Methodology For Manipulation Of Wellhead Pressure Control For The Purpose Of
Recovering Gas To Process In Underbalanced Drilling Applications R.A. Graham, M.S. Culen, Precision Drilling Technology Services Copyright 2004, SPE/IADC Underbalanced Technology Conference and Exhibition
This paper was prepared for presentation at the 2004 SPE/IADC Underbalanced Technology Conference and Exhibition held in Houston, Texas, U.S.A., 1112 October 2004.
This paper was selected for presentation by an SPE/IADC Program Committee following review of information contained in a proposal submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers or the International Association of Drilling Contractors and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, the International Association of Drilling Contractors, their officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers or the International Association of Drilling Contractors is prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300 words; illustrations may not be copied. The proposal must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. Abstract An updated approach to underbalanced well design has been developed with a view to enhancing the ability to recover gas to process (pipeline, gas plant, offset well, temporary storage, etc.). With the growing trend towards reducing greenhouse gas emissions, this technique evaluates the range of anticipated operating conditions (gas inflow and wellhead pressures) as a function of liquid and gas injection rates and the productivity index of the reservoir. Typically, underbalanced drilling projects have been designed to minimize gas injection requirements down the drill string or annulus (concentric injection or parasite injection) for equipment efficiency and erosional considerations. This means that wellhead pressure is typically kept to a minimum, maintaining sufficient pressure to aid in shipping liquids from the separator. An exception to this is where elevated wellhead pressures may be desired to increase bottom hole pressure and reduce inflow from the reservoir, however, the objective in this instance is to control bottom hole pressure. The typical approach to designing for underbalanced drilling parameters is normally based upon a graph of bottom hole pressure on the y-axis as a function of gas rate on the x-axis. It is usual to overlay the hole cleaning parameters based upon minimum annular liquid velocity and minimum and maximum equivalent flow rates through the motor. In addition, the target bottom hole circulating pressure limitations (usually derived as a function of reservoir pressure) define the operating window from which the required liquid and gas injection rates can be determined. This updated approach addresses many issues such as equipment sizing and expected production rates while drilling through to effective pressure management at surface to aid in well stability. Furthermore, the approach lays down the fundamental groundwork for future underbalanced control algorithms. Introduction The approach to well design for underbalanced drilling (UBD) has been fairly well established over the past 10 years. It is typically an iterative design process to establish an operating envelope that meets the desired bottom hole circulating pressure window, while meeting hole cleaning requirements and optimizing power fluid needs of the downhole motor. Most people involved with underbalanced drilling would probably agree that the growth of the underbalanced drilling market has lagged expectations. In part, this is due to the increase in complexity to properly implement the technology, however, the biggest hurdle is still the incremental cost to the operational day rate, which is often disconnected from the ability to add value to the asset. A further obstacle is the growing environmental challenge associated with the flaring of gas during drilling operations. Conversely, growing demand for natural gas and the steady increase in gas prices has fuelled an increasing demand for UBD technology to be applied in more prolific gas reservoirs that were normally off-limits due to the potential for high gas rates at surface and the increased demands placed on surface handling equipment. These challenges have been addressed through the development of enhanced UBD equipment, such as high pressure separators and high pressure rotating control devices to expand the potential operating window. In order to best exploit this new high-spec equipment, an updated approach to underbalanced well design is also required. The following discussion introduces the concept of steady-state and non- steady-state operating conditions and the rationale for manipulation of wellhead pressure and gas-liquid injection rates. The key features of this approach are discussed along with sensitivity analysis and typical application strategies. Finally, several case studies are included to demonstrate the effective application of this evaluation technique. Conventional UBD Well Engineering Figure 1 contains a typical set of curves developed from a steady-state multi-phase flow model for a range of liquid pump rates. The graph shows the operating window constrained by the desired bottom hole circulating pressure 2 IADC/SPE 91220 (BHCP) limits, the minimum liquid annular velocity desired for adequate hole cleaning and the maximum recommended motor throughput. It should be noted that often, the right-hand side of the operating window that is attributed to the maximum recommended motor throughput can be shifted to the right through careful selection of bit nozzle sizes to increase the differential pressure across the bit. This will further compress the gas volume going through the motor and increase the effective gas injection rate that can be comfortably used within the confines of the motor specifications. One of the major challenges associated with the conventional approach is how to account for reservoir influx. While most multi-phase flow models are equipped to facilitate reservoir inflow scenarios, it can be a tedious and confusing exercise to perform a sensitivity analysis that focuses on both of the key independent control parameters liquid injection rate and wellhead pressure. Another challenge associated with multi-phase models and highly prolific wells is the existence of multiple solutions for a given wellhead pressure. This phenomenon has perplexed many UBD engineers when tubing performance curves developed to account for reservoir inflow do not take the form of predictable, smooth curves. This is due to the standard approach of fixing the wellhead pressure and having the computer model iterate to determine the associated bottom hole circulating pressure. By all conventions, this