ACTIVE OR PASSIVE FILTERS? HOW TO CHOOSE THE BEST SOLUTION FOR YOUR INDUSTRIAL INSTALLATION

When harmonics increase, the problem is not just “electrical”: it becomes operational and economic. Overheating grows, nuisance tripping increases, equipment reliability decreases, and installation expansions can change network behaviour over time more than one might expect.

In this context, choosing between an active filter and a passive filter does not mean comparing two products in the abstract. It means understanding which solution is truly consistent with the harmonic spectrum, load variability, installation architecture, and retrofit constraints.

For a general overview of the different families of power factor correction and power quality solutions, the reference remains the dedicated guide (insert link to pillar: Power factor correction and active filters) on traditional power factor correction, SVG, and active filters. Here, instead, the focus is more operational: how to make a sound decision between passive and active, where to install the solution, and which mistakes to avoid.

BEFORE CHOOSING: 4 THINGS YOU NEED TO KNOW

Before even comparing technologies, some basic conditions need to be clarified. This is where many decisions start well, or begin to unravel.

THDI AND SPECTRUM: STABLE OR VARIABLE?

The first real differentiator is not the name of the technology, but the actual behaviour of harmonics over time.

If the spectrum is fairly stable, with few dominant harmonics and predictable loads, a passive solution may make more sense. If, on the other hand, harmonics change according to production cycles, start-ups, shifts, or the evolution of the installation, an active solution tends to be more robust.

It is therefore not enough to know that “THDi is high.” You need to understand:

• how high it is;
• which harmonics dominate;
• whether the profile is stable or changes throughout the day;
• whether the problem is concentrated or distributed.

HOW MANY SOURCES DO YOU HAVE, AND WHERE ARE THEY?

An installation with a single significant harmonic source is very different from one where sources are numerous, scattered, and distributed across multiple panels.

This distinction completely changes the choice:

• few concentrated sources favour more localised approaches;
• many diffuse sources often make a dynamic, modular logic more useful;
• installations that have grown through successive layers almost always require an architectural reading, not just a single-panel view.

3-PHASE OR 3-PHASE + NEUTRAL NETWORK?

This is a question that is often underestimated. With many single-phase loads, LED lighting, switching power supplies, UPS, and distributed offices, the issue is not only distortion across phases: the neutral also comes into play, especially when the 3rd harmonic is significant.

If the network is 3-phase + neutral, ignoring the neutral and triplen harmonics can lead to an incomplete choice. In some cases, the problem observed in the field does not stem so much from reactive power or a single dominant harmonic, but from a combination of single-phase loads stressing the neutral and making overall behaviour more unstable.

RETROFIT AND INSTALLATION CONSTRAINTS

Even the technically correct solution can prove unsuitable if it does not account for the real context:

• limited space available;
• restricted ventilation;
• difficult access;
• inability to accept long plant shutdowns;
• existing power factor correction equipment;
• need to work in successive phases.

In a retrofit, the best filter is not the “highest-performing one in absolute terms,” but the one that resolves the problem without introducing unmanageable side effects or complexity.

WHEN A PASSIVE FILTER IS THE RIGHT CHOICE

A passive filter can be a valid choice, but only when the installation context meets certain precise conditions. It works well when the problem is sufficiently defined, predictable, and stable.

THE 5 CONDITIONS THAT SHOULD HOLD TRUE

A passive filter is generally more suitable when:

• the harmonic spectrum is stable over time;
• there are few dominant, clearly identifiable harmonics;
• loads are fairly predictable;
• the objective is relatively fixed and well-defined;
• the project is managed carefully with attention to coordination, tuning, and operating conditions.

When these conditions are present, a passive solution can offer an effective and targeted response.

Warning

If even one of these conditions fails, especially if the spectrum changes frequently, the neutral is critical, or the installation is continuously evolving, it is worth evaluating an active solution or a hybrid architecture.

WHEN AN ACTIVE FILTER IS THE MORE ROBUST CHOICE

An active filter tends to be the stronger choice when the installation is dynamic, distributed, or difficult to capture with fixed rules. Not because it is always the best solution, but because it handles complexity better.

6 FIELD TRIGGERS THAT POINT TOWARDS AN ACTIVE FILTER

An active filter generally becomes a more compelling choice when one or more of the following conditions apply:

• variable harmonic spectrum during production cycles and load changes;
• many non-linear sources distributed across multiple panels;
• frequent expansions or retrofit situations where unwanted resonance must be avoided;
• significant 3rd harmonic presence and a critical neutral;
• recurring protection trips, hotspots, or overheating;
• need for stable results over time, even as the load mix changes.

