Transmission Line Insulation

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

15-101

Session 2002 © CIGRÉ

DATA BASE FOR 330 - 750 kV TRANSMISSION LINE INSULATION: STRUCTURE, DATA MINING, USE FOR THE PURPOSE OF IMPROVEMENT OF LINE OPERATION A. A. Smirnov, P. I. Romanov∗ North–western Intersystem Power Grid of Russia (Russia)

SUMMARY An implementation of modern information technologies to data collecting and processing at all the Intersystem Power Grid (IPG) management levels is illustrated by taking 330 - 750 kV overhead transmission line (OHL) insulation as an example. To quantify the OHL operation and maintenance comprehensive statistical data are collected and arranged in a common database (DB). The DB contains joint description of electrical and mechanical stresses imposed on the cap insulators. It grounds construction of an empirical distribution function of insulator failures for long service life period. Comprehensive information monitoring of the OHL permits to develop quantitative criteria of the OHL survival, durability, and maintainability absent in practice. Using pessimistic, as well as optimistic predictions based on these criteria creates a foundation for strategic planning at the federal level. Information system of OHL maintenance support makes it possible to make management decisions taking into account risk level, to plan scheduled maintenance, effectively allocate investments, and as a result to optimise OHL maintenance expenditures. KEYWORDS Insulator - Overhead Transmission Line - Maintenance management - Failure - Data Collection - Database 1. INTRODUCTION A suspended cap line insulator is the most numerous and the most open to injury OHL element exposed to extreme mechanical loads and electrical stresses.

System of collection and processing data on these elements has been worked out and accepted about 30-50 years ago and has become obsolete. As a result of changing economic environment for electric utilities, one of the management’s new goals is to improve informational support of the IPG maintenance. Implementation of modern information technologies to data collecting and processing at all the IPG management levels is illustrated by taking 330 750 kV OHL insulation as an example. To quantify the OHL operational capacity indicators comprehensive statistical data are collected and arranged in a common DB. This way DB becomes a management system element. Directing OHL maintenance activity is always based on a management decision. At the level of a single power company a transmission line is regularly inspected to collect updated information about insulator failures or OHL breakdown situation occurrence. Management’s aim at the local level is to plan and organise replacement of failed or damaged insulators on the specific OHL, the specific tower, and the specific string in specific terms. There is no scheduled maintenance on insulators and their expected life span exceeds 25 years [1]. Regular in-service inspections and just in time replacement of failed or damaged insulators is required to ensure failure-free operation of insulator suspensions. At the regional level more general management decisions are based on quantitative and qualitative criteria of the OHL element condition assessment, criteria of maintenance expenditures, and combined technical and economic criteria. At this management level the information received from all power

______________________________________________________________________ ∗ 1/39, Kurchatova Str., P.O. Box 88, 194223, St.Petersburg, Russia

companies is processed and convoluted. This information becomes available and appropriate for the analysis to make strategic decisions at the federal management level. A transition to a new OHL maintenance management system is realised by gradual implementation of new information technologies. Use of bundled soft- and hardware will enable a considerable increase in the amount of initial information and the rate of data processing. It admits putting statistical methods of data processing into practice, as well as employing criteria and estimates based on probability indicators. In longterm perspective it is possible to use these probability indicators for assessment of risk of line failures with regard for monetary losses. 2. PROBLEM STATING AND DB OBJECTIVES OHL pertaining to the North-western IPG of Russia have around 2,250,000 line glass cap insulators comprising suspension (67,000) and tension (27,000) insulator strings. More then 5,000 km of 300 - 750 kV OHL and 16 substations are operated by the North-western IPG and deliver power from nuclear and high-power thermal stations. Geographically the components of the power grid are located within the range of latitude from 53Ο N to 68Ο N. For the region wind speed design values corresponding to the average return period of 25 years vary from 25 to 32 m/s, while ice thickness variations are confined to the interval from 10 to 20 mm. The major part of OHL run in sparsely populated terrain, up to 40 % marshy and difficult to access. The rate of thunderstorms is about 10 - 30 hours per year. Until now data on insulator conditions have been collected by in-service inspections regularly once a year. Surveys of every OHL detect an amount of failed or damaged insulators and their distribution over suspension strings, phases and towers. The data are written into a survey register and then are summarised in the OHL damage register stored in a transmission line division. The OHL maintenance standard rate specifies replacements of failed or damaged insulators. When four failed insulators in a one string for 330 kV OHL are detected, replacement should be performed within a year. Six failed insulators in a one string are to be substituted during a month. However, regulations did not specify an acceptable total amount of failed insulators on the OHL, and also acceptable terms of replacement for one or two failed insulators in a string. The cases considered need the development and approval of criteria for assessment of risk of failed insulator operation.

