OTC 15322 HPHT Well Control; An Integrated Approach
Rolv Rommetveit, Kjell Kåre Fjelde and Bjarne Aas; RF-Rogaland Research, Norman F. Day; BP Exploration (now Taylor-Day Limited) Eric Low; Think-Well Limited and David H. Schwartz, Well Control & System Design (now Petec Houston Inc.)
Copyright 2003, Offshore Technology Conference This paper was prepared for presentation at the 2003 Offshore Technology Conference held in Houston, Texas, U.S.A., 5–8 May 2003. This paper was selected for presentation by an OTC Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Offshore Technology Conference or officers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented.
The frequency of well control incidents is higher than one per HPHT well, and an increasing number of these take place during completion. The Physics of an HPHT Well An HPHT well can be categorized as an integrated physical system. The well is defined as the flow conduit from the mud pump down the drill string, drill bit and up the annulus to the mud pits; with the drilling mud, etc. filling up this flow conduit. Maintaining the control of the well at all times is a question of understanding the physics of the well during changing conditions and using this knowledge to optimize the design of the well, to develop sound drilling procedures and to handle unexpected situations during drilling in an optimal way. Traditional and well proven drilling practices and rules of thumb have been developed by “trial & error”. In many cases, these represent optimal solutions. However, when the drilling situation differs significantly from the traditional, old rules may not apply, and one will need to analyze the problems scientifically to revise the practices. One can use transient computer models with the correct physics built into them to develop new procedures and practices for these wells. Below the main physical parameters and interactions are discussed. Drilling Mud Composition. Drilling mud is a mixture of many components with different properties. The various components react differently to pressure (p) and temperature (T). The gas solubility and mixture properties of more complex hydrocarbon influxes vary significantly. The most common components are water, base oil and weight materials (solids). Other components are chemically active, mix or dissolve into the primary components of the mud for certain time periods during specific operations on the well. The duration of the time period may depend on the operations (for example drill cuttings or kick fluid). Drilling Mud Density. The density of the drilling mud is dependent on both p and T. The total mass of mud in a well will vary with the temperature distribution in the wellbore. The temperature distribution depends on formation virgin temperature and drilling history. The active mud volume in an HPHT well may change significantly when circulation starts or stops even when there
Abstract This paper presents an integrated approach on HPHT well control in narrow pressure margin wells. The approach includes theoretical evaluations, computer simulations with advanced modeling tools, laboratory studies, well control training as well as implementation of learnings in procedures and operations. Examples from the North Sea on the application of advanced transient well control modeling for the planning of HPHT drilling as well as training are described in detail. A special focus is directed towards gas diffusion during drilling operations; with its impact on a number of important parameters like rheology and barite sag. Both theoretical and experimental work is presented. Introduction The drilling of HPHT wells pose special challenges compared to standard wells: • High pressures and temperatures impact mud properties in a dynamic way, and can have effects on well control • Small margins between pore and fracture pressures will prevail in sections of the well • The conditions are above the critical point for the gas/oil/condensate influx; which means that the hydrocarbon influx is infinite soluble in the base oil of the mud. • Hydrocarbon influx will totally mix with the base oil in oil based mud (OBM), and infinite amounts of gas can dissolve in the mud • Drilling of inclined and horizontal wells will make the consequences of barite sag serious • Significant quantities of gas can diffuse into a horizontal section of a well if OBM is used even if the well is overbalanced
which depends on pressure and temperature. i. steel. There is also a difference between oil based mud (OBM) and water based mud (WBM) in the pressure and temperature dependence. due to the temperature driven rheology variations along the wellbore. However in high pressures and temperatures this effect is not large3. • The frictional pressure changes. However. stationary periods etc.
Phase Properties of Mud and Hydrocarbon Influxes. the mixture of mud and influx will be at a thermodynamic equilibrium. This effect is more critical in HPHT wells due to small margins and higher viscosities. mixing will take place between the mud and the influx.2
is no lost circulation or influx. Annular flow will only exist if the influx is allowed to move uncontrolled and expand near the surface. Thermophysical Properties. • Rheology changes can also induce flow regime transitions between laminar flow and turbulence. The dissolution process is not instantaneous.e. the time dependent gel strength increase with temperature for some muds. water and mud. Temperature Effects (Thermal Energy Transport). Mud rheology is not only temperature and pressure dependent. the speed of this diffusion process is driven by the difference in concentration of the gas in the mud and in the formation. wellbore hydraulics and kick probability are needed. In general. formation. The mud density distribution will change due to drilling with different circulation rates. 2. Calculations of dynamic temperature profiles require information about specific heat and thermal conductivity of the materials represented in the well. In general the volume of the gas influx will shrink when the gas molecules move from the free into the dissolved state. but also dependent on shear history. The volume of gas influx changes when it goes into solution. An influx of volatile oil (realistic for HPHT wells) in WBM will release free gas when it is pumped upwards in the well due to the pressure reduction.