should represent a sound approach, since wellhead pressure is operationally the independent (controllable) variable. Furthermore, the conventional approach fails to take into consideration that a portion or the entire operating window highlighted in Figure 1 may not be practically achievable in steady-state conditions. This fact alone could have significant ramifications on the ability to effectively execute the UBD plan. The existence of conditions not suitable for steady-state operations with a fixed wellhead pressure has been recognized. An approach to understanding this phenomenon is to separate the conventional presentation into tubing performance curves overlaid by inflow performance relationship (IPR) curve(s). Viewing Drilling from a Production Point of View Figure 2 can be considered as tubing performance curves for drilling, where the tubing is comprised of the annulus between the production casing and the drill string. In order to evaluate the effects of wells with high productivity indices (PI), it is important to understand the potential operating conditions that may be encountered. This can be done by overlaying the IPR curve over the tubing performance curves and evaluating the intersection points as shown. This demonstrates how manipulation of the liquid rate and wellhead pressure can influence the inflow from the reservoir, where m represents the minimum gas rate that will be encountered in steady-state conditions, and conversely, M is indicative of the maximum gas rate that can be expected within the confines of the tubing performance curves shown. The corresponding y-intercepts of these two points define the range of BHCP that is achievable in steady-state conditions. The challenge with this hybrid approach is that it becomes quite challenging to visualize what is happening dynamically, since the only valid points exist at the intersections of the tubing performance and IPR curve(s). What is needed is a more efficient approach to visualize the dynamics of controlling well parameters using a steady-state model and one that is practical for performing sensitivity analysis. After all, a model is merely a mathematical simplification of a complex process and requires a degree of latitude to permit practical implementation. An Updated Approach to UBD Well Engineering Key Features This updated approach to underbalanced well design centres on analysis of the operating window, taking reservoir inflow into account. This analysis combines the effects of tubing performance and inflow performance and focuses on determining a range of minimum and maximum gas inflow rates that can be expected in order to operate in steady-state operating conditions. The analysis also considers the sensitivity of applying a range of productivity indices. Although this paper focuses on the analysis of gas wells, the phenomenon has also been noted with high GOR oil wells. Figure 3 illustrates this operating envelope analysis technique. The key elements of this presentation are as follows: Maximum Wellhead Pressure (WHP) Threshold Wellhead Pressure to Initiate Reservoir Inflow Bottom Hole Circulating Pressure (BHCP) Window Minimum Gas Flow (steady-state conditions) Maximum Gas Flow (steady-state conditions) Steady-State Region Non-Steady-State Region The curve shown in Figure 3 is developed for a given liquid and gas injection rate by calculating the anticipated wellhead pressure (dependent variable) for a given productivity index over a range of bottom hole circulating pressures (independent variable). One of the advantages of using this technique is that only one iteration is required per point on the curve, since bottom hole pressures are pre-determined and calculations are performed in the direction of flow. The region defined in Figure 3 as the Non-Steady-State region warrants closer investigation. All of points in this region represent technically valid solutions, however, not in the context of steady-state conditions with a fixed wellhead pressure. It is physically impossible to operate in this region without purposely manipulating the wellhead pressure up and down on a continuous basis to maintain a desired bottom hole circulating pressure. Simply maintaining a constant WHP will result in the well being killed (above the curve) or increasing the drawdown and corresponding gas rate (below the curve. Trying to maintain a wellhead pressure above the Maximum WHP as defined by the curve will result in killing the well. IADC/SPE 91220 3 An exception to this is when there is a significant pressure drop across the choke. In this instance, it is possible to achieve steady-state conditions with a fixed choke position (not to be confused with a fixed WHP). In this region, closing the choke will result in a decrease in WHP, while opening the choke will result in an increase in WHP. This would appear to defy the laws of physics; however, it is this unique phenomenon that defines this operating regime. By closing the choke, the bottom hole pressure is increased, resulting in a reduced gas inflow rate and a corresponding decrease in WHP. The reverse is true for opening the choke. Therefore, a fixed choke position will allow the well conditions to stabilize. The degree of pressure drop required across the choke to allow stabilization is the subject of future study. Due to the compressible nature of a multi-phase system and the associated lag time, it is not a simple and straightforward process to evaluate the impact of wellhead pressure manipulation on bottom hole circulating pressure. This can lead to oscillations in bottom hole circulating pressure and resultant oscillation in gas inflow, all resulting in aggravated slugging conditions, making the well more difficult to control. By comparison, when operating in the Steady-State region, it is possible to fix the wellhead pressure and allow well conditions to stabilize. In this region, closing the choke results in a decrease in gas inflow and a corresponding increase in WHP. Opening the choke results in an increase in gas inflow and a corresponding decrease in WHP. If a desired bottom hole circulating pressure is being targeted, wellhead pressure may be manipulated prior to reaching stabilized conditions provided that the MaximumWellhead Pressure value is not exceeded. Where this value is not exactly known (and it normally isnt), caution should be exercised to avoid killing the well. Pre-job engineering taking into consideration the range of productivity expected will go a long way to understanding the practical operating limitations that may be encountered and how to effectively deal with them. Analysis of these curves yields surface equipment boundary conditions for both pressure and flowrate, allowing for informed