21, rue d'Artois, F-75008 Paris http://www.cigre.

org

13-103

Session 2000
CIGR

LIFE CYCLE MANAGEMENT OF CIRCUIT-BREAKERS BY APPLICATION OF RELIABILITY CENTERED MAINTENANCE

by T. ORLOWSKA Polish Power Grid Co (Poland) C. NEUMANN RWE Energie AG (Germany) Summary This paper shows the principles of the developed reliability-centered maintenance strategy (RCM), which takes into account the condition of the equipment and the importance of the equipment in the network. The condition of the equipment can be evaluated on the basis of different criteria. The importance describes the consequence of the equipment failures on the network and the economic. As a result the RCM-strategy leads to the conclusion which equipment has to be maintained first. Besides the theoretical principles of this RCM-strategy the authors describe the experience in Germany. Keywords: Maintenance Strategies - Condition - Importance - Life Cycle 1. Introduction 10 kV: 110 kV: 380 kV: 17,0 years; 18,2 years; 15,8 years. 20 kV: 220 kV: 16,1 years; 22,4 years; G. BALZER* Darmstadt University of Technology (Germany) M. HALFMANN ABB Calor Emag Schaltanlagen AG (Germany) A. STRNAD Innovation Consulting (Germany)

select those which should be maintained first. If the ranking is derived from a reliability-centered maintenance strategy an optimum solution can be stated. In this paper the term maintenance also covers the replacement of equipment as e.g. necessary for renovation. 2. Maintenance strategies

The statistics of German utilities give a rough overview of the age of the equipment, which are installed in the system. For example assuming that the useful lifetime of a circuit-breaker will be about 35 years the new installations of each year can be divided in two parts: extension of the system replacement of old circuit-breakers.

If a calculation is performed for the main voltage levels in Germany, the average age of the circuit-breakers are as follows:

Facing the competitive environment the utilities are forced to rethink their maintenance strategies. Economic considerations point out the need for the calculations which are capable to extend the lifetime of equipment and substations and allows an economic and reliable life cycle management. On this way one of the important tasks of the utilities are: to select the equipment, which should be maintained to make a rankink of the equipment.

The higher age of the 220 kV c.b. demonstrates the less importance of this level in the transmission system of Germany. The distribution of the circuit-breakers concerning the different types are listed in figure 1 between 1966 to 1996. There are still a lot of air blast and minimum oil c.b.'s installed which need more maintenance than the new devices of SF6-type.

The ranking of the equipment gives the opportunity to

* Department of Power Systems, Landgraf-Georg-Str. 4, D-64283 Darmstadt

16000 14000 12000 10000 number 8000 6000 4000 2000 0 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 year 96

other SF6 min. oil air blast

Figure 1: Different types of 110 kV circuit-breakers in West-Germany Generally, the maintenance of equipment or systems summarizes the following steps: Inspection: Determining and assessing the actual condition Servicing: preserving the desired condition Repair: restoring the desired condition Replacement: replacing equipment or systems which are no longer usable wards condition-based maintenance(CBM). Conditionbased maintenance is driven by the technical condition of the equipment. Under this approach, all major parameters are considered in order to determine the technical condition with maximized accuracy. For this reason detailed information via diagnostic methods or monitoring systems should be available. A fourth strategy, which additionally include a reliability-based component, has been under discussion recently, and some applications are already in use /7/. The aim of this approach is to include the influence on the importance of the equipment in the network and the actual condition of the equipment. A maintenance strategy is referred to in this paper as reliability-centered maintenance (RCM) and it has to be noticed that this RCM-method is different considering other RCMapplications which consider equipment only. Figure 2 demonstrates the main conclusion of the different maintenance strategies mentioned above, depending on the two criteria: importance of the equipment considering the network and the condition. The presented evoluation of maintenance strategies including the influencing factors gives a hint to the expense for applying a certain strategy.

Several maintenance strategies are mainly used or in discussion in electrical systems (figure 2) /1/ - /6/: corrective time-based condition-based reliability-centered

In a corrective maintenance (CM) strategy, replacement or repair is performed only if a failure occurred. In the case of equipment where investment costs are low and a fault will have only a minor effect, this procedure may result in the lowest overall costs. This strategy will be mainly used in systems with lower voltages. Only severe failure on certain type of equipment will influence the procedure. A time-based maintenance (TBM) strategy featuring predefined intervals rooted in empirical feedback, where components are replaced after a specified period of use, has been practised as the usual maintenance strategy in electrical power systems for many years. This approach generally produces satisfactory results. It will not, however, be the most cost-effective option in all cases, since the equipment will not usually remain in operation up to the end of the lifetime which is possible. Since some years, however, there has been a developing shift away from time-based maintenance and to-

Figure 2: Evolution process of maintenance strategies

The reliability of supply is influenced by the type of maintenance strategy applied. Figure 3 shows the effects on network reliability comparing the different strategies in the networks nowadays.
Network reliability

measurement results know-how of the service department and so on.