Surge and swab pressures can be more critical due to two factors: Firstly. Multi-Phase Flow of Mud-Hydrocarbon Mixtures. these properties may be obtained from several literature sources1. OBM will behave significantly different from WBM in the case of a kick. and the upper part of the annulus will be heated since hot mud is flowing up the annulus. Influx of a dry gas in OBM will at HPHT conditions be infinite soluble in the base oil. After a kick or with an influx caused by swabbing. the temperature changes are not so large. This normally results in change in the total mud volume in the hole. Pressure variations can be larger in HPHT wells than in standard wells for several reasons: • The hydrostatic pressure varies more due to mud density changes caused by thermal effects. The major part of this will dissolve in the base oil. and hence the rheological variations with temperature are small. Heat transfer and frictional heating depend on mud properties. Mud temperature at a given position in the well changes rapidly depending on the ongoing drilling operation. well fluid. careful evaluations and analysis of the effects of temperature and pressure on rheology. so errors in the estimation of the dynamic circulation pressure have little consequences for well integrity or kick probability. The temperature approaches the geothermal temperature during long stationary periods. cement. Many wells have a large gap between pore pressure and fracture pressure. The bubble point and volume of free gas released will be different than that from the volatile oil alone. When circulation starts. Mud density as well as mud rheology at a given position will change rapidly in this phase. In many cases this is a good approximation. mainly in the drill collar section of the annulus. viscosities in the deepest and hottest parts of the wellbore may be higher (temperature driven effects) and secondly. but is governed by molecular diffusion. The flow of free gas will in general take place in the bubble or slug flow regime. which depend on mud temperature. Little or no data is available for HPHT drilling fluids that are mixtures of several components with different properties. The bubble point of the mixture will govern where in the well one will have free gas flashing from the mud. the lower part of the annulus will be cooled by cold mud from the drill string. Hydrocarbon influxes will in general range from dry gas (methane) through volatile oil and condensate to heavy oil. After an influx in a well. A fraction of the hydrocarbon influx will then dissolve in the mud. for wells with small margins between pore and fracture pressure. the flow of mud-hydrocarbon mixtures upwards in a wellbore is a very complex and transient multi-phase flow process. For shallow wells. The reason for the changes may be one or more of the following effects: • Mud expands or contracts due to temperature variations • Mud expands or is compressed due to pressure variations • The diameter of casing and open hole sections increases or decreases (ballooning) Drilling Mud Rheology. Hence. Note that mud properties and well temperature influence each other mutually. This free gas will expand according to the real gas law. A rapid peak in the bottom hole pressure has been observed as gels are being broken during start up of pumps.
. Influx of volatile oil in OBM will mix totally with the base oil and a new “pseudo base oil” with different PVT properties than the volatile oil will be created. The rheological properties of drilling fluids are often approximated to be independent of pressure and temperature. Pressure Effects. Except for the well fluids. Hydrocarbon gas solubility in the oil phase of the drilling mud is several orders of magnitude larger than in the water phase. Turbulence will create higher frictional pressure losses. If gas diffuses into the well through the mud spurt zone in the wall of the well bore. on the position in the well.
Evaluating the potential for hydrate formation. methane is infinitely soluble in the oil. in particular with increasing water depths. This is usually called dynamic sagging. then formation fluids (water and/or hydrocarbons) will flow into the well. To fully evaluate and take into account the interactions. Flow of Reservoir Fluids into the Well during Underbalanced Conditions. pump rate etc. For the evaluations of lost circulation and underground blow-out scenarios. The variations in drilling parameters (ROP. Diffusion of Reservoir fluids into the Well during Overbalanced Conditions. The rather shallow mud spurt zone created when using oil based mud will enhance the process6. Gas or volatile oil in OBM will have the following special characteristics: • As long as the influx is dissolved. Laboratory Characterization of Mud and Mud-Influx Properties Purpose.. A global transient model for calculation of the impact on downhole pressures from this effect has been developed10. Precipitation of bentonite was observed in Cs/K formate mud. Also. The effect of dynamic sagging may become pronounced in wells with long. lower value. and a high gas peak can be expected when the gas reaches the surface.e. is left without circulation for an extended period of time. it may take place at less than 5% in the non-Newtonian fluids4.) create a very transient downhole situation. A significant loss of weight material from the mud may cause serious problems for the pressure control when the lighter mud reaches sections with small inclination. Dynamic Sagging of Barite. This can create a difficult well control situation. Rheology versus pressure and temperature has been investigated for a number of drilling muds used in HPHT wells. For WBM the pressure effect was much smaller than the temperature effect9 . a special version of the well control model used in this paper has been developed8. this precipitation process proceeds quite slowly. when the circulation rate is low enough for the flow to be laminar and the drill string is not rotated or rotated only slowly. The pressure dependence was much more pronounced for OBM than for WBM. if advanced transient well control and thermohydraulic simulators are used. Darcy’s law regulates the flow rate. dynamic simulators for the well and pressure control related scenarios should be used. The slug flow regime will give a much higher gas slip velocity than the dispersed bubble regime. In order to perform a thorough evaluation of the well control aspects of HPHT drilling. However. This model accounts for the dynamic temperature effects on the hydrate probability7. as well as investigating various well control strategies versus the hydrate formation probability. RPM.
