potential so it has been used as a wind power plant for 70 MW. Furthermore, in the southern part which is a mangrove forest tropical climate with high humidity and rain falls throughout the day. This southern part is the center of South Sulawesi community activities, in Makassar city. Part of transmission network is located in the South region. This area experiences low humidity and rainfall because it is located on the seafront, so it is affected by water vapor from the sea. The Northern Region is dominated by tropical forests with high humidity and rainfall. Air in South Sulawesi both in the north and south is not a perfect insulator, because air contains electrons and ions as a result of various effects such as solar ultraviolet light and sea water evaporation.

The province of South Sulawesi is in the equatorial region which is affected by tropical climates, where the characteristics of the tropical climate is: The temperature is quite high every year, the average air temperature is not less than 18 °C or around 27 °C, during the rainy season or dry season there is no difference that is very far or almost the same, day and night duration looks almost the same, that is around 12 hours a day and about 12 hours a night.

South Sulawesi is a part of Sulawesi Island with an astronomical location in South Sulawesi located at 0 on 12′ South Latitude to 8⁰ North Latitude, and 116⁰ 48′ West Longitude up to 122⁰ 36′ East Longitude. The climate in South Sulawesi is recorded in the South Sulawesi Climatology Station that the temperature throughout 2017 ranges from 23.4 ⁰C – 33.3 ⁰C and the average rainfall is 440 mm to 1322 mm per year. There is significant rainfall in most months of the year. The climatological data are shown in Table 1.

Table 1. Climatology Data of South Sulawesi in 2017 [7]

.
Analysis Model

In this study, the procedures carried out are:

1) The preparation phase, which is the estimation of what component structure will be used to modeling the corona power losses in the South Sulawesi transmission system.

2) Literature study by studying the literature on modeling loss corona power loss, corona effect on power losses in the transmission system, and also the influence of tropical climate on the large corona power losses that occur in a transmission system.

And 3) Data collection.

Basically, power losses in the transmission network result from transmission lines and transformers at the substation. Transformer loss is the amount of losses in the winding and core losses which are expressed in the form of hysteresis and eddy currents [8]. These losses are released in the form of heat energy. The power losses of the transmission line are very dependent on the magnitude of the network electric current and the losses of the corona phenomenon. In addition, the power losses of the transmission line are affected by changes in the configuration of the electrical system because of the consequences of power outages, maintenance and development of the electrical system. So it is very clear that power losses in transmission lines depends on network parameters, load behavior and climate factors.

Corona power losses generally occur at extra high voltages, in climates that experience low pressure, high temperatures, stormy weather, and rain [9]. This is also the result of a larger conductor size, with a rough and uneven surface which results in a lower critical disruptive voltage. Corona power losses do not occur if the distance between conductors is very large. Therefore, high voltage transmission lines are made with two, three or four conductor bundles where the average geometric radius of the conductor is enlarged.

Furthermore, the loss equation due to the corona in the high voltage transmission line is explained. The corona power losses are expressed by the equation which is the result of research conducted by Peek’s [4]:

.

where: Pc = transmission line corona power losses (kW / km / phase), δ = relative air density, f = frequency of the electrical system (Hz), r = radius of the transmission line conductor (cm), D = distance between the transmission line conductor (cm), V = phase voltage to neutral transmission line (kV), Vd = critical voltage (kV). In analyzing the influence of tropical climates due to the corona of high voltage air power losses, a relative air density value is needed where the relative air density is affected by the air pressure and the surrounding temperature of the transmission line conductor. The air density equation is shown in equation (2) [10].

.

where: δ = relative air density, P = air pressure (mmHg), and T = temperature around the transmission line (°C).

Furthermore, for the disruptive voltage shown in equation (3). Disruptive stress arises due to the emergence of electric field strength due to the collision of electrons in the ionization process. The critical voltage is disruptive considering the influence of conductor factors, conductor and environmental surface uniformity as observed by Peek’s. The critical disruptive voltage value at which the corona begins to form is expressed by:

.

where: Vd = critical disruptive voltage per transmission line phase (kV), Ec = penetrating air voltage gradient (kV/cm), m = indefinite factor, r = radius of the transmission line conductor (cm), δ = air density relative, and D = distance between transmission line conductors (cm).

The above equations are used in analyzing the influence of tropical climate on the corona power losses at the transmission line in South Sulawesi.

Tropical Climate Effects on Korona Power Losses

Based on the climate data of South Sulawesi, the trend of annual temperature is flat. As with the characteristics of the tropical climate in general, the temperature of each month does not experience large fluctuations. In January, the average temperature was the coldest compared to other months in one year, which was 26.7 °C. While September is the hottest month in a year, with an average temperature of 28.3 °C. From here it can be seen that January is the coldest month, and September is the hottest month.

Corona power loss calculation data based on temperature data for the calculation of average air density per month taken from Table 1 is shown in Fig.2.

Fig.2 shows corona power losses throughout the year based on changes in temperature, where there are graphs based on maximum temperature per month, minimum temperature per month, and average temperature per month. The average corona power losses for the maximum temperature are 2.02 MW, 1.37 MW for the minimum temperature, and 1.63 MW for the average temperature. In the corona power loss chart averaged per month of 275 kV transmission line between Palopo-Makassar, there is no significant increase in corona losses in each month. From Fig.2 also seen in September is the month that has the largest average corona losses of 2.14 MW.

In each corona loss chart throughout the year, the change in the magnitude of the losses is only 0.23 MW for the maximum temperature graph, 0.14 MW for the minimum temperature graph, and 0.12 MW for the average temperature graph.

Seen in Fig.2, the difference in power between maximum temperature and minimum average temperature is 0.65 MW. If the difference in losses occurs all the time, it will cause the operation of the South Sulawesi system to experience fluctuations in the supply of generating power that must be prepared at least equal to the corona losses.

Fig.2. Corona power losses on the 275 kV transmission line in South Sulawesi based on temperature changes
Corona Power Loss Relationships with Humidity

The temperature fluctuations in South Sulawesi are caused by changes in other climatological parameters, namely relative humidity, solar irradiation time, and rainfall. For this reason, it is shown how the relationship between these parameters is to the change in value of corona power losses throughout the year.

The tendency of relative humidity in South Sulawesi in one year is not much different from the air temperature, which is flat, does not experience significant fluctuations. This is mainly seen from the relative relative humidity each month in one year. Table 1 presents the relative relative humidity value, where the highest humidity in January is 89%, while the lowest relative humidity is in September, which is 70.33%.

Fig.3. Graph of corona power losses and relative humidity on 275 kV transmission lines in South Sulawesi

Judging from Fig.3, it is known that the value of corona power losses has a decreased linear with relative humidity values, where corona power losses tend to increase when the relative humidity decreases. The graph of relative humidity relationship with an increase in the value of corona power losses during the year is shown in Fig.4, where the equation y = -0.0033x + 1.8925 with R2 = 0.2058, where y = corona (MW) power losses, and x = relative humidity (%). So it can described that when the relative humidity around the transmission system decreases, the value of the corona power losses will be greater. In January, with air humidity of 89%, the value of its power losses was only 1.57 MW.

Whereas when the relative humidity drops to 87.67%, the value of the power losses will be greater, which is 1.59 MW.