In addition, rating and weighting factors have to be applied to take account of the influence of each criterion on the overall value c (condition) representing the condition of the equipment. Also, the results of typical measurement carried out on the equipment can be referred to in the assessment. Examples are the operating time, gas and oil analyses, tightness, and synchronism of the contact switching.
NLi CM TBM CBM RCM

Useful lifetime

Figure 3:

Network reliability comparing different maintenance strategies CM: corrective TBM: time-based CBM: condition-based RCM: reliability centered Nli: limit of network reliability

Without a doubt, the RCM strategy offers definite advantages, since besides the condition of an equipment or a system, its importance in the network also crucially influences the maintenance action required. It clearly emerges that network reliability can be upgraded by correct selection of the equipment for maintenance. These equipment will be maintened which have the main influences of the network reliability 3. Reliability-centered maintenance strategy

Both reliability-centered and condition-based maintenance strategies require the same data for assessing the condition of the equipment. Of course, many equipment/network-specific data can be collected for evaluation. In practice, however, they must be restricted to the major influencing variables, for which there is actually utilizable information derived from the process, from relevant statistics, or from manufacturer's documentation. Results are crucially dependent on assessment of the individual criteria involved in evaluating a particular equipment, and the weightings assigned to them. Defining an equipment's importance and assessing the consequences of a fault basically constitute a practical but also subjectively influenced value. In this context, there are numerous different parameters to be considered, e.g.: non-availability of the substation failure rate of the equipment substation configuration interrupted active power kind of customers, social impact financial payment in case of loss energy.

A reliabilitiy-centered maintenance strategy, which combine the two aspects of condition and importance, requires the following procedure: The condition of the equipment has to be determined. The importance of the equipment for the network as a whole must be determined, e.g. the influence of equipment failure on the reliability of supply. Both information inputs must be combined and evaluated in order to specify the optimum sequence of maintenance work for the indi-vidual devices (equipment, substation).

Figure 4 shows the basic procedure for linking the two assessment criteria: condition of the equipment importance of the equipment in the network

The condition, e. g. of a circuit-breaker, can be evaluated on the basis of different criteria for example: age type of circuit-breaker number of short-circuit interruptions number of switching operations experience with this type of circuit-breaker

Both these criteria are combined in an appropriate manner enabling an overall assessment to be arrived at. The overall sequence of maintenance according to figure 4 will lead to a ranking, which equipment has to be maintened first, second, and so on. This will be expressed using the index o combining the two aspects which are mentioned above. The procedure is described in the following. After the values for the c (condition) and i (importance) parameters have been calculated, the results (crosses) can be listed in an X, Y system of coordinates, as shown in figure 5. The c and i axes are scaled in such a way that the c and i values can at maximum assume the value 100.

Figure 4:

Procedure for maintenance planning cR - cM: < cM: servicing, repair no action required

A large value for i signifies that the equipment concerned is of high importance in the network. The vertical axis represents the condition of the circuit-breaker concerned, while the horizontal axis reflects its importance in the network. A cross in the top left-hand corner corresponds to a circuit-breaker which, although in a poor technical condition, would not cause any major consequences if it fails.
c 100 2 cR d1 cM d4 d5 100 i 1 d2 4 d3 5 3

Inside a particular area, the priority for individual measures is obtained by examining the distances d from the straight line. With the aid of the classification achieved in figure 5 the necessary maintenance action can be specified. Basically, it is also possible to specify an area of corrective maintenance (CM) in this scheme, i.e. an area in which all the circuit-breakers needed to be maintained only after a malfunction. These elements are characterized by being below a critical importance threshold, so that a failure has no significant effect on network reliability. In addition, it should be possible in future to subdivide the area " cR - cM" further, so different inspection intervals can be defined. Figure 5 also makes clear, assessing the condition is not of itself sufficient for maintenance planning. For example, servicing of circuit-breaker 5 would be more necessary than for circiut-breakers 1 and 4, although circuitbreaker 1 shows a bad condition. 5. 5.1. Evaluation of criteria General

Figure 5:

Interpretation of the assessment result c condition of the circuit-breaker i importance of the circuit-breaker

