[雙語翻譯]電氣供配電外文翻譯--配電系統(tǒng)可靠性的理論研究分析計算（英文）

1、 1Abstract — Reliability of distribution systems is an important issue in power engineering for both utilities and customers. Pow- er distribution reliability is closely related to individual compo- nent reliability.

2、In this paper, the reliability as measured by cer- tain count indices, mainly the System Average Interruption Du- ration Index (SAIDI) and the System Average Interruption Fre- quency Index (SAIFI), is quantified in terms

3、 of individual distri- bution component reliability. This is an extension of previously reported results. The value of the approach lies in distribution expansion planning in which a radial system may be networked wi

4、th a view to improve reliability. The effect of inserting Distri- buted Generation (DG) in the distribution system is studied in this paper. The result of this insertion may be that the number of outages increases bec

5、ause of the possible failure of the DG hard- ware. Two new indices for this application are proposed: the ratio of DG energy to interrupted energy and the ratio of DG operating duration to outage duration. Index Term

6、s — distribution engineering, reliability, SAIDI, SAIFI, repair time, distributed generation resources. I. ELECTRIC POWER DISTRIBUTION SYSTEM RELIABILITY ELIABILITY is a key issue in the design and opera- tion of electr

7、ic power distribution systems – especially in view of sensitive digitally controlled loads. Con- temporary loads are often digital in nature, and these loads are frequently sensitive to interruptions and, indeed, many

8、 other power quality problems. The customers them- selves are perhaps becoming more sensitive to interruptions due to the possibility of industrial manufacturing interruption, commercial loss of sales, and residential

9、 nuisance. Sophisti- cated control systems may actually exacerbate the impact of service interruptions. Competition in power marketing may be impacted as well: industrial customers may seek to locate at places wher

10、e power system reliability is high. The distribu- tion system accounts for almost 40% of the overall power sys- tem and 80% of customer reliability problems [1]. For these reasons, distribution system design and operat

11、ion is critical for the power industry. The performance of distribution systems may be quanti-fied by measures of voltage regulation and classical power distribution engineering issues. These issues include: evalua- t

12、ion of losses; power factor; overhead vs. underground designs [2-4]; counts of anomalous events [5-9]; and power quality at The authors acknowledge the support of the Power Systems Engineering Research Center (PSerc) u

13、nder grant NSF EEC-0001880 as well as the Future Renewable Electric Energy Distribution Management (FREEDM) Center under grant EEC-0812121. Authors Al-Muhaini, Heydt, and Huynh are with the Department of Elec-trical, C

14、omputer, and Energy Engineering at Arizona State University, Tempe, AZ 85287, and can be reached at {mohammad.almuhaini}, {heydt}, {antho- ny.huynh} @asu.edu. the point of end use [10, 11]. Reference [6] specifically

15、 ad- dresses the value of ‘count indices’ (i.e., counting unplanned and undesired events such as outages or low voltage cases) for the purpose of standardized distribution system planning, and [8] addresses the probab

16、ilistic analysis of these indices. In recent years, the move to the use distributed generation re- sources in the distribution system and their impact on distribu- tion system reliability has also been considered – for

17、 example in [7]. These references are only small samples of the litera- ture since the full literature is voluminous. References [11 – 13] are samples of distribution system engineering analysis and design – an area

18、 of considerable attention for over 100 years. It is important to state the reality that three phase distribu-tion systems often exhibit unbalanced behavior, either in the nominal operating state, or in the faulted s

19、tate. For example, the largest number of distribution faults is single phase in na- ture. Also, nominal operation often entails unbalanced load- ing because the system points of service can not be balanced in real ti

20、me – they can only be approximately balanced by rotation across the three phases in the distribution primary. This rotation of loads (i.e., approximate balancing) is accom- plished in the design phase and in occasiona

21、l subsequent eval- uation of phase loads at the substation supply. References [14, 15] discuss the three phase case. A reliability evaluation study quantifies reliability based on component reliability data and can be

22、used to evaluate past performance and predict future performance of the distribution system. A reliability evaluation study can also identify the problematic components in the system that can impact reliabil- ity. Th

23、e reliability study can also help to predict the reliability performance of the system after any expansion and quantify the impact of adding new components to the system. The number and locations of new components ne

24、eded to improve the reliability indices to certain limits can be identified and studied. Reliability is basically quantified by the probability of a component or system to operate as expected or not to operate as exp

25、ected. The duration and frequency of misoperation are also significant in evaluating the reliability of a device or sys- tem. In this paper, the event count indices will be studied, principally the System Average Inter

26、ruption Duration Index (SAIDI) and the System Average Interruption Frequency In- dex (SAIFI), ????? ? ????? ???????? ?? ??? ?????????????????? ?????? ?? ????????? ?????????????? ? ????? ?????? ?? ?????????????????? ????

