A complete reference for substation engineers covering Merging Units — IEC 61869-9 specification, processing delay limits, sample rates, variant codes, SmpSynch values, conformance classes, holdover mode, PTP/1PPS synchronization, procurement template, and field troubleshooting
Quick Reference
| Item | Value |
|---|---|
| What it does | Digitizes CT/VT signals and publishes Sampled Values on the process bus |
| Standard | IEC 61869-9:2016 (replaces 9-2LE guideline) |
| Transport | Raw Ethernet (Layer 2), EtherType 0x88BA |
| Default sample rate (50 Hz protection) | 4 800 samples/s with 2 ASDU per frame (preferred per IEC 61869-9) |
| Default sample rate (metering / power quality) | 14 400 samples/s with 6 ASDU per frame (preferred) |
| Max processing delay (protection) | 2 ms |
| Max processing delay (quality metering) | 10 ms |
| Time sync (preferred) | IEC/IEEE 61850-9-3 PTP Power Profile |
| Time sync (legacy) | 1PPS on dedicated fiber input |
| Holdover (minimum) | 5 s |
| Multicast MAC range | 01-0C-CD-04-00-00 to 01-0C-CD-04-01-FF |
| APPID range (SV) | 0x4000–0x7FFF |
| Max quantities per stream (100 Mbit/s) | 24 (general) / 8 (metering) |
| Connector | Duplex LC (preferred), BFOC/2.5 (ST, legacy) |
Introduction
A merging unit is the device that sits between conventional instrument transformers and the digital world of IEC 61850. It takes analog current and voltage signals from current transformers (CTs) and voltage transformers (VTs), samples them at a precise rate, time-stamps each sample, and publishes them on the process bus as digital Sampled Values (SV). Every protection IED, bay controller, or meter that needs to know what the power system is doing gets its measurements from the merging unit — not from analog wiring.
This article is the complete reference, verified against the IEC 61869-9:2016 Edition 1.0 standard. It covers what a merging unit does, how the IEC 61850 model represents it, the IEC 61869-9 communication profile, the variant codes, the processing delay limits, the synchronization requirements, conformance classes, ICD examples, and field troubleshooting.
1. What Is a Merging Unit?
A merging unit (MU) is an IEC 61850 process bus device that converts analog current and voltage signals from CTs and VTs into digital Sampled Values (SV) transmitted over Ethernet. It replaces traditional copper wiring between instrument transformers and protection relays with synchronized digital communication, allowing multiple IEDs to share the same measurements over a standard Ethernet network.
2. Why Merging Units Exist
Traditional substations run analog wiring from CTs and VTs to every protection and measurement device in the bay. A bay with five relays means five separate pairs of CT and VT secondary wiring, each carrying the same analog signal. That wiring is expensive, difficult to test, impossible to modify without physical rewiring, and a frequent source of installation errors.
A merging unit eliminates all of that. One device acquires the CT and VT signals for a bay, digitizes them, and publishes them over Ethernet. Any device on the process bus — protection IEDs, bay controllers, meters, fault recorders — receives the same digital sample stream. Adding a new subscriber means adding a software subscription, not running new copper.
The practical benefits: reduced copper wiring, simplified testing, easier engineering changes, standardized digital interface, deterministic timing through synchronization, and significantly reduced commissioning time.
3. What a Merging Unit Does — Step by Step
1. Signal acquisition. The MU connects to the secondary outputs of CTs and VTs. It accepts conventional CT/VT secondary levels — typically 1 A or 5 A for CTs, 100 V or 110 V for VTs — or the low-level analog outputs of electronic instrument transformers.
2. Analog-to-digital conversion. Input signals are sampled by ADCs at a fixed rate. The sampling must be precise and consistent. Jitter directly affects measurement accuracy and protection performance.
3. Anti-aliasing filtering. Before sampling, an anti-aliasing filter limits the input bandwidth. This filter contributes a known delay that the MU accounts for in its timestamps.
4. Time synchronization. Each sample is time-stamped. All merging units in a substation must sample at exactly the same instant for cross-bay protection functions (differential, distance) to work correctly. This requires synchronization to a common time reference — typically PTP per IEC/IEEE 61850-9-3 or a 1PPS optical input.
5. Sampled Value publishing. Digitized, time-stamped samples are packaged into SV messages and published as Ethernet multicast frames on the process bus. Subscribers receive the stream continuously.
6. Status and supervision. The MU monitors its own health and the health of the connected instrument transformers — fuse failures, communication loss, synchronization loss — and makes this information available over the station bus via MMS.