engineering decisions on equipment choice. Furthermore, understanding the potential for the existence of steady-state and non- steady-state behaviour is key to minimizing the potential for fighting the well. A sensitivity analysis should be performed to cover the likely range of productivity anticipated and a range of liquid and gas injection rates. This is an important consideration, especially when using steady-state models to simulate dynamic conditions, such as those experienced in drilling. It is also significant to understand that it is not feasible to accurately characterize the reservoir in a multi-phase model, thus sensitivity analysis is key to understanding practical operating environments. This technique is quite flexible for performing sensitivity analyses, whereas the more conventional tubing performance curves that include productivity make it quite difficult to incorporate the fact that wellhead pressure is likely to increase with increasing flow rate. The curves are, therefore, typically unrealistic and confusing. Sensitivity Analysis This approach typically requires analysis of two distinct sensitivities: (1) varied productivity and (2) gas-liquid ratio (injection). The range of productivity expected in conjunction with the tubing performance (related to geometry) determines the existence and shape/size of the non-steady-state region. A minimum PI with a slope greater than the slope of the tubing performance curve is required to yield this phenomenon (see Figure 4). The second sensitivity analysis is derived through manipulation of the gas-liquid ratio. This is one of the most powerful features of the use of this technique, since this can be used to determine the most feasible approach to minimizing flaring or maximizing gas recovery to process. Figure 5 illustrates how manipulation of the liquid and gas injection rates can be performed on a well with no inflow. This is, effectively, the means for determination of the threshold WHP for gas influx from the reservoir as highlighted in Figure 3. The determination of the maximum WHP is also affected by the liquid and gas injection rates, however, it is also heavily influenced by the productivity index, particularly in more prolific wells. This relationship is useful in the case where gas recovery to process is desired, since it represents the highest wellhead pressure that is achievable for a given drawdown. Once gas influx is seen from the reservoir, the wellhead pressure can only increase for the same injection and drawdown parameters. It may be appropriate at this point to modify the liquid and/or gas injection rates accordingly, should there be some excess wellhead pressure available in the system. Alternatively, a minimum gas injection rate strategy could be utilized until such time as sufficient reservoir influx is achieved to facilitate recovery of the gas to process. At this time, the liquid and gas injection rates can be modified to suit. It is also important to recognize that hole cleaning and maximum motor throughput parameters must also be considered. This is particularly important at low liquid injection and high gas injection rates. Finally, the shape of wellhead pressure vs. bottom hole pressure curve is also heavily influenced by wellbore geometry. Figures 6, 7 and 8 show three different wellbore geometries with identical threshold WHPs. Figure 6 was developed for 5 casing with 2 drill pipe. Figure 7 has 7 casing and 3 drill pipe and Figure 8 uses 9 casing and 5 drill pipe. As can be seen from these figures, the range of bottom hole circulating pressures that can be achieved in steady-state conditions decreases with decreasing hole size. This phenomenon is analogous to becoming tubing limited in tubing performance terminology and is a function of increased friction in the smaller hole geometries. 4 IADC/SPE 91220 Practical Application Strategies The application of this approach to underbalanced well design is focused on two outcomes: (1) minimizing flaring, and (2) maximizing gas recovery. In the first case, injection parameters are adjusted to minimize the amount of gas to be flared. Figures 9 and 10 demonstrate the effects of varying the liquid and gas injection rates. Figure 9 shows the total gas out, while Figure 10 takes injection gas out and simply shows net gas (gas from reservoir). There are several considerations to weigh in determining the most effective injection rate scenario. The typical approach involves minimizing wellhead pressure and minimizing gas injection rates. In the case of using cryogenic nitrogen as the injection source, this may well be a sound approach. However, in the case of using membrane generated nitrogen or natural gas (from a high-pressure supply), the selection of the high liquid rate case appears less attractive. Consideration should be given to the fact that the lower liquid rates will be more stable and less susceptible to slugging due to transient effects (e.g., connections). Also consider that hole cleaning may not be an issue in a prolific gas well. Low liquid rates also decrease the risk of going overbalanced. With the availability of high-pressure separation, a lower liquid injection rate coupled with higher wellhead pressure will further serve to reduce erosion potential by lowering fluid velocities and removing the majority of the energy (the gas) from the system under high pressure at the front end of the system. The use of high wellhead pressures is also not without peril. Consideration must be given to the rotating control device and the strategy for pressure drops to minimize potential for hydrate formation. Selecting the optimum liquid and gas injection rates must, therefore, take into consideration the limitations and capabilities of the equipment to be used for the project. In the case where the goal is maximizing gas recovery, the same approach should be taken. The ability to design purpose-built equipment for every unique job would certainly simplify the recommended approach; however, this is rarely the case in practice. Therefore, it is important to consider the availability and practical operating ranges of any compression equipment and/or separation equipment to be used. In general terms, it is generally more practical for gas recovery to eliminate the requirement for gas compression on the downstream side. Since compression equipment is positive- displacement and the gas out will be transient, it is likely that a certain percentage of gas will need to be flared using this strategy. Depending upon equipment availability and ratings, this may still yield the most practical solution. However, given the considerations discussed above and illustrated in Figures 9 and 10, there may still be merit to using a