When installation behaviour changes frequently, a static solution risks chasing the problem with less flexibility than required.

WHERE TO INSTALL IT: MAIN SWITCHBOARD OR LOCAL PANELS?

This is a very important decision at the design stage. Sometimes the right type of filter is chosen, but installed in the wrong location. This mistake undermines much of the result.

ARCHITECTURE A – ONE CENTRAL FILTER IN THE MAIN SWITCHBOARD

A central filter makes sense when:

• the problem is fairly distributed and a global improvement is needed;
• a more centralised solution is required;
• the installation can be read from a single main supply point;
• harmonic distribution does not require highly localised corrections.

Advantage: overall visibility and centralised intervention.

Limitation: may be less effective if harmonics are heavily concentrated near specific sources or local areas.

ARCHITECTURE B – LOCAL PANEL FILTERS

Installing correction close to the sources is often more effective when the problem originates in specific areas of the installation.

This solution is useful when:

• harmonics are concentrated in one or more sections;
• non-linear loads are clearly identifiable;
• you want to act close to the source of the disturbance;
• the installation behaves very differently from one zone to another.

Advantage: greater local effectiveness and better source control.

Limitation: more distributed project with more points to coordinate.

ARCHITECTURE C – HYBRID: CENTRAL + LOCAL

This is a very sensible approach in installations that have grown through successive layers, with some diffuse disturbance and some highly concentrated disturbance.

Here the objective is not to choose “a single correct location,” but to combine:

• general upstream correction;
• dedicated mitigation where sources are most critical.

Advantage: greater flexibility.

Limitation: requires more thorough analysis and coordination.

ARCHITECTURE D – CRITICAL POINTS AND SENSITIVE LOADS

In some cases, the priority is not to improve the entire network, but to protect a specific line, a sensitive load, or a high-value operational area.

This applies, for example, to:

• lines with sensitive electronics;
• critical process areas;
• sections where even minor anomalies cause shutdowns or significant disturbances.

Here the filter becomes a tool for selective protection, not just general improvement.

QUICK ARCHITECTURE COMPARISON TABLE

APPLICATION SCENARIOS: WHICH CHOICE MAKES THE MOST SENSE

To choose well, it helps to think in terms of scenarios. Inverters are not the only source: harmonics appear in very different contexts, with different constraints.

DATA CENTRES AND UPS

Here the central theme is the relationship between distortion and reliability. Loads are often electronic, continuity is critical, and performance monitoring matters greatly.

In this context, the choice often tends towards solutions capable of managing dynamic conditions and advanced non-linear loads well.

Passive / active / hybrid: why

• Passive: can work only in very stable, well-characterised scenarios;
• Active: often preferable given the variability and sensitivity of the context;
• Hybrid: useful when power quality objectives and specific architectural constraints coexist.

LED LIGHTING, COMPLEX BUILDINGS, AND SINGLE-PHASE LOADS

Here the 3rd harmonic and the neutral become more prominent issues. In technical offices, complex buildings, tertiary structures, or installations with widespread LED lighting, the problem can be highly distributed and less intuitive to interpret.

Passive / active / hybrid: why

• Passive: less suitable when loads are diffuse and heterogeneous;
• Active: often more appropriate when the neutral and triplen harmonics are significant;
• Hybrid: useful when more critical areas coexist alongside more stable zones.

ADVANCED HVAC AND COMPLEX TERTIARY INSTALLATIONS

In modern HVAC systems, conditions change with seasonality, control adjustments, occupancy, and usage cycles. This makes harmonic behaviour less static than it might appear.

Passive / active / hybrid: why

• Passive: may be adequate if the profile remains fairly repetitive;
• Active: more robust when seasonal and operational variability is pronounced;
• Hybrid: useful in complex installations with multiple subsystems and different priorities.

MULTI-PANEL INDUSTRIAL INSTALLATIONS WITH SUCCESSIVE EXPANSIONS

This is one of the most typical retrofit scenarios: the installation has grown over the years, with inverters, UPS, automated lines, secondary panels, and electronic loads added over time. The result is a harmonic spectrum that changes and distributes in a less predictable way.

Passive / active / hybrid: why

• Passive: risks being too rigid if the system continues to evolve;
• Active: often the most solid choice when the load mix changes over time;
• Hybrid: very sensible when some areas have specific problems and others require general improvement.