At the intuitive level it is obvious that even a single failed insulator which is present in a string for a long time can lead to a string failure. Large share of failed insulators on an OHL also reduces reliability of a line. Current standards do not specify engineering or economical methods of risk assessment with the regard for running insulator string of reduced reliability, as well as cost of renovation to reach an acceptable risk level. The problem looks even more important, for 20 % of the OHL in IPG are in service for more then 30 years, age of 50% of the OHL exceeds 20 years. It is evident, that insulator reliability can be determined depending on its location within OHL and taking into account their interconnections with all other line elements. Therefore, information on insulators should be integrated with all other data on the OHL design and condition, mechanical and electric exposures, and air pollution influence on the insulator condition. Accordingly, the data bank designing objective consists in making up an inventory of OHL all components and their functional indicators. This objective is concerned with the necessity of a considerable increase in the amount of information to be processed, its storage in a classified form for obtaining operation indicators needed for solving maintenance problems at three levels of: • operation; • research; • management. To solve operation problems it is necessary to have available and detailed information on the OHL design, elements, and parameters, all their existed modifications, and to have possibility to retrieve any selected information in an appropriate form. Paper medium that is in use at present is not suitable for proper system analysis. In the field of research the task is a development of probabilistic based maintenance parameters, convenient for use in practice, as well as development, testing, and approval of such criteria, that will allow comparison of operation and maintenance quality for the OHL as a whole, and their elements. The main management problem is a transition from employing qualitative estimates and practice of topmanagement decision-making to quantitative estimates based on the probabilistic based assessment of risk of power system element failure. 3. DB STRUCTURE AND INPUT INFORMATION COLLECTION Attempts to copy paper based information on PCs resulted in understanding the possible way of using modern information technologies: GIS, databases, the

Airborne Laser Terrain Mapping system (ALTM) [2], and also problem-tailored software. At present the DB includes: • OHL topographic model, insulators and other OHL element technical performances; • physical composition of the line and the components; • results of scheduled maintenance; • results of emergency maintenance, upgrading and reconstruction; • expenses of every kind of maintenance, upgrading and reconstruction; • OHL operation history (outages, failures, breakdowns with their origin causes analysed, electrical and mechanical stresses, etc.) • data processing software; • results of data statistical treatment; With the OHL monitoring informational system organisation the data received from different power production divisions and specialised organisations in a unified presentation form have been put in the DB. At the design and maintenance stages the presentation form of documentation is reviewed. For the long-term operated OHL technical documentation is updated according to current OHL changes or created again on electronic media. When failed insulators are replaced, its type, exact location, production data, and producer of old and new insulators are registered. The approach gives the opportunity to stack statistical data on insulator reliability and analyse the failure flow dynamics. In DB separate block electrical performances of the OHL availability are continuously registered, besides temporal characteristics of changes in the OHL operating regime: duration of OHL availability, duration of scheduled and breakdown maintenance, and outages. These data is supplemented with information on all OHL failures and lightning strokes. Noting that line imperfection locations are possible to determine in about 60 % of cases, this kind of electrical stress is identified with high precision. With OHL failure initiation short circuit currents characterising electrical loads on the insulators are registered. To refine an electrical stress area the information about estimated and actual line damage position is added to the controlled quantities. Collected measurement results of the tower eathing resistance help to specialise factor in an exponential model of electrical stress decay along the OHL. Documentation is depicted so that after getting the estimated line damage position the necessary information on this section is presented in evident visual form. Optimal route for the crew to reach the line damage position is highlighted on a display map.

For 330 - 750 kV OHL with operation term over 30 years techniques for all line elements condition assessment are developed. Technical documentation is presented in two ways: on paper and electronic media. It was in the process of this work when new specific requirements to approach for documentation arrangement and research execution method became explicit. Implemented technique of laser scanning [2] allows getting more detailed information about the existing OHL condition in the surrounding terrain. The information collected is presented in digital form and makes it possible to measure the OHL geometrical parameters and create a 3D digital line model. Using this data, the conductor and insulator mechanical loads can be calculated. Clearance measurements allows to estimate transmission line capacity and find power resources. To reduce risk and possible damages in obstructive conditions and in places of OHL intersection with engineering constructions checking geodetic mapping is made. In a separate DB block expenditure of time and assets related to location of the OHL damage identification is registered. These data analysis will increase validity of management decisions on maintenance co-ordination of all power company divisions. 4. OHL INSULATOR MAINTENANCE ANALYSIS, FORECAST, AND PLANNING OHL monitoring information system is used for making on-line decisions, but for strategic management decisions it is necessary to add OHL condition trend forecast. Only these both parts can be laid as a foundation of development of measures aimed at power utility management improvement. Fig. 1 depicts interconnection between tactical and strategic information in making management decisions. The OHL element condition assessment is performed at the basis of checking their residual strength against real loads taking into account corrosion wear and mechanical faults. Test calculations of elements of OHL with service life over 30 years give an opportunity to find out strength resources reserved during the OHL design. They can be considered determining basic design criterion according to new standards adopted during this period. Using of such an analysis increases planning efficiency of the OHL scheduled maintenance and reconstruction capacity: according to their technical condition and forecast all OHL are divided into groups. Reconstruction appropriateness, as well as terms and capacities of scheduled maintenance are set for every group. Identification of weak links permits to effectively

allocate investments maintenance [3].