27、?? ?? ????????? ?????????The Reliability of Power Distribution Systems as Calculated Using System Theoretic Concepts Mohammad Al-Muhaini, Student Member, IEEE Gerald T. Heydt, Life Fellow, IEEE Anthony Huynh, Stud

28、ent Member, IEEE R 978-1-4244-6551-4/10/$26.00 ©2010 IEEE3formulas for the case that Tf >> Tr. Note that in typical power distribution engineering, Tf is in the order of tens of thousands of hours and Tr i

29、s in the order of a few hours. Fig. 2 Simple configuration of series and parallel components in a generic system TABLE I EQUIVALENT TIMES OF FAILURE AND REPAIR OF SERIES AND PARALLEL COMPONENTS (ADAPTED FROM [19, 20]) E

30、quivalents Series Parallel eq f T) 2 ( ) 1 () 2 ( ) 1 (f ff fT TT T+) 2 ( ) 1 () 2 ( ) 1 (r rf fT TT T+eq r T) 2 ( ) 1 () 1 ( ) 2 ( ) 2 ( ) 1 (f ff r f rT TT T T T++) 2 ( ) 1 () 2 ( ) 1 (r rr rT TT T+III. THE RELATION

31、SHIP BETWEEN THE COUNT INDICES AND REPAIR / FAILURE TIMES Billinton and Allan [23] show how repair time and failure rate may be used in the radial case to find reliability at distri- bution system buses. In subsequent s

32、ections, we extend the results of Billinton and Allan to distribution systems with dis- tributed generation. The SAIDI and SAIFI event count indices may be graphically analyzed as in Figs. 1 and 2. Noting that SAIFI

33、is the average interruptions frequency per customer and can be calculated by finding the interruption frequency of all buses divided by the number of customers connected in the system, ,T NB1 ii N i AIFSAIFI∑ = =(4) w

34、here Ni is the total number of customers connected in each bus, NT is the total number of customers in the system, and B is the total number of buses. The same idea can be applied for the average duration of all outag

35、es in one year, namely SAIDI, which is simply the summation of the interruption duration of all buses divided by the number of customers connected in the system, .T NB1 ii N i AIDSAIDI∑ = =(5) It is possible to combin

36、e the results of Table I, (2), and (3) to obtain the AID and AIF for a receiving end bus fed by either two series components or two parallel components. This re- sult gives the equivalent AIDeq and equivalent AIFeq (as

37、 ‘seen’ at the receiving bus) as shown in Table II. As in Table I, the equivalent AID and AIF of two simple components in series or parallel assume that Tf >> Tr and the supply bus is 100% relia- ble. The resul

38、ts in Table II are simply obtained using the re- sults of Table I followed by the definition of the equivalent AID and AIF at a power delivery bus being AIF T eqr and ) /( 8760 eq req f T T +respectively. The results

39、shown in Table I agree with Billinton and Allan in [23]; and the results shown in Table II for the radial case are equivalent to that of Billinton and Allan. The parallel case in Table II and the new notation for AID

40、 and AIF, however, are extensions to the cited litera- ture. TABLE II EQUIVALENT AIF AND AID OF A POWER DELIVERY BUS AS A FUNCTION OF THE AID AND AIF OF EACH COMPONENT Equivalents Series supply Parallel supply ?????

41、??? ? ? ???? ????AIF? ? AID?AIF?8760????? ???? ? ???? ????????8760In the foregoing discussion, the three phase detail of the distribution system is not addressed. In some cases, distribu- tion system interruption is

42、done using single pole technology [24]; that is, the faulted phase is isolated and the other phases remain sound. In a sense, the three phase fault case is a worst case since all three phases are isolated and interru

43、pted during a fault. If the loads are distributed across phases A, B, and C, loss of one phase will result in only loss of one-third of the services. IV. TYPICAL CALCULATIONS, RESULTS AND APPLICATIONS Table I implie

44、s that the AID and AIF (as well as other de-rivative indices) may be calculated from circuit topology and values of individual component mean time to failure or mean repair time (Tf, Tr). Equivalently, the AID and AIF

45、 at a deli- very bus in a network can be calculated from the AID and AIF values of individual components’ – that is, the AID and AIF of a delivery bus served by the given individual component using Tr = AID/AIF,

46、 Tf = (8760/AIF )-AID. As an example to demonstrate the effect of different ar-rangements of parallel components, consider two radial feed- ers in parallel with different possible tie connections between them as shown i

47、n Fig. 3. For all the lines in this system, the mean times to failure and repair are listed in Table III. The dotted lines in Fig. 3, denominated as 1, 2, and 3, are potential locations for one added tie line. The A

48、ID, AIF, SAIDI and SAIFI are listed in Tables IV, V and VI for each tie line. These results are found using Table I for both series and parallel connections. Note that the results shown do not include loss of the sup

49、ply power at buses A and E. The results in Tables IV, V and VI show that the minimum AID and AIF for each bus is obtained when the added single tie line is directly connected via the shortest parallel connection and t