Fig.4. Relation of corona power losses and relative humidity on 275 kV transmission lines in South Sulawesi
The Corona Power Loss Relationship with the Duration of Sunshine

Likewise with the sun radiation parameters in a tropical climate. Based on Table 1, the duration of sun exposure in a tropical climate is throughout the day where every day for 12 hours. There are certain months that the duration of the sun’s radiation is slightly disturbed by the presence of clouds, which occurs in January with a figure of 42.0%. While the longest sunshine duration is in August at 83.3%. So it can be ascertained that in August the sky conditions were very bright, only very few clouds covered.

From the graph of the sun irradiation relationship with corona power losses, linear in the same direction is obtained which tends to be the same (Fig.5). The relationship between these two parameters is shown in Fig.6 where the equation based on linear regression is obtained corona losses tend to be directly affected by the duration of sun irradiation in the 275 kV transmission line environment. The equation is y = 0.0018x + 1.513 with R2 = 0.2704, where y = corona power losses (MW), and x = solar irradiation time (%).

Fig.5. Graph of corona power losses and duration of sun irradiation on 275 kV transmission lines in South Sulawesi

When the sun shining for a long time around the transmission line, the corona power losses will also be greater. This is because corona power losses are affected by air temperature which affects relative air density. Therefore, when in January the duration of solar radiation was 42%, the value of its power losses (Pc) was only 1.57 MW. Whereas when the duration of sun exposure is 51.3%, the value of its power losses (Pc) will be even greater, which is 1.59 MW.

Based on the relation of relative humidity and duration of sun irradiation in South Sulawesi, a description of these corona power losses can be made, namely the corona power losses are very dependent on the climatic conditions around the transmission line, where the relative humidity decreases and the duration of sun exposure around the transmission line will accelerate the ionization process so that it causes corona. The decrease in humidity and the duration of sun exposure will affect the air pressure around the transmission line.

Fig.6. Relation of corona power losses and the duration of sun circumference on 275 kV transmission lines in South Sulawesi
Corona Power Loss and Rainfall Relation

Because the relative humidity and the duration of the sun irradiation have shown a relationship pattern with corona power losses approaching linear, then it is reviewed how rainfall affects the magnitude of the corona power losses.

Rain occurs almost all year in tropical climates. Table 1 shows that every month in 2017 there is rain in South Sulawesi. Only 4 months in one year that has little rainfall, ie from August to October. The least rainfall is in August with a value of 440 mm. While in other months it has high rainfall. The highest rainfall is in January with a value of 1322 mm.

Fig.7. Graph of corona power losses and rainfall on 275 kV transmission line in South Sulawesi

If the analysis is based on the relation of rainfall graphs and corona power losses (Fig.7), there is no linear relation pattern, where the pattern of rainfall magnitude fluctuates from large to small which produces a non-linear graph trend. In contrast to corona power losses, the graph tends to be linear. Therefore, changes in rainfall values that occur in the South Sulawesi region for one year are not seen to be directly related to the magnitude of the corona power losses. The largest and smallest value of corona power losses occurs not together with the high and low rainfall values in the South Sulawesi. The relationship between the two parameters is shown in Fig.8, where an equation of the results of linear regression is formed which describes corona power losses not directly affected by the amount of rainfall in the 275 kV transmission line environment. The equation is y = -0.0001x + 66.499 with R2 = 0.0646, where y = corona losses (MW), and x = rainfall(mm).

However, from the results obtained, it can be described that in high and low rainfall conditions that fluctuate throughout the year it affects air humidity, air temperature and air pressure, so that it can affect the high and low air density factors which make the value of corona losses in the transmission line also change in at that time.

Fig.8. Relationship of corona power losses and rainfall on 275 kV transmission line in South Sulawesi
Conclusion

From the results of the analysis and discussion it can be concluded that the corona power loss (Pc) is directly affected by the temperature and air pressure around the 275 kV transmission line. When the temperature rises, which generally occurs in sunny weather conditions, corona power losses will increase, and vice versa. As a result of temperature changes in South Sulawesi, including tropical climates from the maximum temperature to the minimum temperature, a difference in corona power losses of 0.65 MW can occur throughout the year.

The influence of other tropical climate parameters that are quite dominant affecting corona power losses is the duration of sun exposure. The relationship of the duration of sun irradiation with the value of corona power losses is linearly ascending. Similarly relative humidity also influences the magnitude of corona power losses, which are linearly decreasing. The relative humidity and the rainfall which are also tropical climate parameters connected with the magnitude of the corona power losses in South Sulawesi.

Acknowledgment: The authors gratefully acknowledge Indonesia Government of ministry of research and higher education for financial support of this research.

REFERENCES

[1] Bao-hui, Z., Li-yong, W., Wen-hao, Z., De-cai, Z., Feng, Y., Jinfeng, R., Han, X., Gang-liang, Y., 2005. Implementation of power system security and reliability considering risk under environment of electricity market. IEEE/PES Transmission and Distribution Conference & Exhibition: Asia and Pacific Dalian, China.
[2] ESDM Ministry (Indonesia), 2016. PLN electric power supply business plan for 2016 – 2025. Jakarta.
[3] Yahaya, E.A., Jacob, T., Nwohu, M., Abubakar, A., 2013. Power loss due to corona on high voltage transmission lines. IOSR Journal of Electrical and Electronics Engineering (IOSRJEEE), Vol. 8 No. 3, pp 14-19.
[4] Momani, M.A., 2015. Factors affecting corona power losses in Jordan power grid, 2015 Third International Conference on Technological Advances in Electrical, Electronics and Computer Engineering (TAEECE), IEEE.
[5] Kral, V., Rusek, S., Rudolf, L., 2011. Software for calculation of technical losses in transmission network. Przeglad Elektrotechniczny, R. 87 NR 2, pp 91-93.
[6] Loxton, A.E., Britten, A.C., 2002. The measurement and assessment of corona power losses on 400 kV transmission lines. IEEE Africon.
[7] BPS-Statistics of South Sulawesi, 2018. South Sulawesi province in figures. Makassar, Indonesia.
[8] Masoum, A.S., Moses, P.S., 2011. Distribution transformer losses and performance in smart grids with residential plug-in electric vehicles. ISGT, IEEE.
[9] Liu, Y., You, S., Wan, Q., Lu, F., Chen, W., Chen, Y., 2009. UHV AC corona loss measurement and analysis under rain. Proceedings of the 9th International Conference on Properties and Applications of Dielectric Materials, July 19-23, Harbin, China.
[10] Tonmitr, K., Ratanabuntha, T., Tonmitr, N., Kaneko, E., 2016. Reduction of power loss from corona phenomena in high voltage transmission line 115 and 230 kV. Procedia Computer Science, 86, pp 381 – 384.

Authors: Ikhlas Kitta, Salama Manjang, Ida Rachmaniar, Faris Maricar; Electrical Engineering Department, Hasanuddin University; Makassar, South Sulawesi, Indonesia. Address: Jl. Perintis Kemerdekaan Km.10, 90245. Makassar. E-mail: ikhlaskitta@gmail.com, salamamanjang@gmail.com

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 95 NR 1/2019. doi:10.15199/48.2019.01.48

Solving Electric Vehicle Development Challenges

Published by Mark Patrick, EE Power – Technical Articles: Solving Electric Vehicle Development Challenges, July 14, 2022.

