Geographical Information System in the Exploration Environment

The potential of Geographical Information System technology for Exploration and Production operations was recognised by PDO in the early 1990’s and, through its affiliation with the Royal Dutch Shell Group of Companies, led to its participation in the development of a GIS application by Genasys II UK, using its GenaMap/Genius products. The GIS would enable any Shell operating company to implement and build GIS applications around a common user interface and functional model. In 1993, prior to delivery of the common GIS in mid-1994, PDO carried out a pilot GIS project which identified the potential benefits of the technology within Exploration, with target applications being primarily multi-user access to seismic coordinate data and interpretations, prospect and lead portfolio data, and well attributes.

At the same time, PDO began to develop an in-house image processing capability which could be fully integrated with the existing digital mapping environment. The primary objective of this development was the provision of a low cost alternative to photogrammetric mapping in support of seismic acquisition surveys. The popularity of image-based map products increased substantially within the exploration environment, and across the company as a whole, as PDO scientists realised the potential of digital multi-spectral image data to aid geophysical data processing, interpretation, and for operational planning. Processed image data were made available to the PDO GIS user community as raster maps, over which various vector map layers could be displayed.

This paper discusses the implementation approaches adopted by PDO, various techniques used, and describes the benefits of GIS to PDO’s business. The paper will also illustrate how the implementation of GIS, in combination with the development of homogenous corporate databases, is leading to an improvement in data quality, in wider use of common datasets, and better decision-making.

There are many definitions of GIS, but one which is commonly used is: “a system for capturing, storing, checking, manipulating, analysing and displaying data which are spatially referenced to the Earth.” (Maguire et al., 1991).

In practical terms, it can be described as a computer-based application which embraces the disciplines of digital mapping, computer-aided design (CAD), (relational) database management, and image processing. GIS applications rarely excel at any of these, and do not (or should not) pretend to replace the many specialist applications which support these disciplines. However, where a GIS does excel is in the area of spatial analysis, providing an environment to make (ultimately) maps through the integration of spatial data with underlying (linked) attribute information, often residing in external databases.

GIS is more importantly an integrating technology. The integration of information generated by specialist processing software stored in a corporate database results in a map showing the location of licenses, prospects, wells, seismic surveys and reservoir interpretations. Where a GIS differs from other digital mapping applications is in the need for structured topological data for any map layers where spatial analysis is required.

Topology can be defined as “ the study of the characteristics of geometrical objects that are independent of the underlying coordinate system”, and essentially means the ordering of map data into points, lines and area features. Points are single pairs of coordinates, lines are sets of ordered and connected points, and areas are collections of line segments which close to form discrete polygons. Identifiers (called ‘tags’) attached to the map features provide the means to analyse the map data and link it to internal or external spatially-referenced attribute databases (e.g. Oracle). The internal data structure within the GIS, along with software tools, then provides the means to efficiently analyse and navigate using the map features.

Most industrial GIS handle both vector and raster data and provide tools to carry out analysis using either. Remote sensing image data is an example of raster data, dealing with a discrete space having units as squares or pixels, and data values assigned to each pixel representing spectral reflectance.

Numerous GIS are available in the market place, ranging from low-end PC systems to high-end workstation based applications, some fully-packaged to deliver a complete GIS solution which can be used almost immediately, others delivering a toolkit from which a customised GIS application can be built. PDO’s GIS follows the latter approach and has been built for the Shell Group by Fugro-Inpark BV, Netherlands using the GenaMap/Genius GIS products. It is implemented in a multi-user networked IBM RS/6000 environment where user access is through various platforms including high end PCs (via X-vision), Apple Macintosh (via eXodus or Mac-X), Sun and Hewlett Packard workstations (Figure 1 illustrates the main user interface).

The first step in implementing GIS in PDO was to start defining and populating the map layers, which were organised into logical themes, and where necessary linking these to attribute data in databases (mostly Oracle-based) (Figure 2). The approach adopted was to provide a basic set of contextual data (topography, boundaries), and then to gradually add and build new map layers as required, to support the required GIS functionality. By early 1996, the GIS supported approximately 300 map layers, of which 50 had attached database attributes, organised in 13 themes: Topography, Wells, Seismic, Fields, Geology, Oilfield Infrastructure, Exploration Prospects, Safety Environment, Administrative Boundaries, Geophysics, Production, Images and Map Indexes.

The structured coordinate data available within a Microstation-based digital mapping system used in the Exploration Draughting Department, and in legacy databases, made import of data to GIS fairly easy, although the need for strict topologically-structured data required clean-up using topology and feature-tagging tools. Some attribute datasets were available in PC or Apple Mac environments (such as in spreadsheets or PC databases) and were first exported to custom-built Oracle databases. The master source of spatial data is held external to the GIS, either in Corporate databases or in structured Microstation design files.

Effective operating procedures, to manage the data are embedded into the GIS. The design structure ensures compatibility among the procedures and minimises redundant activities. The structure will guide the flow of data into and out of the GIS database. Since much of the data is required in both the GIS and digital mapping environments, maintenance tools are provided to slave to both from these databases (for example scripts accessing the Corporate well database run overnight to create both new GIS maps and Microstation design files with the latest surface and subsurface well data).

