Basic Harmonic Filter Design

Introduction to Harmonic Filter Design

Harmonic filter design is one solution to mitigate harmonics wave distortion or interference in modern electrical systems, particularly in the industrial and commercial sectors. This article will briefly discuss the basic theory, considerations, and procedures for harmonic filter designing. —Omazaki Engineering is an electrical harmonics filter design consultant that also provides harmonic study and harmonic assessment consulting services. Contact us if you need a harmonic filter designer by emailing cs@omazaki.co.id or filling out the contact form.

———————————————

Harmonic Filters and Harmonic Filter Design

Uncontrolled harmonics can reduce system reliability, increase operational costs, and increase the risk of equipment damage. One way to control harmonics is by designing harmonic filters.

A harmonic filter is a device designed to attenuate or reduce harmonic levels in a power system. Harmonic filter design refers to the planning and engineering calculation process used to determine the type, capacity, and configuration of the filter that best suits the load characteristics and harmonic distortion levels in a system. The design must consider efficiency, reliability, and avoid potential system resonance.

The function of installing a harmonic filter is to maintain electrical power quality in accordance with standards, such as IEEE 519. These filters can work by absorbing specific harmonics (at specific frequencies) or actively injecting a harmonic-canceling signal into the system.

———————————————

Types of Harmonic Filters

Passive Harmonic Filter

A passive filter is a type of harmonic filter consisting of passive components such as resistors (R), inductors (L), and capacitors (C). This filter is designed to resonate or absorb harmonics at specific frequencies, especially low-order harmonics (such as the 5th, 7th, or 11th harmonics). Types of passive filters include:

• Single-tuned filter (tuned to a single harmonic frequency)
• High-pass filter (reduces higher-order harmonics)
• Band-pass filter (for a specific frequency range)

Advantages of passive filters:

• Relatively low cost
• Simple structure

Disadvantages:

• Inflexible to load changes
• Risk of causing resonance if not designed properly

Active Harmonic Filter

Active harmonic filters use electronic components such as inverters, sensors, and digital controllers to generate harmonic-canceling currents in real time. The principle of active filters is to detect harmonics in the system and actively inject current with a waveform that is out of phase with the harmonic, thereby negating the harmonics.

Advantages of active filters:

• Capable of handling high-order and fluctuating harmonics
• Flexible for various types of loads
• Can simultaneously correct power factor

Disadvantages:

• Higher cost
• Higher system complexity
• Requires regular maintenance

———————————————

Harmonic Filter Design Considerations

Harmonic filter designing is not simply about selecting the appropriate filter type; it requires comprehensive technical considerations to ensure the system is not only harmonic-free but also efficient, safe, and economical. IEEE Std 1531-2020 outlines several important factors to consider before determining the final harmonic filter configuration.

Reactive Power (kVAR) Requirements

Harmonic filters generally consist of capacitors, reactors, and in some cases, resistors. In addition to reducing harmonic distortion, the use of capacitors in filters also generates capacitive reactive power, which directly impacts the power system.

One of the key aspects of harmonic filter design is knowing the required reactive power (kVAR) and how it affects system voltage control. Capacitors in the filter can also be used to compensate for reactive power, reducing the need for additional external compensation.

However, adding capacitors directly can also trigger voltage fluctuations. According to IEEE Std 1036, it is recommended that voltage fluctuations due to switching kVAR not exceed 2–3% of the nominal voltage. This ensures that the system voltage remains stable and does not negatively impact equipment or overall system performance.

The total kVAR value and the filter step size need to be designed based on the results of the power flow analysis and voltage control requirements. In certain systems, harmonic filters are designed in several switching steps so that the capacitor can be activated gradually according to load requirements and system conditions, while avoiding current surges during startup.

Harmonic Limitation

1) System Limits

Harmonic limits in electric power systems aim to prevent functional disruptions and equipment damage due to electrical waveform distortion. In this regard, IEEE Std 519 serves as the primary reference for determining permissible harmonic distortion thresholds, particularly at the Point of Common Coupling (PCC).

For distribution systems with voltages below 1 kV, total voltage distortion (THD) is generally limited to a maximum of 8%, while for systems above 1 kV, it is limited to a maximum of 5%, depending on the voltage level and network characteristics. For harmonic currents, the permissible Total Demand Distortion (TDD) value ranges from 2.5% to 20%, depending on the harmonic load capacity and the fault current-to-load ratio (Isc/IL).

