Geomechanics modeling for better drilling performance

Rigorous log-based analysis helps operators determine rock mechanical properties. 

Hong (Max) Wang, Mohamed Y. Soliman, Zhaohui Shan, Halliburton; Brian F. Towler, University of Wyoming

Geomechanics modeling is very important for optimizing drilling performance by increasing ROP and minimizing NPT. It provides the basis for bit selection and drilling-parameter optimization, wellbore-stability analysis, and lost-circulation prevention and control. Because of difficulties accessing the rock formation, the modeling largely relies on data availability and quality. To narrow down the error margin, a vast amount of data may be needed for validating the modeling quality.

Using a real case from a cuttings injection project, it is demonstrated what data are needed, which documents contain the data and how the raw data are processed to obtain the needed parameters for further engineering simulation.

INTRODUCTION

Attempts to improve drilling efficiency without integrating geomechanics will not reach the optimum.

However, due to the difficulty of accessing the rock in situ, the cost to determine the rock properties is generally very high. Gaining this knowledge requires acquisition of a batch of data and processing the data for rock properties. Further, simulations are needed to understand the rock behavior under certain drilling conditions.

The major role of geomechanics in drilling currently includes the following:

• Drilling method selection (underbalanced/managed pressure drilling, conventional drilling, casing drilling, etc.)
• Well design for optimized well trajectories and casing programs;
• Drillbit customization and selection for improved ROP and borehole quality;
• Drilling-parameters optimization for high ROP;
• Wellbore-stability indicators for drilling-fluid selection and optimization;
• Wellbore strengthening for higher wellbore-pressure containment and coping with lost circulation.

Well design starts with understanding the subterranean stress and pressure profiles and lithologies. Selecting the best method of drilling can significantly reduce well construction costs, as demonstrated by a progressive reduction in total drilling days in a new drilling area. This occurs after a number of wells have been drilled and performance has been analyzed for better well designs.

Better designs also include optimized casing points and well trajectories including wellbore deviation and azimuth with an understanding of the effects of stress, strength and pressure environments.

Drillbit design, selection and usage should all be tied to the understanding of geomechanics¹ including the lithology, rock strength, shale plasticity, formation pressure, abrasiveness, etc. Bit selection software programs have been created to allow engineers to select the best drillbit for an interval based on the geomechanical properties of the formations. Understanding rock properties is also necessary to help ensure the drillbit achieves the optimal ROP.

Wellbore-stability analysis has long been performed to obtain a minimum mud weight for critical drilling projects. In addition to the mud weight, understanding of hole-collapse mechanisms is equally important for correct mud chemistry, such as selecting the right mud type, preparing an optimized water phase salinity, etc.^2,3 The stress and pressure environment, as well as the rock strength, also play important roles. Because wellbore stability is strongly related to the wellbore orientation, the proposed wellbore deviation and azimuth are needed as well. Rock sensitivity to water and various cations, shale water activities, CEC (Cation Exchange Capacity), etc., together with rock mechanical properties, are all necessary to predict rock behavior under wellbore-stability conditions.

Wellbore strengthening is a new concept^4,5,6 that has led to a series of studies^7,8,9,10 and subsequent field successes.^11,12,13 Wellbore strengthening techniques can help operators achieve higher wellbore-pressure containment and contribute to studies to understand why some wellbores are weaker than others.⁸ Similar to wellbore-stability analysis, designing a wellbore-strengthening job (sometimes referred to as stress cage) requires knowing almost the same types of data as wellbore stability analysis. Knowing the rock parameters, etc., allows us to perform simulations to understand fracture dimensions and determine the best treating materials and engineering design.

Geomechanics can be also important to HSE considerations. One of the important examples is cuttings-injection design. It requires substantial understanding of a geomechanical environment before a good injection simulation can be performed with computer software. Only with the integration of the geomechanics and injection simulation, can a safe disposal domain be designed. Even during the injection process, geomechanical understanding is necessary to explain pressure anomalies and allow us to make smart operational decisions.

Aside from drilling, geomechanics can be extended to completion and production optimization for sand control, stimulation, etc. In a word, geomechanics has played an increasingly important role in drilling and other petroleum-engineering aspects as well.