lower liquid rate than would normally be considered. An example of the options will be included in the following case studies. Case Studies The following three case studies will illustrate the use of this technique to understand well performance and optimize underbalanced well design. The first study demonstrates a real world example of attempting to operate a well in the non steady-state region, prior to the development of the wellhead operability curve. The second study shows how the technique was used to maximise deliverable surface gas pressure to process ultimately minimising auxiliary gas recompression equipment. The third and final study compares recoverable gas pressures from two popular UB methods for gas injection; drill pipe and concentric casing. The case studies represent global projects, and as such units of measurement vary from study to study. Case Study 1 Flow Drilling and Controlling on Gas Rate The first case study illustrates the challenges with operating in the Non-Steady-State region. This well is a normally- pressured sandstone reservoir drilled underbalanced using diesel as the drilling fluid medium. Figure 11 illustrates the wellbore geometry. The primary objective for drilling this well underbalanced was to decrease the overall drilling and completions cost of, and potentially add value to, development wells in the area. Typically, vertical wells have been used to develop this reservoir. If wells produce at low rates, frac stimulation has been used very successfully to increase production. Horizontal wells are widely used throughout the world for development projects due to the increase in reservoir exposure. The main concern with horizontal drilling in this area is the high cost for well stimulation for those wells that do not encounter fractures while drilling. Horizontal wells utilizing UBD techniques have the potential for increasing rates of penetration, bit life and reservoir exposure while reducing formation damage. Figure 12 shows the wellhead pressure vs. bottom hole circulating pressure plots for the well at 3291m and at 3475m. The client was interested in maintaining a maximum gas rate to flare of 10 MMscfd. As can be seen in Figure 12, this means that the well was operated in the Non-Steady-State region with associated operational challenges. As drilling continued and inflow increased, the well became increasingly difficult to manage. Therefore, it was decided to increase the maximum allowable inflow to 12 MMscfd. This resulted in improved operability, however, still required continuous choke manipulation to maintain conditions as stable as possible. Figure 13 illustrates a hybrid view of the same effect at 3291m. This plot highlights the fact that greater stability could be achieved at gas inflow rates between 14 16 MMscfd at a maximum wellhead pressure of 1200 psi. Lower gas inflow rates would mandate continuous choke manipulation to maintain a constant rate. IADC/SPE 91220 5 Figures 14 and 15 illustrate the real-time data plots acquired while drilling, including wellhead pressure, bottom hole circulating pressure and gas rate. Figure 14 shows the conditions around a depth of 3475m, while Figure 15 illustrates the change from controlling the well at 10 MMscfd to 12 MMscfd. These plots show a classic damping profile resulting from an experienced choke operator getting a feel for the well and how to control it. It is important to note that simply increasing wellhead pressure would not have facilitated control of the well. In this case, since the gas was being flared, leaving the choke position fixed would have resulted in stabilizing the well conditions. However, this may have resulted in an increase in the amplitude of the wellhead pressure fluctuation, which could unnerve the choke operator, especially if real-time bottom hole circulating pressure was not available. An alternative strategy to stabilize well conditions would have been to increase the gas rate to approximately 16 MMscfd at 3291m and approximately 20 MMscfd at 3475m. Some benefit may have also been derived through decreasing the liquid rate and increasing the wellhead pressure, however, as can be seen in Figure 16, the benefit is marginal. Figure 17 shows a closer look at the operating conditions highlighted in Figure 12 at 3291m. Points 1 and 3 denote where the choke operator intervened and started to increase and decrease wellhead pressure. Points 2 and 4 denote the apex points where the bottom hole pressure starts to react. The fact that point 2 is below the curve and point 4 above is a function of the lag time effect, which is difficult to account for in steady-state curves. However, note that the intersection of these 4 points lies on the curve (point 5) and corresponds to the gas rate being controlled. Most UBD wells are controlled on bottom hole pressure, however, this well has demonstrated that it is also possible to control on gas rate, although appropriate filtering is required to smooth out the trending. One of the challenges associated with using bottom hole pressure is that the data resolution is very sparse (updates often every 30 seconds to 1 minute). There is also no data available during connections when mud pulse telemetry is used and downhole tools are more susceptible to failures that cannot be remedied until a trip is performed. Real-time gas rate, in contrast can yield filtered (averaged) data at resolutions as good as every second, thus yielding more reliable feedback. However, caution must be exercised when operating in the Non-Steady-State region or near the apex of the wellhead pressure vs. bottom hole pressure curve. Should well control become challenging, the simplest solution is to allow the gas rate to increase until the well is easier to control. Understanding the approximate operating point on the curve will assist in fine tuning an appropriate operating point. Development of the operational curve, when altered to reflect actual well parameters, provides onsite personnel with a map outlining upper and lower boundary conditions for pressure and flow that can be expected while drilling. Case Study 2 Underbalanced Drilling in a Gas Storage Reservoir The second case study involves drilling horizontal wells in a gas storage reservoir. The properties that make this reservoir a good candidate for gas storage (i.e., high porosity and permeability) also make it a good candidate for drilling challenges, such as loss of circulation and differential sticking. In addition, formation damage may play a role in the energy required to inject and retrieve gas. Over many cycles, this may become a significant economic factor. Finally, the ability to minimize formation damage could serve to reduce the required drawdown and thus, reduce