By contrast, a cross in the bottom right-hand corner designates an equipment which is in very good condition. A failure of this device would entail substantial consequences for network operation. The distances d1 to d5 illustrate the sequence in which the individual circuit-breaker must be serviced or replaced. 4. Comparison with previous practise

As explained before, the RCM method considers the condition of the equipment as well as its importance in the system. Therefore this method can be used as a valuable tool for planing of renovation measures. In the following three different types of 420 kV circuitbreakers are considered. The aim of the investigation is to prove the practicability and ability of the computer based RCM-method and the exactness of the assessment. The condition of the circuit-breakers was evaluated by means of a questionnaire. Besides the service life of the breaker some general features as arc extinguishing medium and drive mechanism as well as the service experience and the failure rates are recorded. The analysis can be based on exact values, e.g. gained by measurements and test results, but also on fuzzy values where an estimation or judgement of experts is necessary. The scaling of the condition axis in the first step is more or less arbitrary and the limit value between servicing and replacement is based on a first estimation and has

The evaluation levels cM and cR are entered in figure 5 in accordance with empirical feedback from the network engineers concerned, i.e. as user-specific data. The parallels to abscissa i through cM and cR characteristics limit the areas in which from a technical viewpoint a servicing routine or replacement is required. Area assignment can be defined as follows: 100 - cR: replacement

finally to be confirmed by more thorough investigations on the objects in concern. For definition of importance of the breaker in the system different options are available. The definition applied in these investigations takes into account the non-availability of the equipment and the loss of power caused by the outage of the breaker in question. The non-availability is ascertained from the outage time and the statistical outage and failure rate respectively. The outage time of the breaker depends on the scheme and the layout of the substation and the location of the breaker in the station. The loss of power is evaluated from the real load flow via the breaker in concern. For comparison the value is normalised and related to the maximum value in the network configuration under consideration. The first type of breaker under investigation is a minimum oil circuit-breaker with eight units per pole. These breakers with a spring drive mechanism were taken into service between 1973 and 1978. The second type, a double pressure SF6 breaker with four units per pole, belongs to the first generation of SF6 circuit-breakers and were equipped mainly with hydraulic drive mechanisms, but some with pneumatic drives. These breakers were installed in 1972 and 1973. The third type counts to first generation of SF6 puffer-breaker. It is a four unit breaker with an hydraulic drive mechanism taken into service in the period from 1975 to 1978. 5.2. Minimum oil circuit-breaker The population of this type investigated amounts 34 circuit-breakers in 8 substations. The results are given in figure 6 The condition of the breakers is in the region between 55 and 80, in which the breakers from 1973 exhibit a distinctly worse condition although only some years more in operation. This was mainly due to an essential modification and improvement of the arcing chamber in 1974 resulting in a better performance of the chamber and also in lower maintenance expenditure for this part.
minimum oil breaker 100 80 condition c 60 40 20 0 0 20 40 60 importance i 80 100 120

%. All these latter breakers are installed in generator feeders. From these results one can derive a renovation strategy in four stages. First the three breakers of the before mentioned generator feeders and the two other breakers in a highly exploited substation have to be exchanged. Secondly the breakers in the 50 % range and at last those in the 30 - 40 % range have to be considered. Because of the low importance a renovation of the breakers below the 30 % range can be disregarded and reconsidered when a new inventory of condition and importance for this type of breaker is carried out after a certain period of time. 5.3. SF6-breaker, dual pressure type The population of this type of breaker consists of 21 pieces. The results are presented in figure 7.
SF6 breaker, dual pressure type 100 80 condition c 60 40 20 0 0 20 40 60 80 100 importance i

Figure 7:

Evaluation of 420 kV SF6 breakers, pressure type

Since only breakers of two subsequent years of manufacture are installed the condition varies only in a small range. Compared to the condition level of the minimum oil breakers the average is some what higher, i.e. the condition of this breaker in general is worse which is also in accordance with the overall service experience. The renovation suggested is to exchange all these breakers in a defined sequence beginning with the breakers in the range of importance of 60 % and above. The second section contains the 30 % up to 60 % range and the last the section below 30 %. Because of the reduced reliability, these breakers have to be observed thoroughly, e.g. by yearly recording of the condition. 5.3. SF6-breakers, puffer type The population under investigation of this type comes to 38 breakers. The results can be taken from figure 8. The importance of the population in concern is to more than one half below 35 %. Only a small amount is above 60 %. One of these breakers with 90 % importance belongs to a generator feeder.