In populating GIS, its power as a quality control (QC) tool became evident as artifacts in the datasets were easily located and fixed. Simply as a result of providing the user with the visualisation of data in a map-based view, but also relative to other data types (for example, prospects relative to seismic and wells), led to significant data improvements. In some cases relationships in the data, otherwise unknown, became evident. Also, the fact that the data became shared amongst a large number of users provided an incentive for the data owner to ensure data quality.

This highlights the benefit of using GIS as a means to clean up spatially-related data. GIS enforces data standards: features must have unique codes (tags) to link to external attributes, and must have reliable spatial coordinates to display in a map interface. It can therefore be a catalyst for improving data standards. Furthermore, if there is a need (and there often is) to share data between departments or disciplines where data resides within different corporate databases, then the use of company-wide common keys (tags) is essential (for example standard seismic line names or well names).

Initially development of GIS applications concentrated on 3 data types: prospects, wells and seismic, of which by far the main benefits have been realised from management and clean-up of the prospect portfolio using the GIS and a Shell Group Oracle database application: SPACE/DB. The application allows the external attributed data to be fully integrated into the GIS so that it appears as part of one homogeneous system.

Prospects have been imported in two ways: (1) as point features linked directly to tables within SPACE/DB providing general attributes as well as volumes and risks; and (2) as polygon features layered according to prospect type and stratigraphic objective. Other GIS layers were provided to support the application, such as geological features, play domains, team and ‘prospect champion’ boundaries, as well as wells and seismic. A simple spatial query, for example, would select all seismic lines intersecting the prospect and report these to the user, perhaps for checking that all lines were available to an interpretation project (Figure 3).

A GIS tool to allow the user to build Standard Query Language (SQL) queries to any database and report the results back in the map environment was provided, as well as several standard database reports and exports (Figure 4 shows the prospect module tool). The export tools allow users to create a subset of data in delimited form suitable for import to Excel for further statistical analysis. In addition, a business graphics module allows users to map database attributes as either bars or pie charts on the map (Figures 5 and 6). This tool was particularly useful to managers in harmonising the prospect portfolio across team and domain boundaries - i.e. removing biases resulting from teams or individuals using more or less conservative risk estimates.

A further useful GIS tool provided, named a “Post-it” after the yellow sticker variety, allows the user to interact directly with the map and by pointing to map features, pull up a short summary of either geographic or database attributes of the feature (Figure 7). This was built to be generic and applicable to any map layer.

Laterally specific applications have been built for analysing prospect economics, by loading maps of risk and fracture probability and making spatial queries to these (by back interpolation) and also extracting distance to nearest facilities such as pipelines and gathering stations to help in calculating Unit Technical Cost (UTC). More recently, GIS functionality has been further utilised to analyse prospects as spatial clusters rather than individual points, recognising that the success of one prospect could have a positive economic effect on other, possibly less prospective sites, in the immediate vicinity (Figure 8).

Wells were imported as point feature maps in different layers according to well type (exploration, development, water, etc.) linked to a corporate well database. In addition a 2-D vector map of well trajectories was provided, which when combined with surface infrastructure layers, can be used for well planning purposes (Figure 9).

The means to query, select and report wells according to subsurface horizons penetrated allows the user to locate well attributes according to horizons of interest. Also, tools are provided to export well tops and travel times in a format suitable for import to gridding/contouring systems. Attribute data available linked to GIS layers includes lithostratigraphy, well tops, biozones, biostratigraphy, sidewall samples, and cores.

Seismic layers provided include 2-D seismic and 3-D survey outlines linked to a database and providing access to historical seismic survey coordinate data and attributes. This includes a simple but useful table identifying whether 2-D data are already loaded into the Trace Interpretation Systems (TIS) workstation environment. A future enhancement (currently being implemented) is a link to a corporate seismic interpretation database allowing users to call up and display archived interpretations (including 3-D visualisation) and to export these into TIS projects (Figure 10).

In addition a seismic planning application was built, providing users with map layers showing current and future 2-D and 3-D survey areas linked to attributes such as survey date, crew, acquisition method etc.. Supporting the seismic planning process, full resolution (30 meters) Landsat TM satellite imagery has been loaded allowing geoscientists a first cut review of surface topography and infrastructure affecting the proposed survey (Figure 11).

Other layers available to support seismic planning include military areas, topography and survey control stations, which are used as the basis for establishing local survey height and position control.

Combining prospect and seismic planning layers, users are able to optimise their proposed seismic programme taking into account the current prospect portfolio, extracting prospect coordinates and distances interactively. Also managers can investigate cost benefits and set programme priorities by extracting summed prospectivity within the seismic prospect areas.