A potentially viable approach to solving the development challenges associated with electric vehicles to save both weight and cost.

The growth in EV adoption is a clear and obvious trend, confirmed in several recent industry reports. While the pandemic paused car sales globally, consumers have had time to reflect on their vehicle replacement options, and it appears many are considering an EV as a viable choice. However, many potential obstacles to widespread consumer adoption exist, the most notable being range and charging time anxiety.

Still, thanks to many national and regional government initiatives, and manufacturer special offers, EV sales are set to grow significantly in the coming years.

From the vehicle manufacturer’s perspective, there are still many technical challenges to address moving forward. The initial development of EVs sought to add EV alternatives to existing brand models for the early adopter market. However, sustained and significant growth will require technological advancements on many different fronts.

Battery design, for example, is undergoing major innovation into new chemistries and methods of construction. Although these developments are still in their infancy, there are promising results already. The deployment of a convenient and easy to access EV charging infrastructure requires significant investment too. Range and charging time remain top consumer considerations, along with price, but the key factor impacting these is the vehicle’s weight.

EVs today still maintain a 12 V battery to power non-traction-related functions such as windscreen wipers, seat comfort controls, and infotainment. Some manufacturers are currently replacing a 12 V battery with a 48 V for new models.

The Low Voltage Legacy

Any new vehicle today is equipped with a myriad of electronics-based features, a far cry from when the Hudson Motor Company introduced the concept of a standardized battery in 1918. Today, a sleek touch-controlled infotainment system typically incorporates radio, media players, a GNSS navigation system, smartphone integration, and vehicle status and systems configuration menus. In addition, vehicle occupants can stream music from their smartphone, a high-capacity SD card, or online service. Advanced driver assistance systems (ADAS) use combinations of RADAR, LiDAR, and machine learning-based computer vision to deliver comprehensive driving aids like adaptive cruise control (ACC), blind-spot detection, and emergency braking. The advances in automotive technologies are impressive, but they all share a legacy of the past; the traditional 12 V battery powers them.

For EVs equipped with a 400 V or 800 V battery pack, incorporating an additional battery and the associated power management electronics to power everything highlighted above would appear to incur an unnecessary bill of material cost. Figure 1 illustrates how complex the hybrid and fully EV power architectures have become compared to the internal combustion engine.

In the past, manufacturers packaged individual system functions in separate electronic control units (ECU), each powered with a 12 V supply. This distributed approach to power management and conversion results in high BOM costs. Of course, BOM cost isn’t the only factor to be considered though, since the weight of a 12 V primary battery and all the power components represent a significant payload. For example, the average weight of a 12 V starter battery is 20 kg, which, together with the excess power conversion components, can quickly become significantly increased.

Viewed another way, from the DC/DC power perspective, an electric vehicle power train introduces the need for 50 kW and upwards power conversion and management compared to < 3 kW for a conventional internal combustion engine vehicle. Therefore, achieving reliable and efficient power conversion in the minimum space and with the lowest weight becomes crucial. An EV power architecture needs to support the drive power train, onboard and infrastructure charging, and legacy systems.

Figure 1. A comparison of power architectures used in internal combustion engines, hybrid and fully electric vehicles (source Vicor). Image used courtesy of Bodo’s Power Systems
Automotive Power Deliver Architecture; A Different Approach

A viable approach to solving an electric vehicle’s development challenges proposed by Vicor is to use a virtual 12 V (or 24 V/ 48 V) battery (Figure 2). Rather than rely on a separate 12 V battery, why not create a virtual battery directly from the vehicle’s primary 400 V or 800 V battery pack? With this approach, manufacturers can save weight together with a reduction in the associated, engineering, supply chain and stocking costs.

Figure 2. Implementing an EV power architecture with virtual battery 12 V and 48 V sources (source Vicor). Image used courtesy of Bodo’s Power Systems

By incorporating high-density HV to LV conversion modules into existing sub-systems, the Vicor approach also achieves a higher degree of integration and a reduction in BoM cost – something OEMs wish to achieve too.

The Vicor proposal focuses on three aspects of the power delivery network architecture illustrated in Figure 2: charging, converting, and delivering.

EV Charging: The EV industry is gradually adopting the 800 V operating voltage for battery packs, but much of the EV charging infrastructure deployed is based on the initial 400 V standard. Therefore, any new EV needs to be able to accommodate both voltage levels. Efficient and straightforward bi-directional conversion modules are already available that offer an extremely flexible, high-efficiency and high-density scalable solution for battery-to-charger station compatibility.

Power Conversion: Conversion of an EV’s primary high voltage battery using a high-density automotive-qualified DC/DC module offers considerable weight and space savings for automotive manufacturers. Again, bi-directional power conversion capabilities provide flexibility in power delivery architectural design. The ability to eliminate the need for a 48 V intermediate energy storage, where used, by a virtual 48 V battery from the HV battery, further provides weight and space savings.

Virtual Power Delivery: In newer vehicles, 48 V applications include new drive, steer and brake-by-wire high power systems. Meeting the power delivery requirements of these networks while supporting legacy 12V loads (see Figure 3) with increased power requirements needs careful consideration. Compact high-density module solutions that are smaller and lighter than legacy solutions are available from Vicor.

Figure 3. Supporting legacy 12 V applications through a virtual power architecture from the vehicle’s HV battery (source Vicor). Image used courtesy of Bodo’s Power Systems
Redefining Automotive Power Delivery Architectures

As automotive OEMs grapple with lowering CO2 emissions while increasing vehicle performance and functionality, electric vehicles are proving to be the best option. However, keeping electric vehicle weight to a minimum to achieve a better range is proving to be a challenge. Redefining the architecture of a vehicle’s power delivery network saves both weight and system costs.

This article originally appeared in Bodo’s Power Systems magazine.

Author: Mark Patrick is Head of Technical Marketing EMEA at Mouser Electronics.

Source URL: https://eepower.com/technical-articles/Solving-Electric-Vehicle-Development-Challenges/

Are You Compliant with the IEEE 519-2022 Edition?

Published by Elspec LTD, website: elspec-ltd.com

The IEEE 519 standard defines the voltage and current harmonics distortion criteria for electrical systems design. The IEEE 519-2022 edition replaces the 2014 edition from December 2022.

The IEEE 519-2022 edition includes two important changes:

• Installations with mixed loads and Inverter Based Resources/Distributed Energy Resources
• Even current harmonics limits

New Guidelines for Installations with Mixed Loads and Inverter Based Resources/Distributed Energy

The 2022 edition instructs you whether to follow the IEEE-519 compliance criteria or different standards, as follows

1. IEEE-519 current limits at the point of common coupling (PCC) should be applied if the installation has an Inverted Based Resources (IBRs) or Distributed Energy Resources (DERs), in addition to the loads, and the combined site rated generation is lower than 10% of the annual average load demand.

2. IEEE 1547 or IEEE 2800 (if applicable) should be applied at the point of common coupling (PCC) should be applied if the installation has an IBRs or DERs, in addition to the loads, and the combined site rated generation is higher than 10% of the annual average load demand.

