04 - SPE 63197 - Shallow Gas Kick

Embed Size (px)

Citation preview

SPE 63197 Shallow Gas Kick: Simulation and Analysis for Top Hole Drilling Without a RiserCraig Marken, SPE, RF-Rogaland Research, Svein Hansen, SPE, Petec Software and Services A.S., and Jakup * regaard, SPE, Saga Petroleum A.S .

Copyright 2000, Society of Petroleum Engineers Inc. This paper was prepared for presentation at the 2000 SPE Annual Technical Conference and Exhibition held in Dallas, Texas, 14 October 2000. This paper was selected for presentation by an SPE 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 Society of Petroleum Engineers 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, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers 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. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Although the pilot hole reduces the potential for a shallow gas kick, its presence, even with careful attention to drilling operations and practices, can lead to a false sense of security. This simulator has been a valuable tool for safety analysis and planning of top hole drilling operation where shallow gas zones are a part of the formation prognosis. Introduction Top hole drilling operations without a riser at an offshore location are susceptible to shallow fluid flows, which can not be controlled by shutting in the well. When the fluid originates from a shallow gas zone, safety for the operational personnel and rig become a paramount concern. Controlling and preventing the shallow gas influx has been shown to be one of the more complex and difficult well control challenges during drilling operations [1]. There have been several approaches put forward in addressing the challenges of shallow gas kicks. Operational and planning checks have addressed drilling procedures, well design and rig systems [2]. In this case a simulator model was used for evaluation of the time for a blowout to reach the rotary table and the variation of gas flow with formation wash out. A simulator has also been used to evaluate the performance of diverters during blowouts resulting from shallow gas kicks [3]. Additionally, specialized hardware has been developed. For example there have been demonstrations that a drillstring mounted downhole blowout preventer has been effective for control in replicated gas kicks [4]. Some regulatory agencies, for example the Norwegian Petroleum Directorate [5] have addressed this topic of shallow gas kicks by requiring the drilling of a narrow pilot hole prior to opening to the size suitable to the upper casing program. The perception has been that the smaller hole has a narrower annulus that is beneficial for controlling the kick with dynamic circulating pressure losses. In conjunction with a project assessing the potential of a slender well design, it became important to evaluate the issue of using a pilot hole for safe prevention of a shallow gas kick when entering a shallow gas zone. To do this the integrated drilling software package, Drillbench [6], containing among other application both a kick simulator and an underbalanced drilling simulator [7,8], was used to evaluate several variables and scenarios associated with drilling top hole well sections

Abstract The influx of a shallow gas kick during top hole drilling operations without a riser can result in critical well control and safety problems for both the drilling rig and personnel. With appropriate well planning tools these top hole challenges and possible corrective actions can be addressed prior to and during these drilling operations. Based on modifications to an existing gas kick simulator package, such a well planning tool has been developed and demonstrated for the simulation and analysis of possible shallow gas kick scenarios. Although this tool was initially intended for an analysis of the potential for elimination of the pilot hole, the emphasis of this study became the safety limits of the pilot hole itself. The simulations show the threshold levels for uncontrolled blowouts as a function of gas zone formation permeability and gas pressure when using a 9 5/8 pilot hole. A comparison with a larger 17 1/2 hole has demonstrated the lack of shallow gas kick control at the same permeabilities and even lower gas pressures. Dynamic circulation control of a shallow gas kick as a function of time after penetration of the gas zone, irrespective of when a kick is detected, shows the advantages of the pilot hole. However these advantages must be tempered by the realization of a very short reaction time. The additional hydrostatic pressure resulting from suspended cuttings can make the difference between the control of the shallow gas and uncontrolled kick. After penetrating the gas zone circulating the cuttings out of the annulus to clean the hole can reduce the downhole pressure sufficiently to initiate a blowout while otherwise maintaining the circulation rate that would control the shallow gas.*

