The reservoir engineering manager from a young oil and gas operator had a problem. From his office, the 6-ft (1.8 m), 4-in. highly experienced manager could see the snow on the mountains above Geneva. But his production was not in Switzerland. Instead, it came mainly from a group of old platforms strung together in shallow offshore Nigeria, producing from 47 separate reservoirs. The problem was to decide if it was worthwhile to upgrade the old platforms in the asset.

The company was founded in 1994 as an off shoot of a Geneva-based oil-trading group. It recruited a group of very experienced, technically highly proficient engineers and geologists. They helped build a company approach that focused on getting the key economic and technical decisions right and then implementing them quickly.

Software can help operators apply better reservoir management techniques to their oil and gas fields. (Graphic courtesy of Serafim Ltd.)
In 1998, the young company bought the Nigerian operations of another small operator. On the face of it, the acquired fields were old and in decline, but they had thin oil rims that had been developed with vertical wells, which quickly gassed-out. Maybe modern horizontal wells could revitalize production. Moreover, the reservoir engineering manager and one of his reservoir engineers had noticed something odd about one of the fields, E. Here, the vertical wells had coned-in so much gas that, with the small observed pressure drop, the gas cap must have shrunk a lot. But, in the wells, there was no sign of this. The most likely explanation was that the gas cap, and hence the oil-rim, must be much bigger than suspected.

The horizontal wells and the newly discovered extension of the E field increased production from 4,000 b/d to 12,000 b/d of oil, which is the limit of the oil capacity of the single separator train on the floating production, storage and offloading (FPSO) vessel. Simple spreadsheet calculations quickly showed that it was highly economic to put in a second train, so this was designed, constructed and installed.

There were, of course, reservoir simulation models of the bigger fields, and these helped enormously in understanding the behavior of the oil rims. But a conscious decision was made not to wait for the end of the full history match of the simulator models. If engineering calculations could show that the development decision to be made would not change substantially whatever the final history match was, then the development decision, such as to install the second FPSO train, should be, and was, made and implemented.

Then came the question of the A field platform. This, too, was capacity constrained, limited by its gross liquid and gas handling capacity. The wells flowing to the platform were a mixture of dry, low gas-oil ratio (GOR) and high GOR wells. The options open to the company, therefore, were not limited to increasing capacity but also included possibilities of optimization, for if one prioritized the drier or lower GOR wells, one could possibly produce almost as much oil as after an expensive upgrade.

The answer to this question required setting up a calculation model. But it would be a very complicated calculation for a spreadsheet, so the obvious choice would be an integrated asset model (IAM). Unfortunately, with existing IAMs, it looked as if creating and calibrating a suitable model would be a long process, particularly since none of their reservoir models was particularly suited to describe the behavior of the company assets’ thin oil rim fields.

The experienced engineer thought back to his earlier days studying aeronautical engineering and working as a Shell reservoir engineer before reservoir simulators and integrated asset models were widely in use and when analytical solutions played an important role in decision making. He believed that by using the information he already had it was possible to provide some fairly simple calculations — a linear programming calculation to determine the best use of available capacity and an algorithm to predict approximately the amount of gas coned in by his oil wells in those fields that were without simulator models. Why not find a way of implementing these calculations? A decision was taken to turn to a consultancy specializing in mathematics and reservoir engineering. The instructions were simple, “Help us implement these calculations. Do it quickly and in a way that is more robust and auditable than a spreadsheet.”

The young operator was taking a gamble; software development is notorious for over-spends and failures to deliver working products. But the consultancy was able to quickly modify its existing nodal analysis program to provide the oil company with a tool to solve the A field problem. It supplemented it with a database structure to manage the production profiles (both imported and generated within the program) and the calculations. It allowed defining the surface installations and their capacity, importing production profiles from several sources and running an optimization algorithm to calculate the maximum production under the defined constraints.

The calculations on the A field showed that the proposed upgrades would give significant extra oil production, and the proposed modifications went ahead.

The operator-consultancy cooperation did not end there. Over the past 4 years, the program has been further developed to provide the operator with the tool it requires for nodal analysis, reserves evaluation and for replacing the messy spreadsheets that are so often used for handling production profiles created in simulation or from decline analysis.
The software has also become the forecast database for the entire company. Reservoir engineers of the producing assets decide whether to use decline curve analysis (done within the program) or production profiles generated using other tools. The profiles, defined on a well-by-well basis or reservoir basis, are imported into the database, using the importing tools provided by the software, which can upload production profiles and historical production data either directly using predefined simulator result files, queries in other databases, etc., or via Excel. The software/database then allows combining those profiles and producing forecasts for individual wells, reservoirs, fields, assets or any predefined reporting group. Once the reservoir engineer is happy with the forecast, he or she freezes it, thus disallowing any further modification.

Those forecasts are subsequently submitted to managers for approval. The advantage of using this system is that the approver can quickly examine the forecast, and use the data in the database to study how it was generated. In general, the approval process permits the manager and the engineer to discuss the forecasts and the choices made to generate them. Once a forecast is approved, it becomes the reference used for reserves updates, which are held every quarter. The software allows updating Volts, which is the company’s reserves database, automatically. This process eases some of the pressures of the reserves auditing process since it gives clarity as to how the numbers were generated.

The development of the forecasting tool was the fruit of a close relationship between the operator, which has been able to specify what it wanted from the software, and the mathematicians and programmers writing the code. The application, Serafim FUTURE, is now being released to the wider world.

Our reservoir engineering manager’s office still has a view of the mountains surrounding Geneva, but now his company takes up three floors of a big office block. Its production has grown to 107,000 b/d after big investments in the deepwater O fields development, many wells drilled and new acreage aquired elsewhere in Africa and the Middle East. Oil from E and A fields and the rest of the older assets will gradually diminish in relative importance, but the fields have been an essential part of the growth of the company. And the choice of production forecasting methods has played a small but significant part in enabling and speeding up that growth.

Comments