Fall 2016 Saudi Aramco

THE SAUDI ARAMCO JOURNAL OF TECHNOLOGY Journal A quarterly publication of the Saudi Arabian Oil Company

Contents

Deployment of the World’s Longest Sandstone, Production Nozzle-based ICD, Partially Cemented System in an Offshore Horizontal Well 2 Qadir D. Looni, Mohammad J. Aljubran, Ahmed A. Al-Ramadhan, Marius V. Neacsu, Aly A. Emam and Christian T. Mora of Technology of Development and Field Test of ESP Reliable Power Delivery System 13 Dr. Jinjiang Xiao, Randall A. Shepler, Yhossie S. Windiarto, Rob Fox and Stuart Parkinson

Drilling for the Next Generation of Multilateral Completion Systems 21 Yousif M. Abu Ahmad, Rami F. Saleh, Brett W. Bouldin, Robert J. Turner and Ali Bin Al-Sheikh

Innovative Step Change in Drilling Efficiency for Medium Radius Reentry Deep Gas Wells with a High Build Rate Rotary Steerable System 29 Abdul Halim Ab Hamid, Verdy L. Siregar, Ali N. Al-BinAli, Mohamed E. Khalil, Ayman Ghazzawi, Omar T.A. Ashraf and Muhammad S. Balka

A New Insight on the Impact of Individual Ions on Fluid-Fluid Interactions and SmartWater Recovery 38 Mohammed A. Geer, Dr. Ahmed Gmira, Dr. Ali A. Yousef and Dr. Sultan M. Al-Enezi

Reservoir Stress Path from 4D Coupled High Resolution Geomechanics Model: A Case Study for Jauf Formation, North Ghawar, Saudi Arabia 45 Otto E. Meza Camargo, Dr. Tariq Mahmood and Dr. Ivan Deshenenkov

Implementing the Pressurized Mud Cap Technique for Drilling through Total Loss Zones: A Way to Improve Well Control while Drilling the Reservoir in Oil Well Reentries 61 Khalifah M. Al-Amri, Julio C. Guzman Munoz, Abdulrhman M. Al-Hashim, Ali M. Hassanen and Ayoub Hadj-Moussa

Automatic Well Completion and Reservoir Grid Data Quality Assurance for Reservoir Simulation Models 70 Tariq M. Al-Zahrani, Muath A. Al-Mulla and Mohammed S. Al-Nuaim

Case Study of Intelligent Completion with New Generation Electro-Hydraulic Downhole Control System 79 Suresh Jacob, Nibras A. Abdulbaqi, Chandresh Verma and Rabih Younes

87448araD2R1.indd 1 11/28/16 11:52 AM Deployment of the World’s Longest Sandstone Production, Nozzle-based ICD, Partially Cemented System in an Offshore Horizontal Well Authors: Qadir D. Looni, Mohammad J. Aljubran, Ahmed A. Al-Ramadhan, Marius V. Neacsu, Aly A. Emam and Christian T. Mora

ABSTRACT planning and execution. As a result, the deployment of best practices and state-of-the-art technology was mandatory for Given the current downturn in the oil and gas industry, the overcoming the various drilling challenges. Extensive work smart design and implementation of state-of-the-art drilling was conducted to anticipate and understand these challenges, and completion technologies are key factors toward an opti- including studying and analyzing the historical and offset well mum return on investment. The goal is cost savings, a maxi- data. The major drilling challenges in this project can be sum- mum productivity index, and an effective operational strategy marized as: with minimum risks involved. As part of these efforts, Saudi Aramco successfully deployed the world’s longest 4½” par- • Kicking off from vertical and drilling a curved section to tially off-bottom cemented liner with a sandstone production the target entry in a single run. equalizer system in an offshore field in Saudi Arabia. • Managing hole cleaning and stuck pipe concerns This record was achieved through close monitoring of the through proper drilling and tripping practices. wellbore condition and the creation of an accurate torque and • Optimizing bottom-hole assembly (BHA) and drillstring drag simulation prior to the job. The 7,389 ft open hole was design to manage torque and drag. horizontally drilled in 6⅛” sections and carefully geosteered through the reservoir to yield a 90% net gross ratio. Borehole • Delivering optimum drilling performance through tortuosity and dogleg severity were kept to a minimum. Drilling fit-for-purpose drilling systems selection. and borehole cleaning performance was monitored and en- • Enhancing the downhole equipment reliability to hanced through a real-time cutting recovery analysis. As soon minimize nonproductive time and trips. as the target depth was reached, a reservoir model was built, • Maintaining optimum mud conditions to avoid wellbore using the acquired petrophysical and reservoir data, and em- stability issues and to control shales. ployed toward design optimization. The completion liner was designed to provide the appropriate centralization and stiffness, • Managing drilling dynamics while ensuring proper and to ensure reaching the bottom of the borehole. directional drilling and geosteering. The main challenge in this project was that the well tra- jectory was just below a gas cap. Reservoir mapping led to WELL DELIVERY better understanding of the well placement relative to the gas cap. A sandstone production equalizer system and open Overcoming these challenges was critical to achieve good hole packers were used to divide the wellbore into sections well delivery. This required the implementation of constant and balance influx from high and low permeability zones. communication and teamwork, the use of well-defined best Restricting flow in the desired sections by introducing higher practices and the deployment of top-notch drilling technolo- pressure drops prevents early gas breakthrough from the gies. As a result of these efforts, the following solutions and overlaying gas cap. technologies were adopted: This successful deployment was a result of extensive engi- neering planning as well as operational alertness during side- • Optimizing well design by managing the doglegs and tracking and deployment. This article will provide further BHA compatibility. details on the design and operation phases of this project. • Conducting a local geomechanical study to identify wellbore issues and associated risks. PLANNING • Implementing state-of-the-art technology to ensure It was clear that this well was different from the conventional reliability and directional control, even in deeper drilling operations and required significantly greater care in intervals.

2 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 2 11/28/16 11:51 AM

Fig. 1. Target zone sand, mapped using Periscope HD to monitor the lower boundary (clearly shown) to maintain the maximum TVD and stay inside the target zone as planned. Fig. 1. Target zone sand, mapped using Periscope HD to monitor the lower boundary (clearly shown) to • Deploying PeriScope HD, a multilayer bed boundary GEOMECHANICS maintain the maximum TVD and stay inside the target zone as planned. detection service, to facilitate proactive geosteering and provide more information about the reservoir geometry Workover sidetrack operations in this offshore field typically

and substructure. cut across several shale formation layers to target a shaly • Utilizing cuttings flow meters to monitor hole cleaning reservoir sand. This sandstone reservoir could be at a low pressure due to depletion, meaning the virgin pressure in the effectiveness and wellbore stability. shale layers could be higher than the reservoir pressure.

This difference makes it a challenge to design a mud weight All these solutions and technologies ensured safe and op- (MW) that can both minimize shear failure in the shales, which timumFig. 1. delivery Target of zonethe well sand, and helped mapped to achieve using the Periscope well’s HD to monitor the lower boundary (clearly shown) to objectives. are at a higher pressure, and at the same time minimize differ- maintain the maximum TVD and stay inside the targetential zone sticking as riskplanned. across sandstone reservoirs. Furthermore, since the shale layers are drilled at a high angle before landing the well, they are also prone to bedding plane failure.

Fig. 2. Graph showing that the cuttings recovery (red line) was maintained at around 90% throughout the drilling operation.

Fig. 2. Graph showing that the cuttings recovery (red line) was maintained at around 90% throughout the drilling operation. Fig. 2. Graph showing that the cuttings recovery (red line) was maintained at around 90% throughout the drilling operation. SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 3

87448araD3R1.indd 3 11/28/16 11:27 AM The drilling history of the surrounding wells was reviewed sectioned to the planned sidetrack trajectory to construct a to understand the shale failure mechanisms and to mitigate pre-drill MEM. Using the information gleaned from the drill- the drilling risk. A robust mechanical earth model (MEM) was ing history of surrounding wells, a MW of 84 pounds per constructed using logs from one well of interest, which were cubic foot (pcf) was found optimum to minimize shear and calibrated by correlating the predictions to the drilling history bedding plane failure in the shale as well as differential stick- of surrounding wells. The rock properties were then curtain ing risk in the sandstone. During drilling, the cuttings at the

Fig. Fig. 3. 3. Plots3. PlotsPlots showing showingshowing the delicate control thethe delicateofdelicate drilling operation controlcontrol maintained ofof drillingdrilling throughout operationoperation the job. maintainedmaintained throughoutthroughout thethe job.job. Fig. 3. Plots showing the delicate control of drilling operation maintained throughout the job. 4 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 4 11/28/16 11:27 AM

This zone was isolated as the standoff from GOC was low

Fig. 4. Permeability graph showing the completion setup and design based on the lateral’s standoff distance from the gas cap. Fig. 4. Permeability graph showing the completion setup and design based on the lateral’s standoff distanceshale shaker from were the monitored gas cap. and correlated to the downhole circulating and tripping. These meters were installed in front logs. The well was drilled successfully with a MW of 84 pcf of each shale shaker and connected to software programmed without any drilling issue. to process the raw data coming from the meter and compare them to the theoretically calculated data so as to draw con- This zone was WELL PLACEMENTisolated as the clusions regarding the borehole condition. standoff from This automated analysis ensures a clean hole while drill- GOC was low The well was landed smoothly in the target sand, and the ing and lowers the chances of cuttings accumulations in the bottom boundary was continuously mapped, Fig. 1, to en- annulus. It also provides the foreman and engineers with sure successful geosteering in the target zone, while main- valuable information when they have to decide if a sweep taining the maximum planned true vertical depth (TVD) pill or short wiper trips are needed. Furthermore, it directly until total depth (TD), and keeping to a maximum dogleg of contributes to increasing the overall rate of pick penetration 8.5°/100 ft. It is quite important to watch out for the sand and minimizes the shale exposure, which is a critical parame- The ICD The ICD encourages influx for these intervals, which is layer’s changing dip and thickness,restricts as losing influx track of the ter forfarthest wellbore away stability from maintenance.the GOC. These meters assist in sand would impact the reservoir contactfor this length as well as drilling the hole and lowering the liner to the bottom safely Fig. 4. Permeability graph showinginterval. the completion setup and design based on the lateral’s standoff distanceraise hole frominstability the gasconcerns. cap. Standoff from as the acquisition unit allows the user to set up alarms to GOC is low. alert the field engineers if any anomalies occur. HOLE CLEANINGZone has been As seenShale in Fig. 2, the cuttings recovery was maintained isolated as the at around 90%, which helped in controlling the equivalent standoff from Cuttings analysisGOC is is < a10 tedious ft. and time-consuming task, but circulating density, optimizing pill frequency, monitoring automated approaches can significantly facilitate this pro- wellbore instability and undesired shale cuttings, maintain- cess and yield high quality results1. Cuttings flow meters ing good mud properties, and checking the average hole were deployed to monitor the rock cuttings returning to the size with each pill pumped to ensure no hole enlargement. surface during the different rig operations, such as drilling, This excellent hole cleaning and constant wellbore control

The ICD The ICD encourages influx for these intervals, which is restricts influx farthest away from the GOC. for this interval. Standoff from GOC is low. Fig. 5. The oil influx from the lateral for the different completion scenarios: ICD cases (red, green) and Zone has been Shale base caseisolated (blue). as the standoff from GOC is < 10 ft.

Fig. 5. The oil influx from the lateral for the different completion scenarios: ICD cases (red, green) and base case (blue). Fig. 5. The oil influx from the lateral for the different completion scenarios: ICD cases (red, green) and base case (blue). SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 5

87448araD3R1.indd 5 11/28/16 11:27 AM throughout the job are clearly reflected in plots of the drill- ing operation, Fig. 3.

Zone has been isolated as standoff from the INFLOW CONTROL DEVICE (ICD) COMPLETION DESIGN GOC is > 10 ft.

The ICD restricts influx for this interval as This offshore field, a dome-like anticlinal structure, is part standoff from the GOC is low.

of a larger field complex, which has an active gas drive. This The ICD encourages influx for this interval, which is gas cap introduced another geosteering challenge, given that away from the GOC.

the lateral standoff from the gas-oil contact (GOC) at the

target entry was < 10 ft TVD. Fig.Fig. 6. 6. The The tubing tubing flow profile flow for profile the different for completionthe different scenarios: completion ICD cases scenarios: (red, green) and base case (blue). When designing an inflow control device (ICD) comple- ICD cases (red, green) and base case (blue).

tion in a horizontal well with varying reservoir permeability,

the objective is to maximize influx from the lower permea- section, the lateral was divided into five compartments using Zone has been isolated five open hole packers. In the first compartment, 10 ICDs bility zonesas and standoff restrict from theinflux from the higher permeability zones. The reservoirGOC is > 10 permeability ft. in this well was found to of 1 × 1.6 mm were used to restrict the oil influx coming be in the range of 200 millidarcies (mD) to 800 mD. The from this interval. In the second compartment, 11 ICDs of 4 main challenge in this well was the fact that the GOC was × 2.5 mm were used in the interval, encouraging oil influx. very close to target entry, with < 10 ft TVD difference be- The third compartment was isolated with blank pipes as the The ICD restricts influx tween the two. Designing the forwell this to interval produce as with uniform gamma ray indicated shale in this zone. The fourth compart- standoff from the GOC ment comprised 12 ICDs of 3 × 2.5 mm, and the last com- influx would have led to gas breakthroughis low. from the heel of the well and eventually gas coning, given the high mobil- partment consisted of 15 ICDs of 2 × 2.5 mm nozzle setting, The ICD encourages in both cases encouraging oil influx. ity of gas. This production of gas from the gas capinflux would for this interval, which is Figure 5 shows the oil influx from the lateral, comparing drastically reduce the oil production rate from the well and Fig. 7. The tubing flow profile showing pressure drop for the two ICD scenarios. away from the GOC. the ICD cases (green, red) with the base case using a cased would eventually lead to a decline in reservoir pressure as and perforated liner (blue). the cap lost gas. As a result, the completion strategy was (1) Similarly, Fig. 6 shows the cumulative tubing flow profile to isolate the first section of the lateral, where the standoff from the GOC was < 10 ft TVD; (2) restrict the oil influx in for the ICD cases (green, red) vs. the base case using a cased Fig.the portion 6. The of tubing the well flow where profile the standoff for the was different ~10 ft TVD;completion and scenarios: perforated linerICD (blue).cases (red, green) and base caseand (3) (blue) encourage. oil influx where the standoff was > 10 ft Figure 7 shows the additional pressure drop due to com- TVD, Fig. 4. pletion, which was determined for the two ICD cases and Using this design strategy, the completion was designed compared against the drawdown across the sand face. with 48 nozzle-based ICDs. After isolation of the initial The case selected for adoption in drilling was ICD Case- 2, as this case better achieved the influx objectives of the

Fig. 7. The tubing flow profile showing pressure drop for the two ICD scenarios. Fig. 7. The tubing flow profile showing pressure drop for the two ICD scenarios.

6 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 6 11/28/16 11:27 AM 7” Casing

Baker HMC 4½” Hyd Liner Hanger ZXP Liner Top Packer 7” Casing

Baker HMC 4½” Hyd Liner Hanger ZXP Liner Top Packer 4½” 11.6# Casing

Baker PAC 4½” MPas 4½” MPas 4½” MPas 4½” MPas GPV TOW at Valve Packer Packer Packer Packer Shoe 5,190 ft TD at 4½” 11.6# Casing 12,400 ft

Baker PAC 4½” MPas 4½” MPas 4½” MPas 4½” MPas GPV TOW at Valve Packer Packer Packer Packer Shoe 5,190 ft 6⅛” Open Hole Baker ECPs 4½” 350 Micron Resflow ICD Screen 4½” 350 Micron Resflow ICD Screen MotherboreTD at 12,400 ft Fig. 8. The final well lower completion schematic showing the ICD screens along with the open hole isolation packers as run in the well.

Fig. 8. The final well lower completion schematic showing the ICD screens along with the open 6⅛”hole Open Hole 2 isolationcompletion. packers Figure 8 as is therun final in the lower well.Baker completion ECPs 4½” design, 350 Micron Resflowand ICD drillstring Screen . In4½” addition, 350 Micron there Resflow is ICD a needScreen to maintainMotherbore the showing the ICD screens and packer placements. integrity of the open hole swell packers and the liner’s exter-

nal casing packers, which means rotation of the completion Fig. 8. The final well lower completion schematic showing the ICD screens along with the open hole TORQUE AND DRAG string is not an option. Therefore, torque and drag problems isolation packers as run in the well. can occur. To reduce friction in any well, a good mud pro-

Drag is measured as the difference between the static weight gram design is important. In this well, the friction factor was

of the completion string and the tripping weight. In extended as low as 0.3, and the torque and drag simulation results,

reach holes, horizontal displacement usually is limited be- Fig. 9, ensured that the BHA design was suitable for this cause of frictional forces between the drilling and comple- type of well. tion string and the hole wall. Torque and drag modeling therefore is critical when estimating the capability of the rig

Fig. 9. Torque and drag simulation results.

Fig. 9. Torque and drag simulation results. Fig. 9. Torque and drag simulation results.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 7

87448araD3R1.indd 7 11/28/16 11:27 AM

Fig. 10. Buckling simulation results. Fig. 10. Buckling simulation results. RUN IN HOLE SIMULATION packers, ICD sand screens and the 4½” off-bottom cemented liner hanger system, and cemented successfully on depth. Several simulations were performed to determine the drill- This record was achieved with extensive engineering string was adequate to convey the ICD completion and liner planning, including understanding of the well directional hanger system to TD. Previous experience has shown that and stress orientation, of pre-drilling control points (dogleg, long open holes generate more friction during the tripping well placement, TVD control, etc.) and of the well trajec- run, so there is a greater risk the drillstring will get stuck. tory, which was optimized to avoid the 3D — build and Fortunately in this well, the excellent hole cleaning effort turn — well. Operation alertness during sidetracking was and good simulation guaranteed a successful run to TD very effective: Use of the right MW, optimization of wiper without any issues. The run in hole simulation results were trips to minimize shale exposure time and assessment of hole practically observed during the real run in hole trip with the cleaning using cuttings flow meters were keys to a successful drillstring. operation. The off-bottom cemented liner and ICD screen production equalizer assembly was designed to have flexibil- BUCKLING RISK ity while running in hole so it could reach TD without hang- ing up across any ledges. As previously described, the well trajectory was severe in the section from the window to the end of the heel. As a result, ACKNOWLEDGMENTS the wellbore was likely to generate more buckling on the completion string. This was compensated for by placing a The authors would like to thank the management of Saudi heavyweight drillpipe along this section to prevent helical Aramco and Schlumberger for their support and permission buckling and lockup. Figure 10 shows the simulation gener- to publish this article. ated for worst-case conditions during the run in hole, which This article was prepared for presentation at the 2016 allowed us to predict this scenario. Abu Dhabi International Petroleum Exhibition and Conference (ADIPEC), Abu Dhabi, UAE, November 7-10, CONCLUSIONS 2016.

Saudi Aramco’s Workover Department deployed the world’s REFERENCES longest 4½” off-bottom cemented liner and ICD screen production equalizer system in an offshore well. A total of 1. Marana, A.N., Papa, J.P., Ferreira, M.V.D., Miura, K. and 7,389 ft of lower completion string was run, including MPas Torres, F.A.C.: “An Intelligent System to Detect Drilling

8 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 8 11/28/16 11:27 AM Problems through Drilled-Cuttings-Return Analysis,” BIOGRAPHIES SPE paper 128916, presented at the IADC/SPE Drilling Qadir D. Looni is a Senior Drilling Conference and Exhibition, New Orleans, Louisiana, Engineer with Saudi Aramco’s February 2-4, 2010. Workover Engineering Department. He 2. Denney, D.: “Continuous Improvement Led to the Longest has more than 16 years of experience Horizontal Well,” Journal of Petroleum Technology, Vol. in offshore and onshore drilling, 61, No. 11, November 2009, pp. 55-56. completion, well testing and workover operations, including well planning, engineering and operations monitoring. Qadir’s areas of expertise include sidetracking workover wells to drill long, medium and short radius wells, running complex open completion strings and/or equalizer screens, off-bottom liners, expandable liners and smart completions, in addition to mechanical workovers for repairing casing leaks, wellhead work, changing completions and well testing. During his career, he has been involved in setting several world records in workovers and drilling. This includes the first time worldwide use of “Archer with underreamer,” successful application of an extra-long expandable cased hole liner and deployment of the world’s longest off-bottom liner with an internal control device screen assembly in an offshore environment. In addition to Qadir’s current role, he has worked as a Drilling Superintendent (A) with OMV in Romania, and as a Drilling Foreman and Engineering Supervisor (A) with Saudi Aramco. Qadir has been instrumental in training several young Saudi Drilling Engineers. As a member of the Society of Petroleum Engineers (SPE), he has published and presented several papers to international forums. Qadir received his B.S. degree in Petroleum Engineering from the University of Engineering & Technology, Lahore, Pakistan.

Mohammad J. Aljubran joined Saudi Aramco in mid-2015 as a Petroleum Engineer with the Drilling Technology Team of Saudi Aramco’s Exploration and Petroleum Engineering Center – Advanced Research Center (EXPEC ARC). He published and presented two Society of Petroleum Engineers (SPE) technical papers, was granted publication approval for three more papers, and filed two patent applications in the area of drilling and completion within the first year of his professional career. Mohammad is currently assigned to the Workover Engineering Department as a Workover Engineer. He is designing and planning workover operations in major Saudi Arabia offshore fields, such as Safaniya, Marjan and Zuluf. Mohammad was a lead member of the University of Oklahoma team that won first place at the 2015 SPE Drillbotics competition in automated rig design and construction. In 2015, he received his B.S. degree in Petroleum Engineering from the University of Oklahoma, Norman, OK.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 9

87448araD3R1.indd 9 11/28/16 11:27 AM Ahmed A. Al-Ramadhan joined Saudi Aly A. Emam is a Completions Aramco in 2005 as an Offshore Engineer in the Design and Evaluation Drilling Engineer working in the Services for Clients (DESC) Northern Area Drilling Engineering department with Schlumberger. His Department. Later, he joined the extensive experience includes core Drilling & Workover (D&WO) completions, liner hangers, sand Department as a Workover Engineer control, inflow control devices and and was appointed to be a Workover Engineering Unit smart completions. Aly began his career with Baker Oil Supervisor for offshore workover operations. During his Tools in 2002 as a Field Engineer for cased hole career, Ahmed introduced several innovative procedures completions. Assigned to Egypt, he performed all jobs that helped in salvaging wells under critical conditions, as related to well completions for offshore and onshore fields, well as optimization workover procedures that led to becoming a Completions Technical Engineer in 2005. significant cost reductions. In 2006, Aly joined Schlumberger as a Completions Besides his role as a Supervisor, Ahmed represented his Field Engineer and was assigned to Saudi Arabia. He de- department in several asset team meetings for different signed well completions for offshore and onshore fields offshore fields, and he has also assumed the responsibility and participated in starting the company’s upper and lower of serving as the Operational Excellence Coordinator for completions business in Saudi Arabia. In 2009, he was D&WO. assigned to Egypt, working as a Technical Sales Engineer. Ahmed received his B.S. degree in Applied Mechanical During this time, he was responsible for the sales and mar- Engineering from King Fahd University of Petroleum and keting of completions in the EEG Geomarket (Egypt, Syria, Minerals (KFUPM), Dhahran, Saudi Arabia. Jordan and Iraq). In 2010, Aly moved to Basra, Iraq, working as a Marius V. Neacsu is a Supervisor with Business Development Manager, engaged in starting Saudi Aramco’s Workover business relations and establishing a completions base in Engineering Department. He began Iraq. In 2011, he was assigned to the UAE, working as a his career in 1986 as a Tool Pusher Technical Sales Engineer, responsible for completions sales and Field Supervisor in the Moreni and marketing for Schlumberger throughout the UAE. Drilling Company in Romania. In 2012, Aly moved into his current position, assigned Marius also worked for the same to Saudi Arabia, where he works with Saudi Aramco’s company as a Drilling Engineer Specialist for 10 years. In Workover Engineering Department in a technical support 1998, he began work in Kuwait for Kuwait Oil Company role for completions operations. as a Rig Manager for drilling and workover operations on In 2002, he received his B.S. degree in Mechanical contractor rigs. In 2003, Marius started working for the Engineering from Alexandria University, Alexandria, Waha Oil Company in Libya as a Drilling Foreman. The Egypt. following year, he joined Repsol in Libya as a Drilling and Workover Consultant. After one year, Marius moved on and started working for Saudi Aramco as a Workover Engineer. Since joining Saudi Aramco, he has been heavily involved in major upstream onshore and offshore projects, making substantial contributions to those efforts. In 1986, he received his M.S. degree in Petroleum Engineering from the Oil & Gas Institute, Ploieşti, Romania.

10 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 10 11/28/16 11:27 AM Christian T. Mora is a Completions Engineer for the Design and Evaluation Services for Clients (DESC) department with Schlumberger. His extensive experience includes drill stem testing, tubing conveyed perforating, core completions, slick line, sand control and smart completions. Christian began his career with Schlumberger in 2004 as a Testing Field Engineer for well completions and productivity. His first assignment was in Poza Rica, Mexico, where he performed jobs related to drill stem testing, tubing conveyed perforating and well completions for offshore and onshore fields. In 2007, Christian was assigned to Ecuador as a Completions Field Engineer. He performed all jobs related to well completions and sand control. In 2010, Christian was promoted to Field Service Manager for Completions in Ecuador, where he managed field operations from Coca Base to several different client fields in the Amazon jungle. In 2012, Christian moved to his current position in Saudi Arabia, where he works with Saudi Aramco’s Workover Engineering Department in a technical support role for completions operations. In 2003, Christian received his B.S. degree in Mechanical Engineering from Army Polytechnic School, Quito, Ecuador.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 11

87448araD3R1.indd 11 11/28/16 11:27 AM the wheel is done, the engine needs work Reinventing transportation. We established our newest research center in Detroit, the heart of the U.S. automotive industry, to explore next-generation liquid fuels and innovative engine technologies to reduce emissions. The work being done has the potential to set a new course for the transportation industry and create greener, more efficient cars. There is a lot of talk about climate change and greenhouse gas emissions. At Saudi Aramco, we’re not just part of the conversation; we’re delivering real solutions.

12 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD3R1.indd 12 11/28/16 11:27 AM Development and Field Test of ESP Reliable Power Delivery System

Authors: Dr. Jinjiang Xiao, Randall A. Shepler, Yhossie S. Windiarto, Rob Fox and Stuart Parkinson

ABSTRACT 50% of the electrical failures in one key offshore field were related to the electrical connection below the ESP packer penetrator and to the motor lead extension (MLE), Fig. 1. In high hydrogen sulfide (H2S) and high-pressure/high tem- perature fields, the average run life of electrical submersible Moreover, severe service environments for ESP systems are pumps (ESPs) is still limited to three years. The dismantle generally defined as those applications where the wells have

inspection failure analysis results show that around 50% downhole hydrogen sulfide (H2S) concentrations above 5% of ESP failures are directly or indirectly related to electrical and carbon dioxide concentrations of more than 5%. In delivery problems concentrated at a distance of about 200 ft such applications, corrosion will also affect the run life of between the packer and the motor. This article presents the any equipment. The rate and extension of corrosion, how- results of a collaborative R&D effort to develop and field ever, can differ based on fluid partial pressures, bottom-hole test a reliable power delivery system (RPDS) with the goal temperature and well history — according to the evidence of of extending the average ESP run life from the current three previous premature corrosion. years to 10 years. To alleviate this issue, a joint research project was initi- The development focused on improving the reliability of ated to develop a reliable power delivery system (RPDS) as a key power delivery components, including the packer pene- solution to prolonging ESP run life in a harsh environment. trator, the motor lead extension (MLE) cable and the cable connection with the motor. The design not only integrates SYSTEM DESCRIPTION learnings from advanced completions and subsea technol- ogy, but also includes new concepts, features and materials. The RPDS configuration provides increased reliability over Connections that could be pressure tested in the field were im- the conventional MLE configuration. The key aspect of the plemented to ensure the proper makeup of field connections. RPDS is the fact that all primary interfaces are field testable Factory testing brought together a robust, highly accelerated and metal to metal, vastly reducing the potential failure

life test methodology to simulate a 10-year service life. modes caused by high H2S and rapid gas decompression. The Prototype components were designed, fabricated and scope of development included modification of the motor tested. These components were then integrated and subjected head to accept new feedthrough connectors, metal encased to a rigorous system integration test. After the comprehen- cable, a splice connector below the packer, a packer adapter sive factory tests, a field prototype system was built and and penetrator, and modification of the ESP cable connector. installed in an offshore well. The system was put into oper- Figure 2 shows the RPDS components that were addressed in ation and exceeded the field test’s success criteria of a mini- this R&D project.

mum run life of 180 days. Damaged/Burned For years, engineers and companies have battled with ESP Motor (15%) Intake and Pump Plugging (16%) reliability — specifically with the electrical problems that Round Power lie at the center of many failure causes. This article will dis- Cable (13%) cuss the development and field test of a new generation ESP power delivery technology with the potential of providing an extended run life.

MLE (20%) Wellhead INTRODUCTION Connector (8%)

The reliability of the electrical connections in electric sub- ESP Packer Penetrator (28%) mersible pump (ESP) subsurface equipment is extremely

1 Fig. 1. ESP failure modes. important . Records show, for instance, that approximately Fig. 1. ESP failure modes.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 13

ESP Cable Connector

87448araD4R1.indd 13 11/28/16 12:04 PM

Packer Penetrator

Packer Adapter

Packer

Encapsulated Cable

Splice Connector

Motor Feedthrough Connector

Motor Head

Fig. 2. RPDS stack up diagram showing components addressed in its development.

Damaged/Burned Motor (15%) Intake and Pump Plugging (16%)

Round Power Cable (13%)

MLE (20%) Wellhead Connector (8%)

ESP Packer Penetrator (28%)

Fig.The 1. new ESP feedthrough failure connectors modes. (“feedthrough” is a single-phase connections that feature primary metal-to-metal term used to refer to the components connecting the MLE sealing for the connectors to both the motor head interface with the motor) are based on field proven subsea technology and the new metal encased MLE. The termination to the modified to suit the RPDS’s power rating and architectural motor head was modified with a special insulator and a constraints. The motor head was configured with three motor winding crimp arrangement to accept the individual feedthrough connectors, Fig. 3. The modified motor head ESP Cable Connector also incorporates test ports to verify proper engagement of the metal-to-metal seal during field installation by means of a simple pressure test tool. The electrical feedthrough connectors are constructed in Packer Penetrator such a way to include an electron beam welded (EBW) in-

sulating contact pin assembly, which eliminates the need for Packer Adapter primary elastomeric seals to the main housing. The contact pin is manufactured as an integrally molded polyether ether

Packer ketone (PEEK) assembly, which incorporates a metallic col- lar that optimizes the electrical field and minimizes stress. Encapsulated Cable This collar also facilitates the final EBW process. The con-

Splice Connector nector body includes a metallic differential cone seal to seal into the motor head machining profile, Fig. 4. O-rings are incorporated in the sealing interface to facilitate a pressure test during installation. On the other end of the feedthrough connector, a metal-

to-metal seal is established with the metal encapsulated MLE cable by using a combination of a dual conic swage ferrule Motor Feedthrough Connector and a compression olive, Fig. 5. A retaining ring, torqued to

Motor Head apply compression to the ferrule, enables the setting of the swage using a pre-specified torque value. The compression Fig. 2. RPDS stack up diagram showing components addressed in its development. olive has internal and external O-rings, allowing the metal- Fig. 2. RPDS stack up diagram showing componentsto-metal seal addressed to be tested. in its development. The new MLE has a configuration compatible with each of the three phase connectors, which are metal encased and

connected individually to the motor head. Power is trans- Fig. 3. Motor head designed for the RPDS.

mitted via AWG #4 solid conductors, insulated with PEEK

Fig. 3. Motor head designed for the RPDS. material and protected with a unique fluoropolymer spline

Fig.Fig. 3. 3. Motor Motor head head designed designed for the for RPDS. the RPDS. jacket. The traditional jacket lead was replaced with an Metal cone interface with motor head

Metal cone interface with motor head Metal cone interface with motor head

Fig. 4. Feedthrough connector.

Fig. 4. Feedthrough connector. Fig. 4. Feedthrough connector. Fig. 4. Feedthrough connector.

et encste ce t tree connector conrton coere t te recement Pressure test port Fig. or 6. Metal o orne encapsulated coor MLE cable with three connector configuration, covered Pressure test port with the etreplacement encste 625 or 825 alloy (orange ce color). t tree connector conrton coere t te recement Pressure test port or o orne coor

Individual pressure test port

DualDual metal metalDual cone conemetal seal sealto cone to seal to Individual pressure test port metalmetal encapsulated metalencapsulated encapsulated MLE MLE MLE Primary metal-to-metal seal to encased MLE cable (3 individual terminations)

Fig.Fig. 5.5. 5. Connector’sConnector’s Connector’s m metal-to-metaletal-to-metal metal-to-metal seal seal with sealwith the thewith MLE, MLE, the including includingMLE, a includingtest a testport. port. a test port. Fig. 7.ce Splice connecton connection e eoadded te below cer the packer. Fig. 5. Connector’s metal-to-metal seal with the MLE, including a test port.