. It has been demonstrated in tests performed at RF-Rogaland Research that all mud’s sag during shear flow5. horizontal sections. In the case of OBM at
HPHT conditions. dissolved salts for increase of density. This phenomenon has been studied with local experiments in the laboratory. their viscosity will start at an initially higher value and drop asymptotically over time to a final. Interactions between All Physical Effects during Drilling Operations. No slip will then distribute the influx over a larger portion of the mud. Hydrate formation in the well. can now be done using a recently developed transient well control model. Effects of Hydrocarbons on Mud Properties. knowledge of the physical properties of the fluids involved is of significance. and in instrumented field tests to evaluate the global effect on the gel breaking pressures in HPHT wells. Mud properties will be affected by hydrocarbon influx. Rheology. • The expansion from such a system can be much more violent than from a gas-WBM system due to the combined effect of large expansion and violent flashing near the surface. Sagging of solids may take place even in drilling fluids which use heavy. This will not influence the density of the mud by much. if exposed to a constant shear after some time at rest. Hydrate formation may take place in HPHT wells. to distinguish it from the sagging which may take place in fluids at rest. Both OBM and WBM have been tested10. I. For most OBM pressure and temperature effects nearly cancelled out when pressure increased from atmospheric to 1000 bar and temperature increased from 50 to 150 °C. Weight material will typically sag out of the drilling fluid in long. While this transition is at approximately 25% gas volume fraction for Newtonian fluids. High turbulence and vigorous whipping of the drill string will re-suspend precipitated material.OTC 15322
Transition between bubble and slug flow will depend on the non-Newtonian properties of the mud-influx mixture. The rheology of most drilling fluids shows thixotropy. If a well. realistic input of fluid properties as part of the modelling will greatly enhance the relevance of the work. it will stay rather concentrated. and it may become important only in relatively extreme cases. and substantial amounts of gas can diffuse into the mud. The flow rate is dependent on the permeability and the pressure difference between well and formation. highly inclined sections of a well. which is drilled overbalanced through a gas formation. During drilling operations the wellbore conditions is changing dynamically. If the pressure in the well is lower than the formation pressure during drilling and the formation is productive. gas from the formation will diffuse through mud spurt zone and filter cake and accumulate in the drilling fluid. but the reduction in viscosity will reduce the carrying capacity of cuttings. The effects will be significantly larger in OBM compared to WBM due to the high solubility and mixing properties of hydrocarbons in OBM.
The cell is then left for. and pressure and temperature. methane is infinitely soluble in oil. gas. Typically. Several benefits can be identified:
. experiments in PVT laboratory should be performed with the mud and influx expected. A change in mud volume can be misinterpreted as a gain or a loss while it is actually due to volume change within the mud itself. Detailed information on the mud density development with pressure and temperature is of great importance for the accurate computation of ECD effects during drilling. A specified amount of gas may also be pumped into the cell before or during the heating. with influx of formation fluids in the mud. and are in general a function of p and T. The simulator has been thoroughly validated against downhole pressure and temperature data that has been recorded under controlled circumstances in North Sea HPHT wells. Generally. gas may diffuse into the well even when the well pressure is properly balanced. The drilling temperature profile will be dependent on parameters like mud thermophysical properties. Some published data exist1. 8. Transient Thermohydraulic Model An advanced transient pressure and temperature simulator has been developed. oil etc). that volatile oil mixed in the mud or gas dissolved into the mud will degrade the muds carrying capacity for barite. and has identified the parameters determining the speed of precipitation5. Submodels for all important physical effects are also modeled. However. The situation becomes worse in HPHT wells than in other wells. PVT Properties of Mud-Influx Mixtures. The well control model solves the conservation equations of mass and momentum for the various phases involved during a kick (mud. Models exist to calculate the changes in mud volume. weight material will sag out of drilling fluids during steady shear flow. • Gas solubility and flashing in OBM • Exceeding fracture limits in the well • Prevention / detection of borehole ballooning • Early detection of influxes to minimize the handling of very large gas volumes at surface. and inversely proportional to the viscosity at the given shear rate. • To reduce the increased risk of swabbing in heavy weight muds. formation) is treated separately • Transient modelling of pressure and temperature as functions of time and position • An arbitrary sequence of changes in operational parameters (mud circulation rate. influx type and concentration. the precipitation velocity varies strongly in different muds. cement. etc.) can be simulated. It is important to understand the change in mud volume to be able to identify real kicks. while a piston reduces the volume of the cell from below to prevent boiling in the cell. Mud and formation fluid mixed in controlled fractions are filled into the cell. Dynamic Barite Sag . • Both absolute temperature and temperature variations increase with depth. the sagging velocity is approximately proportional to the local shear rate in the fluid. Today the mud companies characterize the fluids used over the p and T range of interest for the drilling operation. Thermophysical mud properties vary for different mud types. Influx of formation fluids will occur during underbalance. and has been described in detail elsewhere15-18. The following features are important for the simulations that are described in this paper: • Vertical and radial heat transfer • Frictional heating • Radial and vertical discretisation . The content of the cell is then tapped off from the top of the cell in small samples. The density of the samples will tell whether the carrying capacity of the mud has degraded at the static conditions. but there are large differences for different mud formulations also within each group. drill string rotation. The same holds for absolute pressure and pressure variations. RF has developed methods to quantify a mud’s propensity for dynamic sagging. the pit gain is an important parameter for identifying possible kicks. At HPHT conditions. say. e. the content is stirred thoroughly with a magnet stirrer. Most modern drilling muds are composed so that they do not sag at static conditions.4
Density. but as described above. Static Barite Sag. Some of the reasons for this are: • The total mud volume can be larger. This can be investigated in the laboratory with a suitably modified HPHT rheometer.each material (steel. OBMs will sag almost an order of magnitude slower than WBMs. 14. and the total set of equations is solved numerically. The effects of formation fluid influx in the mud on static sagging can be tested in a suitable HPHT cell in the laboratory. The thermophysical properties at HPHT conditions have been studied12. As noted above. Transient Well Control Simulator An advanced well control simulator has been developed for evaluations of well control scenarios and related issues as part of the planning process of well construction7. and therefore variations in active volume due to pressure and temperature variations will be larger. 13.
Well Control Incidents which can be prevented and handled better with improved understanding During drilling of the deeper parts of a well. 24 hours.g. The PVT properties of mud and hydrocarbon influx mixtures vary significantly with mud type (WBM or OBM). well geometry and drilling history.19. Such studies have been reported3. base oil composition. The investigations show large differences in the sagging velocity for different muds. When the cell has reached the nominal pressure and temperature. In order to evaluate the well control effects of an influx. One can expect that adding a mixture of formation fluids to the muds will influence the sagging velocity. this may change. flow rate. However. Such a simulator is a very powerful tool for well control evaluations relevant for HPHT and other wells with small operating margins. Thermophysical Properties.
It should reveal the contribution from different operational phases and specific causal mechanisms and allow identification of critical factors. An important issue for avoiding well control incidents is to have good knowledge of the effective bottom hole pressure. Using this. Such an approach should include evaluations based on probability statements with respect to relevant physical quantities and operational events related to the well that is planned. A 1000 m long horizontal section through the reservoir was planned. a new risk analysis tool with focus on kick and blowout has been developed in order to improve the basis for risk-based well control planning. accurate predictions of the expected ECD’s and fingerprints can be achieved. From this figure. 1 shows the simulated ECD for the Cesium/Potassium formate mud that was planned for use in the Devenick HPHT well. the bottom hole pressure for various pump rates while pulling out of the hole is shown. This can be achieved by using information from offset wells or by using the model during the operation with updated information. The well was classified as an HPHT well with a reservoir temperature of 150 C. Rogaland Reserch was contacted with a request of performing some hydraulic modeling for this particular well. Since the gas influx dissolves in the mud. which enables the planning to include more well and operation-specific information in the analyses and to obtain a more differentiated risk description. Both hydrostatic and frictional pressures will depend on the present well conditions that will change according to the operational state. It is of importance to be able to distinguish such “thermal” driven fingerprints from real influx situations and in that respect. at some stage free gas will occur and a sharp increase in the pit levels can be seen. there will be no signals from the pit tank that an influx is being