Revival: Safety and Reliability in the 90s (1990)

Will past experience or prediction meet our needs?

M.H. Walter, R.F. Cox

• Publisher: Chapman and Hall/CRC
• Year: 2019
• Print ISBN:
  9781138105478
• eBook ISBN:
  9781351357920
• Topic: Business
• Subtopic: Agribusiness
• Index: Business

Reliability-based design is relatively well established in structural design. Its use is less mature in geotechnical design, but there is a steady progression towards reliability-based design as seen in the inclusion of a new Annex D on "Reliability of Geotechnical Structures" in the third edition of ISO 2394. Reliability-based design can be viewed as a simplified form of risk-based design where different consequences of failure are implicitly covered by the adoption of different target reliability indices. Explicit risk management methodologies are required for large geotechnical systems where soil and loading conditions are too varied to be conveniently slotted into a few reliability classes (typically three) and an associated simple discrete tier of target reliability indices.

RELIABILITY ANALYSIS OF THE INDUSTRIAL ELECTRICITY SUPPLY PART I: OUTLINE OF THE PROBLEM.

M.H.J. BOLLEN AND P. MASSEE

Eindhoven University of technology, Faculty of Electrical Engineering P.O. Box 513, 5600 MB Eindhoven, The Netherlands.

ABSTRACT

This paper concerns the industry-owned part of the electric power supply to an industrial load, in short: the industrial electricity supply. In the reliability assessment of the industrial electricity supply we can distinguish four main areas: finding a suitable description for the behaviour of network components; calculating the (stochastic) system behaviour from the known component behaviour; describing the (deterministic) system behaviour during a disturbance; and linking the behaviour of the protection to the system reliability.

After a discussion on the reliability problem of the electricity supply the four main problem areas are discussed in detail.

INTRODUCTION

The high interruption costs associated with modern industrial plants calls for a highly reliable electricity supply. The general design problem can be described as follows: given a connection to the public supply grid and one or more industrial loads (both with their own demands), find the best electrical network to supply energy to the loads. The “best” simply means the cheapest: i.e. the total costs during the economic life-time of the network should be as low as possible. The total costs consist of a deterministic term: the building and exploitation costs that increase with increasing reliability, and a stochastic term: the interruption costs that decrease with increasing reliability. This optimization should take place within the boundary conditions set by the public supply, by the industrial load, by the security and by rules and regulations.

The most important difference between the industrial and the public electricity supply is the economic value of the reliability. The owner of the industrial electricity supply is also the consumer of the electric energy. The building and exploitation costs thus come from the same purse as the interruption costs. This makes the above reliability optimization possible. In the public grid only a small part of the interruption costs come at the expense of the owner, making such an optimization of little value.

Some technical differences between an industrial electricity grid and the average public grid are:

- the large concentration of load in the industrial grid, which leads to short connections;

- the highly predictable time behaviour of the industrial load;

- the large fraction of an industrial load which consists of electrical machines;

- an industrial grid mainly consists of underground cables whereas the majority of public grids (especially the ones being studied) consists of overhead lines.

All this calls for a reliability assessment directed especially towards the industrial electricity supply. Most of the techniques that will be used and developed can be used for the analysis of public distribution grids as well, and, to a lesser degree, also for high-voltage grids.

A reliability analysis has been performed quite often for extended high-voltage grids, where is has been determined to what extent the power stations can generate the energy needed by the users and to what extent the high-voltage grid is able to transport this energy [e.g. 1,2,3]. Much less studies have been performed into the reliability of public distribution grids [e.g. 4]. Only a few studies have been performed into the reliability of the industrial electricity supply [5,6,7,8].

THE RELIABILITY PROBLEM

Figure 1 shows the different aspects related to the reliability of the industrial electricity supply. During the reliability assessment of an (industrial) electricity supply all aspects of Figure 1 should be taken into account.

Network state. The probability of the occurence of a disturbance and the probability that this disturbance leads to an interruption are related to the network state: the electrical loading; the physical loading and the availability.

The electrical loading is the amount of energy being transported from the public supply grid to the industrial load. As an example we distinguish between no-feed (there is no connection to the public grid), no-load (there is a connection to the public grid yet no transport of energy to the load), loaded (there is transport to the load) and the transitions between loaded and not loaded (switching on and switching off of the load). An industrial load often shows an intermittent behaviour (periods of high and low power demands alternate in a regular pattern). In that case a distinction can be made between no-load, half-load, full-load and their transitions.

The physical loading indicates all external phenomena that might influence the chance of failure of the electricity supply. One can think about weather influences (rain, storm, lightning, very dry periods), mechanical activity (digging in the ground) and chemical influence (Two extreme examples: desert sand that sticks to insulators in electrical installations in Oman; unknown chemical compounds that affect the electrical parts of a waste combustion furnace).

The availability indicates which components are available for the transport of electrical energy. If one or more components are not available, the remaining components will, in general, be subjected to a higher loading. The probability of interruption given a disturbance is determined by the availability too. At each moment the availability is determined by the maintenance that is performed at that moment: preventive maintenance as well as repair (corrective maintenance).

