Prof. Nicodemus Abungu Odero

Due to the increased importance of DFIGs in optimization of real and reactive power losses and the maintenance of voltage profile, the general methods of DG placement and sizing in the existing literature cannot be of practical importance in DFIG .In this paper a pure PSO method used in general DG is compared with a HGAPSO in the siting and sizing of DFIG with the objective of minimizing power losses. The corresponding Combined participation factors are assigned using the DFIG Domain Distributed Slack Bus Model and a comparison made on the two schemes of loss minimization.

This paper proposes to determine the feasible optimal solution of the economic load dispatch power systems problem using Particle Swarm Optimization (PSO) considering various generator constraints. The objective of the proposed method is to determine the steady-state operating point which minimizes the fuel cost, while maintaining an acceptable system performance in terms of limits on generator power, line flow,
prohibited operating zone and non linear cost function. Three diff erent inertia weights; a constant inertia weight CIW, a timevarying inertia weight TVIW, and global-local best inertia weight GLbestIW, are considered with the (PSO) algorithm to analyze the impact of inertia weight on the performance of PSO algorithm. The PSO algorithm is simulated for each of the method individually. It is observed that the PSO algorithm with
the proposed inertia weight (GLbestIW) yields better results, both in terms of optimal solution and faster convergence.

Load forecasting refers to the prediction of future load conditions based on present or historical data. This is important especially for transmission planning and economic dispatch. In this paper, an Artificial Neural Network (ANN) is trained using historical data for a sub-station at Ruiru, Kenya and the corresponding loading conditions for the sub-station are used to test its accuracy in forecasting the electrical load when given other parameters.

Though several algorithms for optimizing DG location and size in a network with the aim of reducing system power losses and enhancing better voltage profile have already been proposed, they still suffer from several drawbacks. As a result much can be done in coming up with new algorithms or improving the already existing ones so as to address this important issue more efficiently and effectively. Majority of the proposed algorithms have emphasized on real power losses only in their formulations. They have ignored the reactive power losses which is key in the operation of power systems. In modern practical power systems reactive power injection plays a critical role in voltage stability control, thus the reactive power losses need to be incorporated in
optimizing DG allocation for voltage profile improvement. The results of the few works which have considered reactive power losses in their optimization can be improved by using more recent and accurate algorithms. This research work aimed at solving this problem by proposing a hybrid of GA and IPSO to optimize DG location and size while considering both real and reactive power losses. Both real and reactive power flow and power loss sensitivity factors were utilized in identifying the candidate buses for DG allocation. This reduced the search space for the algorithm and increasing its rate of convergence. This research considers a multi-type DG; type 1 DG (DG generating real power only), type 2 DG (DG generating both real and active
power) and type 3 DG (DG generating real power and absorbing reactive power)."

In present times, the use of DG systems in large amounts in different power distribution systems has become very popular and is growing on with fast speed. Although it is considered that DG reduces losses and improves system voltage profile, this paper shows that this is not always true. The paper presents a GA-IPSO based approach which utilizes combined sensitivity factor analogy to optimally locate and size a multi-type DG in IEEE 57-bus test system with the aim of reducing power losses and improving the voltage profile. The multi-type DG can operate as; type 1 DG (DG generating real power only), type 2 DG (DG generating both real and active power) and type 3 DG (DG generating real power and absorbing reactive power). It further shows that though the system losses are reduced and the voltage profile improved with the location of the first DG, as the number of DGs increases this is not the case. It reaches a point where any further increase in number of DGs in the network results to an increase in power losses and a distortion in voltage profile.

Reduction of system losses and improvement of voltage profile is one of the key aspects in power system operation. Though many methods are used to achieve this aspect, Distributed Generation (DG) has found increased usage nowadays due to its many advantages. Majority of algorithms proposed in this area have emphasized on real power losses only in their formulations. In modern practical power systems reactive power injection plays a critical role in voltage stability control, thus the reactive power losses need to be incorporated in optimizing DG allocation for voltage profile improvement. This paper aims at solving this problem by proposing a hybrid of GA and IPSO to optimize DG location and size while considering both real and reactive power losses. The hybrid technique aims at inheriting the good traits from the two techniques while avoiding the undesirable ones. Arithmetic crossover and mutation has being employed in the proposed algorithm.

