| The equilibrium of power supply and demand is completed in light velocity, but at the same time this process has to comply with KVL and KCL laws as well as lots of requirements for quality and security. Thereby, compared with common commodities, power is extremely complicated and different. Therefore, it is quite difficult to explain accurately the problems of market economy in power system operation only depending upon current economic theory without combining with power system operation principles well. For this reason, this paper will combine with economic theory and aim at power system that is comprised with production, transmission and using section, to study deeply on the economic disciplines reflected from power system operation dispatching with all kinds of constraints being considered, and accelerate the development of power system operation dispatching theory under market circumstance.On the basis of economic theory, two attributes of power commodity, integral and differential attributes, are defined by analyzing the disciplines of power commodity supply, demand and equilibrium. And only one attribute of real-time matching attribute under a more strict condition is defined. On this background, the concept and expression of equal power commodity is presented according to the mathematical optimization theory. Then it is proved that this concept can be fit for resource optimization based on mathematical foundation. In order to obtain and use the eligible power commodity, there are all kinds of resources to be needed, which are called ancillary services currently. Ancillary services problem is one of the difficulties for electricity market implement, and also the studying focus on power system operation dispatching under market circumstance. Besides, equal power commodity includes all these ancillary services. So, the idea to carry out the equivalent measure is presented, and the general model is built to implement the idea which will establish the studying foundation for this paper.With the above ideas and theory as kernel, the studies of this paper will include active power dispatching, reactive power dispatching as well as active and reactive power united dispatching in power system operation, which is the essential theory problem of power system operation dispatching under market circumstance. In addition, aiming at cost covering problem caused by non-convex in power system operation dispatching, the problem of cost compensation under market mechanism is discussed, which makes the studies of this paper become a whole system. The main achievements will be shown as following.Firstly, based on DC power flow, the problems of active power dispatch will be solved, whose contributions have four aspects. (1) In term of active power reserve, the purpose of setting is to cover uncertain factors happening in operation, such as generating uncertainties, transmission network uncertainties and loads uncertainties. By analyzing on the effects of these uncertain factors on power system operation, the cause and effect of uncertain factors leading to reserve is clarified, and the essence of reserve is disclosed, and the responsibility of reserve is cleared. And it is presented that each participant should confirm its reserve responsibility according to the degree of influence on system risk level. The studies indicate that reserve is set for explicit target according to some risk principle. And there is a clear mapping relationship between reserve suppliers and users. To large-scale system, reserve shows the overlapping characteristic. So, if the participants take on their own responsibility and sharing reserve mutually, it will be the best win-win situation on some secure level. (2) Then, if the balance between generation and load is only considered, a new dispatching model of active power with reserve constraint can be presented on certain risk level according to the presented thoughts above. The reserve constraint can clear the targets and responsibility, and make reserve setting cover the anticipatory risk level properly. It can also show the marginal value signals of reserve. Simultaneously, this constraint can clear reserve suppliers and users and their responsibilities accurately. Compared with traditional methods, this idea has an essential difference. It can adapt to the need of open market, and show the further significance of equal power commodity with opportunity cost as reserve value informations. (3) Transmission network is the path of reserve to use. If Transmission network restricts, some capacity of reserve will be invalid. When the uncertainties of transmission network are considered, the phenomenon of reserve transferring is unavoidable. So, according to the given risk principle, the active power dispatching model considering network uncertain factors is built, which is an expansion of the presented model above. This model can modify the reserve setting effectively when the uncertain factors of transmission network happen, and clarify the transferring mechanism of reserve allocated without network restrictions. It transfers the marginal value signal of reserve from system to nodes which produces a spatial difference, and makes the model above more perfect. (4) It is a remarkable characteristic that demand side becomes the initiative participants under market mechanism. In addition, the uncertain events are stochastic. So, the conservative shortcoming exists in the studies above. If the contribution of demand side on reserve and the probability of uncertain events happening is considered, it will bring more objective benefits. Therefore, based on the studies above, with the social expected benefit maximization as objective, the active power dispatching model under probability condition is built by introducing the probability of pre-contingencies happening. This model carries out a further compromise method to deal with benefits and risk, and fits more for market practice. Expected weighting of objectives is obtained