and

time

for

the

OHL

every power company, for every range of voltage (330, 400, 500, 750 kV), insulator types, tension and suspension strings. At the same time for each of the

Informational System

Management

Data Collection

Actual OHL State Assessment Management Decision

Forecast

Predicted OHL State Assessment

Figure 1. Model for management of OHL insulator maintenance Span-wise inspection of vegetation cutover patch in adjacent area reduces risk of short circuit related to tree falls, allows to get on-line information, to supervise and to plan very expensive operations. Transmission line divisions for two years conduct an attempt of implementation of such approach. To assess actual insulation condition all kind of stresses on every string is described in a form of deterministic parameters used in OHL design. To describe insulator failure process including emergency situations a probabilistic phenomenological model of injury accumulation is employed. Tracking time dependence of single insulator failure allows getting an empirical distribution function of accumulated injuries for every insulator of all inspected lines. Well-known shape of reliability dependence on lifetime is a curve consisting of three zones. The first one, being nearly linear increase, corresponds to initial period of operation when evidently fault elements quit out. The second zone characterises the period of stable operation. The third zone corresponds to the beginning of the period when according to statistical rules complex implicit and explicit stresses result in avalanche-like increase in failures. The main prognostic task is to forecast the limits of transition from the second zone to the third one. Evidently, only such a forecast allows reserving time for making strategic management decisions. Only implementation of information technologies in DB gives the possibility to obtain the insulation behaviour for several ensembles: for every OHL, for

ensembles stress loading coefficients (electrical, mechanical, and air pollution), history of loading, climate conditions, tower design, OHL design, OHL responsibility in the IPG are monitored. The DB containing this set of measurable parameters permits to instantly assign any given situation to a type of distributions and to determine crucial moment of time for making a decision. Obviously, it is the prediction that enables to create a delay for strategic management decisions. The DB implementation to practice joins the necessity of operating division staff training, as well as training of management personnel. At the same time a system of activity is needed for developing probabilistic based management criteria with their following approval. 5. CONCLUSIONS An information system is developed for running the maintenance documentation depicting the 330-750 kV OHL isolation life cycle. Modern techniques of data collection and processing are applied at two different levels the North-western regional division of the Federal Power Grid Company of Russia and a single power company. This information is available and appropriate for the analysis at the federal level to make strategic decisions. Developed approach to organisation of the information monitoring of the OHL line insulation enables to analyse, forecast, and plan activity of power utilities,

company divisions, and to considerably raise the level of the OHL maintenance supervision. To describe insulator life cycle all available data on the OHL element electrical and mechanical strengths and stresses are collected in addition to detailed information on insulator type, age and location. Data received in a unified presentation form from different power company functional divisions and research and development institutions are integrated into the information system. In the field of OHL maintenance an inventory of time consumption related to every kind of maintenance of the OHL as a whole and line insulation is arranged. Preliminary patterns of quantitatively evaluated time expenditures related to every kind of maintenance of the OHL as a whole and line insulation are determined in more detail then before. These results have allowed raising a question of reducing maintenance costs only as a consequence of activity co-ordination between the system operation and transmission line divisions. New form of information collection, storage, and processing is being used. Transmission line division personnel have already approved forms used for collection and presentation of data. Now these forms are being improving by transmission line divisions in collaboration with research institution. Complete and representative data on insulator electrical and mechanical strengths and stresses imposed, as well as influence of air pollution create the necessary prerequisites for a statistical treatment of their life cycle. Two timely statistical problems are connected to a large share of insulators. The first one implies determining the dependence of the probability of an insulator string survival on the service time of one or two failed insulator. The second problem concerns the determination of the probability of the OHL failure dependence on the total amount of failed insulators in a line. To solve these problems the DB is created and the data collected over long period of time are statistically treated.

DB allows determining empirical distributions of the periods of OHL availability, scheduled and breakdown maintenance, and outages, classified according to causes of their origin. The probabilistic indicators of the OHL availability will be a basis for making management decisions. Comprehensive information monitoring of the OHL permits to develop quantitative criteria of the OHL survival, durability, and maintainability absent in practice. Using pessimistic, as well as optimistic predictions based on these criteria creates a foundation for strategic planning at the federal level. Information system of OHL maintenance support gives the possibility to make management decisions taking into account risk level, to plan scheduled maintenance, effectively allocate investments, and as a result to optimise OHL maintenance expenditures. REFERENCES [1]

GOST 27661-88. Line Suspension Cap Insulators. Types, Parameters and Sizes. - Moscow, 1988 [2] B. I. Mekhanoshin, Ye. M. Medvedev, L. Yu. Rivin, V. A. Shkaptsov, Ye.A.Yavorsky, Yu. A. Dementyev , R. Urwin, Use of an Airborne Laser Locator to Improve Availability and Quality of Maintenance of Overhead Lines, (CIGRE-2000, Rep 22-204). [3] A. A. Smirnov. Criterion for Making Decisions on Capacity of Overhead Transmission Line Element Maintenance and Reconstruction. (Perspective directions in development of electrical power industry and electrical engineering equipment in 2000–2010, 2000, 1-34, pp. 116-118).