Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61023: Marine Speed Logs (SDME) — Principles, Installation, and Engineering Practice
Abstract: IEC 61023 defines the minimum performance requirements and test methods for marine Speed and Distance Measuring Equipment (SDME), as mandated by SOLAS Chapter V Regulation 19 and aligned with IMO Resolution MSC.96(72). This article examines the engineering principles behind Doppler, electromagnetic, and correlation speed logs; the critical distinction between Speed Through Water (STW) and Speed Over Ground (SOG); sensor installation practices that avoid boundary-layer errors and aeration; calibration procedures; and field-proven troubleshooting strategies for reliable ship speed measurement.
1. Understanding IEC 61023: Scope and Core Requirements
IEC 61023, now in its third edition (2007-06), applies to devices that indicate speed and distance as required by the International Convention for the Safety of Life at Sea (SOLAS). The standard is a direct implementation of IMO Resolution MSC.96(72), with clause numbering cross-referenced throughout the document. The general environmental and EMC testing framework is inherited from IEC 60945, the umbrella standard for maritime navigation equipment.
In engineering terms, SDME is not a standalone sensor—it is a system. Clause 4.1(e) explicitly states that the SDME consists of a sensor, a processing unit that converts the sensor output into speed and distance values, and the display or transmission interfaces that deliver those values to other shipboard equipment. If any one element in this chain is compromised, the entire speed-measurement pipeline fails.
💡 Engineering Insight: Radar plotting aids (ARPA/MARPA) and track control systems (TCS) require speed-through-water data in the fore-and-aft direction. A 2% STW error may seem minor, but over a 6-minute ARPA tracking window, it translates to a positional error of approximately 0.2 cables (37 metres) at 20 knots—enough to shift a CPA/TCPA collision alarm boundary in congested fairways.
1.1 The Six-Component Velocity Model
Annex Figure 1 of IEC 61023 defines a comprehensive ship speed velocity model that many engineers first encounter during type-approval testing. The six components are:
- Vxg — Forward speed over the ground (along the ship’s longitudinal x-axis)
- Vyg — Athwartship (lateral) speed over the ground
- Vxw — Forward speed through the water
- Vyw — Athwartship speed through the water
- Vg — Resultant speed vector over the ground
- Vw — Resultant speed vector through the water
The diagram also introduces Ve (speed due to propeller thrust aligned with the ship’s heading), S (speed due to wind/leeway), and t (speed due to tidal current). The engineering significance is that Vw = Ve + S, while Vg = Vw + t. Understanding this vector addition is essential when diagnosing discrepancies between STW and SOG readings during sea trials.
| Velocity Component | Definition | Typical Sensor | Primary Application |
|---|---|---|---|
| Vxw (Forward STW) | Longitudinal speed relative to adjacent water mass | Electromagnetic sensor, Doppler (water-track) | ARPA collision avoidance, dead reckoning |
| Vxg (Forward SOG) | Longitudinal speed relative to seabed | Doppler (bottom-track), GNSS | Precision navigation, berthing |
| Vyw / Vyg (Lateral) | Speed perpendicular to fore-aft axis | Dual/tri-axis Doppler sonar | Dynamic positioning (DP), berthing |
| Vw / Vg (Resultant) | Magnitude of combined vector | Multi-axis SDME | Situational awareness |
2. How Marine Speed Logs Work: Three Technologies Compared
2.1 Electromagnetic Speed Log
The electromagnetic (EM) speed log operates on Faraday’s principle of electromagnetic induction. A sensor head flush-mounted in the hull generates an alternating magnetic field that penetrates the conductive seawater flowing beneath the vessel. As seawater (a conductor with roughly 3-5 S/m conductivity at 35 ppt salinity) moves through the magnetic field, an electromotive force (EMF) is induced across a pair of sensing electrodes arranged perpendicular to both the field and the flow:
E = B · L · v · k
where B is the magnetic flux density (typically 50-200 μT for a practical sensor), L is the effective electrode separation (0.05-0.15 m), v is the flow velocity, and k is a dimensionless calibration coefficient determined empirically during commissioning.
⚠ Real-World Pitfall: The k-coefficient is not a factory-set constant. It drifts with hull paint thickness (especially after dry-docking with fresh anti-fouling coating), sacrificial anode depletion (which alters the local electric field around the sensor), and seawater conductivity variations. A chemical tanker was once found to read STW 0.3 knots higher than actual after a shipyard period because the new hull coating reduced the effective sensing volume—the accumulated dead-reckoning error exceeded 14 NM over a 48-hour passage.
The fundamental limitation of EM logs is that they cannot measure Speed Over Ground. The magnetic field decays exponentially with distance from the sensor face (skin depth is on the order of millimetres at typical excitation frequencies of 50-400 Hz), so bottom-tracking is physically impossible. This makes EM logs suitable for basic SOLAS compliance but insufficient for applications that demand SOG.