3. If the installation does not have an IBR or DER, IEEE-519 current limits should be applied at the PCC.

IEEE 519-2022: Even Current Harmonics Limits

The IEEE-519 defines the limits of current distortion per harmonic (in percentages of maximum demand load current) and TDD. The harmonics are divided to 5 groups (3rd – 10th, 11th – 16th, 17th – 22nd, 23rd – 34th and 35th – 50th), with different limits to each group of harmonics per rated voltage and ISC/IL ratio.

In the 2014 edition, all the even current harmonics were limited to 25% of their odd counterparts in their respective harmonic group. The 2022 edition is significantly different since only the even harmonics equal or below the 6th harmonic are limited, and these harmonics are limited only to 50% of their odd counterparts in the same harmonic group. The meaning is that all the even harmonics above the 6th harmonics’ values are allowed to be 4 times higher compared to the 2014 edition. i.e., the even harmonics values above the 6th harmonic can be the same as their odd counterparts in their harmonics group.

This might have a huge impact if you exceed the 2014 standard limits, as you may comply to the 2022 edition, avoiding penalties you might suffer from at the 2014 edition period.

An Example: Current Distortion Limits for Systems Rated 120 V – 69 KV

Table 1. (IEEE 519-2014)

.

Table 2. (IEEE 519-2022)

.

Source URL: https://www.elspec-ltd.com/are-you-compliant-with-the-ieee-519-2022-edition/

Understanding the IEEE 519–2014 Standard for Harmonics

Published by Elspec LTD, website: elspec-ltd.com

The IEEE 519-2014 standard defines the voltage and current harmonics distortion criteria for the design of electrical systems. The existed voltage and current waveforms in every part of the system are explained in this standard, and the waveform distortion goals for the system designer are established. The standard is periodically updated as the industry evolves. Since its introduction in 1981, the standard has been updated several times and its latest edition is IEEE 519-2014. Some updates have been made by 2022 (see here). The main terms definitions and statistical evaluation technics are being covered within this current article, as the main changes that have been made in the standard were described in the IEEE-519 2014 edition.

Definitions of Important Terms in the IEEE 519

To understand this document’s aim, the meaning of the following terms applied in this document is written below. The IEEE Standards Dictionary Online should be consulted for other terms not defined below.

1. New Definitions

Maximum demand load current: This current value is enacted at the point of common coupling (PCC) and calculates as the average of the currents corresponding to the peak demand during the previous 12 months.

Notch: A condition, lasting less than ½ cycle, in which the magnitude of the voltage waveform reversed its normal polarity.

Illustration 1: Notches

Point of common coupling (PCC): the point on a public power supply system, electrically closest to a specific load, other loads are, or maybe connected. The PCC is a point located upstream of the regarded installation.

Illustration 2: Point of common coupling (PCC)

2. Redefined Definitions

Short-circuit ratio: in a specific location, the rate of the available short-circuit current, to the load current, in amperes.

Total demand distortion (TDD): The ratio of the root mean square of the harmonic content, including the harmonic components up-to the 50th order. Expressed as a percent of the maximum demand current. Inter-harmonics are specifically excluded. Higher frequencies (harmonics greater than 50) may be added when required.

Total harmonic distortion (THD): The ratio of the root mean square of the harmonic content, including the harmonic components, up-to the 50th order. Expressed as a percent of the fundamental. Inter-harmonics are specifically excluded. Higher frequencies (harmonics greater than 50) may be added when required.

3. Legacy definitions

Harmonic (component): An element of order more than one of the Fourier series of a periodic quantity. For instance, in a 60 Hz system, the harmonic order 3, commonly known as the “third harmonic,” is 180 Hz.

Inter-harmonic (component): Refers to the frequency component of a periodic quantity that isn’t an integer multiple of the frequency in which the supply system operates (for instance, 50 Hz or 60 Hz).

I-T product: The inductive influence is expressed as regards the product of the root-mean-square current magnitude (I), in amperes, times its telephone influence factor (TIF).

kV-T product: Inductive influence expressed as regards the product of root-mean-square voltage magnitude (V), in kilovolts, and times its telephone influence factor (TIF).

Notch depth: The average depth of the line voltage notch from the sine wave of voltage.

Notch area: It is the area of the line voltage notch. It is the product of the notch depth, in volts, times the width of the notch measured in microseconds.

Pulse number: The total number of successive non-simultaneous commutations taking place inside the converter circuit during every cycle when operating without phase control. It is also equal to the principal harmonic order in the direct voltage, i.e., the number of pulses available in the dc output voltage in one cycle of the supply voltage.

Telephone influence factor (TIF): For a voltage or recent wave in an electric supply circuit, the ratio of adding the square root of the squares of the weighted root-mean-square values of every one of the sine-wave components (with alternating current waves both fundamental and harmonic) to the root-mean-square value (unweighted) of the whole wave.er loads are, or maybe connected. The PCC is a point located upstream of the regarded installation.

Differences with the Previous Edition
.
New Measurement Method and Statistical Evaluation Technique

The IEEE 519-2014 introduce a newly measurement methods and statistical evaluation technique to determine compliance with the recommended limits.

Harmonics Measurement Methods

The standard adopt the 10/12 cycles gapless harmonic subgroup measurement from the IEC 61000-4-7. Aggregations of 150/180 cycles (~3sec) and 10min are required for the statistical assessments.

Very short time harmonic measurements: Very short time harmonic values are assessed over a 3-second interval based on an aggregation of 15 consecutive 12 (10) cycle windows for 60 (50) Hz power systems. Individual frequency components are aggregated based on an RMS calculation as shown in Equation (1) where F represents voltage (V) or current (I), n represents the harmonic order, and i is a simple counter. The subscript vs is used to denote “very short.” In all cases, F represents an RMS value.

.

Short time harmonics measurements: Short time harmonic values are assessed over a 10-minute interval based on an aggregation of 200 consecutive very short time values for a specific frequency component. The 200 values are aggregated based on an RMS calculation as shown in Equation (2) where F represents voltage (V) or current (I), n represents the harmonic order, and i is a simple counter. The subscript sh is used to denote “short.” In all cases, F represents an RMS value.

.
Statistical Evaluation

Daily evaluation: It is required to calculate the 99th percentile value (i.e. the value that is exceeded for 1% of the day) of the very short time harmonics values for comparison with the recommend limits.

Weekly evaluation: It is required to calculate the 95th and 99th percentile value (i.e. those values that are exceeded for 5% and 1% of the week) of the short time harmonics values for comparison with the recommend limits.

The chart below display a daily accumulative and relative probability chart of the Total Demand Distortion parameter at resolution of 3sec as taken from PQSCADA Sapphire.

.
The IEEE 519 – 2014 Compliance Criteria

These suggested practice limits are for application at a point of common coupling (PCC) between the system owner or operator and system users. The PCC is often regarded as the point in the power system closest to the user where the system owner or operator could provide services to other users. Usually for service to industrial users, e.g., manufacturing plants through a unique service transformer, the PCC will be at the transformer’s HV side. For most commercial users like office parks, etc., supplied through a usual service transformer, the PCC is commonly at the LV side of the service transformer.

Voltage Distortion Limits

• Daily 99th percentile very short time (3 s) values should be less than 1.5 times the values given in the table below.

• Weekly 95th percentile short time (10 min) values should be less than the values given in the table below.

