Answers
Hierarchies in Large Scale Power Systems:
Because of their complex structures and large sizes, electric power networks are typically monitored and controlled according to their hierarchical structures. Instead of modeling the intricate dynamics of the entire system, the system dynamics is modeled by deriving submodels relevant for each particular sub-process. This is based on observing different time scales over which sub-processes evolve under certain assumptions. The overall system behavior can be fully portrayed by piecing together those simpler, yet essential, elements. The theoretical basis for this type of modeling in large electric power systems was introduced in [8].
The basic submodels are the
(i) primary (local) model at a device level, (ii) secondary (area-wide) level for each administrative area, and (iii) tertiary (global) level representing the interconnected system, Figure 2-1.
The primary control level is entirely decentralized at present. Within this level, controllers respond to the small but fast local disturbances appearing at the terminals of each generator. The speed governor units in electric machines maintain the control of this level. Primary controllers stabilize system dynamics within a very short time constant, T,, i.e., on the second scale, with the performance specification of a minute,
or so.
The secondary control level is decentralized and is particularly useful in analyzingand controlling the dynamic performance within an administrative area (subsystem level). This model represents all generators and large number of loads connected by transmission lines in each administrative area.
The secondary control is implemented at a slower time scale, T,, than that of the primary control (i.e., T, is typically on the several-second scale, with the performance objective over 10 minutes, or so). The secondary control is intended to stabilize system outputs within the administrative area that are disturbed by changes within the area as well as by the changes in neighboring areas. Presently implemented AGC is based on this control structure. Seen from the interconnected system level, each subsystem uses AGC using decentralized
measurements at its own level only. The theoretical tertiary control level is coordinated.
The aggregate tertiary-level models describe the inter-area dynamics among administrative areas and are useful for regulating inter-area variables such as tie-line power flows. These models evolve
on an even slower time constant, Tt, than the secondary level rate, Ts, i.e., on the minute scale. This higher level structure is not currently used in the utility industry.
However, its importance is increasing as the electric power market is changing and becoming more competitive. It is plausible that in the future, decentralized regulation
at the secondary-level would not be sufficient to respond to intense interactions among the areas under an open access environment. The later parts of this thesis provide examples illustrating potential problems of this sort.
Hierarchical level models higher than the tertiary control level described in this topic can also be developed.
For example, in present utility industry, the control centers reset the scheduled values of transmission power among the areas and the phase angle of the slack generator at a much slower rate for economic reasons. The unit commitment procedures such as turning on- and off- the available generators in anticipation of demand on a daily basis, is yet another process of interest. These processes, at least in concepts, could be regarded as evolving at hierarchies beyond the tertiary level. However, these models are beyond the scope of this topic.
а.SH3ィ(-0.s) 2Ha This is the presnt Regitny t uwhich genrotes i opexatii uenu at