2.2 Doppler Speed Log
Doppler speed logs transmit ultrasonic pulses (carrier frequencies typically 100-300 kHz for deep-water operation; modern high-resolution units reach 600 kHz to 1 MHz) into the water column. Two measurement modes are possible:
- Bottom-Track (SOG): The acoustic pulse reflects off the seabed. The Doppler shift Δf between transmitted and received frequencies yields the ship’s velocity relative to the ground. This mode works only when water depth is within the acoustic range of the transducer (typically up to 200-300 m for a 200 kHz system).
- Water-Track (STW): The pulse scatters off suspended particles, plankton, and microbubbles in a water layer 5-20 m below the keel. This yields speed through water but is subject to the scatterer’s own motion relative to the water mass.
The Doppler frequency shift follows:
Δf = 2 · f0 · v · cos(θ) / c
where f0 is the transmitted frequency, v is the relative velocity, θ is the beam angle relative to the horizontal (typically 30° from the vertical, i.e., 60° from the horizontal), and c is the speed of sound in seawater (~1500 m/s, varying with temperature, salinity, and pressure).
The standard Janus configuration uses two (or four) beams angled symmetrically fore and aft. By differencing the Doppler shifts from opposite beams, the effects of heave and pitch are removed to first order. Four-beam (phased-array) Janus systems also capture lateral velocity components, enabling full horizontal velocity vector measurement for dynamic positioning applications.
✅ Design Insight: The Janus beam geometry provides inherent error cancellation. If the ship pitches by a small angle δ, the fore-beam Doppler shift increases by an amount proportional to cos(θ-δ) while the aft-beam shift decreases proportionally. Averaging the two largely cancels the pitch-induced error—but only for small angles. This is why IEC 61023 specifies pitch limits of ±5° and roll limits of ±10° for guaranteed performance.
2.3 Acoustic Correlation Log
The acoustic correlation log (ACL) takes a fundamentally different approach. Instead of measuring Doppler shift, it uses two (or more) receive transducers aligned along the ship’s longitudinal axis. As the ship moves forward, the seabed echo pattern “sweeps” across the receivers: the signal arriving at the aft receiver is a time-delayed version of what the forward receiver saw moments earlier. By computing the cross-correlation between the two received signals and finding the time lag τ that maximises the correlation coefficient, the ship’s speed can be determined as v = d / τ, where d is the known transducer separation.
The ACL has a subtle advantage: it does not require precise knowledge of the beam angle θ, making it less sensitive to installation misalignment and hull flexure. It also performs better over soft seabeds (mud, silt) where Doppler bottom-track can struggle due to weak specular returns.
| Technology | Physical Principle | Measures | Key Advantage | Key Limitation |
|---|---|---|---|---|
| Electromagnetic | Faraday induction: E = B·L·v·k | STW only | Simple, low cost, low power | Water-layer only; conductivity-dependent |
| Doppler | Acoustic Doppler shift: Δf ∝ v·cos(θ) | STW + SOG | Bottom-track in shallow water; multi-axis capability | Costly; deep-water limited to water-track |
| Acoustic Correlation | Cross-correlation of seabed echoes | STW + SOG | Less sensitive to attitude; better over soft seabeds | Complex signal processing; precise transducer alignment needed |
3. Installation, Calibration, and Data Interfaces
3.1 Sensor Location: The Boundary Layer Problem
IEC 61023 Clause 4.5 mandates that no part of the sensor penetrating the hull shall create a risk of water ingress—a safety requirement that seems obvious but drives important mechanical design choices (double O-ring seals, hull isolation valves for retractable sensors, watertight cable glands rated to the vessel’s watertight subdivision standard).
The standard also requires the manufacturer to document installation recommendations, particularly sensor positioning, because location directly determines measurement accuracy. The core challenge is the turbulent boundary layer that forms along the hull:
- Boundary Layer Thickness: On a 200 m vessel at 15 knots, the turbulent boundary layer can reach 30-60 cm thick at midship. The EM or acoustic sensor face must protrude beyond this layer—typically by 15-25 mm for EM sensors, or 150-300 mm for Doppler transducer arrays (often achieved with a faired blister or gate valve mounting).
- Aeration and Bubble Sweep-Down: In heavy seas or at shallow draft, the bow wave entrains air bubbles that are swept along the hull bottom. If the sensor is located in the forward third of the hull, these bubbles can completely blanket the transducer face, causing signal dropout—the speed reading plummets to zero regardless of actual vessel speed. The recommended sensor location is between 1/3 and 1/2 of the vessel length aft of the stem, offset 0.5-1.0 m from the centreline to avoid the keel wake and the bubble stream.
