Optimal reserve management for restructured power generating systems
Introduction
There are various engineering systems in the world. One type of these systems is for the resource and energy supply (such as power and gas supply, etc.). Power generating systems in electric industry are this type of systems. In a more broad sense, such engineering systems can be designated as generating systems. The only function of generating systems is to provide reliable and economical resource and energy to their customers. Back up resource or energy, which is designated as reserve in this paper, is commonly used in generating systems in order to provide adequate supply under unexpected system operating conditions such as generating system failures and load growth. In the restructured power system, a generating system can plan its own reserve and can also share reserve with other generating systems according to their reserve contracts. The reserve structure of a generating system should be determined based on balance between the required reliability and the reserve cost. The objective of reserve management for a generating system is to schedule the reserve at the minimum total system reserve cost while maintaining a required level of supply reliability to its customers.
A generating system usually consists of generating units (GUs) with different capacity, cost and availability. Different generating systems are usually connected together through supply network to form a complex generating system (CGS) as shown in Fig. 1. The supply network is designated as transmitting system (TS) in this paper. The objective of the interconnection is to share the reserve. A GS usually signs the reserve agreements with other GSs to increase its reliability and to reduce the reserve cost. A CGS can operate in different states or performance levels due to random failures. The performance levels of a GS in the CGS are usually determined by the number of units, capacity and availability of each unit, transmitting system and reserve agreements with other GSs. The reserve management of a generating system in a CGS therefore is a complex optimization problem, which belongs to system-structure optimization problems addressed in Ref. [1], where binary-state reliability was considered. A CGS and its GSs and TSs are multi-state systems [10] and can be represented using the corresponding equivalents [3], [4]. A universal generating function introduced in Ref. [2], [5] was used to calculate reliability of a multi-state system. Ref. [6] proposed a system-structure optimization algorithm for multi-state systems. In this algorithm the system components were chosen from a list of available products and were characterized by their performance, availability and cost. The algorithm in Ref. [6] was extended [7] to solve a component redundancy optimization problem for a multi-state system. Selecting redundant elements to maintain a system reliability level was mainly considered in these algorithms. The reserve sharing within a restructured power system through reserve agreements was discussed in Ref. [8]. However, the reserve management of a generating system is a redundant resource problem and the general methods have not been comprehensively developed. The cost of the utilized reserve has not been considered in these techniques.
This paper proposes a technique to optimize the reserve structure of a multi-state GS in a CGS. The objective of the problem is to determine the optimal reserve structure (the set of reserve contracts), which can provide a desired level of system reliability to meet the load upon the minimal reserve cost for a GS. The technique is based on the universal generating function (UGF) and the genetic algorithm. The solution comprises reliability and reserve cost (capacity cost and utilization cost) estimates. A power system-IEEE reliability test system is used to illustrate the technique. In Section 4, a GA with the special encoding scheme that considers the structure of reserve capacity and reserve utilization order is developed for the optimization problem. A fast technique based on UGF evaluating the system reliability is presented in Section 3.
Section snippets
Multi-state system models
In order to determine the optimal reserve structure of a specific GS in a CGS, the GS is represented by a MSGS, all other GSs are represented by the MSRPs, and the network between the MSGS and a MSRP is represented by the corresponding MSTS. The equivalent model for the reserve optimization of a MSGS using these equivalents is shown in Fig. 2, in which there are M MSRPs and M MSTSs.
Assume that the MSGS has Kg states and the available generating capacity for each state ig is . These states
UGF-based reliability evaluation
To solve the optimization problem, the reliability indices have to be determined during optimization process. The procedure for reliability evaluation using UGF has been proved to be very convenient for numerical realization and requires relatively small computational resources [5]. The UGF technique is therefore, used to evaluate the reliability indices. In this section, general definition of a multi-state element or a multi-state system is introduced first and the UGFs for MSGS, MSRPs and
Optimization technique using genetic algorithms
Unlike the exact optimization methods, which usually require complicated procedure to obtain a good single solution, GAs guide a simple evaluation search toward the global optimum in the solution space and operate on a group of solutions. The comprehensive theory and application on GAs for engineering fields are in [11], [12]. The implementations of GAs in reliability engineering have been presented to select component reliability improvement level [13] and to determine optimal
Illustrative example
The restructured IEEE-RTS [16] is used to illustrate the technique. The restructured IEEE-RTS consists of six MSGSs. In this example, the optimal reserve structure of the specific MSGS, which has reserve agreements with other five MSGSs, is determined using the proposed technique. The MSGS owns 6×50 MW hydro units, 5×12 MW oil units, 2×155 MW coal units and 2×400 MW nuclear units. MSRP 1 owns 2×20 MW gas turbine units and 2×76 MW coal units. MSRP 2 owns 3×100 MW oil units. MSRP 3 owns 2×20 MW gas
Conclusions
This paper illustrates a practical technique for optimal reserve management of restructured power generating systems. The technique is to determine reserve providers and the associated reserve capacities from the economical view while satisfying the reliability requirements. The reliability network equivalent, universal generating function and genetic algorithm are the foundations of this technique. The reliability network equivalent method provides a simplified approach to represent generating
References (17)
Handbook of reliability engineering
(1994)Optimal standby problems and a universal generating function
Soviet J Comput Syst Sci
(1987)- et al.