Due to the increased importance of DFIGs in optimization of real and reactive power losses and the maintenance of voltage profile, the general methods of DG placement and sizing in the existing literature cannot be of practical importance in DFIG .In this paper a pure PSO method used in general DG is compared with a HGAPSO in the siting and sizing of DFIG with the objective of minimizing power losses .The corresponding Combined participation factors are assigned using the DFIG Domain Distributed Slack Bus Model and a comparison made on the two schemes of loss minimization. The obtained results for the real and reactive power losses and voltage voltage profile illustrate the DFIG need in the modern power system.

Past literature has attempted to solve the problem of network reconfiguration with Distributed Generators(DGs) without taking into consideration the intermittent renewable at a close proximity. Distribution Network Reconfiguration (ADNR) must account for uncertain behavior of loads and wind when the commercial wind based DG, Doubly Fed Induction Generators (DFIG) supports a significant part of network. In this paper, a new Hybrid Bacterial Foraging and Differential Evolution (HBFDE) algorithm is considered for the ADNR problem with minimum loss and an improved voltage profile. In the HBFDE algorithm the Differential Evolution (DE) algorithm is combined with the Bacterial Foraging (BF) algorithm to overcome slow and premature convergence of BF. Indeed, the proposed algorithm is based on the evolutionary natures of BF and DE, to take their advantage of the compensatory property, and avoid their corresponding drawbacks. In addition, to cope with the uncertainty behavior of loads and wind, a stochastic model is presented to solve the ADNR problem when the uncertainty related to wind and load forecast is modeled in a stochastic framework on scenario approach basis. The proposed algorithm is tested on the IEEE 33 - Bus Radial Distribution Test Systems. The results of the simulation show the effectiveness of proposed algorithm real time and real world optimization problems facing the smart grid.

The problem of power system optimization has become a deciding factor in current power system engineering practice with emphasis on cost and emission reduction. The economic and emission dispatch problem has been addressed in this paper using two efficient optimization methods, Artificial Bee Colony (ABC) and Particle Swarm Optimization (PSO). A hybrid produced from these two algorithms is tested on a 10 generator test system with valve point effects. The results are compared with differential evolution (DE), Strength Pareto Evolutionary Algorithm (SPEA) and Non Sorting Genetic Algorithm (NSGA) and found to be effective on the combined economic and emission dispatch problem.

With the increased penetration of distributed generation into the power distribution system, the traditional
load flow analysis that assumes a single slack bus has become impractical. The existing literature focuses on slack bus placement taking only real power losses into place. However with increasing need of reactive power in maintaining voltage stability at the consumer end; reactive power generation management cannot
be ignored any further. Thus a distributed slack bus model taking into consideration both real and reactive power losses is of paramount importance. The wind based doubly fed induction generator is an attractive option for both real and reactive power loss compensation since it is economically attractive, can be grid
connected and it has the capability to generate and absorb reactive power. A distributed slack bus model
using combined participation factors is developed in this paper to distribute the slack(real and reactive power losses). The combined participation factors are formulated using the method of Lagrange
multipliers and the distributed slack bus model employs a Genetic Algorithm of a Newton Raphson Solver.

Power utility companies are required to supply customers with power within specified voltage limits. Voltage rise in networks with distributed generators therefore poses a challenge. This paper presents a coordinated network controller whose objective is to maintain an optimal voltage profile across the power network. The operations of distributed generators, on-load tap-changing transformers and reactive power sources are controlled. The controller is modelled as an optimisation problem which is solved using Particle Swarm Optimisation. The IEEE 30-bus test network is then used to verify the effectiveness of the controller. The results obtained show that this controller can greatly improve the voltage profile of a power network by varying the parameters of existing generation and voltage control equipment.