Figure 6:

Evaluation of 420 kV minimum oil breakers The importance is mainly in the region below 50 %. But four breakers are classified in the range above 90

SF6 breaker, puffer type

100 80 condition c 60 40 20 0 0 20 40 60 80 100 importance i

Figure 8:

Evaluation of 420 kV SF6 breakers, puffer type

The condition in total is considerably better compared to the breakers dealt with before. It is in the range between 45 % and 55 %. With regard to the condition even the worst breaker of this type is better than the best minimum oil breaker of the same year of manufacture. This proves the superiority of the puffer-type breaker technology and complies with the service experience in general. The small range of variation depends on the years of manufacture close together without any particular change in technology. The results with this breaker show that in the foreseeable future no renovation measures will be necessary. They can also be used to adapt the limit value between normal maintenance activities and renovation. Based on the comparison between the three breaker types under investigation the condition limit value was assumed to 60 %. By means of some more detailed tests and measurements of the breakers condition and consideration of the failures rates of breakers which will be renovated soon the limit value can be fixed more precisely. Nevertheless the method presented shows a satisfactory ability and exactness of assessment. It can surely be improved in the course of time by further experience and data gained with the RCM method. 6. 6.1. Example Circumstances

A German utility (approximately 6,000 km2 service area, 1.6 million private and commercial clients, 1,150 staff members) which had realised the importance of optimising maintenance processes wanted in 1999 to audit, evaluate and optimise the existing assets as well as to analyse and optimise the maintenance processes in co-operation with an internationally active consultancy company within the scope of a model project. The electric supply company followed a CBM strategy. The maintenance processes were in line with this strategy and had been optimised. The aim of the project was to reduce the costs of the whole maintenance area by a further 30% and the duration of the project was nine weeks. 6.2. Analysis phase

The total consult concept served as the strategic framework of analysis - a holistic approach for asset management which does not only consider technical aspects (for instance diagnosis, network analysis), but also takes into account commercial factors (for example investment strategies, budget planning) in the analysis figure 9.
Component Diagnosis
(Equipment Data; Visual Inspections; Measurements; Operation History)

Network Diagnosis
(Load Scenarios, Standards; Supply Concept; Automation Concept, etc.)

System Diagnosis

Development of Scenarios, Technical and Economical Evaluation and Ranking

Maintenance Concept
(Maintenance, RETROFIT, Reinvestment, etc.)

Network Concept
(Expansion, Reduction)

System Concept

Investment-/Budget-Planning

Figure 9:

Consult concept

Some companies are complaining about the increasing pressure of cost savings in asset management which is accompanied by the necessity to maintain a constantly high level of availability. However, the approaches for achieving these goals differ from each other considerably. It is obvious, that an one-sided reduction of the maintenance budget, for example in the course of changing from a Time Based Maintenance (TBM) to a Condition Based Maintenance strategy, is often not enough. Especially in this context, it is much more necessary to optimise maintenance processes so that a strategy can achieve its full potential.

Auditing the available equipment and comparing them to the existing data bases was the first necessary step for achieving the goals of the project - firstly because until then the maintenance and archiving of data depended largely on the person who conducted it, and secondly because certain data (such as condition and degree of maintenance) was not available for all equipment. In order to conduct a cost-effective and quick data acquisition of the condition of equipment on a specific day, the e-Pat solution was used. This solution is a software tool for handheld computers which is used to document the evaluation of condition efficiently on the spot. The required data was acquired with the help of the staff members of the client who had been trained

thoroughly before that. The data which was acquired in this manner was sent via Internet to central computer on a daily basis. In this way it was possible to evaluate more than 8,000 equipment (circuit-breakers, power and instrument transformers, disconnectors and so on) within four weeks, compare the data to the existing data bank of the utility and generate an updated version simultaneously. In addition to that, all critical equipment were identified in this step of the project (critical in the sense that the security of supply depends on them to a large extent in case of a power breakdown). The main data of the h.v. circuit-breakers can be derived from table 1. In total 255 123 kV c.b. and 19 245 kV c.b. were assessed. Type Air blast Minimum SF6 Total Table 1: 123 kV 245 kV Number Time Number Time 96 1952-78 10 1971-78 59 1954-82 1 1980 100 1977-98 8 1971-97 255 1952-98 19 1971-97 Main data of the assessed h.v. circuitbreakers

6.3.

