Specifying harmonic emissions for the connection of offshore wind generation

Offshore wind generation is likely to affect power quality, specifically harmonics, within the transmission or distribution system to which it connects. 

Go to the profile of Gillian Williamson
Aug 22, 2017
0
1
Upvote 0 Comment

Author(s): Gillian Williamson and Z. Emin

Abstract

The system owner/operator undertakes a harmonic assessment study to check that their network subsequent to the connection will operate within permissible harmonic distortion criteria. Based on the results of the assessment subsequent to the connection, they may specify allowable emission levels which may inherently imply that mitigation measures, possibly in the form of harmonic filters, are necessary. This article discusses the assessment methodologies, the likely form of a specification and an iterative compliance review process. A case study is presented which demonstrates the importance of communication between the system owner/operator and offshore wind farm developer.

Introduction

Transmission and distribution system operators have an obligation to maintain network harmonic voltage distortion levels within permissible limits to ensure that the many side effects of harmonics do not become problematic. To maintain acceptable harmonic voltage distortion levels, harmonic assessments are likely to be undertaken as part of the technical assessment associated with the connection of an offshore wind farm development to a transmission or distribution system. This assessment may be undertaken by the network owner, the connecting party or may be sub-contracted. Based on the results of the assessment, harmonic emission limits may be specified in the connection agreement. One of the mitigation measures the connecting party can implement is the installation of harmonic filters. Filter costs and delivery timescales can be significant, therefore it is important that developers are aware of the importance of compliance with harmonic standards and understand the process from assessment, through specification and concluding with a review of the filter design. Ultimately, lack of timely consideration of power quality can delay connection or result in constraints if a filter is not available on time.

This article provides an introduction to harmonic voltage distortion considerations when connecting offshore wind developments. The following topics are discussed:

  • harmonic assessment methodologies,
  • the likely form of a harmonic emission specification and
  • the iterative compliance review process between system operator and connecting party.

Harmonic limits

UK transmission and distribution network operators are required to comply with Engineering Recommendation (ER) G5/4-1 [1] limits in accordance with statutory obligations defined in the Grid and Distribution Codes, respectively.

The Grid Code [2] Connection Condition 6.1.5 (a) Harmonic Content requires that:

(a) Harmonic Content

The electromagnetic compatibility levels for harmonic distortion on the onshore transmission system from all sources under both planned outage and fault outage conditions (unless abnormal conditions prevail), shall comply with the levels shown in the tables of Appendix A of ER G5/4. The electromagnetic compatibility levels for harmonic distortion on an offshore transmission system will be defined in relevant bilateral agreements.

ER G5/4 contains planning criteria which NGET will apply to the connection of non-linear Load to the National Electricity Transmission System, which may result in harmonic emission limits being specified for these loads in the relevant bilateral agreement. The application of the planning criteria will take into account the position of existing and prospective users’ plant and apparatus (and OTSDUW plant and apparatus) in relation to harmonic emissions. Users must ensure that connection of distorting loads to their user systems do not cause any harmonic emission limits specified in the bilateral agreement, or where no such limits are specified, the relevant planning levels specified in Engineering Recommendation G5/4 to be exceeded.

The Distribution Code [3] section DPC 4.2.3.2 (b) requires that:

The harmonic content of a load shall comply with the limits set out in DGD Annex 1, Item 1 ER G5/4-1, ‘Planning levels for harmonic voltage distortion and the connection of non-linear equipment to transmission and distribution systems in the United Kingdom.’

ER G5/4-1 sets planning and compatibility levels for harmonic voltage distortion in the UK and defines how harmonics should be assessed considering harmonic orders up to and including the 50th and the total harmonic distortion. The ER is accompanied by a technical report, Engineering Technical Report (ETR)122 [4], which explains the philosophy of the former. ER G5/4-1 defines compatibility levels with which observed harmonic voltage distortion must always conform. To ensure operation within these compatibility levels, ER G5/4-1 also defines planning levels (at lower or equal values to compatibility) which should be adhered to when performing a harmonic voltage distortion assessment during the planning stages of a connection or system change.

