DVGW Technology Report No. 3/10
Gas installations need to be constantly maintained to ensure their safe and reliable operation and availability. DVGW-Code of Practise G 495 describes how to maintain gas installations and presents different strategies that take into account the operating conditions and loads acting on the individual gas installations. The maintenance intervals thus defined are not, however, fully derived from the latest operational data, which would require capturing all data automatically. This, in turn, is possible only if a gas installation is fitted with the requisite metrology equipment. A DVGW-sponsored research project helped to identify the essential basic information required for implementing condition-based maintenance (CBM) of gas installations with a focus on capturing actual installation condition data as a basis for determining maintenance intervals.The availability and operating safety of gas installations is mainly predicated on maintenance. A custom-designed maintenance strategy that matches the operation and condition of the respective installation will help cut costs and ensure system availability and operating safety at the same time.
The DVGW Code of Practise - a basic document
DVGW Code of Practise G 495 of July 2006 was the first paper to include condition-based maintenance. This Code of Practise focuses on capturing and analysing relevant operational data and on target-performance comparisons on the basis of a new assessment system and the definition of assessment factors (AF) which, in turn, determine maintenance intervals.
While the determination of assessment factors as suggested in Code of Practise G 495 does take into account some key variables like e. g. installation age and mode of operation as well as operating experience, it fails to fully include into the assessment the actual condition of a gas installation.
However, a more comprehensive implementation of CBM requires determining the actual condition of the relevant components and assemblies on the basis of automatically captured and analysed operational data, if possible. The DVGW research project that has just been concluded has prepared the basis for exactly this.
Company survey on CBM
The DVGW subsidiary DBI Gas- und Umwelttechnik published a relevant study that collected data on the failure distribution in gas measurement and pressure reduction stations in order to identify the factors that are relevant for the determination of maintenance intervals. The project pursued the objective of defining those components of gas measurement and pressure reduction stations that are particularly relevant for maintenance, and aimed to devise new approaches to component-specific condition assessment.
The survey began by interviewing a number of company representatives so as to gain an overview of current CBM practises in Germany. It emerged that all interviewees considered the currently observed maintenance intervals appropriate and/or conservative, deeming safety features and the components essential for the functioning of the installations such as, for instance, filters/separators, pre-heaters, flow metres and gas pressure regulators especially relevant for CBM.
The failure data of gas measurement and pressure reduction stations were analysed in order to determine the most failure-prone installation components that are relevant for CBM. The data proved that most failures occurred in heat generators and odorisers, i.e. secondary structures, electrical failures and leaks being the most frequent failures.
New assessment criteria
Current CBM as per Code of Practise G 495 is carried out on the basis of installation assessment, statistical quality control of individual components and assemblies and a failure rate analysis to define maintenance intervals. This requires assessing the condition of the installation regularly so as to determine and, if necessary, adapt the assessment factor AF.
A system of scores was created in the course of the project, which helps to determine an assessment factor on the basis of the "key variables" and "safety and function-relevant components" criteria. While this procedure is based on Code of Practise G 495, it has been amended to comprise additional criteria like the condition changes of safety and function-relevant components, and the method used to establish their condition (see diagram).
The assessment factor as a measure of the condition of a gas installation
In order to be able to adequately take into account the different installation sizes, from cabinet-type installations to inlet stations, gas installations were in a first step classified on the basis of design volume flow and pressure ratio data. This ensures that operators are encouraged to equip large installations with comprehensive metrology and monitoring equipment, while preventing them from disproportionately neglecting smaller installations with rather scant metrological equipment. In determining the total points scored, the installation size is reflected by the different weighting of the three assessment groups (i.e., key variables, safety and function-relevant components).
The assessment factor AF of an installation is then determined on the basis of the system of scores. The AF reflects the condition of an installation and its expected change. As Annex 3 of DVGW Code of Practise G 495 shows, points are awarded for the different characteristics of the key variables (e. g., actual age of the installation, significance ...) or, in other words, for data reflecting the actual situation. Beyond this, points are awarded also for the condition and/or expected condition change of safety and function-relevant components on the one hand, and for the method of establishing the condition on the other.
As far as safety-relevant components are concerned, only the component in the worst condition is considered in the assessment as this is the one that determines the safety of the installation as a whole: e. g. a change in adhesion value or a slam shut valve leakage rate increase gets 100 PA (percentage assessment points) (cf. Table 1); followed by awarding percentage assessment points for the method of establishing the condition, e. g. the criterion gets 25 PA if measurement values are acquired digitally during manual testing with a subsequent evaluation and trend analysis (Table 2).
The mean of the two criteria of “condition” and “establishment of condition” is then formed so that the safety-relevant components consequently score 62.5 PA of the total points scored by the group when determining the assessment factor AF.
Metrological equipment of smaller installations
This assessment procedure is likewise applied to function-relevant components, all of which are assessed, however (i.e. filters/separators, pre-heaters, flow metres, gas pressure regulators).
In order to determine the total points scored and the resulting assessment factor, the percentage assessment points of each group are multiplied by the respective highest number of total points scored. The score key applying to the installation classes takes into account that small installations generally are not fitted with the same metrological equipment as large installations and/or that there is a limit to the economically efficient retrofitting of metrological equipment, whereas large installations frequently enjoy a high degree of automation and RDT equipment so that the focus here is more on the "establishment and change of the condition of safety and function-relevant components".
Component-specific ratios and evaluation approaches permitting statements on the condition change of components have been created for the individual CBM-relevant components; furthermore, the study offers recommendations on the measurement parameters required to establish the condition of the installations.
Inclusion of condition changes
When heeded, the project results permit a more careful performance of condition-based maintenance with greater relevance to the specific properties of individual installations. In other words, the project results thus facilitate the condition-based maintenance of gas measurement and pressure reduction stations, also taking into account the condition changes of safety and function-relevant components, and the method of establishing their condition. Next to creating a damage and failure statistics to identify failure-prone components, the relevant measured values should be recorded and analysed on a regular basis so as to be able to discern trends in changes in condition as early as possible. The results will be integrated into the ongoing review of Code of Practise G 495.
The system of scores is currently being validated; i.e. it is being checked for its user-friendliness, and the weighting of the "key variables" and "safety and function-relevant components" criteria in particular is being tested for its suitability for application in the field.






