This paper presents a load-flow algorithm using the Newton-Raphson method on an extended complex-number plane, in which the real and imaginary parts of each bus volatage are treated as complex numbers respectively. As the results of solving double-dimensional power-flow equations, the solutions converge on real numbers for feasible conditions of power systems, that is, usual system states. If the solutions converge on complex numbers for ill-conditioned system, e. g., heavy-loaded conditions, it can be confirmed that the system state does not have feasible solutions. The authors also develop a method for obtaining feasible system conditions from complex-number solutions calculated in the above-mentioned approach. This algorithm consists of two steps; selecting the specified variable which is the most effective for updating the system state, and estimating a boundary value of the specified variable selected. While the former is based on the so-called sensitivity analysis, the estimation is carried out by approximating the characteristic curve of the relation between imaginary parts of the solutions and the specified values. Numerical examples for various model systems show that the proposed technique works excellently.
We have already made clear that case-based reasoning (CBR) is very useful for the fault restoration support system in electrical power networks. This paper describes the construction method of case-base, adaptation and modification method of a case which is a fundamental process of CBR, and the results of evaluation. The following points were clarified through the simulations of fault restoration under various conditions with electrical power networks of practical scale. (1) Construction method of case-base and the guide to the selection of cases which should be registered in case-base were made clear. Restoration plans which include many cases are registered in case-base. And cases which would take long computation time are additionally registered in case-base. In the application which requires high speed processing, such as fault restoration, there exists an optimum value in the total numbers of cases in case-base. (2) The function of load switching which is essential to perform adaptation and modification of a case flexibly was studied, and embedded it in our system. Inference power was augmented by the implementation of this function, optimum or sub-optimum solutions were obtained. Processing time with CBR is considerably shorter than other methods, and this feature is remarkable with complicated cases.
In order to monitor the voltage stability of an electric power system, various voltage stability indices have so far been proposed. However, they behave quite differently from index to index, and hence, none of them can be an almighty index being effective for a wide variety of voltage instability situations. Therefore, the development of an effective monitoring scheme by the combination of several powerful indices has become one of major topics in power system engineers. In the authors' opinion, minimum requirements for such a scheme should be: (1) to indicate a system-wide stability margin, (2) to show loading margins at critical buses, (3) to provide correct results and (4) fast computation time. It is also desirable that constitunent indices should change linearly with respect to load level and should be continuous even when the reactive power output of a generator reaches the upper limit. This paper proposes an effective monitoring scheme using three indices that satisfy all the above requirements. First, based on an examination of multiple load flow solutions, we have developed a method to find important buses associated with voltage collapse. Then, a new index is proposed which can indicate loading margins to these buses. It is shown that the new index, combined with two of the conventional indices, enables successful monitoring. The linearity and continuity are realized by a normalization operation on these indices. The effectiveness of the proposed approach has been verified through numerical simulations on the IEEE 30, 57, 118 bus test systems.
In recent years, the sizes of Energy Management Systems (EMS) and Supervisory Control and Data Acquisition (SCADA) systems have grown to huge proportions. This is for two reasons: •The power systems to which they are applied have grown in size and complexity. •Their own functions have become more diverse and sophisticated. This raises the following problems: degradation of response time, decreased reliability, limited expandability, less maintainability, increased costs. A functionally distributed system that is characterized by parallel processing and independent subsystems (parallel-independent architecture) will solve these problems. The system is comprised of a group of functional units, each of which runs in parallel, independently and asynchronously. Copies of some programs and power system status data are stored in the relevant functional units, and data-driven architecture is adopted, which eliminates the need for a centralized control mechanism. The feasibility of a functionally distributed system was tested through construction of a prototype. The resugts were satisfactory.
Electric power systems are expanding in size and complexity, and the requirement for energy management system (EMS) is becoming more important. In this computer control system, a single control computer is mainly used as the primary computer and its software is very complicated because of its huge number of small, quick tasks to obtain high response speed. Therefore, much effort is needed to develop and modify the programs, and the responsiveness of this centralized architecture varies greatly when many faults take place in the power system. This paper describes a new distributed architecture for the EMS. Distributed processors execute the functions cooperatively with periodical access to the common bulletin board database in which information about the power system exists. This architecture facilitates the software development and maintenance, and it also enhances the performance by the parallel processing of the distributed functions.
This paper presents a new method of simulating transient responses of multi-machine power system following a large disturbance. The conventional model used in the simulation is composed mostly of synchronous machine models, in which the armature transients are neglected, and steady state transmission system equations. This is because it is very complicated to treat the transmission system and the armature circuits of synchronous machines as differential equations. Naturally the model can not evaluate the effects of these transients neglected. The new method proposed uses the new synchronous machine model and the steady state transmission system model. The former is obtained by modifying slightly the conventional armature transient neglected model. Thus the resultant simulation model has almost the same structure as the conventional one. However it can evaluate approximately the effects of armature transients on the dynamical behaviour of the synchronous machines. For example, it can e valuate easily the angular back swing following a three phase short-circuit fault. In addition, the new method can also estimate the maximum values of developed torque, induced voltage and armature current of synchronous machines at the moment when the disturbance occurred and it was cleared.
The model experimental apparatus for poor contact portion of the gas-insulated switchgear is assumed to a rod-to-plane gap with small arc current in SF6. The breakdown voltage in a gap at gas pressure of 0.1 MPa and small arc current of 1 A have been studied. The effect of the small arc current on the breakdown in the gap was investigated experimentally, and the breakdown mechanism in the gap was studied. Also gas temperature and density were calculated by fluid analysis. It was cleared that the small arc current played an important role on the breakdown in the rod-to-plane gap, and the hot gas produced by an arc was a principal factor of reduction of the breakdown voltage. The decomposition products produced by the arc were absorbed in an activated almina effectively, and the breakdown voltage in the gap increased by setting the activated almina in a chamber. The time history of the breakdown voltage was characterized by six domains, and these breakdown mechanisms were explained by the development of the arc. The schematic diagram of the breakdown mechanism was proposed by the authors.
In thyristor equipment installed near 66kV GIS-cable system, mal-firing of thyristor was generated by surge induced in earthing system during closing operation of GCB. Though earthing impedance was lowered by laying a floor with copper plate, mal-firing could not be suppressed. So investigation using full-scale experimental equipment was made. As the results, the mechanism and method for suppression of this mal-firing are clarified. The resume of results is as follows. (1) High frequency surge induced in earthing system contains two frequency conponents, those are, conponent in the frequency range of several hundred kHz and conponent over 10MHz. (2) The frequency conponent which generates mal-firing is in high frequency range over 10MHz. (3) Condensers attached at joint between GIS tank and power cable sheath and at joint between transformer and power cable sheath are effective for suppressing mal-firing. (4) Earthing mesh is not effective to suppress surge over 10MHz induced in earthing system.
This paper presents experimental studies and theoritical considerations on CO2 evolution phenomena from phosphoric acid fuel cells at start-up processes of the plant. When hydrogen is supplied to cells under an air already supplied condition, a CO2 pulse is detected in the exhaused air. Rise time of the CO2 pulse depends on the flow rate of the hydrogen and coincides with rise time of the cell voltage and the duration of the negative current in a cell. The negative current in the cell increase in its value with time and suddenly change to a positive pulse in regular sequence from the upper stream to the lower stream of the hydrogen and the time lag of their sequence depend on the flow rate of the hydrogen. Measured CO2 volume coincides approximately with an estimated CO2 volume from an electro-chemical equivalent value of the negative current. All these results suggest that the electromotive force of the upper stream portion in the cell causes corrosive current on the lower stream portion of the cell. High flow rate of the hydrogen or low hydrogen content in the fuel gas are not enough to extinguish the CO2 evolution, though increasing flow rate of the hydrogen are effective to decrease the CO2 volume. When the air is supplied to cells under the hydrogen already supplied condition, any trace of the CO2 gas can not be detected. In conclusion, the fuel gas should be supplied prior to the air at start-up processes of phosphoric acid fuel cells.