TYPICAL MISTAKES THAT CAUSE HARMONIC FILTERING PROJECTS TO FAIL

This is often the most practically useful section, because many problems do not stem from the chosen technology, but from incorrect sizing.

SIZING IN KW INSTEAD OF COMPENSATION AMPERES

Harmonic filtering is not sized using the same logic applied to active power or process machinery. If the choice starts from kW rather than the actual harmonic current to be compensated, the risk of error is high.

PLACING EVERYTHING AT THE MOST CONVENIENT POINT, NOT THE RIGHT ONE

Sometimes the main switchboard is chosen simply because it is the most accessible. But if harmonics are concentrated elsewhere, installation convenience can translate into reduced effectiveness.

MEASURING ONLY A “ONE-OFF SNAPSHOT” RATHER THAN THE SPECTRUM OVER TIME

A single instantaneous measurement is useful, but insufficient. If the installation changes across shifts, cycles, seasons, or expansions, the project must be built on a more realistic time-based profile.

FAILING TO COORDINATE FILTERING WITH EXISTING POWER FACTOR CORRECTION

If power factor correction banks are already installed, ignoring them is a mistake. Any intervention on harmonics must be considered alongside what is already in place.

IGNORING THE NEUTRAL AND 3RD HARMONIC IN 3-PHASE + NEUTRAL NETWORKS

This is one of the most common mistakes in contexts with many single-phase loads and LED lighting. Focusing only on the phases means underestimating a significant part of the problem.

NOT VALIDATING THE RESULT WITH A BEFORE/AFTER COMPARISON

Without a comparative measurement, the result remains perceived rather than demonstrated. A well-executed project must also include verification of the benefit achieved.

TECHNICAL CHECKLIST TO USE BEFORE MAKING A DECISION

Before deciding between active, passive, or a hybrid architecture, it is worth gathering at least the following data:

• THDi profile over time;
• harmonic spectrum with dominant orders and variability;
• measurement point or points;
• network type;
• presence of the neutral and possible 3rd harmonic criticality;
• distribution of harmonic sources;
• installation constraints;
• any existing power factor correction;
• primary project objective: compliance, temperature reduction, fewer trips, improved reliability, better power quality.

This information base is needed not only to choose the technology, but also to decide where to install it and with which architectural logic.

FAQ

WHEN IS A PASSIVE FILTER SUFFICIENT?

It is sufficient when the harmonic spectrum is fairly stable, the dominant harmonics are known, loads are predictable, and the project is well coordinated with the rest of the installation. If behaviour changes frequently, it is worth evaluating an active filter or a hybrid solution.

WHEN IS AN ACTIVE FILTER PREFERABLE IN A RETROFIT?

It is often preferable when the installation has grown over time, the spectrum is variable, sources are distributed across multiple panels, and you want to reduce the risk of side effects in an already complex network.

IS IT BETTER TO HAVE ONE LARGE FILTER OR MULTIPLE SMALLER ONES PER PANEL?

It depends on where the harmonics originate. A central filter can be appropriate for overall improvement; multiple local panel filters are often better when sources are concentrated and you want to act close to where the disturbance is generated.

H3: HOW DOES THE CHOICE CHANGE IF I HAVE MANY LED OR SINGLE-PHASE LOADS?

In that case, it becomes very important to assess the 3-phase + neutral network, the neutral conductor, and the 3rd harmonic. When the problem is widespread and strongly involves single-phase loads, a dynamic solution tends to offer greater flexibility.

CAN I KEEP THE EXISTING POWER FACTOR CORRECTION?

Often yes, but only if the new intervention is properly coordinated with it. Ignoring the presence of existing power factor correction is one of the most common mistakes in harmonic filtering projects.

HOW DO I VERIFY THE RESULT?

By comparing measurements before and after the intervention: THDi, spectrum, temperatures, abnormal events, protection trips, and the behaviour of the most sensitive lines or loads. Without a comparative check, the benefit remains difficult to quantify.

THE RIGHT CHOICE COMES FROM REAL CONSTRAINTS, NOT THE “MOST POPULAR” TECHNOLOGY

Between active and passive filters, the winner is not the best-known solution, but the one most consistent with your installation. The harmonic spectrum, its variability, the network type, source distribution, neutral presence, retrofit constraints, and installation architecture matter more than the name of the technology.

That is why the right choice does not start with the question “which is the best filter?”, but with a more useful one: what type of harmonics do you have, where do they originate, and how does your installation change over time?