Under certain conditions, such as temporary disturbances or system startup, higher harmonic limits may be tolerated for a limited time, for example, less than one hour.

2) Equipment Withstand Capacity

In addition to the system, the equipment’s ability to withstand harmonics is also a critical aspect of filter design. High harmonic distortion can lead to overheating, reduced efficiency, and premature equipment failure.

Examples of affected equipment include:

• Transformers, which when used to supply non-sinusoidal loads, will experience harmonic currents, resulting in additional losses and increased temperatures. To prevent overheating, IEEE Std C57.110 provides a method for calculating a derating factor, which adjusts the transformer’s operating capacity based on the level of harmonic distortion. This derating does not require re-rating the transformer, but is used to ensure that the transformer remains operating within safe thermal limits under actual harmonic load conditions.
• Capacitors, as defined in IEEE Std 18, must operate within the following limits:
  • RMS voltage ≤ 110% of nominal value
  • Peak voltage ≤ 120%
  • Current ≤ 135% of nominal current
  • Reactive power ≤ 135% of nominal kVAR

Exceeding these limits can cause dielectric heating, premature aging, and even permanent failure.

Every system component, especially those in the path of harmonic currents, needs to be evaluated for its resistance to distortion. Good harmonic filter design is not only to suppress harmonics, but also to protect system equipment from their impact.

———————————————

Normal System Conditions

Normal power system operating conditions must be thoroughly analyzed to ensure that the designed harmonic filter can meet the desired reactive power (kVAR) requirements and harmonic damping performance. Several important aspects to consider include:

1) System Harmonic Voltages and Currents

All forms of harmonic voltages and currents, both characteristic and non-characteristic, must be taken into account:

• Characteristic harmonics originate from primary harmonic loads such as inverters, VFDs, UPSs, and other electronic control equipment.
• Non-characteristic harmonics, such as even harmonics, triple harmonics (multiples of 3), or harmonics that are not whole multiples of the fundamental frequency, can also arise due to system imbalance or equipment operational defects. These values need to be estimated based on field experience, direct measurements, or system simulations.

In systems connected to the public grid, frequency variations are typically no more than ±0.1 Hz. However, in systems using local generators, variations can be larger, impacting the filter’s tuning accuracy and harmonic suppression performance.

4) Power System Configuration

Variations in system configuration, such as transformer relocations, supply line changes, or medium-voltage feeder reconfigurations, can significantly alter system impedance. Therefore, system simulations should include:

• Complete transformer representation (with proper winding connections)
• Conductor and cable reactance
• Capacitance effects
• Harmonic loads around the filter location.

All harmonic sources electrically connected near the filter location must be included in the model.

5) Loading Conditions

• Changing system load conditions can affect filter performance. These include:
• Changes in harmonic load status (e.g., inverter on/off), Induction motor status,
• Changes in the operation of capacitor banks or other harmonic filters.

It should also be noted that linear (resistive) loads can help dampen harmonic resonances, so they should be included in the calculation.

6) System Voltage Unbalance

Voltage unbalance can lead to increased harmonic injection, especially triplen harmonics, and spread them throughout the system. Therefore, system unbalance must be addressed before installing the filter.

———————————————

Normal Condition of Harmonic Filter

Harmonic filters don’t always operate under ideal conditions, so various variables that affect their performance must be taken into account, such as:

• Component tolerances, such as variations in inductance and capacitance values due to manufacturing
• Changes in ambient temperature, which can decrease capacitance by 0.4–0.8% per 10°C increase
• Capacitor element failure, which can cause tuning changes and overvoltages on the remaining capacitors.

For large filters, the failure of a single element may be tolerable. However, in small filters, the failure of a single unit can drastically degrade filter performance.

System Conditions During Contingencies

Harmonics filter design must not only perform optimally under normal conditions, but also be able to withstand contingency scenarios or temporary or extreme system disturbances. Evaluation of these conditions is crucial to ensure the filter rating remains adequate and prevents overloading. Here are some things to consider:

1) System or Filter Switching

The switching process, whether in the harmonic filter or other system components, can generate significant dynamic transient voltages. For example:

• Turning on a large transformer can generate dangerous voltage spikes for the filter.
• Simultaneous switching on of multiple single-tuned filters, especially at low frequencies, can generate very high transients.
• If a multi-step filter is used, the switching sequence must be controlled to prevent harmful parallel resonance.