Development of strategies

In the next step, the information acquired in the analysis phase was used for developing alternative maintenance strategies. In TBM strategies, the equipment are replaced after a hypothetical period of service life and maintenance is carried out continuously after specific time intervals. A maintenance budget based on this approach corresponds to an index value of 100. CBM strategies, on the other hand, prescribe replacement or maintenance measures only after an equipment has achieved a certain technical condition. This condition is specified with the help of a so called condition index. The condition index is defined differently for different equipment The scenario for a CBM strategy showed a saving potential of 40% compared to that of a TBM strategy. In this connection, it might be interesting to note that the model calculations for the application of a CBM strategy coincided exactly with the amount of the maintenance budget of the client. However, this was not enough for achieving the goal of the project. Assessing the possibilities of approaches using modern Reliability Centered Maintenance (RCM) was an even more important task. Such approaches distinguish themselves through the fact that they not only take the technical condition of the assets into consideration but also their importance in the overall network. The result of the total assessment of the 123 kV and 245 kV circuit-breakers is listed in figure 10. In addition to that, customer specific critical values for the technical condition of the equipment are defined. In the example shown in figure 10, the critical value for a maintenance measure (cM) is fixed at c = 40 whereas the critical value for a replacement or retrofitting measure (cR) is fixed at c = 60. For example according figure 10 an extremely important circuit breaker (i = 100) is in a critical position (c = 56) and must be maintained urgently. Furthermore, a general order for maintenance measures can be derived from the figure. Summarizing the RCM strategy influences the maintenance processes significantly. The optimisation of these maintenance processes offers an additional costcutting potential. In addition to that many companies conduct processes that are too complicated and have interfaces that require manual inputs, making them prone to errors.

The data, which illustrate the operation experience are listed in table 2. Type Air blast Minimum oil SF6 Total Table 2: 123 kV Years Age 3134 32,6 1908 28,9 1290 12,9 6132 24,0 245 kV Years age 251 25,1 19 19,0 95 11,9 365 19,2

Operation time of the h.v. circuitbreakers Years: years of operation Age: average age

In the second step, a thorough condition evaluation of the significant components, particularly the critical components among them, was carried out using modern procedures of diagnosis. The following measurements were carried out in this step of the project to equipment the circuit-breakers: Infrared thermography switching times SF6 humidity and, oil analyses.

The results of the analysis were evaluated and compared to reference values from the internal data bases of the utility. At the end of this phase of the project, a complete "finger-print" of all the significant equipment in the network was available.

development is targeting only the automatic incorporation of time-dependent evaluation parameters. It is decided to integrate this program into maintenance management system. In the final analysis, computerized maintenance planning in a company necessitates a shared database, to serve as a foundation for both maintenance planning and a higher-order information system for the network /8/. References /1/ Moubrary: Reliability centered Maintenance. Butterworth LTd., Oxford (1992) /2/ RCM and Diagnostics. EPRI Seminar on Reliability Centered Maintenance. Newport Beach, CA, USA (1995) /3/ Strnad, A.; Rhsler, H.; et. al.: Strategy for Condition based Maintenance and Updating of Substations. CIGRE 1996, 23-105 /4/ Balzer, G.; Brandl, M.; Strnad, A.; Rhsler, H.; Schnettler, A.: Rechnergesttzte Wartungsstrategie fr Schaltanlagen. ETG-Fachbericht 66(1997), S. 99 - 111 /5/ Balzer, G.; Brandl, M.; Strnad, A.; Rhsler, H.; Schnettler, A.: A computer-aided, reliability-centered maintenance strategy for power networks. ABB-Review 4/1997, p. 21 - 25 /6/ Balzer, G.; Strnad, A.; Schnettler, A.; Schmitt, O.: Decision Supporting Software Tool for an Optimised Maintenance Strategy. CEPSI '98, Thailand, report 3138 /7/ Balzer, G.; Strnad, A.; Schmitt, O: Maintenance Strategy (RCM) for HV Equipment and MV Substation. MatPost 99, Lyon, 18./19. Nov. 1999, H 4.2 /8/ Rhsler, H.; Strnad, A.; Schmitt, O.: Management System for Design, Construction and Maintenance in High-Voltage Networks. Cigre 2000-23

Figure 10: 7.

Assessment of 123 kV and 245 kV circuit-breakers

Conclusion

It makes sense to embed the described procedure in a financial management system and to develop a software tool. The general data and measurements of the process (e. g. substation) should fed into the program to assess the condition and the importance of the equipment. Whereas the main data of the equipment (e. g. type, manufacturer, year) are already listed in network information systems. These values have to be directly imported to the software tool. After the overall assessment a ranking of the equipment will be calculated and the responsible engineer has to handle the order via a commercial software tool. When looking at this optimized process design concept, the time needed to achieve this goal should not be underestimated. Since, however, the procedure for the basic and crucial core process of network component evaluation has already been defined, and an appropriate PC program for this purpose are already available for high voltage equipments on the market. Further