Oct-2012

Updating plant information

Bringing the structures of existing refineries into the digital information age transforms the processes of upgrading and operating

CLIVE WILBY and GARY FARROW
Aveva

Article Summary

When building a new plant, the ability to create its engineering and design information digitally has brought new levels of capability and efficiency, not only in project execution but also in handover, commissioning and lifecycle management. But the majority of the world’s refineries and petrochemical plants were built before such technologies were available. For their operators, the challenge is how to bring these assets into the digital information age to gain the advantages this brings in efficient operations, safety, compliance and upgradeability, and to do so at an economic cost. Only a few years ago this was not practical, but technology has advanced rapidly, bringing down the technical and economic barriers to reverse engineering the digital plant from the master data source of the as-operated one.

There are three aspects to this: the design aspect of capturing the as-operated plant into intelligent 3D models, comparable to those that new plants are created from; the information engineering aspect of capturing and validating the vast amount of other types of engineering and operational information (for instance, electrical and instrumentation); and the knowledge management aspect of capturing the valuable experience of plant operations and maintenance personnel. Achieving these can make the operation of legacy plants as efficient as that of brand-new ones. Around half of the world’s oil production comes from mature facilities that require upgrading and life extension, so here we want to focus on the first two aspects, which particularly affect plant upgradeability.

Capturing the as-operated plant
Rarely, if ever, will an as-operated plant accurately resemble its as-designed state, and the older the plant the wider will be the discrepancies. So, whenever any significant modification is required, the first task must be to survey the affected area. Historically, this has been a difficult, inaccurate, costly and often hazardous operation, particularly where it involves working at height, in congested areas or offshore. This problem has been solved by the development of 3D laser surveying, where both hardware and software have advanced rapidly to a mature level, leading to its wide usage in the oil and gas industry (see Figure 1). 3D design technology vendors provide tools for referencing the raw laser scan survey data in the modelling environment, enabling plant designers to detect clashes between the new design and the as-is condition and create a new design that will fit accurately.

This capability has proved extremely effective, with many contractors able to deliver accurate, right-first-time revamps, minimising project cost, downtime and the hazards of site work. Often, however, the rest of the plant is left partially or completely unsurveyed or, at best, existing only as a 3D point cloud image. While scanner vendors generally offered software that could recognise simple geometric solids in the point cloud, this was a long way from being able to reverse engineer real plant objects. Now, however, the latest technology does enable this, by making use of the 3D geometry data contained within an engineering equipment and piping catalogue.

In use, the software recognises where 3D data points represent, say, a pipe or a beam, and offers the user a choice of the most likely corresponding catalogue items. The user then selects the appropriate choice, for example by choosing the correct pipe specification, and the system generates an intelligent native 3D entity in the design system, superimposed on the as-built data. In this manner, it becomes quick and easy to reverse engineer a laser survey into an accurate, validated and usable 3D model of the physical plant. Importantly, the modelled items can be intelligent, knowing to which other items they are connected, just as if they had been designed in a 3D modelling application from the outset. Once captured in this way, it becomes possible to associate these objects with their counterparts in the P&ID and other information items such as datasheets, plot plans, instrument loop diagrams, single line diagrams and so on, essentially following the normal engineering and design process but in the opposite direction.

The alternative approach to this would be to outsource complete manual remodelling, but this would involve considerable cost and elapsed time to create a model that would still require thorough checking and validation. By employing the appropriate cats & specs data, the semi-automatic solution not only saves a great deal of time and cost, it also generates 
high-quality (that is, specification-based), validated models. The availability of such a reverse engineered model of an as-operating plant greatly facilitates all aspects of plant engineering, not only the design of revamps.

For example, the reverse engineered objects will have accurate weight and centre of gravity data, which can assist in the planning of dismantling and lifting tasks, reducing downtime and site hazards. And just as on new plant projects, the models can also be used for operational purposes such as turnaround planning, plant maintenance, staff training and so on, with the added advantage that the models accurately represent the real plant.

Laser surveying also provides an opportunity to accurately capture the many small-bore pipe runs around a plant (see Figure 2). These are usually field run during construction, so even new plants may not include them in the 3D model and their details may not be accurately recorded in the as-built documentation. But they can be every bit as critical to safe operations as the large-bore pipes.

Inevitably, there are some limitations to what laser modelling can achieve, although the technology is advancing rapidly. It is not practical to reverse engineer electrical and instrumentation systems, for example. But this can largely be addressed by the reverse engineering of information rather than physical objects.

Reverse engineering information
The 3D model is only a subset of all the information that defines the plant. While an individual valve has a visible 3D form with particular dimensions, it also has many invisible attributes and associated information, such as its flow rate, materials specifications, pressure rating, maintenance instructions, parts lists, supplier details and so on. Multiply this by all the tagged objects in a plant and add to that all the information that is not directly tag-related, such as operating procedures, operating limits, process hazard assessment information, process data and so on, and even a modest sized plant has a huge amount of information associated with it. And, of course, all this information exists in different, incompatible formats, both electronic and paper. The result is that the information assets of many older plants are largely unusable or, at best, unreliable. It has been estimated that 30% of the information in a typical refinery becomes out of date within five years. Since it is impossible to know which 30%, you cannot rely on any 
information, so a lot of time is wasted just finding and verifying information in order to carry out a task.