Now a part of Norsk Hydro

2

C. MARKEN, S. HANSEN, J. REGAARD

SPE 63197

with returns to the seabed. Although this simulator package would be applicable with a variety of drilling fluids, in this particular case seawater was used. Recently, some techniques for the use of gel and weighted fluids with returns to the seabed have been discussed for the control of shallow fluid flows [9]. Approach to shallow gas kick simulations The simulator analysis of shallow gas kicks with returns to the sea bed covered a range of variables such as formation permeability, the pressures of the gas zone, and drilling operations to evaluate the applicability of a pilot hole to drilling safety. In its present form the simulator package can simulate drilling operations over a broad range of conditions, and thereby lead to a rather complex analysis of shallow gas kicks. Although analyses at more that one depth of gas zone are beneficial to planning drilling operations, the demonstration of this simulator was limited to one depth in order to illustrate other drilling variables in more detail. This analysis focused on a gas zone that was suggested by seismic studies to be at 644 m depth with a water depth of 412 m. To limit this analysis the drillstring composition was held constant throughout this analysis. This drillstring consisted of 169 meters of 8 drillcollars on 5 drillpipe. Both the 9 7/8 pilot hole bit and the 17 1/2 bit were evaluated using this drillstring. At the beginning of most of the simulations the bit was at 640 meters and drilled the four meters into the top of the gas zone at 644 meters. This allowed some build-up of cuttings into the well annulus without making each simulation excessively long. Although the presence of the cuttings affects the apparent fluid (seawater) density, a steady state concentration of these cuttings was not fully achieved in the earliest simulations. This, in effect, would simulate a clean hole condition a few meters before entering the gas zone. The bottom of the gas zone was presumed to be at 660 m. At a penetration rate of 100 m/hr it would take 9.6 simulated minutes for the bit to pass through the gas zone. With the initial bit position four meters above this zone only 12 simulated minutes were required for the bit to be beneath the gas producing zone. In the one case where the drillcollars were passed through the zone, it took 104 simulated minutes for the drillpipe to be in the gas zone. In the later studies that simulated a steady state concentration of cuttings the initial bit depth was set at 610 m. A steady state concentration was achieved around 15 minutes of simulated time and the 640 m depth was reached in 18 minutes. In this study the variables that were evaluated were gas pressure, formation permeability, and seawater pump rate. In the pilot hole the formation gas pressure ranged from 68 bars, where there was no gas kick observed, up to 70.5 bars where an uncontrolled gas kick was simulated. In the 17 1/2 hole the gas pressure ranged from 67 bars (minor kick) to 68 bars (uncontrolled kick). The formation permeability ranged from 0.5 darcy to 2.0 darcy, both at 0.25 percent porosity. A higher permeability as anticipated in many gas formations would have made the simulated kicks much worse.

The gas used in these simulations had a density of 0.72 kg/m3 at standard pressure and temperature. For the control of a gas kick in progress in the pilot hole the pump rate was increased after a delay of 1.00 to 1.50 min after an initial kick. Note that these are times after an initial kick without any attempt to estimate the time required to detect such a kick either though a downhole pressure tool or by observing gas coming out of the well at the sea bed. As this study was more concerned with the initiation of the gas kick while drilling into the gas zone, it was not the goal of this study to evaluate the development to an uncontrolled well blowout. As described earlier [3] sand production and hole washout becomes important to the analysis of such a blowout. Albeit very critical to hole washout, sand production is not a part of the simulator package at the present time. The results of these simulations are illustrated through the use of bottom hole pressures at the drill bit. Although a driller may more easily interpret these pressures when reported in ECD, the units of bars have been retained for a comparison with the gas pressure in the gas producing zone. The shallow gas pressures in this study range from 67.0 to 72.5 bars, which correspond to an equivalent static density at 644 m of 1060 to 1147 kg/m3, respectively. However, all of the formation pressures and bottom hole pressures are reported in bars. Additionally, as the bottom hole pressures reflect the bit depth these pressures show an increase that is a result of increased hydrostatic head with drilling time. An uncontrolled kick shows a rapid decrease in this pressure, as hydrostatic head is lost. In many cases this hydrostatic head is regained when the gas leaves the well annulus before an uncontrolled kick can develop. Quite often a small influx of connection gas occurs at a drillpipe connection when circulation is temporarily stopped. This connection gas can flow from the annulus before developing into an uncontrolled kick. As in all modeling the interpretation of simulation results, in this case downhole pressure, should be tempered by the assumptions used. In this case several assumptions were made about the drilling conditions and formation. Therefore, these simulations are best interpreted as a warning as to what could be anticipated. Changing the magnitude of the variables could easily lead to situation where uncontrolled blowouts are achieved in most cases. However, such results could fail to illustrate some of the subtle details that can be anticipated during drilling operations and some of the potentials of this simulator. Formation permeability with the pilot hole The effects that the permeability of the gas producing formation has on the gas influx have been simulated at three different permeabilities. Figure 1 illustrates the variation in kick potential in a 9 7/8 pilot hole that is drilled into each of these three 644 meter deep gas zones. The simulated time shown in this and subsequent figures starts when the bit is at 640 meters. In addition, the resultant downhole pressures are shown without regard to implementing any kick control routines (as are many other simulations in this study). In this example all variables including the formation pressure of 70

SPE 63197

SHALLOW GAS KICK: SIMULATION AND ANALYSIS FOR TOP HOLE DRILLING WITHOUT A RISER

3

bars and a seawater circulation rate of 2000 l/min were held constant throughout each simulation. Although the simulation with the 0.5 darcy formation experienced three minor influxes, 1.2 kg of gas at the top of the gas zone, 2.0 kg at the first drillpipe connection and 0.6 kg at the second connection, none led to an uncontrolled situation even at constant pump rate. With the 1.0 darcy formation a slow but steady gas influx started at the top of the gas zone and continued until the first connection for a total of 9.2 kg of gas. With the connection the influx drastically increased and an uncontrolled kick was initiated. With the 2.0 darcy formation simulation an uncontrolled kick started when the bit entered the top of the gas zone. In additional simulations neither the 1.0 darcy nor the 2.0 darcy formation led to an uncontrolled kick at a lower formation pressure of 69.75 bars (the 1.0 darcy simulation illustrated in Figure 2). These examples illustrate the sensitivity of kick potential on a small formation pressure difference (0.25 bar), compared to a difference in formation permeability (1.0 and 2.0 darcy). However, the lower permeabilities give a lesser tendency to kick as clearly illustrated with the 0.5 darcy simulation.76 75 74 0.5 darcy Pressure (bars) 73 72

