A Historical Review of Pre-Commissioning Hydrotest Failures

The concept of proof testing engineering structures has its origins in antiquity. The pre-commissioning hydrostatic test (also known as the pre-service pressure test) has been an important part of the process of commissioning a newly constructed pipeline for over 50 years, since its beginnings in the 1950s in the USA. The purpose of the hydrotest is several-fold: to prove the leak tightness of the pipeline system at a pressure above the design pressure, as a strength (proof) test to identify (fail) defects and sub-standard pipe, and to prove a safety margin above the pipeline design pressure. Historical data, from PARLOC (Pipeline and Riser Loss of Containment), the OPS (Office of Pipeline Safety) 30 day Incident Reports, and the published literature on the number and causes of pre-commissioning hydrotest failures has been reviewed. The historical data covers onshore gas transmission pipelines in the USA and the UK, and gas and liquid pipelines in the North Sea. The data covers the period from 1952 to 2005, although there are significant gaps in the data (e.g. the OPS data for the USA does not report test failures after 1984). In this paper, the historical data is summarised over this period, by year, in terms of the number of failures per km, and trends in the frequency and type of failures are identified. Comparison of USA and UK experience, or onshore and offshore experience, is contentious because of the influences of different design codes, and local custom and practice. The USA and UK pipeline design code requirements for the hydrotest are summarised in the paper, and it is shown that some of the trends in the failure data may be explained by the differences between the codes. Failures during the hydrotest are rare, but occasionally they do occur. The general consensus is that failures during the precommissioning hydrostatic test are now less common, and that failures due to defective line pipe (rather than due to leaking fittings) are rare. The historical data supports this consensus, but it also highlights that it is largely based on anecdotal evidence rather than data and analysis, because information on test failures is not now routinely gathered and published. The results of the historical review demonstrate that understanding the causes and reasons for hydrotest failures is important for learning from past mistakes, and also for identifying those cases where it may be possible to dispense with a pre-commissioning hydrotest. Reliable historical data on hydrotest failures is necessary to quantify trends over time, and to understand the causes of failures. The pipeline industry as a whole is not coherently recording this data. It should be.