Page 24 - Mathematical Models and Algorithms for Power System Optimization
P. 24
14 Chapter 2
(2) Comparing the pumped storage plant and the peak load regulating unit of thermal power
plant: the economical nature of pumped storage plants is that they have higher efficiency
than that of the replaced thermal unit for peak load regulating. When compared to the
economic efficiency of the pumped storage plant, it is also necessary to consider how
much it can be matched with the base-load capacity.
Taking the province’s daily peak-valley load difference as 4000MW as an example, if the peak
load is regulated by the pumped storage plant, the plant needs to have 1200MW installed
capacity, which is about one-third of the peak-valley difference. Thus, it can basically meet the
demand of peak load shifting and give full play to the economic benefits of the daily dispatch
schedule of the power grid. As for the capacity allocation of storage devices in the distributed
power generation system, the difference of peak-valley daily load and the variations of
generator outputs must be fully considered.
Generally, it just takes 100–200s for pumped storage units to generate power from a stopped
state to full load. The units have the characteristics of fast startup speed and power generation.
Normally, the electricity price in the peak period is several times higher than that in the valley
period. The valley price is charged for power consumption of a pumped storage plant while
pumping water, whereas the peak price is paid for the power generation of the pumped storage
plant. Thus, besides the technical factors, peak load regulating with the pumped storage units in
the power grid will bring more efficient economic benefits.
Although the pumped storage units have a significant peak load regulating capacity, in actual
operation the pumped storage power plant must also meet a series of constraints, such as the
initial and end reservoir storage, the maximum and minimum operating storage per hour, etc.
Similar to the constraint of a storage battery’s charge-discharge period, the pumping status
variable of a pumped storage plant must be an integer. Therefore, considering the regional
operation characteristics of the province’s power grid, coupled with the limitations of various
operational constraints, the mathematical model obtained is a large-scale mixed-integer
programming model. It is necessary to select a reasonable and feasible algorithm to solve these
large-scale pumped storage daily economic dispatch optimization problems.
2.1.3 Overview of This Chapter
This chapter studies the optimization model and algorithm for daily multiarea economic
dispatch with a pumped storage plant based on the mixed-integer programming. The objective
of the study is to improve the optimization theoretical knowledge and computer technology
application in daily economic dispatch.
This chapter not only introduces the proposed mathematical model and algorithm but also
presents the basic ideas of the modeling process, which is to transform the practical problems of
complicated multihydro and multithermal power plants and multiarea operations into a