DATA REQUIREMENTS

For different drilling applications, different sets of geomechanical parameters may be needed. However, in general they fall into the following categories:

• Wellbore geometry
• Rock mechanical properties
• Stress and pressure
• Wellbore fluids
• Fluid and rock interaction.

Wellbore geometry accounts for the hole size, deviation, azimuth, length of an openhole interval, etc. Rock mechanical properties may at least include Young’s modulus and Poisson’s ratio. Rock strength, permeability, and other properties may also be needed depending on the interest of investigation.

Stress magnitudes and orientation are very important for the wellbore stability that is a key design criterion for drilling. Formation pressure and fracture gradients define an initial mud weight window. Furthermore, wellbore stability should be considered for this window. If this window is not wide enough, wellbore strengthening may be applied to widen the mud weight window.

Wellbore fluids provide information about the chemistry, rheology and particulate type and concentration. These are all important for understanding wellbore behaviors. Rig daily operations provide information about fluid and rock interaction. This can also be obtained through pressure tests such as Leak-Off Tests (LOT).

DATA COLLECTION

These data are generally obtained through analysis of the following documents:

• Well diagrams
• Formation pressure, mud weight, Frac gradient profile plots
• Drilling, mud and mud logging reports
• Logs
  • Full wave sonic (compressional & shear)/other sonic logs
  • Density/neutron/GR/caliper/borehole image logs
  • Formation pressure tests
  • Other logs
• Region information
  • - Basin study report/seismic map/fault map
• Other information
  • - Core test report, cuttings analysis report
  • - Well testing, hydraulic fracturing report or water injection report
  • - LOT/Extended LOT.

The wide range of data sources normally requires extended knowledge and experience to extract valuable information from these documents.

DATA ANALYSIS

Rock mechanical properties can be obtained through lab tests. However, rock samples are normally not readily available because coring, sample preservation, HTHP testing, etc., can be complicated and costly procedures. A more common methodology uses log-based analysis. Results from lab tests can then be used as calibrations. Table 1 summarizes methods for obtaining various parameters for drilling related geomechanics analysis. Data quality control is always a necessary step for generating correct results.

TABLE 1. Summary for methods to obtain the rock mechanical properties and in situ stresses Click Table to Enlarge

CASE HISTORY

This simulation was done for performance analysis of a cuttings injection project that has been going on for several years. Due to confidentiality, critical information has been removed for this publication.

A geomechanical model must be constructed to provide a base structure for cuttings injection simulations. It is critical that the model reflect the best understood reality as closely as is practical, so that the simulated operational parameters result in output that represents real world performance. This is necessary to ensure safe and efficient cutting disposal.

Geomechanical modeling incorporates pore pressure and effective stress values in developing an understanding of the total active stresses. It is critical to understand the permeability, Young’s modulus, etc., in order clarify total active stresses. Because limited direct data are normally available for potential cuttings injection disposal zones, more expertise is required to understand and describe the geomechanical environment for these operations than for typical hydraulic-fracturing designs.

In hydraulic fracturing, pore pressure is normally known with well-testing methods or production records. However, for cuttings injection it is quite likely that pressure and stresses will have to be determined simultaneously by solving the related correlation equations using the limited known data points for total stresses at given depths.

Stress environment: The big picture. One of the first tasks in developing the geomechanical model was determination of the stress environment in the disposal zone. The pertinent log data were reviewed for quality control as a first step in the process. No log quality issues were identified.

Stepping further back in reviewing the stress in the area, the stress environment was better understood in light of plate tectonics in the area. It was found that the area of interest is in a unique strike/slip environment. Due to colliding movement of the two continental plates, severe stress anisotropy should be expected. The world stress map indicates strike-slip and thrust faulting regimes with the strike in the nearly NW direction. The compression stress component, therefore, is maximized in the NW direction.

This understanding is essential for selecting a good approach to determine the stress environment. (Details and figures are omitted here to protect confidential information.)

Determining the lithology. The lithology was determined using the provided logs. During a log-based analysis, differentiation was achieved primarily with the gamma-ray log and then translated into shale volumes by making use of the mud logs, as well as the other logs provided. The formations of interests were identified as claystone and sandstone.