by the probability of different pre-contingency, which can carry out the increasing of social benefit according to probabilistic risk level.Secondly, this paper studies on the economic principles when active power dispatching is affected by reactive power and voltage. The main contributions will be shown in four aspects as following. (1) In power system operation, the value of reactive power is reflected by the saving of operating expense on normal operating state and the contribution of reactive power supporting on normal or exceptional conditions. Therefore, from the aspects of supply and demand of reactive power, the mathematical description of cost changes caused by reactive power in power system operation is given. The following conclusions are theoretical results that have been rigorously proven. There is no strict boundary for reactive power supply or demand. Reactive power value is decided by whether leading to the changes of active power operation cost or not. If the cost of active power operation increases, the supply or demand of reactive power will be punished. If not, the reward will be given. The shadow of punishment or reward reflects the marginal value information of reactive power. Simultaneously, the supply or demand of reactive power must be finished absolutely. Otherwise, it will not only affect its own benefit, but also influence the others, which is decided by physical disciplines of power system operation. (2) In normal operation state, when the reactive power output of generating unit doesn't restrict its active power output, the increasing cost of active power dispatching caused by reactive power distribution is mainly the opportunity cost produced by networks losses increasing or decreasing, which is caused by voltage changes for the reason of reactive power demand or supply. In such a case, when the active power dispatching mode is confirmed, the reactive power optimization with active power losses minimum as objective is carried out. Using the byproducts of reactive power optimization, a fast approximate modified method for nodal marginal prices is presented. The essence of this modified method is to allocate the increasing cost caused by network losses, including active power distribution and reactive power distribution, which embodies the rewards and punishment mechanism of every participant's contribution on losses. This method also implements measuring the value of reactive power in power system operation dispatching. (3) Further, based on mathematical optimization, according to the tight degree between reactive power distribution and voltage level, three types of solution states are defined, relaxing state, restricted state and critical state. And the changing ruless of increasing cost caused by reactive power on three states are analyzed. It also domonstrates the value produced by supporting system voltage level. This leads to two important results. On relaxing state, reactive power distribution is close to the current economic distribution, and the reactive power flow in the transmission network is the least, and the losses change caused by unit reactive power flow is the least which leads to the least marginal value of reactive power. On restricted state, there must be some nodal voltages at its lower limit. In order to support this situation, the re-distribution of reactive power makes losses increase. And the relevant nodal reactive marginal price will be enhanced, which also indicates the main undertaker for increasing losses. The critical state indicates that the reactive power can satisfy the requirement of system voltage level properly, which will lead to the reactive marginal value hoist suddenly. According to three states, it can define the degree of reactive power supply and demand, and guide the reasonable distribution of reactive power resource. (4) In another situation, when generating unit restricts its output of active power due to reactive power output, for example, some lower loaded units need to operate in leading phase zone, which will lead to the active power restricted by reactive power. According to solving the complete OPF, the essence of opportunity cost caused by reactive power under this situation is deduced. From the results of analysis, the value of reactive power supporting includes not only the opportunity cost caused by network losses, but also the remarkable opportunity cost due to reactive power tying up active power. This situation shows the equivalent weight interchange between active and reactive power, and brings out the essential marginal value.Finally, the studies above on power system dispatching theory under market circumstance all base on the assumption that the model of studying objective satisfies the convex programming conditions. However, for the quasi fixed costs of generating unit, such as no load cost, on-off cost and so on, if clearing the schedule by marginal price, it must lead to the problem of cost compensation and can not carry out the effective market trades. So, based on complete competition theory, two dispatching methods which separately aims at cost minimization and consumer payments minimization are analyzed and compared. The latter doesn't cause the problem of cost compensation, but it can't give the singal for social benefit maximization. And in opposition, the former can realize this goal and also reflect the selected reason for the former in this paper. Hereto, this paper deals with the problem of cost compensation by using the conception of additive cost and gives basic method and idea for deeper study. The advantage is that it doesn't change the marginal information which obtains social benefit maximization and carries out the least losses of social benefit to avoid the blindness of compensating.Therefore, based on effective dealing with the problem of cost compensation, the operation dispatching theory of power system under market circumstance built according to this paper embodies the integrity.
|
PDF Full Text Request