the risk of producing underlying water. The use of stimulation techniques were seen to increase the risk of water production by improving the permeability of the existing path of least resistance, while having little effect on the lower permeability parts of the reservoir that have incurred the greatest drilling induced formation damage. By contrast, successful implementation of UBD technology would prevent the majority of drilling induced formation damage and give the reservoir the best chance of being productive over the greatest percentage of the borehole. In most underbalanced well designs, the reservoir pressure is one of the key parameters. This particular project has the atypical property of having a reservoir pressure that varies as a function of the time of year and how the seasonal weather has impacted demand. In this case, the reservoir pressure typically ranges from a low of 100 Bar to a maximum of 200 Bar. Another feature typical of gas storage reservoirs is that they are normally located in reasonably close proximity to population centres and, as such, significant flaring of gas is not desirable. In addition, any gas flared has intrinsic value, as it was purchased and injected into the storage reservoir. It is, therefore, preferred that as much of the gas as possible is recovered and re-injected into the reservoir or gas gathering system. The three process flow diagrams (Figures 18, 19 and 20) illustrate the three options for drilling that were considered. Figure 18 shows Option 1 the base option to use nitrogen or natural gas as the injected gas medium and simply flaring all the effluent. Figure 19 represents Option 2 the basic configuration required to compress and inject natural gas (available at 40 bar) down the drill string and use a separate compression train to compress the effluent gas to the flow line pressure of 40 bar. Any excess gas would be sent to flare. The amount of gas flared would be a function of the capacity range of the downstream compression equipment. In the event of a gas plant upset, all gas could be directed to flare. Figure 20 illustrates Option 3 the configuration using compression equipment on the injection side only and recovering the majority of the gas to the gas flow line via high pressure separation (> 40 Bar). In this configuration, only minor gas carried over to the low-stage separation would be 6 IADC/SPE 91220 flared. Of course, in the case of a gas plant upset, all gas could be flared as required. Due to the high gas volumes required to drill the 8 open hole, an alternative completion configuration was proposed with a view to reducing costs. Figure 21 shows the original proposed configuration with 8 open hole and the alternative configuration with 6 open hole, however, maintaining the desired 7 completion. The alternative configuration was also attractive in facilitating the use of a 7 downhole isolation valve (DIV), which would allow the deployment of a slotted liner without having to kill the well. The well design strategy looked at minimizing the amount of gas that would need to be flared as well as maximizing the wellhead pressure that could be maintained in order to ensure all of the gas could be separated and sent to process at approximately 40 Bar. Figures 22 and 23 illustrate the sensitivity analysis of the productivity for the 8 and 6 open hole cases respectively. These plots focus on the maximum wellhead pressure and form the basis for the recommendation that the wells be drilled when there is a reservoir pressure greater than 120 125 Bar if gas recovery is targeted. Looking at the PI = 0 curve highlights the minimum bottom hole pressure required to maintain sufficient wellhead pressure to separate the gas and send it to process at 40 Bar. This means that no flaring will be required even when there is no production from the reservoir. In essence, this design approach serves to desensitize the system from the variability of the reservoir properties. As production from the reservoir is realized, the injection rates of the gas and liquid may be altered as desired to optimize the operating ranges of the equipment. By progressing through the well design process, equipment specifications were developed to cover the three options illustrated in Figures 18, 19 and 20. These were further split into Options 1a and 1b, 2a and 2b, and 3a and 3b. The a denotes the 8 open hole configuration while the b denotes the alternative 6 open hole configuration. The key differences between the six options examined were the required gas injection rates, the injection compression duty and the effluent compression duty. A summary of the results is included in Figure 24. This clearly shows Option 1 as the simplest solution, however, this is also the highest cost solution when the cost of the flared natural gas is taken into consideration. While there is little to choose between Options 2 & 3 in terms of total cost, Option 3 clearly represents the most elegant solution. The costs are weighted quite heavily towards the number of compressors required due to equipment availability and subsequent high mobilization charges. However, the availability of equipment must be taken into account and the following table highlights the number of compressors required and ranks the six options in ascending order according to cost.
This table serves to highlight the impact of equipment availability and ratings upon ultimate cost. In the alternative hole geometry, Option 2 (b) ends up being more cost-effective than Option 3 (b) due to the absence of the high-pressure separation module, even though the number of compressors is equal. However, it should be noted that Option 2 is still subject to variations in effluent rate that could result in more gas flaring. Therefore, Option 3 may still end up being more cost-effective if risk were included in the calculations. Option 3 also has the decided advantage of both compressors being used for the same purpose (injection only) and therefore, operational efficiencies in the use of compressors would be expected. In addition, injection rates are predictable, stable and controllable, whereas effluent rates are unpredictable and add complexity to the operation. The use of the wellhead operability curve in this case outlined the solution to efficiently capture gas from a storage reservoir, balancing minimum line pack pressure with economy of surface equipment against variable reservoir pressure. Case Study 3 Concentric Injection vs. Drill String Injection The third case study looks at a couple of wells where gas recovery to process is desired. The first well flows to a low- pressure trunk line at 90 130 psi. Concentric gas injection is often used to facilitate the use of mud pulse MWD