without remedial action. However, in this case there is some time for implementation of such corrective measures. 9 7/8 pilot hole shallow gas kick potential. Formation pressure effects. The range of the potential for shallow gas kicks has been simulated for the 9 7/8 pilot hole. Figure 2 illustrates several scenarios for this pilot hole in the 644 meter gas zone with 1.0 darcy permeability while circulating at 2000 l/min. The formation pressures cover a range of two bars. At 68.00 bars no gas influx was observed either when entering the gas zone or at the drillpipe connections. The changing slope in this curve illustrates the build-up of cuttings in the initial stages of this simulation. When the formation pressure was 69.00 bars a small influx of 1.3 kg of connection gas occurred at the first connection. This gas left the well during the period of the second connection. In a similar manner at a formation pressure of 69.75 bars an influx of 2.0 kg of connection gas occurred at the first connection. However, as discussed above, the 70.00 bar formation lead to an initial sustained kick when the gas zone was penetrated. This was followed by an uncontrolled kick at the first connection. These latter two simulations illustrate how the quarter bar pressure difference may result in very different and potentially dangerous results.76 68.00 bars 75 69.00 bars 74 Pressure (bars) 73 69.75 bars 72 71 70 69

71 70 1.0 darcy 69 2.0 darcy 68 67 0 5 10 15 20 25 Time (min)

70.00 bars 68 67 0 5 10 15 20 25 Time (min)

Figure 1. Kick simulations for the 9 7/8 pilot hole with 70 bar formation pressure and 2000 l/min circulation with varying formation permeability.

Several interesting features are illustrated in the simulations shown in Figure 1. First, some kicks occur in small enough quantity such that they are circulated out of the hole without affecting drilling operations. The downhole pressures simulated in the 0.5 darcy example show some loss in hydrostatic pressure as the connection gas rises after the first and second connections. Even though the pressure loss occurred at the subsequent connection, there is insufficient pressure loss to initiate an uncontrolled kick. Second, as in the 1.0 darcy simulation, a kick, which readily influences the bottom hole pressure, does not lead to an immediate uncontrolled kick. Although this influx formed an uncontrolled kick at the next connection, it is not certain that it would have remain controlled prior to the connection and

Figure 2. Kick simulations for the 9 7/8 pilot hole with 1.0 darcy formation and 2000 l/min circulation with varying formation pressure.

With the same pilot hole drilling conditions, however with the 0.5 darcy formation, the effects of different shallow gas pressures were also simulated. Figure 3 shows the results of these simulations over a two and a half bar pressure range. First, and not illustrated in this figure, is that a 68.00 bar formation reacts the same as with the 1.0 darcy example and does not kick. For the 69.00 bars and 69.50 bars formation pressure simulations an influx of 1.1 kg and 1.8 kg of connection gas, respectively, occurs at the first connection only. However, as described above, the 70.00 bars formation pressure is more complicated, albeit just as controllable

4

C. MARKEN, S. HANSEN, J. REGAARD

SPE 63197

without remedial action. The small 1.1 kg gas kick that arises when entering the 70.00 bars gas zone lasts for about one minute before stopping. This could be explained by increasing hydrostatic head as additional cuttings being suspended in the annulus. There could be a contribution from gas diffusing into the annulus in the immediate vicinity of the wellbore faster than it is replaced by gas deeper in the formation. Both the 2.0 kg and 0.6 kg gas influx at the first and second connections stop when circulation is reestablished. A comparison of the uncontrolled kick at 70.25 bars and 70.50 bars illustrate the faster influx of gas at the higher pressure. This, of course, would lead to shorter reaction time for controlling the kick.76 69.00 bars 74 69.50 bars

at the first connection. As this gas rose up the annulus the resultant loss of hydrostatic head lead to a kick that started 9.6 minutes into the simulation. This last kick developed into the uncontrolled kick illustrated in Figure 4. At a pump rate of 2500 l/min a series of progressively weaker influxes developed through the first four connections. These influxes were 1.7 kg, 1.1 kg, 0.7 kg and 0.6 kg, respectively, for connections one through four. The effects of these successions of kicks on the bottom hole pressure are illustrated by the fluctuations in the simulated pressure. At a pump rate of 3000 l/min very small influxes of 0.59, 0.11 and 0.02 kg were simulated for connections one through three.78 3000 l/min 76 2500 l/min 74 Pressure (bars)

Pressure (bars)

72 70.00 bars 70

72 70 2250 l/min 68

68 70.25 bars 66

66 70.50 bars 64 0 0 5 10 15 20 25 Time (min) 5 10 2000 l/min

64

15

20

25

Time (min)

Figure 3. Kick simulations for the 9 7/8 pilot hole with 0.5 darcy formation and 2000 l/min circulation with varying formation pressure.

Figure 4. Kick simulations for the 9 7/8 pilot hole with 0.5 darcy formation and 70.50 bars formation pressure with varying circulation rate.