14 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

Primary metal-to-metal seal to encased MLE cable (3 individual terminations) 87448araD4R1.indd 14 11/28/16 11:41 AM

ce connecton e eo te cer

cer enetrtor

cer enetrtor

et encste ce t tree connector conrton coere t te recement or o orne coor

Individual pressure test port

Primary metal-to-metal seal to encased MLE cable (3 individual terminations)

ce connecton e eo te cer

Fig. 9. ESP power cable connector. oer ce connector only element to be terminated in the field. The splice connec- tor incorporates the same features and technology with re- Fig. 8. Packer penetrator. spect to metal-to-metal primary sealing to the cable in a way cer enetrtor identical to the motor feedthrough connectors. No potting or INCONEL® tube — 625 or 825 alloy, Fig. 6 — to make epoxy filling is required; metal-to-metal sealing is established the cable immune to rapid decompression effects and highly 10-Year Run between the connectors and the metal encapsulated cableLife at Goal resistant to corrosion effects due to a H S attack. Every in- 2 both ends and is pressure testable during assembly. dividual insulated conductor is metal encapsulated, and the Figure 8 shows the packer penetrator, which was designed

conductors are maintained together to facilitate handling and to mate with the packers of different vendors. It includes an installation in oerthe field ce with low connector profile Monel armor. offset adaptor assembly to avoid interference with the first To facilitate assembly of the system in the field, a “splice” pup joint/tubing installed above the packer. The bottom connection, Fig. 7, was added below the packer. This is the side of the penetrator connects with the metal encapsulated

Breakdown Voltage (kV) Voltage Breakdown 10-Year Run Life Goal

Number of Days (Aging)

t rom cceerte n test otte to roce erton cre

Breakdown Voltage (kV) Voltage Breakdown

Number of Days (Aging)

Fig. 10. Data from accelerated aging test plotted to produce a degradation curve. t rom cceerte n test otte to roce erton cre SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 15

87448araD4R1.indd 15 11/28/16 11:41 AM

connectors and is preassembled in the shop, which makes it Probability of 200 ft MLE Operating Voltage (kV) possible to ship the penetrator to the field with the packer Exceeding 10-Year Run Life and splice connector already terminated, connected and 26 6.26E-4% tested. It incorporates the same key features and sealing phi- 22 31.90% losophy as the motor feedthrough connectors. The upper section of the penetrator mates with the ESP 20 74.10% power cable. Despite this, this side of the connector will be 16 98.70% submerged in completion fluid and won’t see well fluids; it 8 99.99% also incorporates a metal-to-metal seal at the connector in- 4 99.99% terface to the packer penetrator. Figure 9 shows the ESP power cable connector, which is Table 1. A 10-year run life probability vs. operating voltage based on subsea dry mate technology. It interfaces with the a logarithmic scale, which best fits the expected degradation packer penetrator and allows the ESP cable to be connected curve, Fig. 10. This degradation curve was then extrapo- to the penetrator. It was modified to suit the AWG #2 main lated to 10 years, and a run life probability was established cable by adding a new cable seal and termination sleeve. based on the operating voltage. The result is the percent Some specific installation tools were developed as well, in- probability that the unit will operate for 10 years at various cluding a detailed installation and handling procedure for operating voltages. Table 1 shows that at operating voltages termination and testing both at the operational base and in equal to or less than 8 kV, there is a 99.99% probability of the field. exceeding a run life of 10 years. COMPONENT LEVEL TESTING — METAL ENCASED MLE FACTORY SYSTEM INTEGRATION TEST

Encapsulated Cable Cold Bending Test A qualification test of the RPDS was conducted in the factory. This test simulates the bending of the finished MLE over the sheave — the sheave is the wheel used in the installation Endurance Testing

of the ESP. The MLE goes from the spool and through the Operating Voltage (kV) Probability of 200 ft MLE Exceeding sheave before it is connected to the motor during installation This test proved the connector10-Year string’s Run Life functionality as a sys- tem26 under simulated maximum6.26E operating-4% conditions. Figure at the well site. Because in some parts of the world installa- 22 31.90% tion temperatures can be as low as -40 °C, test samples were 1120 is a picture of the endurance74.10% test setup. The test ran for 16 98.70% bent after being stored at -40 °C for a minimum of 24 hours. 408 cycles. The number of cycles99.99% was determined per the life- 4 99.99% This test was performed five times around a 16” mandrel, time testing of the encapsulated cable, which was relevant to which is less than half of the 42” sheave diameter. This e er rn e rot s oertn ote the connector system. Electrical verification tests were taken small reel was used as it was able to fit inside a freezer. Even prior, during — at the start of each cycle — and after the

though this MLE was tested using a 16” mandrel, the mini- test to ensure the connector string’s performance was within

mum bending diameter recommended for this product is still the acceptance criteria. The test parameters were: 42”. Results show no damage or loss of electrical properties because of the cold bending, so this test was considered successful. ESP Cable Connector

Vessel (covered in Encapsulated Accelerated High Temperature insulation) containing connector string Aging Test Motor interface simulation (hose contains oil, so termination components This test simulates material aging, which will occur during are immersed)

the product’s expected 10-year life at a maximum tempera- Fig. 11. nrnceEndurance testtest setsetup. ture of 176 °C. For this test we chose to simulate acceler- ated aging by testing the samples at 216 °C for 5 days, 10 Cycle Voltage (Phase) Current (Phase) days, 20 days, 40 days and 80 days. The Arrhenius model2 1 1-2 3 indicates that for every 10 °C increase in temperature, the

product life is reduced by half. Therefore, if the expected Cycle2 Voltage (Phase) Current1-3 (Phase) 2 1 1-2 3 life is 10 years at 176 °C, the expected life is reduced to 23 1-3 2-32 1 0.625 years (228 days) at 216 °C. By doubling the aging 3 2-3 1 Table 2. The wiring arrangement for the thermal cycles duration between each dataset, we could plot the results on e e rn rrnement or te term cces

16 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD4R1.indd 16 11/28/16 11:41 AM Cycles 1-13 Cycles 14-26 Cycles 27-40 Voltage on Phase 1 and 2 Voltage on Phase 1 and 3 Voltage on Phase 2 and 3 Current on Phase 3 Current on Phase 2 Current on Phase 1

Insulation (MΩ) Resistance

Date Measurement Taken

Fig. 12. IR profile during endurance test. roe rn enrnce test • Maximum working temperature: 150 °C remains well above the acceptance criteria of 10 GΩ. The • Maximum working pressure: 3,000 psi spikes in the profile indicate where the pressure test was un- dertaken before the holding period had stabilized, leading to • Maximum working voltage/current (phase to ground): temperature differences; heat has a direct impact on the IR 2.3 kV at 80 A (4 kV/√3) and so produced the spikes as shown. All test results before, during and after the endurance test were above acceptance The test wiring diagram was changed every 13 thermal cy- criteria. The connector string successfully passed this test. cles to ensure an even distribution of current/voltage across all phases. Table 2 lists the wiring arrangement for these Swage Ferrule Qualification thermal cycles. The temperature range of 66 °C to 150 °C was again set This test was specific to the swage and ferrule metal-to-metal per the encapsulated cable lifetime testing; however, it was sealing arrangement in use on the phase one connectors. This limited at 150 °C by the test vessel’s maximum temperature test was completed on test samples and not on the connector rating. Heating and cooling took approximately 8 hours each, string for practicality reasons. The objective of the test was with a 3-hour hold at each test temperature, which means to determine the swaging capabilities of the dual ferrule so that a heating and cooling cycle took two working days. as to find the average tensile load at which the ferrules slip Pressure cycles were included in the temperature hold period: on the cable. The ferrules were set at 50 lb/ft and put into for each cycle, pressure was decreased to 300 psi and held for the tensometer. Failure mode was cable extrusion through 1 minute, then pressure was increased to 3,000 psi and held the ferrule, with a mean load of 1,403 lb/ft and extension of for 1 minute. For the rest of the endurance test, during the 0.302” determined as being in excess of acceptance criteria. 3-hour hold period at each temperature, five of these pressure The sealing arrangement successfully passed this test.

cycles were undertaken, resulting in five low-pressure and five oc n rton test enc high-pressure cycles. Following these cycles, the pressure re- Power Frequency Testing mained at 3,000 psi. Approximately 155 pressure cycles were

conducted over the duration of the test. The power frequency test is a one-off test designed to confirm Figure 12 charts the insulation resistance (IR) profile the connector string’s capability to withstand a prolonged ap- across the 80 daysVibration of the endurance test’s duration.1.65 The g, 250 Hz pliedmax, HV 20 (4 hours Uo) potential — in radial at 60 Hz. direction. Electrical verification was IR profile is constantShock across all three phases, and69 g,while 1 ittime, perchecked axis by taking IR and continuity readings prior to and degrades across theResonance 40 cycles — Sweepwhich was expected2-200 — Hz, it 3 octavesafter the per test, minute to ensure the connector’s performance remained

e oc n rton test rmeters SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 17

87448araD4R1.indd 17 11/28/16 11:41 AM Cycles 1-13 Cycles 14-26 Cycles 27-40 Voltage on Phase 1 and 2 Voltage on Phase 1 and 3 Voltage on Phase 2 and 3 Current on Phase 3 Current on Phase 2 Current on Phase 1

Insulation (MΩ) Resistance

Date Measurement Taken

roe rn enrnce test

shock exposure, such as might be experienced either during transportation or in service. Figure 13 is a photo of the shock and vibration test bench. Proof that the system stayed within specifications electrically and that the swage and ferrule arrangement maintained its metal-to-metal sealing capability successfully was provided with results from the verification tests conducted before and after the shock and vibration test. Table 3 provides the parameters used in the test. The combined shock and vibration test was conducted along the “X” and “Y” axis. Between these tests, there was an opportunity to make further electrical verification tests. The IR profile and continuity were checked to confirm no degradation.

AC Breakdown Testing

Fig. 13. Shock and vibration test bench. oc n rton test enc This test was to determine the maximum AC voltage the connector string can withstand prior to breakdown. Because within the acceptance criteria. Current leakage was monitored it is a destructive test, it was conducted on a single-phase throughout the 4 hours. Even as the IR exceeded the 10 GΩ only. The voltage was increased, in 2 kV steps, from 10 kV acceptance criteria,Vibration continuity was maintained;1.65 current g, leak250- Hz max, 20 hours — in radial direction. age was steady atShock 2 mA throughout the test and69 within g, 1 actime,- per toaxis 18 kV, at which point the steps were reduced to single kV ceptable limits. ThisResonance test was passed Sweep successfully. 2-200 Hz, 3 octavesincrements. per minute Each step was held for a minimum of 1 minute. The system broke down at 22 kV, above the acceptance cri- eDC Hipot oc Test n rton test rmeters teria of 18.4 kV. Breakdown was attributed to the ESP cable connector. The connector string passed this test successfully. The DC voltage test is a one-off test designed to confirm the connector string’s capability to withstand DC HV (4 Uo) Rated Current Temperature Rise Test for 15 minutes. Electrical verification was checked by tak- ing pre-test and post-test IR and continuity readings. This The rated current temperature rise test is designed to confirm confirmed that the connector’s performance was within the that under maximum ambient operating temperature and acceptance criteria. This test was passed successfully. rated current conditions, the connector’s internal tempera- tures do not exceed the design limits of the insulation mate- Proof Pressure Testing rials used. This test is conducted in two stages, with 4 hours at 80 A followed by a 10-second current surge test at 140 This test was designed to confirm the connector string can A. Contact resistance checks were taken pre-test and post- withstand 1.5x working pressure while maintaining electrical test, to ensure the connector’s performance was within the function. Electrical verification tests — continuity, IR and proof voltage — were conducted before and after this pres- sure test. The test pressure of 4,500 psi was maintained for 15 minutes, with no issues found in either set of electrical checks. This test was successfully passed.

Shock and Vibration Test

This test was to verify that the termination of the cable to the connectors was not affected by random vibration and/or

1.65 g, 250 Hz max, 20 hours — Vibration in radial direction. Shock 69 g, 1 time, per axis Resonance Sweep 2-200 Hz, 3 octaves per minute

Table 3. Shock and vibration test parameters Fig. 14. Motor feedthrough connectors. otor eetro connectors

18 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD4R1.indd 18 11/28/16 11:41 AM

e nstton o te

otor eetro connectors

Fig. 15. Field installation of the RPDS. acceptance criteria. e The glassnstton transition of the o PEEK te insula- CONCLUSIONS tion material used on the connector’s contact pins is 170 °C, whereas the highest recorded temperature during the test was The RPDS maintained electrical integrity, ensuring proper 36 °C. Contact resistance values taken after both the 4-hour ESP performance for more than 250 days. This outcome pro- test and surge test were within acceptance criteria. This test vides enough confidence to install the system not only in this

was successfully passed. offshore field, but also in other fields with high H2S concen- trations. The project demonstrated the successful collabora- FIELD RESULTS tion of bringing subsea and advanced completion connector and penetrator technologies to ESP applications. The field test objective was to assess the performance of the RPDS for a period of at least 180 days to determine the ACKNOWLEDGMENTS robustness of the feedthrough connection and the usability of the system. The idea was also to determine the challenges The authors would like to thank management of Saudi that the field personnel might encounter during the handling, Aramco and Schlumberger/OneSubsea for their permission to installation and operation of this MLE and motor head sys- publish this article. tem, and to assess their effect on the short-term operational This article was presented at the SPE Saudi Arabia Section performance of the product. Moreover, the RPDS had to Annual Technical Symposium and Exhibition, al-Khobar, maintain good electrical readings during the test period. In Saudi Arabia, April 25-28, 2016. case an ESP failure arose, it was important that it not be as- sociated with the RPDS. REFERENCES The system was installed in an offshore well in February 2015 and has successfully operated for more than 250 days, 1. Al-Sadah, H.: “ESP Data Analysis to Enhance Electrical which exceeds the field test’s success criteria of a minimum Submersible Pump Run Life at Saudi Arabian Fields,” SPE 180-day run life. Figure 14 is a photo of the actual motor paper 173703, presented at the SPE Middle East Artificial feedthrough connectors. Figure 15 shows the field installa- Lift Conference and Exhibition, Manama, Bahrain, tion of the RPDS. The downhole equipment configuration November 26-27, 2014. includes a pump with 57 stages, a LSBSB-BSBSB severe ser- 2. Spahi, S. and Parsley, P.: “Reliable High Power ESP vice protector and a 300 HP motor. The ESP with the RPDS Tubing Hanger Connector Systems for Deepwater was function tested in March 2015 at various speeds and an Downhole Applications,” OTC paper 24138, presented estimated rate of 5,000 bbl of fluid per day (BFPD). at the Offshore Technology Conference, Houston, Texas, Upon rig movement, the ESP was commissioned and is May 6-9, 2013. running at an average of 55 Hz. During the life of the ESP, several trips were made to the subsea well for various reasons; however, the ESP, electrical connection system and RPDS still indicated a healthy system. The ESP is still running and con- tinuously producing between 5,000 BFPD to 8,000 BFPD.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 19

87448araD4R1.indd 19 11/28/16 11:41 AM BIOGRAPHIES Rob Fox joined Diamould Ltd. in 2006 and began working as the Dr. Jinjiang X. Xiao is a Petroleum OneSubsea Diamould Lead Engineer Engineering Consultant working in on the Reliable Power Delivery Saudi Aramco’s Exploration and System project, where he was Petroleum Engineering Center – responsible for the concept, design Advanced Research Center (EXPEC and qualification through to ARC). He is currently the focus area installation of the connector system. His previous work champion for artificial lift. experience includes working in the nuclear power Prior to joining Saudi Aramco in 2003, Jinjiang spent industry. During the last 10 years, Rob has progressed 10 years with Amoco and later BP-Amoco, working from being a Designer to serving as a Lead Engineer, on multiphase flow, flow assurance and deepwater working predominantly on power supply systems, both production engineering. on application engineering projects and new technology He received both his M.S. and Ph.D. degrees in development projects. Some of the recent projects he has Petroleum Engineering from the University of Tulsa, been involved in are located in the North Sea, South Tulsa, OK. China Sea and offshore Brazil. Rob is currently working toward gaining chartered Randall “Randy” A. Shepler is a engineer status. Petroleum Engineering Specialist In 2004, he received his B.S. degree in Mechanical working in the Artificial Lift Unit Engineering from Manchester Metropolitan University, within Saudi Aramco’s Northern Manchester, U.K. Area Production Engineering & Well Services Department. In addition to Stuart Parkinson is an Oil and Gas his 35 years of upstream experience, Connector Specialist currently Randy has authored and coauthored numerous technical working at OneSubsea, a papers with a primary emphasis on electric submersible Schlumberger company. Over the last pump (ESP) systems and associated completions and 24 years, he has worked on several production. technology world firsts related to Additionally, he has served as a chairman, committee electrical and optical connector member and master class training instructor for numerous systems for subsea, wellhead and downhole applications. Society of Petroleum Engineers (SPE) and PRAXIS During his career, Stuart has held several engineering and workshops and symposiums. management positions within Tronic (Expro), Diamould, Randy also has three patents and numerous Schlumberger and OneSubsea, in each position patent-pending inventions pertaining to his upstream introducing a number of unique and patented areas of expertise. technologies developed and implemented for field use. He received his B.S. degree in Petroleum Engineering He holds several patents for high voltage subsea wet and a B.S. degree in Business Administration from West mate connector technology. Virginia University, Morgantown, WV. Stuart received a Higher National Diploma from the University of Lancashire, Lancashire, U.K., and also holds Yhossie S. Windiarto is a Petroleum qualifications in offshore corrosion and metallurgy from Engineer working in the Artificial Lift Cranfield University, Cranfield, U.K. Unit within Saudi Aramco’s Production & Facility Development Department. Prior to joining Saudi Aramco in 2012, he worked for Schlumberger’s Artificial Lift Unit in several locations, including Oman, Indonesia and Saudi Arabia. In 1999, Yhossie received his B.S. degree in Petroleum Engineering from the University of National Development “Veteran” Yogyakarta, Indonesia.

20 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD4R1.indd 20 11/28/16 11:41 AM Drilling for the Next Generation of Multilateral Completion Systems

Authors: Yousif M. Abu Ahmad, Rami F. Saleh, Brett W. Bouldin, Robert J. Turner and Ali Bin Al-Sheikh

ABSTRACT interdependent and require continuous coordination, this article summarizes the modifications made to the drilling The objective of drilling a well is to accommodate the com- procedures to achieve the well’s objective. The emphasis is pletion string, which has been designed to optimize reservoir mainly on three operations: (1) Underreaming while drilling production. Streamlining a completion design for a mature the 8½” × 9” hole section, (2) Hole preparation prior to field maximizes the chances of achieving the well objective running the open hole completion, and (3) Implementation with minimal drilling challenges. When introducing a new of a unique cable avoidance procedure during conventional completion technology, that streamlining effort requires window milling operations. reevaluating the drilling procedures and practices to accom- modate the modified completion. This is especially true when BACKGROUND running new completion designs into an open hole, as op- Lateral 1

posed to a cased hole completion. The Saudi Aramco drilling Saudi Aramco has many Lateral 0 oil Lateralproducing 1 tight car- team was faced with the challenge of executing a multilateral Lateral 2

well plan that involved running three open hole lateral com- bonateLateral 0formations that pletions. The intent was to segment the open hole laterals respond Lateral 2 well to horizon- and then provide active control and monitoring capabilities tally drilled, open hole for each segment. Such a completion requires a complex completions. Experience power and telemetry system that can deliver electric power has shown that increased to all the segmented lateral completions, including the ones reservoir contact with a that are deployed through the milled windows. The open lower unit drawdown is hole laterals also had to be conditioned properly to success- an effective way to in- fully run the modified completion components in the open crease recovery and mini- 7” Liner hole without restriction and land the completions at the pre- mize 7” Liner water and gas coning planned target depths. in these wells. Maximum In this modification, the main power supply and telem- reservoir contact (MRC) etry conduit between the upper completion and the open wells were introduced in hole completions is a cable cemented behind the 7” liner. 2002 as commingled tri- lateral wells with over 5 This presented several challenges, such as how to run and 9⅝” Casing km of reservoir contact, 9⅝” Casing cement the liner and mill the windows without intersecting Figs. 1 and 2. By 2004, in- the cable, which would jeopardize the success of the entire Fig. 1. A plan view of a typical MRC well.

completion. The team developed new procedures to properly telligent completions were Fig.Fig. 1. 1. A Aplan plan view view ofof aa typical MRC MRC well. well. execute these operations and succeeded in maintaining the added with the placement

integrity of the power system and landing the completions

at depth. This was confirmed by conducting tests at various

stages throughout the implementation process. An integrated project, such as this one, requires focused

collaboration between the drilling and completion teams.

Whereas the traditional practice is to drill to total depth and

then complete the well, this unique completion design requires

a repeated drill-and-complete sequence as the well progresses.

In addition to highlighting the key factors in managing and implementing such a well, where various disciplines are Fig.Fig. 2.2. Plan of aa standardstandard trilateraltrilateral well. well.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 21

Fig. 2. Plan of a standard trilateral well.

87448araD5R1.indd 21 11/28/16 12:07 PM

of hydraulic internal control valves (H-ICVs) in the mother- development of the ManaraTM completion system, it determined bore to control each lateral1. Over time, MRC wells further that to supply power to, and communicate with, this many increased their length to over 10 km, but the number of ICVs tools downhole, a digital electrical ICV and umbilical telemetry was limited to four or five because each ICV required con- system was needed. The solution was to “branch” the umbil- nection to one or more hydraulic umbilical cables, extending ical cables into the laterals and utilize an inductive coupler. To from downhole to the surface. Also, packers and wellheads work, the solution required a downhole A/C power source and had a limited number of feedthrough penetrations to accom- communications telemetry. To ensure the reliability of the in- modate these umbilical cables. ductive coupler and the electric umbilical “backbone,” the sys- Early studies involving the production logs of over 60 tem was made completely analog; no digital circuitry was used single horizontal wells of 1 km in length showed an average between the surface box and the downhole ICVs. flow contribution from 85% of the horizontal section with Almost immediately afterward, Schlumberger began de- a generally uniform flow profile2. Additional contribution velopment of a next-generation ICV completion, called the was recorded from sections beyond the 1 km length, though Manara Station. The design called for an all-electric flow with decreasing returns1, 3. This was further confirmed on a control ICV and a flow monitoring mandrel to be integrated field basis by a trend showing diminished productivity as a into a single downhole module. Additional capabilities later function of increased reservoir contact; as reservoir contact included a venturi flow meter, a capacitive water cut sen- increased, compartment length also increased, with decreased sor and a variable choking flow orifice. While the Manara returns. Clearly this problem needed to be solved; if compart- Station itself was a very challenging module to design, build ments were limited to a 1 km length and reservoir contacts and qualify, it was a relatively straightforward module to began to exceed 10 km, over 10 ICVs per well would be deploy in the field. The inductive coupler was another matter required, not just four or five. Since it was more efficient to completely. The inductive coupler connector system required drill fewer, longer laterals than many short laterals, it would the co-location of male and female coils within +/- 2”. be necessary to compartmentalize the laterals as well as the Multilateral selective casing nipples and locks normally used motherbore. And even further increases in reservoir contact for whipstock location and orientation were used to accu- were desired — to 20 km and beyond. As the number of rately position the inductive couplers. The use of these tools drainage points in the well increased, flow measurement and was the main factor driving design of the completion needed control had to meet new high definition requirements if they to successfully implement an ERC Manara well, Fig. 3. were to achieve well production optimization. All these new To identify operational uncertainties prior to running the requirements made it overwhelmingly clear that a fundamen- full-scale system in a multilateral well, several sub-trials were tally new type of intelligent completion was needed. conducted to evaluate the well construction process, identify In 2007, Saudi Aramco initiated a project, called extreme risks and function test the system. Table 1 is a summary of reservoir contact (ERC), designed to expand the capabilities the trials. of intelligent completions to 50 or more control valves, with The Manara completion design offers independent mon- many valves per lateral4. As Schlumberger, in response, began itoring and control within the lateral segments, which are

Fig. 3. Completion schematic.

22 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD5R1.indd 22 11/28/16 12:07 PM Trial Objective Results 1 • Construction of wired junction. Established the procedure for window milling. • Run open hole completion in 10,000 ft lateral. Established the FF requirements for running the open 2 • Run two prototype Manara Stations in open hole. hole completion. Successfully function tested the system • Test connectivity of the well to the facility. connectivity. Tracked the liner and intercepted the electric umbilical 3 • Full system deployment. during window construction. Successfully implemented the cable avoidance technique 4 • Test cable avoidance technique. on three laterals. • Deploy full ManaraTM completion system in 5 Successfully deployed and function tested the system. motherbore. • Full ManaraTM completion system deployment in Successfully deployed the system in a trilateral well and 6 trilateral well. effectively function tested all the components.

Table 1. Summary of all field trials conducted

isolated with swell packers. The innovative design requires location has been identified, each lateral is drilled and one the deployment of what has been coined a “wired liner.” Manara completion, containing several stations, is installed The liner itself is used as a power and telemetry conduit per lateral. Figure 5 shows the general makeup of the lateral between the upper completion andCable the lateralwrapped segments, around pro- the linercompletion, to with the two Manara stations compartmental- viding a permanent path for ACallow power for and window communication. milling withoutized by a swell packer. The liner was “wired” by addingintercepting the following the elements cable. to the string: WIRED LINER

• Inductive couplers, which allow communication with The wired liner is equipped with an electric umbilical cable the lateral completions through induction. wrapped around the exterior of the liner string. The cable is • An electric umbilical, which interconnects the inductive secured to the liner with clamps that prevent radial and axial couplers, providing a means for power transmission and movement by the cable. To accommodate the increase in the ICC with communicationInductive between the segments and the upper effective diameter of the liner, the 8½” hole was underreamed completion.Coupler ICC with ICC with ICC with Inductive Inductive Inductive • An index casing coupling (ICC), which givesCoupler a Coupler Coupler

rotational and depth reference for use in locating the umbilical. Fig. 4. Liner configuration wired with inductive couplers and a wrapped umbilical (figure provided by Schlumberger).The wired liner is considered the backbone of the com- Cable wrapped around the liner to allow for window milling without pletion as it permits lateral branching while maintaining intercepting the cable. connectivity with the upper completion, Fig. 4. This new concept of a connected liner, however, presents several chal-

lenges when it comes to running and cementing the liner, ICC with Inductive Coupler ICC with ICC with ICC with window milling and bottom-hole assembly (BHA) separation Inductive Inductive Inductive Coupler Coupler Coupler

from the liner as it drills the lateral without intercepting the

electric umbilical. Fig.Fig. 4. 4. Liner Liner configuration configuration wired with inductivewired with couplers inductive and a wrappedcouplers umbilical and a(figure wrapped provided umbilical by Schlumberger).(figure provided by Schlumberger).

Once the liner is deployed and the umbilical cable

Manara Swell Manara Male SwellSwell ElectricElectric Manara Swell Manara LocatorLocator Packer Male Station Packer Station Coupler PackerPacker UmbilicalUmbilical Station Packer Station Packer Coupler

Male Swell Electric Manara Swell Manara Packer Male Swell Electric Manara Swell Manara LocatorLocator Umbilical Station Packer Station Packer Coupler PackerPacker Umbilical Station Packer Station

Fig. 5. Lateral completion showing the ICC locator, a coupler, the umbilical, and Manara Stations separated by swell packers (figure provided by Schlumberger). Fig. 5. Lateral completion showing the ICC locator, a coupler, the umbilical, and Manara Stations separated by swell packers (figure provided by Schlumberger). Fig. 5. Lateral completion showing the ICC locator, a coupler, the umbilical, and Manara Stations separated by swell packers (figure provided by Schlumberger). SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 23

87448araD5R1.indd 23 11/28/16 12:07 PM

to 9”. Enlarging the hole was necessary for several reasons During the first run, an anchor packer is run in hole, and related to the integrity and successful deployment of the wired then oriented and spaced out in reference to the cable loca- liner. First of all, when pumping the cement for the new liner tion. Ideally, the window should be cut 180° from the cable, in the smaller hole, the annular restrictions created by the on the opposite side of the liner, minimizing the chances of clamps and cable will result in high equivalent circulating den- contact with the umbilical. The whipstock is set in the sec- sities acting on the borehole, which may result in fracturing ond run and the window milling operation is initiated. the formation and lead to severe losses. Another reason for Once the window depth and orientation are selected, the hole enlargement is the presence of reactive shales; enlarging window can be milled with minimal chances of intercepting the hole will help to overcome the time dependent swelling of the cable. The challenge arises when trying to separate from the shales during liner deployment and keep them from creat- the liner with no reliable directional readings due to the ing an obstruction. Finally, the additional clearance between magnetic interference created by the BHA’s proximity to the the open hole and the liner will accommodate the increase in metallic body of the liner. This essentially creates a blind the effective diameter of the liner and its accessories. spot for the first ± 200 ft of drilling beyond the window. In Trial 3, a two-run operation was conducted to enlarge The risk materialized in Trial 3, where the window was suc- the 8½” hole. First, the hole was drilled, then a separate cessfully milled but separation was not achieved. The BHA underreaming run was performed to enlarge the hole. This instead tracked the liner and intercepted the umbilical, lead- was found to be time-consuming and provided no additional ing to aborting the project and running a conventional intel- benefits with regards to hole quality. For the subject well, ligent completion. To address the lack of directional readings the drilling and underreaming were performed in a single during separation, a method for monitoring the separation in run, resulting in a more than 50% reduction in time when real time needed to be established. compared to the two-run method. Based on the lessons learned from previous wells, a fit-for- The specific 9” gauge of the enlarged hole was selected purpose procedure was implemented to ensure proper separa- based on previous trials. When one well was enlarged to tion from the liner without damaging the umbilical cable. At 10”, it created an excessive gap between the liner and open the core of the procedure is the ability to qualitatively monitor hole, resulting in poor cement distribution. The excessive ce- the separation of the drilling BHA from the liner in real time, ment thickness on one side of the liner may contribute to the enabling any immediate adjustments in the directional drilling BHA tracking the liner during window milling and subse- BHA steering ratio as needed. This was achieved by deploying quent lateral drilling operations. This scenario was observed a near bit, real-time resistivity tool in the directional drilling during Trial 3, when the BHA drilled the cement around the BHA to monitor the resistivity trends. The tool registers the liner, rather than the harder compressive strength rock, and difference between the liner resistivity and formation resistiv- so tracked the liner, which resulted in intercepting the elec- ity as the drilling BHA progresses in depth, Fig. 6. The mon- tric umbilical. This led to aborting the completion plan and itoring employed a matrix to accelerate the decision making running the contingency conventional intelligent completion. process while drilling past the window. Based on the lessons learned from the trials, the under- The first point, Pt 1, shows no resistivity reading, indi- reamed hole size was reduced to 9”, which provides enough cating that the tool is in the liner. Once the tool exits the clearance to run the liner with all its accessories without window, a signature spike in resistivity is observed, Pt 2. jeopardizing the cement quality. This spike is the first positive indication of separation from WINDOW MILLING AND CABLE AVOIDANCE the liner. The resistivity should then move closer to the formation resistivity as drilling progresses. Successful sep- The fact that an umbilical cable is wrapped around the liner aration is signaled when the resistivity reading matches the makes it challenging to mill the window while keeping the formation readings. If a trend is observed of declining or sta- cable intact. One of the functions of the ICCs is to permit bilizing resistivity prior to reaching the target value, several the identification of the cable location. Dedicated cable lo- Casing Resistivity at Formation cating runs were conducted for each of the ICCs. Based on Resistivity Window Exit Resistivity the information, the whipstock orientation was selected with a contingency for each window. Because the depth restriction of the windows limited the options of window orientation, a proper window milling procedure needed to be developed and followed to ensure successful drilling exit and separation from the liner and its umbilical cable in the first milling at- tempt. This procedure is described as follows. Once the cable’s location behind the liner is confirmed, Fig. 6. Real-time resistivity readings as a lateral is drilled. window milling is carried out using a two-trip process. Fig. 6. Real-time resistivity readings as a lateral is drilled.

24 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD5R1.indd 24 11/28/16 12:07 PM

Fig. 7. Tripping the pipe with circulation allowed the cuttings to be carried to the surface improving the FF by about 0.025.