systems that do not perform well under high gas volume fraction environments. The second well is a high-pressure well that flows to a high-pressure trunk line at 600 1000 psi. This high-pressure well typically only uses drill string gas injection until flow from the reservoir is achieved. Figure 25 shows the range of wellhead pressures achievable for the low-pressure well in concentric injection and drill string injection scenarios. While gas recovery appears to be marginally feasible under the concentric gas injection scenario, the operating window becomes significantly larger under a drill string injection scenario. In addition, while the concentric casing injection case requires 3000 scfm gas injection, drill string injection can be accomplished with a IADC/SPE 91220 7 third or less of this gas volume. A higher drawdown can be achieved while reducing the amount of gas to surface. This will have the effect of reducing the potential for erosion or permitting the use of smaller line sizes. Consideration must also be given to the potential for slugging when using concentric injection, due to the potential for U- tubing in the concentric annulus. This is caused by liquid entering the concentric annulus until sufficient gas pressure is built up to reverse the flow back into the primary annulus. Mykytiw et al discusses methodology to mitigate this phenomenon by determining the critical gas injection rate required. 1 In Figure 26, some reduced liquid injection rates are evaluated with a view to separating the gas with a high-pressure separator and recovering directly to process. This illustrates that to ensure all injected gas is recovered to process at PI = 0, a gas injection rate approaching 3500 scfm is required. However, as the well starts to produce gas, this injection rate can be reduced to approximately 500 scfm (for the PI assumed) and gas recovery maintained. This approach removes the uncertainty over PI and ensures the ability to recover all gas to process. Although 3500 scfm appears to be a significant gas injection rate, if the gas is taken from the same process system at 1000 psi and compressed to the desired injection pressure of 4000 psi, only 450 BHP (335 kW) of compression power is required. This compares quite favourably to 2800 BHP (2100 kW) of power required for a membrane generated nitrogen system. It should be noted that these higher gas injection rates will not permit the use of MWD systems using mud pulse telemetry, but rather would require the use of electromagnetic or hard-wire telemetry, as these are not affected by the higher gas volume fractions inside the drill string. Significant progress has been made in the range of implementation of EM-MWD technology. 2 It still represents the data telemetry technology of choice for underbalanced drilling operations. This example highlights the benefits of drill string injection for optimizing the potential for gas recovery while managing hole cleaning and motor performance requirements. The reduction of the liquid injection rate also reduces the impact of transient effects such as those caused by making connections or encountering increased gas production from the reservoir. Conclusions The availability of higher spec equipment such as high- pressure rotating control devices and high-pressure separators has highlighted a need for an updated approach to detailed underbalanced well design. Higher levels of optimization are now possible in the areas of gas recovery and erosion control that will make underbalanced drilling safer and more cost- effective. This updated approach to underbalanced well design offers a simpler approach to performing sensitivity analysis on varying well productivity and liquid-gas injection rates. This method should be used in conjunction with more conventional underbalanced well design techniques, however, it provides an improved presentation for evaluation of the operating conditions as a function of the independent controllable variables wellhead pressure and liquid and gas injection rates. The use of this evaluation method highlights the potential for considering lower liquid injection rates with corresponding higher gas injection rates in order to optimize the ability to recover gas to process. Using a higher gas-volume fraction also provides the benefit of expanding the operating envelope and reducing the impact of transient effects caused by interruptions in circulation (e.g., connections) or encountering a sudden increase in gas influx from the reservoir. This technique also provides an elegant means to understand operating conditions using standard steady-state flow models. This understanding and the ability to more accurately predict operating regimes will improve the feasibility of developing more automated choke control systems that will improve the stability of operating conditions and reduce the risk of accidentally killing the well. More work is required with dynamic models in order to bridge the cause and effect relationships between steady-state operating conditions and transient effects. This will promote a better understanding of transient behaviour and minimize the need for a high degree of accuracy when using models in a real-time environment. References 1. C.G. Mykytiw, I.A. Davidson and P.J. Frink, Design and Operational Considerations to Maintain Underbalanced Conditions with Concentric Casing Injection, paper IADC/SPE 81631 presented at IADC/SPE Underbalanced Technology Conference and Exhibition, Houston, 25 26 March 2003. 2. P. Brett, D. Weisbeck and R. Graham, Innovative Technology Advances Use of Electromagnetic MWD Offshore in Southern North Sea, paper IADC/SPE 81628 presented at IADC/SPE Underbalanced Technology Conference and Exhibition, Houston, 25 26 March 2003. 8 IADC/SPE 91220 Nomenclature UBD = Underbalanced Drilling Gas Production Rate B o t t o m
H o l e
C i r c u l a t i n g
P r e s s u r e
m In c r e a s in g L iq u id R a t e o r W H P Non-steady-state Steady-state M BHCP Range 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 3000 3500 4000 BHCP (psi) W H P ( p s i) 0 500 1000 1500 2000 2500 G a s In je c t io n R a t e ( s c f m ) 50gpm/ 1800 scfm 100 gpm/ 1700scfm 150 gpm/ 1500scfm 200 gpm/ 1000scfm Reservoir Pressure BHCP = Bottom Hole Circulating Pressure PI = Productivity Index IPR = Inflow Performance Relationship WHP = Wellhead Pressure MMscfd = Million Standard Cubic Feet per Day DIV = Downhole Isolation Valve MWD = Measurement While Drilling Scfm = Standard Cubic Feet per Minute BHP = Brake Horsepower EM = Electromagnetic Acknowledgements The Authors wish to express their appreciation to Northland Energy, a division of Precision Drilling Technology Services, for their permission to publish this paper.
