Reliability Evaluation Of Composite Power Computer Science Essay

Published Date: 02 Nov 2017

Ankur Malhotra1 , Rahul Agrawal, Student MIEEE2,Dr. S.K. Bharadwaj, Member IEEE3, Dr. D. P. Kothari, Fellow IEEE4, Reena Jharaniya5

1 M. Tech Scholar, VITS Indore (M.P.), India,2,5 Asst. Prof., VITS Indore (M.P.), India

3 Professor, Department of Electrical Engineering , MANIT Bhopal (M.P.), India

4 Director General, J.B. Group of Educational Institutions Hyderabad(A.P.), India

[email protected]

[email protected],

[email protected]

[email protected],

Abstract – Reliability evaluation has been broadly used for power system planning and operations [1], since they are capable of incorporating various system uncertainties such as equipment failures as well as random variations in generation capacity or load demand. Meanwhile, power systems are becoming more complicated, which leads to highly nonlinear problems in their reliability evaluation. The purpose of assigning the reliability of a composite power system is to estimate the ability of the system to perform its function of transporting the energy provided to the bulk supply points. This paper gives an introduction to reliability evaluation of composite power system. Paper also includes description of reliability evaluation parameters at different levels and different methods for reliability evaluation of power system.

Keywords-Reliability, Power System, Monto Carlo Simulation

Introduction

A very basic definition of reliability defines reliability as a characteristic of any device which shows its capability to perform its adequate function under given conditions. Reliability is one of the very important concerns in power system. The system planners and designers are always concern with power system reliability. The need of reliability evaluation for composite power system is required because of increasing complexity of systems, cost competitiveness, to have alternate designs of system, cost benefit analysis and to study effects of operations and maintenance procedures. The evaluation of reliability of composite system is also necessary because the main objective of the system is to supply electrical energy continuously at low cost and with good quality and as the today’s scenario is of privatization and having a tough competition with others so each company as well as consumer wants a reliable system. The reliability evaluation can be done on the basis of four basic parts- Probability, adequate function, specified period & given environment [25]. Probability provides a numerical measure to the reliability whose scale is between 0 and 1. It is a mathematical concept. Adequate function in power system means the sufficient, desired and proper function of the complete system.. The adequacy term is all about the system capability to meet the load demand in any condition. The adequate function is again the numerical measure which has to evaluate with respect to adequate performance of system. Time period is one of the criteria to measure the reliability. Reliability decreases from 1 to 0 from time t = 0 to time t = ∞ i.e. reliability is a decreasing function of time. Reliability of a system also depends on environment in which it is functioning.

There are two main factors System capability or system adequacy and system security. Sometimes these two factors are used to declare the system as reliable or unreliable. In other words we can say that sometimes on the above two parameters the system reliability depends.

System capability is in concern to satisfy the requirement of power or to meet the load demand. The system capability is all about the ability of system or we can say it’s all about availability of resources at the generating end just to satisfy the load demand at consumer end.. The generating station capability is always maintained above the peak load demand and is decided by doing forecasting of the load by system designers. The system should be adequate even after considering the planned or unplanned outages of the generating units and other equipments which forms the system

At the same time along with this the system should be secure i.e. it should be capable to survive under normal conditions as well as in conditions of sudden loading. It may be over loading it may be under loading. Variation in parameters viz. voltage variation, current variation, frequency variation and even the thermal parameter variation also affects the system stability and hence system security and reliability respectively. All these parameters should remain in permissible limit. After fulfilling all these conditions i.e. adequacy and security the system can be considered under the category of a reliable system.

In today’s time lots of steps are taken to make the system capable with an increased rate of security. One of the methods to improve reliability is grid connection and as a advancement in this the latest technique is smart grid. There are many techniques to evaluate the reliability but in all the methods we need to calculate the reliability indices. Reliability indices like loss of load probability i.e. [LOLP] loss of load expectation (LOLE) & expected energy not supplied (EENS) etc.

As here we are dealing with reliability evaluation of composite power system so a composite power system can be divided in many operating states in terms of the capacity available to fulfil demand subject to the satisfaction of security limits (line flows and voltage limit). Hence, the evaluation of a reliability index for a composite system is very much computationally demanding. The three basic functional zones are those of generation, transmission and distribution as shown in figure-1.

These functional zones can be combined to form hierarchical levels (HL) for conducting system reliability analysis [4]. Reliability assessment at HL I is concerned with the generation facilities [6]. Reliability assessment at HL II considers the generation and transmission as a composite system. The effect of load growth, configuration changes and facility additions can be studied and reliability indices can be evaluated for the overall system, as well as for the individual buses [7-10]. All three of the functional zones are involved in an HL III assessment. The main objective of an HL III study is to conduct adequacy assessment at consumer load points [4].