Fig. 8. Reaming trips smoothened the borehole and enabled more cuttings to be lifted to the surface, improving the FF by about 0.05.

contingencies can be applied, such as increasing the steering • Reaming trip. ratio on the directional assembly, or changing the directional • Increased fluids viscosity. BHA from a rotary steerable system to a steerable motor for a more aggressive build. • Use of high performance lubricant. This cable avoidance method when milling the window and separating from the liner was implemented successfully The team’s objective was to understand the effect of each for the two windows on the subject well, and had proved of these elements on the FF. The results clearly showed that successful three times on the previous trial well, giving suffi- wiper trips, combined with circulation at specific critical cient evidence that the technique is reliable. Once separation points, and reaming both had the most impact in reducing the is achieved, the lateral is drilled to target depth by imple- FFs, Figs. 7 and 8, respectively. This can be mainly attributed menting field specific drilling practices. to the agitation and removal of the accumulated cuttings. The increase in fluids viscosity also reduced the FF, but HOLE PREPARATION to a lesser extent when compared to the wiper trips and reaming, Fig. 9. The viscosity increase reduces the drag ex- Routine operations in the field do not require running open perienced by the drillstring by enhancing the fluid’s cuttings hole completions. Therefore, hole geometry and friction suspension capabilities and preventing the accumulation of factors (FFs) are not a concern once the lateral is drilled and cuttings on the low side of the wellbore. the drilling assembly is pulled out of the hole.Resistivity This is not atthe The use of a high performanceFormation lubricant was also evalu- Casing ated and found to consistently reduce the FF by 0.05. The case for the subject Resistivitywell. Based on the simulations,Window a low Exit FF Resistivity was targeted as required to successfully land the completion addition of the lubricant was used as a contingency when all at the required depth. To keep the FF low entails drilling other techniques fell short of achieving the low FFs needed a smooth wellbore with minimal doglegs and few varia- to accommodate running the completion in the open hole. tions in diameter while maximizing hole cleaning efficiency. The drilling drive mechanism and drilling practices play a SUCCESSFUL COLLABORATION big role when it comes to hole quality and hole cleaning. Though drilling practices in the subject field are well estab- The deployment of the new system necessitated a revision of lished, the objective had to shift from maximizing the rate of the entire drilling and completion dynamic. Principally, the penetration to minimizing the FF5. conventional drilling operations for completing the process The targeted low FF was accomplished by applying sev- of well construction was changed to a drill-and-complete eral techniques. To understand the effectiveness of each of cycle for each lateral. This meant drilling and completing the the techniques, FF readings were taken and compared after lateral, testing the completion system prior to moving on to each one was implemented. This was time-consuming but the next lateral, and then repeating the process. The nature necessary to identify best practices for future wells. The of the project required a unique collaboration, as the line study focused on the following hole preparation techniques: between drilling and completion was blurred, making the Fig. 6. Real-time resistivity readings as a lateralresponsibilities is drilled. and targets interdependent for the sake of the • Wiper trips — tripping pipe without rotating — and project’s success. circulating at critical locations. One example would be window milling. In common

Fig. 7. Tripping the pipe with circulation allowed the cuttings to be carried to the surface improving the FF by about 0.025.

Fig. 7. Tripping the pipe with circulation allowed the cuttingsSAUDI ARAMCO to be JOURNAL carried OF TECHNOLOGY to the surfaceFALL 2016 25 improving the FF by about 0.025.

87448araD5R1.indd 25 11/28/16 12:07 PM

Fig. 8. Reaming trips smoothened the borehole and enabled more cuttings to be lifted to the surface, improving the FF by about 0.05.

Casing Resistivity at Formation Resistivity Window Exit Resistivity

Fig. 6. Real-time resistivity readings as a lateral is drilled.

Fig. 7. Tripping the pipe with circulation allowed the cuttings to be carried to the surface improving the FF by about 0.025.

Fig. 8. Reaming trips smoothened the borehole and enabled more cuttings to be lifted to the surface, improving the FF by about 0.05. Fig. 8. Reaming trips smoothened the borehole and enabled more cuttings to be lifted to the surface, improving the FF by about 0.05.

Fig. 9. Increasing the fluid viscosity gave better cutting suspension, which slightly improved the FF. Fig. 9. Increasing the fluid viscosity gave better cutting suspension, which slightly improved the FF. practice the whipstock engineer monitors the milling oper- CONCLUSIONS ation to ensure success. In the case of the subject well, both the multilateral engineer and the completion engineer were One of the primary goals of the team was to establish spe- involved in the window milling operation to ensure that cific procedures and practices that can be implemented on the completion passed through the window. In many cases, future wells. Through the lessons learned from the subject the completion requirements directed the drilling practices, well and the preceding trials, the team was able to produce necessitating a new level of compromise between different an effective step-by-step methodology for drilling similar segments within the team. This was crucial, for example, wells and set the standards that will open the door for fur- to achieving the desired low FF in the 6⅛” open hole. The ther efficiency and optimization. drilling practices, clean out operations and BHA design

where optimized to reduce the FF while sacrificing the rate ACKNOWLEDGMENTS of penetration. The stated examples represent a common theme seen The authors would like to thank the management of Saudi throughout the project. Collaboration and alignment of Aramco and Schlumberger for their support and permission goals and objectives among the parties was achieved with to publish this article. effective communication involving extensive meetings and This article was presented at the Offshore Technology correspondence. Conference, Houston, Texas, May 2-5, 2016.

26 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD5R1.indd 26 11/28/16 12:07 PM

REFERENCES BIOGRAPHIES

Yousif M. Abu Ahmad joined Saudi 1. Salamy, S.P., Al-Mubarak, H.K., Hembling, D.E. and Aramco in 2005 as a Tool Pusher Al-Ghamdi, M.S.: “Deployed Smart Technologies Enablers and Foreman working in the South- for Improving Performance in Tight Reservoirs — Case: ern Area Drilling Operations Depart- Shaybah Field, Saudi Arabia,” SPE paper 99281, presented ment. His field experience includes at the Intelligent Energy Conference and Exhibition, onshore, offshore and workover op- Amsterdam, The Netherlands, April 11-13, 2006. erations. In 2010, Yousif was selected 2. Saleri, N.G.: “Maximum Reservoir Contact Wells: for a one year out-of-Kingdom developmental assignment with Baker Hughes, gaining practical experience from Rewriting the Rules of the Subsurface,” paper presented working in U.S. drilling operations. at the SPE Gulf Coast Section, Houston, Texas, September He has 11 years of experience, including work as a 26, 2002. Rig Foreman, Drilling Optimization Engineer, Workover 3. Saleri, N.G., Salamy, S.P., Mubarak, H.K., Sadler, R.K., Engineer and Senior Drilling Engineer in Saudi Aramco’s Dossary, A.S. and Muraikhi, A.J.: “SHAYBAH-220: Exploration and Oil Drilling Engineering Department. A Maximum Reservoir Contact (MRC) Well and Its Yousif is currently on a developmental assignment with the Implications for Developing Tight Facies Reservoirs,” Contract Administration Division as a Contract Advisor. SPE paper 81487, presented at the Middle East Oil Show, He was a key member in the Manara team and was a major contributor to the planning, execution and Bahrain, June 9-12, 2003. successful delivery of the first extreme reservoir contact 4. Saggaf, M.M.: “A Vision for Future Upstream (ERC Manara) well in the world. Technologies,” Journal of Petroleum Technology, Vol. 60, Yousif has published several technical papers on No. 3, March 2008, pp. 54-98. drilling optimization, technologies and field studies. He is 5. Bouldin, B.W., Verma, C., Dyer, S., Singh, P. and Oliveira, an active member of the Society of Petroleum Engineers T.: “Powering Through a Lateral Junction for ERC (SPE) and is a certified SPE Petroleum Engineer. In 2004, he received his B.S. degree in Science in Mining Wells — Is It Really a Step Too Far?” SPE paper 160856, Engineering from the University of Arizona, Tucson, AZ. presented at the SPE Saudi Arabia Section Technical Symposium and Exhibition, al-Khobar, Saudi Arabia, April Rami F. Saleh is a Drilling Engineer- 8-11, 2012. ing Division Head in the Exploration and Oil Drilling Engineering Depart- ment of Saudi Aramco. He has over 15 years of experience in drilling en- gineering and operations. This experi- ence ranges from onshore and offshore rigs to deep gas drilling, exploration and shale gas drilling, and oil increment development, including the giant Khurais field, and most recently, the Shaybah field expansion increment. Rami has published several technical papers related to drilling optimization, and he participated in several international technical panels on the subject of drilling automation. He is an active member of the Society of Petroleum Engineers (SPE) and the Saudi Association of Energy Economics, and is a certified SPE Petroleum Engineer. Rami received his B.S. degree in Mechanical Engineering from Tufts University, Medford, MA, and his M.S. degree in Petroleum Engineering from Institut Francais du Petrole (IFP), Rueil-Malmaison, France.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 27

87448araD5R1.indd 27 11/28/16 12:07 PM Brett W. Bouldin is a Petroleum Engi- neer Consultant with Saudi Aramco and has been with the company for 6 years. He has over 33 total years of product development experience in the completions industry, first with Baker Hughes, then as a founding member of WellDynamics/Halliburton. Brett currently ini- tiates and manages completions development projects for the Exploration and Petroleum Engineering Center – Ad- vanced Research Center (EXPEC ARC), focusing on new tools that improve production recovery, mainly dealing with next-generation intelligent completion systems. He has authored or coauthored 7 technical papers and 35 U.S. patents. Brett received his B.S. degree in Industrial Engineering from Texas A&M University, College Station, TX, and is a Registered Professional Engineer in Texas.

Robert “Rob” J. Turner is a Petroleum Engineering Specialist in the Advanced Completions focus area of the Produc- tion Technology Team of Saudi Aram- co’s Exploration and Petroleum Engineering Center – Advanced Research Center (EXPEC ARC). During his 30 years in the oil industry, he has worked in the U.K., Australia and Southeast Asia for operators Shell, BHP Billiton and Chev- ron. This has enabled Rob to gain experience in land, platform and subsea operations for a variety of oil and gas, brownfield and greenfield projects. He has held positions in all life cycle stages of hydrocarbon development, including project devel- opment, completion operations and reservoir management. For the last 10 years, Rob has specialized in smart fields, from smart well justification, functional specification, project engineering and installation to commissioning and reservoir management. Prior to joining Saudi Aramco, he was the smart fields technical expert for Brunei Shell Petroleum, the leading operator in the Asian region for smart well installations. Rob received his B.S. degree in Chemical Engineering from Leeds University, Leeds, U.K., and a M.Eng. degree in Petroleum Engineering from Heriot-Watt University, Edinburgh, U.K.

Ali Bin Al-Sheikh is a Completion Project Engineer working for Schlum- berger in Houston, TX. His 9 years of oil field experience include running intelligent completions, permanent downhole monitoring systems and new completions technology. For the past 5 years, Ali has been working as a Manara Project Manager, supporting the engineering team and the execu- tion team as they carry out all Manara installation-related operations. He received his B.S. degree in Mechanical Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.

28 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD5R1.indd 28 11/28/16 12:07 PM Innovative Step Change in Drilling Efficiency for Medium Radius Reentry Deep Gas Wells with a High Build Rate Rotary Steerable System

Authors: Abdul Halim Ab Hamid, Verdy L. Siregar, Ali N. Al-BinAli, Mohamed E. Khalil, Ayman Ghazzawi, Omar T.A. Ashraf and Muhammad S. Balka

ABSTRACT of these tools can only be used outside the zone of magnetic interference once sufficient separation from the motherbore Generally, deep gas workover and reentry wells in Saudi has been achieved. Moreover, consistent doglegs of more than Arabia are kicked off in the Sudair formation through a 14°/100 ft were recorded using the HRSS; the maximum dog- whipstock because the overlying base Jilh dolomite (BJD) leg was 17.44°/100 ft. formation can flow with high pressure, which jeopardizes Since then, this concept has been applied successfully to well control. Whipstocks are set deep in the 9⅝” casing, other vertical reentry wells and at existing inclinations in the after which the 8⅜” and 5⅞” holes are drilled to access the 8⅜” and 5⅞” sections in Saudi Arabia and worldwide. The targeted lower carbonate and sand reservoirs. Deeper kick- scope of the article is limited to Saudi Arabian deep gas wells offs also avoid contact across the water-bearing Carbonate only. The average ROP across the build section showed a A reservoir and instead aim for displacement across the 137% improvement over the ROP for conventional motor Carbonate B or C reservoirs. Isolation from the Carbonate A bottom-hole assemblies (BHAs) across similar build sections. reservoir is important for multistage fracturing completions Eliminating the 8⅜” section, avoiding the hazards of drilling as they are still not proven for maintaining the long-term in the Jilh and Sudair formations, saving the motor trip to isolation of water-bearing zones. kickoff from the whipstock and improving the ROP resulted Regardless of the deeper whipstock setting, the high dogleg in significant savings. This step change in drilling performance requirements of such wells exceed the capabilities of conven- was realized by acquiring a thorough understanding of local tional rotary steerable systems (RSSs). Conventional steerable drilling conditions and conducting the in-depth analysis that motors with a high bend housing and the capacity to achieve enabled efficient execution. 70% to 80% of the sliding mode of drilling have been the only option to drill wells with such high dogleg severity (DLS) — INTRODUCTION 100 ft. Drilling medium radius wells with a conventional motor assembly, however, requires multiple runs, several wiper trips Most of the deep gas wells in Saudi Arabia are drilled in the to clean the hole and multiple reaming trips before running gigantic South Ghawar field. The main gas producing zones are the Late Permian Carbonate B and C stacked carbonate the liner. These operations show poor drilling efficiency due to reservoirs1. Wells drilled here incorporate three types of cas- their slow rates of penetration (ROP) and numerous bit trips. ing designs, namely MK1, K1 and K2. Figure 1 shows the A high build rate rotary steerable system (HRSS) was in- troduced as a solution to such challenges in these workover wells’ 8⅜” and 5⅞” sections. While the HRSS technology has been used before, this was the first time the HRSS kicked off vertically from a whipstock in a well in Saudi Arabia, setting a worldwide milestone. The new technology allowed the kickoff point to be pushed further into the Sudair for- mation near the Sudair dolomite, reducing the costly risk from Jilh pressure. The step change also provided the option to slim the hole by eliminating the 8⅜” hole size, so kickoff was done in the 7” liner. Deployment of the HRSS, which allowed direct kickoff from a whipstock set vertically, eliminated the need for a ded- icated steerable motor assembly run. Direct kickoff also meant eliminating the need for a gyro tool, used to enhance the

Fig. 1. The differences between the K2, K1 and MK-1 casing designs. steerability of conventional RSS tools; the steering capacity Fig. 1. The differences between the K2, K1 and MK-1 casing designs.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 29 300

250

87448araD6R1.indd 29 200 11/28/16 12:25 PM

150

100 Time (hrs) 50

0 Well F Well G Well H Well J Well K

Time to kick off (hrs) Time between POOH and P/U RSS (hrs)

Fig. 2. Time taken to drill sidetracks in typical reentry wells using a conventional motor, RSS BHAs, and gyro tools.

difference among the three casing designs. In a typical K2 well found in the Jilh formation isn’t present. It can be drilled with profile, the 13⅜” casing is set from the surface downhole 30 a relatively low dogleg severity (DLS) of 3°/100 ft to 5°/100 ft into the base Jilh dolomite (BJD) formation, followed by ft. The 8⅜” hole section can be 2,000 ft to 4,000 ft in length the drilling of a 12” section into the Sudair formation to the with the true vertical depth (TVD) at around 10,000 ft to top of the carbonates, and then the drilling of an 8⅜” section 13,000 ft. The 5⅞” section can be drilled horizontally, pene- across the carbonates and sand reservoirs. trating the potential reservoir for a length of about 2,000 ft to The MK1 design, which is the “slim design,” sets a 9⅝” 3,000 ft. Some of the challenges presented by this design are casing from the surface into the BJD formation. From there, an the cutting of a window through the 9⅝” and 13⅜” casings, 8⅜” section is drilled to the top of the carbonates, after which and the hazards of further drilling through the unstable BJD a 5⅞” section is drilled into the target carbonates and sand res- formation pressures, which can be avoided if the kickoff is ini- ervoirs. The MK1 design is cost-effective, which is characteristic tiated deeper in the Sudair or Sudair dolomite formations. of slim designs, since there is one less hole size to drill and one The second option, the deeper kickoff design, is more less casing size. On average, it takes approximately 100 days to challenging because the intended target has to be achieved drill a K2 design compared to 80 days for a MK1 design2. in a relatively short TVD margin. The required DLS is more The MK1 design obviously makes better financial sense com- demanding: in excess of 10°/100 ft. Furthermore, the kickoff pared to the K2 design, but occasionally higher pressures are is usually done in the Sudair formation, which is plastic shale encountered in the BJD formation, which then limit drilling of having the ability to ball up the bit, reducing DLS capabilities the 8⅜” section to the deeper target Carbonate B or C forma- up to 30% and reducing the rate of penetration (ROP) more tions. The K2 design is particularly advantageous in these situ- than 50%. The second option becomes even more challenging ations because it provides the opportunity to set an additional in the case of an MK1 design where Jilh pressure has been en- casing string. That means, later on when these wells are deemed countered, which exacerbates the difficulty of drilling multiple no longer productive, though they are in a known good reser- formations with a high mud weight (MW). In such compli- voir, reentry to drill for different targets or install other comple- cated scenarios, two options are available: tion types becomes a viable objective to pursue. By drilling reentry wells, almost 60% of a new well’s cost 1. Milling the 7” liner below the 9⅝” casing shoe to create can be saved. To design a conventional reentry sidetrack well room for re-drilling the 8⅜” section. This option risks and avoid underbalanced drilling options, it is important to opening up the borehole to the high-pressure Jilh forma- understand the initial well design — either MK1 or K2 — that tion. Should Jilh pressure be encountered, drilling has to be was used to drill the pilot hole. The following design options are preceded with high MW and often can lead to losses in the possible: deeper Sudair or Sudair dolomite formations. The added risks of drilling with extreme overbalance in carbonates 1. Reentry well design with two hole sizes, meaning the cased make this option infeasible. hole is sidetracked from inside the 9⅝” casing by setting a 2. The other option is to kickoff deeper in the Carbonate B whipstock. The sidetracking usually takes place in the BJD or C formations and land in the target reservoir. This op- formation, and the 8⅜” hole size is drilled to land inside tion, with its limited TVD availability, effectively requires the gas-bearing Carbonate B or C formation, where the 7” drilling a short radius well with a DLS above 30°/100 liner will be set and cemented. The drilling is then contin- ft. Alternatively, kicking off at a shallower depth in the ued in a 5⅞” open hole to the well’s total depth (TD) in Sudair or Sudair dolomite allows a single hole size of 5⅞” the zone of interest. to be drilled. 2. Reentry design in one or two hole sizes with a deeper side- track, meaning that the sidetracking from inside the 7” or Traditionally, these high doglegs were handled using a di- 9⅝” liner using a whipstock occurs at a greater depth, cor- rectional steerable motor assembly with high bend settings. responding to deeper formations, i.e., the Sudair and Sudair Design option 1, with the cased hole sidetracked from inside dolomite. The well can be drilled either as a 8⅜” then 5⅞” the 9⅝” casing in the BJD formation, even with its reduced hole or as a single 5⅞” hole size to TD. This option is usually DLS requirement, needed a minimum of two runs to drill with selected to avoid the potential Jilh high pressure that was a conventional steerable motor assembly or a combination encountered in the 12” section of the original wellbore — of steerable motor and conventional rotary steerable system or in the offset wells — by sidetracking at a greater depth. (RSS) assembly. At the sidetrack/kickoff point, however, with less than 20 ft of the rathole, the conventional RSS tools CHALLENGES were still under magnetic interference near the casing and whipstock. For this reason, at least 300 ft to 400 ft had to Each design has a particular set of challenges. The first design be drilled with a steerable motor assembly to gain separation is easier to pursue in cases where the high pressure sometimes from the pilot hole and reach a depth free from magnetic

30 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD6R1.indd 30 11/28/16 12:25 PM interference where the RSS could be used properly. Design op- from the Sudair or Sudair dolomite formations, saves up to 4 tion 2, with the deeper kickoff, was only possible using con- days by avoiding the well control issues of the BJD formation; ventional steerable motor assemblies and had multiple risks of because the formation is already covered by a casing, the in- heavy sliding and a high number of trips. flow is avoided. On the other hand, if the kickoff is performed In addition to these challenges, sidetracking from the right below the BJD formation and at the top of the Sudair whipstock set at near vertical inclinations (less than 3°) had formation, the well is still vulnerable to flow. A minimum of issues when gyro tools — single shot or gyro steering — were 2 days can be lost in a well control situation if the well flows required for motor steerability because of magnetic interfer- after the window has been opened in the casing with the ence from the casing and whipstock. Tri-cone insert (TCI) placement of a whipstock or after drilling has commenced. bits were used with single shot gyro tools to provide better In some instances, 3 or 4 days were lost trying to control the tool face control, compared to that achieved with a polycrys- influx while at the same time cure the losses. In the five wells talline diamond cutter fixed bit, while steering in the correct presented as a case study in this article, the potential lost time direction was assured by marking a scribe line on the drill- fighting Jilh influx was avoided by opening the window deep pipe. This results in a slower ROP as well as limiting the total into the Sudair or Sudair dolomite. number of revolutions on the bit. In addition, it requires an Following is a summary of the well design strategy, which extra trip while exiting the whipstock. Issues with gyro tool has been followed successfully to drill five wells with a deep availability and reliability, with tracking the casing because of kickoff directly from a whipstock using the HRSS. poor cement quality around the liner or casing, and with the Due to the inherently complex nature of these workover additional time taken to kickoff in the Sudair formation were reentry wells, a detailed technical analysis first has to be some of the further challenges that underscored the need for conducted to determine the point of sidetrack initiation, well an alternate and more efficient solution. directional plan, type of whipstock, length of the rathole, MW and mud properties, and design of the HRSS bottom-hole as- METHODOLOGY sembly (BHA) configuration, including hydraulics; also to be completed are a torque and drag analysis, tripping road maps, Given the challenge of trying to avoid reentering the BJD for- the design of drilling and tripping procedures to prevent me- mation, the reduced ROP due to plastic shale in the Sudair chanical and differential sticking mechanisms, and a strategy formation, and issues with vertical whipstock kickoffs, Saudi to execute the operations. Aramco embarked on a technology driven strategy to reduce The Saudi Aramco Engineering team and a directional drill- the number of days to complete a conventional reentry work- ing company worked together to define the deepest point for over by minimizing both the visible and the hidden nonproduc- each sidetrack. They determined this by working on different tive time. This article presents the successful deployment and trajectory scenarios, ensuring that it was deep enough to avoid field application of a high build rate rotary steerable system Jilh influx, while at the same time able to achieve the target (HRSS) tool that is able to directly exit from the whipstock. It entry, landing and reservoir objectives. draws a comparative analysis between the traditional modes of A detailed offset well analysis study is next undertaken by whipstock exit using steerable motors and the HRSS option. An closely studying the events of the pilot hole to determine the analysis of the HRSS’s dogleg capabilities, ROPs, steerability feasibility of drilling the reentry workover well as a single 5⅞” and stability was done to compare these capabilities with those hole size or as a dual size — 8⅜” for drilling the curve and of traditional drive systems. Lessons learned from the first deep landing in the target Carbonate B or C reservoir, and 5⅞” for gas vertical whipstock exit for a medium radius well are de- horizontally drilling the lateral in the target reservoir. Based scribed. Also, the case histories of the five wells using the tech- on the final choice of the completion design, the casing design nology are discussed as a measure to demonstrate the number is evaluated (not covered under the scope of this article). Once of days saved and the reduced nonproductive time. the basic framework of the well design is set, an estimate of This strategy has made it possible to eliminate the addi- the maximum DLS requirement in each formation is weighed tional runs required to gain separation from the casing and to against the capabilities of different tool configurations. The exit the zone of magnetic interference, while at the same time bit design and selection process is conducted by running sim- achieving deeper kickoffs with high doglegs in deep gas wells. ulation scenarios of BHA configurations vs. bit design in the integrated drill design software. The selection of the bit is DESIGN based on simulation results showing consistency in ROP and DLS capability. The HRSS has the potential to increase the efficiency of well Once the bit and configuration of the HRSS is decided delivery, saving a minimum of 5 days by drilling more effi- based on the tool capability and DLS requirements, it is im- ciently with a faster ROP and also by eliminating the time portant to run the cement bond and/or variable density logs needed to kickoff with a conventional steerable motor assem- to get an estimate of the cement quality around the casing bly. This requires changes to the well plan. Kicking off deeper, before opening the window. This information helps prepare

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 31

87448araD6R1.indd 31 11/28/16 12:25 PM

Time between Pull Out Well Hole Size Time Required to Kickoff Comments of Hole and Pick Up RSS Gyro tool failure led to tracking 7” liner Well F 8 ” 216 65.5 ⅜ below window, TCI bit had slow ROP Well G 8⅜” 30 27.5 Smooth kickoff Well H 8⅜” 19 21 Smooth kickoff Well J 8⅜” 43 No RSS used Difficulties kicking off

Well K 5⅞” 43 No RSS used Difficulties kicking off

Fig. 1. The differences between the K2, K1 and MK-1 casing designs. Table 1. Average time taken to sidetrack from the motherbore using a whipstock with the help of a conventional steerable motor assembly and gyro tool when kicking off from a vertical inclination.

weight on bit is pushed from the beginning to steer away from 300 the motherbore. The advantage of having near bit inclination 250 is that it provides real-time assurance of separation away 200 from the pilot hole, even though the azimuth readings are still 150 affected by magnetic interference. A proper fatigue manage- 100

Time (hrs) ment process has to be followed prior to drilling to ensure the 50 tubulars are in good condition to resist the bending stresses 0 Well F Well G Well H Well J Wellassociated K with use in a high DLS environment.

Time to kick off (hrs) Time between POOH and P/U RSS (hrs)

RESULTS Fig.Fig. 2. 2. Time Time taken taken to drill to drillsidetracks sidetracks in typical in reentrytypical wells reentry using wells a conventional using a conventional motor, RSS BHAs, and Time Time between Pull gyromotor, tools. RSS BHAs, and gyro tools. Well Hole Size Required to Out of Hole and Comments Table 1 describesKickoff the averagePick Up time RSS taken to sidetrack from Gyro tool failure led to tracking 7” liner contingency plans for high mud, in case flow is encountered Well F 8⅜” 216 65.5 the motherbore in five wells using a whipstockbelow window, TCI with bit had slowthe ROP help due to channeling in the cement. A dedicated run for gamma Well G 8⅜” 30 27.5 Smooth kickoff Wellof Ha conventional8⅜” 19 steerable motor21 assembly Smoothand kickoffa gyro tool; ray and/or casing collar location is carried out before the Wellit alsoJ depicts8⅜” the43 time takenNo RSS for used a round Difficultiestrip — kicking pulling off whipstock setting to correlate the depth and avoid setting the Well K 5⅞” 43 No RSS used Difficulties kicking off out of hole and picking up the RSS — to change the BHA whipstock in front of a casing collar. A gyro tool is used to Table 1. Average time taken to sidetrack from the motherbore using a whipstock with the help of a conventional steerable motor assembly and gyro tool when kicking off from a vertical inclination from a steerable motor to the RSS option. Figure 2 plots the orient the whipstock in the desired direction of kickoff in sit- time data for the five wells in Table 1. Table 2 shows the uations where the well is vertical, while a measurement while drilling tool can be used for orientation if the well inclination ROP Footage 10.0 2000 is more than 5°. 9.0 1800

At this time, the window is cut and a rathole is drilled, 8.0 1600 with its length determined by the distance of the top stabilizer 7.0 1400 6.0 1200

on the HRSS tool and the bit. The rathole length has to be (ROP ft/hr)

5.0 1000 (ft) kept short enough to prevent losing TVD room, needed to 9.0 4.0 800 7.3 7.4

6.6 Footage ease the dogleg requirement, while at the same time it has to 3.0 6.0 600

of Penetration 2.0 400 be long enough to ensure that once drilling has started, the 3.1 metal-to-metal contact between the stabilizer and the window Rate 1.0 200 0.0 0 Well MR1 Well MR2 Well MR3 Well MR4 Well MR5 Average and casing does not cause damage to the HRSS tool. The road

map is set up for the initial settings to be deployed to kick- Fig.Fig. 3. 3.Average Average ROP andROP footage and achieved footage in fiveachieved medium inradius five wells medium drilled with radius motor BHA.wells drilled

off the sidetrack in the required direction and make sure the with motor BHA.

Medium Radius Wells Drilled with Motor BHA Well Hole Size Footage Hours ROP (ft/hr) Kickoff Formation Medium Radius Wells DrilledMR1 with8⅜” Motor1,595 BHA 218.5 7.30 Sudair MR2 8⅜” 1,000 135 7.41 Carbonate C Well Hole Size Footage Hours MR3 8⅜”ROP (ft/hr)518 165.5 3.13 KickOffCarbonate Formation B MR4 8⅜” 1,429 159 8.99 BJD MR5 8⅜” 1,797 299.5 6.00 Carbonate D MR1 8⅜” 1,595 218.5 7.30 Average ROP = 6.56 ft/hr Sudair

MR2 8 ” 1,000 135 Table 2. Average ROP7.41 of medium radius wells drilled using a conventionalCarbonate steerable motor C assembly and ⅜ cutting across formations from BJD to Khuff MR3 8⅜” 518 165.5 3.13 Carbonate B MR4 8⅜” 1,429 159 8.99 BJD MR5 8⅜” 1,797 299.5 6.00 Carbonate D Average ROP = 6.56 ft/hr

Table 2. Average ROP of medium radius wells drilled using a conventional steerable motor assembly and cutting across formations from BJD to Khuff.

32 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD6R1.indd 32 11/28/16 12:25 PM Reentry Wells with Direct Whipstock Kickoff and High DLS Well Hole Size Start Incl End Inl Footage ROP (ft/hr) Max DLS Runs A 5⅞” 2° 48° 579 15.8 17.4 1 B 5⅞” 19° 84° 1,187 16.7 10.7 1 C 8⅜” 32° 74.85° 1,424 14.2 6 1 D 8⅜” 3° 41° 576 15.8 10.3 1 E 8⅜” 2.2° 81.5° 1,359 15.3 9.3 1 Average ROP = 15.6 ft/hr

Table 3. Average ROP and DLS on the five reentry wells drilled using the HRSS tool across formations varying from Sudair to the Carbonate B reservoir.

assembly, while Fig. 3 provides a graph of the data. Table 3 ROP Footage

17.0 1600 shows the average ROP and maximum DLS of drilling the

16.5 1400 five reentry wells using the HRSS, while Fig. 4 illustrates the 16.0 ft/hr) 1200 ROP and footage data. The HRSS shows an average ROP 15.5 1000 (ROP 15.0 of 15.6 ft/hr compared to 6.56 ft/hr with a conventional

800 (ft) 14.5 16.7 600 steerable motor assembly. This represents a 137% increase 15.6 14.0 15.3

15.8 Footage Penetration 15.8 400 13.5 in ROP over the conventional steerable motor option. It 14.2 13.0 200 also means that an estimated 3.7 days of drilling time can Rate of 12.5 0 Well A Well B Well C Well D Well E Average be saved if the curve of the medium radius well of 1,000 ft

Fig. 4. 4. ROP ROP and footageand footage achieved achievedin wells drilled in withwells the drilledHRSS tool, with kicking the off HRSS directly fromtool, a whipstock.kicking has to be drilled. In addition, direct savings of at least 40 Wells A and B were kicked off in the 5⅞” section, while Wells C, D and E were kicked off in the 8⅜” off directly from a whipstock. Wells A and B were kicked off in the 5 ” section, section. ⅞ hours or 1.6 days were achieved due to the quick exit di- while Wells C, D and E were kicked off in the 8 ” section. ⅜ rectly from the whipstock since a dedicated motor run was

time spent and average ROP of drilling five medium radius no longer needed to achieve separation from the motherbore

wells in the same set of formations using the steerable motor and move away from the zones of magnetic interference to Reentry Wells with Direct Whipstock Kickoff and High DLS Well Hole Start Incl End Incl Footage ROP (ft/hr) Max DLS Runs Size A 5⅞” 2° 48° 579 15.8 17.4 1 B 5⅞” 19° 84° 1,187 16.7 10.7 1 C 8⅜” 32° 74.85° 1,424 14.2 6 1 D 8⅜” 3° 41° 576 15.8 10.3 1 E 8⅜” 2.2° 81.5° 1,359 15.3 9.3 1 Average ROP = 15.6 ft/hr

Table 3. Average ROP and DLS on the five reentry wells drilled using the HRSS tool across formations varying from Sudair to the Carbonate B reservoir

Fig. 5. Results achieved in Well-A. Fig. 6. Results achieved in Well-B. Fig. 5. Results achieved in Well-A. Fig. 6. Results achieved in Well-B.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 33

87448araD6R1.indd 33 11/28/16 12:25 PM

FigFig.. 77.. Results achieved achieved in Well-C.in Well-C. FigFig.. 8.8. ResultsResults achieved achieved in Well-D.in Well-D.

obtain reliable azimuth readings. Furthermore, the use of a gyro tool for a purely steering purpose was eliminated since orienting of the tool face is taken care of by the HRSS tool directly. The summary in Table 3 of the maximum doglegs achieved on the five wells drilled with the deep kickoff and direct exit from the whipstock using the HRSS tool is ac- companied by summary graphs of the doglegs vs. gamma ray data and ROP, Figs. 5 through 9. Figure 5 depicts the results achieved in Well-A. The objec- tive of this well was to sidetrack as a 5⅞” section by kicking off deep and to achieve a high build rate in a formation not friendly toward the planned dogleg. A conventional approach would have required aggressive motor bend settings to achieve the planned build rates; however, this well was drilled with the HRSS tool, which consistently delivered on average 15°/100 ft DLS. Optimum drilling parameters were applied to achieve maximum ROP. Where a conventional motor BHA would yield an ROP of 6 ft/hr in a plastic shale formation characterized by high gamma ray count (green), the HRSS tool showed a consistent ROP of approximately 20 ft/hr. Figure 6 shows the results of Well-B. This was a 3D profile well, turning the well toward the northwest direction while building inclination at the same time. The HRSS tool handled the required DLS, and the 7” liner was run smoothly without

any issues of tortuosity. The HRSS tool was configured with Fig. 9. Results achieved in Well-E. low output capability and delivered on average 8°/100 ft DLS Fig. 9. Results achieved in Well-E.