Determining the closure pressure. The closure pressure was analyzed with the SQRT Time Plot, G-Function Plot, GdP/dG Plot. This is demonstrated with the LOT No. 1 data. Analysis plots using LOT data were performed using the above three methods.

A Pressure Decline Analysis (PDA) was performed for the LOT No. 1 on closure pressure at the depth of 7,907 ft (2,410 m) with the three methods. These results are very close to one another with an average of 15.9 ppg, which is the same as determined directly from the LOT raw-data plot. Many times, these results don’t agree with each other and one has to be cautious to select the correct result. With the same method, some other pressure decline data from cuttings slurry injection were analyzed and the summary of some the data are listed in Table 2.

TABLE 2. Summary of closure-pressures-based PDA

In Table 2, there are two data sets with significant variance from the others. The first one is the determination of a 15.87 ppg equivalency from LOT No. 1. This is a significantly lower value than the average. It is possibly due to the strong stress anisotropy which creates a low tangential stress area at and near the wellbore. This effect would not be seen again after the fracture has grown into the far-field area. Apparently in this case, the fracture initiation pressure is smaller than the far-field minimum horizontal stress.

The second notable variance is the 18.98 ppg result for Injection No. 2. This may be caused by solids accumulation inside the fracture that might possibly increase the stress at closure. Cleaning the solid deposit or creating a new fracture would reduce the closure pressure back to values approaching the far field stress-minimum horizontal stress. Subsequent batches showed reduced pressures and the variance that has occurred during the life of the injection project indicate some significant transient dynamics of solids placement and movement in the disposal zone. After eliminating the two data sets in question, an average of 0.94 psi/ft or 18.12 ppge minimum horizontal stress was determined at injection depth.

Determination of permeability from the available data remains approximate. LOTs were first performed on the originally targeted claystone formation at the base of the injection zone. Due to the limited nature of the pressure decline data available, a pressure decline analysis method developed by Soliman14 was applied to the test data to determine the permeability. This method can also be used to obtain a formation pressure estimate. It requires identification of the flow regime with a -tdP/dt vs. t plot, from which a constant for calculating permeability can be defined with (Pi-Pw)tⁿ vs. t. The pore pressure can also be determined at the same time.

Two LOT tests were performed at 7,907 ft with a brine fluid of 9.1 ppg. Using Soliman’s method, a claystone permeability of 0.14 mD was determined from LOT No. 1, as shown in Fig. 1. The bleed-back volume from the LOT No. 1 test was 1.2 bbl after 40 min. of shut-in. A total of 6.5 barrels were injected at a rate of 0.3 bpm. This volume of fluid loss indicates a relative larger permeability than normally expected in claystone formations. This permeability should be the average of the matrix and natural fractures if they exist.

However, due to the short shut-in time, the interpreted permeability can be higher than actual permeability. The interpreted value can be used as an upper bound for the claystone-formation permeability for the cuttings-injection simulation. Neither the SP log, nor resistivity logs, show good indications of permeability of any formations. The sandstone formations may have relatively low permeability.

Determining the pressure. The pressure decline analyses were performed on the test and injection data in order to better understand the nature of in situ pressure before and during the injection operations. Analysis was performed with data acquired from LOT No. 1. The results were either obtained by the Horner or Soliman’s methods. These analyses were performed to determine pressure in the claystone formations.

While the provided resistivity logs and density logs reveal no indication of significant abnormally pressured zones, analyses do indicate pressures consistently higher than a normal pressure gradient of 0.45 psi/ft. However, this is inconsistent with a review of the mud weights used in drilling the well to provide an estimate of pore pressure in the sandstone. The low mud weights used indicate a normal pressure gradient in the sandstone formation, which would tend to indicate higher permeability and formation extent, which would have allowed equilibration of pressure over geologic time. A typically very low permeability claystone might be expected to be sufficiently non-transmissive such that pressure would not be dissipated, however, the permeability of the claystone has been previously estimated to be approximately 0.14 mD. With that permeability, the pressure gradients in the claystone and sandstone formations should be basically identical.

Based on these findings, the pore pressure gradient is approximately normal and a 0.455 psi/ft pressure gradient was assumed in both sandstone and claystone formations at the depths of interests.