To prevent excessive voltage surges, adequate switching delays must be implemented. IEEE Standard 18 recommends:

• A minimum of 5 minutes for medium or high voltage capacitors (>1000V)
• A minimum of 1 minute for low voltage capacitors (≤1000V), to allow the residual voltage on the capacitor to drop below 50 V before restarting.

If the system requires faster switching, the following can be used:

• Discharge devices
• Switching devices with resistors/insertion reactors
• or Switches that close at near-zero voltage across the open terminals to avoid transient surges.

2) Use of Filters Tuned to the Same Frequency

If more than one harmonic filter is installed in the same location and tuned for the same harmonic frequency, it is important to ensure that the harmonic current is distributed proportionally between the filters. Impedance imbalances between filters can cause excessive current concentration in only one filter.

3) System Frequency Variations

Under contingency conditions, frequency variations can be much greater than normal, especially when the system is powered by a local generator. These variations can shift the filter tuning point and compromise harmonic damping performance.

4) System Configuration Changes

Scenarios such as partial system outages, changes in source connections, or even the loss of one or more filters need to be analyzed. Configuration shifts can cause harmonic resonances to shift to new frequencies, making some equipment more susceptible to harmonics.

5) Increase in Uncharacteristic Harmonics

In emergency situations, unusual harmonics (such as even or triple harmonics) can increase above normal values. This value should be included as a rating in the filter design to ensure safety during extreme conditions.

6) Unidentified Harmonic Sources

Not all harmonic sources can be identified initially. In complex systems, harmonics from hidden or future loads often arise. Therefore, for design safety, a margin factor should be added to account for harmonics from unknown or future sources.

———————————————

Filter Installation Location

In general, harmonic filters can be installed directly on individual devices (for example, on non-linear loads like VFDs or UPSs), or on a common bus serving multiple loads simultaneously. Each approach has its own advantages and technical considerations.

Filters can also be placed on low-voltage systems, such as 480 V, or on medium- to high-voltage systems, such as 4.16 kV or 12.47 kV, depending on the location of the harmonic source and the coverage area to be protected.

Location selection should consider:

• The site’s ability to achieve acceptable harmonic voltage and current limits for the system
• The impact of harmonic current flow on equipment and conductors, including potential power losses and conductor heating, which can affect the reliability of the power distribution system.

With proper placement, harmonic filters can provide maximum protection without causing side effects such as resonance or local overloading.

———————————————

Harmonic Filter Configuration

Harmonic filters can be designed in various configurations, for both grounded and ungrounded systems. Configuration selection is not only based on attenuation effectiveness, but also depends on:

• Availability of equipment and hardware in the field
• Ease of installation and maintenance
• Protection system considerations.

The key principle is to ensure that the selected configuration not only effectively attenuates harmonics but also does not introduce new resonance points in the system.

Utilizing Existing Capacitor Banks

In many cases, power systems already have existing capacitor banks used for power factor correction. To increase cost and space efficiency, these capacitors are often considered as part of a harmonic filter.

However, not all capacitor banks are suitable for this purpose. Factors to consider include:

• The capacitor’s voltage, current, and kVAR ratings must be sufficient to withstand harmonics
• The increase in fundamental and harmonic voltages that the capacitor will experience must be taken into account
• In many cases, existing capacitors can only be used if the unit was overrated during initial installation.

———————————————

Harmonic Filter Design Procedure

Effective harmonic filter design requires not only a technical understanding of harmonics but also a systematic approach to determining appropriate specifications and configurations. IEEE Std 1531-2020 provides a step-by-step guide to produce a harmonic filter that is reliable, efficient, and meets power quality standards.

The following are the main steps in the harmonic filter design procedure:

Step 1: Determining the kVAR Size of the Harmonic Filter Bank

In addition to its function of damping harmonics, the harmonic filter also plays a role in providing capacitive reactive power to the power system. This reactive power is useful for improving the power factor and stabilizing the voltage, especially when the system is under heavy load conditions.

This reactive capacity measurement is known as “effective kVAR,” which is the actual kVAR value supplied by the harmonic filter after subtracting the reactance effects of the series reactors within the filter. This value is always smaller than the nameplate kVAR value of the filter capacitor due to the voltage drop caused by the resistance and inductance in the filter reactors.

The kVAR value is generally determined using power flow analysis software to accurately adjust the system’s reactive power requirements. Several factors to consider when determining the kVAR size include:

• The number of switching steps in the harmonic filter capacitor
• The range of system voltage variations, both under normal and unbalanced conditions
• The range of load variations that may occur during system operation
• The power system configuration, both under normal and contingency conditions, including the system being planned.

