Pore pressure monitoring optimizes deepwater wells

Exploration drilling in deep water exposes operators to high risks and significant drilling challenges, especially when they encounter narrow mud weight windows. Pressure-related events can cause serious safety hazards and costly nonproductive time (NPT) due to kicks, overpulls, losses, stuckpipe, or blowouts. The formation fracture gradient fixes the upper limit of the safe mud weight window, while pore pressure establishes the lower limit. In greater water depths, fracture gradients tend to be lower. When pore pressures also are higher, there is little margin for error.

Success in such high-risk wells depends both on thorough predrill planning and real-time pore pressure monitoring and intervention while drilling. Recently, Eni Indonesia achieved all of its targets without suffering any pressure-related NPT in four critical deepwater exploration wells, one of which had an operational window so narrow that it might have been undrillable without real-time monitoring.

Predrill pore pressure study

Eni Indonesia planned to drill four exploration wells in relatively unexplored blocks in the Kutei and Tarakan basins offshore East Kalimantan in water depths from 427 m to 825 m (1,400 ft to 2,700 ft). All four wells targeted new, deeper gas-charged reservoirs, not all of which had been encountered in previous drilling. Offset well information was limited, which increased potential risks and uncertainties.

In deeper water, fracture gradients are lower. When pore pressures are higher, the safe mud weight window can be extremely narrow. Avoiding NPT requires thorough predrill planning and real-time pore pressure monitoring. (Images courtesy of Schlumberger)

Eni teamed up with Schlumberger Data & Consulting Services (DCS) to conduct a detailed predrill pore pressure study. It was important not only to predict the presence of overpressure but also to better understand its cause. Following a thorough audit of all available well and seismic information, specialists generated pore pressure models for each offset well using two independent methods to increase confidence. One method, described by Eaton (1975), uses a normal compaction trend line. The other, described by Bowers (1995), involves establishing seismic velocity and effective stress.

Base shale pressures were estimated using seismic velocities and offset well data. Models were calibrated against direct measurements of formation pressure, leak-off tests, and pore pressure-related drilling events observed in drilling reports and logs. Then predrill pore pressure models were constructed for each of the four planned well locations using different methods for shales and permeable formations and taking into account the buoyancy effect of gas. Finally, Eni and Schlumberger engineers and geoscientists agreed on predrill fracture gradients and pore pressure estimates to guide casing design, mud weight programs, and contingency plans.

Target sand in the planned T2 well (left) was 5 km (3 miles) away and 458 m (1,500 ft) updip from an overpressure zone in the offset well. Due to the buoyancy effect of gas, pore pressure in the T2 was expected (and found) to be even higher.

The T2 exploration well in the Tarakan basin proved the most challenging well in the campaign. According to seismic interpretation, target reservoirs in the T2 were laterally continuous with gas-charged sands in an offset well. The offset well had a safe mud weight window up to 2 lb/gal. While measured formation pressure was elevated in the offset well, the predrill study predicted additional overpressure of nearly 1.5 lb/gal in the planned T2 well due to gas buoyancy. Since this well was drilled at a water depth of 825 m, low fracture gradients were estimated, which further constrained the mud weight window.

Real-time pore pressure monitoring

While drilling the critical reservoir zones, real-time pore pressure estimates were calculated to update the predrill model and predict pressures several hundred meters ahead of the bit. Unexpected changes in pore pressure would have to be addressed quickly to optimize safe drilling practices without incurring pressure-related NPT.

Because the predrill model indicated that the first target sand in the T2 exploration well would be overpressured, drillers were prepared to increase mud weight as necessary. Real-time pore pressure computations based on streaming LWD data showed a gradual rise, as predicted, which increased the team's confidence in the validity of the predrill model. About 50 m (150 ft) above the first overpressured target sand, a sudden increase in gas readings also was observed. The 16-in. casing was set, and mud weight was increased accordingly. While drilling the 14 3/ 4 -in. section through the first sand, a high-pressure increase closely matching the predrill model was observed. This proved the sand was in pressure communication with the offset well and suggested that drillers could expect higher pressures in deeper gas-charged target sands yet to be drilled. Because the pressure in the first reservoir was very close to the current limit of the operational window, mud weight could not be increased safely without first casing the hole. Hence, 13 3/ 8 -in. casing was set earlier than originally planned, and mud weight was increased. Because the 13 3/ 8 - in. casing had been short-landed, 11 3/ 4 -in. casing, which had been held in contingency, was added to the program.

Below the 11 3/ 4 -in. casing shoe, real-time estimates of pore pressures in the shales were somewhat higher than predicted, which provided advance warning that pressures in the sands might be higher as well. As a result, ROP was reduced, and pore pressure engineers carefully monitored events for any signs that pressures were increasing. However, no further high gas readings or pressure gains were observed. Although the operational window remained extremely narrow, real-time pore pressure monitoring and continuous model updates enabled drillers to successfully maintain mud weights slightly above predrill estimates while drilling through the remaining target sands.

Post-drill review

Eni's drilling campaign began with rather limited information on pore pressures and fracture gradients. The predrill study had established that overpressures in target reservoirs were due to under-compaction rather than fluid expansion or tectonic stresses and would increase updip due to the buoyancy effect of gas. The LWD data and real-time pore pressure-while-drilling data enabled even more reliable estimates. Finally, post-drilling reviews were conducted on each well, during which additional hard data were acquired through wire-line services. Real-time pore pressure models were calibrated accordingly and updated by integrating these wireline data, thereby enhancing Eni's knowledge base for future drilling operations in the area.

Acknowledgement

The authors thank Eni Indonesia and Schlumberger management for permission to publish this work. This article is based on a paper published in the 35th Indonesian Petroleum Association Annual Convention & Exhibition, IPA11-E-210, and at SPE APOGCE, Jakarta (SPE-147914).