AADE-05-NTCE-51.

Well Control Simulation A Tool for Engineering and Operations

Fred Ng, Wild Well Control, Inc.

This paper was prepared for presentation at the AADE 2005 National Technical Conference and Exhibition, held at the Wyndam Greenspoint in Houston, Texas, April 5-7, 2005. This conference was sponsored by the Houston Chapter of the American Association of Drilling Engineers. The information presented in this paper does not reflect any position, claim or endorsement made or implied by the American Association of Drilling Engineers, their officers or members. Questions concerning the content of this paper should be directed to the individuals listed as author/s of this work.

Abstract
Well Control Simulation is becoming a very useful tool for managing risks and making critical well control decisions. The stakes are especially high in expensive and increasingly popular projects such as ultra deep water or deep gas drilling. In well planning, it is used for evaluating well design by modeling and developing circulating kick tolerance, based on casing shoe fracture gradient, gas handling capacity of surface equipment, or other specified limits. In operations, it is used to support well control decisions with updated kick tolerance when changes occur in shoe fracture gradient, mud weight range, or size and depth of hole and casing. For special well control operations beyond the conventional Drillers or Wait and Weight procedures, it can be used to evaluate options such as interrupting the kill to change mud weight, pump rate and other conditions during a kill; as well as extended shut in and volumetric kill. The modeling is based on complete transient two phase flow, and results are displayed in simple graphics that are easy to understand and use in the field.

Besides the standard drillers method or wait and weight circulations, a transient simulator of this kind allows changes in operating conditions to be made at any point during the well control process. So it can handle special procedures such as extended shut in of a kick, changing mud weight, pump rate, taking additional influx or other interruptions during a kill. Other useful features made possible by this kind of simulation include: Two Phase versus Single Bubble Modeling A single bubble model assumes that the kick influx occurs as a single phase and remains so as it is circulated up the wellbore. This tends to produce conservative results, which is usually preferable. However, for deep and or high pressure wells, the results can sometimes be too conservative and not realistic, making it almost impossible to design a well. A multiphase model provides results that are more realistic, which is especially important for deep and high pressured wells.. Choke Line Friction Pressure (CLFP) limit For drilling in deep water, say over 1,000, CLFP in a long choke line through the water column can sometimes be high enough to effectively over ride the effect of the adjustable choke, as often happens when a gas kick is circulated close to the surface. In order to maintain the proper pressures, it will then be necessary to use an additional choke or kill line, slow down the kill pumping rate, stop the kill to bleed down the pressure, or take all three of these actions. A transient simulation allows the detection of such occurrences. The simulation can be paused to execute one or more of the above responses, and continuation of the simulation will then enable the effects of these actions to be evaluated. Separator and Flare Line Limits The simulation includes effects of pressure drop in a gas flaring system, including the separator, flare line and mud leg. If the gas flaring rate exceeds system capacity, the simulation will indicate that the mud leg has been blown. Usual response options are slowing the pump rate, shutting down the pump to bleed gas, or by passing the separator. The simulation can be

Introduction
Recent advances in computer technology have made well control simulators available as tools for drilling engineering and operations. This paper describes the application of one such simulator for well planning, drilling operations as well as for evaluating special well control operations.

The Simulator
A fully time transient, two phase flow simulator provides a number of technology improvement over conventional steady state models. It can simulate and track the actual full behavior of a kick, starting with influx of the kick, flow check, shut in, the subsequent kill and venting of gas on the surface. Instead of just seeing the results, as in the case of a steady state model, the entire kick and kill process can be visualized.

WELL CONTROL SIMULATION A TOOL FOR ENGINEERING AND OPERATIONS

paused to implement one or more of these actions, and continuation of the simulation will allow evaluating the effects of such actions. The simulator used for modeling in this paper is the Kick program from the Drillbench suite of drilling simulation software. It was developed by Norways Rogaland Research Institute and licensed through Scandpower Petroleum Technology. It runs on Windows desk top and lap top computers.

Applications
This paper describes the application of such simulations to the following phases of a drilling project: Well Planning The simulations can be used to evaluate well design by modeling and developing circulating kick tolerance, based on both casing shoe fracture gradient as well as gas handling capacity of surface equipment. Results for each hole section can be summarized in simple charts for use at the office or in the field. Drilling Operations Support well control decisions with updated kick tolerance charts and simulations when changes occur in shoe fracture gradient, mud weight range, casing or hole sizes and depths. Special Well Control Operations Modeling and evaluation of alternatives either for contingency planning or in response to an actual well control situation. Examples of such options include changing mud weight, pump rate and other conditions during a kill; changing choke, lines and gas separator sizes; extended shut in before or during a kill, and volumetric kill.