Harmonic effects of an offshore wind farm connection

Offshore wind generation will affect power quality, in particular harmonics, within the transmission or distribution system to which it connects. This is because of the significant harmonic component of the currents of the inverter connected wind turbines, and associated reactive power compensation, often in the form of a static var compensator. A change in harmonic voltages also comes about because connecting the offshore system modifies the overall system impedance, and this change can alter background harmonic voltage distortion levels. The impedance and susceptance of normally long subsea cables, along with the impedance of any onshore shunt reactive power compensation, dominates the impedance of an offshore wind farm electrical system. Connecting such an impedance in parallel with that of the transmission or distribution system can result in significant changes in the harmonic impedance of the system; shifting existing resonances to lower frequencies and even creating new resonances. Consequently, background harmonic levels because of the multitude of harmonic sources connected throughout an electrical network can change simply because of the connection of the passive elements associated with an offshore wind farm development.

Planning stage harmonic assessment

To assist distribution and transmission system operators in assessing compliance with ER G5/4-1, a three-stage assessment process is outlined as part of the recommendation. Different procedures apply at different voltages and connection ratings to ensure that the assessment is in proportion to the impact of the connection. The most onerous Stage 3 process requiring determination of harmonic distortion voltages based on a harmonic impedance model is applicable to Points of Common Coupling (PCC) at 33 kV or above, and therefore probably all offshore wind farm connections.

In a Stage 3 assessment, the additional distortion because of the new connection is considered along with the measured background voltage distortion in the existing system as shown in Fig 1.

Fig 1: ER G5/4-1 Stage 3 assessment process

Incremental harmonic voltage distortion

It is always advisable to use a full system model to assess the incremental harmonic voltage distortion because of the new harmonic current emission source. Representation of system components and the harmonic source in the new network arrangement enables the additional voltage distortion to be calculated for each individual harmonic based upon the network harmonic impedance. Such a detailed representation is necessary because although normally inductive at nominal frequency, a network may be capacitive at higher harmonic frequencies. As the frequency increases and the network impedance changes from inductive to capacitive, resonances can occur resulting in excessive harmonic voltage distortions.

Resonances are just one reason why accurate system representation is required. Parameters to allow detailed modelling of the harmonic performance of the transmission and distribution systems should be available to the system operators. Therefore, for accurate results it is essential that similarly realistic information is used for the offshore wind farm system to be connected. Assumptions or the use of ‘worst case’ results can lead to overly pessimistic results and onerous filter requirements. Where possible it is recommended that turbine generator harmonic currents, measured in accordance with IEC61400-21 [5], are used.

A detailed model facilitates checking compliance with harmonic voltage distortion planning levels at not only the PCC, but also other affected network locations because some locations within the network may be more susceptible to harmonic distortion because of resonances. All planned system configurations should be checked with a model, since a network may have a different harmonic impedance profile for each different network arrangement, different loading and different generation scenario.

Measuring background harmonic voltage distortion

Measurement of background harmonic voltage distortion at the PCC and other remote network nodes is required for Stage 3 assessments. It is recommended that instruments and measurement methodologies comply with the latest versions of BS EN 61000-4-7 [6] and BS EN 61000-4-30 [7]. In accordance with ETR122, 3-second and 10-min characteristics of harmonics up to the 50th harmonic should be measured over a 7-day period to observe daily fluctuations and the appropriate values used in the assessment. 3-second values are more applicable to very short-term burst effects and hence the 10-min values are more suitable for the assessment of wind farm connections. The probabilistic nature of distortion should be taken into consideration by analysing measured raw data to determine the 95% values, that is, values for which 95% of the recordings are below. These 95% values should be used in the assessment of compliance with ER G5/4-1.

