Executive Summary
A South African metallurgy plant needed a power-quality retrofit for a 2,500 kW electric arc furnace (EAF). The coordinated solution combined three 500 kvar Static Var Generators (SVGs) with two 400 A Active Harmonic Filters (AHFs). Recorded project results show power factor improving from 0.65 to 0.95–0.97 and THDi falling from above 21.5% to below 3%.
Project at a Glance
| Project item | Details |
|---|---|
| Location | South Africa |
| Industry | Metallurgy / steel processing |
| Critical load | 2,500 kW electric arc furnace |
| Baseline conditions | Power factor 0.65; THDi above 21.5%; thermal stress and nuisance tripping reported |
| Installed solution | 3 × 500 kvar SVG (1,500 kvar total) + 2 × 400 A AHF (800 A total) |
| Recorded results | Power factor 0.95–0.97; THDi below 3%; apparent power reduced from 3,846 kVA to 2,632 kVA |
1. The Power-Quality Challenge
Electric arc furnaces create rapidly changing, nonlinear loads. During arc ignition and melting, the plant recorded large reactive-power swings and characteristic harmonic current. The retrofit therefore had to address power factor and harmonic distortion as separate but coordinated engineering problems.
Low power factor: Average measured power factor was 0.65, increasing current and apparent-power demand for the same useful power.
Harmonic current: Project measurements recorded THDi above 21.5%, with prominent low-order components associated with the unstable three-phase arc.
Operational symptoms: The site reported feeder-cable heating, motor noise and repeated protective-relay trips on production lines.
Harsh environment: Conductive metal dust, graphite powder and elevated temperatures increased the enclosure and cooling requirements.
Eskom guidance identifies 0.96 as the nominal power-factor level below which an additional reactive-power charge may apply, depending on the customer tariff. The project target was therefore set at 0.95 or higher, with final readings assessed at the grid-side point of measurement.
External reference: Eskom Power Factor guidance
2. Reactive-Power Assessment and Equipment Sizing
2.1 Theoretical Compensation Requirement
For a 2,500 kW load, improving power factor from 0.65 to 0.95 gives the following theoretical full-load requirement:
Step 1 — Baseline angle: tan φ₁ = √(1 − 0.65²) / 0.65 ≈ 1.1691
Step 2 — Target angle: tan φ₂ = √(1 − 0.95²) / 0.95 ≈ 0.3287
Step 3 — Required compensation: Qc = P × (tan φ₁ − tan φ₂) = 2,500 × (1.1691 − 0.3287) ≈ 2,101 kvar
This value is a full-load theoretical reference, not a substitute for time-series measurements. EAF demand changes across the melting cycle, so final equipment sizing must also consider the measured duty cycle, simultaneous load, compensation already available, harmonic current and the required margin.
Calculation reference: Eskom Power Factor Sizing Guide
2.2 Deployed SVG Configuration
Installed capacity Three 500 kvar floor-standing SVG units connected in parallel, providing 1,500 kvar of dynamic reactive-power capacity.
The 1,500 kvar deployed rating is below the 2,101 kvar theoretical full-load value. According to the project design basis, the final selection was made from the measured operating profile rather than nameplate power alone. Parallel modules also allowed staged output and continued partial compensation if one unit was unavailable. The recorded 0.95–0.97 grid-side power factor should be treated as the commissioning result for the measured operating conditions, not as a guarantee for every future load state.
2.3 Deployed AHF Configuration
Installed capacity Two 400 A floor-standing AHFs connected in parallel, providing 800 A of harmonic-compensation capacity.
The AHFs were configured to compensate the measured harmonic-current spectrum, with particular attention to the dominant low-order harmonics. The active filters inject compensating current dynamically, avoiding the fixed tuning of a passive LC filter. Final settings were verified at the project assessment point under representative furnace operation.
Product information: Sinava Static Var Generator | Sinava Active Harmonic Filter
3. Why the Project Used Coordinated SVG and AHF Equipment
| Project requirement | Primary response | Engineering purpose |
|---|---|---|
| Rapid reactive-power variation | SVG | Continuously supplies or absorbs reactive current to improve power factor and support voltage. |
| Harmonic current | AHF | Measures harmonic current and injects compensating current in real time. |
| Changing furnace stages | Parallel modular control | Allows output to follow the operating profile and supports staged capacity. |
| Maintenance availability | Multiple modules | Maintains partial compensation when one module is unavailable; it does not provide full-capacity redundancy. |
Traditional capacitor banks can be effective for stable base reactive-power demand, but step switching may not track an EAF’s fastest load changes. Capacitors also require harmonic and resonance assessment before application. For this project, the engineering team selected a fully power-electronic SVG + AHF arrangement to control reactive power and harmonic current independently.
4. Enclosure Design for Conductive Dust
The metallurgy environment required more than standard indoor cabinet construction. The project specification separated sensitive control electronics from the main cooling-air path and added measures intended to reduce conductive-dust ingress.