As the 0.5 darcy formation is not as prone to a gas kick and would present a best case system for study, this formation was used for further evaluations. It should be anticipated that the use of a 1.0 darcy permeability (and of course even higher permeability) would yield larger kicks and result in shorter reaction times. Therefore, the remaining examples in this study are under the more favorable circumstances. Fluid Circulation Rate. The effects of seawater circulation rate were evaluated in the 9 7/8 inch pilot hole with the 0.5 darcy formation at 70.50 bars of pressure, which is the worst case illustrated in Figure 3. Figure 4 shows the results of these simulations using a constant circulation rate ranging from 2000 l/min to 3000 l/min. When using the 2250 l/min circulation rate from the beginning of the simulation instead of the original 2000 l/min an uncontrolled circulation also was realized. However, this kick was much more complicated and took longer time to reach an uncontrolled state. First a small kick of 3.0 kg of gas started when the bit penetrated the 644 meter gas zone. This kick lasted about 1.5 minutes before stopping. A second influx of an additional 3.0 kg took place

The simulations at these different pump rates illustrate how different and complex gas kicks can be under varying conditions. These kicks can come all at once or be the result of a series of progressive kicks dependent on drilling conditions. They also illustrate how an uncontrolled kick can cascade from an initial smaller kick that remains in the annulus. 17 1/2 hole shallow gas kick potential Formation pressure effects. The possibility of eliminating a pilot hole, which is intended for control of shallow gas kicks, required a comparison with a larger hole. For such a comparison a 17 1/2 top hole was evaluated under the same conditions as the above pilot hole. The same drillstring configuration, 100 m/hr penetration rate, 412 meter seabed, 644 meter gas zone and 0.5 darcy formation permeability were used in this evaluation. With this larger hole the seawater circulation rate was increased to 3000 l/min. Figure 5 illustrates the results of the simulated top hole drilling at varying formation pressures. Although the above pilot hole did not yield a gas kick at 68.00 bars, this 17 1/2 hole had a very definite uncontrolled kick at this same pressure. Even at the lower pressure of 67.75 bars a lower intensity but still

SPE 63197

SHALLOW GAS KICK: SIMULATION AND ANALYSIS FOR TOP HOLE DRILLING WITHOUT A RISER

5

Pressure (bars)

uncontrolled kick resulted. At both of these formation pressure simulations the kicks started when the bit penetrated the gas formation. It was not until the reservoir pressure was set to 67.50 bars before this larger hole did not lead to an uncontrolled kick. Even though this reservoir pressure was only a quarter bar lower than the uncontrolled kick scenario the kick totaled 7.5 kg of gas. This limited kick also started when the bit penetrated the gas zone. At 67.00 bars gas formation pressure there was no kick simulated.80 78 76 74 Pressure (bars) 72 67.50 bars 70 68 66 64 68.00 bars 62 60 0 5 10 15 Time (min) 20 25 30 67.75 bars 67.00 bars

68 67 66 6000 l/min 65 5000 l/min 64 3000 l/min 63 62 61 60 0 2 4 Time (min) 6 8 10

Figure 6. Kick simulations for the 17 1/2 hole with 0.5 darcy formation and 67.75 bars formation pressure with varying circulation rate.

Figure 5. Kick simulations for the 17 1/2 hole with 0.5 darcy formation and 3000 l/min circulation with varying formation pressure.

This 17 1/2 hole was significantly more prone to a kick than the 9 7/8 pilot hole. In this larger hole there was only three-quarter of a bar pressure difference between no kick and a completely uncontrolled kick. The pressure at which the uncontrolled kick occurred with this larger hole was 2.5 bars lower than for the pilot hole. However the formation pressure where no kick took place was only one bar lower. Remember that the seawater circulation rate was 3000 l/min for the larger hole and only 2000 l/min for the pilot hole. Fluid Circulation Rate. Increasing the circulation rates in the 17 1/2 hole had a little effect on the gas kick simulations in the 67.75 bar formation. The simulations illustrated in Figure 6 indicated that doubling the pump rate to 6000 l/min (the maximum rate that simulator will accept) still does not prevent an uncontrolled kick when the bit enters into the gas zone with this larger hole. If this increased pump rate will not prevent a shallow gas kick from starting, increasing the pump rate once the kick starts will not, under normal circumstances, dynamically control the kick. These simulations show that the 17 1/2 hole will not compete with the 9 7/8 pilot hole both for preventing and controlling a shallow gas kick. Although this larger hole will not compete, it must be remembered that using the smaller pilot hole is not totally reliable at preventing the shallow kick in the first place.