- Thruster and Propeller Interference: Bow thrusters generate intense turbulent jets that can extend 5-10 m aft along the hull. Sensors should be positioned at least 3-5 m longitudinally from any thruster tunnel opening.
✅ Best Practice: For merchant vessels (bulk carriers, tankers, container ships), the optimal sensor location is in the forward half of the hull, roughly 1/3 to 1/2 LBP (length between perpendiculars) aft of the stem, on the flat bottom plating, offset 0.5-1 m to port or starboard of the keel. This region has a moderate boundary layer thickness (typically < 20 cm), low aeration probability, and avoids the propeller wake entirely.
3.2 Accuracy Requirements and Calibration
IEC 61023 Clause 4.3 sets the accuracy bar:
| Parameter | Requirement | Test Reference |
|---|---|---|
| Digital speed display | ±2% of ship speed or ±0.2 kn, whichever is greater | Clause 5.12.1: Incremental testing 1 kn to 5 kn, then 5 kn steps to max speed |
| Analogue speed display | ±2.5% of ship speed or ±0.25 kn, whichever is greater | Clause 5.12.1 |
| Digital data output | ±2% of ship speed or ±0.2 kn, whichever is greater | Clause 5.12.1 |
| Distance run (1 hour) | ±2% of distance or ±0.2 NM per hour, whichever is greater | Clause 5.12.2: 5 kn for ≥60 min, 5 kn increments to max speed, ≥5 NM each |
| Contact closure (pulse output) | One closure per 0.005 NM, pulse width ≥50 ms, interval within ±2% | Clause 5.9.1: ≥10 kn constant speed, ≥10 consecutive closures |
Field calibration typically employs the double-run method: the vessel steams at a constant RPM and heading over a measured baseline (two precisely surveyed transit marks or GNSS waypoints) in both directions, cancelling the effect of tidal current. The calibration run should cover at least 5 NM to ensure statistical significance of the accumulated distance measurement. For Doppler systems with bottom-track capability, a direct comparison with DGPS-derived SOG is also acceptable, provided the vessel maintains straight-line, constant-speed motion for at least 15 minutes in calm sea conditions (< Sea State 2).
🔴 Common Calibration Failure: The most frequent cause of invalid sea-trial calibration is failure to account for cross-track current. If the calibration baseline is not aligned with the prevailing tidal stream, the vessel must crab (maintain a drift angle) to stay on track, and the longitudinal speed sensor reads the cosine component of the true through-water speed—an error that can exceed 1% even at small drift angles (3-5°). Always orient the calibration run parallel to the predicted tidal stream direction.
3.3 NMEA Data Output: VBW, VLW, and Pulse Interfaces
IEC 61023 Clause 4.2(d)(ii) requires SDME to support at least the following IEC 61162-1 (NMEA 0183) sentences:
- VBW — Dual Ground/Water Speed: Transmits longitudinal and transverse speed over ground and through water, along with data-validity status flags. Example:
$--VBW,12.3,0.5,A,10.8,0.3,A,*hhwhere the first pair (12.3, 0.5) is SOG longitudinal/transverse, the second pair (10.8, 0.3) is STW longitudinal/transverse, and “A” indicates valid data. - VLW — Distance Traveled Through Water: Transmits total accumulated water-track distance in nautical miles, with a range of 0 to at least 9,999.9 NM in 0.1 NM increments. Example:
$--VLW,15234.5,N,15234.5,N*hh.
Additionally, Clause 4.2(d)(i) mandates a hardware contact-closure output: one dry-contact pulse per 0.005 NM of distance run, with a minimum pulse width of 50 ms. This legacy interface, dating back to the electromechanical taffrail log era, remains critical because many radar processors, ECDIS installations, and autopilot track-control systems still use it as their primary speed input for dead reckoning. It also serves as a GNSS-independent speed source—valuable for resilience against GPS jamming or spoofing.