4. Merging Unit Hardware Architecture
A standard MU per IEC 61869-9 consists of:
| Block | Function |
|---|---|
| Primary sensors | Conventional CTs/VTs or low-power instrument transformers (LPIT) |
| Secondary converters (SC) | Analog-to-digital conversion blocks for each input channel |
| Anti-aliasing filter | Removes frequency content above the Nyquist limit |
| Processing unit | Scaling, message formatting, ASDU assembly |
| Clock input | PTP via Ethernet ports or dedicated 1PPS optical input |
| Digital output | Ethernet port(s) for SV stream(s) — typically 100BASE-FX fiber |
| Power supply | Auxiliary DC/AC supply (typically 80–300 V DC or 100–250 V AC) |
The MU’s two main signal interfaces are: (a) the instrument transformer primary on one side, (b) the digital output connector on the other. Everything in between is the MU’s responsibility.
5. Stand-alone Merging Unit (SAMU) vs Built-in MU
IEC 61869-9 distinguishes two physical realizations:
| Type | Description | Standard |
|---|---|---|
| Built-in MU | MU electronics integrated directly into an electronic LPIT | IEC 61869-7, -8, -12, -14, -15 |
| Stand-alone MU (SAMU) | Separate product accepting analog instrument transformer outputs | IEC 61869-13 |
The output of a SAMU and a built-in MU should be indistinguishable on the wire — both produce IEC 61869-9 compliant SV streams. SAMUs typically have lower end-to-end accuracy because they cascade the separately-rated CT/VT accuracy with the SAMU’s own accuracy.
6. IEC 61850 Model Inside a Merging Unit
IEC 61850 models a merging unit as a physical device containing one or more logical devices, each with a defined set of logical nodes.
Logical Device Naming
Each logical device name follows: xxxxMUnn where xxxx is the configurable IED name and MUnn is the LD instance (e.g., BAY01MU01).
LPHD — Physical Device Information
Every physical device has an LPHD logical node. IEC 61869-9 extends LPHD with mandatory nameplate data objects:
| Data Object | Type | Description |
|---|---|---|
PhyNam | DPL | Physical nameplate (vendor, model, serial number, hardware/software revisions, manufacture date) |
PhyHealth | Health | Overall hardware operational status |
NamVariant | VSD | Semicolon-separated list of supported variant codes (see Section 6) |
NamHzRtg | VSD | Supported nominal frequencies (e.g., dc;50;60) |
NamAuxVRtg | VSD | Rated auxiliary power supply voltages (e.g., 80-300 dc;100-250 ac) |
NamHoldRtg | VSD | Rated holdover time in seconds (e.g., 10) |
NamMaxDlRtg | VSD | Maximum processing delay time in microseconds (e.g., 1500) |
These attributes are mandatory and read-only. They define the MU’s electronic nameplate.
LLN0 — Logical Node Zero
LLN0 hosts the Sampled Value Control Blocks (MSVCB or USVCB) — one per SV stream — and the data sets that define which samples are published.
TCTR — Current Transformer Logical Node
TCTR represents one current measurement channel. A separate instance is used for each phase, typically four total (A, B, C, N).
The mandatory data object is AmpSv of type SAV (Sampled Analogue Value), carrying the sampled current value. IEC 61869-9 adds:
| Data Object | Description |
|---|---|
NamAccRtg | Accuracy class rating in format <measuring>/<protection> (e.g., 0.2S/5P20) |
NamARtg | Rated primary currents in amperes (e.g., 400;800;1200) |
NamClipRtg | Ratio of clipping limit to rated primary current × √2 (e.g., 20) |
TVTR — Voltage Transformer Logical Node
TVTR represents one voltage measurement channel. The mandatory data object is VolSv of type SAV. IEC 61869-9 adds:
| Data Object | Description |
|---|---|
NamAccRtg | Accuracy class rating in format <measuring>/<protection> (e.g., 0.5/3P) |
NamVRtg | Rated primary voltage in volts (e.g., for 300 kV/√3 → 173000) |
NamClipRtg | Ratio of clipping limit to rated primary voltage × √2 (e.g., 2) |
MMXU — Measurement Logical Node (Optional)
Some MUs implement MMXU, which calculates power system quantities from sampled values — phase currents, phase-to-earth voltages, active power, reactive power, frequency. These calculated values are published over the station bus via MMS. MMXU is optional.