transported towards surface. • Develop revised procedures if current procedures are not optimal • Improve crew training and make it more relevant Examples from HPHT Wells Spring 2001. KickRisk. Fig. 2. The well was initially planned with OBM. very accurate modeling results can be achieved if measured PVT data of the mud is available and used in the modeling process. Uncertainty is expressed through probabilities related to factors at a detailed system level and is propagated in kick and blowout probabilities through the logic of the model structure20-23. In addition. In the Devenick study. This can be related to a fingerprint that is expected to be observed in the pit tanks where a slight expansion should be seen. In Fig. The KickRisk model is used as part of the planning of HPHT wells in Norway to identify critical risk factors and kick risk reducing measures. • Confirm best practice procedures. but a switch to Formate brine was found necessary at a later stage.OTC 15322
• • • • • •
Provide more realistic kick tolerances compared to using simple single bubble models Evaluate different kick scenarios Evaluate and develop well control procedures Investigate the possibility for hydrate formation Separator desing Training
Quantitative Analysis of Kick Risk Analysis of well-specific kick probability should also be part of an integrated well control approach for HPHT wells. it shows how the total static mud weight tends to decrease during a connection due to a net temperature increase in the well. The Objectives of Modeling Work in Planning an HPHT Well The objectives and motivation to perform modeling work in the planning phase of an HPHT well are multiple: • Address the concern over the high number of well control incidents during drilling HPHT wells • Address the fact that a growing number of well control incidents in HPHT wells take place during completion • Identify any well control risk for the particular well • Confirm that existing procedures are according to best practice. The kick is first detected when it is almost at surface and the drilling crew have a very short time to activate well control
. the frictional pressures for various flowrates can be determined.
BP’s motivation for performing the modeling work was: • There was a need for evaluating the combination of HPHT and horizontal well control procedures . dynamic modeling can be a valuable tool. the incident of an undetected kick can cause problems. In addition. One of the concerns was an expected high pressure zone just above the reservoir formation. Focus in the well control studies were: • Kick Tolerances • Undetected Kicks • Kill methods • Comparing kick behaviour in OBM vs brine • Surface flow behaviour In oil based mud. temperature distribution and swab pressures. The idea behind this tool is more detailed system modeling. and a planned TD at 4613 m TVD. • Identify any specific well control risks • Evaluate casing design – Kick tolerances • Address the change of fluid system from OBM to brine (initially motivated from the planned switch to brine during completion and the increasing number of kick incidents during completion) • Improve crew training and make it more relevant • Address the concern for gas diffusion in OBM in the long horizontal section Thermohydraulic Modeling. Kick Modelling. However. BP Aberdeen was in the planning process of the Devenick HPHT well in the North Sea. which could cause well control incidents and had a direct impact on the casing design. both oil based mud and brine was modelled and compared with respect to ECD’s. There is a clear need for pumping out of the hole to maintain an overbalance. When comparing field results with the study results it became clear that specifying accurate inlet and outlet mud temperatures was very important.
A kick taken in OBM will dissolve and remain at bottom if circulation is not resumed. as methane is infinitely soluble in oil at HPHT conditions. It also became clear that the existing volumetric well control procedures were not adequate since they did not account for inclination effects and hole geometry changes. Rheology Changes due to Formation Influx. In Fig. rig crew must initiate well control procedures very quickly in order to avoid fracturing the casing shoe. with an initial diffusion rate of about 0. 4 show the situation when a 4 barrel undetected kick is circulated up prior to cementing the 9-7/8” casing. The results indicated clearly that this would have involved unacceptable risks. the situation is quite different if a free gas kick is taken in a brine mud. Hence.5” and 5”. the instrument can be pressurized with formation gas. a sharp increase in surface pressures can be seen. However. substantial amounts of gas can become dissolved in the OBM. The well pressure is 724 bar and the temperature 145 ºC. Most of such effects can be investigated in the laboratory. both for cuttings and for weight material. new procedures were developed and verified with the kick model. an example of a kick tolerance curve is given. when the well is left quiescent for some days or weeks. If this happens. With a suitable modified HPHT rheometer. A method to calculate the magnitude of the diffusion rate was described in (6). and a very viscous fluid is left in the well bore. Due to the relatively small solubility of methane in brine. Theoretical Evaluations of Gas Diffusion. 10.
