Drilling into the data - part two

Subsea well integrity monitoring

All the reported measurements were carried out using the sensor system shown in Figure 5. The illustration shows the set-up in 2017. From 2019, the system also includes a strain sensor kit.

Figure 5. The SWIM monitoring system.

The measurement system consists of three integrated motion units (IMU) measuring accelerations and rotation rates at 10 Hz sampling frequency: the riser sensor, LMRP sensor and wellhead sensor.

The accelerations and rotational rates are converted into angles. Additionally, some integrity parameters for the well are calculated. These integrity parameters assume that drilling with a subsea BOP for most exploration wells can be expressed as a simplified pendulum model.

The following parameters were calculated:

• The well rotational stiffness.
• The BOP resonance period.
• The BOP damping ratio.
• The wellhead rotational depth.

The integrity parameters and the measured response are stored and structured for each well, allowing for the use of this data for future operations.

Comparing upfront design analysis with measured response

All relevant data is stored in a structural storage or data reservoir. Design input, soil site survey data, analysis models, analyses results, and measured responses are stored together with relevant metadata.

The benefit of using the same vendor for measuring and performing upfront analysis for several exploration wells with the same vessel is that the measured response can be utilised to improve the analysis models for future campaigns.

It should be noted that the comparison and improvement of the analysis model should be done on parameters that are rig and riser specific, not on parameters influenced by the well response since operations are carried out on various wells.

One example of such parameter is the BOP stiffness, where the BOP is often modelled as an infinite stiff beam in global riser analyses. A significant difference between the wellhead and the LMRP sensor dynamics was observed. By using the measured response from the LMRP sensor and the wellhead sensor for several wells, the BOP stiffness can be estimated. The ratio between the LMRP and the wellhead measured angle standard deviation is presented together with analyses results in Figure 6.

Figure 6. Ratio of LMRP to wellhead angle from analyis (Hs = 2.5 – 4.5, Tp = 3.5 – 20.5) and measurements.

The initial riser model was too stiff, and by reducing the bending stiffness of the BOP a better match between analyses and measurements can be achieved.

Re-entering an existing exploration well

Further benefits from measurements can be achieved when re-entering an existing exploration well. In this case, the measured response can be used to establish the actual stiffness of the well, removing the need to make conservative estimates from the uncertainty from the site survey. By combining the integrity parameters (BOP resonance period, rotational depth and well stiffness) the softer soil support scenarios can be prevented. An example of this is shown in Figure 7.

Figure 7. Dynamic amplification over the BOP. 

The plot shows three different soil profiles (lower bound, upper bound and a best estimate curve), all established using the soil spring formulation from API RP 2GEO.⁶ Two different amplification spectra from the measurements are presented (the stiffest and the softest hour from the measurement campaign), as the measured response showed a softening behaviour over time.

The knowledge of the actual well support was then used in the upfront conductor analysis to improve the operational limits for this operation. The effect on the maximum dynamic bending moment from the analyses of the different soil profiles is presented in Figure 8.

Figure 8. Consequence on the conductor bending moment, the lower bound soil profile.

The maximum observed conductor bending moment was reduced by 35%, and the bending moment in the first conductor connector was reduced to close to zero. This had a significant impact on the operational limits as well as the fatigue prediction of the connector for the upcoming operation. The improved fatigue prediction was enabling the operation, as the initial analyses of the re-entry of the well showed unacceptable fatigue damage.

Conclusion

Using an approach where structured data from design analyses is combined with measured data from the actual operation is beneficial both for the ongoing operation and for future conductor designs.

A pre-requisite for such an approach is to store all relevant information in a structural manner with relevant metadata. This enables a learning process whereby knowledge from the operational phase can be fed into the design of future wells to improve the operational window and reduce conservatism in well conductor design.

This includes both structural and fatigue performance of the wellhead system. This will potentially allow for reduced conductor dimensions, reduce the number of conductor joints, allow for less complex connectors and simplify offshore operations; in total, reducing the investment and operating costs.

In the case of a re-entry of a subsea well, the measurements can be used to rule out the most conservative estimates of the soil support, increasing the allowable vessel offsets and increasing the allowable wave height for the upcoming operation. This again may reduce time waiting on weather and reduce costs related to station keeping. Further on, enhanced models will provide improved wellhead fatigue estimates, which extend the well service life and enable increased production.

For wells with challenging soil support conditions that lead to limitations on the well operations (Hs or offset limitations), the measured response can be used to rule out the most conservative soil estimates, extending the operational windows and reducing the cost of the operation.

4Subsea is constantly developing analyses methods and measuring algorithms in order to deliver more value to the customer. The SWIM measurement system, for example, now includes strain sensors, which enables direct measurement of bending moment. This will again reduce the conservatism in the estimates of bending loads imposed by the riser and BOP on the top of the subsea well.⁷

The company is currently performing more sophisticated soil models for conductor analyses. The methodology used to compare global riser and BOP response with the measurements has been significantly improved. This now includes a more extensive set of system state parameters and how these can be extracted from the measurements.⁷ Headquartered in Asker, Norway, the company has also developed a methodology where these state parameters can be used directly as input to a global riser analysis to obtain more accurate prediction of the riser response.⁸

References

1. H. Holden, I. Marre, H. G Holm, and R. Næss, “Enabling Drilling Operations Using Structured Data and Measured Response,” in OTC, OTC-27726-PT, 2017
2. H. Holden, M. Russo and P. Bjønnes, "A simplified methodology for comparing fatigue loading on subsea wellheads," in ASME International Conference on Offshore Mechanics and Arctic Engineering, OMAE2013-11529, 2013.
3. NORSOK, "NORSOK standard U-001, Subsea production systems," Edition 4, 2015.
4. P. Jeanjean, "Re-Assessment of P-Y Curves for Soft Clays from Centrifuge Testing and Finite Element Modeling," in Offshore Technology Conference, OTC-20158, Houston, Texas, 2009.
5. H. Holden, H. G. Holm, Y. Zhang, V. Smith and R. Nress, "Calibration of soil spring models for conductor analysis based on field measurements," in Proceedings of the ASME 2017 36th International Conference on Ocean Offshore and Arctic Engineering, Trondheim, 2017.
6. ANSI/API, "RECOMMENDED PRACTICE 2GEO, Geotechnical and Foundation Design Considerations," Addendum 1, 2014.
7. H. Holden, V. Martinsen, Å. Grønningsæter: "Using a new, autonomous, and easy to install strain sensor for monitoring of dynamic loads on a subsea BOP in order to reduce conservatism and increase service life of subsea equipment." in Offshore Technology Conference, OTC-30862-MS.
8. A. Cetin, H. Holden, V. R. Solum: "Robust method for wellhead loads estimation based on lower stack motion measurements" in International Conference on Ocean, Offshore & Arctic Engineering, OMAE2020-6486

This is part two of a two-part article. Part one is available to read here: https://www.oilfieldtechnology.com/offshore-and-subsea/21042020/drilling-into-the-data--part-one/

Read the article online at: https://www.oilfieldtechnology.com/offshore-and-subsea/22042020/drilling-into-the-data--part-two/