IEC 61803: Determination of Power Losses in HVDC Converter Stations

Key Insight
IEC 61803 establishes a unified methodology for calculating and measuring power losses in HVDC converter stations, covering every component from thyristor valves to auxiliary systems, enabling accurate efficiency comparison across competing designs and technologies.

1. Scope and Loss Classification Framework

IEC 61803, first published in 1999 and updated in 2016, provides a comprehensive framework for determining the total power losses of HVDC converter stations operating under both line-commutated (LCC) and voltage-source (VSC) converter topologies. The standard classifies losses into five primary categories: converter transformer losses, valve losses, harmonic filter and reactive compensation losses, DC smoothing reactor losses, and auxiliary power consumption. This systematic classification enables utilities and manufacturers to assign loss costs with high precision in project tenders and performance guarantees.

The standard mandates that loss determination be performed at defined reference operating points — typically nominal DC power, nominal DC voltage, and specified AC system conditions — ensuring comparability across different bids. For bipolar configurations, the losses are calculated for one pole and then doubled, with adjustments for the neutral bus and electrode line losses.

Engineering Insight: In a typical 800 kV, 8000 MW HVDC link, total converter station losses account for approximately 0.5–0.8% of transmitted power per station. Over a 30-year project life, a 0.1% difference in loss rate can represent millions of dollars in capitalized loss cost — which is why IEC 61803’s rigorous methodology is the contractual cornerstone of every major HVDC project.

2. Component-Level Loss Calculation Methods

2.1 Converter Transformer Losses

Converter transformers operate under combined AC voltage and DC current stress, with significant harmonic content. IEC 61803 specifies that load losses (I²R losses in windings) must be calculated considering the actual current spectrum — including fundamental and characteristic harmonics (12-pulse: 11th, 13th, 23rd, 25th, etc.). No-load losses are determined from the applied voltage waveform, which includes DC offset components. The standard provides correction factors for stray losses in structural parts due to harmonic frequencies, as eddy-current losses increase approximately with the square of frequency.

Loss Category Key Components Typical Share of Total Losses Primary Calculation Method
Converter transformer Winding I²R, core no-load, stray losses 35–45% Spectrum-based harmonic summation
Thyristor/IGBT valves Conduction, switching, snubber, leakage 25–30% Statistical switching energy integration
AC harmonic filters Capacitor dielectric, reactor, resistor 10–15% Filter branch current summation
DC smoothing reactor Winding, core, stray losses 5–8% DC + harmonic current method
Auxiliary systems Cooling pumps, fans, controls, lighting 5–10% Measured input power summation

2.2 Valve Losses

For LCC-HVDC, valve losses comprise conduction losses (dominated by the thyristor on-state voltage drop, typically 1.5–2.5 V per device), switching losses (turn-on and turn-off energy, though these are relatively small for line-commutated thyristors operating at 50/60 Hz), snubber circuit losses, and gate unit power consumption. For VSC-HVDC using IGBTs or IGCTs, switching losses become substantially more significant due to higher switching frequencies (1–2 kHz for modular multilevel converters, though individual cell switching can be lower). IEC 61803 requires statistical methods that account for the distribution of valve currents and voltages across the hundreds of series-connected devices in a typical high-voltage valve.

Design Warning: Snubber circuit losses are frequently underestimated in early design phases. For a 12-pulse LCC valve group operating at 400 kV DC, the cumulative RC snubber losses across all series-connected thyristor levels can reach 100–150 kW per valve — equivalent to 2–3% of total station losses. Optimizing snubber component values using the standard’s recommended calculation method is essential for minimizing these losses without compromising voltage sharing.

3. Measurement and Type Testing Requirements

IEC 61803 distinguishes between calculated losses (used during the design phase and tender evaluation) and measured losses (verified during type testing and commissioning). For prototype validation, the standard defines a back-to-back test configuration where two identical converter stations are operated with one rectifying and the other inverting, sharing a common DC bus. The net power input from the AC system represents the total system losses, which are then apportioned to each station. This method requires careful instrumentation with class 0.2 accuracy for voltage and current transformers and class 0.5 for power analyzers.

The latest edition also addresses the unique challenges of measuring losses in VSC-HVDC systems, where the high-frequency switching components create measurement bandwidth requirements up to 50 kHz. Rogowski coils and optical current transducers are preferred over conventional current transformers for these measurements due to their superior high-frequency response and absence of magnetic saturation.

4. Frequently Asked Questions

Q1: What is the difference between “declared losses” and “guaranteed losses” in an HVDC contract?

Declared losses are the manufacturer’s best-estimate losses provided during the tender stage, calculated per IEC 61803 methodology. Guaranteed losses are contractually binding maximum values, typically set 3–5% above declared losses. If measured losses during commissioning exceed guaranteed values, the contract usually includes liquidated damages proportional to the present value of the excess energy loss over the project lifetime.

Q2: How do VSC-HVDC losses compare to LCC-HVDC losses?

Modern VSC-HVDC systems (using modular multilevel converters, MMC) have total station losses approximately 15–25% higher than equivalent LCC systems — roughly 1.0–1.2% per station for VSC versus 0.5–0.8% for LCC. The difference is primarily due to higher semiconductor switching losses in VSC valves and the additional losses in the submodule capacitors. However, VSC offers operational advantages such as black-start capability and independent reactive power control that often outweigh the efficiency penalty.

Q3: Can IEC 61803 be applied to multi-terminal HVDC systems?

Yes, but with modifications. The standard’s framework extends to multi-terminal configurations by treating each converter station individually and adding the DC network losses (cable/overhead line losses) separately. For multi-terminal systems, coordinating loss measurement across all stations simultaneously is challenging, and the standard recommends using synchronized phasor measurement units (PMUs) for time-aligned power readings.

Q4: What accuracy class of instrumentation does IEC 61803 require for factory acceptance testing?

For formal loss measurement during type testing, the standard requires current and voltage transformers with accuracy class 0.2 or better, and power analyzers with class 0.5 or better. All instruments must be calibrated within 12 months of the test date. Temperature measurements must be accurate to within 0.5 °C, as winding resistance (and therefore I²R losses) varies approximately 0.4% per °C for copper conductors.

© 2026 TNLab — HVDC Converter Station Loss Analysis Reference

Leave a Reply

Your email address will not be published. Required fields are marked *