7. Variant Codes Explained (FfSsIiUu Notation)
IEC 61869-9 defines a human-readable notation to describe MU capabilities concisely:
FfSsIiUu
| Letter | Meaning |
|---|---|
| F f | Digital output sample rate in samples/second |
| S s | Number of ASDUs (samples) per SV message |
| I i | Number of current quantities per ASDU |
| U u | Number of voltage quantities per ASDU |
Examples
| Variant Code | Description |
|---|---|
F4000S1I4U4 | 9-2LE MSVCB01 — 4000 sps × 1 ASDU × 4 currents + 4 voltages — for 50 Hz legacy systems |
F4800S1I4U4 | 9-2LE MSVCB01 — for 60 Hz legacy systems |
F4800S2I4U4 | Preferred — 4800 sps × 2 ASDU per frame × 4 + 4 — protection and measurement |
F14400S6I4U4 | Preferred — 14400 sps × 6 ASDU × 4 + 4 — quality metering |
F12800S8I4U4 | 9-2LE MSVCB02 — deprecated, 50 Hz only |
F15360S8I4U4 | 9-2LE MSVCB02 — deprecated, 60 Hz only |
F96000S1I4U4 | Preferred — high-bandwidth DC control applications |
A merging unit’s NamVariant data object lists all supported codes. The active configuration is in the MU’s SCL file.
8. Standard Sample Rates (IEC 61869-9 Table 902)
| Sample Rate (Hz) | ASDUs / Frame | Frames/sec | Application |
|---|---|---|---|
| 4000 | 1 | 4000 | 9-2LE backward compatibility (50 Hz) |
| 4800 | 1 | 4800 | 9-2LE backward compatibility (60 Hz), or 50 Hz with 96 samples/cycle |
| 4800 | 2 | 2400 | Preferred for protection and measurement |
| 5760 | 1 | 5760 | 60 Hz backward compatible with 96 samples/cycle |
| 12800 | 8 | 1600 | Deprecated (9-2LE 50 Hz MSVCB02) |
| 14400 | 6 | 2400 | Preferred for quality metering |
| 15360 | 8 | 1920 | Deprecated (9-2LE 60 Hz MSVCB02) |
| 96000 | 1 | 96000 | Preferred for high-bandwidth DC control |
Key insight: The preferred rates (4800 with 2 ASDU, 14400 with 6 ASDU) both produce 2400 frames/second — a consistent network rate regardless of power system frequency (50 or 60 Hz). This simplifies network sizing.
9. Maximum Processing Delay Limits (Table 901)
Processing delay time (tpd) is the difference between the time encoded by the SmpCnt field in a SV message and the time the message appears at the digital output. Per IEC 61869-9:
| Application Class | Maximum Processing Delay |
|---|---|
| Quality metering applications | 10 ms |
| Protective and measuring applications | 2 ms |
| Time-critical low-bandwidth DC control | 100 µs |
| High-bandwidth DC control | 25 µs |
The delay limit is measured at the MU output and does not include network or switching delays. Those are the system integrator’s responsibility.
💡 Why this matters: The delay adds directly to the relay’s fault detection time. Phase error itself is independent of processing delay because the encoded timestamp — not the message arrival time — is used for phasor reconstruction. So while delay doesn’t degrade accuracy, it does add to protection operating time.
The MU manufacturer declares the maximum processing delay in the LPHD.NamMaxDlRtg data object.
10. The Sampled Value Frame — What Goes on the Wire
Each SV message is a standard Ethernet frame:
| Field | Value / Range | Purpose |
|---|---|---|
| EtherType | 0x88BA | Registered for IEC 61850-9-2 Sampled Values |
| Destination MAC | 01-0C-CD-04-00-00 to 01-0C-CD-04-01-FF | Multicast range reserved for SV |
| VLAN tag (802.1Q) | Mandatory, priority 4 default | Separates SV from GOOSE traffic |
| APPID | 0x4000–0x7FFF (default 0x4000) | 16-bit application identifier |
| APDU | ASN.1 BER encoded | The payload |
APDU Contents
| Field | Description |
|---|---|
noASDU | Number of ASDUs concatenated in this frame |
svID | Stream identifier (matches MSVCB.MsvID) |
datset | Data set reference (optional, recommended FALSE per IEC 61869-9) |
smpCnt | 16-bit sample counter; resets to 0 at sync pulse |
confRev | Configuration revision counter |
smpSynch | Synchronization status (see Section 12) |
smpRate | Sample rate (optional) |
sample | The actual data values (instMag.i + quality) |
Multicast MAC Address Breakdown
01 - 0C - CD - 04 - xx - xx
└──┴──┴──┴── IEEE-assigned for IEC 61850
04 = multicast Sampled Values
(01 = GOOSE, 02 = GSSE)
xx-xx = stream-specific, assigned during engineering11. AmpSv and VolSv Scaling Factors
IEC 61869-9 fixes the SAV scaling to enable interoperability. The mandatory values are:
AmpSv (Current) — Table 904
| Attribute | Value |
|---|---|
AmpSv.units.SIUnit | 5 (code for ampere) |
AmpSv.units.multiplier | 0 |
AmpSv.sVC.offset | 0 |
AmpSv.sVC.scaleFactor | 0.001 |
AmpSv.instMag.i | Count of milliamperes |
So a value of 100000 in AmpSv.instMag.i represents 100.000 A (100 000 mA × 0.001 = 100 A).