Threats. Everything that threatens the interruption-free operation of the electricity supply is described by the term “threats”. A distinction can be made between ageing (an increase of the momentary failure rate with increasing age), transient overload (a short overvoltage or overcurrent due to a switching operation, a lightning stroke or a short circuit), temporary overload (a long-lasting overcurrent or overvoltage e.g. due to one cable being out of operation), and damage (e.g. of the insulation of a cable). The appearence of these threats is influenced by many other events and aspects. Figure 1 shows some of them.

The degree of ageing is influenced by the (normal) electrical loading, the frequency and severity of transient and temporary overloads, the physical loading and by quantity and quality of the preventive maintenance. Quite general it can be stated that increasing the loading will lead to an increased ageing. No such general relation is available for preventive maintenance, although preventive maintenance is intended to decrease the failure rate.

Transient overload occurs or may occur due to a short circuit, due to transitions in the normal electrical loading (from no-load to half-load to full-load and back), due to an intervention of the protection (this can lead to overvoltages but it reduces the severity of overcurrents), due to a component being taken into operation after repair or after preventive maintenance, and due to a component being taken out of operation for preventive maintenance. Transient overvoltages can further enter the industrial network from the public supply grid.

Temporary overload will generally not occur during normal operation. The design will be such that the highest normal loading is not an overload. Due to extreme weather (a long-lasting period of drougth) or chemical influence (the before mentioned Arabian desert sand; or Dutch sea water) the permitted loading can go down. The dry weather in combination with a high ambient temperature prevents a cable from loosing its heat. As the temperature should not exceed a certain upper level, the maximum permitted current shrinks. In a simular way the maximum permitted voltage shrinks due to the influence of the sea water.

A temporary overload that will probably appear more often is related to the availability. In case one or more network components are not available the others will suffer a higher loading. Whether this is considered an overload depends on the dimensioning of the components and the (“arbitrary”) boundary between no-overload and overload.

A temporary overload can also be due to a change in voltage of the public grid. An increase in voltage might cause a temporary overvoltage, a decrease an overcurrent (the electrical power demand of the industrial load will be nearly constant).

Damage can occur due to physical loading, but also during repair or preventive maintenance. Preventive maintenance can also repair the damage.

Faults. Further on we will distinguish between disturbances and interruptions (sometimes referred to as interruption of electricity supply or interruption of plant operation). A disturbance is any deviation from the normal state. A disturbance can be “on purpose” (e.g. taking a component out of operation for preventive maintenance) or “by accident” (a short-circuit, an overload, or an incorrect intervention by a protective device).

Figure 1: Visual summary of the different aspects of the reliability of the electricity supply.

The disturbances can be divided into faults and non-faults. A fault is a situation that is so dangerous to the components and/or to the electricity supply somewhere else, that we are willing to disconnect one or more components, even if this causes an interruption of the electricity supply. The best-known example is the short circuit. The currents appearing during a short circuit are such high that components would suffer considerable damage if no intervention would occur. As normal electricity supply is impossible during a short circuit, no-intervention would lead to an interruption for shure. The result of an intervention will not be worst than no-intervention.

The weighing is more difficult in case of an overload. The electricity supply will not be (directly) influenced by the overload, the component c...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Preface
• Table of Contents
• List of Contributors
• Operability Studies and Hazard Analysis in the Management of Safety
• Strengthening the Link Between Management Factors and Quantified Risk Assessment
• “SURVIVE”: A Safety Analysis Method for a Survey of Rule Violation Incentives and Effects
• Wise Men Learn by Others Harms, Fools by Their Own: Organisational Barriers to Learning the Lessons from Major Accidents
• Organisational, Management and Human Factors in Quantified Risk Assessment: A Theoretical and Empirical Basis for Modification of Risk Estimates
• Human Factors Technologies for Safety and Reliability in the ’90s
• A Methodology for Applying Traditional Safety Analysis Techniques to Systems Handling Crowd Flows
• The Development of a Human Factors Strategy for the Design and Assessment of Plant and Equipment
• A Resources-Flexible Approach to Human Reliability Assessment for PRA
• Assessing the Safety of Existing Plants - a Case Study
• Operability Study Adaptations in Selected Field Work Areas
• Some Aspects of the Use of Quantified Risk Assessment for Decision-Making
• The Application of Root Cause Analysis to Incident Investigation to Reduce the Frequency of Major Accidents
• Management of Safety through Lessons from Case Histories
• Aircraft Crash Studies in AEA Technology
• The Selection of Field Component Reliability Data for Use in Nuclear Safety Studies
• Process Visibility - the Key to Software Reliability Assessment
• Some Results of the Alvey Software Reliability Modelling Project
• The Acceptance of Software Quality as a Contributor to Safety
• Enhanced Partial Beta Factor Method for Quantifying Dependent Failures
• A Bayesian Analysis of Pollution Incidents through Time
• An Expert System for Integrating the Physical Modelling and Reliability Assessment of Engineering Systems
• HARRIS - a PC-Based Data System to Support Risk and Reliability Studies
• Reliability Analysis of the Industrial Electricity Supply Part I: Outline of the Problem
• Reliability Analysis of the Industrial Electricity Supply Part II: Two Examples
• Probabilistic Sensitivity Analysis of Materials for Structural Fatigue and Fracture