Analyses showed a gradual increase in pressure with increasing injection time, indicating the additive effects of continued waste injection. Therefore, the early data best represents the far initial-reservoir pressure. For a large body of the fluid injection, an increasingly longer shut-in time should be expected to reach an estimate of the original formation pressure, especially with low formation permeability.

Determining the minimum horizontal stress. Many different methods for interpreting horizontal stresses have limitations, and they all need calibration with some other data source, such as pressure tests. The method used herein is an “effective-stress” method, which assumes is a correlation exists between the vertical and horizontal stresses in shale. The method requires the knowledge of the overburden stress. The vertical stress was assumed to be 1.0 psi/ft, where the density log was not available for depths shallower than 1,000 m. The interpreted stresses are calibrated to the LOT results. Refer to Track “PS ppg” in Fig. 2 for minimum horizontal stress designated as Shm_ppg.

Determining Young’s Modulus and Poisson’s Ratio

Young’s Modulus and Poisson’s Ratio were derived using a synthetic sonic log based on the following equations:

Conversion to static values was performed by taking porosity into consideration. The analysis indicates that the Young’s Modulus is about 1.5 and 2.0 million psi for claystone and sandstone, respectively, and Poisson’s Ratio is about 0.3 for both claystone and sandstone. See “Mechanical Properties” in Fig. 2.

Simulation results. With the geomechanical model created, a multiple batch simulation was performed to see whether the result will match with recorded data.

Figure 3 shows the result of a 7,400 min. multiple batch injection simulation. It is the bottomhole injection over time. The recorded field data for the same period of time is also shown. It can be seen that they match very well.

CONCLUSIONS AND SUMMARY

Geomechanics modeling can provide the basis for improving drilling performance in various aspects.

• Data categories are summarized and data documentations are listed for data collection.
• Analysis for geomechanics modeling requires a wide range of knowledge and experience to extract and integrate the data into the model.
• A general process was demonstrated with an actual case history for cuttings injection performance analysis.

The case study indicates that the new method for pore-pressure and permeability analysis was easy to use.  

ACKNOWLEDGMENT

The authors thank Halliburton for granting permission to publish the article. This article was prepared for presentation at the 2008 AADE Fluids Conference and Exhibition held at the Wyndham Greenspoint Hotel, Houston, Texas, April 8–9, 2008.

