Sankey and Chord Flow Charts

Sankey and Chord charts are two complementary visualization approaches for making complex system relationships visible. Sankey emphasizes the quantity and magnitude of flow transferred from a source to a destination, while Chord reveals the intensity of relationships and interactions between components in a circular structure.

SankeyFlow and quantity
ChordRelationship and interaction
SCADAEvent and system analysis
HPPEnergy and maintenance insight
Key distinction in EOS LIVE DASH: The question β€œHow much flowed or was lost?” is addressed more directly by Sankey, while β€œWho is related to whom, and how strongly?” is addressed more directly by Chord.

Sankey Charts

Energy Flow Sankey – Illustrative HPP Model
Reservoir Potential energy Turbine Hydraulic losses Generator / Grid Losses Energy flow Electrical energy

Sankey charts show the source β†’ transformation β†’ destination path of flows such as energy, mass, cost, or data together with the magnitude of the flow. The most distinctive characteristic is that the thickness of the flow line is proportional to the quantity being transferred.

Basic Components

  • Nodes: sources, intermediate transformation points, and final outputs/consumption points.
  • Links: directional connections; thickness can represent quantity, while color can represent flow type.
  • Flow direction: can be arranged from left to right or from the center outward.

Main Types

  • Energy Flow Sankey: shows energy transformations.
  • Loss / Efficiency Sankey: emphasizes where losses occur and their impact within the transformation chain.
  • Process Sankey: shows flows between industrial process stages.
  • Alarm / Event Sankey: presents relationships such as alarm β†’ cause β†’ result using a flow-based structure.

Sankey – SCADA and Energy System Applications

Power Generation Facilities

  • Fuel energy β†’ mechanical energy β†’ electrical energy.
  • Separation of net generation, internal consumption, and losses.

Thermal and Gas Power Plants

  • Fuel β†’ boiler β†’ turbine β†’ condenser β†’ losses.
  • Stage-by-stage examination of the efficiency chain.

Distribution and Grid

  • Generation β†’ transformer β†’ line β†’ consumption.
  • Separation of technical and non-technical losses.

Hydroelectric Power Plant – Water β†’ Energy Conversion

Objective: Make visible the stages at which hydraulic energy is lost during its conversion into electrical energy.

Flow: Reservoir potential energy β†’ hydraulic losses β†’ turbine mechanical energy β†’ mechanical losses β†’ generator electrical energy β†’ electrical losses β†’ energy delivered to the grid.

Technical Insight: If the thickest loss branch is assumed to represent penstock friction loss, it may help prioritize hydraulic design, surface condition, or cleaning activities. This is an illustrative analytical scenario, not an actual field result.

Advanced Sankey Scenarios for HPPs

Alarm Flow Sankey

Total alarms β†’ operator intervention β†’ alarm reset β†’ downtime β†’ maintenance β†’ recurring alarm

This view analyzes not only the number of alarms, but also what outcomes they lead to and whether they recur after intervention.

Example Insight: If a significant proportion of alarms recur after intervention, this may indicate that the root cause has not been permanently eliminated. The 40% value given in the source is illustrative.

Failure β†’ Downtime β†’ Energy Loss

Failure types β†’ downtime duration β†’ energy loss (MWh)

This makes it possible to evaluate failure frequency together with generation loss. Gate failures that occur infrequently but cause long downtime can become critical through this approach.

Maintenance Type β†’ Impact

Mechanical maintenance β†’ vibration reduction β†’ efficiency increase
Hydraulic maintenance β†’ oil alarms β†’ gate response time

This helps relate maintenance activities not only to their duration, but also to the operational indicators they affect.

Flow Rate β†’ Operating Mode β†’ Alarm

Low / medium / high flow β†’ normal / load following / operation under alarm conditions β†’ alarm types

An increase in flow thickness associated with vibration or cavitation alarms under high-flow conditions can provide a technical indication for reviewing operating limits.

Unit β†’ Problem β†’ Maintenance

Unit-1 / Unit-2 / Unit-3 β†’ mechanical / hydraulic / electrical problems β†’ corresponding maintenance type

This makes it possible to identify characteristic problem clusters by unit and see how they translate into maintenance activities.

