簡(jiǎn)介：在開(kāi)發(fā)深水油田初期，�？松�美孚公司就使用四維地震模擬技術(shù)對(duì)該油田Kizomba A進(jìn)行規(guī)劃，現(xiàn)在它已經(jīng)收回采用四維模擬技術(shù)的成本，開(kāi)始獲益。

Asset planning that considers 4-D monitoring from the very beginning pays dividends for one company’s deepwater fields.

Some technologies seem to pop up out of nowhere or at least gain attention very suddenly. Others develop at a seemingly glacial pace until one day everyone realizes just how valuable a part of the toolkit they really are.

Time-lapse (4-D) seismic is probably a good example of this latter group. Although the technology got a lot of fanfare in the late ’90s and was even showcased on a couple of television news shows, it remained the purview of a select few for several years because of its inherent complexities, namely the difficulty in repeating acquisition geometries and in determining whether changes between surveys were actually reflecting anything of importance happening in the reservoir.

Gas released from solution as the reservoir pressure declines illuminates connected volumes of a reservoir in the 4-D data at Kizomba A. The data also show effective water sweep in one reservoir interval and fingering in another zone

E&P talked to Dave Johnston, geophysics coordinator with ExxonMobil Production Co. He recently highlighted his company’s experience in the deep water as an AAPG Distinguished Lecturer. Johnston has been focused on reservoir characterization for his entire career and has evolved into an expert on 4-D.

The early days

Beginning as a researcher at ExxonMobil, Johnston studied rock physics to understand how heterogeneities at the pore scale influenced the seismic and electrical properties of reservoirs. From there he moved to predicting lithology based on seismic velocity analysis and then to predicting reservoir properties such as porosity from 3-D seismic data.

“It was a natural progression to go on to 4-D seismic,” Johnston said. “Instead of focusing on just the static properties of the reservoir, I started thinking more about how the dynamic properties influence recovery of hydrocarbons, how they can be monitored, and how we can use that information to maximize value for our fields.”

While references to 4-D date back to at least the 1960s, Johnston said that seismic technology was not well enough developed to image the types of subtle changes that occur in a reservoir over time. It took the maturation of 3-D technology in the 1980s to enable the technique to take off.

It also took a major change in mindset. Johnston said that early data comparisons had to be done with existing datasets that might not have been acquired or processed in the same way. “These were science experiments that we used to demonstrate some of the key concepts of 4-D technology,” he said. “It wasn’t until we could start planning for seismic reservoir monitoring and convince our business units to acquire dedicated 4-D surveys over their fields that we could demonstrate this technology could really add value.”

The tipping point

Deepwater developments are among the chief candidates for ExxonMobil 4-D surveys today. It was not always so. Johnston said the litmus test was the Jotun field in the North Sea. The field had reached its production peak rapidly, stayed on plateau for only a short time, and began a rapid decline. Water cuts were extremely high in some wells, nonexistent in others.

“It was clear that the model for depleting that reservoir was incorrect,” he said. “There was a lot of incentive for drilling infill wells, so we were able to convince the business unit to shoot a 4-D survey.”

However, two wells were drilled before the survey results were available. Both were failures. One well was drilled in a zone that the model indicated should have been unswept, but the area was depleted. The second well failed to find productive sands.

Once the 4-D survey results were available, all of the wells drilled based on that information were successes. “I think Jotun convinced ExxonMobil management that 4-D could add value to our assets and that it was something we should consider for broader commercialization,” Johnston said.

Current 4-D strategy

Years of working with 4-D data has created a workflow within ExxonMobil that begins the discussion of time-lapse surveys as early as the exploration phase of a project. Johnston refers to this as “part of overall field geophysics lifecycle planning.

“When we discover a field, we want to make sure right upfront that we obtain the data that allow us to evaluate whether or not 4-D seismic might be a viable technology in the future,” he said. This involves a screening process to examine rock and fluid properties as well as seismic data quality issues. This information is used to high-grade the assets and determine which ones stand the greatest chance of 4-D success.

When those fields enter development planning, and an initial geologic model, flow simulation model, and depletion plan are complete, a more detailed feasibility study is conducted to predict the 4-D response in the future. This can be used to determine optimal repeat times for collecting 4-D data and to estimate the potential reservoir management benefit.

“Ideally, all of that information is put in the development and surveillance plans for the field. Time-lapse surveys are included in the field budget so that when the asset is transferred to ExxonMobil Production Company, everyone is aware that 4-D is the plan,” he said. “Because of the length of time that it takes to acquire seismic vessels these days, we want to make sure that we are in a position to start procurement as early as possible.”

He added that 4-D is particularly useful for deepwater fields, where drilling and intervention costs can be prohibitive. In West Africa, all of ExxonMobil’s fields are developed in partnership with national oil companies. These companies often require produced gas to be injected into the field for future development. “Knowing where the gas is going is critical to prepare the field for gas production and to prevent gas recycling, which is costly and wasteful,” Johnston said.

The reservoirs are also complex, with heterogeneities that have a big impact on sweep efficiency and complex channel systems that offer the potential for both vertical and lateral connectivity. There’s also compartmentalization that can affect the performance of producer/injector pairings. A good 4-D survey can help explain these complexities.
Beyond that, it may be the only surveillance tool available. Johnston said that many of the fields are being completed using subsea wellheads tied back to a platform or floating production, storage, and offloading vessel (FPSO), meaning that any kind of downhole surveillance or intervention is extremely difficult.

“Given the environment that we’re working in, the way these fields are being developed, and the high cost of intervention, 4-D seismic can be a relatively easy sell,” he said.

The facilities themselves pose another problem. Since baseline surveys are often shot prior to development, the acquisition grid encounters no obstacles. Repeat surveys have to work around the platform or FPSO along with any other vessels operating in the area.

Usually a source vessel is placed on the opposite side of the installation from the streamer vessel to undershoot it. But this geometry is very different from the baseline survey. For a couple of fields ExxonMobil has acquired baseline surveys with the anticipated obstruction in mind so that the boats steer around an imaginary facility, making repeatability less of an issue in the future.

Overall, Johnston said, challenges include obstructions, currents, aggressive drilling schedules, and the need for quick data processing turn-around. But the benefits are worth the trouble.

“For us, there are significant rewards for 4-D seismic application in the deepwater environment,” he said. “These include efficient gas management, optimizing our infill well locations, identifying new opportunities for infill wells, and improving our reservoir model. I like to say that all models are wrong, but some are less wrong than others. By using 4-D to condition our models, we can make them more predictive.

“If we have a more predictive model, we can more effectively manage the reservoir.