As the comparison of the 17 1/2 hole results with the 9 7/8 pilot hole study shows a large difference in the influx and control of the shallow gas kick, some baseline data of these two examples were generated. The simulator generated baseline downhole pressures at the top of the 644 meter gas zone both under static conditions and at the different circulation rates used in this analysis. These downhole pressures, shown in Table 1, were determined without the presence of drill cutting in the annulus and at a zero penetration rate. Additionally, the above 17 1/2 hole simulations used the same drillstring as the earlier 9 7/8 pilot hole study. As these 8 drillcollars are not expected to be use in this size hole during drilling operations, this table also includes baseline downhole pressure for the more normal 9 1/2 drillcollars used with the 17 1/2 bit. In addition baseline data for 11 1/4 collars that represent the largest collars that could, in practice, be used for these drilling operations. Table 1. Static and Circulating pressures at 644 meters Hole/bit 9 7/8 17 1/2 17 1/2 17 1/2 (inches) Drillcollars 8 8 9 1/2 11 1/4 (inches) Pump rate Pressure Pressure Pressure Pressure (l/min) (bars) (bars) (bars) (bars) 0 66.17 66.17 66.17 66.17 2000 68.16 ---2250 68.63 ---2500 69.15 ---3000 70.31 66.20 66.21 66.22 5000 -66.25 66.27 66.31 6000 -66.29 66.31 66.37

6

C. MARKEN, S. HANSEN, J. REGAARD

SPE 63197

The simulated pressures in Table 1 reveal two relevant points. First point, the pressures show that the annular dimensions of the pilot hole have a much greater effect on the bottom hole circulating pressure than do the dimensions of the larger hole. The pilot hole shows a baseline pressure increase of over 3 bars within in the 3000 l/min circulation range. While over the 6000 l/min circulation range of the larger hole the pressure increase ranged from 0.12 bars to 0.20 bars in the 17 1/2 hole. The differences between these baseline data help to clarify the simulated performance of the two hole sizes. Second point, the 0.02 bars maximum pressure difference between the use of the 8 drillcollars and the more normal 9 1/2 collars in the 17 1/2 hole is not so great as to be a factor in the above analysis. Dynamic circulation kick control in a 9 7/8 pilot hole To illustrate the use of the 9 7/8 pilot hole for dynamic control of a shallow gas kick the previously described kick at 70.25 bar formation pressure (see Figure 3) was used. This example was selected since it was the slower of the two uncontrolled shallow gas kicks. This uncontrolled kick occurred as soon as the bit entered the gas bearing formation while circulating seawater at 2000 l/min. To dynamically control the kick the circulation rate was increased to 4000 l/min and simultaneously reducing the ROP to zero. The simulations illustrated in Figure 7 demonstrate the results of this dynamic kick control procedure when it was initiated at various times after the initial occurrence of the kick. It needs to be emphasized that these times are from the initial kick and not some time after the kick would be otherwise detectable.

instability remained for about ten minutes after implementing the routine. This instability is a direct result of the gas circulating out of the annulus. The application of this routine only one minute after the kick shows that the sooner the dynamic control start the better the success of preventing an uncontrolled shallow gas kick with this pilot hole. The rapid detection of the kick is essential in safely implementing such a dynamic kick control routine. The quick response required to control such a kick is probably best done by vigilantly monitoring the downhole pressure. However, using such a high pump rate also requires carefully monitoring of downhole pressure for possible washouts of the pilot hole, which can result in a reduction of the downhole pressure. This latter point may be moot when the kick produces sand. Hole cleaning effects on shallow gas kicks The forces that keep cuttings suspended in the annulus add to the pressure exerted on the bottom of the hole. This additional pressure can be influential in the prevention of a gas kick. As mentioned earlier the above simulations were not done with the cuttings concentration in steady state equilibrium. As the simulation represents an ideal situation the exact effects of the cuttings in actual drilling operations can only be inferred through these simulations. For that reason, and because of the time it takes to achieve the steady state in the simulation process, it was judged that the steady state was not necessary for the above analysis However, as the presence of cuttings will influence the bottom hole pressure and the potential for a shallow gas kick, it was necessary to run some simulations under steady state conditions. Figure 8 illustrates the difference in gas kick potential in the 9 7/8 pilot hole for the case where the gas formation pressure was 70.25 bars. This was the lowest pressure where an uncontrolled gas kick was achieved as shown in Figure 3. The plots in that figure start when the bit is four meters above the 644 meters gas zone. However, in Figure 8 both simulations were started at a bit depth 610 meters, which was 18 minutes simulation time before reaching the 640 meter depth was reached. As it took around 15 minutes of simulation time to reach an equilibrium cuttings concentration in the annulus the steady state was achieved before the 640 meter depth was reached. The steady state is clearly illustrated in the cuttings laden simulation, cuttings present, shown in Figure 8. As there was no gas kick in the presence of these cuttings, the bottom hole pressure shows a steady increase with bit depth. To verify the effect of cuttings suspension, the clean hole example was achieved through a simulated hole cleaning process after the 18 minute of simulation. When the cuttings were removed the simulation was continued with a 100 m/h ROP and 2000 l/min circulation. From that point the accumulation of cuttings in the annulus resumed. The resultant curve is the same as the earlier mentioned curve in Figure 3 for 70.25 bars formation pressure. These two simulations illustrated in Figure 8 dramatically show the effects of the cuttings in the annulus. The steady state cuttings concentration simulation indicate that there was

76 1.00 min 74

Pressure (bars)

72 1.25 min 70

68

1.50 min

66 none 64 0 5 Time (min) 10 15

Figure 7. Dynamic circulation control of a shallow gas kick in the 9 7/8 pilot hole by increasing the pump rate from 2000 l/min to 4000 l/min.