VolSv (Voltage) — Table 906
| Attribute | Value |
|---|---|
VolSv.units.SIUnit | 29 (code for volt) |
VolSv.units.multiplier | 0 |
VolSv.sVC.offset | 0 |
VolSv.sVC.scaleFactor | 0.01 |
VolSv.instMag.i | Count of centivolts |
A value of 1000000 in VolSv.instMag.i represents 10 000.00 V (1 000 000 cV × 0.01 = 10 000 V).
The values are stored as signed 32-bit integers (INT32), giving a dynamic range of approximately ±2.147 MA peak for current and ±21.47 kV (per centivolt) peak for voltage.
12. Quality Attribute Behavior
The SAV common data class includes a quality attribute. IEC 61869-9 constrains its behavior:
| Quality Flag | When Set |
|---|---|
validity = good | Normal operation, sample meets accuracy class |
validity = questionable | Set when inaccurate flag is true, or outOfRange is true |
validity = invalid | Quantity not provided (e.g., unused channels in a legacy 4×4 data set) |
inaccurate = true | IT supervision detected error other than sync loss; sample doesn’t meet accuracy class |
failure = true | IT supervision detected error; sample is unusable (e.g., hardware fault) |
outOfRange = true | Input exceeded clipping limits; remains true until input returns within limits AND accuracy recovers |
oldData, inconsistent, operatorBlocked, oscillatory, badReference, overflow | Set to FALSE per IEC 61869-9 |
source | Always set to process |
test | Set per IEC 61850-7-3 (test mode) |
The MU shall always send its best estimate of the primary value, even when quality is questionable. Subscriber applications decide how to use questionable values.
⚠️ Important distinction: The
qualityattribute refers to the sample value quality. TheSmpSynchattribute (Section 12) refers to time synchronization quality. A sample can havegoodquality whileSmpSynch = 0(not synchronized) — useful for applications that don’t need cross-bay alignment.
13. SmpSynch Attribute Values
The SmpSynch attribute in each SV message tells subscribers about the time synchronization source:
| Value | Meaning | When to Trust for Cross-Bay Functions |
|---|---|---|
| 0 | Not synchronized (free-running, or holdover expired) | ❌ Do not use for differential protection or synchrophasors |
| 1 | Synchronized to a local area clock (unspecified ID) | ✅ Only if you know all MUs receive from the same local clock |
| 2 | Synchronized to a global area clock (GPS, NTP, NIST) | ✅ All MUs synced to global clocks are aligned |
| 3–4 | Reserved | — |
| 5–254 | Synchronized to specific local area clock with this ID | ✅ Same ID = same clock = aligned |
| 255 | Reserved | — |
Practical Implications
- Differential protection requires
SmpSynch ≥ 1AND verification that endpoints share the same clock source - Synchrophasor (PMU) applications require
SmpSynch = 2(global clock) for cross-substation alignment - Local current measurements in a single bay work with any
SmpSynchvalue (samples from the same MU are always aligned to each other) - Distance protection without remote infeed considerations can use
SmpSynch = 0if both polarity references come from the same MU
14. Time Synchronization — PTP vs 1PPS
The MU must accept an external synchronization signal. The accuracy of the time signal is expected to be better than ±1 µs for accuracy characterization.
PTP per IEC/IEEE 61850-9-3
The preferred method. PTP distributes time over the Ethernet network with sub-microsecond accuracy. The merging unit operates as a PTP slave. For details, see: PTP Power Profile Explained (IEC/IEEE 61850-9-3).
PTP clockClass | Treat as |
|---|---|
| 6 or 7 | Global area clock → SmpSynch = 2 |
| Any other | Local area clock → SmpSynch = 1 (or specific 5–254) |
1PPS Synchronization (Legacy)
For legacy systems, the MU accepts a 1 pulse per second signal:
| Parameter | Specification |
|---|---|
| Signal type | Optical on graded-index 62.5/125 µm glass fiber |
| Clock rate | One pulse per second |
| Change of second | Rising edge low → high |
Pulse duration tH | 10 µs to 500 ms |
| Rise/fall times (10–90 %) | Up to 200 ns |
| Jitter | ±2 µs maximum |
| Optical wavelength | 820–860 nm |
| Receiving power | −12 dBm max, −24 dBm min (while high) |
| Connector | BFOC/2.5 (ST) — future tech may be used |
The 1PPS signal carries no source information, so it is always interpreted as a global area clock (SmpSynch = 2).