Overall results for the well show that, for kick volumes of up to 85 bbl in size, kick tolerance ranges from 0.2 to 2.5 ppg in magnitude. Choke pressure seen in these containable kicks ranged up to 3200 pounds per square inch (psi), and surface gas rate ranged up to 3.9 million cubic feet per day (mmcfd) at kill rates of up to 300 gallons per minute (gpm), with a pit volume gas expansion of up to 85 barrels ( bbl). Casing pressure increased up to 3000 psi above mud column hydrostatic while circulating out containable kicks with a mud weight of 12.7 ppg in the 12-1/4x14-3/4 interval. It should be noted that these maximum values result from containable kicks within the volume range investigated. Larger volume kicks and kicks that break down the casing shoes can result in substantially higher pressures at the surface and in the wellbore, larger expansion pit volumes and higher surface gas rates. These can only be evaluated with additional modeling on a case by case basis. Results Simulations were run for the maximum planned depth of each hole interval. Drillers Method is used to circulate the kicks in order to provide conservative results. The four intervals involved were: 16x20 hole drilled below 18 casing from 11200 to 15800'. 14-3/4x17-1/2 hole drilled below 16 liner from 15800' to 20800' 12-1/4x14-3/4 hole drilled below 13-5/8 casing from 20800' to 21800'MD. 10-5/8 hole drilled below 11-7/8 liner from 21800' to 27000'.

Well Planning
Well Plan Summary This is a typical summary of well control evaluation for a well plan. The project involves a 27,000 well to be drilled in a water depth of some 7000. Simulations were performed to evaluate circulating kick tolerances based on formation fracture strength at the casing shoe for critical well sections below 11,180 MD (measured depth). Although Formation Integrity Test (FIT) is used in this case, predicted Fracture Gradient (FG) or Leak Off Test (LOT) value can also be used. The four sections consists of the 16x20, 14-3/4x17-1/2, 12-1/4x14-3/4, and 10-5/8 holes, all of which are to be drilled with synthetic based mud (SBM). Based on FIT values, two hole sections are identified as having kick tolerances below the common practice minimum of 0.5 pound per gallon (ppg) while drilling with maximum mud weight at total depth (TD) of the section. These are the 16x20 and 14-3/4x17-1/2 holes, where minimum shut in kick tolerances are 0.22 and 0.31 ppg respectively.

Figure 1 shows the wellbore schematic, as well as locations of the observations points where wellbore pressure is displayed. Up to five observation points can be designated for each simulation, which are typically chosen to be at the mud line, tops of cement, tops of liners or other points of interest. Results from the simulations are shown for all the four hole sections in the summary display of Circulating Kick Tolerance plots in Figure 2, while details for a typical interval are shown in a separate plot in Figure 3. Following are high lights from these results: Kick tolerance These are simulated for each hole section to evaluate the wells ability to contain the shut in and circulation of kicks without exceeding formation fracture pressure at the respective casing shoe of the interval. Results are plotted in Figures 2 and 3, where kick magnitude and volume combinations below a kick tolerance curve for the given mud weight can be shut in and circulated out without exceeding formation fracture pressure. Those combinations above the curve will result in exceeding fracture pressure at some point in

AADE-05-NTCE-51

the kill process. Each plot involves the minimum and maximum mud weights planned for the hole interval. The data points on these curves are also labeled with the equivalent bottom hole pore pressure for each point. Note that, for any given hole section, while kick tolerance is always higher for a lighter mud weight, the equivalent pore pressure of the corresponding containable kick is lower or equal to that for the higher mud weight. In other word, a gain in kick tolerance from using a lower mud weight actually results in a lower, or at best the same, pore pressure that can be handled. Kick volume range for these simulations have been limited to about 80 bbl, since this should cover the range of kicks normally expected when drilling operations are conducted by reasonably well trained personnel. The table below is a summary of the range of containable kicks as determined from the simulations: Maximum expected pressure, pit volume expansion and surface gas rate These result from kick combinations that are represented by the red stars in the kick tolerance plots, and their values are also labeled in the plots. Time Histories and Press at observation points Figures 4 contain a typical time history plot of a red star run referred to above. Besides wellbore pressure at the observation points, other parameters displayed in these plots include casing and drill pipe pressures, pit volumes and surface gas rates. It also shows pressure at the casing shoe and how it compares with formation fracture (FIT) pressure. These figures represent typical time history illustrations of simulated well kicks from time of influx to gas at surface. Casing Pressure at Observation Points Figure 4 also contains time history plots that show maximum wellbore pressure at the designated observation points. These observation points are typically chosen at the mud line, tops of cement, tops of liners or other points of interest, where such pressure information can be critical in the evaluation of casing burst design. Separator and flare lines The simulator provides a warning whenever the rate of gas and mud discharge results in a back pressure that would blow the specified mud leg column height in the separator system. Specifications used for the separator system in this report are based on typical values for a deep water offshore drilling operation. The system appears to be more than adequate to handle the gas kicks at the simulated kill rates of up to 300 gpm, as the warning was not tripped by any of the simulation runs.

