Compute Reliability Indices Research Articles

Reliability-Based Code Calibration for Design of Wood Members Using Load and Resistance Factor Design

The procedures underlying the load and resistance factor design (LRFD) format for engineered wood construction combine elements of reliability analysis with the experience gained from decades of successful use of allowable stress design procedures. However, the judgments applied in much of the early development or LRFD hinged directly on the load factors and underlying load distribution assumptions in the 1993 version of the ASCE 7 load standard. Since that time, load factors in ASCE 7 have changed several times and additional information is available regarding load distributions. Assumptions regarding resistance statistics also have evolved during the past decade. For example, early reliability analyses were based on resistance distributions derived from individual data sets. Unfortunately, this approach is confounded by often significant differences that occur between data sets. This approach also neglects the self-calibrating nature of engineered wood product specifications. This paper examines the range of computed reliability indices that result from updating the assumed load statistics and assumed resistance statistics. Several load cases involving dead, live, snow, and wind load are examined. The reference structural configuration is a simple flexural member. The results show that the range of computed reliability indices is most significantly influenced by regional variations in load statistics. While these results also show slight differences in computed reliability indices from one engineered wood product to another, these differences are not judged to be significant. The results of this study will not only assist the Structural Engineering Institute Design of Engineered Wood Construction Standards Committee in their upcoming revision of ASCE 16, but can also serve as a template for other material standards committees (steel, concrete) as they update their LRFD specifications to reflect evolutionary changes in ASCE 7.

An efficient decomposition approach for multi-area production costing

This paper presents a new approach for multi-area production costing called SEG-GSIDES (segmented global simultaneous decomposition-simulation). It is based on the SIDES (simultaneous decomposition-simulation) approach which has been developed for computing reliability indices in a multi-area configuration. The global state space is obtained from the maximum number of units committed in each area and thus contains the state space corresponding to any set of committed units. In the GSIDES approach, the global state space is decomposed and the resulting sets are used to compute the required indices for any subset of units committed. The energy supplied by each unit is calculated by the difference between the expected unserved energy before and after unit commitment. In some cases it may be possible to obtain indices by decomposition alone but generally a decomposition phase is followed by a simulation phase. When the units committed are close to base loaded units, their contribution to unserved energy from the decomposition phase is low. The major contribution in this case comes from the simulation phase and this introduces an approximation due to convergence. To reduce this approximation the decomposition is performedsequentially on different segments of the global state space. In this way, the major part of the indices is obtained from the decomposition phase and this leads to a more accurate determination of energy supplied by the units. The concepts are illustrated using a small example system. Case studies are presented for the three area and four area IEEE-RTS systems.

A probabilistic method for establishing the transformer capacity in substations based on the loss of transformer expected life

A method is presented for computing reliability indices, probability, duration frequency, and the loss of transformer expected life. The probabilistic model takes into account the variables affecting the loss of transformer expected life, daily load curve, transformer forced outage rate, outage duration, occurrence time of forced outage and expected overload duration. Comparing the results of this model with those of Monte Carlo simulation an acceptable degree of accuracy is achieved. From the reliability indices, a reliability criterion is obtained for rating transformers in substations, which have traditionally been rated on the basis of conservative deterministic criteria.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Reliability Calculation under Planned Maintenance

This paper provides an extension to methods for computing reliability indices in the case of planned maintenance, in particular, a new formula for the frequency of system failure due to overlapping of two outages, forced and planned, is derived for two components. A numerical example illustrates the possibilities of the proposed approach.

A Probabilistic Method for Establishing the Transformer Capacity Reserve Criteria for HV/MV Substations

This paper provides an extension of methods for computing reliability indices for substations which takes into account the permissible overload of the power transformer, the daily load cycle and emergency loading. Based on the reliability indices derived in this paper it is possible to determine a reliability criterion for the rating of the substation, the number of transformers and the reserve factor required, which have traditionally been based on conservative deterministic criteria.

Reliability of axially loaded columns

The reliability of current and proposed design formulae for axially loaded columns is determined by the first-order second-moment probabilistic method. Computed reliability indices are compared to target values, and the consequences of changing the resistance factor are examined. The basic behavioural model includes the effects of residual stress, initial crookedness and small end-restraint.

Extensions to the bounded-set theoretical approach to reliability modelling and calculations for composite generation-transmission systems using a DC load-flow model

A bounded-set theoretical approach is given for evaluating the loss of load probability and frequency of failure of a composite generation-transmission power system represented by a DC load-flow model. The computed reliability indices are in general upper bounds to the corresponding set of reliability indices obtained by using a linear network flow modelling of the composite power system. A specialized parametric linear programming method was also developed for implementation of the set decomposition solution strategy.