After 1.5 minutes into the kick it was too late to dynamically control the gas kick by doubling the pump rate to 4000 l/min. Initiating this same routine 15 seconds earlier was successful at controlling this kick. However, the bottom hole pressure fluctuation in this simulation shows that some

SPE 63197

SHALLOW GAS KICK: SIMULATION AND ANALYSIS FOR TOP HOLE DRILLING WITHOUT A RISER

7

no gas influx, even at the drillpipe connections. As previously shown the clean hole simulation resulted in an uncontrolled kick as soon as the bit penetrated the gas bearing zone. The presence of suspended cuttings made the difference between no kick and an uncontrolled kick.Pressure (bars) 75 74 73 Pressure (bars) 72 71 cuttings present 70 69 clean hole 68 67 66 18 20 22 24 26 28 30 32 Time (min)

78 77 70.50 bars 76 75 74 73 72 71 72.00 bars 70 72.50 bars 69 68 18 22 26 30 34 38 42 46 50 54 Time (min) 71.00 bars 71.50 bars

Figure 9. Kick simulations for the 9 7/8 pilot hole with steady state cuttings concentration prior to entering the gas zone at varying formation pressure.

Figure 8. Comparison of the effects of steady state cuttings concentration in the pilot hole annulus with that of an initial clean hole without cuttings using a 70.25 bars formation pressure, 100 m/hr ROP and 2000 l/min circulation.

The effects that suspended cuttings have on the gas kick potential as the pressure of the gas zone changes were also simulated. Figure 9 contains the results of simulation in the pilot hole under the steady state cuttings conditions described above. Again the simulation time shown in this figure starts when the bit has reached 640 meters and the cuttings in the simulation are already in steady state equilibrium from the 100 m/hr ROP with 2000 l/min seawater circulation. At 70.75 bars of formation pressure there was still no gas influx observed either when penetration the gas zone or at the connections. When increasing the pressure 0.25 bars to 71.00 bars in the gas zone there were a series of connection gas flows ranging from 0.3 kg to 0.8 kg for the first five connections after the bit penetrated the gas zone. However, none of these flows resulted in an uncontrolled kick. With a formation pressure of 71.50 bars there was a more complicated scenario. At this pressure the first influx of 1.7 kg of gas occurred at the first drillpipe connection after entering the gas zone. A second influx of 2.6 kg of gas occurred at the second connection. The gas in the annulus from the first kick decreased the hydrostatic head and accentuated this second kick. Almost immediately (less than a minute of simulated time) after this second kick the decreased hydrostatic head from the gas of both the first and second kick resulted in a third kick. As the gas moved up the annulus this third kick became uncontrolled.

An increase in the formation pressure to 72.00 bars in the presence of the steady state concentration of cuttings also resulted in an uncontrolled gas kick. Under these conditions a 3.0 kg gas flow occurred with the first connection. As this gas rose up the annulus and the hydrostatic pressure decreased, a kick took place four minutes after the first connection or seven minutes after entering the gas zone. As the hydrostatic pressure further decreased this kick became uncontrolled. With an additional 0.50 bar to 72.50 bars of formation pressure an uncontrolled gas kick formed as soon as the bit penetrated the gas zone at 644 meters. Comparison of theses results with the earlier simulation where the cuttings concentration were building-up when the gas zone was entered (see Figure 3) reveals several interesting differences. Most obvious is the higher formation pressure that is required for the kicks with the higher concentration of cuttings. Additionally, with steady state cuttings the simulations indicate that one bar formation pressure (71.00 bar vs. 72.00 bar) can be the difference between controlled gas kicks and an uncontrolled kick when the gas zone in reached. The earlier simulation without reaching steady state cuttings concentration this difference could be reduced to a quarter bar (70.00 bar vs. 70.25 bar). Whether such a difference takes place in practice is uncertain. However, these differences in simulation show the possibility of rather different kick scenarios at different levels of cuttings loading during drilling operations. They also illustrate the difference in kick potential with hole cleaning. These results show the effects of increased hydrostatic pressure from suspended cuttings and decreased hydrostatic pressure for gas flow up the annulus. As the gas pocket is drilled with an influx into the annulus there will inevitably be a reduction in the downhole pressures caused by the suspended cuttings. With the suspended cuttings the influx will need to be larger before reaching an uncontrolled kick situation.

8

C. MARKEN, S. HANSEN, J. REGAARD

SPE 63197

An attempt was made to determine if a kick would occur when the drill collars passed through the gas bearing zone. The larger dimensions of the annulus of the drill pipe result in a reduction in circulating pressure loss. Figure 10 shows an extension of the simulation with 70.00 bars formation pressure with 2000 l/min circulation and 100 m/hr penetration that is a part of pilot hole simulations in Figure 3. This particular simulation was selected because it was the highest formation pressure that did not sustain an uncontrolled kick. As the top of the 169 meters of drillcollars passed into the gas zone after 104 minutes of simulation time no kick was observed. However, when the drilling was stopped at 110 minutes of simulation time and circulation continued at 2000 l/min, a bottom hole pressure drop took place during the hole cleaning process. Four minutes later at 114 minutes of simulation time a gas kick took place. Without remedial action such as increasing the pump rate this developed into an uncontrolled kick.94Drill Collars pass top of gas zone Drilling stops with continued circulation

decreased while cuttings are removed. At a point (approximately 31 minutes of simulation time) before all of the cuttings are removed from the annulus an uncontrolled gas kick occurs.74

