Perfered ways to configure GOOSE messages - ABB SSC600

Configuring a GOOSE publisher

Creating a GOOSE data set with the IEC 61850 Configuration tool:
The sending data set is defined with the GOOSE control block. With the IEDs of this product series, the sending GOOSE data set can have a maximum of 20 data attributes to minimize the message-handling load in the receiving and sending IEDs.
All data sets must be configured under the logical node LLN0 and must be provided with names unique within the IED. The IEDs allow a maximum of four GOOSE control blocks, which effectively limits the IED to four data sets for GOOSE, as there is a one-to-one correspondence between the GOOSE control blocks and GOOSE data sets. Typically, it is sufficient to define a single data set and control block for an application. However, it is recommended to use a separate data set and corresponding control block for analog values.

A maximum of 80 data attributes can be added to IED's GOOSE data sets. Recommendation is to divide attribute amount to 20 per GOOSE data set, for maximum performance in sender/receiver. After creating the GOOSE data sets, define the data set entries (data attributes or data objects) for the data sets. If quality data attributes are added to a data set, they must be located after the status value of the corresponding data object.

After discussing with experts, particularly from DNV, we have updated the information regarding whether to use one large dataset or several small datasets for sending GOOSE messages.

• Sending GOOSE Messages with One Large Dataset:
  Using a large dataset allows users to send all the data points in a single GOOSE message. This approach simplifies and speeds up the configuration process of the GOOSE message, as all data points are consolidated into one transmission. This method can be particularly advantageous in systems where ease of configuration and a unified transmission are prioritized.
  However, there are trade-offs. The receiver needs to process all the data points contained in the large dataset, which can be more time-consuming. The receiver must parse through the entire dataset, potentially leading to slower data handling and execution times. This delay is primarily due to the additional time required to decode and process a more extensive dataset. Despite these concerns, it is important to note that transmitting GOOSE messages still occurs within microseconds, which is generally sufficient for many applications. Thus, using one large dataset is suitable in scenarios where timing is not critically sensitive and where configuration simplicity is a priority.

• Sending GOOSE Messages with Multiple Small Datasets:
  On the other hand, using multiple small datasets provides a clear and organized overview of the system's configuration. This method makes it easier for engineers to review and understand the setup, as each dataset is smaller and more manageable. Smaller datasets can be configured and maintained with greater precision, reducing the likelihood of errors during configuration.
  For the receiver, processing multiple smaller datasets can be more efficient. Each dataset contains fewer data points, which means the receiver can quickly decode and execute the necessary information without the need to parse through an extensive dataset. This can lead to faster response times and improved system performance, especially in time-critical applications where rapid data handling is essential.
  Additionally, having multiple smaller datasets allows for more granular control and flexibility in system configuration. Engineers can update or modify specific datasets without affecting the entire system, facilitating easier maintenance and updates. This modular approach can also enhance system resilience, as issues in one dataset do not necessarily impact others.

In conclusion, the choice between one large dataset and multiple small datasets for sending GOOSE messages depends on the specific needs and constraints of the system. One large dataset simplifies configuration but may slow down processing, whereas multiple small datasets offer clearer configuration, faster processing, and greater flexibility. The decision should consider factors such as the criticality of timing, ease of configuration, and the overall system architecture.