Modified background harmonic voltage distortion

In the harmonic assessment, the network impedance at the PCC is taken to be affected by the new connection and so the background voltage distortion measured there is modified to reflect the change in impedance. It is possible to assume that the connection only affects the point of connection and has a lesser effect on the background harmonic voltage distortion at all other locations in the network. However, the decision to extend the analysis such that it covers modification of background harmonic voltage distortion at other locations can only be determined by looking at each case individually.

The methodology for calculating the expected modified background harmonics at the PCC can be based on the assumption that the existing background harmonics are because of a constant voltage source. On this basis, it can be shown that the background harmonics at the PCC change by the same factor as the self-impedance at the point of connection, known as an amplification factor. The expected modified harmonics can then be calculated by multiplying the measured initial background harmonic voltage distortion values by the calculated amplification factors at each individual harmonic. Self-impedance values are normally calculated as part of frequency sweeps using power systems analysis software with harmonic capabilities. The existing background harmonic voltage distortion can also be recreated by modelling multiple harmonic current sources, but this can prove to be extremely difficult especially in highly meshed networks and should only be attempted when precise locations of major harmonic current sources are known.

Aggregation

The overall harmonic voltage distortion is calculated from the incremental harmonic voltage distortion and the appropriate background using the summation method defined in section 8.3.2 of ER G5/4-1. For the particular harmonic with the largest summated magnitude, it is assumed that the background harmonic distortion and the incremental distortion because of the new harmonic source peak at the same time and are in phase and so are added linearly. All other harmonics are assumed to have 90° phase difference between the background and the new distortion and are therefore added by taking the square root of the sum of squares.

Harmonic specification

The initial limits included in a specification within a connection agreement are given on a need basis. Not all available rooms at each individual harmonic is provided to the margin connecting party. This ensures that fair access is given to other customers who may also apply to connect at the same point before the development is connected, commissioned and the revised background can be measured.

Incremental and aggregate harmonic voltage emission limits can be expected in the specification. It is also worth remembering that this specification is for harmonic performance purposes only and that the connecting party may choose to rate its equipment, in particular for harmonic filters, at this level of harmonics or slightly more.

Specified incremental emission limits

Incremental limits may be defined as a table of the harmonic voltage distortions allowed at the point of connection based on only the harmonic current injection from the new connection. These limits will be associated with keeping the harmonics at the PCC and other remote network locations within permitted planning levels. If a predicted specific harmonic voltage distortion at a remote location exceeds the ER G5/4-1 planning level, then the maximum harmonic current injection from the new connection corresponding to the planning level at that limiting location must be established. The harmonic voltage distortion at the point of connection calculated based on this reduced harmonic current injection then becomes the specified incremental harmonic voltage emission included in the specification at the particular harmonic in question.

If it is found that the aggregate harmonic voltage distortion at the point of common coupling exceeds the ER G5/4-1 planning levels, then the emission limit on the specified incremental voltage distortion at the harmonic in question is the difference between the planning level and the modified background level of that harmonic.

If the voltage distortions at a particular harmonic are within planning levels at the PCC and all remote locations, then the specified incremental harmonic emission is that at the point of connection due to the unconstrained harmonic current.

Specified aggregate limits

A table of the maximum aggregate harmonic voltage emissions is included in the specification to bound the effects of the connection including how it changes existing background harmonics. This table is based upon the calculated aggregate voltage distortions. If the maximum calculated aggregate voltage distortion at a particular harmonic order is less than the ER G5/4-1 planning level, then the value is included in the specification unchanged, so defining the allowable maximum overall voltage distortion because of the new connection at that harmonic. For the harmonics where the calculated overall voltage distortion exceeds ER G5/4-1 planning levels, the maximum voltage distortion is restricted to the appropriate ER G5/4-1 planning level. In such cases, the allowable voltage distortion at the harmonic orders close to (at least one up and two down) the problematic harmonic can be increased to allow some margin for the filter design.