34 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD6R1.indd 34 11/28/16 12:25 PM using 70% to 100% of the tool’s actual DLS capability for these savings are potentially more if the ROP using a TCI bit the first target formation. It later delivered the same results on a conventional steerable motor to drill a medium radius with reduced settings equivalent to only 50% of the tool’s well is compared to the ROP of the HRSS tool. actual DLS capability. The well was kicked off and landed There are also hidden savings that come from not having in the same run, cutting across multiple formations as shown to pick up the motor off-bottom to orient the toolface or not through the change in gamma ray (green). The well was having to wait for a gyro survey, which increases the chances drilled as a single hole size, eliminating the 8⅜” section and of getting the drillstring stuck differentially in high overbal- avoiding the high-pressure formation. anced carbonates by remaining stationary in one position. Figure 7 shows the results of Well-C. The low inclination That wait is eliminated as the HRSS can be used to kickoff kickoff concept was applied here for the first time in the ⅜8 ” with the normal drilling parameters from the very beginning. section. This was planned as a 3D well, diverging away from These advantages of technological innovation coupled with the original hole. The well was delivered as planned by diverg- the proper engineering of well design make it possible to ing from the motherbore, then turning and aligning toward drill complex reentry wells more efficiently in tough drilling the target. Based on previous experience, the tool was con- environments. figured with a low gain DLS output. Throughout the section length, consistently 4°/100 ft to 5°/100 ft DLS was achieved, ACKNOWLEDGMENTS and the section was drilled to the planned target depth along the required inclination and azimuth. The authors would like to thank the management of Saudi Figure 8 shows the results of Well-D. The objective was to Aramco and Schlumberger for their support and permission to achieve a planned build rate in the 8⅜” section to mitigate the publish this article. drilling risks associated with the sliding BHA. The tool was This article was presented at the SPE Saudi Arabia Section programmed to handle the planned DLS by steering at max- Annual Technical Symposium and Exhibition, al-Khobar, imum capability, followed by a low dogleg interval with the Saudi Arabia, April 25-28, 2016. HRSS on 40% to 50% steering capability settings. While the tool was drilling the transition zone, high torsional vibrations REFERENCES were recorded. These were mitigated by increasing the surface RPM without affecting the tool’s performance. 1. Al-Khamees, S.A., Okwa, H.D., Verma, J.K. and Ganda, Figure 9 shows the performance of Well-E. This well was S.: “The First Successful Short-Radius Reentry Well in kicked off deep in the 8⅜” section. Usually the high gamma Deep Gas Drilling in Saudi Arabia,” SPE paper 139855, ray formation has a lower response to drilling wells with DLS. presented at the SPE/IADC Drilling Conference and Subsequently, as observed, a change of formation to high or Exhibition, Amsterdam, The Netherlands, March 1-3, low gamma ray did not affect the good DLS response from 2011. the HRSS. The original pilot hole was drilled vertically and 2. Thomas, S.P., Mukherjee, T.S., Ezi, P.C., Alfonzo Briceno, had encountered high-pressure zones. These were avoided by L.A., Verma, J.K., Ganda, S., et al.: “A Novel Approach the deeper sidetrack. to Drilling 8⅜” Medium-Radius Curve Section in Deep Gas Drilling in Saudi Arabia: Successful Introduction of CONCLUSIONS High Build Rate Rotary Steerable System,” SPE paper 168072, presented at the SPE Saudi Arabia Section Annual With the ability to kickoff directly from a vertically set whip- Technical Symposium and Exhibition, al-Khobar, Saudi stock, as well as whipstocks at a higher inclination, the HRSS Arabia, May 19-22, 2013. technology opens up multiple opportunities for sidetracking medium radius wells. This includes the ability to kickoff the well at deeper points to avoid the high-pressure Jilh forma- tion, thereby making previously uneconomical reentry side- tracks viable. The risky option of milling the existing 7” liners to avoid the high-pressure zone and then kicking off deeper can also be avoided, because the higher DLS capability of the HRRS allows deeper kickoffs and landing. Efficient drilling with the HRRS at its higher ROP enables a minimum savings of 5 days of rig time compared to simi- lar wells drilled with a conventional motor assembly. Wells drilled with HRSS have also been observed to provide a better borehole quality as there have been no issues when running wireline logging or a liner after reaching those wells’ TD. All

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 35

87448araD6R1.indd 35 11/28/16 12:25 PM BIOGRAPHIES Verdy L. Siregar joined Saudi Aramco in 2011 as a Drilling Engineer for the Abdul Halim Ab Hamid joined Saudi Gas Workover Engineering Unit in the Aramco in 2005 and is currently a Gas Drilling Engineering Department. Supervisor for the Gas Workover He is working extensively on Engineering Unit in the Gas Drilling numerous gas workover projects for Engineering Department. During his reentry sidetracks, and on safety career with Saudi Aramco, he and his related workovers and mechanical repairs, which include team of engineers have worked on some casing-casing annulus communication repair. Verdy is numerous gas workover projects for reentry sidetracks, also involved in trials of some new technology tools within safety-related workovers and mechanical repairs. The the Gas Drilling Engineering Department, the high DLS primary mission for Abdul Halim and his team is to RSS tool, mechanical tubing puncher, and off-bottom expedite the recovery of potential gas wells while cementing multistage fracturing combo tool, to name a few. continuously striving for effective solutions to minimize He started his career in the oil and gas industry as a costs and still ensure safe operations. Part of his work Completion Engineer with Total Indonésie in 1997. Verdy includes special projects for casing-casing annulus repair, held several different positions afterwards within the same multistage fracture (MSF) completions and deployment of organization, including Drilling Foreman, Stimulation the off-bottom cementing tool: MSF combo technology. Engineer, and Coiled Tubing Drilling Engineer until his Abdul Halim started his career in the oil and gas international assignments as a Drilling Engineer with industry as a Logging Engineer with Western Atlas in Total Gabon and as a Drilling Engineer and Operational 1997 and later worked for ExxonMobil Malaysia as a Safety Leader with Total Cameroun. He spent the last Workover/Well Service engineer for 10 years. 10 months prior to joining Saudi Aramco working as the He has written and coauthored several Society Head Drilling Engineering for Delta Operation with Total of Petroleum Engineers (SPE) articles relating to the Indonésie in Balikpapan, Indonesia. application of new technologies in drilling and completion. Verdy received his B.S. degree in Petroleum Engineering Abdul Halim received his B.S. degree in Electrical from Trisakti University, Jakarta, Indonesia, his M.S. Engineering from the University of Missouri, Columbia, degree in Mineral Economics from Colorado School of MO. Mines, Golden, CO and a Diplom-Ingenieur degree in Petroleum Economics from École Nationale Supérieure du Pétrole et des Moteurs (Institut Français du Pétrole), Rueil Malmaison, France.

Ali N. Al-BinAli is a Gas Drilling Engineering General Supervisor in Saudi Aramco’s Gas Drilling Engineering Department. He has 18 years of experience in drilling and workovers, in both gas and oil. Ali was appointed a Workover Engineering Supervisor in 2008 and then went on to be the Workover Engineering General Supervisor from 2011 to 2014. Afterwards, he moved to deep high-pressure/high temperature gas drilling and workovers, which has been his focus from January 2015 to date. Ali has worked as an Engineering Division Head for all kinds of wells, including onshore, offshore, water, oil and deep high-pressure/high temperature gas wells, as well as underbalanced drilling. He received his B.S. degree in Mechanical Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.

36 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD6R1.indd 36 11/28/16 12:25 PM Mohamed E. Khalil joined Saudi Omar T.A. Ashraf is currently Aramco in 2002. He is a Drilling and working as a Customer Services Workover Engineering Supervisor for Leader for Pakistan at GE Power Saudi Aramco’s Gas Drilling Services. Prior to this, he worked with Department in ‘Udhailiyah. From this the Schlumberger Drilling and position, Mohamed has been leading Measurements segment from 2006 to engineering efforts toward deep gas 2016 in Saudi Arabia. Omar has held drilling and reentry optimization, such as short and various field and office-based engineering and operational medium radius reentries, snubbing operations, and dealing roles involving the entire suite of drilling and workover with advanced completion and workover types. projects in Manifa, Shaybah, and ‘Udhailiyah. Prior to joining Saudi Aramco, he was with the In the past, Omar has served with Alstom at their Khalda Petroleum Company in Egypt for more than 15 corporate headquarters in Paris, where he gained years, working in the Drilling and Workover Engineering experience in global sustainability initiatives, corporate Department. responsibility, and renewable energy startups. In 2006, Mohamed received the Top Achievers in He has published several Society of Petroleum Engineers Drilling Supervision award from the Murchison Drilling (SPE) papers on the subjects of: drilling and workover School. He is a member of the Society of Petroleum operations optimization, innovative solutions for drilling Engineers (SPE). efficiency, introduction and deployment of new technology, In 1985, Mohamed received his B.S. degree in and cost reduction through differentiated value addition Petroleum Engineering from Cairo University, Giza, Egypt. services. Omar received his B.S. degree in Mechanical Ayman Al-Ghazzawi is a Lead Senior Engineering from National University of Sciences and Drilling Engineer in Schlumberger. Technology, Islamabad, Pakistan. He also has three M.S. Having more than 10 years of degrees in Management and Engineering of Energy and experience with Schlumberger, he has Environment from Ecole des Mines (EMN), France; Royal held a variety of positions in different Institute of Technology (KTH), Sweden; and Universidad locations: Measurements and Logging Politecnica de Madrid, Spain. While Drilling Field Engineer in Malaysia; Directional Drilling Field Engineer and Muhammad S. Balka joined Real-Time Operations Support Center Engineer in Qatar; Schlumberger in 2001, working as a in-house Drilling Engineer for Al-Khafji Joint Operations Directional Driller in the Middle East and Wellbore Surveying Specialist in Saudi Arabia. and North Africa. From August 2006 For the past 3 years, Ayman has worked as the Lead to July 2010, he was based in Drilling Engineer with Saudi Aramco, primarily in Shenzhen, China as the Lead Senior exploration projects for gas and unconventional gas in the Drilling Engineer for China and Northern Area. He is also one of the specialists in wellbore Japan. During this period, Muhammad worked on surveying and anti-collision within the Middle East deepwater and technically demanding projects for different geographic location. clients in the South China Sea. In his next assignment, Ayman received his B.S. degree with honors in Chemical from August 2010 to May 2014, Muhammad was assigned Engineering from King Fahd University of Petroleum and as the Drilling Engineering Manager for Schlumberger Minerals (KFUPM), Dhahran, Saudi Arabia. Drilling and Measurement (D&M), providing drilling He is a member of the Society of Petroleum Engineers engineering and technical support to clients in Thailand, (SPE). Myanmar and Vietnam, while based in Bangkok. He is now with the working as a Senior Drilling Engineer for Drilling Optimization in Saudi Arabia, where he has been since June 2014. Muhammad supports Saudi Aramco in exploration projects for gas and unconventional gas in the Northern Area. He started his career working as a Geophysicist from 1995 to 1996, and from 1997 to 2001 as a Data Engineer in Pakistan and the Middle East. In 1994, Muhammad received his M.S. degree in Geophysics from Quaid-e-Azam University, Islamabad, Pakistan. He is a member of the Society of Petroleum Engineers (SPE).

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 37

87448araD6R1.indd 37 11/28/16 12:25 PM A New Insight on the Impact of Individual Ions on Fluid-Fluid Interactions and SmartWater Recovery

Authors: Mohammed A. Geer, Dr. Ahmed Gmira, Dr. Ali A. Yousef and Dr. Sultan M. Al-Enezi

ABSTRACT lately and has proven the efficiency of this recovery method. Extensive research on oil-brine-rock systems has shown the SmartWater flooding has proven to be an effective and suc- impact of ionic makeup and salinity on wettability alteration cessful recovery method for carbonates. With SmartWater and oil recovery. flooding, the injected water alters the carbonate rock wettabil- The consensus in the industry is that salinity affects oil- ity to produce incremental oil. Core-scale displacement exper- brine-rock systems by rock wettability alteration. A series of iments have demonstrated significant incremental recoveries in spontaneous imbibition tests were conducted1, 2 on carbonate both secondary and tertiary modes. Single well chemical tracer rocks, and the results attributed the observed oil recovery tests have demonstrated this potential in the field at a scale increase to the interplay of determining ions — magnesium larger than that available in the laboratory. Still, the underly- 2+ 2- 2+ (Mg ), sulfate (SO4 ) and calcium (Ca ). SmartWater injec- ing mechanisms responsible for the SmartWater alteration of tion in carbonates has been widely investigated in several pre- carbonate wettability are not well understood. The objective vious studies, which showed the positive impact of seawater of this work is to understand the effects of individual mon- dilution on oil recoveries by employing numerous techniques: ovalent and divalent ions on brine-oil interactions, and their from measuring interfacial tension (IFT), contact angle and role in the observed alteration of carbonate wettability. zeta potential, to coreflooding3-7 and nuclear magnetic reso- In previous studies, we investigated liquid-rock interactions nance techniques8. The latest fundamental research findings and their role in wettability alteration. At fixed salinities, concern the detrimental effects of monovalent ions, such as monovalent and divalent ions were found to have different sodium (Na+) and chlorine (Cl-), and the key role played by effects on the calcite surface potential. In this study, we have multivalent ions, such as Ca2+, Mg2+ and SO 2-, in addition to investigated the liquid-liquid interactions. We performed in- 4 the connectivity enhancement between micropores and mac- terfacial tension (IFT) measurements between oil and brines ropores due to anhydrite dissolution. It has been shown that of fixed salinities but varying ionic compositions, and we col- dissolved ions are temperature dependent thermodynamically lected IFT data at different temperature conditions. and that sulfate ions gain in efficiency at a higher temperature That the different SmartWater recipes have exhibited differ- ent IFT values at fixed salinities indicates the varying effects of Component Amount ions on fluid-fluid interactions. For instance, SmartWater reci- Saturates 39.2% pes composed exclusively of magnesium (Mg) cations exhibited Aromatics 48.3% a remarkably low level of IFT values. Other SmartWater recipes Resins 7% with sodium (Na) or calcium (Ca) cations exhibited comparable Asphaltenes 5.5% IFT stabilization levels, while SmartWater recipes that are solely composed of sulfate anions have resulted in higher IFT values. Total Acid Number 0.25 mg KOH/g oil Those results will be integrated at further stages with mea- Properties Amount Saturation Pressure surements of the zeta potentials and contact angles acquired 1,804 across brines of variant ionic compositions. This systematic (psia at 212 °F) Stock Tank Oil Gravity integration will eventually allow for a clearer distinction 0.3 (°API at 60 °F) of the ions’ effects, which will help to better optimize the Dead Oil Density at SmartWater recipe. 0.545 Room Temperature (lb/cf) Dead Oil Viscosity at INTRODUCTION 0.1459 Room Temperature (cP)

Enhancing oil recovery by adjusting the ionic composition Table 1. The components and properties of the crude oil used in conducting this and salinity of the injected water has been widely investigated experiment

38 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD7R1.indd 38 11/28/16 12:37 PM 9 — 70 °C . Recent zeta potential measurements on carbon- calcium chloride (CaCl2), sodium chloride (NaCl) and sodium 10 ates have identified the contribution of individual ions and sulfate (Na2SO4), Table 2. All brine compositions had a simi- SmartWater recipes in altering the rock surface charges, which lar total dissolved solids content of 5,000 ppm. is considered a key mechanism in rock wettability alteration. The present work is a first step in a series of studies IFT Apparatus striving to decouple the fluid-fluid interaction caused by SmartWater injection through an evaluation of oil-brine IFT A pendant drop tensiometer — model IFT-10 — manufactured characteristics. IFT between oil and brine has been studied by Temco Inc. (USA) was used for the IFT measurements. since the 1960s11. In addition, the impact of variant operating The pendant drop cell is a cylindrical stainless steel chamber pressure and temperature on IFT between oil and brines is by with a 40 cm3 volume. A stainless steel needle is placed at the now well understood12, 13. In this work, we introduce a new bottom of the cell to generate pendant drops. The chamber is perspective by measuring the IFT between Arab-D oil and equipped with appropriately sealed borosilicate glass windows, multiple brine compositions, solely composed of key individ- which allow a view of the inner space during the tensiometer’s ual monovalent and divalent ions, at various temperatures to operation. The light beam source, located at one side of the investigate the role of these SmartWater flooding constituents visualization axis, is a halogen bulb covered by a white diffuser. on the fluid-fluid interaction at a macro level. The camera, located at the opposite side of the visualization axis, is a Ramé-Hart F1 series connected to a desktop computer EXPERIMENTAL PROCEDURE through a frame grabber card. The overall setup sits on top of a vibration-free table to en- Fluids and Material sure accurate measurements, Fig. 113. The temperature control system is mainly composed of a water bath that circulates a An Arab-D stock tank crude oil was used in conducting this cooling fluid in two internal loops and a temperature con- study. Table 1 lists the properties of the crude oil used. This troller to set the measurement temperature conditions. A Dell crude oil was filtered to remove any contaminants, and then computer desktop was used to acquire the digital image of the it was vented at room temperature to prevent any gas evo- oil pendent drop and to perform the subsequent drop image lution during the measurements at elevated temperatures. analysis, digitization and computation. DropImage Software Four different synthetic brine compositions were used in the from Ramé-Hart was used and the IFT was calculated by us- experiment. Each type of brine was composed solely of one of ing the Young-Laplace equation.

the following key components: magnesium chloride (MgCl2), IFT MEASUREMENT PROCEDURE Brine Brine Composition Salt Mass (g/l)

MgCl2 MgCl2.6H2O 12.30 The experimental procedure for determining IFT is as follows: The cell is heated to the desired experimental temperature CaCl2 CaCl2.2H2O 7.63 NaCl NaCl 5.76 until stabilization at that temperature. The brine is pumped through a stainless steel tube to the needle tip, then a 17 µl to Na SO Na SO 5.76 2 4 2 4 20 µl crude oil droplet is generated inside the cell. IFT values Table 2. Composition of the four brines used in the experiment, with each brine and drop volumes are monitored immediately and monitoring is composed of one of the following salts: MgCl , 6H2O, CaCl .2H O, NaCl 2 2 2 continues for 1 hour to allow the two fluids to reach equilib- and Na2SO4 rium, with the software capturing and processing an image

every 30 seconds. The IFT cell body and its various parts afterward are cleaned thoroughly with toluene, acetone and deionized water, followed by air drying. Lamp Camera RESULTS AND DISCUSSION IFT Cell Temperature Controller The IFT phenomena in a freshly formed interface is a dynamic process, and time is required to reach equilibrium between

Water Bath two immiscible fluids. Crude oil contains thousands of com- ponents, including polar fractions that interact closely with

Vibration Free Table the brine solution. Diffusion and adsorption and/or desorption are the main processes acting at the interface, and therefore enough time is allowed for the fluids to reach equilibrium14. Fig. 1. The IFT-10 pendant drop tensiometer apparatus used to measure the IFT The recurrent phenomenon observed was a rapid drop in between oil and multiple brine compositions for this set of experiments13. IFT within the first seconds of forming an oil-brine interface,

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 39

87448araD7R1.indd 39 11/28/16 12:37 PM

followed by a slow decay until equilibrium was reached. The These measurements reveal a distinctive variation in initial

quick decay is related to the quick adsorption, occurring at and equilibrium IFT values as a function of both the brine composition and temperature. The brine composed of MgCl the interface, of the active components of the crude oil — 2 mainly polar components. This was followed by a diffusion only showed the lowest initial and equilibrium IFT values, Fig. 4. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual process acting within the bulk of the two phases, a much SmartWatermeasured floodingat 25 °C key at ions 34 conducted mN/m and at 90 29.4 ¡ C. mN/m, respectively, 15 slower process . Fig. 7. The CaCl2 brine followed, measured at 25 °C with ini-

Figures 2 to 4 show the dynamic behavior of IFT as a func- tial and equilibrium IFT values at 38 mN/m and 32.6 mN/m, respectively, Fig. 8. The NaCl brine had a stabilization level at tion of time (1 hour) for crude oil and the single ion brines at

25 °C, 50 °C and 90 °C, respectively, all at atmospheric pres-

sure. Initial and equilibrated IFT values depend strongly on the

brine’s ionic content. Of note is the sensitivity of the IFT values

12, to temperature, in accordance with previous reported studies 13, and how the IFT decreases as the temperature increases for

both initial and equilibrated values, Figs. 5 and 6, respectively.

Fig. 2. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual Fig. 5. Initial values of IFT for the four brine compositions. Note that MgCl has Fig. 5. Initial values of IFT for the four brine compositions. Note that MgCl2 has the lowest IFT values SmartWater flooding key ions conducted at 25 ¡ C. 2 acrossthe lowest all temperatures,IFT values across CaCl all temperatures, and NaCl have CaCl comparable2 and NaCl havevalues, comparable while Na SO has the highest among 2 2 4 allvalues, constituents. while Na SO has the highest among all constituents. 2 4

Fig. 2. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual SmartWater flooding key ions conducted at 25 °C. Fig. 2. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual SmartWater flooding key ions conducted at 25 ¡ C.

Fig. 6. Equilibrium values of IFT for the four brine compositions. Note that MgCl2 has the lowest across all

temperatures, CaCl2 and NaCl have comparable values, while Na2SO4 has the highest among all

constituents.

Fig. 6. Equilibrium values of IFT for the four brine compositions. Note that Fig. 6. Equilibrium values of IFT for the four brine compositions. Note that MgCl2 has the lowest across all MgCl has the lowest across all temperatures, CaCl and NaCl have comparable temperatures,2 CaCl and NaCl have comparable 2values, while Na SO has the highest among all 2 2 4 constituents.values, while Na2SO4 has the highest among all constituents. Fig. 3. IFT vs. time plot for the fluid-fluid interface between oil and the four brines

Figcomposed. 3. IFT vs. of individualtime plot SmartWaterfor the fluid -floodingfluid interface key ions between conducted oil andat 50 the°C. four brines composed of individual

SmartWater flooding key ions conducted at 50 ¡ C.

Fig . 3. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual

SmartWater flooding key ions conducted at 50 ¡ C.

Fig. 7. The brine composed of MgCl only showed the lowest initial and equilibrium Fig. 4. IFT vs. time plot for the fluid-fluid interface between oil and the four 2 Fig. 4. IFT vs. time plot for the fluid-fluid interface between oil and the four brines composed of individual brines composed of individual SmartWater flooding key ions conducted at 90 °C. IFT value, measured at 25 °C at 34 mN/m and 29.4 mN/m, respectively. SmartWater flooding key ions conducted at 90 ¡ C. Fig. 7. The brine composed of MgCl2 only showed the lowest initial and equilibrium IFT value, measured atFig. 25 7.¡ CThe at brine 34 mN/m composed and of 29.4 MgCl 2mN/m, only showed respectively. the lowest initial and equilibrium IFT value, measured at 25 ¡ C at 34 mN/m and 29.4 mN/m, respectively. 40 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD7R1.indd 40 11/28/16 12:37 PM

Fig. 5. Initial values of IFT for the four brine compositions. Note that MgCl2 has the lowest IFT values across all temperatures, CaCl2 and NaCl have comparable values, while Na2SO4 has the highest among all constituents.

the lowest IFT, followed by CaCl and NaCl with comparable 2 IFT stabilization levels, and then Na SO with the highest IFT 2 4

values at all temperatures. Having the same sequence of stabi-

lization across all operating temperature conditions (25 °C, 50

°C and 90 °C) confirms the sensitivity of IFT to the ionic com-

position. Also, the measurement repeatability defies any sus-

picion of measurement anomalies. The results also showed a

linear trend of decreasing IFT values in the case of MgCl and 2 CaCl , while this behavior was not seen in NaCl and Na SO . 2 2 4

In those brines, the IFT values tended to flatten at higher tem-

Fig. 8. The brine composed of CaCl only showed the next lowestperatures, initial and and the equilibrium difference between IFT value these, values measured at 2 measured at 25 ¡ C at 38 mN/m and 32.6 mN/m, respectively.50 °C and 90 °C becomes less pronounced. Fig. 8. The brine composed of CaCl2 only showed the next lowest initial and equilibrium IFT value, Fig. 8. The brine composed of CaCl only showed the next lowest initial and This sequence of IFT values seen across all temperature measured at 25 ¡ C at 382 mN/m and 32.6 mN/m, respectively. Fig.equilibrium 8. The IFT brinevalue, measured composed at 25 °C ofat 38CaCl mN/m2 andonly 32.6 showed mN/m, respectively. the next lowest initial and equilibrium IFT value, runs highlights the important role of ions when it comes to measured at 25 ¡ C at 38 mN/m and 32.6 mN/m, respectively. designing the composition of brines used for SmartWater

flooding applications. Lower IFT values will eventually lead

to lower capillary pressure at the interface of the wetting (oil)

and non-wetting (brine) phases, leading to further interaction.

Since rock wettability alteration is the main recovery mecha-

nism in SmartWater flooding processes, this enhanced inter-

action between the wetting and the non-wetting phases will

intensify the exchange of the ions between the SmartWater

flooding and the oil at the rock surface. This in turn will lead

to further release of the carboxylic molecules adsorbed at the

rock surface and eventually result in an incremental oil recov-

2+ ery. In all test runs, Mg ions have shown the lowest level of

Fig. 9. The NaCl brine had a stabilization level at 25 ¡ C similarIFT to values, that of and the this CaCl fact brine,suggests although that optimized its formulations 2 starting Fig. 9. The point NaCl brinewas had higher a stabilization at 42.8 level mN/m. at 25 °C similar to that of the of the SmartWater ionic content to contain more Mg ions Fig.CaCl 9. brine, The although NaCl its brine starting ha pointd awas stabilization higher at 42.8 mN/m. level at 25 ¡ C similar to that of the CaCl2 brine, although its 2 may lead to further incremental oil recovery, compared to starting point was higher at 42.8 mN/m. Fig. 9. The NaCl brine had a stabilization level at 25 ¡ C similar to that of the CaCl2 brine, although its the recovery achieved with the current ionic composition of starting point was higher at 42.8 mN/m. SmartWater.

Reducing sulfate ions in the SmartWater ionic constituents

may also lead to optimized SmartWater flooding performance.

The obstructive role of the sulfate ions seen in these sets of

IFT measurements is in agreement with results of the zeta

potential10 and streaming potential16 measurements already

published.

2+ The low IFT values exhibited by Mg can be attributed to

its relatively low molecular weight, in addition to the weak

ionic bond in MgCl compared to bonds in the other tested 2 salts. Although Mg comes second after sodium in atomic

weight, the MgCl bond-dissociating energy — an indication 2 Fig. 10. The brine that is composed mainly of Na SO had the highest 2 4 of the ionic bond strength — when dissolved in water is 318 Fig.stabilization 10. The level brine at 25 °C, that with is an composedIFT value of 44.5 mainly mN/m, andof Nait had2SO a very4 ha d the highest stabilization level at 25 ¡ C, with an 17 IFThigh value starting ofpoint 44.5 at 55.6 mN/m mN/m., and it had a very high starting point atkJ/mol 55.6 mN/m.measured at 298 K, which is a much weaker bond Fig. 10. The brine that is composed mainly of Na2SO4 had the highest stabilization level at 25 ¡ C, with an when compared to CaCl and NaCl, with energies of 398 kJ/ IFT value of 44.5 mN/m, and it had a very high starting point at 55.6 mN/m. 2 Fig. 25 °C10. similar The brine to that that of theis composed CaCl brine, mainly although of Naits starting2SO4 ha d the highest stabilization level at 25 ¡ C, with an 2 mol and 410 kJ/mol, respectively. This weak bond will break IFTpoint value was ofhigher 44.5 at mN/m 42.8 mN/m,, and it Fig. had 9. a Finally, very high the startingbrine that point atfirst, 55.6 allowing mN/m. Mg2+ to easily get to the oil interface, interact

is composed mainly of Na SO had the highest stabilization and reach equilibrium with the oil phase, all faster than the 2 4 level at 25 °C, with an IFT value of 44.5 mN/m, and it had a other ions. This can be seen clearly — previously shown in very high starting point at 55.6 mN/m, Fig. 10. Figs. 2 to 4 — across all operating temperature conditions. This IFT sensitivity to the ionic composition of the brines MgCl2 establishes a flat stabilization level of IFT values vs. was seen in the other set of measurements conducted at 50 time — a clear sign of equilibrium — much earlier than all the

°C and 90 °C. The same sequence of stabilization levels was other salts. Na2SO4, with the highest bond-dissociating energy

recorded at both temperatures, where MgCl2 came first with compared to the other brines at 1,384 kJ/mol, has the highest

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 41

87448araD7R1.indd 41 11/28/16 12:37 PM initial and equilibrium IFT values across all temperatures. necessary to detect and quantify the incremental oil recovery In general, as the temperature increases, the equilibrium due to different ionic compositions of brine. between the oil phase and the brine is reached faster, and as a result the IFT vs. time plot exhibits the stabilization level is ACKNOWLEDGMENTS

achieved much earlier. While both NaCl and CaCl2 showed very similar IFT values at the stabilization level, the plots also The authors wish to thank the management of Saudi Aramco

show that CaCl2 reached equilibrium first at 25 °C, while for their support and permission to publish this article. NaCl equilibrated with the oil phase faster at 50 °C and 90 °C. This phenomenon may indicate that under higher tem- REFERENCES perature conditions, Na+ ions become more active and are triggered to interact with the oil phase and reach equilibrium 1. Austad, A. and Standnes, D.C.: “Spontaneous Imbibition faster than Ca2+ at those higher temperatures. of Water into Oil-Wet Carbonates,” Journal of Petroleum Science Engineering, Vol. 39, Nos. 3-4, September 2003, CONCLUSIONS pp. 363-603. 2. Austad, T., Strand, S., Madland, M.V., Puntervold, Based on these sets of IFT measurements between oil and T. and Korsnes, R.L.: “Seawater in Chalk: An EOR multiple compositions of brine, the following conclusions and Compaction Fluid,” SPE Reservoir Evaluation & were drawn: Engineering, Vol. 11, No. 4, August 2008, pp. 648-654. 3. Yousef, A.A., Al-Saleh, S., Al-Kaabi, A.U. and Al-Jawfi, • The IFT of oil and/or brine is sensitive to brine ionic M.S.: “Laboratory Investigation of Novel Oil Recovery composition, in addition to the well-known sensitivity Method for Carbonate Reservoirs,” SPE paper 137634, to temperature conditions. presented at the Canadian Unconventional Resources and • SmartWater recipes composed exclusively of Mg cations International Petroleum Conference, Calgary, Alberta, exhibited a remarkably low level of IFT values. Other Canada, October 19-21, 2010. SmartWater recipes with sodium or calcium cations 4. Yousef, A.A., Al-Saleh, S.H. and Al-Jawfi, M.S.: exhibited comparable IFT stabilization levels, while “SmartWater Flooding for Carbonate Reservoirs: Salinity SmartWater recipes solely composed of sulfate anions and Role of Ions,” SPE paper 141082, presented at the SPE resulted in higher IFT values. Middle East Oil and Gas Show and Conference, Manama, Bahrain, September 25-28, 2011. • The weak ionic bond in MgCl2, compared to the bonds in other constituents of the SmartWater flooding 5. Yousef, A.A., Al-Saleh, S.H. and Al-Jawfi, M.S.: “The composition, might be the reason behind the low Impact of the Injection Water Chemistry on Oil Recovery values of IFT achieved through that brine. In this case, from Carbonate Reservoirs,” SPE paper 154077, presented