Table 1. (IEEE 519-2014)

.

Current Distortion Limits

• Daily 99th percentile very short time (3 s) harmonic currents should be less than 2.0 times the values given in the tables below.

• Weekly 99th percentile short time (10 min) harmonic currents should be less than 1.5 times the values given in tables below.

• Weekly 95th percentile short time (10 min) harmonic currents should be less than the values given in tables below.

Table 2. (IEEE 519-2014) Current distortion limits for systems rated 120 V – 69 kV

.

Table 3. (IEEE 519-2014) Current distortion limits for systems rated 69 kV – 161 kV

.

Table 4 (IEEE 519-2014) Current distortion limits for systems rated > 161 kVa

.

a Even harmonics are limited to 25% of the odd harmonic limits above
b Current distortions that result in a dc offset, e.g., half-wave converters, are not allowed
c All power generation equipment is limited to these vales of current distortion, regardless of actual ISC/IL.
ISC = maximum short circuit current at PCC
IL = maximum demand load current (fundamental frequency component) at PCC

Source URL: https://www.elspec-ltd.com/ieee-519-2014-standard-for-harmonics/

Assessment of the Impact of the Micro Wind Turbine on the Power Quality in the Distribution Network

Published by Marek GAŁA, Andrzej JĄDERKO, Politechnika Częstochowska, Wydział Elektryczny

Abstract. The article presents the principles of measurements and assessment of power quality characteristics, with the power supplied by the micro wind turbine connected to the distribution network. It describes the basic technical parameters of the vertical axis micro wind turbine 10 kW and the characteristics of its output as a function of wind speed. Besides, it shows selected results of measurements of parameters characterizing the power quality in the node of micro wind turbine of 10 kW connection to the 400 V network.

Streszczenie. W artykule przedstawiono zasady pomiarów i oceny jakości energii dostarczanej przez mikroturbinę wiatrową podłączoną do sieci dystrybucyjnej. Opisano podstawowe parametry techniczne mikroturbiny wiatrowej o pionowej osi obrotu i mocy 10 kW oraz charakterystykę jej mocy wyjściowej w funkcji prędkości wiatru. Pokazano również wybrane wyniki pomiarów parametrów charakteryzujących jakość energii w węźle przyłączenia mikroturbiny wiatrowej do sieci dystrybucyjnej 400 V. (Ocena wpływu pracy mikroturbiny wiatrowej na jakość energii elektrycznej w sieci dystrybucyjnej niskiego napięcia).

Keywords: vertical micro wind turbine, power quality, distribution network
Słowa kluczowe: mikroturbina wiatrowa o pionowej osi obrotu, jakość energii, sieć dystrybucyjna

Introduction

The current regulations applicable to non-business energy users state that a microgeneration plant can be connected free of charge to the distribution grid after reporting such an intention to a regional distribution company. Additionally, various funds can be obtained to finance investments into Renewable Energy Sources (RES). These factors are responsible for the visibly growing interest in RES, especially photovoltaic systems and wind turbines, equipped with inverter systems, control systems and protection systems [1, 2, 12].

The massive increase in the number of microgeneration plants can however cause significant problems for the distribution grid, including aggravation of energy quality. Because of this, microgeneration plants connected to the grid should meet a number of technological requirements, as well as conditions specified by grid operators in the relevant instructions, e.g. [9] and [3], in accordance to applicable standards and regulations [4, 5, 6, 10].

In the next sections of this paper characteristics of the wind turbine MEW-10 are presented, followed by selected measurement results of electrical energy quality generated by this unit.

Characteristic of a wind turbine

The wind turbine type MEW-10 consists of a vertical axis wind turbine (VAWT) with a three-blade rotor of the H-Darrieus type, a disc slow-rotation permanent magnet synchronous generator (PMSG) and a controller together with protection systems. The rated power obtained by the wind speed 12 m/s is 10 kW. If the wind speed exceeds this value, the power is constrained by the control system [11]. The maximal rotational speed of the rotor is about 140 rpm. A number of empirically obtained characteristics of the MEW-10 are presented in Table 1.

Table 1. Basic characteristics of the MEW-10 wind turbine

.

The cross-section area of the rotor (wind circle) is 30 m2, with the blade height of 6 m and the rotor diameter of 5 m. A blade cross-section is a symmetrical assembly of standard airfoils type NACA 0021. A wind turbine of the type investigated in the paper is presented in Figure 1.

A power inverter creates current load on the plant generator, which causes torque on the shaft, adjusting the rotational turbine speed to the current wind speed, in this way ensuring the maximal power coefficient. A grid-tie inverter couples the generator with the power grid, generating three voltage waveforms synchronized with the grid phase voltages.

Fig.1. Vertical wind turbine of type MEW-10
Measurements of the quality of wind turbine-generated energy

The measurements of energy quality at a connection node of the wind turbine in an consumer internal grid were carried out in the first half of July 2018, by means of an energy quality analyzer PQ-Box 200, meeting the requirements of the standard [8] with respect to class A. The measurements were intended to verify if energy generated by the wind turbine meets the requirements specified in [3, 4, 5, 9]. Below are presented selected results of power parameters and energy quality parameters collected over a week period of observation, with a 10- minute period of data aggregation, tA = 600 s.

Figure 2 presents the mean square (rms) values of phase current, the maximum value of which was Imax = 3.3A, and the rms value at tA = 0.2 s was Imax 0.2s = 12.23 A. The visible asymmetry of currents is caused by the current Iinv = 0.29 A flowing through the power inverter control system from the phase L3. Figure 3 presents the values of the active power P and the reactive power Q. The minimal value of the active power was Pmin = -1,8 kW (Pmin 0.2s = – 8,66 kW – the case of maximum power generation by the wind turbine). The working inverter consumes the power of about Pinv = 37.5 kW. As can be seen, the plant has significant demand for reactive power: Qmin = -1,5 kvar (Qmin 0.2s = -5,44 kvar) – Fig. 3.

Fig.2. Root mean square values of the currents IL1, IL2, IL3 at the node connection node
Fig.3. Active power P and reactive power Q of micro wind turbine during one week of measurements

Figure 4 presents the values of voltage THD coefficients. No voltage distortion exceeding the admissible level was observed: THD U ∊ 〈1.88, 3.05〉%.

Fig.4. THD UL1, THD UL2, THD UL3 of phase voltage in the node of micro wind turbine

Figure 5 presents the momentary currents recorded at maximal power generation, i.e. P = Pmin. The current deformation was assessed by obtaining harmonic rms values for n = 2,…,50 and comparing them to the values specified in the relevant standards [4, 6]. The results obtained are presented in Table 2. As can be noted, the plant does not cause higher harmonics of values exceeding the admissible limit level to flow through the connection node of the wind turbine.

Fig.5. Momentary values of currents IL1, IL2, IL3 at Pmin

Table 2. Comparison of the measured results of higher harmonics with values specified in the standard

.

The value of THD I for Pmin was THD IPmin = 7.11%. Curves representing the variation of the voltage and current asymmetry coefficients were obtained on the basis of the direct components (U1, I1) and inverse components (U2, I2):

.