4. Reliability Engineering and Troubleshooting
4.1 Common Failure Modes and Root Causes
| Symptom | Likely Cause | Diagnostic Approach | Corrective Action |
|---|---|---|---|
| Erratic, noisy speed readings | Marine fouling on sensor face; damaged cable shielding | Diver inspection of sensor; megger test of cable insulation; continuity check on shield | Clean sensor face with plastic scraper (never steel—it scratches the acoustic window); repair or replace cable |
| Speed drops to zero at high vessel speed | Aeration/bubble blanking; poor sensor placement | Reproduce fault at different speeds and sea states; correlate with pitch/heave conditions | Extend sensor protrusion or add fairing; relocate sensor further aft on the hull |
| STW consistently lower than expected | Sensor recessed within boundary layer; coating or deposit build-up on sensor face | Measure sensor face protrusion relative to hull plating; compare with reference STW (e.g., towed log) | Install fairing block to raise sensor into free stream; clean and recalibrate |
| Doppler loses bottom-track in deep water | Depth exceeds acoustic range; soft mud seabed with weak backscatter | Check echo sounder depth; verify seabed type on chart | Switch to water-track mode; in deep ocean, combine with GNSS SOG |
| NMEA data dropouts, checksum errors | Baud rate mismatch; loose terminations; common-mode noise on signal lines | Monitor with NMEA analyser tool; check wiring continuity and termination | Verify consistent baud rate, data bits, parity; use shielded twisted pair with single-ended grounding |
| Distance accumulation does not match GPS | Contact-closure circuit fault; counter scaling factor mismatch | Verify 200 pulses per NM scaling (1 pulse / 0.005 NM) | Reconfigure receiving equipment scaling factor; check relay contact resistance |
4.2 Preventive Maintenance Strategy
💡 Field Experience: In tropical and subtropical waters, barnacle growth on sensor faces can degrade speed accuracy by more than 0.5 knots within a single week. Vessel operators should include quarterly underwater hull inspections (including sensor-face visual check and cleaning) in their Planned Maintenance System (PMS). Between dry-dockings, an in-water ultrasonic cleaning of the sensor face every 3-6 months is recommended for vessels trading in high-fouling regions. For vessels at extended anchorage, power up the SDME periodically (at least once every two weeks) to prevent electrode polarisation in EM sensors and discourage marine growth from establishing on static surfaces.
Q1: How do I choose between an electromagnetic log and a Doppler log?
The decision hinges on whether your operation requires Speed Over Ground (SOG). If you only need STW for basic SOLAS compliance and ARPA input, an EM log offers excellent value (typically USD 3,000-8,000 installed). If your vessel operates in shallow coastal waters, narrow channels, or requires berthing assistance, a Doppler log with bottom-track capability (USD 15,000-40,000) provides the SOG data needed for precision manoeuvring. Dynamic positioning (DP) vessels almost always require a multi-axis Doppler system to capture lateral velocity and rate-of-turn information. A useful rule of thumb: if your vessel’s operating profile includes more than 30% of time in waters shallower than 200 m, the investment in Doppler bottom-track capability will pay for itself in improved situational awareness.
Q2: Why does my speed log read differently in shallow water versus deep water?
This is the shallow-water effect. In water depths less than approximately 3-5 times the vessel draft, the flow beneath the hull is constrained between the ship’s bottom and the seabed, creating a Venturi-like acceleration of the water passing under the keel. An STW sensor measuring this accelerated flow will read higher than the true free-stream speed through water. IEC 61023 Clause 4.1(a) explicitly states that STW-measuring devices must meet the performance standard only in water depths greater than 3 m beneath the keel—below this depth, the standard acknowledges that shallow-water effects can produce systematic errors. If your log reads lower than expected in shallow water, the more likely cause is sediment disturbance from the propeller wash reducing the effective sensing volume, not a water-depth effect per se.
Q3: What changed between IEC 61023:1999 and the current 2007 edition?
The third edition introduced three substantive technical changes aligned with IMO Resolution MSC.96(72) of 2000. First, the minimum under-keel water depth for ground-referenced SDME was reduced from 3 m to 2 m, reflecting improved sensor sensitivity in modern Doppler systems. Second, the accuracy requirement for analogue speed displays was relaxed from 2% to 2.5% or 0.25 knots, acknowledging the inherent reading uncertainty of analogue meter movements on a vibrating ship’s bridge. Third, and most significantly for system integrators, the standard introduced a mandatory requirement for a serial digital interface conforming to IEC 61162-1 (NMEA 0183), specifically requiring support for the VBW and VLW sentences. This reflected the industry-wide transition from discrete analogue and pulse-only interfaces to integrated digital navigation data networks.
Q4: What is the contact-closure pulse output actually used for in modern bridge systems?
The 0.005 NM pulse output (minimum 50 ms width) is essentially a hardwired distance integrator that remains relevant for three reasons. First, it feeds the ARPA/MARPA tracking processor, which accumulates pulses to compute ship’s speed for target-motion vector resolution—many legacy radar processors still preference this hardwired input over NMEA serial data because it has lower latency and no parsing overhead. Second, it provides dead-reckoning input to ECDIS when GNSS is unavailable or suspect—the pulse train, combined with gyrocompass heading, enables continued position estimation independent of satellite signals. Third, it serves as a cyber-resilient navigation input: unlike NMEA 0183 serial data, a contact-closure pulse chain cannot be spoofed, jammed, or corrupted by software faults. In an era of growing concern about GNSS vulnerability, this simple electromechanical interface provides a valuable layer of navigation system diversity.
