Inferensys

Glossary

Phasor Data Concentrator (PDC)

A Phasor Data Concentrator (PDC) is a node that aggregates and time-aligns streaming synchrophasor data from multiple PMUs for local archiving or forwarding to higher-level systems.
Data scientist building training data pipeline on laptop, data preprocessing visible, technical workspace.
SYNCHROPHASOR DATA AGGREGATION

What is a Phasor Data Concentrator (PDC)?

A Phasor Data Concentrator (PDC) is a hardware or software node that aggregates, time-aligns, and processes streaming synchrophasor data from multiple Phasor Measurement Units (PMUs) to create a coherent, system-wide data stream for wide-area monitoring and control applications.

A Phasor Data Concentrator (PDC) functions as a critical middleware layer in the synchrophasor data hierarchy. It ingests asynchronous, time-stamped data streams from multiple downstream Phasor Measurement Units (PMUs) or lower-tier PDCs, buffers the incoming frames, and correlates them using their precise GPS-derived timestamps. The PDC outputs a single, time-synchronized, aggregated data stream, effectively compensating for network latency and packet jitter to provide a consistent, simultaneous snapshot of grid conditions across a wide geographic area.

Beyond basic aggregation, a PDC performs essential data quality checks, discarding corrupted or late-arriving frames based on configurable wait-time limits. It supports multiple communication protocols, including IEEE C37.118.2, for both input and output streams, enabling seamless integration into a Wide-Area Monitoring System (WAMS). By providing a unified, high-resolution data feed, the PDC serves as the foundational data backbone for real-time small-signal stability analysis, oscillation detection, and archival storage for post-event forensic analysis.

PHASOR DATA CONCENTRATOR

Core Functional Characteristics

A Phasor Data Concentrator (PDC) is the middleware node that aggregates, time-aligns, and processes streaming synchrophasor data from multiple PMUs, functioning as the critical data integration layer between substation measurement and wide-area monitoring applications.

01

Time Alignment & Latency Normalization

The PDC's primary function is to buffer incoming asynchronous synchrophasor streams and align them to a common UTC time tag using the SOC (Second of Century) timestamp. It compensates for variable network latency—fiber, microwave, or MPLS—by implementing a configurable wait-time window (typically 50-200 ms). Frames arriving after the window expire are discarded and flagged as late data. The output is a coherent, time-correlated dataset where all phasors correspond to the exact same reporting instant, enabling valid cross-comparison of phase angles across hundreds of miles.

< 1 µs
Alignment Accuracy
50-200 ms
Typical Wait Window
02

Multi-Protocol Ingestion & Output

A PDC acts as a protocol translation gateway. It ingests data via IEEE C37.118.2 (legacy synchrophasor protocol) and IEC 61850-90-5 (routable GOOSE/SV for synchrophasors), then outputs concentrated streams to higher-level systems. Output protocols include:

  • IEEE C37.118.2 for upstream PDCs or WAMS
  • IEC 61850-90-5 for substation LAN integration
  • STTP (Streaming Telemetry Transport Protocol) for low-latency, loss-tolerant streaming to cloud analytics This multi-protocol capability allows a single PDC to bridge legacy PMU fleets with modern digital substation architectures.
03

Data Validation & Quality Flagging

The PDC performs real-time quality assurance on every incoming frame before forwarding. It inspects the STAT word (a 16-bit flag field per IEEE C37.118) for:

  • Data valid/invalid indicators
  • PMU sync status (locked to GPS vs. holdover)
  • Trigger reason (manual, disturbance, forced)
  • Time quality and leap-second flags Frames with invalid data or unsynchronized timestamps are either dropped or forwarded with a degraded quality flag, preventing corrupted measurements from triggering false alarms in downstream oscillation detection or state estimation applications.
04

Downsampling & Rate Conversion

PMUs often report at high rates (50/60 frames per second for protection-class, 10/25 fps for measurement-class), but control room applications may only require 30 fps or 1 fps data. The PDC performs intelligent downsampling by selecting the most recent valid frame within each output interval, rather than simple decimation. This preserves the freshest data while reducing bandwidth to upstream systems. Some PDCs also support anti-aliasing filtering during rate conversion to prevent the introduction of spurious low-frequency artifacts in downsampled oscillation data.

05

Local Archiving & Disturbance Recording

Beyond real-time forwarding, the PDC functions as a local historian for high-resolution grid data. It buffers continuous synchrophasor streams to solid-state storage, typically maintaining:

  • Circular buffers for continuous recording (days to weeks)
  • Triggered event files captured around disturbances (pre-trigger + post-trigger)
  • COMTRADE file generation for relay event analysis This local storage ensures that forensic data survives network outages and provides the raw material for Prony analysis, ringdown analysis, and post-mortem disturbance investigations without requiring continuous WAN bandwidth.
06

Cascading PDC Hierarchies

PDCs are designed to operate in a hierarchical tree topology for scalability across large interconnections:

  • Substation PDC: Aggregates 4-8 local PMUs, performs first-level concentration
  • Regional PDC: Aggregates outputs from multiple substation PDCs, covering a utility control area
  • Super PDC / Central PDC: Aggregates regional PDC streams for ISO/RTO-level wide-area visibility Each level re-applies time alignment with progressively larger wait windows to compensate for cumulative latency. This architecture allows a single Wide-Area Monitoring System (WAMS) to ingest coherent data from thousands of PMUs across an entire interconnection without overwhelming any single node.
PHASOR DATA CONCENTRATOR ESSENTIALS

Frequently Asked Questions

Clear, technically precise answers to the most common questions about Phasor Data Concentrator architecture, functionality, and deployment in wide-area monitoring systems.

A Phasor Data Concentrator (PDC) is a node in a synchrophasor network that aggregates streaming data from multiple Phasor Measurement Units (PMUs) and other downstream PDCs, time-aligns the measurements using their GPS-synchronized timestamps, and outputs a single, coherent, time-synchronized data stream. The PDC buffers incoming frames, waits for all expected data with the same timestamp to arrive within a configurable maximum wait time, performs interpolation if necessary for missing or late packets, and then assembles a consolidated output frame. This process resolves the inherent communication latency differences between PMUs located at different substations, ensuring that a Wide-Area Monitoring System (WAMS) receives a unified snapshot of grid state at each reporting interval, typically 30, 60, or 120 frames per second.

HIERARCHY COMPARISON

PDC vs. Super PDC vs. PMU

Functional differentiation between the measurement unit, the local data aggregator, and the regional concentrator in a synchrophasor architecture.

FeaturePMUPDCSuper PDC

Primary Function

Measures and timestamps electrical phasors

Aggregates and time-aligns multiple PMU streams

Aggregates and correlates multiple PDC outputs

Hierarchical Level

Substation bay / field level

Substation or local control center

Regional transmission operator / reliability coordinator

Input Data Source

CT/PT secondary waveforms

IEEE C37.118 streams from PMUs

IEEE C37.118 streams from subordinate PDCs

Time Alignment

GPS-synchronized sampling at source

Re-aligns streams by GPS timestamps into a coherent dataset

Synchronizes datasets across wide geographic areas

Output Data Rate

10-120 frames per second

10-60 frames per second (aggregated)

10-30 frames per second (system-wide)

Latency Budget

< 50 microseconds (measurement)

< 100 milliseconds (concentration)

< 500 milliseconds (regional aggregation)

Data Archiving

Oscillation Detection

Local mode monitoring

Inter-area mode monitoring

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.