Tutorial Overview On the Modeling and Control of Power Transmission Networks

A large-scale electric power system is a rather complicated interconnection of many systems. At one end there are generation units that convert available energy sources (either fossil, nuclear or renewable) into electric energy. This generated power must then be transported to load centers via high voltage AC transmission lines. The received energy is then delivered to individual customers via a distribution network. During operation, one of the highest priorities is the availability of high quality power to all consumers at all times. Such reliability is difficult to achieve, because of two challenges: (a) Consumer demand is continually changing and cannot be controlled by the power provider, and (b) With current storage technology, electric energy is virtually impossible to store in large quantities. Consequently, a provider must track consumer demand to send dispatch commands to the generators. Furthermore, these dispatch commands must be aware of the power flow limitations imposed by the transmission and distribution network. Failure to implement a suitable dispatch policy will result in poor power quality (erratic voltage levels) and in the worst case power outages.

Poor power quality is addressed by the implementation of a power management strategy, which considers small fluctuations in power demand over small time-scales (seconds to minutes). Fundamental to power management is the concept of spinning reserves. The idea is that a generator can quickly increase or decrease power output within a reasonably sized window of power conditions. Power outages can be prevented by energy management, which considers large demand changes over large time-scales. Fundamental to energy management is the fairly predictable nature of consumer demand. The basic idea is that generation facilities can be scheduled to provide power within specific time periods.

As renewable sources, such as wind and solar, are introduced to the system, additional burden will be placed on the transmission network. Renewable sources tend to be located great distances from load centers and will likely require significant expansion of the transmission hardware. In addition, renewable sources are inherently intermittent and non-dispatchable. With capacity factors of 30 to 40%, power output from renewable sources will vary from zero to three times the average. Short term variations in power output (due to wind gusts and small clouds) are also expected. It is further noted that the power production gaps created by renewable sources must be made up by the remaining dispatchable sources.

In this paper we will present a tutorial overview of the modeling and control system methods used in the operation of power transmission networks. In particular, effort will be made to highlight parallels between these methods and those used in the chemical process industry.