15. Holdover Mode and Free-Running Mode
Holdover Mode
When the external synchronization signal is lost, the MU enters holdover mode:
- The MU continues sending samples
- Sample timing must remain within the measuring accuracy class
- Minimum holdover duration: 5 seconds under stable temperature
SmpSynchremains unchangedSmpCntincrements and wraps as if sync were present
The actual holdover duration is declared in LPHD.NamHoldRtg. Premium MUs offer holdover up to several hours with TCXO/OCXO oscillators.
Free-Running Mode
When the MU has never been synced, or holdover has expired:
- Samples are still sent
- Sample rate maximum deviation from nominal: ±100 × 10⁻⁶ (±100 ppm)
SmpSynch = 0SmpCntincrements and wraps as if sync were present
In free-running mode, samples from the same MU remain aligned to each other (local applications still work), but cannot be used for cross-bay functions.
Time Adjustment Behavior
When sync is restored or a time jump occurs, the MU adjusts as follows:
- The sampling jumps from old time to new time between two consecutive samples
- The sample interval over the jump: no more than 1.5× nominal, no shorter than 0.5× nominal
SmpCntis discontinuous over jumps larger than what can be accommodated by an off-nominal interval- The sample immediately after the jump has the adjusted
SmpCntandSmpSynchvalues
Subscriber applications using sampled values during a sync state change should be designed to cope with this transition.
16. Conformance Classes a/b/c/d
IEC 61869-9 defines four conformance classes that determine which IEC 61850 services the MU implements:
| Class | What It Adds |
|---|---|
| a | Minimum: just publishes Sampled Values (Multicast SVC) |
| b | Class a + GOOSE publisher/subscriber |
| c | Class b + IEC 61850 information model with self-description (logical device, logical nodes, data, data sets, server/client services, GetDataValues, etc.) |
| d | Class c + file transfer + buffered/unbuffered reporting |
What This Means in Practice
- Class a — minimum to publish SV. Cannot be configured remotely. No MMS. Used for fixed-function MUs.
- Class b — adds GOOSE for binary status. Common in simple bay-level MUs.
- Class c — adds full IEC 61850 client/server. The MU appears in System Configurator tools, supports SetDataValues, supports remote diagnosis. This is the common procurement target.
- Class d — adds file transfer (for SCL upload/download) and reporting (for event logs). High-end MUs.
The MU declares its conformance class in its ICD/PICS file. For procurement, specify class c minimum unless you have a specific reason to accept class a or b.
17. Type Tests Required by IEC 61869-9
The standard defines specific type tests every MU must pass:
| Test (per IEC 61869-9) | What It Verifies |
|---|---|
| 7.2.6 — Accuracy test | Ratio error, phase error, and composite error using full-cycle DFT |
| 7.2.901 — Digital output conformance | Full IEC 61850-10 conformance and performance tests |
| 7.2.902 — Maximum processing delay | Delay measurement over ≥1 minute; verify ≤ declared NamMaxDlRtg; verify holdover-to-free-running transition |
| 7.2.903 — Loss of synchronization | Holdover duration; SmpSynch transitions; sample rate ±100 ppm in free-running |
| 7.2.904 — 1PPS test | Jitter compliance; SmpSynch changes correctly with/without 1PPS |
For procurement, demand the full type test report from an accredited laboratory — not vendor self-declaration.
18. Dual Accuracy Class Rating
IEC 61869-9 mandates that protection-rated instrument transformers and protection-capable SAMU channels carry dual accuracy class ratings — one for measuring, one for protection.
Format
<measuring class> / <protection class> <KALF>
Where KALF is the accuracy limit factor.
Examples
| Rating | Meaning |
|---|---|
0.2S | 0.2S measuring class only (no protection rating) |
0.2S/5P20 | 0.2S measuring class AND 5P protection class with KALF=20 |
0.5/3P | 0.5 measuring class AND 3P protection class |
0.5/5P40 | 0.5 measuring AND 5P40 protection |
Dual rating is reported in:
- The IT nameplate
- The MU’s
TCTR.NamAccRtgorTVTR.NamAccRtgdata objects
This format reflects real-world use where protection CTs are also used for measurement and indication.
19. TCTR and TVTR Naming Convention
IEC 61869-9 defines a precise naming convention for TCTR and TVTR instances:
TCTR Naming
InnpTCTRn where:
| Element | Meaning | Example |
|---|---|---|
Inn | Current measurement point number (01–99) — unique within the bay | I02 |
p | Phase identification (A, B, C, N for AC; A=pole1, B=pole2, N=earth for DC) | A |
TCTR | Logical node class | TCTR |
n | Instance number (1–99) — unique within the logical device, fixed by manufacturer | 4 |
Example: I02ATCTR4 = current measurement point 02, phase A, connected to TCTR instance 4.
