scope:

1.2 INTRODUCTION TO CODES

In most countries the control of structural safety is administered by government regulatory agencies that enforce very broad legislative provisions regarding public safety. These agencies specify various codes and standards that must be satisfied to meet these statutory requirements.

There are many reasons why structure failures occur. In the case of offshore platforms, failure can be caused by facilities malfunction, accidental damage, blowouts, environmental overload, operational errors, fabrication errors, material defects, etc. To avoid failures, conservative design and operation practices have evolved through the years that have been based on performance experience, analysis, and testing of critical components. A recent addition to design practice is reliability engineering. The purpose of reliability engineering is to assess and quantify any known risks of failure in light of prevailing practices, and available field and experimental data. The eventual aim of reliability engineering is to select safety strategies so that structures will be acceptably safe and economical.

In general design practice it is only necessary for the engineer to demonstrate that (1) the design falls within the scope of a required code and (2) that all the provisions of that code are satisfied. In so doing, the designer does not directly address the question of structure safety. By meeting the code the designer is usually presumed to have satisfied the relevant statutory requirements; the responsibility for assessing safety margins then falls directly on the code writers.

Anderson (7) has raised the question of what happens if there is no code appropriate to the design, or that the code provisions cannot be reasonably applied to a particular structure? In a great many cases in the offshore petroleum industry, the use of standard codes is not appropriate, and the engineer must directly and explicitly address the questions of safety and standards expected of a prudent offshore operator. These questions can become a major factor in certification, and the certifying authority may require justification of certain engineering decisions. In most cases, the engineering problem is cast in terms of the closest code provisions, and a good deal of judgment is used in their application by both the designer and the certifying authority. In offshore engineering, for example, the environmental loadings needed therein are not in the scope of most standard building codes. For that reason, the API developed a recommended practice (RP 2A), not a code, to provide guidelines that consolidate the judgment and experience of industry engineers. At present this recommended practice is not explicitly reliability based; however, an alternative practice that is reliability based has been developed and should be available for public comment in mid-1989. These notes attempt to clarify and explain the provisions of this document as well as instruct the reader on its use.

The purpose of the reliability approach is to provide a rational means of assessing safety in engineering terms and assist in code development. In fact, codes like the American Concrete Institute (ACI), American Institute of Steel Construction (AISC), the Canadian Building codes, and most foreign codes have recently been reformulated in a reliability based format. The format chosen allows "factors of safety" to be specified by the code committees in terms of the uncertainties in the existing resistance and load predictions and the desired target reliability. For example, codes may now apply different target reliabilities for serviceability design versus limit state component failure design. Some codes may also explicitly or implicitly account for structural system reliability, in addition to the reliability of individual components.

Unfortunately, it is currently impossible to assess the reliability of platforms or any other structure in an absolute actuarial sense. For example, "this platform has a 1.2 × 10−3 chance of failure during 20 years." These types of statements are made, but they may be orders of magnitude apart from reality. There are simply too many unknown factors that affect reliability; each of which may have a large influence on the final result. At its present state of development, reliability engineering can only address a subset of disciplines perceived to be the most relevant to the designer and, therefore, those for which the most information has been developed.

Specific applications of reliability methods to offshore platforms have been carried out by various offshore operators (7-12). However, these are not routine calculations and the results may vary widely among the various analysts. It suffices to say that, at their present stage of development, reliability results cast in a probabilistic sense are not suitable for interactions with regulatory bodies. That is, one cannot say offshore structure risks are greater, or less than comparable industrial exposures or other public exposures such as automobile driving or mountain climbing. At present, the main use of reliability engineering is limited to two areas. The first is code oriented calibration and safety index formatting. This means the application of reliability methods in specification oriented research and development, rather than directly in routine design calculation. The second area is in decision strategies; for example, in comparing alternative concepts for a particular design or in evaluating remedial strategies for existing systems. The reliability model is often a good framework for aiding this decision process.