Existing procedures could in fact make the well go underbalanced.6
procedures. either through swabbing or through diffusion. These are pure geometrical effects. as a result the 9-7/8” casing was not deepened. 3 and Fig. well control procedures had to be initiated quickly since well pressure would build up rapidly during shut-in due to fast migration rates. in the 8. and a following pressure control problem assigned to the wrong causes. The calculations assume a shallow mud spurt zone of 1 cm. well control simulations showed that there were no benefits in using the Wait and Weigh kill method when drilling the 12 ¼” hole section. it was decided to use the more advanced dynamic kick simulator for updating the tolerances when necessary and to base the casing design on this. For instance. better data became available and kick tolerances were updated. in this way the effect of dissolved gas can be investigated. formation gas can diffuse into a quiescent well even if the well is properly overbalanced. Design Modifications. The consequences of formation fluids being mixed into the drilling mud. and uniform bore hole and formation properties for the whole well length. some uncertainty existed whether it was possible to drill through this high pressure formation without setting the 9 7/8” casing first. one should be aware that the measurements might become corrupted from the weight material sagging out of the fluid. 11 and 12 show results from calculations for the HPHT Devenick well. we can also observe that the surface pressures can vary due to geometrical changes. It is of importance to know how different muds behave relative to formation fluid influx of different kinds. In this case. Even during closed-in conditions. Swab calculations confirmed the need for maintaining circulation to minimize the risk of swabbing. These effects can be investigated in standard or modified HPHT rheometers. Through close collaboration. the modeling tools were used during the drilling operation. gas migration will take place and lead to increased well pressures. The effects of different mixtures of formation
. As the influx leaves the horizontal section and migrates upwards. they may be masked by other processes. There were also doubts about the strength of the formation around the 13 3/8” casing shoe. Modifications to Procedures. Initial kick tolerance results showed that the single bubble model approach used by BP was too conservative. like temperature effects. Fig. using actual field data to assess the feasibility of setting the casing deeper. Enhanced sagging due to Formation Influx. Laboratory Experiments. oil and water phase may be squeezed into the formation. As the drilling operation proceeded. The kick model was used to validate existing procedures.9 bbl/hour. However. horizontal well and performed the calculations for the two wellbore diameters of 8. Fig. As mentioned above. If a kick in brine occurred. Hence. Due to the expected high pressure zone just above the formation. but it warns that the effect may be substantial even for a more realistic cases. This may affect the carrying capacity and the mud viscosity. From these figures. in WBM the effect will not be large enough to result in serious problems. The example may thus represent a worst case. may be a serious degradation of mud quality. After a while it enters a region with larger volumes and the kick length is shortened which temporarily reduce the surface pressures. Influxes taken in brine will typically lead to larger pressures and gas volumes at surface during the well kill compared to that seen for an OBM. 5 shows the situation. Another issue in the study was to keep focus on the possible change of fluid system (from OBM to brine) and the impact this had on well control behaviour. Figs. The effect may be especially large in HPHT wells with OBM.5” well one may have a volume of about 35 bbl methane at downhole conditions diffused into the well after the first 3 days. One may add that even if the effects seem large. As was demonstrated. An example is shown in Figs. Simulation results showed that the rig crew had to be very careful and avoid undetected influxes in oil based mud. Dissolved gas or other formation fluids mixed into the drilling mud may degrade the mud’s carrying capacity. The initial modeling work focused on providing kick tolerances for validating the planned casing design. If larger amounts of gas diffuse into the OBM of a non-circulating well the water-in-oil emulsion may even break down. One has assumed a 1000 m long. The gas influx may thus remain undetected. 6 to 9 where we have compared two well control scenarios and assumed a perfect kill situation. Mud and non-volatile influx fluids can be mixed before the rheometer is brought to HPHT conditions. As the figures show.