Mg2+ ions were able to disassociate easily in the brine at the SPE EOR Conference at Oil and Gas West Asia, and quickly reach out and interact with the fluid-fluid Muscat, Oman, April 16-18, 2012. interface. As a result, our macroscopic approach used 6. Yousef, A.A., Al-Saleh, S.H. and Al-Jawfi, M.S.: here suggests that optimized recipes of the SmartWater “Improved/Enhanced Oil Recovery from Carbonate

containing more MgCl2 may result in increased oil Reservoirs by Tuning Injection Water Salinity and Ionic recovery. Content,” SPE paper 154076, presented at the SPE Improved Oil Recovery Symposium, Tulsa, Oklahoma, • Although NaCl and CaCl2 showed comparable IFT April 14-18, 2012. stabilization values, CaCl2 was able to stabilize faster under room temperature conditions, while 7. Yousef, A.A. and Ayirala, S.C.: “Optimization Study of a NaCl reached equilibrium first at higher operating Novel Water-Ionic Technology for Smart Waterflooding temperatures. This suggests that sodium ion activity Application in Carbonate Reservoirs,” Oil and Gas increases under elevated temperature conditions. Facilities, Vol. 3, No. 5, October 2014, pp. 72-82. 8. Kwak, H.T., Yousef, A.A. and Al-Saleh, S.H.: The way forward beyond these IFT measurements is to “New Insights on the Role of Multivalent Ions in study the effect of pressure on IFT values while varying Water-Carbonate Rock Interactions,” SPE paper 169112, the ionic content of the SmartWater. A set of contact angle presented at the SPE Improved Oil Recovery Symposium, measurements on the same set of brine compositions is also Tulsa, Oklahoma, April 12-16, 2014. necessary to confirm the wettability alteration trend and to di- 9. Yi, Z. and Sarma, H.K.: “Improving Waterflood Recovery rectly relate the impact of the individual brine constituents to Efficiency in Carbonate Reservoirs through Salinity the recovery mechanism. A set of coreflooding experiments is Variations and Ionic Exchanges: A Promising Low-Cost

42 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD7R1.indd 42 11/28/16 12:37 PM ‘Smart Waterflood’ Approach,” SPE paper 161631, presented at the Abu Dhabi International Petroleum Conference and Exhibition, Abu Dhabi, UAE, December 11-14, 2012. 10. Alotaibi, M.B. and Yousef, A.A.: “The Impact of Dissolved Species on the Reservoir Fluids and Rock Interactions in Carbonates,” SPE paper 177983, presented at the SPE Saudi Arabia Section Annual Technical Symposium and Exhibition, al-Khobar, Saudi Arabia, April 21-23, 2015. 11. Wagner, O.R. and Leach, R.O.: “Effect of Interfacial Tension on Displacement Efficiency,” Society of Petroleum Engineers Journal, Vol. 6, No. 4, December 1966, pp. 335-344. 12. Alvarez, E., Vazquez, G., Sanchez-Vilas, M., Sanjurjo, B. and Navaza, J.M.: “Surface Tension of Organic Acids + Water Binary Mixtures from 20 °C to 50 °C,” Journal of Chemical & Engineering Data, Vol. 42, No. 5, September 1997, pp. 957-960. 13. Okasha, T.M. and Alshiwaish, A.: “Effect of Brine Salinity on Interfacial Tension in Arab-D Carbonate Reservoir, Saudi Arabia,” SPE paper 119600, presented at the SPE Middle East Oil and Gas Show and Conference, Manama, Bahrain, March 15-18, 2009. 14. Xu, W.: “Experimental Investigation of Dynamic Interfacial Interactions at Reservoir Conditions,” M.S. Thesis, Louisiana State University, Baton Rouge, Louisiana, 2005. 15. Li, J., Wang, W. and Gu, Y.: “Dynamic Interfacial Tension Phenomenon and Wettability Alteration of Crude Oil-Rock-Alkaline-Surfactant Solution Systems,” SPE paper 90207, presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, September 26-29, 2004. 16. Al Mahrouqi, D.A., Vinogradov, J. and Jackson, M.D.: “Understanding Controlled Salinity Waterflooding in Carbonates Using Streaming Potential Measurements,” SPE paper 177242, presented at the SPE Latin American and Caribbean Petroleum Engineering Conference, Quito, Ecuador, November 18-20, 2015. 17. Zakarian, A.: “Bond Dissociation Energy,” 2014, available at: https://labs.chem.ucsb.edu/zakarian/ armen/11—bonddissociationenergy.pdf. Last accessed January 22, 2016.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 43

87448araD7R1.indd 43 11/28/16 12:37 PM BIOGRAPHIES Dr. Ali A. Yousef is the Chief Technologist of Reservoir Engineering Mohammed A. Geer has been Technology and Team lead of the working for Saudi Aramco’s Improved/Enhanced Oil Recovery Exploration and Petroleum (IOR/EOR) research program in the Engineering Center – Advanced Saudi Aramco Upstream Advanced Research Center (EXPEC ARC) for Research Center. He has more than the past 3 years, during which time he 24 years of experience in upstream research and participated in and led the macro-scale technology. Since joining Saudi Aramco, Ali has been research activities focused on SmartWater flooding. Prior involved in applied research projects on IOR, to this assignment, he worked for the Reservoir waterflooding and EOR. He played a pivotal role in Management Department managing the water injection planning, developing and implementing the EOR roadmap system in ‘Uthmaniyah field. for the company. Ali is currently leading more than 50 Mohammed has published multiple papers in the field EOR scientists, engineers, and technicians. Those EOR of SmartWater flooding, and he has filed and been granted researchers are dedicated to the development of various several patents. EOR processes, including SmartWater, carbon dioxide, In 2015, he won the Iain Hillier Academic Award, and chemical EOR technologies as well as other novel given by the London Petrophysical Society for the best processes with the clear target of meeting the company petrophysics-related research. objective. He was the impetus for Saudi Aramco’s Mohammed received his B.S. degree in Petroleum SmartWater revolution: a patented award-winning Engineering from King Fahd University of Petroleum and technology that was proven in the laboratory and Minerals (KFUPM), Dhahran, Saudi Arabia, and his M.S. demonstrated in the field through single well tests. degree in Petroleum Engineering from Imperial College Ali received the Society of Petroleum Engineers London, London, U.K. (SPE) 2016 prestigious IOR Pioneer Award at the IOR conference in Tulsa, OK for his pioneering contributions Dr. Ahmed Gmira is a Petroleum made to the advancement of enhanced water flooding Scientist with the SmartWater Team processes in carbonates. of Saudi Aramco’s Exploration and He has written over 60 technical papers and reports Petroleum Engineering Center – and has more than eight patents. Ali is currently an active Advanced Research Center (EXPEC member of SPE and has chaired several SPE workshops ARC). His main interests are and forums, helped organize several petroleum engineering enhanced oil recovery, SmartWater related conferences, and taught courses on IOR/EOR and flooding, fluids-fluids interfaces and fluids-rocks interfaces. waterflooding. He is considered a worldwide authority in Ahmed joined Saudi Aramco in April 2015. Prior to the field of IOR/EOR. this, he worked as a Research Scientist in the Schlumberger Ali received his B.S. degree in Chemical Engineering Research Centers in Dhahran, Saudi Arabia, and in Rio from King Fahd University of Petroleum and Minerals de Janeiro, Brazil. He also worked as a Research Fellow (KFUPM), Dhahran, Saudi Arabia, and his M.S. and in a postdoctoral position with the Department of Physics Ph.D. degrees, both in Petroleum Engineering, from the at the Norwegian University of Science and Technology, University of Texas at Austin, Austin, TX. Trondheim, Norway. Ahmed received his Ph.D. degree in Physico-Chemistry Dr. Sultan M. Al-Enezi is a Petroleum from the University of Orléans, Orléans, France. Engineer with the Reservoir Engineering Technology Division of Saudi Aramco’s Exploration and Petroleum Engineering Center – Advanced Research Center (EXPEC ARC). His research interests include enhanced oil recovery, digital rock physics, petrophysics and fluid flow in porous media. Sultan received his B.S. degree in Industrial Chemistry from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia. He also received his M.S. and Ph.D. degrees in Petroleum and Natural Gas Engineering from Pennsylvania State University, University Park, PA.

44 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD7R1.indd 44 11/28/16 12:37 PM Reservoir Stress Path from 4D Coupled High Resolution Geomechanics Model: A Case Study for Jauf Formation, North Ghawar, Saudi Arabia

Authors: Otto E. Meza Camargo, Dr. Tariq Mahmood and Dr. Ivan Deshenenkov

ABSTRACT the grid model using the following:

This study presents the full in situ stress tensor results, i.e., ΔShmin = 0.8055*ΔPP(Pore_Pressure) + 0.3762 (1) orientation and magnitude, for the Jauf formation at pre- and

post-production conditions, and their use to build a 4D cou- The Shmin stress average maps were generated from pre- and pled one-way geomechanics model. These results were con- post-production conditions in three stratigraphic levels (zones) strained by using existing data from wireline logs, downhole of the Jauf formation — Upper, Middle and Lower. This measurements and laboratory tests. After building 1D to 3D study concluded that the Upper Jauf is relatively more stressed Mechanical Earth Models (MEMs) using data from 27 wells, than the Middle Jauf and Lower Jauf. Breakdown pressure the 3D model was used as input to the 4D coupled model for maps for these three stratigraphic levels were also generated years A, B, C and D. and are presented in this study. This study concluded that the Jauf formation in the study area is characterized by a strike-slip-faulting regime in which INTRODUCTION

the maximum horizontal stress (SHmax) is the largest principal The main objective of this study was to define the full in situ stress, i.e., SHmax > vertical stress (Sv) > minimum horizontal stress tensors at pre- and post-production conditions for the stress (Shmin). The SHmax orientation, N75°E, was constrained by using borehole image logs. Jauf formation in the North Ghawar area and to build a 4D coupled one-way geomechanics model for the years A, B, C The calibrated stress models were established based on and D. 1D to 3D geomechanics models were built utilizing all poro-elastic equations, fracture closure pressures (FCPs), core available data for 27 wells across the study area, and the 3D data, wellbore stability models and drilling events depicting geomechanics model was used as a prerequisite input for the an average anisotropy ratio of approximately 1.2 to 1.4 (max- 4D coupled one-way geomechanics model. imum principal stress magnitude)/(minimum principal stress The 4D calibrated model will be used to identify field-scale magnitude). The stress model at pre-production conditions hydraulic fracturing “sweet spots” and to optimize hydraulic showed values for the pore pressure gradient of approximately

~0.62 psi/ft, for the Shmin gradient of ~0.71 psi/ft to 0.95 psi/ft

and for the SHmax gradient of ~1.3 psi/ft to 1.4 psi/ft. The 3D geomechanics high resolution grid was created for elastic properties and rock strength parameters propagation; the latter was driven by the total porosity (PHIT) model as a controlling parameter. The range of the estimated values are:

• Young’s modulus from 1.2 Mpsi to 6.0 Mpsi. • Poisson’s ratio from 0.24 to 0.38. • Unconfined compressional strength (UCS) from 6.0 Kpsi to 16.0 Kpsi.

In the 4D coupled model, the FCP values from hydraulic

fractures were used to calibrate the Shmin at post-production conditions over years A, B, C and D. The predicted stress Fig. 1. S direction in study area (N75°E ± 10°) from borehole breakouts and model showed a good match with the FCP over the 27 wells Hmax Figure 1 used in this study. The reservoir stress path was defined over drilling-induced fractures in the Jauf formation (Upper, Middle and Lower).

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 45

87448araD8R1.indd 45 11/28/16 12:56 PM

Well Locations with WBI

Breakout Breakout-Strike

σmax≈ ~75°

Induce Fracture- Azimuth Induce Fracture

LogsLogs Corrections Corrections for for32 wells32 wells Fig. 1. SHmax direction in study area (N75°E ± 10°) from borehole breakouts and drilling-induced fractures in the Jauf formation (Upper, Middle and Lower).

Gr Lithology AI PHT Density CAL DT DTS ROCK PHYSICS ANALYSIS

Well log corrections were performed for the sonic and density logs of all 27 wells using core information and multilinear regression equation analysis. Results from one of the wells are illustrated in Figs. 2 and 3.

Rock Physics Modeling

Rock physics models are mathematical equations based on physical principles that are used to generate P and S veloci- ties based on rock structure, composition and properties. By setting some of the parameters, i.e., clay content, sand bulk and shear modulus, the equations can be solved to model ve- High locities in missingPHIT(v/v) vs Vpdata(m/s) / VQUAintervals.(v/v) The mainHigh advantagePHIT(v/v) vs AIof (Kg.m/cm3.s) this / VQUA(v/v) 200 ft approach is that all relations between elastic properties and AI AI PHIT PHITRHOBRHOBDCAL DCAL DT DTDTSM DTSM rock quality are preserved.

The rock physics modelIncreasing isPorosity focused onAI predicting dynamic Increasing Porosity Increasing Vshale Increasing Vshale compressionalVP velocity (Vp) and shear velocity (Vs) as accu- VP rately as possible, producing results suitable for further me- Fig.Fig. 2.2. Well Well log log corrections corrections were were performed performed for sonic for and sonic density and logs, density using corelogs, using core information and information and multilinear regression equation analysis. chanical modeling. Using the porosity and clay content from Low

multilinear regression equation analysis. Low Low Porosity wellLow logs, an analysisPorosity was performedHigh to determine which rock High 0 1 0 1 physics models are most suitable. The advanced differential High PHIT(v/v) vs Vp (m/s) / VQUA(v/v) High PHIT(v/v) vs AI (Kg.m/cm3.s) / VQUA(v/v) effective medium modeling described by Prasad and Nur (2003)3 was used to estimate P and S velocities. A tangential High High PHIT(v/v) vs Vs (m/s) / VQUA(v/v) PHIT(v/v) vs RHOB (G/C3) / VQUA(v/v) shear factor was introduced to obtain an optimal match with

Increasing Porosity AI Increasing Porosity Increasing Vshale Increasing Vshale Increasing Vshale VP observed Vp/Vs ratios in the sandstones, since contact theory VP is known to overpredict shear wave velocities by neglecting Increasing Porosity

Low rotational freedom and slip at grain contacts. Low RHOB VS Increasing Vshale Low Porosity Low Porosity High High 0 1 Increasing Porosity 0 1 Vp and Vs are functions of porosity, clay content, differ- ential pressure and saturation. The setup of the input pa- Low rametersLow was completed iteratively to find the best solid clay High High PHIT(v/v) vs Vs (m/s) / VQUA(v/v) PHIT(v/v) vs RHOB (G/C3) / VQUA(v/v) Low Porosity High Low Porosity High properties0 for this dataset. This can1 be 0thought of as inverting 1 Increasing Vshale for solid clay elasticity, assuming that all other properties are

Increasing Porosity known and that our model is correct. Calculated Vp and Vs Fig. 3. Cross-plots were used to analyze the trends and relationship between different rock parameters RHOB VS Increasing Vshale were then calibrated on dynamic mechanical properties de- Increasing Porosity with respect to the PHIT. rived from the core analysis to obtain the best fit between all

Low Low

Low Porosity High Low Porosity High 0 1 0 1

High PRD(v/v) vs AI (Kg.m/cm3.s) / PHIT(v/v)

Quartz Fig. 3.3. Cross-plotsCross-plots were were used used to analyzeto analyze the trends the trends and relationship and relationship between between different rock parameters withdifferent respect rock toparameters the PHIT. with respect to the PHIT.

fracture design. In addition, the 4D geomechanics model will be used in well placement optimization and in the analysis of AI High PRD(v/v) vs AI (Kg.m/cm3.s) / PHIT(v/v) wellbore stability for horizontalQuartz wells.

The maximum horizontal stress (SHmax) direction, N75°E ± 10°, was inferred from drilling-induced tensile fractures and borehole breakouts1, 2 identified in the borehole image logs. Increasing Porosity Calcite FigureAI 1 shows the results of the stress direction analysis. No Low

significant azimuthal rotation of the principal horizontal stress Low PR Dynamic High was identified across the study area. 0 0.2

Increasing Porosity Calcite Fig. 4. Cross-plots between the Poisson’s ratio and acoustic impedance showing Fig. 4. Cross-plots between the Poisson’s ratio and acoustic impedance showing the scatter of data over the scatter of data over the Quartz zone. Low the Quartz zone. Low PR Dynamic High 0 0.2 46 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY Fig. 4. Cross-plots between the Poisson’s ratio and acoustic impedance showing the scatter of data over the Quartz zone.

87448araD8R1.indd 46 11/28/16 12:56 PM WELL XX_A WELLWELL XX_E XX_B WELL XX_C

Density Density Density DTS DTS DTS Gr Lithology DT AI PR YMD PHT CAL Gr Lithology DT AI PR YMD PHT CAL Gr Lithology DT AI PR YMD PHT CAL

200 ft

Fig. 5. Ultrasonic core test data was used to calibrate the dynamic rock properties derivate from sonic and density logs. TimeDate vs vs Pressure Gradient Gradient i ltsoni oe test t ws se to libte1.200 the ni o popeties eite o soni n High ensit los 1.000

0.800

Pore Pressure Initial 5000 5000 5000 5000 0.600 Gradient from XX_J Gradient Low Leak Off Test Pressure Leak Off Test Pressure Fracture Close Pressure 0.400 Fracture Close Pressure BHSP BHSP Pressure Fracture Close Pressure MDT Gradient Pressure MDT 0.200 Pore_Pressure_3D simulations 7000 7000 7000 7000

0.000Low PP≈~ 0.49 19-Apr-2001 14-Jan-2004 10-Oct-2006 06-Jul-2009 01-Apr-2012 27-Dec-2014 22-Sep-2017 PP≈~ 0.49 PP≈~ 0.49 PP≈~ 0.49 Low High TimeTime

9000 9000 9000 9000 Fig. 7. Pore pressure derived from 3D simulations and FCP plotted over time, ishow Poe well-defined pesse trends.eie oThe FCP-derived siltions “stress n Ppath” plotte is interpreted oe tie to showresult welleine tens he Pderived “stress path” is interpreted to result from depletions. from depletions.

TVD (ft.) TVD (ft.) TVD (ft.) TVD (ft.) available data, Figs. 4 and 5. 11000 11000 11000 11000

TVD The dynamic rock properties obtained from ultrasonic core KUFF Unit ~ KUFF Unit ~ KUFF Unit ~ KUFF Unit ~ tests were used to calibrate the dynamic Poisson’s ratio (PRD) and dynamic Young’s modulus (YMD) for Well XX_A, Well YMD(PSI) vs YMS (PSI) YMD(PSI) vs YMS (PSI) 13000 13000 13000 13000 XX_E and Well XX_C. BKDC Unit ~ BKDC Unit ~ BKDC Unit ~ BKDC Unit ~

JAUF Unit ~ JAUF Unit ~ JAUF Unit ~ JAUFPORE Unit ~ PRESSURE AND MINIFRAC DATA 2000 ft

15000 15000 15000 15000 The pore pressure model at pre-production conditions was Fig. 6. FCP shows a wide dispersion effect (from 0.71 psi/ft to 0.98 psi/ft), which generated from direct measurements obtained through the can be due to depletions. High Shmin≈~ 0.71 modular formationPP≈~ 0.62 dynamicsShmin≈~ 0.71 tester and from bottom-hole PP≈~ 0.62 Shmin≈~ 0.71 PP≈~Figure 0.62 PP≈~Shmin≈~ 0.62 0.71 6 0 0.2 0.4 0.6 0.8 01Low1.20.2Pore0 0.4 Pressure5000 0.610000 0.8Grad 150001 High1.2 20000 0 5000 10000 15000 20000 SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 47 Pressure Gradient (Psi/ft.) Pressure Gradient Pressure(Psi/ft.) (Psi.) Pressure (Psi.) i iil oe n eltionship

i P shows wie ispesion eet o psit to psit whih n be e to epletions

87448araD8R1.indd 47 11/28/16 12:56 PM

TimeDate vs vs PressurePressure Gradient Gradient

1.200 High

1.000

0.800

Pore Pressure Initial 0.600 Gradient from XX_J Gradient

0.400

Pressure Fracture Close Pressure Gradient Pressure 0.200 Pore_Pressure_3D simulations

0.000Low 19-Apr-2001 14-Jan-2004 10-Oct-2006 06-Jul-2009 01-Apr-2012 27-Dec-2014 22-Sep-2017 Low High TimeTime

i Poe pesse eie o siltions n P plotte oe tie show welleine tens he Pderived “stress path” is interpreted to result from depletions.

static pressure (BHSP) estimated from hydraulic fracturing in

0.6 0.6 the Jauf formation, Fig. 6. Eaton’s method was used for the YMDRHOB(PSI) (G/C3)vs vsYMSPRS (PSI)(ratio) YMD(PSI)PHI (v/v)vsvsYMSPRS (ratio)(PSI) pore pressure estimation, and the results were calibrated with 0.5 0.5 the modular formation dynamics tester measurements, BHSP, 0.6 0.6 RHOB(G/C3) vs PRS (ratio) PHI (v/v) vs PRS (ratio) mud weight (MW) and drilling events. 0.4 0.4 0.5 0.5 The BHSP estimated from hydraulic fracturing in the Jauf 0.3 0.3 formation shows a wide variability, changing from 0.58 psi/ft 0.4 0.4 to 0.62 psi/ft. This effect may be due to the depletions in the PRS (ratio) PRS 0.2 (ratio) PRS 0.2 zones where the hydraulic fractures were performed. 0.3 0.3 0.1 The minimum stress values were estimated from the lea- PRS = -0.0891*Density + 0.4841 0.1

PRS (ratio) PRS (ratio) PRS PRS = 0.1844*PHIT + 0.2515 0.2 R² = 0.024 0.2 R² = 0.012 koff test and the fracture closure pressure (FCP). These also 0 0 show a large variability across the study area, with values 0.11.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 0.10.00 0.05 0.10 0.15 0.20 0.25 0.30 PRS = -0.0891*Density + 0.4841Density (g/c3 ) PorosityPRS (v/v) = 0.1844*PHIT + 0.2515 ranging from 0.71 psi/ft to 0.98 psi/ft. The initial pore pres- R² = 0.024 R² = 0.012 0 0 sure gradient in Well XX_J can be estimated as 0.62 psi/ft. 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 0.00 0.05 0.10 0.15 0.20 0.25 0.30 The lower values are predicted because hydraulic fractures 0.6 Density (g/c3 ) 0.5 Porosity (v/v) Fig. 8. Tri-axial core, YMS and YMD relationship. YMS(MPSI) vs PRS (ratio) 0.45 PRD(ratio) vs PRS (ratio) were performed over the depletion zones, which can reduce 0.5 i iil oe n eltionship 0.4 the FCP, Fig. 7. 0.6 0.5 0.35 0.4 (MPSI) (ratio) (ratio) (ratio) YMS vs PRS 0.45 PRD vs PRS 0.5 0.3 ROCK MECHANICAL PROPERTIES CORRELATIONS 0.4 0.3 0.25 PRS (ratio) PRS (ratio) PRS 0.35 0.4 0.2 Values for the YMD, shear modulus, bulk modulus and 0.2 0.3 0.15 Poisson’s ratio of the rock were generated from the compres- 0.3 0.25 0.1 (ratio) PRS (ratio) PRS 0.1 sional sonic, shear sonic and density logs. Converted static PRS = 1E-08*YMS + 0.2473 0.2 PRS = 1.1068*PRD 0.05 0.2 R² = 0.069 R² = 0.55 properties using empirical correlations from triaxial tests were 0 0.15 0 0 2000000 4000000 6000000 8000000 10000000 12000000 also estimated. From these triaxial tests, a relationship was de- 0.1 0.1 0 0.1 0.2 0.3 0.4 0.5 YMS (Psi )PRS = 1E-08*YMS + 0.2473 PRD (ratio ) PRS = 1.1068*PRD rived between the static Young’s modulus (YMS) and YMD, 0.05 R² = 0.069 R² = 0.55 0 Fig. 8. A derivative of the equation is: 0 0 2000000 4000000 6000000 8000000 10000000 12000000 YMS (Psi ) i. . Static0 Poisson’s 0.1 0.2 ratio plotted 0.3 with 0.4 the PHI 0.5 densit and . here are no diret relationships PRD (ratio ) 1.51606 identified eteen the parameters. YMS = 0.0001646742* YMD (2) Fig. 9. Static Poisson’s ratio plotted with the PHIT, density and YMS. There are

no directCore relationshipsCore –CoreLog – Log– correlations Logidentified correlations correlations between the parameters. i. .Core StaticCore – Log – Log Poisson’s correlations correlations ratio plotted with the PHI densitThe and PRD . was estimated here arewith no the diretporo-elastic relationships equation, and identified eteen the parameters. WELL XX_A WELL XX_C WELL XX_B CoreCoreCore –Core –Log Log– Log– correlations Logcorrelations correlations correlations Core – LogDensity correlations TS Density TS Density TS DTS UCS DTS UCS DTS UCS Gr Lithology DT PRDWELLYMD YMSXX_APRS F_AN PHT CAL Gr Lithology DT PRD YMDWELLYMS XX_BPRS F_AN PHT CAL Gr Lithology DT PRD YMDWELLYMS XX_CPRS F_AN PHT CAL

Density TS Density TS Density TS DTS UCS DTS UCS DTS UCS Gr Lithology DT PRD YMD YMS PRS F_AN PHT CAL Gr Lithology DT PRD YMD YMS PRS F_AN PHT CAL Gr Lithology DT PRD YMD YMS PRS F_AN PHT CAL

200 ft

RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDPRDYMDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSUCSPHITPHITUCSDCALPHITPHITDCALDCALPHITDCALDCAL F_AngF_AngF_AngF_AngF_Ang RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDYMDPRDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSPHITPHITUCSDCALPHITDCALDCALPHIT DCAL 200 ft F_AngF_AngF_AngF_AngF_Ang

RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDPRDYMDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSUCSPHITUCSPHITDCALPHITPHITDCALPHITDCALDCALDCAL RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDPRDYMDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSUCSPHITPHITUCSDCALPHITPHITDCALDCALPHITDCALDCAL F_AngF_AngF_AngF_AngF_Ang F_AngF_AngF_AngF_AngF_Ang RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDYMDPRDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSPHITPHITUCSDCALPHITDCALDCALPHIT DCAL F_AngF_AngF_AngF_AngF_Ang Fig. 10. Well XX_A, Well XX_B and Well XX_C with static rock mechanics properties modeled using dynamic properties and core correlations. RHOBRHOBRHOBRHOBPRDRHOBPRDPRDPRDYMDPRDYMDYMDYMDYMSYMSYMDYMSYMSPRSPRSYMSPRSPRSUCSUCSPRSUCSUCSPHITUCSPHITDCALPHITPHITDCALPHITDCALDCALDCAL i. . ell ell F_AngF_Ang andF_AngF_AngF_Ang ell ith stati ro mehanis properties modeled usin dnami properties and ore orrelations. 48 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

i. . ell ell and ell ith stati ro mehanis properties modeled usin dnami properties and ore orrelations. 87448araD8R1.indd 48 11/28/16 12:56 PM

vs vs Low TVD 2000 ft

SV_Gradient≈~ 1.1 Psi/ft SV_Gradient≈~ 1.1 Psi/ft High

Low SV High

i ertical stress was estiated ro the densit lo

WELL XX_D

Mineralogy MW SV Shear Failure Density UCS Shmin Kick DTS PRS TS SHmax Losses SH/Sh Gr SW DT YMS F_Ang PP Breakdown CAL PHIT Shmin BHI

MW MW

FCP FCP BHSP BHSP

200 ft

BreakoutBreakout

SH/Sh SH/Sh SV SV

RHOB RHOBYMS PRSYMS PRSUCS F_AngUCS F_AngPore PoreShmin ShminSHmax SHmax DCAL DCALPHIT PHITShmin Shmin PressurePressure grad grad

Fig. 11. Vertical stress was estimated from the density log. Fig. 12. Stress model and wellbore stability model for Well XX_D. Figure 11 i Stress odel and wellore stailit odel or ell the static Poisson’s ratio was defined based on several cor- poro-elastic and horizontal strain stress models, where the

relations from the triaxial core tests using the porosity, bulk Shmin and SHmax at each depth depended upon: density, YMS and PRD results. No direct relationship was inferred between these properties, Figs. 9 and 10. • Mechanical properties • Pore pressure Static_Poisson_ratio = 1.15*Dynamic_Poisson_Ratio (3) • Vertical (overburden) stress The unconfined compressional strength (UCS) was esti-

mated using the multi-correlation between the YMS and total The Shmin was also constrained by the FCP and leakoff porosity (PHIT). The equations’ result was: test data4. The stress profiles were plotted together with MW windows.

UCS = 0.001405728*(YMS) - 19821.21*(PHIT) + The Shmin, as estimated from the hydraulic fracture data, 5828.503 (4) was found to lie in the range of approximately ~0.85 psi/ft to ~0.98 psi/ft.

1D MECHANICAL EARTH MODEL (MEM) The SHmax was estimated using the poro-elastic model, the wellbore stability model and drilling events. The wellbore The Jauf formation is characterized by a strike-slip-faulting stability model was also calibrated with the borehole image interpretation — identifying drilling-induced tensile fractures regime in which the SHmax is the largest principal stress — SHmax and breakouts. The estimated gradient of maximum principal > vertical stress (Sv) > minimum horizontal stress (Shmin). The horizontal stress magnitude is approximately ~1.3 psi/ft. orientation, N75°E, of the SHmax was inferred from borehole breakouts and drilling-induced tensile fractures interpreted from borehole images. Stress Calibrations Using Drilling Events

Vertical Stress Figure 12 shows the final wellbore stability calculation for Well XX_D. No major drilling issues were observed in the The overburden, or vertical stress, was evaluated using the wells drilled through the Jauf formation except for Well XX_B density logs, which showed an average gradient of approxi- (drilled in 2009), which reported a gas kick. The MW used to mately ~1.1 psi/ft, Fig. 11. drill through this well was 9.8 pounds per gallon. There is good agreement between the calculated breakouts and the actual Horizontal Stresses breakouts as interpreted from the borehole images5. Figure 12 also shows the main results from the 1D geome-

The Shmin and SHmax profiles were estimated using the chanics process as follows:

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 49

87448araD8R1.indd 49 11/28/16 12:56 PM Porosity (v/v) Cube Static Poisson’s Ratio Cube (ratio)

Fig. 13. Geomechanics grid orientation following the S direction. Figure 13 Hmax

Fig. 15. Modeled 3D static Poisson’sFigure ratio. 15

Porosity (v/v) Cube Static Young’s Modulus Cube (psi)

Porosity (v/v) Cube Unconfined compressional strength (Psi) Cube

High High Fig. 14. Cross-plot over the 3D geomechanicsFigure 14 grid properties extrapolations.

• Track 1. Gamma ray. PHIT PR • Track 2. Stratigraphy through the reservoir section.

• Track 3. Measured depth. Low Low • Track 4. Mineralogical model estimated from

petrophysics interpretation. Fig. 16. Modeled 3D YMS. Figure 16 • Track 5. Water saturation estimated from petrophysics Porosity (v/v) Cube UCS Cube (psi) interpretation.

Z-values: RHOBR_Cor effective_Porosity, [ft3/ft3] • Track 6. Density, compressional slowness and shear 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 Density [g/cm3] 2.8000 20000 2.7000High 2.6000 2.5000 2.4000 slowness. 18000 2.3000Low vs High 2.2000 16000

• Track 7. Static rock properties — Poisson’s ratio (light 16000 18000 20000 14000 14000 UCS, [psi]

green) and Young’s modulus (pink). 12000 12000 10000 10000 UCS

• Track 8. Rock strength properties — UCS (light green), 8000 6000

friction angle (dark blue) and tensile strength (orange). 6000 8000 4000

Low4000 • Track 9. Calculated stress profiles — pore pressure High 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 Low Low PHIT High UCS (dark blue), S (green), S (red), Sv (black), FCP Symbol legend hmin Hmax effective_Porosity vs. UCS vs. RHOBR_Cor (All cells) effective_Porosity vs. UCS vs. RHOBR_Cor (Upscaled) (red dot) and BHSP (pink dot). i odeled S Matrix Case • Track 10. Calculation of the stable MW window; limits are calculated for kicks (gray), breakouts (red), mud Fig. 17. Modeled 3D UCS. Figure 17 losses (blue) and formation breakdown (purple), and the drilling MW (dark yellow line) is also shown. Method:Method Gravity :GravityPressure Pressure • Track 11. Caliper logs. Sh Gradient: 0.71 psi/ft • Track 12. PHIT. Sh Gradiend :0.71 Psi/ft

• Track 13. Ratio between S /S (red) and the S SH/Sh: 1.35 Hmax hmin hmin SH/Sh :1.35 gradient (pink). Sh Azimuth: 165° • Track 14. Breakout from borehole image interpretation Sh Azimuth :165 Degree (red). Fig. 18. Boundary conditions used on the 3D MEM.

50 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY i ondar conditions sed on the

87448araD8R1.indd 50 11/28/16 12:56 PM Maximum Horizontal Stress Tensor Maximum Horizontal Stress Magnitude (psi)

Fig. 19. SHmax tensor stress estimated from geomechanics simulations.

3D GEOMECHANICS PROPERTY MODEL FigureThe 19 rock mechanics properties equations (Eqns. 2 and 3) were applied to estimate the YMS, Fig. 16, and the static 3D Grid Model Poisson’s ratio based on the dynamic properties. The rock strength properties’ UCS was estimated using Eqn. 4, Fig. 17.

