Effect of communication network infrastructure on load frequency control

Sudipto Bhowmik

Deregulation in the power sector has raised several new issues. Of particular interest to the studies here are those relating to communication infrastructure. Traditional methods for providing communications and control need to be reviewed. One such service is load frequency control, which now needs to incorporate bilateral contracts between generation companies and loads. For frequency control to be effective, a secure, reliable and open communication network needs to be implemented. As of now, few load frequency simulation models include the effects due to the communication channel. The effects of communication were not a major concern in the vertical utility structure as utilities were cohesive in their actions and the control locations was relatively fixed and known. Thus, signal decimation was achieved using dedicated links. However with deregulation in the power generation sector, a necessityfor enhanced communication infrastructure to support increasingly varied services and dynamic generation-load geographical structure, is inevitable. A new distributed and duplex communication system based on datagram packet switching using common off the shelf (COTS) hardware seems to offer the most features and support for most of the services envisaged. This gives rise to a need to define the parameters and variability introduced with this new communication topology. In this thesis, communication models, based on the requirement and structure of the load frequency control, are proposed based on modern queuing theory. These models help approximately define bounds for delay times in the network layer. The approximations are adequate for including the effects of network behavior in a robust load frequency control model. Simulations of constant delays and random delays are performed on different configurations of generators involved in pure load frequency control as well as different bilateral participations. The effects of these and their consequences are tabulated. Simulations integrating the AGC model with a data network simulator were performed and the effects of various network parameters on the automatic generator control are subsequently investigated. Results indicate the possibility of failures arising from communication system problems and suggest the importance of precise communication models. In addition to these modeling issues, data and information security design issues pertinent to the load frequency control signals are presented.

Effect of communication network infrastructure on load frequency control

Files and links (1)

Abstract

Metrics

Details