The rig crews felt well prepared and were knowledgeable on what to expect from fluid behaviours in the well. Australia. Eymard. A rig crew training course was arranged prior to the operation. Perth. Confidential.. R. They confidently and successfully applied complex kick prevention procedures avoiding any major well control events. R. 6. and White. The effort put into well control planning. 1990 5. The rig crew was encouraged to come up with scenarios. focusing both on hydraulic calculations and on kick scenarios. Rommetveit. B.: “Gas Rise Velocities During Kicks”. PVT Measurements of Base Oils. Modelling Support during Drilling Operation. which could be run during training. depths. B.. The decision to use CsK formate mud and not OBM through the reservoir was made for a variety of reason. and Tarr. and Merlo. Petersen. Sixth edition. Aas. Johnson. And Lekvam. ROP was a major issue and the reduced ECD with CsK made a very significant contribution in improving ROP. SPE 20431. • The verification of modified well control procedures through modeling mitigated concerns over changing to a fluid with very different behaviours. a Tool for Design of Complex Well Control Situations”. procedures for this have been established during earlier investigations at RF5. N.implementing learnings in procedures and operations . B. 2002. Its Impact on a Horizontal HPHT Well”. the PC based modeling software used in the study was run live for the rig crew. This approach consist of . and Merlo. Sterri.training . covered the specific drilling practices and well control procedures that were to be used on the well and the results of the well control study.OTC 15322
fluids can be investigated in a suitably modified HPHT rheometer. Bjørkevoll. B. Model output/results were discussed/combined with general HPHT experience. kick tolerances were recalculated. K.. J. Paper SPE/IADC
. Rommetveit. Conclusions An integrated approach to HPHT well control has been presented in this paper. J. Bjørkevoll. the development of new procedures. and Larsen. A. “Gas Diffusion. H. 3.QRA analysis of kick risk . • The thorough understanding of CsK fluid behaviours.: “Kick with Lost Circulation Simulator.. W. Devenick was a major operational success for BP. Amsterdam. Gjeraldstveit. D. SPE 77474. During critical phases of the operation. D.. RF report PRC K36/86. A.theoretical evaluations . S. Deciding to use CsK Formate Mud and not OBM Through Reservoir. E. H. Perry’s Chemical Engineers’ Handbook.D. Rommetveit. Rommetveit. 10.. R. Focus was also on giving the crew a feel for the differences between OBM and WBM behaviours during well control. Operational Success with no Well Control Problems. N.S. Green. It included: • Company personnel • Senior rig crew to derrickman • Main service company personnel • Key shore based personnel In the course. 1997. Corre. • Virtual elimination of the risk of gas diffusion as mechanisim for gas influx into the horizontal well bore. temperature data and updated FIT information. • The reservoir formation was exceedingly hard (2000 bar UCRS). • Elimination of the barite sag.: ”Computing the Danger of Hydrate Formation Using a Modified Dynamic Kick Simulator”. HPHT Training. and Guenot. The model was calibrated with the most recent data for survey. B. K. 27 February1 March 2001 8. K. K. 1984. B.operational support Use of the main elements of this approach has contributed to the successful drilling of a horizontal HPHT well with a minimum of well control problems. Acknowledgements The authors are grateful to BP and RF-Rogaland Research for the permission to publish this paper. The Netherlands. increasing the understanding of HPHT well control phenomena and crew training resulted in a successful well. being the first HPHT horizontal well drilled. Development of Small Scale Instrumentation for Testing of Barite Sag in Drilling Muds. A. Gard. Aas. A significant achievement considering the swab risk in horizontal hole sections with heavy weight mud. "Numerical Computation of Temperature Distribution in a Wellbore While Drilling". References
1. In addition. 7. Low. 2. A.. Bahrain. mud weights and rheology. kick tolerances were updated to provide input for the future decisions. Bradley. “Transient Gel Breaking Model for Critical Wells Applications with Field Data Verification”. SPE/IADC 39282.laboratory investigations . R. McGraw Hill 1984. SPE 49956 presented at the 1998 SPE Asia Pacific Oil & Gas Conference and Exhibition.. Petersen. 4. “Temperature and Pressure Effects on Drilling Fluid Rheology and ECD in Very Deep Wells”. A. This simulator was programmed from the modeling studies to enhance realistic training. a rig floor well control simulator was used extensively. especially during well control from modeling studies. 12-14 October 1998 9. These included: • Return permeability tests on core samples indicated a substantial improvement with CsK.. SPE 13208. Rogaland Research report RF-1999/263. SPE/IADC 67749 presented at the SPE/IADC Drilling Conference. During drilling operations. Bjørkevoll. and Aas. MDT log results showed that the horizontal reservoir section had been successfully drilled with an overbalance of only 200 psi. B. J.computer simulations and scenario developments with advanced modeling tools . The course also contained some HPHT theory.
1984. Then the pump is shut off. and Rommetveit.First the rate is 425 gpm – circulating until steady state. 15. Presented at the 59th SPE Fall Conference.
Bottomhole pressure when swabbing at 60 ft\min
11200 Pressure (psi) 11000 10800 10600 10400 10200 10000 0 100 200 300 400 500 Pump rate (gpm) Bottomhole pressure when swabbing Static Mud Weight (1. SPE Drilling Engineering March 1988." SPE 14182.. T. and Bjørkevoll.... A. Ravenna.. June 1997. March 11-13. B. Ekrann. 7–8 March 2001.. Texas. "Predicting Mud ECD in HPHT Wells". P.
23. Elsevier Reliability Engineering and System Safety 61 (1998) 53-63. “A practical Case of Assessing Subjective Probabilities – A Discussion of Concepts and Evaluation of Methods”. Vefring. K. and Maglione.. Nilsen. R.
Figure 1 – ECD for a 1.