The geomechanics grid was oriented to follow the SHmax direc- tion, N75°E, Fig. 13. The vertical and horizontal resolutions 3D IN SITU STRESS MODEL from the geomodel in the geomechanics grid were retained to capture the wide variability in the model. The geomechanics 3D Stress Estimations grid dimensions are: The first stage of a 3D stress analysis involves calculating • Total number of cells: 12 million. stresses that represent the pre-production conditions through- out the reservoir and its surroundings. Due to the complex • xinc = 100 m; yinc = 100 m; zinc ≈ 4.5 ft. variations in structure and properties within the model, the stress equilibrium must be solved numerically. A finite element Propagations of Geomechanics Properties method was used to determine the required solution, produc- ing a 3D map of stress magnitudes and orientations that vary The propagations of the properties were performed using the both laterally and vertically. The model then uses the structure porosity model because it provided good correlation between and rock mechanical properties defined in the preceding sec- Vp, Vs and density. A cokriging algorithm was used to ex- tions, together with the loads that govern stresses — gravita- trapolate the well log data, which included Vp, Vs and bulk tional, pore pressures and boundary conditions — to simulate density, Fig. 14. the initial stress state of the field. The 3D static rock properties were estimated based on the The results are calibrated to in situ stress profiles from the rock mechanics properties equations defined from the triaxial 1D MEMs. To address a number of uncertainties in the input core test, with the range of the static Poisson’s ratio being data, such as the assigning of fault stiffness and the propaga- modeled from ~0.24 to ~0.38, Fig. 15. tion of mechanical properties, a number of simulations were

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 51

87448araD8R1.indd 51 11/28/16 12:56 PM Fig. 20. Comparison between the 1D stress model (black) and the 3D stress model (S in green, Sv in red and S in pink) for Well XX_J. hmin Hmax Fig. 21. Comparison between the 1D wellbore stability model (track 3 in dotted lines) and the 3D wellbore stability model (track 3 in filled colors) for Well XX_J. performed to assess the sensitivity to those parameters, Fig. 18.

The goal is therefore to ensure that the general magnitude Stress Initializations Figure 21 Figure 20 and trend of the stresses is honored, while also expecting dif- ferences in the details. The comparison is valuable, however, The model was initialized on year A as describing pre-pro- because it provides a way of checking that the overall cali- duction conditions based on two cases: Matrix and Matrix + bration of the 3D model is correct, particularly regarding the Faults. Both models considered the same boundary conditions, choice of boundary conditions. and the maximum, minimum and vertical stresses were com- In general, the match between the stresses obtained from puted as tensor stress, Fig. 19. the 1D MEM (stress model) and the 3D MEM (stress model) is good at initial conditions, Fig. 20. In the Matrix case, the Comparisons between the 1D and 3D Stress Models relatively good match was achieved in 11 wells in the Jauf formation. When comparing the stresses between the 1D and 3D MEMs, it is not expected that all three principal stresses of the 3D MEMs will match those of the 1D MEMs exactly. One of the Comparison of 1D and 3D Wellbore Stability Models advantages of the 3D model is that it accounts for interactions that the 1D approach is unable to, for example: The 3D wellbore stability model was estimated using the 3D geomechanics model and compared with the 1D wellbore stability results. In both cases, the match was consistent under • Vertical stress being different from the 1D integration of the boundary conditions defined for the 3D geomechanics overburden weight. simulations model. Some of the wells were considered de- • Magnitude and orientation changes of stress due to pleted in the 1D MEM (wellbore stability model) for calibra- nearby faults. tion proposes. The estimated kick (pore pressure gradient), • Consistent stress and strain variation between strong shear failure (breakouts), mud loss and breakdown pressure and weak layers. all showed a good match, Fig. 21, Track 3.

52 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD8R1.indd 52 11/28/16 12:56 PM Fault cells TensorTensor stress stress in in the fault’s fault’s path path

Slight rotation

Slight rotation

Fig. 22. Fault simulations case, showing the normal and shear stress for cells within faults. Jauf Formation: Shedgum Field 9000

High8500 • Fault shear stiffness: 2.35E+06 KPa/m.

8000 7500 • Friction angle: 20°. 7000 6500 • Cohesion: 1 Kpa. 6000

5500

5000 4500 From the borehole image interpretations, no significant (Psig) Pressure 4000

Pressure 5 Well 3500 rotation of stress-induced features — drilling-induced tensile 3000 Pore Stress Step Stress Step Stress Step StressFigure Step 22 Static 2500 A B D fractures and breakouts — was observed along depth, sug- 2000 C 1500 gesting either that the influence of faults and fractures was 1000 Static Pressure (Psig) 500 minimal or that there are no minor faults in the vicinity of Low 0 Low Date High these wells. Although some slight rotations can be identified in 1-Jan-05 1-Jan-06 1-Jan-07 1-Jan-08 1-Jan-09 1-Jan-10 1-Jan-11 1-Jan-12 1-Jan-15 1-Jan-13 1-Jan-14 1-Jan-02 1-Jan-03 1-Jan-04 1-Sep-05 1-Sep-06 1-Sep-07 1-Sep-08 1-Sep-09 1-Sep-10 1-Sep-11 1-Sep-12 1-Sep-13 1-Sep-14 1-Sep-01 1-Sep-02 1-Sep-03 1-Sep-04 1-Aug-15 1-May-05 1-May-06 1-May-07 1-May-08 1-May-09 1-May-10 1-May-11 1-May-12 1-May-13 1-May-14 1-May-15 1-May-02 1-May-03 1-May-04 the tensor stress model for the faults, these are still inside the Fig. 23. Pore pressure and stress steps defined for years A, B, C and D. i Pore pressre and stress steps deined or ears and S range. The normal stress and shear stress were computed Hmax for each cell in the faults showing stable conditions, Fig. 22. WELL XX_J Faults StabilityHigh Analysis WELL XX_J Upper 4D COUPLEDShmin PP ONE-WAY GEOMECHANICS MODEL Shmin To evaluate the fault stability6 at pre-production conditions, WELL XX_D Low Middle the normal stiffness and shear stiffness were estimated — un- Stress Step and Pore Pressure Changes PP Shmin der the same boundary conditions — using the following geo- mechanics fault properties: The stress step calibrations were selected from the pore pres- FCP FCP sure profile to simulate the stress conditions over years A, B, • Fault normal stiffness: 7.67E+06 KPa/m. C and D, Fig. 23. FCP

WELL XX_D @A SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 53

i crosssection showin the Shin initialiation copared with the P hdralic ractre oer 87448araD8R1.inddear 53 or ear 11/28/16 12:56 PM

Fig. 24. 3D cross-section showing the Shmin initialization, compared with the FCP (hydraulic fracture over year A) for year A.

Fig. 25. 3D cross-section through Wells XX_D to XX_J, showing the Shmin at pre-production conditions and over time.

Stress Variations over Time In year D, the FCPs from Well XX_Z and Well XX_K were

compared with the 3D Shmin, showing a good match, Figs. 26 The stress calibration from pre- to post-production conditions and 27. was performed taking the FCP measured from the wells over time7. In Figs. 24 and 25, the FCPs from Well XX_D and Reservoir Stress Path

Well XX_J were compared with the 3D Shmin simulation for initial conditions (year A), reaching a good match. In normal depletion behavior, the effective vertical stress

54 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD8R1.indd 54 11/28/16 12:56 PM Fig. 26. 3D cross-section showing the Shmin, compared with the FCP (hydraulic fracture year D) for year D.

Fig. 27. 3D cross-section showing the Shmin through Well XX_Z and Well XX_K at pre-production conditions and over time.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 55

87448araD8R1.indd 55 11/28/16 12:56 PM Upper Upper Middle Middle WELL XX_K WELL XX_K

WELL XX_Z WELL XX_Z

High High

Shmin Shmin @A @B Low Low

Upper Upper Middle Middle WELL XX_K WELL XX_K

WELL XX_Z WELL XX_Z

High High

Shmin Shmin

Low @C @D Low

i crosssection showin the Shin throh ell and ell at preprodction conditions and oer tie

vs Pore Pressure (Psi/ft) vs FCP (Psi/ft) 1.200 1.200High High

1.100 1.000

1.000

0.800 ΔShmin=0.8055*ΔPore_Pressure + 0.3762 0.900 R2 =0.5548

0.600 0.800

Shmin FCP Shmin

0.700FCP 0.400 Middle Lower

0.600 Pressure Gradient Fracture Close Pressure 0.200 Pore_Pressure_3D simulations 0.500

Low 0.000Low 19-Apr-2001 14-Jan-2004 10-Oct-2006 06-Jul-2009 01-Apr-2012 27-Dec-2014 22-Sep-2017 0.400 Low High Low Time 0.20 0.30 0.40 0.50 0.60PorePore 0.70 Pressure Pressure 0.80 0.90 1.00 1.10High 1.20

Fig. 28. The pore pressure from 3D simulation results and FCP plotted over time, Fig. 29. Reservoir stress path defined by the relationship between hminS /Pp and FCP ishowing he the possiblepore pressre trend for the ro reservoir silation stress path. reslts and P plotted oerithrough tie to eseroir theshowin reservoir stressthe depletions. possile path deined the relationship etween ShinPp and P throh to the trend or the reseroir stress path reseroir depletions

Z-values: Zones (hierarchy) 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600

PRESSURE, [psi] TOTAL_EFFSTR_HMIN[3]_2014, Z-values: Zones (hierarchy) 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600

Zones (hierarchy) Zones (hierarchy) Upper_Jauf Upper_Jauf 13600 High Middle_Jauf 13600 High Middle_Jauf Lower_Jauf Lower_Jauf 13600 13600 12800 12800 12800 12800 12000 12000 12000 12000 [psi] TOTAL_EFFSTR_HMIN[1]_2008 TOTAL_EFFSTR_HMIN[3]_2014 11200 11200 [psi] 11200 11200 Shmin_3D Shmin_3D 10400 10400 10400 10400 9600 9600 9600 9600 8800 Shmin=0.8055* Pore_ Pressure +5578 @A 8800 Shmin=0.8055* Pore_ Pressure +5578 @B 8800 Low8800 Low

5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600 TOTAL_EFFSTR_HMIN[0]_2002, High Low PRESSURE,Pore Pressure [psi] High Low Pore Pressure Symbol legend Symbol legend PRESSURE vs . TOTAL_EFFSTR_HMIN[0]_2002 vs . Zones (hierarchy) (All cells) TOTAL_EFFSTR_HMIN_3_2014_vs_PRESSURE TOTAL_EFFSTR_HMIN_3_2014_vs_PRESSURE PRESSURE vs . TOTAL_EFFSTR_HMIN[1]_2008 vs . Zones (hierarchy) (All cells)

Z-values: Zones (hierarchy) Z-values: Zones (hierarchy)

PRESSURE, [psi] TOTAL_EFFSTR_HMIN[3]_2014, PRESSURE, [psi] 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600

Zones (hierarchy) Zones (hierarchy) Upper_Jauf Upper_Jauf Middle_Jauf 13600 High Middle_Jauf 13600 High Lower_Jauf Lower_Jauf 13600 13600 TOTAL_EFFSTR_HMIN[3]_2014, 12800 12800 12800 12800 12000 12000 12000 12000 [psi] TOTAL_EFFSTR_HMIN[2]_2011 11200 11200 11200 11200 Shmin_3D Shmin_3D 10400 10400 10400 10400 [psi] 9600 9600 9600 9600

8800 Shmin=0.8055* Pore_ Pressure +5578 8800

8800 Shmin=0.8055* Pore_ Pressure +5578 @C 8800 Low Low @D 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 9600 5200 5600 6000 6400 6800 7200 7600 8000 8400 8800 9200 High9600 Low Pore Pressure High Low Pore Pressure Symbol legend Symbol legend TOTAL_EFFSTR_HMIN_3_2014_vs_PRESSURE PRESSURE vs . TOTAL_EFFSTR_HMIN[3]_2014 vs . Zones (hierarchy) (All cells) PRESSURE vs . TOTAL_EFFSTR_HMIN[2]_2011 vs . Zones (hierarchy) (All cells) TOTAL_EFFSTR_HMIN_3_2014_vs_PRESSURE i eseroir stress path deined the relation etween Pp and Shin ro silations oer depletions tie

Fig. 30. Reservoir stress path defined by the relation between Pp 3D and S 3D from simulations, over depletions time. Figurehmin 30 56 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD8R1.indd 56 11/28/16 12:56 PM Year A

Fig. 31. Year A S average stress map for the Upper (left), Middle (middle) and Lower Jauf (right). hmin Figure 31 increases at the same rate that pore pressure decreases because Stress Maps the total vertical stress is unaffected by changes in pore pres- sure. Subsequently, the effective horizontal stress increases more Shmin stress average maps were generated for each stress step slowly during depletion than the pore pressure decreases. The and the three levels defined for the Jauf formation, showing total horizontal stress decreases with pore pressure following important changes throughout the formation. 8-10 the Shmin/pore pressure coupling ratio , Fig. 28. For the pre-production conditions (year A), Fig. 31, the Figure 29 shows the relationship defined between the S / hmin Shmin stress average map shows: pore pressure ratio with a value of 0.805 and the FCP, which can be considered as a normal range for a strike-slip-faulting • In the Upper Jauf section, high stress values from the

regime. It was assumed that the drop in Shmin with the pore southwest area, e.g., Well XX_L, and lower values from

pressure is nonrecoverable and that the Shmin does not increase the central part of the reservoir, e.g., Well XX_D. with a re-pressurization process. • In the Middle Jauf section, lower stress in comparison In the 4D coupled one-way geomechanics model, the with the Upper Jauf section, but still with high stress in stresses obtained for the three stratigraphy levels defined for the southwest area — Well XX_L. the Jauf formation — Upper, Middle and Lower — show al- • In the Lower Jauf section, high stress in the southwest most the same behavior regarding the pore pressure and Shmin relationship. This can be deduced from Fig. 30, where the area but less stress in the central area, e.g., Well XX_D. Middle and the Lower Jauf are overlapping. During the depletion process over years A, B, C and D, the The post-production conditions for years B, C and D show

pore pressure and Shmin follow the stress path previously defined important reductions in the stress due to depletions in the in Fig. 29, reaching in year D depleted zones where the pore Middle Jauf. The stress map over the Lower Jauf also shows pressure values are less than the original pressures. Following significant reductions in the central part of the reservoir near the stress path defined at post-production conditions, separate Well XX_D. The southwest part keeps the high stress values for trends were identified for the Upper, Middle and Lower Jauf. the Upper, Middle and Lower Jauf around Well XX_L, Fig. 32.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 57

87448araD8R1.indd 57 11/28/16 12:56 PM Year D

Fig. 32. Year D S average stress map for the Upper (left), Middle (middle) and Lower Jauf (right). hmin Figure 32 CONCLUSIONS gave a good match with the FCP over the 27 wells where data was available, with the reservoir stress path defined over the The Jauf formation in the study area is characterized by a grid model by the following equation:

strike-slip-faulting regime in which the SHmax is the largest S = 0.8055* PP + 0.3762 (5) principal stress, i.e., SHmax > Sv > Shmin. The SHmax orientation, Δ hmin Δ (Pore_Pressure) N75°E, was constrained by using borehole image logs.

The calibrated stress models were established based on po- The Shmin stress average maps were generated from pre- to ro-elastic equations, FCP, core data, wellbore stability models post-production conditions for the three stratigraphy levels — and drilling events showing an average anisotropy ratio of ap- Upper, Middle and Lower Jauf — showing more stress in the proximately 1.2 to 1.4 (maximum principal stress magnitude)/ Upper Jauf in comparison with the Lower and Middle Jauf (minimum principal stress magnitude). The stress model at sections. The southwest area shows the highest stress values pre-production conditions showed values in the pore pressure of the three levels, > ~13,000 psi, even at post-production conditions. gradient of ~0.62 psi/ft, the Shmin gradient of ~0.71 psi/ft to

0.95 psi/ft, and the SHmax of around ~1.3 psi/ft to 1.4 psi/ft. The high resolution 3D geomechanics grid was used to ACKNOWLEDGMENTS propagate the elastic properties and rock strength using the PHIT model as the main drive; the ranges of estimated values The authors would like to thank the management of Saudi were as follows: Young’s modulus from 1.2 Mpsi to 6.0 Mpsi, Aramco for their support and permission to publish this ar- Poisson’s ratio from 0.24 to 0.38, and UCS primarily from ticle. The authors would also like to thank Khaqan Khan, 6.0 Kpsi to 16.0 Kpsi. Jubril Oluwa and Thamer A. Sulaimani for their contributions In the 4D coupled model, the FCP values from the hydrau- to this project.

lic fractures were used to calibrate the Shmin at post-production conditions over years B, C and D. The predicted stress model

58 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD8R1.indd 58 11/28/16 12:56 PM REFERENCES BIOGRAPHIES

Otto E. Meza Carmargo joined Saudi 1. Bell, J.S.: “Practical Methods for Estimating In Situ Stresses Aramco in 2014 as a Geologist for Borehole Stability Applications in Sedimentary Basins,” Engineer working in the Exploration Journal of Petroleum Science and Engineering, Vol. 38, Technical Services Department. He Nos. 3-4, June 2003, pp. 111-119. has 10 years of experience in the oil 2. Prensky, S.: “Borehole Breakouts and In Situ Rock Stress industry, and prior to joining Saudi — A Review,” The Log Analyst, Vol. 33, No. 3, 1992, pp. Aramco, he worked in a variety of 304-312. geological positions across the Middle East region, Brazil, Peru and the U.K. Otto’s experience includes several 3. Prasad, M. and Nur, A.: “Velocity and Attenuation aspects of reservoir characterization and integrated Anisotropy in Reservoir Rocks,” SEG paper 2003-1652, geomechanics modeling for conventional and presented at the Society of Exploration Geophysicists unconventional reservoirs. International Exposition and Annual Meeting, Dallas, He received his B.S. degree in Geological Engineering Texas, October 26-31, 2003. from the University of San Marcos, Lima, Peru. 4. Raaen, A.M., Skomedal, E., Kjørholt, H., Markestad, P. and Økland, D.: “Stress Determination from Hydraulic Dr. Tariq Mahmood joined Saudi Fracturing Tests: The System Stiffness Approach,” Aramco in 2008 and is currently working as a Geological Specialist/ International Journal of Rock Mechanics and Mining Geomechanics Team Leader in the Sciences, Vol. 38, No. 4, June 2001, pp. 529-541. Exploration Technical Services 5. Kirsch, E.G.: Die Theorie der Elastizität und die Department. He began his career in Bedürfnisse der Festigkeitslehre, Springer, Dresden, 1996 working in Perth, Australia, for Germany, 1898, 11 p. Z&S Geoscience/Baker Hughes, specializing in fractures/ 6. Zoback, M.D.: Reservoir Geomechanics, Cambridge faults characterization from borehole images. Tariq has University Press, Cambridge, U.K., 2010, 461 p. provided consultancies to the major oil companies in the Asia Pacific region and Australia, including Shell Brunei, 7. Koutsabeloulis, N.C., Heffer, K.J. and Wong, S.: Petronas, Chevron, Santos, Apache, Woodside, etc. “Numerical Geomechanics in Reservoir Engineering,” He is a member of the American Association of in H.J. Siriwardane and M.M. Zaman (eds.), Computer Petroleum Geologists (AAPG), the European Association Methods and Advances in Geomechanics, A.A. Balkema, of Geoscientists and Engineers (EAGE) and the Dhahran Rotterdam, The Netherlands, 1994. Geosciences Society (DGS). 8. Santarelli, F.J., Tronvoll, J.T., Svennekjaier, M., Skeie, H., In 1996, Tariq received his Ph.D. degree in Structural Geology from the University of Adelaide, Adelaide, Henriksen, R. and Bratli, R.K.: “Reservoir Stress Path: The Australia. Atlas: 3D Analogue Modelling of Extensional Depletion and Rebound,” SPE paper 47350, presented at Fault Systems Plus Field Applications (1995), published the SPE/ISRM Rock Mechanics in Petroleum Engineering, by the University of Adelaide, et al., included his Ph.D. Trondheim, Norway, July 8-10, 1998. research. 9. Hillis, R.R.: “Coupled Changes in Pore Pressure and Stress in Oil Fields and Sedimentary Basins,” Petroleum Geoscience, Vol. 7, No. 4, December 2001, pp. 419-425. 10. Breckels, I.M. and van Eekelen, H.A.M.: “Relationship between Horizontal Stress and Depth in Sedimentary Basins,” Journal of Petroleum Technology, Vol. 34, No. 9, September 1982, pp. 2191-2199.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 59

87448araD8R1.indd 59 11/28/16 12:56 PM Dr. Ivan Deshenenkov is a Geologist with the Exploration Technical Services Department. He joined Saudi Aramco in 2013 with over 9 years of petrophysical experience in both service companies and exploration and production companies in Russia, the U.S. and France, specializing in petrophysics, rock physics, digital rock physics and special core analysis. Ivan holds several patents and has authored more than 30 technical papers. During his career, he has received several awards, including the Russian President Grant, the Society of Petroleum Engineers’ (SPE) STAR Fellowship and the American Association of Petroleum Geologists’ (AAPG) Gustavus E. Archie Memorial Grant for research work. He is a member of AAPG, the European Association of Geoscientists and Engineers (EAGE) and the Dhahran Geosciences Society (DGS). In 2013, Ivan received his Ph.D. degree (with honors) in Petrophysics and Petroleum Engineering, with a concentration in capillary pressures, relative permeability analysis and reservoir production forecasting, from Gubkin Russian State University of Oil and Gas, Moscow, Russia.

60 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD8R1.indd 60 11/28/16 12:56 PM Implementing the Pressurized Mud Cap Technique for Drilling through Total Loss Zones: A Way to Improve Well Control while Drilling the Reservoir in Oil Well Reentries

Authors: Khalifah M. Al-Amri, Julio C. Guzman Munoz, Abdulrhman M. Al-Hashim, Ali M. Hassanen and Ayoub Hadj-Moussa

ABSTRACT INTRODUCTION

Drilling through total loss zones without returns at the surface Reentering old wells to access bypassed portions of the is a common practice for operating companies in the Middle reservoir is one of the most common jobs in the workover East. This is known as the mud cap drilling technique, a operation. Workover engineers spend a great deal of time and “blind drilling” technique that has been used for a long time effort to come up with designs to accommodate the conditions in the region. The lessons learned throughout these years have of the existing wellbore so the reentry well can reach the produced well established procedures and “rules of thumb” required targets in a cost-effective manner. for estimating the mud density, volume and pumping schedule Currently, one of the most common reentry designs applied of the mud cap for different situations. in one of the main oil fields in Saudi Arabia is the long radius The application of this mud cap (blind) drilling technique sidetrack. The wells following this design are initiated by becomes more challenging when drilling through oil reservoirs cutting a window in the existing 7” casing, then drilling a

with a high hydrogen sulfide (H2S) content and through a 6⅛” curve section with dogleg severities around 5°/100 ft combination of zones in the same hole section with pressure through the overlying formations to land the well horizontally regimes that are significantly different. This situation is often at the top of the reservoir. Drilling continues using geosteering the case when drilling wells used to reenter oil producer wells techniques with logging while drilling tools until the well in Saudi Arabia fields. Under these circumstances, the main reaches the planned total depth (TD). Finally, the opened well control strategy is to maintain enough of a mud cap to upper formations are isolated from the producing lateral by prevent any migration of the hazardous gases and/or liquid means of a cemented off-bottom liner, Fig. 1a. hydrocarbons to the surface. The main drilling hazards found during the workover This approach often results in a conservative design reentries come from having different pressure regimes requiring large mud cap densities and volumes, which in combined in the same hole section. These hazards are as- turn generates huge expenditures in mud material and puts sociated with (1) the relatively high pressure in the water tremendous stress on logistics and transportation. producer, Formation-A, and (2) the nature of Formation-B, an The pressurized mud cap technique offers a safe alternative oil reservoir that is highly fractured in some zones of the field. to blind drilling. This technique uses regular managed These two create conditions for experiencing differentially pressure drilling (MPD) equipment to monitor the behavior stuck pipe or a total circulation loss. Moreover, dealing with of the reservoir at all times, thereby improving the safety of total circulation loss while drilling through the oil reservoir the operation by having accurate well control. This approach can produce dangerous well control situations, which are

also realizes substantial cost savings by optimizing the use of severely aggravated by the high hydrogen sulfide (H2S) content drilling fluid materials and other resources associated with the found in the reservoir fluids. mixing and pumping of the mud cap. Conventional managed pressure drilling (MPD) has been This article describes the successful implementation of this used in the past to optimize the overbalance applied to the technique for reentering oil wells in an onshore field in Saudi lower pressure zone and to reduce the risk of getting the Arabia. The article also provides details on the equipment drillpipe differentially stuck. But conventional MPD is not and procedures utilized to maintain a full column of fluid applicable once total losses are encountered. This led to the under pressure by keeping the MPD choke fully closed while idea of using the pressurized mud cap technique, a variant drilling without any returns at the surface. The results, when of MPD, as an alternative to the traditional blind drilling compared with the conventional mud cap (blind) drilling to continue drilling the reentry under total circulation loss technique, show significant improvements in safety (well conditions, all while providing accurate well control and control), cost efficiency and logistics. minimizing fluid lost to the reservoir.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 61

87448araD9R1.indd 61 11/28/16 1:22 PM Fig. 1a. A typical oil reentry well, showing a window is cut into the existing 7” casing. Fig. 1b. Cross flow from the higher pressure formation into the low-pressure formation.

CONCEPTUAL FRAMEWORK THE PRESSURIZED MUD CAP TECHNIQUE

Conventional MPD, understood as the “constant” bottom- Once the total loss of drilling fluid is encountered, the hole pressure (BHP) variation of MPD1, does not solve the pressurized mud cap technique uses MPD equipment — challenges presented by workover reentry operations in Saudi drillpipe check valves, rotating head and choke — to maintain Arabia. The main reason is that to maintain a constant BHP, a full column of viscous fluid with a density lighter than that it is necessary to have a “closed system.” The regular closed required to control the pore pressure found in the highly circulating system existing in any ordinary drilling operation fractured formation. In this way, a positive pressure is is lost once the drilling fluid starts massively leaking into the maintained in the MPD choke at the surface. downhole formations. Drilling continues with a sacrificial amount of fluid being Initially, it was thought that the conventional MPD was pumped through the string to drive directional tools and lift going to help prevent these drilling fluid losses by keeping the cuttings to the loss zone; the MPD choke remains fully the wellbore pressure profile within the mud weight (MW) closed. The positive choke pressure permits operators to window and walking along the line between well kicks monitor the well to detect any hydrocarbons migrating to the and losses. This approach was attempted but failed to have surface, any increase in drilling fluid losses or any evidence positive results. It became clear that conventional MPD was that the formation is plugged. not the right application for this environment due to the It is important to emphasize that the pressurized mud cap pre-existence of karst type fractures with extremely high approach will work only if the formation fracture or karst is permeability. big enough to receive all the fluid and cuttings being dumped Pressurized mud cap drilling is another variation of MPD1, as a result of the change in drilling technique. If the formation and a quick literature review confirmed the suitability of the starts getting plugged, the operation will have to switch to technique for drilling through highly fractured hydrocarbon conventional MPD. 2, 3 zones with H2S content . It was decided to take a hybrid In this particular field, operators suspected that the high- approach to the implementation of MPD in the oil reentry pressure formation, Formation-A, had started to cross wells: start with conventional MPD and then switch to flow and to dump formation water into the oil reservoir, pressurized mud cap drilling once total loss of the drilling Formation-B, as represented in Fig. 1b. At first, they expected fluid is encountered. The two main objectives of this effort to use the pressurized mud cap technique only for drilling and are to (1) reduce the frequency and severity of differentially tripping; however, as the crew became familiar with it, the stuck pipe and well control events, and (2) minimize the total implementation was extended for the logging and completion amount of drilling fluid used during the reentry. phases, including running and cementing the off-bottom liner.

62 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD9R1.indd 62 11/28/16 1:22 PM

Probability Severity (Impact) Low: Incident/scenario being analyzed has not Low: Less than 12 hours rig lost time. occurred in the region or frequency is less than one incident every five years. Medium: Incident/scenario being analyzed has Medium: Between 12 hours and 24 hours rig lost occurred in the region and frequency is more than time. one incident per every 5 years and less than once yearly. High: Incident/scenario being analyzed has a high High: Rig lost time is more than 24 hours. potential for occurrence or historical data shows a frequency of more than one incident per year.

Probability Severity (Impact)

Low: Incident/scenario being High analyzed has not occurred in the Low: Less than 12 region or frequency is less than hours rig lost time. one incident every five years.

Medium Medium: Incident/scenario being analyzed has occurred in the Medium: Between 12 region and frequency is more hours and 24 hours than one incident per every 5 rig lost time. years and less than once yearly. Low High: Incident/scenario being analyzed has a high potential High: Rig lost time is for occurrence or historical data more than 24 hours. Low Medium High shows a frequency of more than one incident per year.

Fig. 2. The matrix for probability and impact risks. Fig. 2. The matrix for probability and impact risks.

RISK ASSESSMENT AND FINAL PROCEDURE Uncontrolled Flow to the Surface — Rotating

Control Device (RCD) Failure

The peer review and risk assessment process identified

the following main issues for the implementation of the The probability of uncontrolled flow to the surface was

pressurized mud cap technique. Some of the issues are related considered a low risk; there are no records of catastrophic to conventional MPD, but are mentioned here due to their rotating control device (RCD) failures causing a sudden relevance for the pressurized mud cap technique as well. For release of fluids to the atmosphere in regular MPD operations simplification purposes, the risk scenarios listed here are based in Saudi Aramco or in the region. Obviously, the severity was on a 3 × 3 Risk Matrix, Fig. 2. deemed as a high risk due to personnel considerations. Rig well control equipment would take care of the situation to Differentially Stuck Pipe in Low-Pressure and/ contain the momentary release, but to maintain the severity or High Permeability Formations to as low a risk as was reasonably practical, it was decided to keep the surface working pressure at a maximum of 500 psi The concern was that once operators started implementing at any given time. It is important to mention that the pressure the pressurized mud cap, the clear, solids-free fluid used as the rating for the RCD used is 2,500 psi, and the rig blowout “mud” cap was going to deplete the bridging agent — regular preventer stack is rated at 5,000 psi.

calcium carbonate (CaCO3) — in the formations uphole of Formation-A and Formation-B, promoting the tendency for Drillpipe Check Valve Failure having the drillstring differentially stuck; the probability of this happening was considered a medium risk. Even though In the event of a check valve failure, flow or pressure through such depletion is not normally observed in regular blind the drillpipe would occur during the connection. Because drilling while pumping the mud cap, in this new scenario, several failures have been associated with check valves in the fluid acting as the mud cap would not have any of the MPD operations in the region, the probability was considered bridging agents common in regular mud caps. a high risk. The severity was a medium risk, mainly associated The severity was considered a medium risk as well, since with the lost time spent tripping and replacing the valves. it was hoped that the differential sticking situation would be As a risk mitigation factor, it was decided to use the valves managed by manipulating the pressure in the backside with provided by the MPD company, which have a better track the MPD choke. To mitigate this risk, the decision was made record than the valves provided by the rig and/or the di- to use resilient bridging material — graphite and gilsonite — rectional drilling company. Also, a minimum of two of these to set a good durable cake in these upper formations. The use valves were requested in the final procedure. In the event of of these damaging bridging agents would then be discontinued a drillpipe check valve failure, the procedure to follow was before starting to drill the reservoir. to kill the well, switch to the conventional mud cap (blind) drilling technique and trip out of the hole to fix the problem.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 63

87448araD9R1.indd 63 11/28/16 1:22 PM Drillstring Washout or Failure Tripping

Pressure containment is lost if the drillstring fails. The The trip for conventional MPD is performed with the well probability of this happening was considered a medium to low killed. The trip for the pressurized mud cap application can be risk, and the severity was deemed as a medium to low risk performed using the rotating head and pressurized fluid in the as well, mainly associated with the rig lost time and the need backside all the way to “light pipe” depth. At this point, the to kill the well by bullheading. To mitigate the risk, it was well is controlled by bullheading kill mud, and the operation decided to perform a full drillstring inspection to the DS-1 Cat switches to conventional mud cap (blind) drilling techniques, 3-5 level prior to beginning with the MPD and pressurized which involves frequent pumping of a certain volume of mud mud cap operation. In the event this situation happened, the cap that is heavier than the kill drilling fluid. The empirical contingency was to kill the well by bullheading and continue best practice, learned over the years, is to pump (for the mud remedial operations with the conventional mud cap (blind) cap) 60 barrels (bbl) per hour of fluid that is 10 pounds per drilling technique. cubic foot (pcf) heavier than the kill mud. The main tripping risks — RCD failure and drillpipe

Hydrocarbons and H2S Migrating to the Surface check valve failure — have already been mentioned, as well as their respective mitigations. Additionally, the light pipe Preventing such a migration is actually an advantage of the depth for the drillstring would have to be calculated prior to pressurized mud cap technique compared with the regular tripping and after any pressure change observed during the mud cap (blind) drilling technique. The probability of having trip, making sure that the calculation included a 100% safety

hydrocarbons and H2S migrate uphole to the surface was factor. The procedure involves pumping viscosified water in considered as a medium to high risk. The severity was deemed the backside to replace the drillstring volume being pulled out as a medium risk since having the RCD rigged up was going of the hole. to actually mitigate the consequences of having any gas or It is necessary to keep in mind that the operating company, oil cut mud to surface — when compared with the regular like others in the region, has extensive experience in drilling blind drilling. The risk mitigation strategy was basically the without returns to the surface using the conventional mud concept of the pressurized mud cap itself. With it, the rig team cap (blind) drilling technique. Because of this, risks related to was going to be able to monitor the behavior in the drillpipe drilling mechanics and hole cleaning were not included for casing annulus, thereby detecting any gas or hydrocarbon this analysis since those have already been addressed in the migration in time to take the corrective actions well before blind drilling practices. the hazardous fluids rose to the surface. The main additional A common operational risk faced when using any of the barrier used to reduce the chances of having gas migration MPD variations is the drilling crew’s and rig team’s lack to the surface was the utilization of viscosified water — 25 of familiarity with MPD equipment and procedures. The lb/100 ft2 to 30 lb/100 ft2 yield point — as the fluid in the MPD service provider thereby played a very important role pressurized mud cap. in training all crews on-site. The numerous pre-job meetings held with the key members of the well site management team Formation Plugging were also crucial to explain the objectives and details of the application. For this particular case, the footage drilled in In the event the loss zone(s) get plugged and returns are the “dead” rock found in the formations above the high- regained, the cuttings would be trapped in the annulus pressure zone — Formation-A — was used to get the crews space since the choke is fully closed. This would severely familiar with the connections, tripping and directional survey affect the hole cleaning capabilities and create a big risk of procedures as well as with the MPD equipment. hole pack off. Also, since the pressurized mud cap — water The simplified procedure for the application of the — would be substantially underbalanced with the higher pressurized mud cap technique for drilling through a total loss pressure Formation-A, the pressure at the surface might zone is summarized as follows: increase to values higher than desirable. The probability of having this situation develop was considered a low risk, 1. Once total losses are observed during conventional MPD based on the extensive experience with conventional blind operations, close the MPD choke fully and completely fill drilling; the severity was deemed as a high risk, mainly for the annular space until positive pressure is observed in the the potential stuck pipe and well control issues. As a point of casing pressure MPD gauge. Keep filling until 100 psi to risk mitigation, the procedure included alarms set to flag any 150 psi casing pressure is reached. Continue drilling with abnormal increase in casing pressure; two full hole volumes of the choke fully closed and pump sacrificial water through kill mud were ready to be used at any time; and the full MPD the drillstring; also pump a 30 bbl hi-vis water pill for package was primed to switch to conventional MPD in case every stand drilled to assist with hole cleaning. full circulation was regained.