The global rise in energy demand has resulted to the over exploitation of both renewable and non renewable energy sources. Most feasible hydroelectric power (HEP) plants sites have been exploited and the current focus is on harnessing energy from small HEP plants which have low head and flow velocity rendering them unsuitable for HEP generation. Previous research work focused on improving the turbine shape and efficiency; designing better water intake, improving the generator and development of turbines suitable for low heads. The main aim of this research was to optimize the power generated by low head small hydro plants through the use of hydraulic ram pump (hydram) to boost the water pressure before it impinges on the turbine. In the current work, a smallHEP prototype system was designed fabricated and test runs conducted. The prototype comprised of; a low head water reservoir, a hydraulic ram pump which was used to increase the head of the water emanating from a low head source, a high head reservoir mounted at a the most optimal height based on the hydram flow rate and pressure considerations and a double cup pelton wheel turbine suitably designed to extract power from the water jet. A drive pipe was used to connect thehydram pump to the low head reservoir while the delivery pipe connected the pump to the high head reservoir. Water from the high head reservoir was used to turn the pelton turbine which was coupled to a generator. The flow rate in the drive pipe and the delivery pipe as well as the pressurein the hydram were optimized by adjusting the waste valve stroke length. It was observed that the hydram was able to pump water to a higher head which then increased the power produced by the turbine.

The problem of power system optimization has become a deciding factor in current power system engineering practice with emphasis on cost and emission reduction. The economic and emission dispatch problem has been addressed in this paper using two efficient optimization methods, Artificial Bee Colony (ABC) and Particle Swarm Optimization (PSO). A hybrid produced from these two algorithms is used on the 30-bus 6 generator IEEE test system. The results are compared with ABC, Fuzzy Controlled Genetic Algorithm (FCGA) and Non Sorting Genetic Algorithm (NSGA-II) and found to be effective on the combined
economic and emission dispatch problem.

Power utility companies are required to supply customers with power within specified voltage limits. Voltage rise in networks with distributed generators therefore poses a challenge. This paper presents a coordinated network controller whose objective is to maintain an optimal voltage profile across the power network. The operations of distributed generators, on-load tap-changing transformers and reactive power sources are controlled. The controller is modelled as an optimisation problem which is solved using Particle Swarm Optimisation. The IEEE 30-bus test network is then used to verify the effectiveness of the controller. The results obtained show that this controller can greatly improve the voltage profile of a power network by varying the parameters of existing generation and voltage control equipment.

The 21st Century Power Transformer is produced by combining modern high voltage cross-linked polyethylene (XLPE) cable technology with conventional transformer. The technique of solid insulation is adopted in the new dry transformer so that the pollution from leakage of insulating oil can be avoided, and so XLPE cable-winding transformer is very suitable in environment sensitive places such as populous cities, hydropower stations, and underground caver and so on. This paper is meant to show that the marriage of the well-proven high voltage power cable technology with transformer technology sets a new standard in improving power system voltage stability.

Power utility companies are required to supply customers with power within specified voltage limits. Voltage rise in networks with distributed generators therefore poses a challenge. This paper presents a coordinated network controller whose objective is to maintain an optimal voltage profile across the power network. The operations of distributed generators, on-load tap-changing transformers and reactive power sources are controlled. The controller is modelled as an optimisation problem which is solved using Particle Swarm Optimisation. The IEEE 30-bus test network is then used to verify the effectiveness of the controller. The results obtained show that this controller can greatly improve the voltage profile of a power network by varying the parameters of existing generation and voltage control equipment.

Power utility companies are required to supply customers with power within specified voltage limits. Voltage rise in networks with distributed generators therefore poses a challenge. This paper presents a coordinated network controller whose objective is to maintain an optimal voltage profile across the power network. The operations of distributed generators, on-load tap-changing transformers and reactive power sources are controlled. The controller is modelled as an optimisation problem which is solved using Particle Swarm Optimisation. The IEEE 30-bus test network is then used to verify the effectiveness of the controller. The results obtained show that this controller can greatly improve the voltage profile of a power network by varying the parameters of existing generation and voltage control equipment.

Concerted efforts have been made towards developing more elaborate techniques for solving aperture coupling problems. The majority of these techniques, however, deal with apertures of regular shapes and, in each case, only a particular problem has been solved. It is only with the development of numerical methods, such as the Method of Moments and Finite Element Method that it has become possible to treat irregularly shaped apertures.

However, each of the above methods has its own advantages and disadvantages when applied to different problems. The Method of Moments is an integral equation method which handles unbounded problems very effectively but becomes computationally intensive when material and structural inhomogeneities exist. In contrast, the true power of the finite element method is revealed in three-dimensional volume formulations in the presence of material and structural inhomogeneities. The method requires less computer time and storage because of its sparse and banded matrix. The matrix filling time is also negligible when simple basis functions are used. For complex basis functions, the matrix filling time can be significant.