73 Pressure (bars)

72

71

70 18 20 22 24 26 Time (min) 28 30 32 34

Figure 11. Shallow Gas Kick while hole cleaning prior to making a drillpipe connectionPressure (bars) 90

86Cuttings removed from hole with circulation continuing

82 90 100 110 Time (min) 120 130

Figure 10. Shallow Gas Kick after Drillcollars Pass the Top of the Gas Zone

This simulation illustrates the consequences that may be anticipated with the removal of cuttings from the annulus. Although good hole cleaning is a concern when drilling a pilot hole with seawater, attention needs to be given to the possibility that during the dynamic suspension of the cutting in sea water can be masking a possible shallow gas kick. This can be a particular concern when stopping the circulation before the cuttings have been clean out of the hole can result in settling of the cuttings and stuck drillpipe. The potential of taking a shallow gas kick while hole cleaning prior to completing a connection is illustrated in Figure 11. This figure shows the same simulation with the 71.00 bar gas zone illustrated in Figure 9 with the exception that prior to the second connection after entering the gas zone the hole was cleaned by continuing the seawater circulation at 2000 l/min. When this circulation is continued prior to stopping to make the connection the bottom hole pressure

Discussion Through the use of this simulator package many different scenarios for shallow gas kick in a top hole well section with returns to the seabed could be evaluated. However, within this limited study, and with the caveat of accepting these resultant simulations, several observations have been made that are important to drilling operations. A direct comparison of the shallow gas kick potential in the 9 7/8 pilot hole relative to a 17 1/2 top hole can be made (compare Figure 3 with Figure 5 and Figure 4 with Figure 6). There are two advantages to the pilot hole. First, circulation in the pilot hole can obviously contain the higher gas pressure in the formation than the circulation in the larger hole. For the examples simulated the pilot hole does not result in an uncontrolled kick until a threshold formation pressure of 70.25 bars while the same kind of kick could occur at a threshold of 67.75 bars in the larger hole. With a pressure difference of this magnitude, this can be an important operational factor during drilling of the top hole section. More importantly, these simulations show that at the threshold pressure for an uncontrolled kick, the pilot hole can be successful at preventing the kick through the use of a higher circulation rate before penetrating the gas zone. These two situations alone make it difficult to bypass the pilot hole where there is a high probability of a shallow gas zone. A crucial factor for the dynamic control of a kick in a pilot hole is the response time before corrective action is taken. The simulation shown in Figure 7 shows that a matter of fractions of a minute could make the difference between containing a shallow gas kick and realizing an uncontrolled kick. Therefore the time between the start of the kick and the

SPE 63197

SHALLOW GAS KICK: SIMULATION AND ANALYSIS FOR TOP HOLE DRILLING WITHOUT A RISER

9

rig crew being aware of the kick is critical. In preventing an uncontrolled kick with downhole MWD instrumentation it is important that this technique can detect the kick without ambiguity. Even more important is the ability to transmit this information to the surface with sufficient frequency to minimize the crews reaction time. At present, data transmissions methods send data at intervals that can be larger than the fraction of a minute reaction time needed to make the difference in controlling the kick. Other techniques such as reliance on an ROV for observing gas leaving the well may not lead to sufficient reaction time as it could already be to late when the gas is observed at the wellhead. For the pilot hole example illustrated in Figure 7 it can take up to 2.5 minutes after penetrating the gas zone for the gas front to reach the seabed and be visually observable. This is well after the 1.5 minutes of gas influx at 2000 l/min fluid circulation rate and would lead to an uncontrolled kick even with the circulation rate increase to a 4000 l/min. Even if the circulation rate could be increased to 4000 l/min at the moment the gas zone were reached it would take 1.25 minutes for a gas front to reach the seabed. The simulations in this study suggest that hole cleaning can be an important factor in the development of a shallow gas kick. The additional hydrostatic head of the cuttings in the circulating seawater increase the downhole pressures exerted on the formation. Figure 8 dramatically indicates that the additional downhole pressure could make the difference between no observable gas kick and an uncontrolled gas kick. Where all other variables are held constant this difference is dependent on nothing more than the quantity of cuttings suspended in the annulus at the time that the bit enters the gas producing formation. A consequence of this difference is illustrated in the simulation shown in Figure 11. In this case an uncontrolled gas kick was simulated some time after safely entering the gas zone. Once inside this gas zone continued circulation to clean the hole of cuttings prior to making a connection reduced the downhole pressure to the point where a gas kick occurred as long as the circulation rate was held constant. Such a possibility must be a concern when cleaning a hole during drilling operations. Because of limitations in the present state of the development of this simulator package it was not possible to evaluate additional effect of the occurrence of a connection gas during the connection of a drillpipe during drilling operations. In the current state of development this software has a fixed routine that simulates the connection. During this routine the pump is stopped with a subsequent pressure drop (not really noticeable in the 17 1/2" hole simulation of Figures 5 and 6). The time for this connection is fixed (at approximately 15 s) and can not be extended to evaluate the effects of a longer connection time. Additionally, during this time period there appears to be no settling of the suspended cuttings. Although the connections have been shown to be critical during several of the simulated kick scenarios, neglecting the time and settling factor can change the simulated kicks. As the connection time is probably short,