A good design will consider all of the above factors so that the capacitor bank is not only effective in supplying reactive power and damping harmonics, but also safe and stable under various operating conditions.

Step 2: Determining the Initial Harmonic Filter Tuning

Once the kVAR capacity is determined, the next step in harmonic filter design is to determine the initial tuning frequency, which is the frequency at which the filter has minimum impedance and is most effective in absorbing harmonic currents. This initial tuning aims to reduce current and voltage distortion to meet established power quality standards.

In general, tuning is performed on the harmonic with the highest amplitude. For example, if the 5th and 7th order harmonics are dominant, then a single filter tuned close to the 5th harmonic may be sufficient to significantly reduce distortion. However, further evaluation through harmonic simulation is necessary to determine whether a single filter is sufficient or whether additional filters are needed.

It is important to note that filter tuning is not performed precisely at the harmonic frequency. There are two main reasons why tuning slightly below the harmonic frequency is preferred:

• If tuning precisely at the harmonic frequency, all the harmonic current at that frequency will flow through the filter. This forces the filter to be designed larger and more expensive than actually necessary.
• Precise tuning can create parallel resonance with the system impedance. This interaction can shift the system’s resonance point to a frequency adjacent to the dominant harmonic, resulting in harmonic amplification instead of filtering, and even causing damaging voltage spikes.

Factors That Causes Resonance

Some factors that can cause the resonance point to deviate from its original design include:

• Capacitor element failure, either due to internal or external fuse blowing, which changes the total capacitance and shifts the resonance frequency up or down.
• Manufacturing tolerances and temperature changes, which can affect the L (reactor) and C (capacitor) values.
• Power system configuration variations, such as transformer outages, supply line changes, or overhead-to-underground network conversions, all of which change the system impedance and affect the position of the parallel resonance.

Therefore, it is recommended that filters be tuned approximately 3% to 15% below the target harmonic frequency. For example, for the 5th harmonic (300 Hz in a 60 Hz system), the filter can be tuned around 282 Hz (≈ 4.7 times 60 Hz), rather than exactly 300 Hz.

In systems with multiple harmonic filters, this tuning strategy still needs to be thoroughly reviewed to avoid overlapping effects between filters. The entire harmonic spectrum at the filter installation point, both under normal and contingency conditions, must be analyzed.

An alternative approach is to avoid harmonics rather than absorb them. This is applicable to systems that are less sensitive to distortion but want to protect capacitors from excessive harmonic currents or prevent resonance. In this case, tuning is performed below the characteristic harmonic (e.g., the 4.3rd or 4.7th), and the filter can be ungrounded to avoid triplen resonances such as the 3rd harmonic.

Step 3: Optimize the Filter Configuration to Meet Harmonic Guidelines

Once the initial tuning frequency is determined, the next important step in the harmonic filter design process is to optimize the harmonic filter configuration so that the system meets the harmonic distortion limits recommended by standards. One key reference in this regard is IEEE Std 519, which sets maximum limits for harmonic voltage and current distortion at various voltage system levels and connection points (PCCs).

Harmonic filters must be designed to remain effective under:

• Normal power system conditions
• Abnormal or contingency scenarios, such as partial blackouts, network switching, or sudden load changes.

To achieve this, a thorough harmonic analysis is required, typically performed using harmonic simulation software. Manual calculations are only suitable for simple systems. In more complex systems, numerical simulations are required to evaluate filter performance across the entire harmonic frequency spectrum and under various operating conditions.

Important Factor

Several important factors to consider when optimizing filter configurations include:

• The number of filter switching steps (if filters are used in a cascade),
• The likelihood of failure (outage) of one filter, especially if there is more than one filter in the system,
• System voltage variations during normal and extreme conditions,
• Load fluctuations, including nonlinear loads,
• Changes in power system configuration, both under normal conditions and during contingencies,
• Detuning effects due to:
  • Changes in system frequency,
  • Component manufacturing tolerances (L, C),
  • Changes in capacitance values due to extreme temperatures,
  • Partial capacitor unit failure,
• Characteristic and uncharacteristic harmonics, including even and triple harmonics,
• Background harmonics of the existing system.

If simulation results indicate that harmonic distortion is still above standard limits, several solutions can be considered:

• Retuning the filter frequency slightly downward to avoid parallel resonance with the system,
• Adding a new filter with a different tuning (multi-tuned filter),
• Increasing the filter capacity (kVAR), which automatically widens the filter bandwidth and compensates for component tolerances.