Time β†’ Event Flow

Morning startup β†’ alarm β†’ intervention β†’ downtime β†’ maintenance β†’ restart

Training Value: Showing how events transform into one another within a single flow provides a powerful visual approach for operator training and event investigations.

Chord Charts

Alarm and Subsystem Interactions – Illustrative Chord
Relationship intensity Hydraulic Mechanical Electrical Instrumentation

Chord charts show mutual relationships and interactions between components within a circular structure. Unlike Sankey, the primary focus is not one-way flow quantity, but relationship intensity and many-to-many connections.

Basic Components

  • Nodes: system components, equipment, alarm types, sections, and subsystems.
  • Chords: represent connections between two nodes; thickness can indicate frequency, quantity, or interaction intensity.
  • Circular scale: the arc length of a node can be used as a visual indicator of the total interaction volume.

Chord Types

  • Symmetric Chord: shows an A ↔ B relationship bidirectionally based on total interaction.
  • Directed Chord: highlights the source β†’ target direction using color or gradients and can be used for cause-and-effect analysis.
  • Group-Based Chord: groups subsystems or clusters to reduce complexity.

Chord – SCADA and Energy System Applications

General Application Areas

  • Power generation facilities: alarm triggering relationships between subsystems; interactions between turbines, boilers, and auxiliary systems.
  • Switchyards and transmission: switching and protection relationships between buses, lines, and transformers.
  • Thermal and gas power plants: failure types ↔ affected equipment; emission violations ↔ process stages.

Powerful Analyses for Hydroelectric Power Plants

  • Interactions between alarm types: hydraulic ↔ mechanical ↔ electrical ↔ instrumentation ↔ safety alarms.
  • Alarm propagation between subsystems: turbine, generator, hydraulic system, gates, and switchyard/transformer.
  • Unit Γ— problem type: Unit-1/2/3 and mechanical, hydraulic, electrical, and instrumentation problems.
  • Alarm β†’ intervention β†’ result: cross-relationships between alarm types, operator intervention, reset, downtime, and maintenance.
Example Interpretation: A prominently thick hydraulic ↔ mechanical chord may suggest that hydraulic-originated problems could be strongly associated with mechanical alarm generation; root cause analysis can then be examined more deeply on the hydraulic side.

Chord – Maintenance and Operational Relationships

Maintenance Type Γ— Reduced Alarm Type

Mechanical maintenance β†’ vibration alarms
Hydraulic maintenance β†’ oil temperature alarms
Electrical maintenance β†’ related electrical alarm groups

This analysis helps visually evaluate the relationship between maintenance activities and reductions in alarm behavior. Thus, it is possible to examine not only whether maintenance was β€œperformed,” but also which problem it affected and to what extent.

Flow Range Γ— Problem Type

Low flow / medium flow / high flow ↔ cavitation / vibration / overheating / alarm-downtime

Technical Interpretation: A prominent high-flow ↔ cavitation chord may indicate a relationship between operating conditions and cavitation events that requires further investigation.

Unit Γ— Problem Type

A dominant Unit-2 ↔ hydraulic chord may indicate that the characteristic problem of the corresponding unit is concentrated around the hydraulic system.

Analytical Warning: A thick chord is not direct evidence of causality. It should be evaluated together with SCADA timestamps, event records, maintenance records, and process conditions.

Comparison of Sankey and Chord

Feature Sankey Chord EOS SCADA Example
Primary focus Flow quantity Relationship / interaction Energy flow / alarm relationship
Structure Source β†’ destination Circular nodes Energy chain / subsystem network
Thickness Transferred quantity Relationship intensity MWh / alarm / frequency
Strongest question How much went where? Who is related to whom? Loss / root cause
HPP example Water β†’ electricity β†’ loss Alarm ↔ subsystem Efficiency / maintenance / alarm analysis
Combined Use: Sankey can be used to examine energy, loss, or event flows within a system, while Chord can be used to examine the intensity of relationships between components in the same system. Together, they enable complex SCADA data to be evaluated from two different analytical perspectives.
← Back to Previous Content