Impedance loci

The incremental and aggregate harmonic voltage specifications should be accompanied with information defining the impedance of the network at the point of connection. Such impedances are necessary to enable filter design to fulfil the allowable voltage distortions included in the specification. Owing to the aforementioned range of impedances as a result of variations in the system operation, a set of network harmonic impedance envelopes is often the most appropriate way of presenting this information. These envelopes are formed for each harmonic by plotting the system impedance for each system configuration in the R– X plane. Some plots may be formed for a small band of harmonics if the range of impedances is similar. It is particularly important to define the boundaries of these ranges such that resonance points do not coincide with the boundaries.

Harmonic performance compliance

Specification of harmonic emission limits is based upon all planned network configurations, but checking this is not practicable because it is difficult to measure the harmonics corresponding to all system configurations. Harmonic measurements may be taken before a connection is made to demonstrate the satisfactory operation prior to the connection and provide a baseline. Recordings after the connection should be made to continually monitor compliance with harmonic emission limits as network configuration varies.

In the meantime any filter design would be checked for satisfactory performance by review of the design studies or independent analysis. It is important to understand that it is not the responsibility of the transmission and distribution system owner/operator to endorse or reject the design of filter(s) but rather the responsibility lies with the connecting party to demonstrate compliance with the given harmonic performance specification. It is possible that the performance of the system with the filter may be established to be outside of the harmonic emission limits included in the specification. However, this is likely to be acceptable as long as the predicted harmonic voltage distortions are within ER G5/4-1 planning levels. Since not all available margin is provided when defining the emission limits in the specification, it is possible to alter the limits such that the expected levels with the filter become the new limits for those harmonic orders at which non-compliance of the original limits is noted. This is where proper communication between the developer and system operator is essential because performance with the filter should not pose any adverse effect to the system, provided that the calculated worst-case values are not exceeded in practice.

Conclusion

To ensure continuing network compliance with system voltage waveform criteria, connection of an offshore wind development is likely to require a harmonic assessment as described within this paper. Although specialist knowledge and software will probably be required for the study, a good appreciation and understanding of the process and underlying principles will be helpful to ensure an effective and efficient assessment.

It should be noted that many of the aspects of the ER G5/4-1 Stage 3 assessment discussed here are not only applicable to offshore wind farms. Similar approaches can be applied to the harmonic assessment of other non-linear connections, such as HVDC [8].

The importance and timing of harmonic assessment should be recognised so that the design and construction of any necessary filters can be completed in line with the overall construction programme. Co-operation between the connecting party and system operator is required at all stages; to ensure an appropriate accurate assessment based on realistic data rather than extreme worst-case pessimistic information that could lead to unnecessarily excessive mitigation. Also, co-operation and communication are required when assessing compliance of a proposed filter since it may be necessary to flex the original specification of emission limits in a connection agreement as long as they remain within allowable planning levels.

It was discussed that specifications are made on a need basis and that if the expected distortion is expected to exceed the limitations then the specification is capped at 100% of the planning level. This ‘first come, first served’ principle is advantageous for the connecting party, but can be problematic for subsequent connectees. It has already been suggested that there is a need to change this approach [8] and a review of ER G5/4-1 is expected soon.

Acknowledgments

The authors would like to express their gratitude and thanks to Parsons Brinckerhoff for its support and permission to publish this paper. The case study given is based upon [9] and hence the authors would also like to extend the author's gratitude and thanks to ENW and Vattenfall in allowing us to publish their study results.

References

  1. Engineering Recommendation G5/4-1: ‘Planning levels for harmonic voltage and the connection of non-linear equipment to transmission systems and distribution networks in the United Kingdom’ (Energy Networks Association, October 2005) .
  2. The Grid Code Issue 5 Revision 5–5 November 2013.
  3. The Distribution Code Issue 20, September 2013.
  4. ETR122 Guide to the Application of ER G5/4 in the Assessment of Harmonic Voltage Distortion and Connection of Non-linear Equipment, Electricity Association, December 2002.
  5. IEC 61400-21 ed2.0 Wind turbines – Part 21: Measurement and assessment of power quality characteristics of grid connected wind turbines, 2008.
  6. BS EN 61000-4-7:2002+A1:2009, 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. BS EN 61000-4-30:2009, Electromagnetic compatibility (EMC) – Part 4-30: Testing and measurement techniques – Power quality measurement methods.
  8. Emin Z. Ghassemi F. Price J. J.: ‘Harmonic performance requirements of an HVDC connection; network owner perspective’. Nineth Int. Conf. on AC and DC Power Transmission, ACDC 2010, London, United Kingdom, 20–21 October 2010.
  9. Emin Z. Fernandez F. Poeller M. Williamson G. E.: ‘Harmonic distortion specification and compliance of offshore wind generation’. Tenth Int. Conf. on AC and DC Power Transmission, ACDC 2012, Birmingham, United Kingdom, 4–5 December 2012.