LITERATURE CITED

¹ Caicedo, H. U., Calhoun, W. M. and R. T. Ewy, “Unique ROP predictor using bit-specific coefficient of sliding friction and mechanical efficiency as a function of confined compressive strength impacts drilling performance,” SPE/IADC 92576 presented at the SPE/IADC Drilling Conference, Amsterdam, The Netherlands, February 23-25, 2005.
² Rojas, J. C., Clark, D. E. and J. Zhang, “Stressed shale drilling strategy-Water activity design improves drilling performance,” SPE 102498, presented at the 2006 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, September 24-27, 2006.
³ Mody, F. K. and A. H. Hale, “Borehole-stability model to couple the mechanics and chemistry of drilling-fluid/shale interactions,” JPT, November 1993, p. 1093.
⁴ Aston, M. S., Alberty, M. R., de Jong, H. J. and K. Armagost, “Drilling fluids for wellbore strengthening,” SPE/IADC 87130 presented at the 2004 IADC/SPE Drilling Conference, Dallas, Texas, March 2-4, 2004.
⁵ Alberty, M. W. and M. R. McLean, “A physical model for stress cages”, SPE 90493 presented at 2004 SPE Annual Technical Conference and Exhibition, Houston, Texas, September 26-29, 2004.
⁶ Sweatman, R. E., Kessler, C.W. and J. M. Hillier, “New solutions to remedy lost circulation, crossflows, and underground blowouts,” SPE/IADC 37671 presented the 1997 SPE/IADC Drilling Conference, Amsterdam, The Netherlands, March 4-6 1997.
⁷ Wang, H., Sweatman, R., Engelman, B., Deeg, W., Whitfill, D., Soliman, M. Y., and B. F. Towler, “The key to successfully applying today’s lost circulation solutions,” SPE 95895 presented at the 2005 SPE Annual Technical Conference and Exhibition, Dallas, Texas, October 9-12, 2005.
⁸ Wang, H., Towler, B. F. and S. Y. Mohamed, “Fractured wellbore stress analysis - can sealing micro-cracks really strengthen a wellbore?” paper SPE/IADC 104947 presented at 2007 SPE/IADC Drilling Conference, Amsterdam, The Netherlands, 20-22 February 2007.
⁹ Wang, H., Towler, B. F. and Mohamed, S. Y.: “Near Wellbore Stress Analysis and Wellbore Strengthening for Drilling Depleted Formations.” SPE 102719 presented at the 2007 SPE Rocky Mountain Oil & Gas Technology Symposium, Denver, Colorado, April 16-18, 2007.
¹⁰ Wang, H., Soliman, M. Y. and B. F. Towler, “Investigation of factors for strengthening a wellbore by propping fractures,” IADC/SPE 112629 presented at the 2008 IADC/SPE Drilling Conference held in Orlando, Florida, U.S.A., March 4-6, 2008.
¹¹ Whitfill, D., Wang, H., Jamison, D. E. and A. Angove-Rogers, “Preventing lost circulation requires planning ahead,” SPE 108647 presented at the 2007 International Oil Conference and Exhibition, Veracruz, Mexico, June 27-30, 2007.
¹² Whitfill, D., Jamison, D. E., Wang, H., and C. Thaemlitz, “New design models and materials provide engineered solutions to lost circulation,” SPE 101693 presented at the 2006 Russian Oil and Gas Technical Conference and Exhibition, Moscow, Russia, October 3-6, 2006.
¹³ Wang, H., Soliman, M. Y., Whitfill, D. L. and B. F. Towler, “Case histories show wellbore strengthening as a cost-effective option for drilling with narrow mud weight windows,” AADE-08-DF-HO-17 presented at the 2008 AADE Fluids Conference and Exhibition held at the Wyndham Greenspoint Hotel, Houston, Texas, April 8-9, 2008.
¹⁴ Soliman, M. Y., Craig, D., Bartko, K., Rahim, Z., Ansah, J. and D. Adams, “New method for determination of formation permeability, reservoir pressure, and fracture properties from a MiniFrac test,” ARMA/USRMS 05-658 presented at Alaska Rocks 2005, The 40th U.S. Symposium on Rock Mechanics (USRMS): Rock Mechanics for Energy, Mineral and Infrastructure Development in the Northern Regions, Anchorage, Alaska, June 25-29, 2005.

THE AUTHORS
	Hong (Max) Wang is a global technical advisor for Baroid Fluid Services at Halliburton. Max has acquired extensive land and offshore experience with more than 20 papers on the subject published. He is a professional petroleum engineer with a PhD in petroleum engineering from the University of Wyoming and an MS degree in chemical engineering from South China University of Technology. He is an active member of SPE, DEA and AADE.
	Mohamed Y. Soliman is chief reservoir engineer with Halliburton Energy Services at its Houston center. Dr. Soliman received a BS in petroleum engineering degree with top honors from Cairo University in 1971. He also earned MS and PhD degrees from Stanford University in 1975 and 1978 respectively. Dr. Soliman has written over one hundred technical papers in areas of well test analysis, fracturing, Conformance, and numerical simulation. He also holds thirteen US patents.
	Dr. Zhaohui Shan, SPE, is a principle technical professional at Halliburton, Security DBS Drill Bits. He joined Halliburton 3 years ago. He has worked for more than 10 years in industries and specialized in fracture and fatigue analysis of materials and structures and design optimizations. He has about 25 publications in journals and conferences. He earned a PhD degree in mechanical engineering from Hong Kong University of Science and Technology.
	Brian Towler is a professor of chemical and petroleum engineering at the University of Wyoming and College of Engineering and Applied Sciences Fellow for Hydrocarbon Energy Resources. He was previously Department Head of Chemical and Petroleum Engineering from 2004-2008. He has a BE and PhD in Chemical Engineering from the University of Queensland in Australia. He was instrumental in setting up the Queensland section of the SPE in 1985 and was chairman of the section in 1988. He is a registered Professional petroleum engineer in the state of Wyoming.