64 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD9R1.indd 64 11/28/16 1:22 PM a. Fill the hole with viscosified water — XC polymer at 3. If total losses are observed and it is not possible to follow 1.5 ppb + water — to obtain a viscosity of 25 lb/100 ft2 the above procedures, maintain positive casing pressure to 30 lb/100 ft2. with one of the two actions: b. Have the MPD engineer record the exact amount of a. Pump 3 bpm with viscosified water. water used to fill up the annulus space. b. If it is not possible to keep up with viscosified water, 2. Continue drilling with the choke fully closed while switch to plain water and pump a 30 bbl pill of vis- observing the pressure behavior in the casing pressure cosified water every hour. MPD gauge. This pressure should remain stabilized. a. If the casing pressure drops below 50 psi, pump vis- ACTUAL IMPLEMENTATION cosified water at 1 barrel per minute (bpm) to bring the pressure back to the previous value, then carry on Operations commenced in the selected well. Figure 3 shows drilling to TD. the MPD equipment that was rigged up after cutting the b. If the casing pressure starts steadily increasing and window in the 7” liner. A good pressure profile was obtained does not show signs of stabilization while the drillpipe while drilling through the upper interval. The high-pressure pressure remains steady at the same value, bullhead zone was detected in Formation-A, and an exact pore pressure the annulus volume with viscosified water to the top value was obtained for the maximum pressure found in this of Formation-A at 8 bpm, then stop bullheading and formation. This value was used later on for all killing fluid observe the pressure behavior. density and mud cap density calculations. • If the pressure is below 500 psi — the risk Drilling in a conventional MPD mode continued. The well mitigation limit set by management — carry on was landed as planned into the reservoir in Formation-B, and drilling while monitoring the casing pressure. total losses were encountered approximately 400 ft after the • If the pressure is greater than 500 psi, stop drilling target entry point. A total of 2,488 ft were drilled horizontally and perform calculations for the new MW to — via geosteering — after hitting the total loss zone. The be used as a pressurized mud cap. Bullhead the procedure described previously was utilized to drill the total annular space with the new pressurized mud cap loss section using a pressurized mud cap. The surface casing — at a heavier MW — and continue drilling, closely pressure remained very steady at ± 300 psi, and no major monitoring the casing pressure and repeating the changes were observed during connections. Figure 4 is a previous procedures as needed. sample of the pressurized mud cap log. The drilling BHA was c. If the drillpipe pressure increases at the same time that pulled out of the hole while maintaining the pressurized mud the casing pressure does, the loss zone is probably cap, and the well was killed, when the bit was approximately getting plugged. Attempt to establish full circulation 2,000 ft from the surface. After killing the well, conventional with conventional MPD and then continue drilling. mud cap (blind) drilling procedures were followed for laying

RCD Main flow line

Microflux control manifold GasGas toto ventvent

Trip tank fill up

Trip tank pump

Fig. 3. MPD equipment configuration used for the pressurized mud cap technique implementation. Fig. 3. MPD equipment configuration used for the pressurized mud cap technique implementation.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 65

87448araD9R1.indd 65 11/28/16 1:22 PM

TOTAL LOSSES: Zero Returns (Red Line)

TOTAL LOSSES: Surface Pressure to Zero

(Red Line)

Pressurized Cap: Surface Pressure Steady at ±300 psi. (Red Line)

Fig. 4. Actual MPD log while drilling using the pressurized mud cap technique. Fig. 4. Actual MPD log while drilling using the pressurized mud cap technique. down the drilling BHA and picking up the next reaming BHA. The off-bottom liner was then run with the pressurized A pressurized mud cap was used again for the reaming trip mud cap, and returns were partially regained after stringing and the logging run. Memory pipe conveyed logs were run the seal assemblies into the lower completion. Full returns using the RCD and pressurized mud cap procedures. Also were obtained with MPD techniques, and the liner cementing the lower completion, consisting of inflow control devices job was performed successfully, with full returns, keeping and mechanical open hole packers, was run while using the control of the BHP with the MPD equipment. pressurized mud cap.

66 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD9R1.indd 66 11/28/16 1:22 PM RESULTS CONCLUSIONS AND OBSERVATIONS

All objectives were met and the actual results significantly Conventional drilling approaches or even conventional MPD exceeded all expectations for this first trial of the pressurized techniques won’t work in drilling scenarios like the one mud cap. presented in this article. When the dominant cause of the When compared with the conventional mud cap (blind) total losses is pre-existing karst type fractures, any attempt drilling technique, the pressurized mud cap offered a much to regain circulation before continuing drilling will require better control of the well and provided the ability to monitor an excessive amount of time and materials and presents very BHP and reservoir fluid behavior at all times. limited chances to succeed. The only options left are either to The main immediate cost benefit was obtained from continue drilling with the mud cap (blind) drilling technique eliminating the cost of the saved mud. A total of 17,000 bbl or to set a sacrificial casing if the well design allows it. of 95 pcf mud cap were saved in this well, thanks to the Pressurized mud cap drilling was demonstrated in this trial pressurized mud cap implementation. to be a safer and more cost-effective alternative to the con- The value goes beyond the monetary savings of these ventional mud cap (blind) drilling technique. The main benefit mud cap barrels themselves. The associated logistics and obtained with the application of this technique is the ability materials for preparing the 17,000 bbl, involving about 3,500 to accurately monitor the pressure profile in the wellbore by

metric tons of calcium chloride and CaCO3, represent ±190 means of the indications seen at the casing pressure gauge, truckloads and round trips to the rig site — approximately thanks to having a full column of mud filling the annulus 76,000 km of heavy traffic avoided on public roads. Also, space. This capability allows the drilling crew to react in time there was a corresponding reduction in operational exposure to take the corresponding corrective actions well before any for the rig personnel, who did not have to mix this large hazardous fluids reach the surface. amount of drilling fluid at 60 bbl per hour. This added benefit In addition to the huge cost savings that come from should not be understated since the mixing process involves achieving a reduction in the mud cap volumes of about numerous heavy load movements and other hazards. 70% and the obvious operational safety improvements In terms of the rate of penetration (ROP), there was not related to well control, there were at least three further a noticeable improvement during the application of the important benefits observed during the implementation of pressurized mud cap; the ROP in the lateral is very dependent this technique: on the amount of the intersected pay zone, and the observed ROP in the interval below the loss zone was the same as other • A dramatic reduction in the usage of fresh water for wells drilled in the area. It is fair to say, however, that while mixing mud. using the MPD in the upper section — when full returns were • Improved safety at the rig site since the mud mixing still in place — the observed ROP was about 33% higher than requirements were significantly reduced. in the offset wells. With these encouraging results, the technique was applied • A significant reduction in transportation requirements, in one more reentry well and later in three newly drilled wells. decreasing exposure to motor vehicle accidents. The results were consistent with the observations made in the first trial well. A diligent risk management process helped the team to develop a robust procedure with virtually all the contingencies FUTURE DEVELOPMENTS well covered. Many of the identified risks were easily mitigated by having the possibility to bullhead the well due to Currently, the pressurized mud cap technique is the operator’s the high permeability of the fractures. preferred way to drill reentry wells in zones where total losses are expected. The main urgency now is to bring the technique ACKNOWLEDGMENTS to offshore rigs, where the potential savings are even higher. During this first stage of the implementation, the operating The authors would like to thank the management of Saudi company is keeping the full MPD setup. This definitely Aramco, Halliburton and Weatherford for their support and permission to publish this article. The authors would also like offers a high degree of flexibility, especially useful to obtain to thank the operations and field personnel involved in the all the information in the zones above the lost circulation first implementation of the pressurized mud cap technique zone. Currently, the operator’s engineering team and the in Saudi Arabia, especially Askar S. Al-Hajri, Gabriel M. service providers are working on alternatives to optimize the Morava, Joshi Ainnikal and Hussain Ghazzaly. equipment requirements. Reducing the amount of equipment This article was presented at the SPE/IADC Middle East needed for a pressurized mud cap will help to enable an easier Drilling Technology Conference and Exhibition, Abu Dhabi, implementation of the technique and to capture more oppor- UAE, January 26-28, 2016. tunities for its application.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 67

87448araD9R1.indd 67 11/28/16 1:22 PM REFERENCES BIOGRAPHIES

Khalifah M. Al-Amri is a General 1. Chopty, J., Jaipersad, M., Arnone, M. and Sardo, A.: Supervisor at the Gas Drilling “Development of Technological Applications of Managed Engineering Department; he is th Pressure Drilling,” paper presented at the 10 Offshore currently in charge of the Offshore Mediterranean Conference and Exhibition, Ravenna, Italy, Gas Division, handling the drilling March 22-25, 2011. engineering for the most critical and 2. Colbert, J.W. and Medley, G.: “Light Annular Mud challenging offshore gas wells in Saudi Cap Drilling — A Well Control Technique for Naturally Aramco. During Khalifah’s more than 18 year-long career, Fractured Formations,” SPE paper 77352, presented at he has built in-depth knowledge and solid experience in all aspects of onshore and offshore well construction and the SPE Annual Technical Conference and Exhibition, San intervention processes for both oil and gas wells. Khalifah’s Antonio, Texas, September 29 - October 2, 2002. specialized expertise is in oil and gas workover engineering 3. Sweep, M.N., Bailey, J.M. and Stone, C.R.: “Closed Hole and operations, covering all procedures required during the Circulation Drilling: Case Study of Drilling a High-Pressure entire life cycle of the wells, high-pressure/high temperature Fractured Reservoir — Tengiz Field, Republic of drilling, underbalanced coiled tubing drilling and snubbing Kazakhstan,” SPE paper 79850, presented at the SPE/ operations. IADC Drilling Conference, Amsterdam, The Netherlands, He received his B.S. and M.S. degrees in Petroleum February 19-21, 2003. Engineering (with Second Honor) from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.

Julio C. Guzman Munoz is an Engineering Supervisor at the Workover Engineering Department in Saudi Aramco. During his 13-year career in the oil and gas industry, he has gained extensive experience working on several projects associated with coiled tubing drilling, extended reach drilling, high-pressure/high temperature gas wells, underbalanced drilling and workover operations in Venezuela, the United States, Mexico, Bahrain and Saudi Arabia. Currently, Julio is providing engineering support for several key initiatives for Saudi Aramco Drilling & Workover. Julio received his B.S. degree in Petroleum Engineering from the Universidad Industrial de Santander, Bucaramanga, Santan der, Colombia, in 1997.

Abdulrhman M. Al-Hashim is leading a unit in Saudi Aramco’s Workover Engineering Department. Previously, he worked in various departments within the Drilling & Workover (D&WO) Department, gaining extensive experience in onshore and offshore D&WO engineering. Abdulrhman has also worked on the Manifa Workover Campaign, the D&WO Environmental Stewardship program, a lump sum turnkey (LSTK) reentry contract and multiple other critical projects. He received his B.S. degree in Mechanical Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia. Abdulrhman is member of the Society of Petroleum Engineers (SPE).

68 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD9R1.indd 68 11/28/16 1:22 PM Ali M. Hassanen works for Halliburton in Saudi Arabia as a Drilling/Well Engineer (Design and Operation) for reentry wells. He has 11 years of experience in the oil and gas industry. Previously, Ali worked with the Khalda Petroleum Company (Apache Egypt) in drilling and workover for land wells and well intervention for oil, gas and water wells. His work experience also included time with the El Paso Egypt Production Company where he was involved in an exploration drilling project. Ali then moved over to work with TransGlobe Energy Corporation doing the same type of work. Finally, he worked for the Petrol Amir Petroleum Company in both drilling and workover operations prior to his current position.

Ayoub Hadj-Moussa is currently the Country Product Line Manager of Secure Drilling Services for Weatherford Saudi Arabia. With over 9 years of oil field experience, he specializes in controlled pressure drilling operations, such as underbalanced drilling (UBD) and managed pressure drilling (MPD). In 2005 Ayoub joined Weatherford, starting his oil field career as a Data Acquisition Engineer in Underbalanced Drilling Operations in Hassi Messaoud, Algeria, gaining valuable hands-on experience. Since then, Ayoub has worked on a broad range of UBD and MPD operations in various countries in the MENA region. In 2010 he relocated to Saudi Arabia to work on various projects, such as the Deep Gas Drilling Coil Tubing UBD project, the High-Pressure Jilh Formation Depletion project and the MPD project for Saudi Aramco’s Drilling and Workover Department. Ayoub has coauthored various Society of Petroleum Engineers (SPE) papers in the field of MPD. In 2004, he received his B.S. degree in Systems Engineering from Carleton University, Ottawa, Ontario, Canada.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 69

87448araD9R1.indd 69 11/28/16 1:22 PM Automatic Well Completion and Reservoir Grid Data Quality Assurance for Reservoir Simulation Models

Authors: Tariq M. Al-Zahrani, Muath A. Al-Mulla and Mohammed S. Al-Nuaim

ABSTRACT geologists and geophysicists develop structure models, geo- logical horizons and geological models for the reservoir; pet- Reservoir simulation engineers have to go through an intense rophysicists develop initial saturation and initial oil in place and time-consuming process to manually quality check large maps; and reservoir engineers provide the field development amounts of data, which consist of wells, logs and completion plan — all of which represent interpreted data. The raw and events going back for more than 50 years of history. It is interpreted data are then integrated when constructing the fundamental that data used in building reservoir simulation reservoir simulation model2. To date this had been a notably models be quality controlled and cross-checked to eliminate cumbersome process. inconsistencies during the history match process. Building reservoir simulation models is a challenging (and This article provides an overview of a new simulation data tedious) task, given that simulation models are essentially quality workflow design and outlines the development of a an integration of all collected geological and engineering state-of-the-art software package that will guide engineers multidisciplinary data. After building the model, reservoir through the quality checking process, seen as the first stage of simulation engineers have to go through an intense and simulation model construction. time-consuming process to manually quality check the large The workflow software package developed here covers most amounts of data, which consist of wells, logs and completion of the data required for building simulation models. The prepa- events going back for more than 50 years. It is fundamental ration and validation of data are conducted using tailored, that data used in building reservoir simulation models be automated workflows that guide engineers through the quality quality controlled and cross-checked to eliminate inconsisten- control steps in a streamlined manner. At every step, specific cies and problems during the history match process. The tool data is reviewed and validated based on some well-developed described in this article strives to develop new capabilities that quality control (QC) criteria. In case of any inconsistencies, the allow simulation engineers to easily prepare, validate, store software issues relevant error and/or warning messages and rec- and retrieve geological model related data, e.g., grids, well ommends possible remedial solutions or actions. rates, historical events, perforations, deviation surveys, etc., in This tool eliminates tedious informal manual effort and a secure and user-friendly manner. results in high quality simulation models that have been By using this automated reservoir grid and well completion checked and verified through formal automated workflows. data workflow, reservoir engineers can easily validate the The new workflow/software reduces the length of time spent simulation related data and so reduce the time spent in the on the checking process from weeks to days or even hours, history matching process, leading to a more accurate reservoir and thereby minimizes the overall time needed to perform history matching.

INTRODUCTION

During the production life cycle of oil and gas extraction from reservoir fields in geological formations, certain stages follow in a specific order: exploration, appraisal, reservoir develop- ment, production decline and abandonment of the reservoir. Massive amounts of data are collected in the process, such as seismic, well logs, core data and production data, all rep- resenting raw data1. In addition, during the life of the field development, several disciplines develop studies to increase understanding of different parts of the reservoir. For example, Fig. 1. Cross section where the colors indicate geological properties in a reservoir. Fig. 1. Cross section where the colors indicate geological properties in a reservoir. 70 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD10R1.indd 70 11/28/16 1:54 PM

Fig. 2. The nine QC steps.

Fig. 3. Example of the mismatch of well completions with model grids.

model. What follows is a better prediction of gas and/or oil actions of interpreting and integrating quality controlled data production. This tool will reduce the time spent in quality at different scales from a variety of sources and vintages are checking simulation related data from months to days. prerequisites for preparing a comprehensive reservoir model. Therefore, the validation of the geological model at the very GEOLOGICAL MODEL VALIDATION beginning of the larger overall process is a very important part of constructing the simulation model3. The geological model aggregates all the concepts and static measurements of reservoir properties into a full reservoir char- THE QUALITY CONTROL (QC) TOOL acterization. The geological model not only provides the geo- logical basis of the reservoir simulation model, but also serves As mentioned, it is time-consuming and critical work to check to assist with well planning and volumetric calculations, among all geological parameters. This is where the idea of this tool other uses. The geological model is constructed by integrating came from. This in-house developed system will assist in im- the reservoir geometry with rock characterization data3. proving the model quality, enforce best practices and reduce The geological model depicts the distribution of rock types the turnaround time required for building and conducting and properties in the reservoir, Fig. 1. It also includes geological reservoir simulation studies. The workflow consists of nine features, such as faults, fractures and stratiforms. Rock prop- QC steps, Fig. 2, that allow simulation engineers to easily val- erties such as porosity, permeability, and thickness of reservoir idate the geological model, especially the data related to the strata are described in the geologic model as well. Engineering grid and well completions, e.g., grid orientation, grid origin, data, such as pressures, fluid densities, free water levels, satura- flipped cells, grid horizon, well trajectories and completions, tions, etc., may also be included in the geologic model. Because and faults orientation, by using one of the pre-processing and the model is an interpretation, and interpretation is necessarily post-processing platforms. the result of the geologist’s judgment, knowledge and experi- The following are just a few of the aforementioned issues ence, the geological model may have some issues4. that reservoir simulation engineers can face when checking the Interpretation is often a step toward a more accurate and geological model: reliable geologic model, developing further as new data is gathered and new concepts are formulated. The geologic • When the well completions are integrated with the model is not a final product, but a milepost on the way to model grids, they sometimes do not share the same obtaining a description of the “actual” reservoir. The two depth zone, Fig. 3.

Fig. 2. The nine QC steps.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 71

87448araD10R1.indd 71 11/28/16 1:54 PM

Fig. 1. Cross section where the colors indicate geological properties in a reservoir.

Fig. 4. Example of the mismatch of well tops with grid horizons.

• When the horizons in the geological model are integrated with the well tops, sometimes those well tops do not overlay the horizons correctly, Fig. 4. Fig. 4. Example of the mismatch of well tops with grid horizons. • When the well trajectories are integrated with the well tops and the grids, they sometimes do not intersect with each other, Fig. 5.

Actually, building a simulation model is a good opportu- nity for the simulation engineer to go to a second level of QC to see the data from an integration perspective.

Fig. 5. Examples of the mismatch of well trajectories with grids (left) and well Fig.NINE QC2. STEPS The nine QC steps. Fig.tops (right).5. Examples of the mismatch of well trajectories with grids (left) and well tops (right).

At every step, specific data is reviewed and validated based Grid Validation on some well-developed QC criteria. The following discusses these nine QC steps. Validation of the grid requires checking the origin and lo- cation of the grid. This is the first step to performing QC of any other data. In this first QC step, the tool validates the reservoir grid index order automatically,Well A and it checks the origin of the grid by considering all possible reasons, which is implemented to generate the grid or grid affected because of the loading process by using a different format2. An output result is displayed in text format showing a 3D environment with the base horizon and the origin of the grid indicated with a point at the cell where I=1, J=1 and K=1. The tool validates Fig. 5.the Examples top left cell, whether of the it is (1,1,1) mismatch or not, then of it validateswell trajectories with grids (left) and well tops (right). the grid location by importing a well-known location well. Validating the grid location can be performed with reference

Fig. 6. Grid origin checked with one well location in 3D environment.

Fig. 3. Example of the mismatch of well completions with model grids. Well A Fig. 3. Example of the mismatch of well completions with model grids.

Fig. 4. Example of the mismatch of well tops with grid horizons. Fig. 4. Example of the mismatch of well topsFig. 6. Grid with origin checkedgrid with horizons. one well location in 3D environment. Fig. 6. Grid origin checked with one well location in 3D environment. 72 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD10R1.indd 72 11/28/16 1:54 PM

Fig. 5. Examples of the mismatch of well trajectories with grids (left) and well tops (right).

Well A

Fig. 6. Grid origin checked with one well location in 3D environment.

Well Tops Horizon which the tool provides statistical results of how many well Well Tops Horizon tops have a distance of more than the assigned threshold for each grid horizon. Also, the tool displays all the well tops that do not intersect with grid horizons in 3D, Fig. 7.

Grid Faults Fig. 7. Horizons checked against well tops.

Fig. 7. Horizons checked against well tops. In the third QC step, the tool automatically validates the grid Fig. 7. Horizons checked against well tops. faults against the interpreted faults to ensure that the grid

faults are modeled correctly, Fig. 8. All the grid faults should intersect with the imported interpreted faults. The simula- tion engineers are allowed to select a threshold, then the tool provides statistical results of how many grid faults have a distance of more than the assigned threshold for imported interpreted data. Also, the tool displays all the grid faults that do not intersect with imported faults in 3D.

Geometry of the Grid Cells

In the fourth QC step, the tool highlights all the simulation grids having issues related to cell angle, cell bulk volume, cell Fig. 8. Interpreted faults and model grid faults, highlighting issues graphically. Fig. 8. Interpreted faults and model grid faults, highlighting issues graphically.volume contrast and cells inside out that will significantly im-

to any of a number of existing objects, like wells, faults, etc., pact both the simulation convergence and the simulation com-

Fig.or to8. Interpreteda range of faults grids, and Fig. model 6. grid faults, highlighting issues graphically. puting time. This step is about checking the geometry of the

Property Units Total Total Total Threshold Bad Mingridimum cells Max andimum assuring Average that Status the criteriaDisplay related to the simulator Minimum Maximum Average Cells 3 GridCell Vol. Horizons ft 52,329 507,243,264 53,167,167 0 0 grid0 cell are0 met. The0 tool displays all the cells having issues PropertyCell Angle Unitsdeg Total 0.00 Total 0.64 Total 0.15Threshold0.81 Bad25 Minimum0 Maximum0.00 Average 0.00Status 0.00Display0.52 Minimum Maximum Average Cells in 3D, based on the threshold put forward by the simulation Areal 3 Cell Vol. ft 52,329 507,243,264 53,167,167 0 0 0 0 engineers,0 and provides full statistics for, and the status of, InCell the Angle second deg QC step,0.09 the tool8.96 automatically 4.16 validates25 the 0 0.00 0.00 0.00 Cell Angle deg 0.00 0.64 0.150.81 25 0 0.00 0.00 0.000.52 Vertical each cell’s properties, Table 1. gridAreal horizons against the well tops to ensure that the grid Cell Cell Angle deg 0.09 8.96 4.16 25 0 0.00 0.00 The0.00 properties that are generated in checking the geometry horizonsDimension are ftmodeled 635.93 correctly. 1,916.05 This step917.70 determines 0.00 the ac-2.295 635.93 1,916.05 917.70 Vertical X curacyX of the horizons in the grid by comparing them against of the grid cell as listed in Table 1 are discussed next. Cell theDimension well tops,ft ft which635.93708.69 are 1,916.05deemed1,974.32 to917.70 be interpreted 1,041.620.00 with2.2950.0 high635.932.295 1,916.05708.69 917.701,974.32 X 1,041.62 X XY confidence. The tool automatically picks the horizons of the Cell Volume. This refers to the bulk grid cell block volume; Cell Vol. 0.00 1 0.81 0.75 581 0.00 0.75 0.52 X DimensionContrast ft 708.69 1,974.32 1,041.62 0.0 2.295 708.69 1,974.32the porosity1,041.62 andX saturations are not accounted for in the cal- gridY provided for the grid validation — previously described. Cell Inside 0.00 0.00 0.00 0 0 culation.0.00 The0.00 minimum0.00 and maximum cell volume values of AllCellOut Vol. the well tops0.00 should intersect1 with0.81 the horizons.0.75 581 The simu0.00 - 0.75 0.52 X Contrast the entire grid are reported in the “Total Min.” and “Total lation engineers are allowed to select a threshold in feet, after Cell Inside 0.00 0.00 0.00 0 0 0.00 0.00 0.00 TableOut 1. Statistics describing the results of the cell validation step Total Total Total Bad Property Units Threshold Min. Max. Average Status Display Minimum Maximum Average Cells Table 1. Statistics describing the results of the cell validation step Cell Vol. ft3 52,329 507,243,264 53,167,167 0 0 0 0 0

Cell Angle deg 0.00 0.64 0.150.81 25 0 0.00 0.00 0.000.52 Areal Cell Angle deg 0.09 8.96 4.16 25 0 0.00 0.00 0.00 Vertical Cell Dimension ft 635.93 1,916.05 917.70 0.00 2.295 635.93 1,916.05 917.70 X X Cell Dimension ft 708.69 1,974.32 1,041.62 0.0 2.295 708.69 1,974.32 1,041.62 X Y Cell Vol. 0.00 1 0.81 0.75 581 0.00 0.75 0.52 X Contrast Cell Inside 0.00 0.00 0.00 0 0 0.00 0.00 0.00 Out

Table 1. Statistics describing the results of the cell validation step

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 73

87448araD10R1.indd 73 11/28/16 1:54 PM Max.” columns, respectively, and an arithmetic average of all the total grid block volume cell volume values in the entire grid is calculated and reported is of the correct sign, then under the “Total Average” column. this may not affect the abil- ity of the simulator to solve Cell Angle Areal. This property, along with the “Cell Angle the flow equations. It is Vertical,” is generated to investigate orthogonality issues that recommended that engineers may rise during simulation. The property represents the worst investigate the causes of any angle deviation from the ideal 90° angle in the areal plane. All inside out cells. Inside out eight angles in a grid cell block in the XY plane, Fig. 9, are

computed and compared to an ideal orthogonal angle, after cells usually occur because Fig. 9. The eight angle positions. which the deviation from 90° is reported. For example, both there is too much distortion Fig. 9. The eight angle positions. angles of 75° and 105° are reported as deviated by 15°, and in some of the input data or

the maximum deviation value is assigned for the grid block. A because one or more control lines are incorrectly classified as cell angle of 0° means that the cell is areally orthogonal. “I” lines when they should be “J”Property lines, and Name vice-versa. Threshold Description After reviewing the statistics, the engineer needs to defineMinimum cell volume allowed. Cell Volume Cell Angle Vertical. This property, along with the “Cell Angle threshold values for each property. Notice that these valuesAny values less will be reported Maximum cell angle deviation Areal,” is generated to investigate orthogonality issues that are treated differently as explainedCell in Angle Table Areal 2, which alsoallowed. pro- may rise during simulation. The property represents the worst vides recommended threshold values for best performanceAny valuesby greater will be reported. angle deviation from the ideal 90° angle in vertical planes. All the simulator. Maximum cell angle deviation 16 angles in a grid cell block in the XZ and YZ planes are Cell Angle Vertical allowed. computed and compared to an ideal orthogonal angle, after Property Name Threshold DescriptionAny values greater will be reported. which the deviation from 90° is reported. For example, both Cell dimension X direction to honor. angles of 75° and 105° are reported as deviated by 15°, and MinimumCell Dimension cell volume X allowed.Any values outside the range will be reported. the maximum deviation value is assigned for the grid block. A Cell Volume Cell dimension Y direction to honor. cell angle of 0° means that the cell is vertically orthogonal. Any valuesCell Dimension less will beY reportedAny values outside the range will be reported.

Maximum cell angle deviationMaximum cell volume contract Cell Dimension. This is the cell size in meters. Cell Volumeallowed. allowed. Contrast Cell Angle Areal Any values Less will be reported. Cell Volume Contrast. This property is generated to investigate Any values greater will be Maximum cell inside-out allowed. any throughput issue that might occur in the simulation. The Cell Insidereported. Out Any values greater will be reported. QC tool computes the ratio of the volume of a specific grid Maximum cell angle deviation block and all its neighboring cells’ volumes, and then reports Table 2.allowed. Cell’s threshold description the highest values. The ideal value for volume contrast is 1. Cell Angle Vertical Any values greater will be reported. Cell Inside Out. To measure the quality of the simulation grid block geometry, the tool uses a temporary fine grid of micro- Cell dimension X direction to honor. cells. Assuming that these microcells are defined by trilinear Cell Dimension X mapping, the QC tool calculates the Jacobian at the eight Any values outside the range will corners and at the center points of the microcells. It then uses be reported. those calculations to construct an M by M by M grid inside Cell dimension Y direction to each simulation grid block, M being a resolution parameter honor. Cell Dimension Y controlled by the QC tool. The total number of times that the Any valuesFig. 10. outside Well tops the check rangeed will against horizons. Jacobian is negative is then reported. When the grid is good, be reported. the reported result is zero. In most cases, the values are all, or Maximum cell volume contract almost all, zeros. Only when an inverted coordinate system Cell Volume allowed. has been used at the start will a larger number be the result. Contrast A grid is not good when the result is different in different Any values Less will be reported. grid blocks. There is no known rigorous test for an inside out hexahedral with bilinear surfaces1. Maximum cell inside-out allowed. Even if none of the cells is inside out, many of them can be Cell Inside Out much distorted. This is to be expected, depending on the input Any values greater will be data. The converse can also happen: a cell is flagged as inside reported. out but does not appear to be so. In these cases, the Jacobian still will have changed sign somewhere inside the grid block. If Table 2. Cell’s threshold description

74 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD10R1.indd 74 11/28/16 1:54 PM

Fig. 9. The eight angle positions.

Trajectories

Property Name Threshold Description Grids Minimum cell volume allowed. Cell Volume Any values less will be reported Maximum cell angle deviation

Cell Angle Areal allowed.

Any values greater will be reported. Maximum cell angle deviation Cell Angle Vertical allowed. Any values greater will be reported. Cell dimension X direction to honor. Cell Dimension X Any values outside the range will be reported. Cell dimension Y direction to honor. Fig. 11. QC step for the well trajectories. Cell Dimension Y Any values outside the range will be reported.

Maximum cell volume contract

Cell Volume allowed. Contrast Any values Less will be reported. Maximum cell inside-out allowed. Cell Inside Out Any values greater will be reported.

Table 2. Cell’s threshold description

Perforations

Fig. 10. Well tops checked against horizons. Fig. 10. Well tops checked against horizons.

Trajectories Grids

Fig. 11. QC step for the well trajectories. Completions Fig.Well 11.Tops QC step for the well trajectories.