Case study: Ormonde Offshore Wind Farm

The connection of Ormonde Offshore Wind Farm (OOWF) involved following the harmonic assessment, specification and review process as described in [9]. A Stage 3 assessment according to ER G5/4-1 [1], requiring determination of harmonic voltages based on a harmonic impedance model, was necessary for the assessment of the OOWF development.

OOWF, located 10 km off Barrow-in-Furness in the Irish Sea, comprises thirty 5 MW offshore wind turbines. It is connected to the Electricity North West Limited (ENWL) power network at Heysham 132 kV. A 132 kV subsea cable 41.5 km long connects the offshore substation to a cable interconnection facility on shore, from where a further 2.8 km land cable makes the connection at Heysham 132 kV substation.

The local connectivity is depicted in Fig 2. To avoid being overly cautious, the assessment considered single outage scenarios, specifically an outage in either the ENWL or National Grid (NGET) system.

Fig 2: Schematic showing the connectivity of the Ormonde wind farm system

At the time of the harmonic assessment, the connection of Walney offshore wind farm at Heysham was also being planned and had been the subject of NGET harmonic assessments. Two connections at the same point and at the same time added some complexities, but also provided a temporary advantage. It meant that some delays were caused because the assessments were sequential and the Ormonde assessment could not start until the Walney assessment was completed. Also, the Ormonde assessment was complicated by having to consider an outage of the Walney wind farm connection meaning that the studies needed to be repeated for both with and without the Walney network connected. However, further down the line when installation of the Ormonde filter was delayed, the close proximity of the Walney system meant that Ormonde could agree with ENWL temporary conditional operating regimes to ensure harmonic voltage distortion compliance before the Ormonde filter was installed.

Data from the system operators and also the manufacturers of the new equipment was used to develop a detailed representation of the wind farm power network connected to the ENWL system with a simple equivalent of the transmission system. This model was used to evaluate the incremental harmonic voltage distortion because of the new harmonic current sources in the wind farm. Utilising measurements of the background harmonics in the existing system, the same system representation was used to evaluate the expected modified background harmonic voltage distortion at the PCC based on amplification factors. Power system simulation software with harmonic analysis capability was utilised to calculate the voltage distortion from the new harmonic source and the self-impedance values used within the determination of the modified background harmonic voltage distortion. However, some subsequent processing of this data was then required to calculate the overall prospective harmonic voltage distortion (in this case a spread sheet).

The resultant harmonic specification for the Ormonde connection included the maximum allowable aggregate (Va) and incremental (Vi) harmonic voltage distortion and is given along with the associated background (Vb) harmonic voltage distortion in Table 1. For most harmonic numbers the specified allowable harmonic distortion was defined by the maximum voltage distortion calculated as part of the assessment since the results were less than the ER G5/4-1 planning limits. However, the study results indicated a resonance around the 5th harmonic resulting in the predicted overall harmonic distortion exceeding the ER G5/4-1 planning levels at the 5th and 6th harmonic, thus indicating the need for mitigation. Consequently, the specified allowable voltage distortion was limited to ER G5/4-1 planning limits at the 3rd to 7th harmonics.