0 500 1000 1500 2000 2500 3000 3500 0 1 2 3 4 5 6 Gas Flow Rate (MMscfd) B H C P ( p s i) 7 Max Motor Throughput Min AVliq (hole cleaning) Est. Reservoir Pressure Target BHCP Min BHCP 50 gpm 100gpm 150gpm 200 gpm 250 gpm Operating Envelope Figure 1: Conventional UB Well Engineering
Figure 2: Tubing Performance with IPR Overlay Operating Envelope Analysis BHCP W H P G a s
I n f l o w
( I n j e c t e d
p l u s
P r o d u c e d )
WHP vs BHCP Gas Inflow Max Gas Inflow Non-Steady State Steady State Min Gas Inflow Threshold WHP to Initiate Gas Inflow Maximum WHP for Steady State Operating Conditions Steady State Min BHCP BHCP Window Figure 3: Operating Envelope Analysis Technique
Reservoir Inflow B H C P PI = 1 PI = 2 PI = 3 PI = 5 PI = 10 PI = 20 Slope of PI > Slope of Tbg Performance Curve Slope of PI < Slope of Tbg Performance Curve
Figure 4: Slope of Tubing Performance vs. PI
Figure 5: Manipulation of Liquid and Gas Injection IADC/SPE 91220 9
0 250 500 750 1000 500 1000 1500 2000 2500 3000 3500 BHCP (psi) W H P ( p s i) 0 10 20 30 40 T o t a l G a s R a t e ( M M s c f d ) PI =0 PI =3 PI =5 PI =10 PI =20 Figure 6: PI Sensitivity Analysis in 5 Casing
0 250 500 750 1000 500 1000 1500 2000 2500 3000 3500 BHCP (psi) W H P ( p s i) 0 10 20 30 40 T o t a l G a s R a t e ( M M s c f d ) PI =0 PI =3 PI =5 PI =10 PI =20 Figure 7: PI Sensitivity Analysis in 7 Casing
0 250 500 750 1000 500 1000 1500 2000 2500 3000 3500 BHCP (psi) W H P ( p s i) 0 10 20 30 40 T o t a l G a s R a t e ( M M s c f d ) PI =0 PI =3 PI =5 PI =10 PI =20
Figure 8: PI Sensitivity Analysis in 9 Casing
0 250 500 750 1000 1250 1500 1750 1250 1500 1750 2000 2250 2500 2750 3000 BHCP (psi) W H P ( p s i) 0 5 10 15 20 25 30 35 T o t a l G a s R a t e ( M M s c f d ) 50gpm/ 3000scfm 100 gpm/ 1500scfm 150 gpm/ 1000scfm 200 gpm/ 500scfm
Figure 9: Gas-Liquid Rate Sensitivity Analysis Total Gas
0 250 500 750 1000 1250 1500 1750 1250 1500 1750 2000 2250 2500 2750 3000 BHCP (psi) W H P ( p s i) 0 5 10 15 20 25 30 35 N e t G a s R a t e ( M M s c f d ) 50gpm/ 3000scfm 100 gpm/ 1500scfm 150 gpm/ 1000scfm 200 gpm/ 500scfm
Figure 10: Gas-Liquid Rate Sensitivity Analysis Net Gas
Bottom 7 Liner (3064 mMD) 6 Open Hole 600 m A p p r o x .
2 8 4 0 m
t o
2 8 6 5
m
T V D
TMD = 3664 m 5 DP & 3 DP KOP 2037m 2-5 o /30m 4 Turbine Gas BHP = 3569 psi BHT = 242 deg F (static) Sandstone Fluid: Diesel Resevoir Static Pressure: 3569psi Target BHCP: 3369 psi
9 5/8 Casing @ 2662 m MD
Top 7 Liner (2502 m MD) Figure 11: Wellbore Geometry Case Study 1 10 IADC/SPE 91220
0 250 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 BHCP (psi) 0 10 500 750 1000 1250 1500 W H P
( p s i ) 20 30 40 50 60 T o t a l G a s
R a t e
( M M s c f d ) WHP @ 3291m WHP @ 3475m Gas Rate @ 3291m Gas Rate @ 3475m 12 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 5 10 15 20 25 30 35 40 Reservoir Inflow (MMscfd) B H C P
( p s i ) 1200 psi WHP 800 psi 1000 psi 0 500 1000 1500 2000 2500 3000 3500 4000 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00 Time P r e s s u r e
( p s i ) 0 2 4 6 8 10 12 14 16 G a s R a t e
( M M s c f d ) BHCP WHP Gas Rate 0 500 1000 1500 2000 2500 3000 3500 4000 7/4/2003 12:00 8/4/2003 00:00 8/4/2003 12:00 9/4/200300:00 9/4/200312:00 10/4/2003 00:00 Time P r e s s u r e ( p s i ) 0 2 4 6 8 10 12 14 16 G a s
R a t e
( M M s c f d ) BHCP WHP Gas Rate
Figure 12: Operating Points @ 3291m and 3475m
Figure 13: Tubing Performance with IPR Overlay @ 3291m
Figure 14: Real-Time Data @ 3425m 3475m
Figure 15: Changing Controlled Gas Influx Rate
0 250 500 750 1000 1250 1500 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 BHCP (psi) W H P ( p s i ) 0 10 20 30 40 50 60 T o t a l G a s
R a t e ( M M s c f d ) WHP @ 285 gpm WHP @ 200 gpm Gas Rate @ 285 gpm Gas Rate @ 200 gpm 12
Figure 16: Effect of Decreasing Liquid Injection
Figure 17: Choke Operator Intervention Analysis 1000 1050 1100 1150 1200 1250 1300 3300 3325 3350 3375 3400 3425 3450 BHCP (psi) W H P ( p s i) 0 5 10 15 20 25 30 T o t a l G a s R a t e