Several other map-based applications are in general use in PDO’s Exploration Department, ranging from legacy mainframe-based applications, to cartographic draughting systems (Microstation), gridding/contouring applications (Landmark’s ZMAP+), image processing (ER_MAPPER and Intergraph ISI-2), CAD applications (Canvas, Adobe), as well as systems supporting Trace Interpretation Systems, such as Landmark’s GeoDataWorks. Most of these applications are available to a limited user community or used on a ‘project’ basis, but require access to a standard set of basic spatial map data whether it be general contextual data (administrative boundaries, topography) or core exploration data such as seismic and wells. Maintaining the currency of data in such a wide portfolio of applications becomes increasingly difficult.

The current role of GIS amongst these applications is to provide user-friendly means for all Explorationists to visualise their data on their desktop, make spatial or attribute-based selections, do simple spatial analysis (proximity, buffering), make working maps, and report results or export data to other applications for further analysis. Final, cartographic quality maps remain in the domain of the drawing office using its Map Management System (Microstation-based) and the challenge for data managers is to ensure the availability, currency and efficient maintenance of data in the two environments (GIS and Microstation). The digitising and editing of map features remains in the drawing office domain and no attempt has been made to provide this functionality in GIS as, for example, Microstation graphic editing tools remain more advanced and productive than those available in the GIS.

GIS is also seen as providing the role of a spatial data warehouse providing dynamic links or tools to transfer spatial data into other project-based applications thereby ensuring users get user-friendly access to the latest data, as well as enforcing data standards through the GIS’ dynamic links to corporate attribute databases. To date links have already been built to export data to both Landmark and Zycor.

To date GIS has been used to support a wide variety of applications and already some significant benefits have been realised. Whilst it has made some routine jobs easier and therefore saved interpreters and technical assistants valuable time, it has also made available certain data and analyses that previously had not been contemplated because of the lack of functionality or data. Its main benefits however are in the areas of:

• - data visualisation through a common user-friendly map interface,

• - user friendly access to corporate data (data sharing) in a straightforward, reliable and timely way,

• - investigating data relationships through the spatial component,

• - improving data quality,

• - enforcing data management and standards, and

• - integrate with specialist software systems.

It has played a major role in cleaning up the prospect inventory and supporting management in preparing and optimising the exploration drilling programme. Similarly the ease with which users can now extract reliable data required for prospect economics is providing significant time savings.

Interpreters also report major time savings (days/per project) in locating well and seismic data using the dynamic links to corporate databases which were otherwise not accessible on the desktop in a user-friendly environment. In particular, tracking down all available seismic data, and its vintage was previously a time consuming task.

For new personnel, GIS navigation and search tools (find well x or seismic line y) provide a simple way of locating data, whilst the richness of data layers enables them to become quickly familiar with the environment and exploration history in their project area.

New applications are frequently arising, for example:

• - GIS was used extensively to support relinquishment decision-making in analysing prospectivity, investment history (wells, seismic, infrastructure) within target relinquishment.

• - Extracting road distances and drive times for the road network to support new journey management safety iniatives. This used road types (tarmac single carriageway, major graded, minor tracks, etc.) and allowable speeds to extract drive times, posted on a road map, a task that would have been impractical manually.

• - Reporting sensitivities for new well locations or seismic prospects.

• - Prepare elevation profiles using an underlying digital terrain model for microwave and pipeline route design (Figure 11).

Whilst it is mostly a success story, GIS technology is still rather immature and requires specialist expertise to operate, develop and maintain. In addition, strong Oracle database management skills, as well as a flexible and supportive network and system management are required to maintain the various live links to databases across a distributed network in a multi-user environment. For some end users GIS is sometimes seen as just one more desktop mapping application to learn, and for the occasional user, the application could still be made more user-friendly.

The successful implementation of GIS technology at PDO has provided users with access to valuable core information via a uniform map based interface. Visualisation of various exploration data such as wells, seismic and prospects, in a map-based view has led to significant improvements to the spatial definition of these data, and has led to the enforcement of data standards. The ability to link uniquely coded map-based information to various attribute data stored externally in an Oracle database, and to build SQL queries on the external data, has provided an additional catalyst to standardise corporate data storage. Furthermore, GIS is seen as a significant aid to efficient management of prospect portfolios and optimisation of drilling programs.

The evolution of GIS will continue in the short and medium term, in order to satisfy both new exploration requirements and to develop new applications for the benefit of the rest of PDO. Future developments will include implementation of closer links between spatial and attribute data, implementation of value-added raster data processes within the GIS environment, and improvement of cartographic tools. The development of new applications in support of non-exploration activites are expected to include pipeline and flowline facility management, general asset management, and emergency response.

The author acknowledges the contribution of Marc van Nes and Frank Peerdeman of Fugro-Inpark BV, and various PDO geoscientists, notably Angus McCoss, Mark Wilson and Graeme Smith. The Ministry of Petroleum and Minerals of the Sultanate of Oman is thanked for permission to publish this paper. The author also thanks Gulf PetroLink for improving the graphical displays.

Roger D. Abel is currently a Topographer with Shell International Petroleum Company with assignments in New Zealand, Holland, Oman and currently Nigeria. Roger also worked with Gardline Surveys UK as Hydrographic Surveyor between 1978 and 1980. Roger holds a BSc (Honors) in Land Surveying Sciences.