A suitably implemented hybrid method takes advantage of the strengths of the individual methods constituting it while avoiding their weaknesses. This research therefore, as one of its objectives, has developed a hybrid method that combines the method of moments and the finite element method (MOM/FEM). The analysis is based upon the "generalized network formulation" for aperture problems. The cavity region is subdivided into tetrahedral elements resulting in triangular elements on the surfaces of the apertures.

In this work, a hybrid MOM/FEM solution procedure for the general problem of apertures of arbitrary shapes in thick conducting screens and waveguide walls has been developed and used in the analysis of a variety of representative problems. Appropriate modeling of the aperture/cavity has been carried out using tetrahedral and triangular elements. Suitably defined sets of basis functions have been integrated into the formulation which is capable of accurately evaluating fields of apertures of arbitrary shape. The problem has been formulated by invoking the equivalence principles and utilizing boundary conditions on the apertures/cavity to derive equations which have then been transformed into matrices that are then solved numerically by simulation on a digital computer. The finite element method, employing reliable vector formulation, has been employed in the computation of the interior admittance matrix. Here, edge elements or tetrahedra in which the degrees of freedom are assigned to the edges rather than the nodes are utilized. This resulted in the avoidance of nonphysical or spurious modes, a difficulty that arises when node-based elements are used. Based on the preceding formulation, extensive computation of various parameters for apertures/cavities of various shapes has been done and results presented. The two main classes of problems treated in this study comprise apertures of arbitrary shape in thick conducting screens and waveguide-backed apertures.

Power utility companies are required to supply customers with power within specified voltage limits. Voltage rise in networks with distributed generators therefore poses a challenge. This paper presents a coordinated network controller whose objective is to maintain an optimal voltage profile across the power network. The operations of distributed generators, on-load tap-changing transformers and reactive power sources are controlled. The controller is modelled as an optimisation problem which is solved using Particle Swarm Optimisation. The IEEE 30-bus test network is then used to verify the effectiveness of the controller. The results obtained show that this controller can greatly improve the voltage profile of a power network by varying the parameters of existing generation and voltage control equipment.

The paper deals with an extension of the previous work appearing in a past issue of this transaction by the authors on hybrid FEM/MOM technique for analyzing transmission properties of arbitrarily shaped apertures on a thick conducting screen. In the present work, the effect of placing different dielectric material slabs in the conducting screen cavity on the electromagnetic transmission parameters is first analyzed and, then, the effect of interchanging the positions of these dielectric slabs relative to the incident field. Validation results for rectangular and cross-shaped slots are presented. Close agreement between our data and published data is observed.  Further data has been generated for rectangular, circular, diamond-shaped and cross-shaped apertures.

This paper presents a summary of the current status of Electrical Power Engineering Education at Kenyan universities, followed by a summary of the situation on the same at universities world-wide. Against a backdrop of expected changes in the Kenyan power industry, the paper discusses the industry expectations of a power systems graduate currently and in the future, making a case for the revision of the current power engineering syllabi, concluding with recommendations and strategies for making the required changes.

In this paper a hybrid numerical technique is presented, suitable for analyzing transmission properties of an arbitrarily shaped slot in a thick conducting plane. The slot is  excited by an electromagnetic source of arbitrary orientation. The analysis of the problem is based on the "generalized network formulation" for aperture problems. The problem is solved using the method of moments(MOM) and the finite element method(FEM) in a hybrid format. The finite element method is applicable to inhomogeneously filled slots of arbitrary shape while the method of moments is used for solving the electromagnetic fields in unbounded regions of the slot. The cavity region has been subdivided into tetrahedral elements resulting in triangular elements on the surfaces of the apertures.  Validation results for rectangular slots are presented. Close agreement between our data and published results is observed.  Thereafter, new data has been generated for cross-shaped, H-shaped and circular apertures.

In this paper a hybrid numerical technique is presented, suitable for analyzing transmission properties of an arbitrarily shaped slot in a thick conducting plane. The slot is  excited by an electromagnetic source of arbitrary orientation. The analysis of the problem is based on the "generalized network formulation" for aperture problems. The problem is solved using the method of moments(MOM) and the finite element method(FEM) in a hybrid format. The finite element method is applicable to inhomogeneously filled slots of arbitrary shape while the method of moments is used for solving the electromagnetic fields in unbounded regions of the slot. The cavity region has been subdivided into tetrahedral elements resulting in triangular elements on the surfaces of the apertures.  Validation results for rectangular slots are presented. Close agreement between our data and published results is observed.  Thereafter, new data has been generated for cross-shaped, H-shaped and circular apertures.

