The Distributed Network Protocol (DNP3) is a widely used SCADA communication protocol designed to enable reliable data exchange between master stations and outstations. It is optimized for transmitting small, structured data packets efficiently and reliably in industrial environments
A key strength of DNP3 lies in its object-based data model, which organizes data into object groups, variations, and indices. This structured approach enables flexible, efficient, and interoperable communication across devices from different vendors.
Table of Contents
1. DNP3 Object-Based Data Model
DNP3 uses an object library at the application layer to represent all data exchanged between devices.
According to the DNP3 reference manual:
- Data is organized into object groups (types of data)
- Each object is further defined by variations (format of data)
- Individual data points are identified by indexes
This model allows:
- Standardized communication
- Efficient data transmission
- Support for multiple data types in a single message
2. DNP3 Data Objects
DNP3 data objects represent real-world measurements and control points stored in device databases.
Typical data types include:
- Binary inputs (e.g., breaker status)
- Analog inputs (e.g., voltage, temperature)
- Counters (e.g., energy usage)
- Control outputs
As explained in the DNP3 primer:
- These data points are stored as arrays in outstation databases
- Each element is identified by a zero-based index
This structure allows efficient access and transmission of large numbers of data points.
3. Object Groups (Data Categories)
DNP3 organizes objects into groups, where each group represents a category of data.
Examples:
- Binary Inputs
- Analog Inputs
- Counters
- Binary Outputs
- Analog Outputs
Groups define what type of data is being communicated, not how it is formatted.
The DNP3 manual describes this as:
Data objects are “broadly grouped together into Object Groups such as Binary Input Objects and Analog Input Objects”
4. Variations (Data Representation)
Each object group contains multiple variations, which define how the data is encoded.
Variations specify:
- Data format (integer, floating point)
- Size (16-bit, 32-bit, etc.)
- Presence of additional information (e.g., flags, timestamps)
For example, analog data may be transmitted as:
- Integer values
- Floating-point values
- With or without status flags
This flexibility allows DNP3 to:
- Optimize bandwidth
- Match device capabilities
- Provide detailed or compact data formats
5. Object Headers and Message Structure
DNP3 transmits data objects using object headers, which are part of the application layer message.
Each object header contains:
- Group number
- Variation number
- Qualifier field
- Range or count of objects
The diagram clearly shows:
- Object headers precede the actual data
- Multiple objects can be included in a single message
This structure allows:
- Efficient bulk data transfer
- Flexible addressing of data points
6. Indexing of Data Points
Each data object is identified by an index (point number).
- Indexes are zero-based
- Represent specific field devices or measurements
- Used to access data within object groups
From the primer:
“Element numbers are called point indexes… the lowest element is always identified as zero”
This enables scalable systems with thousands of data points.
7. Static and Event Objects
DNP3 defines two main types of data:
Static Objects
- Represent current values
- Retrieved using polling
- Belong to Class 0
Event Objects
- Represent changes in data
- Stored in event buffers
- Assigned to Class 1, 2, or 3 (priority levels)
From the DNP3 manual:
- Event data is returned when the master polls a specific class
- Only changed data is transmitted, improving efficiency
8. Class-Based Data Organization
DNP3 assigns data objects to classes:
- Class 0 → Static data
- Class 1 → High-priority events
- Class 2 → Medium-priority events
- Class 3 → Low-priority events
This allows:
- Priority-based polling
- Efficient bandwidth usage
- Faster reporting of critical events
9. Efficient Data Transmission
The object model supports efficient communication by:
- Sending only changed (event) data
- Allowing multiple object types in one message
- Supporting unsolicited responses
DNP3 systems can operate in:
- Polling mode (master requests data)
- Quiescent mode (report-by-exception)
In quiescent mode:
- Outstations send data only when changes occur
- Communication remains idle otherwise
10. Practical Example
In a substation SCADA system:
- Binary Input → Breaker ON/OFF
- Analog Input → Voltage level
- Counter → Energy consumption
Each data point:
- Belongs to a group
- Has a variation (format)
- Is identified by an index
- May be assigned a class
The master station retrieves:
- All static data (Class 0)
- Only changed event data (Class 1–3)
Conclusion
The DNP3 data model—based on objects, groups, variations, and indices—provides a structured and flexible framework for representing industrial data. This model enables:
- Efficient communication
- Scalable system design
- Interoperability across vendors
- Reliable data handling in SCADA systems
By separating data type (groups) from data format (variations), DNP3 achieves both flexibility and efficiency, making it one of the most effective protocols for industrial automation.