h

Vi

Va

Vb

h

Vi

Va

Vb

2

0.16

0.16

0.055

27

0.1

0.1

0.050

3

1.72

2.0

0.159

28

0.1

0.2

0.050

4

0.8

0.8

0.056

29

0.1

0.13

0.050

5

0.6

2

0.408

30

0.1

0.1

0.050

6

0.47

0.5

0.064

31

0.1

0.11

0.050

7

1.75

2.0

0.326

32

0.1

0.1

0.050

8

0.27

0.28

0.050

33

0.1

0.18

0.050

9

0.10

0.1

0.050

34

0.1

0.18

0.050

10

0.19

0.31

0.050

35

0.1

0.1

0.050

11

0.10

0.10

0.050

36

0.1

0.1

0.050

12

0.10

0.10

0.050

37

0.1

0.1

0.050

13

0.10

0.10

0.050

38

0.1

0.1

0.050

14

0.10

0.10

0.050

39

0.1

0.1

0.050

15

0.26

0.27

0.050

40

0.1

0.1

0.050

16

0.16

0.2

0.050

41

0.1

0.1

0.050

17

0.38

0.49

0.050

42

0.1

0.2

0.050

18

0.1

0.2

0.050

43

0.1

0.491

0.050

19

0.1

0.1

0.050

44

0.1

0.2

0.050

20

0.1

0.1

0.050

45

0.12

0.2

0.078

21

0.1

0.1

0.050

46

0.1

0.2

0.050

22

0.1

0.1

0.050

47

0.1

0.27

0.125

23

0.1

0.22

0.050

48

0.1

0.12

0.050

24

0.1

0.1

0.050

49

0.1

0.1

0.050

25

0.1

0.1

0.050

50

0.1

0.1

0.050

26

0.1

0.1

0.050

 Table 1: Specified harmonic voltage distortion limits and background as a percentage of fundamental

The resonance around the 5th harmonic was simply because of the combination of system impedances, and these also led to predicted harmonic voltage distortions being above the planning levels at the 16th, 18th and between the 42nd and 46th harmonics. Consequently, the specified allowable distortion was capped at the ER G5/4-1 planning limits at these harmonic numbers. However, these assessment results were considered less of a concern and possibly overly pessimistic since the underlying measurements of existing background are typically inaccurate particularly at higher harmonic orders and sensitivity studies including the skin effect in the long subsea and land cables indicated lower harmonic voltage distortions.

The specified limits presented in Table 1 were accompanied with the set of network harmonic impedance envelopes, applicable to the point of connection, presented in Figs 3–6.

Fig 3: Network impedance envelopes seen from Heysham 132 kV for 2 ≤ h ≤ 7 without Walney

Fig 4: Network impedance envelopes seen from Heysham 132 kV for 8 ≤ h ≤ 50 without Walney

Fig 5: Network impedance envelopes seen from Heysham 132 kV for 2 ≤ h ≤ 7 with Walney

Fig 6: Network impedance envelopes seen from Heysham 132 kV for 8 ≤ h ≤ 50 with Walney

During the harmonic assessment process close co-operation ensured that the harmonic specification was based upon the most suitable parameters and was adjusted to match the expected performance of the wind farm system with the planned filter design.

Lessons Learned

  1. Recognise the requirement for a harmonic assessment.
  2. Include the design, construction and installation of potential harmonic filtering in the offshore wind farm development programme.
  3. Obtain the most up-to-date design parameters for the wind farm and the power network to perform a harmonic assessment in time to meet the construction and installation programme.
  4. Develop good co-operation between the network operator and wind farm developer to ensure an effective and efficient assessment, specification and review process.
  5. Understand the importance of realistic data for an appropriate specification.
Go to the profile of Gillian Williamson

Gillian Williamson

Principal power systems engineer , Parsons Brinckerhoff

1 Comments

Go to the profile of Stephen Sommerville
Stephen Sommerville 8 months ago

Hi, interesting article. I would be interested to know some more about the impedance loci, i have seen these shown in passing in several text books, but without any real explanation of how to read them and how they are applied practically.

Could you recommend any good reference books/studies/papers on this?