TVTR Naming
UnnpTVTRn with the same rules:
Example: U01ATVTR1 = voltage measurement point 01, phase A, TVTR instance 1.
Why This Matters
This naming binds the substation section (physical equipment) of the SCL to the IED section (logical implementation). A system configurator uses this binding to know which TCTR instance reads which physical CT phase.
20. 9-2LE vs IEC 61869-9
IEC 61850-9-2LE — The Implementation Guideline
9-2LE is not an IEC standard. It is a vendor implementation guideline developed by the UCA International Users Group to address the lack of specific implementation constraints in the original IEC 61850-9-2. The original 9-2 left too many degrees of freedom (data set structure, sample rate, ASDU count), causing interoperability issues.
9-2LE solved this with fixed conventions:
- Fixed data set: 8 samples per ASDU (Ia, Ib, Ic, In, Va, Vb, Vc, Vn) each with quality
- Fixed sample rates: 80 samples/period (protection), 256 samples/period (metering)
noASDU = 1— one ASDU per frame- Mandatory SmpSynch
9-2LE became the de facto standard from the 2000s through early 2010s. Most legacy MUs implement 9-2LE.
IEC 61869-9 — The Formal Standard (2016)
IEC 61869-9 is a proper IEC standard, part of the IEC 61869 series on instrument transformers. It replaces 9-2LE with a formally standardized profile. Key differences:
| Aspect | 9-2LE | IEC 61869-9 |
|---|---|---|
| Status | UCA guideline | Formal IEC standard |
| Data set | Fixed 4I + 4U | Configurable, with variant codes |
| Sample rates | 80/256 per period | Multiple preferred rates (4800, 14400, 96000 sps) |
| ASDUs per frame | 1 | Configurable (1, 2, 6, 8) |
| Time sync | Implementation-specific | IEC/IEEE 61850-9-3 PTP mandatory option |
| Dual accuracy class | — | Mandatory for protection MUs |
| Nameplate (LPHD) | Basic | Extended with NamVariant, NamHoldRtg, NamMaxDlRtg, etc. |
| Backward compatibility | — | Yes — supports F4000S1I4U4, F4800S1I4U4 variants for legacy |
Interoperability Considerations
A 9-2LE-only MU and an IEC 61869-9 IED may interoperate if the IEC 61869-9 device supports the legacy variant (F4000S1I4U4 or F4800S1I4U4). The data set structure is compatible for these variants. For other variants (F4800S2I4U4, F14400S6I4U4), both ends must support IEC 61869-9.
Procurement Recommendation
- New projects: Specify IEC 61869-9 explicitly.
- Brownfield upgrades: Verify firmware support for both legacy and new variants.
- Mixed-vendor systems: Confirm each device’s PICS file declares the required variant codes.
21. Network Bandwidth Sizing
SV streams generate continuous, high-rate traffic. Bandwidth planning is a required engineering step.
Per-Stream Bandwidth (Typical)
| Variant | Frames/s | Frame Size | Bandwidth |
|---|---|---|---|
| F4000S1I4U4 (9-2LE 50 Hz) | 4 000 | ~150 B | ~5 Mbit/s |
| F4800S1I4U4 (9-2LE 60 Hz) | 4 800 | ~150 B | ~6 Mbit/s |
| F4800S2I4U4 (preferred protection) | 2 400 | ~300 B | ~6 Mbit/s |
| F14400S6I4U4 (preferred metering) | 2 400 | ~900 B | ~17 Mbit/s |
| F96000S1I4U4 (high-bw DC) | 96 000 | ~150 B | ~120 Mbit/s |
Maximum Quantities per Stream (100 Mbit/s networks)
Per IEC 61869-9, on 100 Mbit/s networks:
| Application | Max Quantities (I + U) |
|---|---|
| General measuring and protection | 24 |
| Quality metering | 8 |
| DC control applications | 24 |
| Minimum (any) | 1 |
These limits prevent excessively long Ethernet frames from blocking the network. No specific limits apply to 1 Gbit/s and faster networks. DC instrument transformer outputs may require point-to-point connections on Gigabit links.
Why Gigabit Ethernet Is Often Preferred
100 Mbit/s handles SV traffic in small substations, but margin disappears as bay count grows. A 100 Mbit/s segment shared between 10+ SV streams, GOOSE, and management traffic approaches its practical limit. Gigabit Ethernet is increasingly the baseline for process bus design — not because individual streams demand it, but because aggregate traffic and future expansion headroom do.