When an insulator is placed in an electrode gap, its surface charge field distorts the geometric field, leading to a flashover at a voltage value dependent on the degree of the gap field distortion. This paper reports on studies conducted to determine the relationship between flashover voltages and electric filed distribution along solid insulator surfaces. The surface electric field distribution along different insulator profiles was determined using the Finite-Difference method, and the flashover of the actual profile model measured. The obtained results show that each profile had a peak surface electric field and the higher the peak value, the lower the flashover voltage. The correlation curve for peak electric field and flashover voltage was developed using a curve fitting technique based on the Nelder-Mead simplex algorithm.

In this paper a hybrid numerical technique is presented, suitable for analyzing transmission properties of an arbitrarily shaped slot in a thick conducting plane. The slot is  excited by an electromagnetic source of arbitrary orientation. The analysis of the problem is based on the "generalized network formulation" for aperture problems. The problem is solved using the method of moments(MOM) and the finite element method(FEM) in a hybrid format. The finite element method is applicable to inhomogeneously filled slots of arbitrary shape while the method of moments is used for solving the electromagnetic fields in unbounded regions of the slot. The cavity region has been subdivided into tetrahedral elements resulting in triangular elements on the surfaces of the apertures.  Validation results for rectangular slots are presented. Close agreement between our data and published results is observed.  Thereafter, new data has been generated for cross-shaped, H-shaped and circular apertures.

For my thesis I did a problem formulation and then wrote a computer program to help speedily analyze various insulator profiles for use at high voltages. The program when fed the profile would output the potential and electric field patterns around the high voltage insulator, in addition to predicting it's flashover voltage. Validation of the model was obtained through practical measurement in a high voltage laboratory. Profiles that would insulate very high voltages were arrived at this way in a relatively short time.

When an insulator is placed in an electrode gap, its surface charge field distorts the geometric field, leading to a flashover at a voltage value dependent on the degree of the gap field distortion. This paper reports on studies conducted to determine the relationship between flashover voltages and electric filed distribution along solid insulator surfaces. The surface electric field distribution along different insulator profiles was determined using the Finite-Difference method, and the flashover of the actual profile model measured. The obtained results show that each profile had a peak surface electric field and the higher the peak value, the lower the flashover voltage. The correlation curve for peak electric field and flashover voltage was developed using a curve fitting technique based on the Nelder-Mead simplex algorithm.

For my final year undergraduate project, I developed tariffs for various classes of consumers of KPLC electrical energy using the marginal costing technique. This is a forward looking accounting method that takes into account expansion plans and how the power system is to be operated as demand increases. Expansion plans, which would entail capacity/inestment costs, operation and maintenance costs forecasts, administration and general costs forecasts, total kilowatt costs, and peak and off-peak energy and fuel costs, spring from load forecasts. This resulted in an efficient resource allocation scheme in which power prices to consumers are related to the resource costs of changes in consumption, i.e, the addition of a new consumer or an increase in consumption of an existing consumer will impose additional costs on the enterprise, while a reduction in consumption will save costs. These alteration in costs are the ones that need to be reflected in tariffs. The change in the cost to a consumer of altering his electrical behavior will then mirror the change in the cost to the enterprise. This not only results in fair and equitable tariffs, but also ones that inherently have incentives to better consumption patterns by consumers.

This marginal costing technique contrasts sharply with the accounting (traditional) one which is based on examining the records of past expenditure thus becoming backward looking. Such prorated accounting costs are quite different from the costs relevant to resource allocation and creates the illusion that resources which can be used or saved are as cheap or as expensive as in the past, i.e, resources are as abundant or as restricted as in the past. On the one hand this may cause over investment and waste; on the other, it may lead to under investment and unnecessary scarcity. In addition, if the past holds a number of poor projects, the sunk costs of mistakes, if reflected in prices, will overstate the costs to the consumer of extra consumption, which is not efficient. Tariff schedules and the various simplifications thereof are derived by spreading total accounting costs among consumers. This generates tariffs which relate to average rather than to marginal costs.

.