Protected control chamber: Core control and driver components were housed in an IP54-rated compartment with sealed door interfaces and slight positive pressure.
Separated cooling path: A rear forced-air duct carried cooling air across the heatsink path without routing workshop air across the control boards.
Filter maintenance: Metal-mesh filtration and periodic fan-reversal logic were specified to limit dust accumulation on the cooling path.
PCB protection: Control boards received conformal coating as an additional barrier against dust and moisture.
These features reduce environmental exposure, but they do not eliminate inspection, filter cleaning or preventive maintenance requirements.
5. Installation and Commissioning
5.1 Grid-Side Closed-Loop CT Location
The main sampling CTs were positioned downstream of the transformer low-voltage main breaker and upstream of the SVG, AHF and EAF branches. This allowed the controller to measure the residual grid current after compensation.
Connection sequence Transformer → Main breaker → Main CT sampling point → SVG branches → AHF branches → EAF load
5.2 CT Wiring and Signal Integrity
CT ratio: The project specified 5,000/5 A CTs to accommodate the approximately 4,000 A nominal LV current and furnace inrush margin; final ratio and accuracy class must match the approved design.
Cable routing: Shielded twisted-pair CT secondary wiring was routed separately from high-current EAF power cables, with single-point shield grounding.
Parallel-unit coordination: A centralized signal-distribution arrangement supplied consistent CT references to the three SVGs and two AHFs.
5.3 Commissioning Checks
Phase sequence and CT polarity: Verify voltage phase sequence, CT direction and measured phase angle before enabling compensation.
Parallel load sharing: Confirm communication, current sharing and protection settings across all SVG and AHF modules.
Representative EAF test: Record power factor and harmonic current through relevant furnace stages, then compare grid-side results with the agreed project targets.
6. Recorded Post-Retrofit Results
The following values come from the supplied project case material. They should be supported by the commissioning report or analyzer exports before public attribution as independently verified measurements.
| Metric | Before | After | Interpretation |
|---|---|---|---|
| Power factor | 0.65 | 0.95–0.97 | Improved grid-side power factor under measured operating conditions. |
| Apparent power demand | 3,846 kVA | 2,632 kVA | Calculated reduction of approximately 1,214 kVA at 2,500 kW. |
| Current distortion (THDi) | >21.5% | <3.0% | Below the project’s 5% commissioning target. |
| Operational reliability | Heating and nuisance trips reported | Thermal stress and nuisance tripping reduced | Outcome should be supported by maintenance and event records. |
Important “After” values describe the recorded commissioning conditions. They are not universal performance guarantees; results depend on system impedance, load profile, CT installation, control settings and equipment sizing.
7. Engineering Lessons from the Project
Measure before selecting equipment. Nameplate power alone is not enough for a fluctuating EAF load; time-series kvar and harmonic-current data are essential.
Separate the problems. Use SVG capacity for reactive-power control and AHF capacity for harmonic-current mitigation; do not treat THDi and power factor as the same metric.
Treat CT installation as part of the control system. Incorrect CT location, polarity or routing can undermine otherwise correct equipment selection.
Design for the environment. Dust management, cooling-path separation and maintainability are central to reliability in metallurgy plants.
Frequently Asked Questions
Why can an electric arc furnace require both SVG and AHF equipment?
An SVG primarily controls dynamic reactive power and power factor, while an AHF primarily compensates harmonic current. An EAF can create both problems at the same time, so the two devices are sized and controlled for different electrical objectives.
Why was the installed SVG capacity lower than the theoretical 2,101 kvar value?
The 2,101 kvar figure is a full-load calculation based on 2,500 kW and a power-factor change from 0.65 to 0.95. The project’s 1,500 kvar selection was based on the measured operating profile and duty cycle. The recorded result applies to those measured conditions.
What does grid-side closed-loop CT sampling mean?
The main CTs are installed upstream of the compensation branches so they measure the residual current drawn from the grid. The controllers then adjust SVG and AHF output using the actual grid-side result.
Does an active solution remove all resonance risk?
No engineering design should claim zero risk without a system study. Active devices avoid the fixed tuned branches used by passive filters, but the complete network still requires impedance, protection and control-coordination review.
What data is needed for a similar EAF project?
Useful inputs include the single-line diagram, system voltage, transformer and short-circuit data, EAF operating profile, kvar trend, power factor, harmonic spectrum, CT details, capacitor banks, event records and the target assessment point.
Conclusion
This South African EAF retrofit shows why reactive-power compensation and harmonic mitigation should be engineered as coordinated but distinct functions. Three parallel SVG units addressed the changing kvar demand, while two parallel AHFs targeted the measured harmonic current. The recorded results demonstrate the potential of a correctly measured, sized and commissioned solution in a demanding metallurgy environment.
Planning a similar project? Send Sinava Power your single-line diagram, load profile and power-quality measurements for a site-specific review. Contact Sinava Power