DSO and the Distribution System Platform (DSP)
Distribution grids are evolving, so should distribution utilities. The evolvement from Distribution Network Operator (DNO) to Distribution System Operator (DSO) needs changes in all aspects of the operation. A DSO should operate a system, which includes planning, the grid, generation, load and market, hence the Distribution System Platform.
The DSO role has evolved out of two kinds of concerns:
- the problem of managing high penetration DER and “prosumers” for both bulk system and distribution operations while maintaining distribution reliability
- the emerging problem of bypassing, as the middle layers in grid systems might be skipped and for example a few megawatt scale generation facilities get direct connection to upper layers.
The first led to new models for the roles and responsibilities of distribution operators and system operators, while the second led to new views about multi-scale grid coordination and control.
The evolvement from Distribution Network Operator (DNO) to Distribution System Operator (DSO) needs changes in all aspects of the operation.
Distribution System Platform: What are the goals and How we should realize them (Courtesy of DOE-OE, Modern Distribution Grid Report)
The grid management system of the DSP
Grid management solutions have traditionally been centralized, communicating to remote subsystems and equipment as needed. Due to recent various trends emergence, like DER, new loading schemes, and emergence of many different players like aggregators, microgrids, etc., the need for changes in control system structure has become apparent.
Meantime, the ability of the grid to resist faulty situations and to recover quickly in the face of extreme events (grid resilience), whether they are caused by nature of human, gained a great deal of importance.
Centralized or decentralized?
There is a debate about weather the new control system architecture as the Central Nervous System (CNS) of the Distribution System Platform (DSP) goes central or distributed or a combination of both.
In central control scheme, massive amounts of data must flow to a central location; substantial amounts of computation must be done to determine everything from voltage regulator set-points to DER dispatches and then the commands must be distributed back to the grid and edge devices.
In the more distributed approaches, data flows are more local and decentralized computing, operating under a coordination framework solves the control problems in a distributed fashion, with each part acting as a team member rather than as a slave to a central master.
The choice between these approaches is structural (architectural) and has a massive impact on many other downstream design decisions.
The centralized control structure has the advantage of familiarity and well known performance, but was not intended to deal with vast numbers of edge distributed devices and systems. This form was developed in the days before significant DER penetration began to take hold and was appropriate at the time, but it suffers from limitations in its ability to scale up both communications and computation to levels needed to handle large numbers of intelligent interacting edge devices.
The choice between centralized or decentralized approaches is structural (architectural) and has a massive impact on many other downstream design decisions.
Market optimized or Operation optimized?
At the other hand there are two approaches for grid optimizations: market vs. control mechanisms: two extreme views about grid and DER management: one says that a well-designed market with the “right rules” will provide prices that make everything work. The other says that a proper optimal control formulation will make everything work. The key issue is to understand where each mechanism fits and how they work together.
In either case the issue of interoperability standards and the processes, systems, and devices that go with them must be integral to a grid modernization strategy.
In order to cover both approaches, a combination of edge processing and central control seems to be effective as it can interact at central level with market system, while has better performance as there will be intelligent, processing eyes and ears at the edge level, handling most of the local computing at the edge, like optimizing a micro grid, EV charging, or even fault handling locally, under main CC supervision.
In order to cover market and operation optimization approaches, a combination of edge processing and central control seems to be effective
More functions, more complexity
And finally is the issue of increasing number of distribution level functions. New functions are increasingly connected through the grid, adding complexity as well as hidden control coupling through grid electrical physics even when they appear on the surface to be independent. The complexity of and number of such functions, makes shifting the grid management paradigm to a central/decentral combination. A main control center supervising, and coordinating autonomous grid edge intelligence.
This also in line with growing number of interconnected endpoints to be managed, sensed, and/or controlled. Large numbers of devices with embedded processing and communication capabilities are increasing the potential efficiency of the grid. However, these intelligent devices must be managed in terms of provisioning, accounting, security, and function to achieve the benefits and mitigate potential operational risks.
The complexity of and number of functions required by a DSP, makes shifting the grid management paradigm to a central/decentral combination inevitable. A main control center supervising, and coordinating autonomous grid edge intelligence.
The DSP CNS system
In the theory of distributed control, a key issue is the mechanism that keeps decentralized and autonomous elements focused on solving a common problem.
Coordinator nodes provide the distributed processing that aligns local elements by exchanging coordination signals in well-defined patterns
The coordinator nodes themselves have core functionality defined by the mathematics, but serve additional purposes as well.
Coordinator nodes provide northbound and southbound communications for inter-layer coordination as well as peer-to-peer communications for intra-layer coordination.
Layered Hierarchical (“Laminar”) Control/Coordination Structure (Courtesy of DOE-OE, Modern Distribution Grid Report)
In the theory of distributed control, a key issue is the mechanism that keeps decentralized and autonomous elements focused on solving a common problem.
The DSP grid management model
The DSP grid management model presumes that the DSO has substantial responsibility for coordinating DER operation in its service territory and handles the interface to the bulk system Transmission System Operator (TSO).
Key properties of Laminar Coordination Frameworks
- Extensibility
- Boundary deference
- Local objective support (selfish optimization)
- Constraint fusion
- Scalability
- Securability
For more information on Modern Distribution Grid Architecture visit: https://gridarchitecture.pnnl.gov/modern-grid-distribution-project.aspx
More Modern Grid Management Insights at: