Turbine early valve actuation is effective to suppress the first swing, but it has a defect that it enlarges the succeeding swings, and that it makes the generator step out at the recovery of mechanical input. In this paper, one method of excitation control in operation of the early valve actuation was proposed. For the first swing, the excitation was kept at the ceiling voltage to avoid the step out. For the succeeding swings, it was controlled to damp them as fast as possible, and to avoid the step out at the recovery of the mechanical input. The control was switched at the peak of the first swing, and it was calculated according to the maximum principle of Pontryagin after that. The stability limit of the first swing is determined by the ceiling voltage of excitation, but the damping of the swings is caused by the rise and fall of the internal voltage E'q. Especially, the fall of E'q in the first backward swing plays an important role, and the bottom voltage of excitation determines the damping characteristic. In order to enlarge the stability limit, it is necessary to lower the bottom voltage for improvement of the damping characteristic as well as to raise the ceiling voltage for enlargement of the first swing stability limit. In the example system, it was shown that the system can be stabilized up to the steady state stability limit. Based on the optimal control, a feedback control was introduced and shown that it can produce good results close to the optimal control. Some comparison with the high-speed excitation with PSS was also made.
An optimal short-range scheduling of multi-reservoir hydro-thermal power systems considering runoff prediction is discussed in this paper. Using Prasad's rainfall-runoff model and the Kalman filter based second order filtering technique, the hour-by-hour runoff characteristic is estimated for a given short-range forecasted rainfall pattern. Unlike the usual hydro scheduling problem in which a single reservoir operating condition is specified at the end of the scheduling period, various end point conditions are examined to deal with the predicted runoff characteristic and avoid possible water spillages from reservoirs. Dynamic Programming Successive Approximations is successfully applied to get optimum operating strategies for many end point conditions at a time. A rapid convergence characteristic coupled with a simple computation algorithm makes the proposed method one of the atractive approaches for scheduling of complicated multi-reservoir hydro-thermal systems considering runoff prediction.
This paper presents a load frequency control system using a refined controller for interconnected power systems. The controller consists of integral compensator and state feedback. It has been proven that any transfer function between a reference input and a controlled output or between a disturbance and a controlled output is attained by this controller, if it is attainable by realizable linear compensation. Parameters of the controller are determined upon the optimal regulator theory and additional calculation. The controller can reduce time deviation and inadvertent interchange without information about tie-lines as well as deviations of frequency and net interchange power. This structural feature makes it possible to design the control system for each regulating power plant instead of for the control center. If some state variables are not available, the control system can be designed with the remainder of state variables and a dynamic compensator. Simulation results show that the control system is superior to the TBC system.
A self extinguishing principle of GCB without puffer action (“Autopuffer GCB”) using an external magnetic field was investigated for improvement of the interrupting capability in the small current region (several kA). Properties of the rotating speed of an arc, the arc voltage and gas pressure rise in an expansion space under the external magnetic field were studied experimentally. The relation between the interrupting capability and the radial magnetic flux density in the “Autopuffer GCB” with a permanent magnet (“Magnet-assisted autopuffer GCB”) was obtained experimentally and also analyzed numerically. The results are summarized as follows, (1) The arc column is driven in the azimuthal direction by the radial component of the external magnetic field. (2) The rotating speed of the arc driven by the radial magnetic field of a permanent magnet remains constant in the arc current range below several kA, while it decreases with an increasing current in the current above several kA. (3) Since the arc rotation raises the arc voltage and gas pressure in the expansion space, the gas flow from the expansion space to the arc increases at around current zero, then the interrupting capability is improved in the small current region. (4) It is confirmed that the “Magnet-assisted autopuffer GCB” has a good interrupting capability not only for the large current region but also the small current region.
In order to clarify the grounding conditions for protecting workers on the de-energized circuit from induction due to the live circuit and the required current capacity of grounding conductor for a 1, 000kV double-circuit power transmission line, the phenomena caused by the electromagnetic induction resulting from the live circuit are discussed here following Part 1, Electrostatic Induction. The results are as summarized below. (1) The electromagnetic induction current Ig flowing through a worker and the electromagnetic induction current Ig0 flowing through the grounding conductor can be clearly discussed by dividing the various power line conditions into three main factors, and also by using the results of electromagnetic induction current calculations of a power line having an infinite length. (2) The zone, where Ig does not exceed 1 mA, is within 1 km from the grounding point on the de-energized circuit, due to the grounding resistance (0.1Ω) of substations at both ends of the line and the difference in the phase configulations on the line when the currents I1 at the live circuit is a constant 1 kA through the line. (3) In the same manner, the current value of Ig0 is determined by the grounding resistance of substations at both ends of the line and the difference in the phase configulations on the line. Also, Ig0 is approximately 10 A per 1 kA of I1. (4) The current Ig and Ig0 due to the effects of electrostatic induction, and Ig and Ig0 due to electromagnetic induction act in such a manner that the former Ig is added to the latter Ig. This is also the case with Ig0. Thus, Ig and Ig0 in the actual power line can be respectively evaluated as the sum of absolute values for both the former and the latter values.
As a pumped-storage power station is generally located far from load center, it frequently has power system stability problem, especially at pumping operation mode. Capacity of units constructed in recent years is as large as a few hundreds MW, and water head is as high as 500m. The period of water pressure oscillation of such high head station becomes longer and closer to power system swing. At such condition the output is affected by the conduit system dynamics and becomes different from analytical results using the classical model which assumes constant torque or output during power system disturbance at pumping operation mode. This report describes the newly-developed simplified pumped-storage plant model based on the elastic theory of the conduit system and the complete characteristics of the turbine, which has necessary accuracy to be analyzed being connected to bulk power system. Comparison with the real plant test result verifies the validity of the developed model. The interaction with power system is examined by linearlized analysis and simulation study.
It is generally known that insertion of a barrier under nonuniform field in the air increases the flashover voltage, which is called a barrier effect. There are many uncertainties, however, about the barrier effect of an SF6 gas that remain to be clarified. We therefore examined the influences of the barrier shapes and positions on the flashover voltage by inserting a barrier between a hemisphere capped rode and a plate in the SF6 gas under near atmospheric pressure for the lightning impulse voltages and AC voltages (50Hz). As a result, the following findings were obtained. (1) Insertion of a disk-shaped barrier reduces the flashover voltage; a cup-shaped barrier whose surface curves almost along the equipotential line increased the flashover voltage by 30 to 40%. (2) Trapped charges on the barrier surface is of the same polarity as the applied voltage and the magnitude of positive charge was greater than that of negative charge. It was also found that there is a correlation between the amount of trapped charge and the flashover voltage. (3) As mentioned above, application of a barrier under nonuniform field in the SF6 gas of atmospheric pressure is considered to help reduce the size of, for example, a cubic type gas-insulated switchgear.
This paper presents experimental studies and theoretical considerations on local overheat phenomena of catalysts on some components of a phosphoric acid fuel cells stack exposed to leakage combustible gas. By blowing air contained with few % of hydrogen, catalysts having low heat wasting capability such as catalysts on a PTFE film can be heated up to glitter and found to set flame to the inflammable PTFE film. On the other hand, catalysts on massive carbon structure can not heat the carbon even at the local overheat and glitter conditions of the catalysts. In order to prevent overheat damage of stack components, suppression of hydrogen concentrartion in air, exclusion of useless large catalyst pieces on components and careful usage of infiammble thermal insulator such as PTFE film are proposed. Application of new detecting method of catalysts on stack components is also recommended.
For the purpose of obtaining usable energy from the sun, a proportional plus integral plus derivative (PID) controller is used in almost all the solar thermal energy utilization systems. However, it is difficult to collect the heat continuously close to a prescribed temperature using a PID controller because the solar radiation is often interrupted by passing clouds. We investigated, therefore, a Model Reference Adaptive Control (MRAC) system. In order to demonstrate its effectiveness, we constructed a MRAC system and introduced it into the collector loop of a solar system. Typical experimental results were already reported and it was shown that the MRAC algorithm was suitable for controlling a system affected by irregular disturbance in the incident solar flux. On the basis of these results, this paper gives experimental comparison between the different parameter adaptation algorithm of the MRAC and simulation for robustness by difference of order of transfer function between the reference model and the plant. From these results, we define that controllability of various adaptation algorithms of the MRAC is nearly same and is superior to the PID and that control of the MRAC system goes well when the order of reference model is equal to or higher than that of the plant.