lengthening this time would only exacerbate the kick, as it would allow more gas to enter the annulus. Settling of cuttings during a connection could also exacerbate the kick. During this study no attempt was made to simulate the effects of the periodic circulation of a viscous pill to enhance hole cleaning operations. Conclusions Through the use of the simulator the applicability of a 9 7/8 pilot hole for top hole riserless drilling with the potential for a shallow gas kick has been evaluated. Although the simulations can not reproduce precise drilling conditions and operations, they are useful at pointing out potential sources of problems and evaluating possible actions to minimizing their effects. Through the simulations done in this study several conclusions can be made. 1. A simulation tool has been developed that can be valuable in evaluating shallow gas kick in both the well planning stage and during drilling operations. 2. Within the limits of the simulations it is difficult to justify the elimination of the pilot hole when drilling in an unknown formation or a formation with the high potential for a shallow gas zone. 3. A 9 7/8 pilot hole has several advantages over a 17 1/2" as a higher gas pressure zone can be drilled without sustaining an uncontrolled gas kick. 4. Likewise the formation pressure margin between no kick and an uncontrolled gas kick is greater for the pilot hole that with the larger hole. 5. Increasing the circulation rate of the seawater fluid in the 9 7/8 hole can control or prevent a gas kick where a similar operation in the 17 1/2" hole may still lead to an uncontrolled kick situation. 6. Even though a gas zone can be entered during the drilling process without an influx of gas into the well, a subsequent drilling operation, such as making a connection, may lead to a gas kick. 7. Likewise, hole cleaning, which removes cuttings from the annulus and reduces the downhole pressure, may result in an uncontrolled kick even though no gas influx occurred when the gas zone was penetrated. 8. The potentially short reaction time between the initial influx of gas and a corrective operation to prevent an uncontrolled kick can require a downhole detection system with high frequency transmission of the information to the surface. 9. Even though a pilot hole can be beneficial for the prevention and control of a shallow gas kick, its use can still lead to a false sense on security and uncontrolled kick situations can easily occur when inadequate attention is given to drilling practices. Nomenclature ECD = equivalent circulation density ROP = rate of penetration

10

C. MARKEN, S. HANSEN, J. REGAARD

SPE 63197

Acknowledgements We thank RF-Rogaland Research, Petec Software and Services, Saga Petroleum (more recently a part of Norsk Hydro) and Norsk Hydro for allowing us to produce this paper. We would also like to thank Saga Petroleum for the financial support that made this study possible. Special thanks go to many of the staff of the RF-Group, particularly those at RF-Rogaland Research and Petec Software and Services, who have developed the applications that are contained in the Drillbench software package. References1. Adams, N. J. and Kuhlman, L. G.: Case History Analysis of Shallow Gas Blowouts, paper SPE 19917 presented at the 1990 IADC/SPE Drilling Conference, Houston, Texas, February 27March 2, 1990. 2. Starrett, M. P., Hill, A. D. and Sepehrnoori, K., A ShallowGas-Kick Simulator Including Diverter Performance, SPE Drilling Engineering, (March 1990) 79-85. 3. Murray, S. J., Williamson, M. D., Gilham, S., Atkins, W. S. and Thorogood, J. L., Well Design for Shallow Gas, paper SPE 29343 presented at the 1995 SPE/IADC Drilling Conference, Amsterdam, February 28-March 2, 1995. 4. Sangesland, S., Sivertsen, A. and Hoirth, E., Downhole Blowout Preventer, paper SPE 21962 presented at the 1991 SPE/IADC Drilling Conference, Amsterdam, March 11-13, 1991. 5. Anonymous, Regulation relating to drilling and well activities and geological data collection in the petroleum activities, Norwegian Petroleum Directorate, 1998. 6. Staff at Petec, Drillbench 2.1 User Guide and Documentation, Petec Software and Services, 2000. 7. Vefring, E. H., Wang, Z., Gaard, S., and Bach, G. F., An Advanced Kick Simulator for High Angle and Horizontal Wells-Part I, paper SPE 29345 presented at the 1995 SPE/IADC Drilling Conference, Amsterdam, February 28March 2, 1995. 8. Rommetvent, R., Vefring, E. H., Wang, Z., Fieseman, T and Faure, A. M., A Dynamic Model for Underbalanced Drilling With Coiled Tubing, paper SPE 29363 presented at the 1995 SPE/IADC Drilling Conference, Amsterdam, February 28March 2, 1995. 9. von Flatern, R., Muds made to measure, Offshore Engineer, September, 1999.

SI Metric Conversion Factors bar F ft in lbm 1.0* E + 05 = Pa (F 32)/1.8 = C 3.048* E 01 = m 2.540* E 02 = m 4.535 924 E 01 = kg

*Conversion factor is exact