Step 4: Determining Filter Component Ratings

Once the harmonic filter configuration and performance have been optimized, the next step is to determine the ratings for each major component in the filter. This process involves determining the technical specifications of the capacitors, reactors, resistors (if used), and switching devices.

The following details each component that must be determined:

4.1) Harmonic Filter Capacitors

Capacitors are the main components that generate reactive power at the fundamental frequency and are directly exposed to harmonics. Capacitor ratings are typically determined by the manufacturer based on:

• System harmonic spectrum
• Transient and dynamic voltages due to switching or disturbances
• System var requirements
• System data provided by the user.

There are three types of voltages that must be considered when determining capacitor ratings:

• Steady-state voltage (including harmonics)
• Transient voltage (lasting less than half a cycle)
• Dynamic voltage (can last several seconds).

In many single-tuned applications, steady-state voltage is often the primary basis for determining the rating. Transient voltage is usually not critical unless multiple filters with different resonant frequencies are connected to the same bus.

For safety, the actual peak voltage (fundamental + harmonics) applied to a capacitor should not exceed 100% of the capacitor’s rated peak voltage. This voltage is calculated from the worst-case harmonic current spectrum and the total capacitive reactance. All significant harmonics must be included in the calculation.

The fundamental frequency current for a wye-connected harmonic filter can be calculated using the equation:

Where:

• VS is the voltage across the harmonic filter-capacitor-harmonic filter-reactor circuit
• XC is the capacitive reactance at the fundamental frequency
• XL is the inductive reactance at the fundamental frequency

Harmonic currents are part of the harmonic filter design criteria. The total RMS current in a harmonic filter is calculated as follows:

The RMS current through the capacitor must be less than 135% of the nominal current based on the kVAR rating and the capacitor voltage. This value must also remain within the capacitor’s fuse capacity. Generally, current limits are not a major limiting factor except for high-order harmonics.

4.2) Harmonic Filter Reactor

The filter reactor (inductor) rating is determined after tuning is established through harmonic analysis. Parameters that must be determined include:

• Inductance value (L)
• Q ratio (X/R at the tuning frequency)
• Acceptable inductance tolerance and Q value
• Harmonic and fundamental currents under normal and contingency conditions.

The reactor rating must be sufficient to carry harmonic currents continuously and withstand heating due to resistive losses.

4.3) Harmonic Filter Resistor (if used)

In high-pass filters, resistors are used to adjust the shape of the impedance curve and limit high resonances. Resistor parameters that must be determined include:

• Resistance value (R)
• Allowable resistance tolerance
• Maximum acceptable series inductance in the resistor
• Power rating based on harmonic energy dissipation
• Harmonic and fundamental currents in various operating scenarios.

4.4) Circuit Breaker or Switch

The short-circuit current rating of the switching device must be calculated based on system data. This value is not the same as the short-circuit current of the harmonic filter reactor, but rather represents the fault current on the bus where the switch is installed.

Please note:

• Capacitor switches do not have to interrupt fault currents, unlike circuit breakers.
• However, the switch must be able to withstand fault currents during the turn-on (close-and-latch) process.
• Capacitive current switching requirements must account for the worst-case combination of maximum system voltage, capacitance tolerance, and the presence of harmonics.
• Current transformers (CTs) and protective relays must remain accurate under high harmonic conditions.

4.5) Switching Transients

Transients resulting from switching are often difficult to predict precisely due to their random and rapidly damping nature. Therefore, it is recommended that designers be provided with transient study data, for example in the form of oscillograms or time-to-harmonic curves.

For design purposes, it is helpful to have information on:

• The magnitude of harmonic currents flowing through each filter leg (capacitor, reactor, resistor),
• The current distribution based on the harmonic sequence
• The current-versus-time profile during the worst-case transient conditions.

This information allows designers to determine the final component rating that will withstand the worst-case conditions that may occur in the system.

———————————————

References

• IEEE 3002.8-2018: Recommended Practice for Conducting Harmonic Studies and Analysis of Industrial and Commercial Power Systems
• IEEE 1531-2020: IEEE Guide for the Application and Specification of Harmonic Filters
• IEEE Std 519-2014: Recommended Practice and Requirements for Harmonic Control in Electric Power Systems

———————————————

Contact Omazaki Engineering if you are looking for a designer or consultant for harmonic filter design or design services or harmonic filter design for electrical systems of industrial and commercial facilities in Indonesia and Southeast Asia.

———————————————