II. Reliability Indices

Reliability indices can be classified under two categories: [24]

i) Deterministic indices

ii) Probabilistic indices

Deterministic indices show the postulated conditions. They do not show the system reliability and are also not responsible for system reliability. Hence for planning these systems are having very limited use.

Different deterministic reliability indices are:

Percent reserve margin: This shows the excess of installed generating capacity over annual peak load.

Reserve margin in terms of largest unit.

These parameters do not reflect data like unit size and outage rate etc.

Probabilistic indices show the reliability of system and are necessary part for reliability evaluation. These parameters can also impact the system reliability.

Different probabilistic indices are:

LOLP: Loss of load probability: This LOLP may be defined as probability of loss of load exceeding the available generation capacity.

LOLE: This reliability index shows the time when insufficient generating capacity is there to serve peak load.

Expected unserved energy or Expected energy not supplied (EENS): this index measures the expected amount of energy which is failed to serve or supply to consumer because of shortage in basic energy supply.

Composite system reliability indices are also categorized into system based indices and load point indices. Both sets of indices complement each other. System based indices provide an appreciation of global system adequacy and can be used by planners and managers for comparing the adequacies of different systems. However, these indices cannot be used to assess the adequacy of particular system load points. Therefore, load point indices are required to assess the reliability of load points, which is useful for benchmarking load points and identifying weak areas in the system. Load point indices can be used to identify the contribution of each load point to bulk system unreliability.

The system based reliability indices are:

Bulk power interruption index

Average number of load curtailment

Average energy curtailment

Average number of voltage violation

Maximum system load curtailed under contingency condition

Maximum energy not supplied under contingency condition.

The load point indices include the evaluation of:

Probability of failure

Failure frequency

Average of voltage violation

Average load curtailed

Expected energy not supplied

Average duration of load curtailment

With the help of above available reliability indices the reliability of the composite power system is calculated, this calculation will be done with the help of any one of the method of reliability evaluation.

Wenyuan Li [19] has already considered one of the reliability indices to calculate reliability i.e. aging effect. He took BC Hydro North Metro system as a experimental model and shows the impact on reliability due to aging effect. Hagkwen Kim, Chanan Singh [20] also considers aging effect for calculating reliability and calculates LOLE which is one of the reliability indice. Not only this. reliability of power system not completely dependent on real power. Some where it also depends on reactive power. Peng Wang and et al [21] calculate reliability indices by considering the reactive power shortage. In the above discussion one of the reliability evaluation indices is probability of failure and in this we can consider the failure of protection devices. The same was done by Xingbin Yu [22]. In this the system reliability was evaluated by considering the failure of the protection system which is considered to be one of the biggest reason for failure of the system.

III. Composite System Reliability Evaluation Methodology

Reliability of a composite power system can be obtained by various methods. To evaluate reliability there are two basic and important models must be considered in order to perform composite system reliability evaluation. These are the component outage model and the load model. Combinations of the component outage models form the contingencies which are convolved with the load model and the reliability indices are produced.

Independent Outage Models

A component is considered to be on outage when it is unavailable to perform its intended function. A component outage, however, may or may not cause load interruption. Independent outage events including the outage of two or more components are referred to as overlapping outages. The basic component model used in these applications is the two-state representation shown in Figure 2, in which the component is assumed to be either up or down. The rate of departure from the component up state to its down state is the component failure rate λ. The restoration of the component to its operating state is denoted by another transition rate, termed the component repair rate μ. The actual restoration process could be high or low speed automatic reclosure, repair or simple replacement of the failed component by a spare. Different restoration rates are associated with each of these activities. The two parameters, λ and μ can be expressed in terms of Mean Time To Failure (MTTF) and Mean Time To Repair (MTTR) respectively, where, MTTF is the reciprocal of λ and MTTR is the reciprocal of μ.

Load Model

Load model is another important element that is required to perform composite system reliability analysis. This model can be a constant load model and in this case the resulting indices are the annualised indices or it can be variable load. Variable loads can be rearranged to produce the load duration curve which is usually used for power system analysis. The load duration curve can be represented using a multi steps load model [4].

Other then above basic models there are two more main methods which are widely used for reliability evaluation

Analytical Methods

Computational or Simulation Methods

Both approaches are able to calculate reliability indices. Both methods use the adequacy & security of a system state using a power flow methods for calculating reliability. AC load flow or DC load flow [8-9] may be used depending upon the needs of the study.