Figure 6 presents the voltage coefficient curve. With the maximal value of the coefficient αUmax = 0.32%, it is significantly smaller than the limit value of 2%. It was also observed that the value of current asymmetry coefficient varies from 1.78% at Pmin to 100% when no power is generated and only the phase L3 is under load due to the power inverter being powered from this circuit. Figure 7 presents the variation of the indices Plt The values of the index Plt are within the interval 0.25 – 0.50, whereas the values of the index Pst are included in the interval 0.07 – 0.67. According to [8], the values of the indices Pst and Plt do not exceed the limit values, as specified in [3]:

.
Fig.6. Voltage asymmetry coefficient αU
Fig.7. Indices of long-term flicker severity Plt L1, Plt L2, Plt L3 in the node of micro wind turbine

The power inverter control system requires continuous supply of active energy E+ (blue), even when it is working at wind speeds below the turbine start-up speed. The amount of active energy consumed during the test week was 5,36 kWh and the amount of reactive energy was EQ (purple) (7.50 kvarh) – Fig. 8. The total net amount of energy supplied to the grid was E+ = 2.08 kWh (green), but is has to be kept in mind that the measurements were taken during summer, and the wind speed attested at that time was not optimum for the operation of the wind turbine.

Fig.8. Energy: E – active energy supplied to the grid (red), E+ – active energy consumed by the power inverter (blue), E+ – active energy generated by the wind turbine plant (green); EQ – reactive energy consumed from the grid by the wing turbine plant (purple)

In order to analyze in detail the plant’s demand for reactive power, additional measurements were carried out, with the consideration of the aggregation time tA = 1 s – Fig. 9. Besides, Figures 10 and 11 present momentary values of currents and voltages, respectively, during the charging of capacitors in the intermediary circuit of the power inverter in the microgeneration plant.

Fig.9. Current I, active power P and reactive power Q during the charging of the capacitors in the power inverter intermediary circuit; tA = 1 s
Fig.10. Momentary currents IL1, IL2, IL3 during connecting capacitors into the intermediary power inverter circuit
Fig.11. Momentary voltages UL1, UL2, UL3 during connecting capacitors into the intermediary power inverter circuit
Conclusions

The measurements carried out for the sake of the present study indicate that the operation of the wind turbine does not cause voltage changes exceeding 3%, nor does it cause voltage asymmetry, voltage fluctuations or current harmonics exceeding admissible limit levels. During the charging of the capacitors in the power inverter circuit, impulse currents with momentary values reaching 90 A occurred, which caused additional voltage drop at the grid impedance and contributed to momentary voltage distortion, as shown in Fig. 11. This phenomenon was not however found to interfere with the operation of any devices at the consumer side. Still, it needs to be further scrutinized by the manufacturer of the power inverter with the view to optimizing the control algorithm. Besides, another set of verification measurements should be carried out at wind speeds ensuring generating the rated power of the wind turbine.

REFERENCES

[1] Act on Power Law of 10 April 1997, Journal of Laws of 1997 no 54, item 348, with later amendments (Ustawa z dnia 10 kwietnia 1997 r. Prawo energetyczne, Dz. U. z 1997 r., nr 54, poz. 348 z późn. zm.).
[2] Act on Renewable Energy Sources of 20 February 2015, Journal of Laws of 2015, item 478 (Ustawa z dnia 20 lutego 2015 r. o odnawialnych źródłach energii (Dz. U. z 2015 r., poz. 478).
[3] Connection criteria and technical requirements for microgeneration plants and small-scale generation plants connected to the LV distribution network (Kryteria przyłączania oraz wymagania techniczne dla mikroinstalacji i małych instalacji przyłączanych do sieci dystrybucyjnej niskiego napięcia) TAURON Dystrybucja S.A., Krakow, July 18, 2016
[4] EN 50438 Requirements for micro-generating plants to be connected in parallel with public low-voltage distribution networks
[5] IEC 61000-3-2:2014 Electromagnetic compatibility (EMC) – Part 3-2: Limits – Limits for harmonic current emissions (equipment input current ≤ 16 A per phase).
[6] IEC 61000-4-7:2002+A1:2008 Electromagnetic compatibility (EMC) – Part 4-7: Testing and measurement techniques – General guide on harmonics and interharmonics measurements and instrumentation, for power supply systems and equipment connected thereto.
[7] IEC 61000-4-15:2010 Electromagnetic compatibility (EMC) – Part 4-15: Testing and measurement techniques – Flickermeter Functional and design specifications.
[8] IEC 61000-4-30:2015 Electromagnetic compatibility (EMC) – Part 4-30: Testing and measurement techniques – Power quality measurement methods.
[9] Instructions for Distribution Network Traffic and Exploitation applicable since 01.01.2014, TAURON Dystrybucja S.A. (Instrukcja Ruchu i Eksploatacji Sieci Dystrybucyjnej TAURON Dystrybucja S.A. obowiązująca od dnia 01.01.2014 r.).
[10] The Ministry of Economy ordinance on the detailed conditions of the power system operation, Journal of Laws of 2007, no, 93, item 623 with later amendments (Rozporządzenie Ministra Gospodarki z dnia 4 maja 2007 r. w sprawie szczegółowych warunków funkcjonowania systemu elektroenergetycznego, Dz. U. z 2007 r., nr 93, poz. 623 z późn. zm.).
[11] Turbine MEW-10 catalogue description
[12] Sobierajski M., Rojewski W. ”Conditions for connecting microgeneration plants to the LV grid vs. legal regulations,” (Warunki przyłączania mikrogeneracji do sieci niskiego napięcia w świetle obowiązujących przepisów), Zeszyty Naukowe Wydziału Elektrotechniki i Automatyki Politechniki Gdańskiej, nr 33/2016, pp. 79-82

Authors: dr inż. Andrzej Jąderko, Politechnika Częstochowska, Wydział Elektryczny, Al. Armii Krajowej 17, 42-200 Częstochowa, e-mail: aj@el.pcz.czest.pl
dr inż. Marek Gała, Politechnika Częstochowska, Wydział Elektryczny, Al. Armii Krajowej 17, 42-200 Częstochowa, e-mail:m.gala@el.pcz.czest.pl

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 95 NR 1/2019. doi:10.15199/48.2019.01.09

HARMONICS: Understanding the Facts – Part 3

Published by Richard P. Bingham

Abstract. Understanding what is important to know about harmonics can be challenging for those without extensive electrical engineering backgrounds. This is third and final part of a three part series. This part will provide details on what causes harmonic problems and suggested solutions.

What they look like

One recent survey showed the percentage the total electrical consumption by non-linear loads will double from the year 1985 to 2000. The AC-DC converter used in the switching-type power supplies found in most personal computers and peripheral equipment, such as printers, is an example of a non-linear load. While they offer many benefits in size, weight and cost, the large increase of equipment using this type of power supply over the past fifteen years is largely responsible for the increased attention to harmonics.

Figure 1 shows how the first stage of a switching-type power supply works. The AC voltage is converted into a DC voltage, which is further converted into other voltages that the equipment needs to run. The rectifier consists of semi-conductor devices (such as diodes) that only conduct current in one direction. In order to do so, the voltage on the one end must be greater than the other end. These devices feed current into a capacitor, where the voltage value on the capacitor at any time depends on how much energy is being taken out by the rest of the power supply.