Drilling Operations
The kick tolerances above are applicable only to the specified mud weight and well scenarios. If there are changes in hole or casing sizes, fracture gradient, FIT results, well depth, shoe setting depth or mud weight, then additional model runs will be needed to develop new or at least updated sets of kick tolerance curves. Note that the model indicates a shoe failure even if the well bore pressure is just a few psi above the fracture pressure. Since actual fracture pressures are not usually known to this degree of accuracy, this should be taken into consideration when interpreting the plot. If the well encounters a kick combination that is close to the curve, a model run can be made based on specific and updated parameters. Output from such a run would be similar to those in Figure 4, and would provide a more accurate evaluation of well control conditions. Once the model has been set up in the well planning phase, any of the above updates can be provided in a matter of minutes. This near real time information can be very helpful for managing a drilling operation, especially for exploratory wells which often involve significant uncertainties in such well parameters.

Special Well Control Operations

Beyond the conventional Drillers Method and Wait and Weight Method for circulating out a kick, these simulations can also be used to evaluate a number of special well control operations, as described in the examples below, where water based mud (WBM) is used in order to show the effects of rapid gas migration:

Extended shut in This demonstrates effects of rapid migration up the hole for a gas kick in water based mud. As shown in Figure 5, after shutting in for 4 hours without bleeding the annulus, the gas has migrated to 7000 (from an initial 13000), and wellhead pressure rose from the initial shut in value of 2000 psi to 4000 psi. Pressure at the shoe, which was initially 200 psi below fracture, exceeded it after a shut in of only 18 minutes. Volumetric Kill In this run, the same kick was kept shut in. As shown in Figure 6, this time the gas was allowed to migrate, and was brought to the surface using the volumetric kill method, in which annulus pressure was bled off to maintain constant drill pipe pressure. Pressure at the casing shoe was kept below fracture pressure during the kill process.

www.wildwell.com

Tel 281.784.4700 USA

WELL CONTROL SIMULATION A TOOL FOR ENGINEERING AND OPERATIONS

Examples in Risk Management

Following are two recent cases in which application of well control simulation had a significant impact on risk mitigation decisions: Casing burst design A 13000 intermediate string of 13-5/8 casing was originally planned for a 20,000 well. Burst design for the casing was evaluated based on simulated pressures at the observation points similar to those shown in Figure 4. Results indicated that the burst design was too marginal to meet the companys design criteria when casing wear was taken into account. The design was changed to a string of 14 casing to provide thicker wall pipe for drilling the intervals below. Insurance premium A number of underwriters for well control insurance are now requiring some of their clients to obtain well control evaluation using this type of simulation. This helps to lower the risk exposure for the underwriters, who often in turn reduce the insurance premiums for the operators that comply with this evaluation. In a recent case, one company was reported to have saved over $120,000 in well control insurance premiums in the course of drilling five wells.

Conclusions
Much of petroleum exploration and production is about risk management, and well control is an ever present and potentially costly issue. Well control simulation is a useful tool for helping to manage these risks. When applied to well planning, it serves to identify potential well control problems and mitigate these risks by developing and optimizing well design. It can also be applied to evaluate special well control operations and thus provide contingency provisions in the event that such problems do occur.

Acknowledgements
The Author wishes to thank AADE, Wild Well Control, Inc. and Scandpower for their co-operation and for the opportunity to present this paper to the drilling community.

Casing Program Diameter (in)

0

2,000

4,000

6,000

8,000

10,000

AADE-05-NTCE-51

12,000

Measured depth (ft)

14,000 16,000 18,000 20,000 22,000 24,000

www.wildwell.com
26,000

Tel 281.784.4700 USA

Figure 1 Well bore schematic and observation points

14 x 17 hole at 20800 with 16 liner at 8500-15800, 13.8 FIT, 300 gpm kill rate

16 x 20 hole at 15800 with 18 shoe at 11200, 11.8 FIT, 300 gpm kill rate

10 5/8 hole at 27000 with 11 7/8 liner at 20500-21800, 16.6 FIT, 200 gpm kill rate

WELL CONTROL SIMULATION A TOOL FOR ENGINEERING AND OPERATIONS

12 x 14 : hole at 21800 with 13 5/8 shoe at 20800, 15.3 FIT, 250 gpm kill rate

Figure 2 Summary of circulating kick tolerance plots.

AADE-05-NTCE-51

Figure 3 Typical circulating kick tolerance plot for a hole section

Figure 4 Typical time history graphic output, 80 bbl x 0.6 ppg kick with 15.0 ppg mud, see red star in Figure 3.

www.wildwell.com

Tel 281.784.4700 USA

WELL CONTROL SIMULATION A TOOL FOR ENGINEERING AND OPERATIONS

Figure 5 50 bbl x 1.0 ppg kick, 17.0 ppg WBM, 6-1/2, 4 hour shut in on initial kick.

Figure 6 50 bbl x 1.0 ppg kick, 17.0 ppg WBM, 6-1/2, volumetric method to bleed gas to surface while maintaining constant drill pipe pressure.