22. PRP and HSR Redundancy on the Process Bus
Protection-grade process bus requires zero-recovery-time redundancy. Two approaches:
PRP (Parallel Redundancy Protocol)
The MU has two Ethernet ports — one to LAN A, one to LAN B. Every SV frame is sent simultaneously on both LANs. The subscriber receives both, processes the first, discards the duplicate. Zero recovery time on any single LAN failure.
HSR (High-availability Seamless Redundancy)
The MU is part of a ring. Every SV frame is sent in both ring directions. The receiver gets two copies; the second is discarded.
Bandwidth Impact
In both PRP and HSR, every SV frame is duplicated. The effective bandwidth doubles on each network path. For a 10-bay substation at the preferred F4800S2I4U4 rate on PRP:
- Per LAN: ~60 Mbit/s SV traffic
- Plus GOOSE, MMS, PTP — easily approaching the 100 Mbit/s limit
- Gigabit Ethernet strongly recommended
For details:
23. Procurement Specification Template
Use this checklist when writing a tender specification for a merging unit:
Required Certifications
- IEC 61869-9:2016 compliance, verified by third-party type test report
- IEC 61869-6 (general LPIT requirements)
- Relevant product-specific standard: IEC 61869-7 (EVT), -8 (ECT), -12 (combined), -13 (SAMU), -14 (DC current), -15 (DC voltage)
- IEC 61850-3 environmental and EMC requirements
- IEC/IEEE 61850-9-3 PTP Power Profile support
Required Variant Code Support
State explicitly which variants must be supported, for example:
F4800S2I4U4— preferred for protection/measurementF14400S6I4U4— preferred for quality meteringF4000S1I4U4— 9-2LE backward compatibility (50 Hz systems)F4800S1I4U4— 9-2LE backward compatibility (60 Hz systems)
Performance Requirements
- Maximum processing delay ≤ 2 ms for protection (≤ 10 ms acceptable for metering only)
- Holdover duration ≥ 5 s minimum; specify higher (e.g., 10 s, 1 min, 1 hour) per application
- Free-running rate stability ≤ ±100 ppm
- Sample timing accuracy ≤ ±1 µs from absolute time (when synchronized)
- Dual accuracy class rating (e.g., 0.2S/5P20)
Synchronization Requirements
- PTP per IEC/IEEE 61850-9-3 mandatory
- PRP/HSR doubly-attached PTP support if redundancy is required
- 1PPS optical input (mandatory or optional per project)
Conformance Class
- Minimum Conformance Class c (logical device, data, data sets, server/client) — recommended default
- Class d only if file transfer and buffered reporting are needed
Network Requirements
- Two Ethernet ports with duplex LC connectors (preferred)
- 100BASE-FX minimum; 1000BASE-LX recommended for PRP/HSR or large substations
- PRP and/or HSR support per IEC 62439-3
Documentation Required
- Type test report from accredited third-party laboratory
- PICS, PIXIT, MICS (per IEC 61850-10)
- ICD file with full SCL declaring all supported variants
- Maximum number of supported MSVCB instances
- Maximum number of concurrent subscribers
- Detailed processing delay specification (best case, worst case)
Hardware Requirements
- Operating temperature range (e.g., −40 °C to +85 °C for class 2 per IEC 61850-3)
- Dual redundant power supplies (e.g., 80–300 V DC + 100–250 V AC)
- Auto-detection of fuse failure (TVTR.FuFail)
- Visible LED indications: power on, in service, alarm, communication status, test mode
- Test signal generation capability (disabled by default)
24. Field Troubleshooting: Common Failure Patterns
When working on site, these are the patterns to check first.
🔴 Pattern 1: No SV stream received by IED
Likely causes:
SvEna = FALSEin the MSVCB- Cable problem between MU and switch
- Wrong VLAN tag (IED expects VLAN N, MU sends VLAN M)
- VLAN priority mismatch breaking QoS filtering
Action: Verify SvEna with system configurator tool. Capture with Wireshark on the IED-side switch port. Check VLAN configuration on both ends.
🔴 Pattern 2: SmpSynch = 0 in steady state
Likely causes:
- PTP master unreachable (network path broken, master failed)
- PTP master clockClass ≠ 6 or 7 (holdover or worse)
- Wrong PTP domain configured
- PTP profile mismatch (using default IEEE 1588 profile instead of IEC/IEEE 61850-9-3)
- 1PPS cable disconnected or laser failed
Action: Check PTP master diagnostics. Verify domain (default 0; some substations use 93 per IEC/IEEE 61850-9-3 recommendation). Capture PTP traffic with Wireshark.
🔴 Pattern 3: SmpSynch toggles between 0 and 2
Likely causes:
- PTP master entering/exiting holdover
- Network path with marginal link quality
- Misconfigured PRP/HSR redundancy losing one path
Action: Check master clock source (GPS antenna). Monitor master’s clockClass over time. Verify PRP RedBox/DAN ports are both up.