( M M s c f d ) WHP @ 3291m Gas Rate @ 3291m 3 4 1 2 5 IADC/SPE 91220 11
Figure 18: Low-Pressure Separation & Flaring
Figure 19: Low-Pressure Separation & Recovery
Figure 20: High-Pressure Separation & Recovery
Figure 21: Original Configuration & Alternative
Figure 22: PI Sensitivity Analysis 8 Open Hole
Figure 23: PI Sensitivity Analysis 6 Open Hole
Choke Manifold 4-Phase Separator (low stage) Liquid Knockout Gauge Tank DAT Lab CPD Package Natural Gas Supply (@ 40 bar) Solids Liquid Well Mud Handling System Injection Compression
Gas to Flare 2-Phase Separator (hi stage) To Gas Flow Line (40 bar) Surf ace Facilities: Opt ion 3 Choke Manifold 4-Phase Separator Effluent Gas Compression Gauge Tank DAT Lab CPD Package Natural Gas Supply (@ 40 bar) Solids Liquid Well Mud Handling System Injection Compression Excess Gas to Flare To Gas Flow Line (40 bar) Surf ace Faci lities: Option Surface Facil ities: Option 2 2 N2 or Gas from Choke Manifold 4 Phase Separator Gauge Tank DAT Lab CPD Package Pipeline Solids Liquid Well Mud Handling System Injection Compression Surface Facilit ies: Option 1 Surface Facilities: Opt ion 1 Gas to Flare 2970m TVD 7 slotted liner in 8 open hole 2914m TVD 9 / 10 casing in 12 open hole 2469m TVD 13 casing in 17 open hole 880m TVD 100m TVD 18 casing in 23 open hole 2970m TVD 4 5 slotted liner in 6 open hole 2914m TVD 7 liner in 8 open hole 2469m TVD 9 / 10 casing in 12 open hole 880m TVD 100m TVD 13 casing in 17 open hole 7 tubing 7 tubing (7 drilling liner with DIV during drilling) 2270m TVD 2713m TVD 2713m TVD Operating Envelope PI Sensitivity Analysis in 8.5" O.H. with Qliq = 450 lpm / Qgas - 260 m 3 /min Pres = 150 bar 0 10 20 30 40 50 60 70 80 80 85 90 95 100 105 110 115 120 125 130 135 140 145 150 155 BHCP (bar) W H P
( b a r ) 0 10000 20000 30000 40000 50000 60000 70000 80000 G a s R a t e ( In f lo w + In je c t e d ) [ m 3 /h r ] WHP vs BHCP @ PI = 0 m3/hr/bar/m WHP vs BHCP @ PI = 25 WHP vs BHCP @ PI = 7 WHP vs BHCP @ PI = 3.5 Total Gas Rate @ PI = 0 m3/hr/bar/m Total Gas Rate @ PI = 25 Total Gas Rate @ PI = 7 Total Gas Rate @ PI = 3.5 Operating Envelope PI Sensitivity Analysis in 6" O.H. with Qliq = 230 lpm / Qgas - 125 m 3 /min Pres = 150 bar 0 10 20 30 40 50 60 70 80 85 90 95 100 105 110 115 120 125 130 135 140 145 150 155 BHCP (bar) W H P
( b a r ) 0 10000 20000 30000 40000 50000 60000 70000 80000 G a s R a t e ( In f lo w + In je c t e d ) [ m 3 /h r ] WHP vs BHCP @ PI = 0 m3/hr/bar/m WHP vs BHCP @ PI = 25 WHP vs BHCP @ PI = 7 WHP vs BHCP @ PI = 3.5 Total Gas Rate @PI = 0 m3/hr/bar/m Total Gas Rate @PI = 25 Total Gas Rate @PI = 7 Total Gas Rate @PI = 3.5 12 IADC/SPE 91220
Figure 24: Summary of Gas Injection & Compression Duty Requirements
Operating Envelope Analysis Pres = 2170 psi / PI = 0.017 Mscfd/psi/ft 0 100 200 300 400 500 600 700 800 900 1000 1700 1800 1900 2000 2100 2200 2300 BHCP (psi) W H P
( p s i ) 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 G a s R a t e ( In f lo w + In je c t e d ) [ M M s c f d ] WHP vs BHCP - 150gpm / 1000 scfm WHP vs BHCP - 120 gpm / 2000 scfm WHP vs BHCP - 90 gpm / 3400 scfm WHP vs BHCP - 70 gpm / 4500 scfm WHP vs BHCP - 200 gpm / 3000 scfm cc injection Total Gas Rate - 150 gpm / 1000 scfm Total Gas Rate - 120 gpm / 2000 scfm Total Gas Rate - 90 gpm / 3400 scfm Total Gas Rate - 70 gpm / 4500 scfm Total Gas Rate - 200 gpm / 3000 scfm cc injection Reservoir Pressure
Figure 25: Concentric Injection vs. Drill String Injection
Operating Envelope Analysis Pres = 3600 psi, PI - 0.02 Mscfd/psi/ft 0 200 400 600 800 1000 1200 1400 1600 1800 2000 3000 3100 3200 3300 3400 3500 3600 3700 3800 3900 4000 BHCP (psi) W H P
( p s i ) 0 5 10 15 20 25 30 35 40 45 50 G a s R a t e ( In f lo w + In je c t e d ) [ M M s c f d ] WHP vs BHCP - 120 gpm / 3500 scfm - PI=0 WHP vs BHCP - 120 gpm / 3500 scfm WHP vs BHCP - 150 gpm / 1600 scfm WHP vs BHCP - 180 gpm / 500 scfm Total Gas Rate - 120 gpm / 3500 scfm - PI=0 Total Gas Rate - 120 gpm / 3500 scfm Total Gas Rate - 150 gpm / 1600 scfm Total Gas Rate - 180 gpm / 500 scfm Reservoir Pressure