Daniel Kullmann

Fault tolerant aggregation for power system services

Exploiting the flexibility in distributed energy resources (DER) is seen as an important contribution to allow high penetrations of renewable generation in electrical power systems. However, the present control infrastructure in power systems is not well suited for the integration of a very large number of small units. A common approach is to aggregate a portfolio of such units together and expose them to the power system as a single large virtual unit. In order to realize the vision of a Smart Grid, concepts for flexible, resilient and reliable aggregation infrastructures are required. This paper presents such a concept while focusing on the aspect of resilience and fault tolerance. The proposed concept makes use of a multi-level election algorithm to transparently manage the addition, removal, failure and reorganization of units. It has been implemented and tested as a proof-of-concept on the distributed smart grid test bed SYSLAB at the Technical University of Denmark.

Pattern-based automatic translation of structured power system data to functional models for decision support applications

Improved information and insight for decision support in operations and design are central promises of a smart grid. Well-structured information about the composition of power systems is increasingly becoming available in the domain, e.g. due to standard information models (e.g. CIM or IEC61850) or otherwise structured databases. More measurements and data do not automatically improve decisions, but there is an opportunity to capitalize on this information for decision support. With suitable reasoning strategies data can be contextualized and decision-relevant events can be promoted and identified. This paper presents an approach to link available structured power system data directly to a functional representation suitable for diagnostic reasoning. The translation method is applied to test cases also illustrating decision support.

Asynchronous control of Distributed Energy Resources using behaviour descriptions

The activation of power system services from small DER is a key to building electrical power systems with a high penetrations of renewable generation. Reliable, real-time and low-cost communication to a large number of DER units is required for controlling these resources, but difficult to achieve. This paper presents an alternative to traditional closed-loop control which has the potential to relax the real-time requirements on the communication link between supervisory controller and DER unit. Examples are given, and experimental results are provided as a proof-of-concept.

Implementation and test of demand response using behaviour descriptions

The term Smart Grid describes the effort to enable the integration of large numbers of renewable distributed energy resources into the power grid. The fluctuations inherent in renewable energy resources imply the need to also integrate the demand side actively into the control of the power system. For this effort to succeed, a new control infrastructure has to be put into place. The power system is a distributed system, and it needs a sophisticated communication framework to cope with communication problems, such as delays and failures. Recently, behaviour descriptions have been proposed as an alternative to synchronous communication in power systems, especially with small distributed energy resources. This article presents an implementation of behaviour descriptions and an experiment that has been carried out to evaluate the feasibility of such an approach.