1. Analytical Methods can be divided in to

Contingency Enumeration Methods (CEM),

State Enumeration Methods (SEM) and Markov Cut Set method. One more fault analysis method is fault tree analysis method and minimal cut set method is one of the part or we can say its the first step of fault tree analysis method to evaluate reliability. Based on the mathematical analysis in all analytical methods a mathematical model is prepared and generally all models are based on markov models. Yong Liu & Singh, C [23] proposed a method to evaluate reliability of composite power system. In that they used markov cut set method which was based on DC optimal power flow. The minimal cut set is analysed using this DC optimal power flow which was limited upto a particular order and then they used markov cut set metod to calculate reliability indices. Yong Liu & Singh, C [26] again use minimal cut set method approach as analytical methodology to evaluate reliability of composite power system but for a short duration of time. Cost outage is also one of effect of reliability. Choi, J.S. and et al [32] use analytical approach to determine the outage cost in a composite power system as a effect of reliability. To evaluate this cost outage they use the analytical method using effective load duration curve.

2. Computational Methods can be divided in to

Monte Carlo Simulation (MCS) & Artificial Intelligence (AI) based on the iteration technique. Monte carlo method can be applied on time varying load to evaluate reliability. For showing the time varying load load duration curve can be used. Monte carlo itself consist of sequential technique and non sequential methods such as state sampling and state transition sampling.

Monte Carlo simulation methods estimate the indices by simulating the actual process and random behaviour of the system. The method, therefore, treats the problem as a series of experiments. In general, if complex operating conditions are not considered and/or the failure probabilities of components are small ( ie., the system is very reliable), then analytical techniques are usually more efficient. When complex operating conditions are concerned and/or the number of severe events is relatively large, Monte Carlo methods are often preferable [5]. The basic sampling procedure can be conducted by assuming that the behaviour of each component can be categorized by a uniform distribution under {0,1}. In the case of a two-state component representation, the probability of outage is the component forced unavailability. It is also assumed that component outages are independent events. The basic information obtained by this component outage is only unavailability of component. The other information by this type of model is availability of component. This gives only average data information. If we increase the information by increasing the no of states than it is going to be quite complex. To remove this complexity S. A. Khaparde, K. Bhattacharyya [30] gives the combination of fuzzy loagic and neural network to evaluate reliability with more accurate power system model and more accurate reliability indices. A. R. Abdelaziz [31] also evaluate reliability on fuzzy based power system. Jaeseok Choi et al [27] presents a concept of composite power system effective load duration curves using monte carlo method. This effective load duration curve is important for generation and transmission both. The reliability indices for composite power system are calculated using this effective load duration curve with monte carlo method. Sometimes the load is completely uncurtained and for such type of problems fuzzy logic can be used. J. Tome Saraiva et al[28] consider the uncertainty in load and define it by fuzzy numbers using monte carlo algorithm. This fuzzy technology is specially used for long term planning where there is uncertainty in load in future. J. He, Y. Sun, L. Cheng et al used a hybrid method to evaluate reliability. They use the combinational advantage of state enumeration method and monte carlo simulation. The advantage of this combination is that SEM is used to solve lower order contingencies and MCS for higher order contingencies. As such there are lots of computational methods but out of those computational methods genetic algorithm is again one of the good computational methods which makes the reliability evaluation easier and reduces the efforts as compared to analytical methodology. Samaan, N.; Singh, C. [33] used genetic algorithm method to evaluate reliability. Genetic algorithm is a computational method which gives the result having resemblance with the actual system results. To improve the accuracy we can increase the number of algorithm generations. In this the actual reliability indices for complete system are calculated using reliability indice data for each load separately. The GA method takes the failure status of each load separately and then combined it to get the annual result. Lingfeng Wang; Singh, C.[34] used this GA method in parallel to speed up the computational process. All of the above described methods for reliability evaluation are used on different test systems. Most of them are used on IEEE reliability test system. The IEEE reliability test system (RTS) is of two versions.RTS-79 which was the first test system of IEEE. RTS-96 has introduced as an advanced version of reliability test system from IEEE. In this there are some changes with respect to evaluation methodologies [35]. R. Billinton also developed a reliability test system which is known as RBTS i.e. Roy Billinton test system. This test system is also used in many research works to evaluate the reliability of system.

IV. Acknowledgement

Its quite difficult to express thanks in words. I would like to thanks all my respected faculties for their precious guidance and time. I want to thank all net lovers who are continuously updating their work on net which works as a path showing candle for us to take our work upto a new destination.

V. Conclusion

This paper summarizes the various aspects for the reliability evaluation of composite power system. This introduces the various reliability indices used for reliability evaluation. The load point indices like LOLP, EENS etc. play an important role in reliability evaluation of composite power system. For reliability evaluation as test systems we have IEEE RTS and RBTS. IEEE RTS 96 is one of the advanced test system with different evaluation methods as compared to IEEE RTS 79. For evaluation Simulation and analytical methods are available. Simulation methods are considered to be superior because it reduces the computational time as compared to analytical methods as well as it gives more accurate system. Monte carlo is one of the good simulation technique for large systems. GA is also a good option for reliability evaluation. Fuzzy logic is also in use for reliability evaluation with unpredictable load conditions.