Figure 1. Typical AC-DC Converter

\When the input voltage (Vi) is higher than voltage on the capacitor (Vc), the diode will conduct current through it. This results in a current waveform as shown in Figure 2, and harmonic spectrum in Figure 3. Obviously, this is not a pure sinusoidal waveform with only a 60 Hz frequency component.

Figure 2. Current Waveform
Figure 3. Harmonic Spectrum of Current Waveform Shown in Figure 2.

Figure 3. Harmonic Spectrum of Current Waveform Shown in Figure 2. If the rectifier had only been a half-wave rectifier, the waveform would only have every other current pulse, and the harmonic spectrum would be different. Whereas the above harmonic spectrum contains only odd harmonics for current, the spectrum for the current of a half wave rectified circuit would only have even harmonics.

Certain types of loads also generate typical harmonic spectrum signatures, that can point the investigator towards the source. This is related to the number of pulses, or paths of conduction. The general equation is h = ( n*p ) +/- 1, where h is the harmonic number, n is any integer (1,2,3,..) and p is the number of pulses in the circuit. Table 1 shows examples of such. The magnitude decreases as the ratio of 1/h (1/3, 1/5, 1/7, 1/9,…).

Table 1. Typical Harmonics Found for Different Converters.

.

When transformers are first energized, the current drawn is different from the steady state condition. This is caused by the inrush of the magnetizing current. The harmonics during this period varies over time. Some harmonics have a negligible value for part of the time, and then increase for a while before returning to basically zero. An unbalanced transformer (where either the output current, winding impedance, or input voltage on each leg are not equal) will cause harmonics, as will overvoltage saturation of a transformer.

Fluorescent lights can be the source of harmonics, as the ballasts are non-linear inductors. The third harmonic is the predominate harmonic in this case. (See Table 2) As previously mentioned, the third harmonic current from each phase in a four-wire wye or star system will be additive in the neutral, instead of canceling out Some of the newer electronic ballasts have very significant harmonic problems, as they operate somewhat like a switching power supply, but can result in current harmonic distortion levels over 30%.

Table 2. Sample of Harmonic Values for Fluorescent lighting [4].

.

The process of melting metal in an electric arc furnace can result in large currents that are comprised of the fundamental, interharmonic, and subharmonic frequencies being drawn from the electric power grid. These levels can be quite high during the melt-down phase, and usually effect the voltage waveform.

How do you get rid of them

Care should be undertaken to make sure that the corrective action taken to minimize the harmonic problems don’t actually make the system worse. This can happen as the result of resonance between harmonic filters, PF correcting capacitors and the system impedance. Examples of ways to minimize the harmonic problems include:

• Isolating harmonic pollution devices on separate circuits with or without the use of harmonic filters. – Loads can be relocated to try to balance the system better.

• Phase shifted transformers, such as “zig-zag transformers”, can be used to cancel out specific harmonics by making one voltage circuit 180 degrees out-of-phase from another.

• Neutral conductors should be properly sized according to the latest NEC-1996 requirements covering such. Where as the neutral may have been undersized in the past, it may now be necessary to run a second neutral wire that is the same size as the phase conductors. This is particularly important with some modular office partition-type walls, which can exhibit high impedance values.

• The operating limits of transformers and motors should be derated, in accordance with industry standards from IEEE, ANSI and NEMA on such.

• Use of higher pulse converters, such as 24-pulse rectifiers, can eliminate lower harmonic values, but at the expense of creating higher harmonic values.

Summary

Harmonics are here to stay. But the amount of harmonic voltage and current levels that a system can tolerate is dependent on the equipment and the source. Ongoing preventive maintenance programs that include harmonic monitoring can detect problems in the making, eliminating costly failures. Knowing what your system harmonic levels presently are, what the effect of new equipment being added will due to these levels, and how much of an increase in harmonic levels that your system can tolerate are valuable pieces of information that are readily attainable from modern power quality/harmonic analyzer monitoring equipment.

References

National Electrical Code – NEC-1996, National Fire Protection Association

Blog posts: 
HARMONICS: Understanding the Facts – Part 1,
HARMONICS: Understanding the Facts – Part 2

HARMONICS: Understanding the Facts – Part 2

Published by Richard P. Bingham

Abstract. Understanding what is important to know about harmonics can be challenging for those without extensive electrical engineering backgrounds. In this two part series, this second article will help to clarify why you need to be concerned about them, how and where to find them, and when they are a problem.

Why Worry About Harmonics

The presence of harmonics does not mean that the factory or office cannot run properly. Like other power quality phenomena, it depends on the “stiffness” of the power distribution system and the susceptibility of the equipment when operating in the presence of the harmonics. One factory may be the source of high harmonics but be able to operate properly. This harmonic pollution is often carried back onto the electric utility distribution system, and may effect neighboring facilities on the same system which are more susceptible.

There are a number of different types of equipment that may experience misoperations or failures due to high harmonic voltage and/or current levels:

• Excessive neutral current, resulting in overheated neutrals. The currents of triplen harmonics, especially the odd harmonics, (3rd, 9th, 15th,…) are actually additive in the neutral of three phase wye circuits. This is because the harmonic number multiplied by the 120 degree phase shift between the three phases is a integer multiple of 360 degrees, or one complete cycle. This puts the harmonics from each of the three phase conductors “in-phase” with each other in the neutral, as shown in Figure 1.

Figure 1. Additive Third Harmonics [1]

• Incorrect reading meters, including induction disc-type W-hr meters and averaging type current meters.

• Reduced true PF, where PF= Watts/VA.

• Overheated transformers, especially delta windings where triplen harmonics generated on the load side of a delta-wye transformer will circulate in the primary side. Some type of losses go up as the square of harmonic value (such as skin effect and eddy current losses). This is also true for solenoid coils and lighting ballasts.

• Positive, negative, and zero sequence voltages on motors and generators. These are voltages at a particular frequency that try to rotate the motor forward, backward, or neither (just heats up the motor), respectively. As shown in Table 1, the voltage of a particular frequency in a balanced system harmonics can have either a positive (fundamental, 4th, 7th,…), negative (2nd, 5th, 8th…) or zero (3rd, 6th, 9th,…) sequencing value.

Table 1. Harmonic Sequencing Values in Balanced Systems.

.

• Nuisance operation of protective devices, including false tripping of relays and failure of a UPS to transfer properly, especially if the controls incorporate zero-crossing sensing circuits.

• Bearing failure from shaft currents through uninsulated bearings of electric motors.

• Blown-fuses on PF correction caps, due to high voltage and currents from resonance with line impedance.

• Mis-operation or failure of electronic equipment – Light flicker results when there are voltage subharmonics in the range of 1-30Hz. The human eye is most sensitive at 8.8Hz, where just a 0.5% variation in the RMS voltage is noticeable with some types of lighting.[2]

Where They Come From

The amount of voltage harmonics will often depend on the amount of harmonic currents being drawn by the load, and the source impedance, which includes all of the wiring and transformers back to the source of the electricity. Ohm’s Law says that Voltage equals Current multiplied by Impedance. This is true for harmonic values as well. If the source harmonic impedance is very low (often referred to as a “stiff” system) then the harmonic currents will result in lower harmonic voltages than if the source impedance were high (such as found with some types of isolation transformers). The impedance of an inductive device goes up as the frequency goes up, while the impedance goes down for capacitive devices for higher harmonics.