. Rommetveit. Lund Bøe. "Prediction of Downhole Pressures and Temperatures in HPHT Well". and Guarneri. "A Simulator for Gas Kicks in Oil-Based Drilling Muds. T. Berg Andersen. The Netherlands. and Rommetveit. Vefring. Monger. and Maglione.H. Rommetveit. 22/23 November 1995. "Simulator accurately predicts conditions in HPHT wells". O’Bryan. T. Bjørkevoll. R. Sept. June 9-12.-25.. B. Guarneri. p77. before it increase to 80.62 sg Formate mud in the 8 ½” section. Anfinsen. ”Risk-Based Well Control Planning: The Integration of Random and Known Quantities in a Computerized Risk Management Tool”. SPE 13208..785 412 E-03= m3 in x 2.-T. Rogaland Research 1997 (Confidential). 20. Øystein..54 E+00= cm lbm x 4. Anfinsen. 1994..159 E 00 = m3
Figure 3 – A four barrel kick is taken when running the 9 7/8” casing. L. Sandøy. Berg Andersen. Kristiansand 4-6 October 1994. R.048 E-01 = m gal x 3. B. P. New Orleans. T. 22.535 924 E-01= kg psi x 6. ”An Experimental Study of Gas Solubility in Oil-Based Drilling Fluids”.. and Rommetveit R. T. E. SPE 35967. 1993 Bjørkevoll.H. L. 120 and 425 gpm. Safety & Environment.
16. Presented at OMC93.T.. Italy. Eymarol. R. E. Gas Solubility and Undersaturated Compressibility Measurements of Drilling Fluids. SPE 68447.
21. and Aadnøy. Houston. Circulation rate 300 gpm. 1985. "Changes in Active Volume Due to Variations in Pressure and Temperature in HPHT Wells". “On Risk Interpretation and the Levels of Detail in Quantitative Blowout Risk Modeling”. “Stochastic Modeling for the analysis of blowout risk in exploration drilling”. L. A.-T. Bourgoyne. R. Labes. B. Corre. R.G. Oct 10. “Numerical Computation of Temperature Distribution in a Wellbore while Drilling”. D.
79843 presented at the SPE/IADC Drilling Conference in Amsterdam. Nilsen. Texas. M.
19. Aven.S.P. R. 22. Anfinsen.
17.. and Kopeck.H. S.
12. K. Las Vegas. Paper presented at the 7th Northern European Drilling Conference. Bjørkevoll. Vefring. 19-21 February 2003. E.62 sg) Open Top
Figure 2 – Shows how important it is to maintain circulation during swabbing operations to omit underbalanced conditions. Conference paper presented at the 9th Annual Conference in Aberdeen. K. Report RF-1997/294.8
11. Presented at the 1985 SPE Fall Conference.
14. Presented at the SPE/ICoTA Coiled Tubing Roundtable held in Houston. 3342.L. B. Offshore Drilling Technology. Oil & Gas Journal. Berg Andersen..S.. No expansion in pit before the influx boils out of solution. “Thermophysical Properties at HPHT Conditions”. p. and Guenot.
SI Metric Conversion Factors ft x 3.894 757 E+00= kPa bbl x 0.-T. K.. 13. A. Rogaland Research 1999 (Confidential). Proceedings of the ESREL ’97 Conference. Report RF-1999/178.S. 18. Lisbon. A. Aven.S. Presented at the 1996 International Conference on Health.
Note the pressure build up even at closed-in conditions. The kick is taken in the horizontal section.
Figure 8 – Gas flow rate out for the 100 barrel kick in oil based mud.
Figure 9 – Gas flow rate for the 100 barrel kick in brine.
Pressure changes caused by geometry differences (inclination and flow area)
Increase caused by gas expansion
Figure 6 – Choke pressure development for a 100 barrel kick in OBM. a 20 barrel kick has been taken in brine. Note that the surface flow rates are significantly larger and bottomsup time is significantly lower.
Figure 7 – Choke pressure development for a 100 barrel kick in brine – Note significant larger pressures.
Figure 5 – In this case.OTC 15322
Figure 4 – Position of influx when the sharp increase in pitgain is observed caused by presence of free gas.
8020 P re ssure a t Ca sing shoe (psi) 8000 7980 7960 7940 7920 7900 7880 7860 7840 7820 0 20 40 60 80 100 1.Accumulated gas volume during the first days.75.9 0.5 0.4 0.Casing shoe pressure for different kick sizes using reservoir pressures of 1.5'' w ellbore 5'' w ellbore
Figure 12 -Volume flow of gas during the first days.6 0.
Gas innflux rate
.77 s g Fracturation pres s ure 1.0 0 2
Gas rate (bbl/hour)
8.75 s g 1.78 s g
K ick S i z e (b b ls)
Figure 10 .7 0.77 and 1.3 0. 1.
Accumulated innflux of gas
60 Gas volume (bbl) 50 40 30 20 10 0 0 2 Time (days) 4 6
8.2 0.1 0.5'' w ellbore 5'' w ellbore
Figure 11 .8 0.78 sg.