🔴 Pattern 4: confRev mismatch on subscribers
Likely causes:
- MU configuration was changed after IEDs were commissioned
- Data set modification (added/removed/reordered members)
- SCD file regenerated without updating IED firmware
Action: Compare confRev in the MU’s MSVCB against confRev configured in each subscriber. Re-engineer all subscribers if MU configuration changed.
🔴 Pattern 5: Wrong values on subscriber
Likely causes:
- Scaling factor wrong on subscriber (expecting different units)
- Wrong CT/VT ratio configured on subscriber
- Wrong phase assignment (subscriber reading TCTR2 thinking it’s phase A)
- Endian/format mismatch (unlikely with proper ASN.1 BER decoding)
- Sample data set order differs from subscriber expectation
Action: Verify scaling: AmpSv.instMag.i is in milliamperes (×0.001 for amps), VolSv.instMag.i is in centivolts (×0.01 for volts). Verify TCTR/TVTR instance-to-phase mapping in the substation section of the SCL.
🔴 Pattern 6: APPID conflict
Likely causes:
- Two MUs configured with the same APPID
- Default APPID
0x4000left in place (indicates lack of configuration)
Action: Each SV stream needs a unique APPID in the range 0x4000–0x7FFF. Use system configurator tool to enforce uniqueness. Wireshark capture shows duplicate APPIDs immediately.
🔴 Pattern 7: SV traffic causes network problems
Likely causes:
- Network bandwidth saturation (too many high-rate streams on 100 Mbit/s)
- Switch CPU overload from multicast snooping
- Improper VLAN segmentation (SV in same broadcast domain as MMS)
- PRP/HSR doubling traffic without bandwidth headroom
Action: Calculate total SV bandwidth (Section 20). Upgrade to Gigabit Ethernet if approaching 50% of link capacity. Implement IGMP snooping or static multicast filters. Separate SV VLAN from station bus.
🔴 Pattern 8: Differential protection mis-operates after restart
Likely causes:
- One MU is in free-running mode (SmpSynch=0)
- Two MUs synced to different local clocks (SmpSynch=1 but different IDs)
- Time jump > sample interval during PTP re-lock causing SmpCnt discontinuity
Action: Verify both MU endpoints have SmpSynch=2 (or matching IDs in 5–254 range). Check PTP master coverage. Configure protection to use blocking logic during sync state changes.
🔴 Pattern 9: outOfRange flag stays true
Likely causes:
- Persistent fault current/voltage exceeding clipping limit
- Hardware ADC saturation issue
- Primary CT/VT saturation
Action: Compare measured value against NamClipRtg × NamARtg × √2. If genuine overload, this is expected behavior. If sustained without fault, suspect hardware.
🔴 Pattern 10: Cannot ping MU’s IP
This is expected and normal. The MU’s primary output (SV stream) is on Ethertype 0x88BA — a Layer 2 protocol with no IP. If the MU also has MMS support (Conformance Class c or d), it should respond to ping on its station bus address.
Action: Use the MU’s MMS-side IP address, not the process bus address. Verify the MU supports IP on the station bus side per its conformance class declaration.
Summary
A merging unit is the foundation of the IEC 61850 process bus. It digitizes CT/VT signals and publishes them as Sampled Values that any number of subscribers can consume — replacing analog wiring with a deterministic, time-aligned digital interface.
The key things to remember:
- The current standard is IEC 61869-9:2016, which formally replaces the older 9-2LE guideline
- SV streams use Layer 2 Ethernet on EtherType 0x88BA, multicast MAC
01-0C-CD-04-xx-xx, VLAN priority 4, APPID range0x4000–0x7FFF - Variant codes use the FfSsIiUu notation — preferred rates are
F4800S2I4U4for protection/measurement andF14400S6I4U4for metering - Maximum processing delay is application-dependent: 2 ms for protection, 10 ms for metering, 100 µs for low-bw DC control, 25 µs for high-bw DC control
- AmpSv is scaled in milliamperes (scaleFactor 0.001), VolSv is scaled in centivolts (scaleFactor 0.01)
- SmpSynch values: 0=free-running, 1=local area clock, 2=global area clock, 5–254=specific local clock ID
- Preferred time sync is PTP per IEC/IEEE 61850-9-3; 1PPS is the legacy alternative
- Holdover minimum: 5 s; free-running rate stability: ±100 ppm
- Conformance class c minimum for procurement (full IEC 61850 server/client)
- For procurement, specify IEC 61869-9 explicitly with required variant codes — not just “9-2LE” or “IEC 61850”
- Always demand a third-party type test report, not vendor self-declaration