How this electricity is used by the different types of loads can have an effect on the “purity” of the voltage waveform. Some loads cause the voltage and current waveforms to lose this pure sine wave appearance and become distorted. Depending on the type of load and system impedances, the waveform may consist of predominately harmonics.

“The main sources of harmonic current are at present the phase angle controlled rectifiers and inverters.” [3] These are often called static power converters. These devices take AC power and convert it to DC, then sometimes back to AC power at the same or different frequency based on the firing scheme. The firing scheme refers to the controlling mechanism that determines how and when current is conducted. One major variation is the phase angle at which conduction begins and ends.

Power converters come in different sizes. Low power, AC voltage regulators for light dimmers and small induction motors adjust the phase angle or point on the wave where conduction occurs. Medium power converters are used for motor control in manufacturing and railroad applications, and include such equipment as ASDs (adjustable speed drives) and VFDs (variable frequency drives). Metal reduction operations, like electric arc furnaces, and high voltage DC transmission employ large power converters, in the 2-20MVA rating.

Where to look for them

Wherever the aforementioned equipment is used, one can suspect that harmonics are present. Like any power quality investigation, the search can begin at the equipment effected by the problem or at the point-of-common-coupling (PCC), where the utility service meets the building distribution system. If only one piece of equipment is effected (or suspected as being the producer), it is often easier to start the monitoring process there. If the source is suspected to be from the utility service side (such is the case when there is a neighboring factory that is known to generate high harmonics), then monitoring usually begins at the PCC.

How do you find them

Hand-held harmonic meters can be useful tools for making spot checks for known harmonic problems. However, harmonic values will often change during the day, as different loads are turned on and off within the facility or in other facilities on the same electric utility distribution system. This requires the use of a harmonic monitor or power quality monitor with harmonic capabilities (such as shown in Figure 2), which can record the harmonic values over a period of time.

Figure 2. Power Quality Monitor with Harmonic Analysis

The phase voltages and currents, as well as the neutral-to-ground voltage and neutral current should be monitored, where possible. This will aid in pinpointing problems, or detecting marginal systems. Monitoring the neutral will often show a high 3rd harmonic value, indicating the presence of non-linear loads in the facility.

Typically, monitoring will last for one business cycle. A business cycle is how long it takes for the normal operation of the plant to repeat itself. For example, if a plant runs three identical shifts, seven days a week, then a business cycle would be eight hours. More typically, a business cycle is one week, as different operations take place on a Monday, when the plant equipment is restarted after being off over the weekend, then on a Wednesday, or a Saturday, when only a skeleton crew may be working.

In order to be able to analyze complex signals that have many different frequencies present, a number of mathematical methods were developed. One of the more popular is called the Fourier Transform. Duplicating the mathematical steps required in a microprocessor or computer-based instrument is quite difficult. So more compatible processes, called the FFT for Fast Fourier Transform, or DFT for Discrete Fourier Transform, are used. These methods only work properly if the signal is composed of only the fundamental and harmonic frequencies in a certain frequency range (called the Nyquist frequency, which is one-half of the sampling frequency). The frequency values must not change during the measurement period. Failure of these rules to be maintained can result in mis-information.

For example, if a voltage waveform is comprised of 60Hz and 200Hz signals, the FFT cannot directly see the 200Hz. It only knows 60, 120, 180, 240,…, which are often called “bins”. The result would be that the energy of the 200Hz signal would appear partially in the 180Hz bin, and partially in the 240Hz bin. An FFT-based processer could show a voltage value of 115V at 60Hz, 18V at the 3rd harmonic, and 12V at the 4th harmonic, when it really should have been 30V at 200Hz. A spectrum analyzer can also exhibit a similar problem, as it places the energy in its own machine-dependent bins. In addition, a spectrum analyzer takes only a snap-shot in time of the harmonics, which are best analyzed on an averaged, steady-state waveform.

When are they a problem

To determine what is normal or acceptable levels, a number of standards have been developed by various organizations. ANSI/IEEE C57.110 Recommended Practice for Establishing Transformer Compatibility When Supplying Non-sinusoidal Load Currents [5] is a useful document for determining how much a transformer should be derated from its nameplate rating when operating in the presence of harmonics. There are two parameters typically used, called K-factor and TDF (transformer derating factor). Some power quality harmonic monitors will automatically calculate these values.

IEEE 519-1992 Recommended Practices and Requirements for Harmonic Control in Electrical Power Systems [6] provides guidelines from determining what are acceptable limits. The harmonic limits for current depend on the ratio of Short Circuit Current (SCC) at PCC (or how stiff it is) to average Load Current of maximum demand over 1 year, as illustrated in Table 2. Note how the limit decreases at the higher harmonic values, and increases with larger ratios.

Table 2. Current Harmonic Limits as per IEEE 519-1992.

.

For voltage harmonics, the voltage level of the system is used to determine the limits, as shown in Table 3. At the higher voltages, more customers will be effected, hence, the lower limits.

Table 3. Voltage Harmonic Limits as per IEEE 519-1992.

.

The European Community has also developed susceptibility and emission limits for harmonics. Formerly known as the 555-2 standard for appliances of less than 16 A, a more encompassing set of standards under IEC 1000-4-7 are now in effect.

Most electrical loads (except half-wave rectifiers) produce symmetrical current waveforms, which means that the positive half of the waveform looks like a mirror image of the negative half. This results in only odd harmonic values being present. Even harmonics will disrupt this half-wave symmetry. The presence of these even harmonics should cause the investigator to suspect there is a half-wave rectifier on the circuit. This also result from a full wave rectifier when one side of the rectifier has blown or damaged components. Early detection of this condition in a UPS system can prevent a complete failure when the load is switched onto back-up power.

References

[1] Power Line Harmonic Problems – Causes and Cures, Dranetz Technologies, December 1994.
[2] NFPA 70B Recommended Practice for Electrical Equipment Maintenance – Chapter
24, National Fire Protection Association, Quincy MA, 1994.
[3]J. Arrillega et.al. Power System Harmonics, John Wiley and Sons, 1985.
[4]Heydt, GT, Electric Power Quality, Stars in the Circle Publication, Indianapolis, 1991, pg 240.
[5] ANSI/IEEE C57.110 Recommended Practice for Establishing Transformer Compatibility When Supplying Nonsinusoidal Load Currents
[6] IEEE 519-1992 Recommended Practices and Requirements for Harmonic Control in Electrical Power Systems National Electrical Code – NEC-1996, National Fire Protection Association

About the Author: Richard P. Bingham is currently the Chief Technologist for Dranetz Technologies, Inc., having previously been the Vice-President of Engineering and Strategic Planning. He has been with the company since 1977, following completion of his BSEE at the University of Dayton. Richard also has an MSEE in Computer Architecture and Programming from Rutgers University. He is a member of IEEE Power Engineering Society and Tau Beta Pi, the Engineering Honor Society. Richard is currently working with the NFPA 70B committee on Power Quality and several IEEE committees related to IEEE 1159, and has written and presented numerous papers and seminars in the electric utility and power quality instrumentation fields.

Blog post: HARMONICS: Understanding the Facts – Part 1