In the fifth QC step, the tool ensures that the well trajecto- ries are located correctly by comparing them with the well

tops’ locations. This step is about checking the well trajectory Fig. 12. Well completions checked against perforations. locations by comparing the intersections of synthetic well perforation events. They should share the same depth zone, tops with the grid horizons against the current well tops that Fig. 12. Well completions checked against perforations. and the tool displays all the well completion events that do have been imported. They should intersect, and based on the not intersect with the perforation events in 3D, Fig. 12. threshold provided by the simulation engineer, the tool pro- vides statistical results showing how many well trajectories have distance more than the threshold. All the well trajectories Well Logs (Production) with issues can be displayed in 3D, Fig. 10. In the eighth QC step, the Well Trajectories tool checks if the well com- Perforations pletion events are located In the next two QC steps, multiple validations are executed correctly by comparing them at the same time. The first uses simulation perforations to against the well’s production validate well trajectories, assuming that the perforations are of logs. Production logs are im- high quality, and the second uses imported completion events ported and checked against that are cross-checked with the perforations for any inconsis- the completion event, and tencies. In this QC step, the tool checks if the well trajectories the tool reports any incon- are located correctly by comparing them against the assigned sistencies, like observed grids — i.e., the perforation events. They should intersect, and production in the production the tool displays all the trajectories that do not intersect with logs within intervals that the perforations in 3D, Fig. 11. are not perforated. The well completion events should Well Completion lower the production log depth zone. Based on the In the seventh QC step, the tool checks if the well completion threshold provided by the Fig. 13. Production log checked against Completions Fig. 13. Productionthe completion events.log checked against the completion events. events are located correctly by comparing them against the simulation engineer, the tool

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 75

Water Injection Rate Observed

Fig. 12. Well completions checked against perforations. 87448araD10R1.indd 75 11/28/16 1:54 PM

Rate Flow Liquid

Fig. 14. Observed data checked against the completion events.

Fig. 13. Production log checked against the completion events.

Water Injection Rate Observed

Rate Flow

Fig. 13. Production log checked against the completionLiquid events.

Water Injection Rate Observed

Rate Flow Liquid Liquid Flow Rate

Fig. 14. Observed data checked against the completion events.

lists and displays all the wells with issues, Fig. 13,Fig. showing 14. Observed decision data made checked in the against QC process. the Thiscompletion will enhance events. the com- production log vs. completion. munication and collaboration among simulation engineers by allowing them to save, access and review previous decisions. Observed Data The tool also provides automatic KPIs summarizing the simu-

lation model data quality, displayed in statistical and graphi-

In the ninth QC step, the tool validates the well completion cal forms, Fig. 15.

Fig.events 14. against Observed the observeddata checked data. against The observed the completion production events. data is imported and cross-checked against the completion data over CONCLUSIONS time; the purpose is to highlight any production that coincides

with a closed well, Fig. 14. The tool first checks the opening and This workflow contains nine QC steps that streamline all the closure of the wells for production against the observed data workflows that are currently being used by simulation engi- measurement, then it issues messages with the well’s name and neers, and the QC tool automates the processes applied in the

corresponding error, e.g., flowing wells without an open flow grid and well completion workflow. The tool also highlights section, completed wells without a flow rate, etc. all the simulation model data inconsistencies, and it issues a warning message and specifies proper actions needed to resolve REPORTING AND KPI the relevant errors. Even more, this tool enforces reservoir simulation best prac- In addition, this system provides a means of transferring tices by not allowing the user to go further without finishing the accumulated knowledge by capturing every activity and the current QC step successfully. It gives flexibility within this

Fig. 15. KPI report sample highlighting the statistics of the well tops and cell volumes. Fig. 15. KPI report samplesample highlightinghighlighting thethe statisticsstatistics ofof thethe well well tops tops and and cell cell volumes. volumes.

76 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD10R1.indd 76 11/28/16 1:54 PM stricture by letting the user activate the next QC step after pro- ACKNOWLEDGMENTS viding a proper business and technical justification. Tests and quality checks of the static model using this tool The authors would like to thank the management of Saudi before the reservoir simulation process will ensure the consis- Aramco for their support and permission to publish this tency and good quality of the reservoir model. It offers many article. advantages: This article was presented at the SPE Reservoir Characterization and Simulation Conference and Exhibition, • Eliminates the manual effort. Abu Dhabi, UAE, September 14-16, 2015. • Reduces the turnaround time. REFERENCES • Increases the accuracy.

• Reduces the history matching process. 1. Aziz, K. and Settari, A.: Petroleum Reservoir Simulation, • Enforces reservoir simulation best practices. Applied Science Publishers, London, U.K., 1979, 476 p. 2. Shirzai, A.F., Solonitsyn, S. and Kuvaev, I.: “Integrated Such an automated QC tool requires less work, which saves Geological and Engineering Uncertainty Analysis a lot of time during the history match phase and significantly Workflow, Lower Permian Carbonate Reservoir, improves the static models. It enhances the robustness of the Timan-Pechora Basin, Russia,” SPE paper 136322, production forecasts based on these models. It must be noted presented at the SPE Russian Oil and Gas Conference and as well that the proposed workflows can be applied in con- Exhibition, Moscow, Russia, October 26-28, 2010. junction with any geostatistical simulation method. Therefore, 3. Aljenaibi, F.S., Salameh, L.A., Recham, R., Albadi, B.S. the geological modeling process chosen by the geologist will and Meziani, S.: “Best Practice for Static and Dynamic always be preserved, as long as it is consistent with dynamic Modeling and Simulation History Match Case — Model data. Only the input parameters will be adjusted later to better QA/QC Criteria for Reliable Predictive Mode,” SPE paper account for constraints defined by the reservoir engineer. 148279, presented at the SPE Reservoir Characterization and Simulation Conference and Exhibition, Abu Dhabi, UAE, October 9-11, 2011. 4. Flint, S.S. and Bryant, I.D.: The Geological Modeling of Hydrocarbon Reservoirs and Outcrop Analogues, Blackwell Scientific Publications, Oxford, U.K., January 1993, 269 p.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 77

87448araD10R1.indd 77 11/28/16 1:54 PM BIOGRAPHIES

Tareq M. Al-Zahrani joined Saudi Aramco in October 2002. He is a Petroleum Engineer working in the Northern Area Reservoir Management Department. Tareq has 13 years of experience, mainly in reservoir engineering and reservoir management, and he has a proven track record of success within various organizations in Saudi Aramco. He has authored and coauthored many Society of Petroleum Engineers (SPE) papers. In 2002, Tareq received his B.S. degree in Petroleum Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.

Muath A. Al-Mulla is a Reservoir Simulation Engineer working in Saudi Aramco’s Reservoir Simulation & Description Department. He is currently working on constructing high level integrated reservoir modeling and studies, conducting reservoir and well performance data analysis, and providing technical support to the optimization of field development plans for the major operated oil fields in Saudi Arabia. He received his B.S. degree in Petroleum Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.

Mohammed S. Al-Nuaim joined Saudi Aramco in mid-2000 as a Chemical Engineer working in the Ras Tanura Refinery Projects Division. In 2003 he was moved to the ECC/PEASD Simulation Systems Division to work as a Petroleum Engineer and System Analyst for two years. Currently, Mohammed is leading the Business Analysis Group in the Intelligent Field and Producing Systems Division of the Southern Area Production Engineering Department. His 15 years of experience include working on several key projects related to reservoir simulation, i.e., assisting in the development of history matching and simulation model data accuracy systems. Mohammed has filed two patents with the U.S. Patent Office and has four more under processing. He received his B.S. degree in Chemical Engineering from King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia, and an M.S. degree in Petroleum Engineering from the University of New South Wales, Sydney, Australia.

78 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD10R1.indd 78 11/28/16 1:54 PM Case Study of Intelligent Completion with New Generation Electro-Hydraulic Downhole Control System

Authors: Suresh Jacob, Nibras A. Abdulbaqi, Chandresh Verma and Rabih Younes

ABSTRACT sensors are powered from the surface using small diameter control lines. An early generation of intelligent completion Extended reach wells with thousands of feet of open hole systems relied heavily on electronics, where multiple downhole reservoir contact and multilateral wells with eight to 10 laterals valves were operated through a single electric line from the are becoming common practice as the industry develops tight surface1. These systems used a downhole motor or a hydraulic and difficult reservoirs. Over the years, drilling and well pump to generate the motive force. The downhole electronics construction technologies have made significant advances in the were later simplified by replacing the motor and pumps field of geosteering, multilateral junctions, etc., to make these with simpler electro-hydraulic solenoid systems that uses complicated wells not just a reality but also commonplace. The high-pressure hydraulic fluid from the surface through control industry has recently introduced several new well completion lines. These hybrid systems had only solenoid valves downhole technologies for the downhole monitoring and inflow control to select the valve to be operated. Both the electronic and of these extended reach and multilateral wells. Over the last the hybrid electro-hydraulic valves had a very fast operation 10 years, Saudi Aramco has drilled and completed hundreds time and provided excellent feedback for the operation. of multilateral wells with intelligent completions for real-time Unfortunately, the electric cable, connectors, and electronics downhole monitoring and for remote valve operation to control used in these systems were highly susceptible to electrical inflow from each lateral. These technologies have improved failures due to water ingress. The low reliability of such well performance and reduced well intervention. electronics in harsh downhole conditions and the high cost of This article presents a case study examining the design, these systems forced the industry away from electrical systems. planning, installation, and operation of Saudi Aramco’s The industry therefore moved from electronically first intelligent completion to be operated with a combined operated systems to purely hydraulic systems. Valves were electro-hydraulic control system. New technologies like this operated with high-pressure hydraulic fluid supplied from system are expected to improve the operation and inflow from the surface through small diameter lines to the valves. The multi-zone wells by increasing the number of downhole inflow industry leveraged its experience with hydraulically operated control zones and faster operation of downhole valves. subsurface safety valves and mechanical sliding side doors Conventional intelligent completion systems use downhole to make highly reliable, hydraulically operated intelligent valves that are operated by hydraulic pressure. The pressure completion valves. In addition to their high reliability, is supplied from the surface through dedicated hydraulic lines these valves were less expensive than previous electronically running to each of the downhole valves. The wellhead and operated valves. Over time, hydraulically operated valves operations equipment constraints impose a maximum limit of became mainstream in intelligent completions. Currently, five downhole valves in a well. nearly 1,000 wells have been completed in the industry with The system in this case study uses an electro-hydraulic hydraulically operated intelligent completions. control module at each of the downhole valves. The system The hydraulic system, while being highly reliable and more can control up to 12 intelligent completion tools with the use cost-effective than the electronic system, had its own set of of only two hydraulic lines and one electrical line from the drawbacks, i.e., it took a longer time to operate each downhole surface. Technology like this allows control of inflow from each valve, and the system could operate only a limited number of of the laterals, enhancing the performance of the completion. downhole valves. Given the hydraulic system’s complicated downhole fluid metering and fluid directing systems, in some EVOLUTION OF INTELLIGENT COMPLETIONS cases it took hours to operate a single downhole valve. This TECHNOLOGY delay in valve operation created additional rig time for the installation of an intelligent completion. In regions with high rig Downhole inflow control valves (ICVs) and permanent sensors costs, the additional time negatively affected the economics of are critical elements of intelligent completions. The valves and those completions. The long operating time for the valves also

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 79

87448araD11R1.indd 79 11/28/16 2:21 PM One ICV for two laterals

One ICV for two laterals

Fig. 1. Schematic showing a well completion with a hydraulic intelligent completion — limited to a maximum of five downhole valves.

Fig. 1. Schematic showing a well completion with a hydraulic intelligent completion — limited to a maximum of five downhole valves. Fig. 1. Schematic showing a well completion with a hydraulic intelligent completion — limited to a maximum of five downhole valves. immediate feedback of valve positions. Recently the industry has leveraged the advances in high temperature electronics to develop new electronically operated intelligent completions that overcome the limitations of the previous electrical systems3, 4. The following sections of this article describe this new generation electro-hydraulic system that uses a single electric line and two hydraulic lines to operate up to 12 downhole ICVs. The single electric line and two hydraulic lines are connected to all the downhole valves. The two hydraulic lines are connected, one each, to the open and closed sides of the downhole ICV, while the electric line is connected to a solenoid valve in the ICV. Each of the solenoid valves has a different address, and each can be electrically activated from the surface independently from the others. Activating a particular solenoid valve allows the high-pressure hydraulic pressure in the control lines to operate the ICV for the zone, without affecting the other connected ICVs. Fig. 2. Schematic of the electro-hydraulic intelligent completion showing Fig.Valve 2. B operation.Schematic of the electro-hydraulic intelligent completion showingIn summary, Valve B operation.the electric system is used to select the specific valve to be operated, and the hydraulic system provides the

motive force to operate the valve. Figure 2 shows the valve reduced the frequency of their operation during production. The hydraulic intelligent completion system required a operation using the electro-hydraulic system. Valve B is being dedicated hydraulic line for each downhole valve. The maximum operated. After the solenoid valve is energized for that zone, number of valves in a well was five, due to the operation high-pressure fluid from the surface in open line 2 (purple), constraints of multiple lines and the limited number of ports on flows through the solenoid valve in Unit B to Valve B. This the wellhead and packers. This created a situation where a well high-pressure fluid operates Valve B, and the fluid from the would be drilled with eight laterals but completed with only backside of the piston on this valve (pink) returns to the Fig.five downhole 2. Schematic valves, Fig. of 1.the This electro-hydraulic limitation forced adjacent intelligent surface completion through lineshowing 1. Valves Valve A and BC operation.are not operated laterals to be commingled through one valve2. Premature gas during this process as the solenoid valves in Unit A and Unit

or water breakthrough on one lateral would require its valve C are not activated and the ICVs are hydraulically isolated..

to be closed, thereby reducing inflow from the adjacent lateral, The solenoid is activated and kept open during the which is produced through the same valve. New technologies duration of the ICV operation, it is de-activated and closed were developed to increase the number of downhole valves when the ICV operation is complete. The downhole valves while maintaining the same number of downhole cables. may be closed, opened fully or operated to an intermediate choke setting by controlling the time period the solenoid is NEW GENERATION ELECTRO-HYDRAULIC kept open. INTELLIGENT COMPLETIONS Figure 3 shows a comparison of operating times for the hydraulic and electro-hydraulic downhole valves. The As stated earlier, the electronically operated intelligent electro-hydraulic valves’ operation is much faster. completion offers several benefits, like faster operation and The solenoid valve unit used in the system is physically

80 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD11R1.indd 80 11/28/16 2:21 PM Electro-hydraulic module

ICV

Electro-hydraulic module (enlarged)

Fig. 3. The electro-hydraulic control module and ICV as separate components.

Zone 1 Zone 2 Electro-hydraulic Zone 1 Actuation Electro-hydraulic Zone 2 Actuation

Fig. 3. Comparison of operating times for hydraulic valves (left) and electro-hydraulic valves (right). Fig. 4. Comparison of operating times for hydraulic valves (left) and electro-hydraulic valves (right).

Electro-hydraulicElectro-hydraulic module module

Actuation ICV

Electro-hydraulicElectro-hydraulic modulemodule (enlarged)(enlarged)

Fig. 4. The electro-hydraulic control module and hydraulically operated ICV as separate components. Fig. 3. The electro-hydraulic control module and ICV as separate components. separate from the valve, with only hydraulic connections WELL CONSTRUCTION AND COMPLETION DESIGN

between the solenoid module and the hydraulically operated

valves, Fig. 4. This physical separation of the valve from the An existing single lateral well with a 7” casing and a 4½”

solenoid control module allowed the continued use of an liner was converted to a multilateral well to test the electro- industry proven hydraulic ICV. The only new equipment in hydraulic intelligent completion. Figure 5 shows the initial Zone 1 Zone 2 Electro-hydraulic Zone 1 Actuation the system is solenoid valve module. All the other components well condition and the proposed multilateral. The workover, are taken from existing, field proven technologies. The system drilling and completion sequenceElectro included-hydraulic retrieving Zone the 2 Actuation uses a fluid filled electric cable with positive pressure to reduce existing tubing, milling the upper section of the 4½” liner to risk of water ingress into the electric chamber and subsequent reach deep enough within the 7” liner to cut a new window, electrical failures. drilling a 6⅛” open hole, installing a 5½” expandable liner,

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 81

87448araD11R1.indd 81 11/28/16 2:21 PM

Fig. 4. Comparison of operating times for hydraulic valves (left) and electro-hydraulic valves (right).

drilling the TAML classified level-2 laterals from the 5½” of the electro-hydraulic valves and the reduced number of expandable liner and installing the intelligent completion. control lines. After installation, the valves were tested to The number of laterals was reduced from nine to six due confirm their proper operation. The electro-hydraulic system to operational requirements. Figure 6 shows the intelligent enabled the use of an individual downhole valve for each completion installed. lateral, so the flow from two or more laterals did not need A downhole valve with 10 discrete choke settings was used to be commingled. In addition to the downhole valves, a in each of the zones. Figure 7 shows a zoomed-in view of downhole fiber optic cable was also installed for inflow the choke trim section of the ICV, fully open position (left), monitoring using distributed acoustic sensing. closed (center) and position 3 (right). PRODUCTION OPERATIONS COMPLETION INSTALLATION Production from the well commenced three months after The six-zone intelligent completion with downhole valves, installation of the electro-hydraulic system. First, each of the packers, sensors, etc., was successfully installed without any laterals was flowed back individually to verify operation of downtime. The time to deploy the electro-hydraulic valve the downhole valves and flow back of the completion brine system was much shorter than the deployment time for and removal of any drilling mud. Once a lateral was cleaned normal hydraulic valves. This was due to the faster operation

Fig. 5. Detailed drawing of the single lateral well before workover (left), and the planned well trajectory and laterals (right).

Fig.Fig. 7. 7.Detailed Detailed views views ofof thethe intelligent completion completion valve valve showing showing the fully the fullyopen open position (left), the fully closed position (center) and the valve set to position 3 (right). Fig.Fig. 5. 5. Detailed Detailed drawing drawing of the of single the singlelateral welllateral before well workover before (left), workover and the (left), planned and well the trajectory position (left), the fully closed position (center) and the valve set to position 3 andplanned laterals well (right). trajectory and laterals (right). (right).

Tubing Hanger

Tubing Hanger

Nipple 5 Nipple

PDHMS

PDHMS PDHMS 25 SSD

Packer 6 2 PDHMS Electro-hydraulic ICV 6 5 SSD Dual PDHMS 7” 26# Casing ShoePacker 6 Packer 5 Packer 4 Packer 3 Packer 2 Packer 1 5 5½” Liner Shoe Electro-hydraulic ICV 6 L5 L L L2 L L

DTS TAS and Bullnose

Electro-hydraulic ICV 5 Electro-hydraulic ICV 4 Electro-hydraulic ICV 3 Electro-hydraulic ICV 2 Electro-hydraulic ICV 1 Dual PDHMS Fig. 8. Percentage inflow contribution from each lateral with all the downhole valves in a fully open

position. Fig. 6. Schematic of7” the 26# intelligent Casing completion Shoe with the electro-hydraulic controls installed in six lateral wells. Packer 4 Packer 5 Packer 3 Packer 2 Packer 1

5½” Liner Shoe

DTS TAS and Bullnose

Electro-hydraulic ICV 5 Electro-hydraulic ICV 4 Electro-hydraulic ICV 3 Electro-hydraulic ICV 2 Electro-hydraulic ICV 1

Fig. 6. Schematic of the intelligent completion with the electro-hydraulic controls installed in six lateral wells.

Fig.82 6. FALL Schematic 2016 SAUDI ARAMCO of the JOURNAL intelligent OF TECHNOLOGY completion with the electro-hydraulic controls installed in six lateral wells.

87448araD11R1.indd 82 11/28/16 2:21 PM

35

30

25

20

15

10

5

Percentage Inflow from Each Lateral Each from Inflow Percentage 0 L-5 (Pos-10) L-4 (Pos-10) L-3 (Pos-10) L-2 (Pos-10) L-1 (Pos-10) L-0 (Pos-10) Downhole Valve Choke Setting for Each Lateral (a)

35 18

16 30 14 25 12

20 10

15 8 6 10 4 5 2 Percentage Inflow from Each Lateral Each from Inflow Percentage Percentage Inflow from Each Lateral Each from Inflow Percentage 0 0 L-5 (Pos-10) L-4 (Pos-10) L-3 (Pos-10) L-2 (Pos-10) L-1 (Pos-10) L-0 (Pos-10) L-5 (Pos-7) L-4 (Pos-4) L-3 (Pos-8) L-2 (Pos-8) L-1 (Pos-9) L-0 (Pos-10)

Downhole Valve Choke Setting for Each Lateral (a) Downhole Valve Choke Setting for Each Lateral (b)

Fig. 8(a). Percentage inflow contribution from each lateral with all the downhole valves in a fully open position, and (b) Balanced inflow contribution with downhole

valves in18 different choke settings. 16 Saudi Aramco: Company General Use up and14 had started producing dry oil, it was extensively flow ACKNOWLEDGMENTS

tested 12to measure reservoir pressure, gas-oil ratio, water cut

and to10 estimate productivity. The authors wish to thank the management of Saudi Aramco 8 and Halliburton for their support and permission to publish WELL PERFORMANCE6 MODELING this article. The authors also acknowledge the valuable 4 support of Manesh Mathew, Ameen Al-Zubail, Elias Garcia The data2 collected from the lateral flow tests was used to and Savio Saldanha at Halliburton for their contribution to Percentage Inflow from Each Lateral Each from Inflow Percentage simulate0 well and lateral inflow performance at different well L-5 (Pos-7) L-4 (Pos-4) L-3 (Pos-8) L-2 (Pos-8) L-1 (Pos-9) L-0 (Pos-10) the successful installation. conditions. The simulations show that due to variations in Downhole Valve Choke Setting for Each Lateral (b) This article was presented at the SPE Saudi Arabia Section productivity and completions geometry, some laterals will Annual Technical Symposium and Exhibition, al-Khobar, dominate inflow while others will have low contributions. Saudi Arabia, April 25-28, 2016. The results shown in Fig. 8a indicated that laterals 5 and 3 dominated the inflow when the well was produced with all the Saudi Aramco: Company General Use REFERENCES downhole valves in the fully open position. This dominance prevents weaker laterals, like the mainbore and lateral 4, from meeting the target rates. Long-term production in this 1. Shaw, J.: “Comparison of Downhole Control System condition can cause premature water or gas breakthrough in Technologies for Intelligent Completions,” SPE paper the laterals with the higher contribution, which can ultimately 147547, presented at the Canadian Unconventional affect the life of the well. Resources Conference, Calgary, Alberta, Canada, The production objective of balancing inflow from all the November 15-17, 2011. laterals is achieved by reducing the rate from the stronger laterals 2. Jacob, S., Bellaci, I.J., Nazarenko, P. and Joseph, P.: and increasing the rate from the weaker laterals. The rate from “Designing, Planning and Installation of an 8-Zone the stronger laterals can be reduced by operating the downhole All-Electric Intelligent Completion System in an Extreme choke from the fully open position — position 10 — to a smaller Reservoir Contact Well,” SPE paper 176811, presented at position. The rate from the weaker laterals can be increased the Middle East Intelligent Oil and Gas Conference and by keeping the ICV fully open and increasing the pressure Exhibition, Abu Dhabi, UAE, September 15-16, 2015. drawdown of these laterals. Figure 8b shows the balanced inflow 3. Garcia, E. and Saldanha, S.: “Electrohydraulic ICV Control from all laterals by following the above process. System: A Novel Approach to Multizonal Control,” OTC paper 26816, presented at the Offshore Technology CONCLUSIONS Conference Asia, Kuala Lumpur, Malaysia, March 22-25, 2016. Saudi Aramco’s first electro-hydraulic intelligent completion was successfully installed in a six-lateral multilateral well. 4. Garcia, E. and Mathew, M.: “Delivering Selective Interval The technology performed as designed during installation and Control: Electro-Hydraulic Intelligent Completion subsequent production. The electro-hydraulic system allows Enhances Reservoir Management for Multilateral Well,” the number of downhole zones in a well to be increased, and SPE paper 178168, presented at the SPE/IADC Middle it improves efficiency by reducing the time required to operate East Drilling Technology Conference and Exhibition, Abu the valves. Dhabi, UAE, January 26-28, 2016.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 83

87448araD11R1.indd 83 11/28/16 2:21 PM BIOGRAPHIES Rabih Younes is currently working as the Acting Production Engineering Suresh Jacob is a Petroleum Unit Supervisor in Saudi Aramco’s Engineering Specialist in Saudi Northern Area Production Aramco. He works primarily on Engineering Department. His advanced well completions, i.e., experience includes working in both intelligent completions, permanent the conventional and the downhole gauges, inflow control unconventional gas production sectors of the industry. devices and integrated solutions. Rabih has extensive experience in the areas of well testing, Suresh has led or participated in the design, field well modeling and production allocation and forecasting. installation and production operation for several new well He has over 10 years of industry experience in service and completion technologies. He has over 20 years of industry operating companies, including work in Australia and the experience in service and operating companies, including Middle East. work in North America, the Middle East and the Asia In 2003, Rabih received his B.S. degree in Geological Pacific. Engineering from the Royal Melbourne Institute of Suresh received his B.S. degree in Mechanical Technology, Melbourne, Victoria, Australia. In 2004, he Engineering from the University of Kerala, Kerala, received his M.S. degree in Petroleum Engineering from India, and his M.S. degree in Petroleum Engineering from Curtin University, Perth, Western Australia. Texas A&M University, College Station, TX. Suresh is the recipient of the 2015 Society of Petroleum Engineers (SPE) Middle East Regional Completions Optimization and Technology Award for his contribution to technical expertise in well completions.

Nibras A. Abdulbaqi is a Petroleum Engineer working for Saudi Aramco’s Reservoir Management Department in Shaybah, as part of the team under the Manifa Reservoir Management Division. In 2011, she received her B.S. degree in Petroleum Engineering from the University of Oklahoma, Norman, OK, and in 2014, Nibras received her M.S. degree in Petroleum Engineering, with a concentration in Smart Oil Field Technologies, from the University of Southern California, Los Angeles, CA. While at USC, she worked on conventional and unconventional reservoirs for reservoir performance forecasting and prediction.

Chandresh Verma is a Supervisor with Saudi Aramco’s Drilling and Workover Engineering Department. He joined Saudi Aramco in 2006 after working at India’s premier E&P company, Oil & Natural Gas Corporation, from 1984-2006. Chandresh has 32+ years of oil field experience in onshore and offshore drilling operations as well as in drilling engineering. In 1982, he received his B.S. degree in Mechanical Engineering from the University of Indore, Madhya Pradesh, India.

84 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD11R1.indd 84 11/28/16 2:21 PM SUBSCRIPTION ORDER FORM

To begin receiving the Saudi Aramco Journal of Technology at no charge, please complete this form.

Please print clearly.

Name ______

Title ______

Organization ______

Address ______

City ______

State/Province ______

Postal code ______

Country ______

E-mail address ______

Number of copies ______

TO ORDER

By phone/email:

Saudi Aramco Public Relations Department JOT Distribution +966-013-876-0498 [email protected]

By mail:

Saudi Aramco Public Relations Department JOT Distribution Box 5000 Dhahran 31311 Saudi Arabia

Current issues, select back issues and multiple copies of some issues are available upon request.

The Saudi Aramco Journal of Technology is published by the Saudi Aramco Public Relations Department, Saudi Arabian Oil Company, Dhahran, Saudi Arabia.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY FALL 2016 85

87448araD12R1.indd 85 11/28/16 2:23 PM GUIDELINES FOR SUBMITTING AN ARTICLE Acknowledgments TO THE SAUDI ARAMCO JOURNAL OF TECHNOLOGY Use to thank those who helped make the article possible. These guidelines are designed to simplify and help standardize Illustrations/tables/photos and explanatory text submissions. They need not be followed rigorously. If you have additional questions, please feel free to contact us at Submit these separately. Do not place in the text. Positioning Public Relations. Our address and phone numbers are listed in the text may be indicated with placeholders. Initial on page 75. submission may include copies of originals; however, publication will require the originals. When possible, submit Length both electronic versions, printouts and/or slides. Color is Varies, but an average of 2,500-3,500 words, plus preferable. illustrations/photos and captions. Maximum length should File formats be 5,000 words. Articles in excess will be shortened. Illustration files with .EPS extensions work best. Other What to send acceptable extensions are .TIFF, .JPEG and .PICT. Send text in Microsoft Word format via email or on disc, Permission(s) to reprint, if appropriate plus one hard copy. Send illustrations/photos and captions separately but concurrently, both as email or as hard copy Previously published articles are acceptable but can be (more information follows under file formats). published only with written permission from the copyright holder. Procedure Author(s)/contributor(s) Notification of acceptance is usually within three weeks after the submission deadline. The article will be edited for style Please include a brief biographical statement. and clarity and returned to the author for review. All articles Submission/Acceptance Procedures are subject to the company’s normal review. No paper can be published without a signature at the manager level or above. Papers are submitted on a competitive basis and are evaluated by an editorial review board comprised of various department Format managers and subject matter experts. Following initial No single article need include all of the following parts. The selection, authors whose papers have been accepted for type of article and subject covered will determine which parts publication will be notified by email. to include. Papers submitted for a particular issue but not accepted for that issue will be carried forward as submissions for Working title subsequent issues, unless the author specifically requests Abstract in writing that there be no further consideration. Papers previously published or presented may be submitted. Usually 100-150 words to summarize the main points. Submit articles to: Introduction Editor Different from the abstract in that it “sets the stage” for the The Saudi Aramco Journal of Technology content of the article, rather than telling the reader what it C-11B, Room AN-1080 is about. North Admin Building #175 Dhahran 31311, Saudi Arabia Main body Tel: +966-013-876-0498 E-mail: [email protected] May incorporate subtitles, artwork, photos, etc. Submission deadlines Conclusion/summary

Assessment of results or restatement of points in introduction. Issue Paper submission deadline Release date Endnotes/references/bibliography Spring 2017 November 30, 2016 March 31, 2017 Summer 2017 February 28, 2017 June 30, 2017 Use only when essential. Use author/date citation method Fall 2017 May 31, 2017 September 30, 2017 Winter 2017 August 30, 2017 December 31, 2017 in the main body. Numbered footnotes or endnotes will be converted. Include complete publication information. Standard is The Associated Press Stylebook, 51st ed. and Webster’s New World College Dictionary, 5th ed.

86 FALL 2016 SAUDI ARAMCO JOURNAL OF TECHNOLOGY

87448araD12R1.indd 86 11/28/16 2:23 PM Additional Content Available Online at: www.saudiaramco.com/jot

An Experimental Investigation of Performance Evaluations for Seawater and CO2 Injection Using Dual Core Methodology at Reservoir Conditions Xianmin Zhou, Fawaz M. Al-Otaibi, AlMohannad A. Al-Hashboul and Dr. Sunil L. Kokal ABSTRACT

In this study, a novel methodology of dual coreflooding testing was first developed and described in detail, and then applied to evaluate the displacement efficiency and performance of seawater and supercritical carbon dioxide (sc-CO2) injection in tests of two carbonate core plugs with different permeabilities at reservoir conditions. The dual coreflooding apparatus, which consists of seven components, can be used for the individual or simultaneous injection of one- or two-phase fluids into core plugs of various length arranged in a single or dual coreflooding configuration.

Chemostratigraphic Approach: A Tool to Unravel the Stratigraphy of the Permo-Carboniferous Unayzah Group and Basal Khuff Clastics Member, Central Saudi Arabia Dr. Mohamed Soua ABSTRACT

The Permo-Carboniferous Unayzah Group is generally lacking in high resolution biostratigraphic control and fails to produce a stratigraphic correlation using lithostratigraphy, due to its large similarities with sandstones encountered in both the Devonian and the Silurian sections. This study strives to propose correlation schemes for the Permo-Carboniferous Unayzah Group in central Saudi Arabia and to define the Unayzah Group and basal Khuff clastics (BKC) boundaries based on chemostratigraphic analysis.

Development and Evaluation of Gel-based Conformance Control for a High Salinity and High Temperature Carbonate Dr. Jinxun Wang, Dr. Abdulkarim M. Al-Sofi and Abdullah M. Al-Boqmi ABSTRACT

The extreme heterogeneity of carbonate reservoirs, featuring fracture corridors and super-permeability thief zones, challenges the efficient sweep of oil in both secondary and tertiary recovery operations. In such reservoirs, conformance control is crucial to ensure injected water and any enhanced oil recovery (EOR) chemicals contact the remaining oil in an optimal manner with minimal throughput. Gel-based conformance control has been successfully applied in both sandstone and carbonate reservoirs. In-depth conformance control in high temperature reservoirs is still a challenge, due to severe gel syneresis and the associated significant reduction in gelation time.

Scale Deposition and Removal in Khuff Sour Gas Wells Dr. Qiwei Wang, Dr. Syed R. Zaidi, Jairo A. Leal Jauregui, Irfan Syafii, Amro E. Mukhles, Dr. Tao Chen, Dr. Fakuen F. Chang and Mauricio A. Espinosa

ABSTRACT

Scale formation has been a persistent challenge in many producing sour gas wells from the Khuff reservoir in Saudi Arabia. Accumulation of scale deposits on downhole tubulars and in wellhead manifolds interferes with field operation, limits well accessibility and decreases well productivity. Extensive efforts have been devoted to understanding the scale deposition process and to developing a cost-effective mitigation strategy. This article discusses the up-to-date knowledge regarding scale formation in these prolific gas wells and presents the descaling technologies deployed in them.

11/28/16 3:04 PM