Unreliability cost assessment of an electric power system using reliability network equivalent approaches
IEEE Trans Power Syst
(2002) - et al.
Reliability assessment of a restructured power system using reliability network equivalent techniques
IEE Proc-Gener Transm Distrib
(2003) Universal generating function
Soviet J Comput Syst Sci
(1986)- et al.
Redundancy optimization for series-parallel multi-state systems
IEEE Trans Reliab
(1998) Redundancy optimization for multi-state systems with fixed resource-requirements and unreliability sources
IEEE Trans Reliab
(2001)- et al.
Reliability assessment of a restructured power system considering the reserve agreements
IEEE Trans Power Syst
(2004)
Cited by (23)
Predictive maintenance scheduling for multiple power equipment based on data-driven fault prediction
2022, Computers and Industrial EngineeringCitation Excerpt :In order to meet the growing demand for electricity, the scale of power plants in China is expanding, and the number of power equipment is also increasing rapidly, which brings some difficulties to the maintenance and management of equipment (Tripathi, 2018; Benjamin and Hobbs, 2007). Especially in complex and uncertain environment, the operating condition of equipment will change with various external accidents, and it often occurs that multiple equipment need to be maintained/repaired in the same period (Ding et al., 2006). Considering the limitation of maintenance resources, as well as the requirements of power output and power supply continuity, it is of great academic research and practical application value for the maintenance scheduling of multiple power equipment operating in abnormal conditions (Alaswad and Xiang, 2017; Liang and Parlikad, 2020).
Optimal loading and protection of multi-state systems considering performance sharing mechanism
2016, Reliability Engineering and System SafetyCitation Excerpt :It is not realistic to examine all possible solutions exhaustively. In the literature, the genetic algorithm (GA) has been proven efficient in solving large-scale reliability optimization problems [5,9,11,42,47–49]. The details of GA can be found in [50].
Using inverse Lz-transform for obtaining compact stochastic model of complex power station for short-term risk evaluation
2015, Reliability Engineering and System SafetyCitation Excerpt :A numerical solution for this problem can be obtained by using Genetic Algorithm (GA). The GAs are the most popular heuristic algorithms for solving different kinds of reliability optimization problems for MSS [9, 16, 20]. In Fig. 1, one can observe a power station consisting of n generating units connected to common switchgear.
Availability modeling and optimization of dynamic multi-state series-parallel systems with random reconfiguration
2014, Reliability Engineering and System SafetyCitation Excerpt :Based on this approach, an optimization problem is formulated with the consideration of the availability of the system and its operation costs. Though there are different approaches, such as Pareto dominance [36,34,25] and weighted sum [7] to deal with multi-objective optimization problems, an usual way is to optimize one objective constrained by the other ones [5,23,30–32,41]. In particular, this paper aims to minimize the total system cost with the constraint that system availability is greater than a predetermined level.
A multi-state model for the reliability assessment of a distributed generation system via universal generating function
2012, Reliability Engineering and System SafetyCitation Excerpt :For more realistic reliability assessment of power systems, multi-state models, which have been widely applied to resolve system reliability assessment problems [33,38], are being introduced to describe the random behaviors of the generation sources [8] and the degradation/repair of the components [35]. Compared to binary-state models, the multi-state models offer greater flexibility in the description of system state evolution, for more accurate approximation of the real-world power systems [14,50]. In this paper, we consider a relatively complete electrical network with distributed generation and describe it as a multi-state stochastic system by means of an analytical probabilistic model.