Kinetic Hydrate Inhibition at Pressures up to 760 Bar in Deep Water

Apr 22, 2010 - 4036 Stavanger, Norway. Jan Erik Iversen. International Research Institute of Stavanger (IRIS), P.O. Box 8046, 4068 Stavanger, Norway...
0 downloads 0 Views 2MB Size
Energy Fuels 2010, 24, 3003–3013 Published on Web 04/22/2010

: DOI:10.1021/ef9016152

Kinetic Hydrate Inhibition at Pressures up to 760 Bar in Deep Water Drilling Fluids Malcolm A. Kelland* Department of Mathematics and Natural Sciences, Faculty of Science and Technology, University of Stavanger, 4036 Stavanger, Norway

Jan Erik Iversen International Research Institute of Stavanger (IRIS), P.O. Box 8046, 4068 Stavanger, Norway Received December 30, 2009. Revised Manuscript Received March 21, 2010

Kinetic hydrate inhibitors (KHIs) have been used successfully in the field for about the past 15 years to prevent gas hydrate formation mostly in gas- and oilfield production lines. They work by delaying the nucleation and often also the growth of gas hydrate crystals for periods of time dependent on the subcooling in the system and the absolute pressure. In this paper, we have investigated whether KHIs can be used at very high pressures, particularly for use in ultra-deep-water completion and water-based drilling fluids (WBMs) with high salinity. Compatibility tests were carried out on various brines with several commercial KHIs. Laboratory hydrate equilibrium experiments were conducted on WBMs up to 800 bar. KHI performance tests were carried out in a 500 mL stirred steel autoclave at pressures up to 760 bar. These are much higher pressures than have ever been previously reported for any KHI studies. With the previous knowledge that not only the subcooling but the absolute pressure (70-150 bar) affects the performance of KHIs, our results indicate that, even at the highest pressures investigated, KHIs can still give substantial delays in hydrate formation as long as the subcooling is not too severe and the brine salinity is high.

situations, or minimization of the impact of hydrate formation such as elimination of potential hydrate blockages.2,8-11 Completion fluids can also suffer the same problem if gas is mixed with the fluid under cold, high pressure conditions. Oil-based muds (OBMs) are generally thought to be more effective at preventing hydrate problems than water-based muds (WBMs). This may be due to the lack of reported problems in the field when drilling in deep water with OBMs, although some laboratories have unpublished claims of having obtained hydrate plugs in autoclaves using OBMs, including our own laboratory. The best WBMs today are not adequate to prevent hydrate formation in all ultra-deep-water drilling situations, where the pressure can be several hundred bars at the sea bed and temperatures range from -2 °C to þ4 °C depending on the location (e.g., Gulf of Mexico, Norwegian North Sea, West Africa, offshore Brazil, etc.).12,13 Blends of thermodynamic inhibitors such as salts and glycols have a strong but limited ability to prevent hydrate formation. The addition of more salts (e.g., sodium chloride) is often impossible since the

Introduction Kinetic hydrate inhibitors (KHIs) are a class of low dosage hydrate inhibitors (LDHI) that have been in commercial use in the oil and gas industry for about 15 years.1-3 KHIs are water-soluble polymers, often with added synergists that improve their performance. KHIs delay the nucleation and usually also the crystal growth of gas hydrates. The nucleation delay time (induction time), which is the most critical factor for field operations, is dependent on the subcooling (ΔT) in the system;the higher the subcooling, the lower the induction time. The absolute pressure (tested in the range 70-150 bar) has also been shown to be an important factor and is therefore very relevant for this very high pressure study.4-7 In deep water drilling, the low temperature and high pressure favors gas hydrate formation in the drilling fluid if gas is present. The problem that needs to be solved regards the prevention of hydrates from forming during well control *To whom correspondence should be addressed. E-mail: malcolm. [email protected]. (1) Kelland, M. A. Production Chemicals for the Oil and Gas Industry; CRC Press (Taylor & Francis Group): Boca Raton, FL, June 2009. (2) Sloan, Jr. E. D.; Koh, C. A. Clathrate Hydrates of Natural Gases, 3rd ed.; CRC Press, Taylor & Francis Group: Boca Raton, Fl, 2008. (3) Kelland, M. A. Energy Fuels 2006, 20, 825. (4) Kelland, M. A.; Svartaas, T. M.; Dybvik, L. Ann. N.Y. Acad. Sci. 2000, 912, 744-752. (5) Arjmandi, M.; Tohidi, B.; Danesh, A.; Todd, A. C. Chem. Eng. Sci. 2005, 60, 1313–1321. (6) Peytavy, J.-L.; Glenat, P.; Bourg, P. IPTC 11233. In Proceedings of the International Petroleum Technology Conference, December 4-6, 2007, Dubai, U.A.E. (7) Kelland, M. A.; Mønig, K.; Iversen, J. E.; Lekvam, K. A Feasibility Study for the Use of Kinetic Hydrate Inhibitors in Deep Water Drilling Fluids, In Proceedings of the 6th International Conference on Gas Hydrates, July 6-10, 2008, Vancouver, Canada. r 2010 American Chemical Society

(8) Zamora, M.; Broussard, P.; Stephens, M. In Proceedings of the SPE International Petroleum Conference and Exhibition, Villahermosa, Tabasco, Mexico, February 1-3, 2000; SPE 59019. (9) Reyna, E. M.; Stewart, S. R. In Proceedings of the SPE/IADC Drilling Conference, February 27-March 1, Amsterdam, The Netherlands; SPE/IADC 67746. (10) Barker, J. W.; Gomez R. K. J. Petrol. Tech. 1989, March, 297. (11) Castro, G. T.; Terry, A. P.; Ferreira, L. V.; Ribeiro, G. S. In Proceedings of the Offshore Technology Conference, Houston, TX, 1998; OTC008777. (12) Ebeltoft, E.; Yousif, M.; Sørgard, E. SPE Drilling Completion 2001, 16 (1), p 19; SPE 68207. (13) Sørgard, E.; Alteras, E.; Fimreite, G.; Dzialowksi, A.; Svanes, G. S. In Proceedings of the SPE/IADC Drilling Conference, February 27-March 1,2001, Amsterdam, The Netherlands; SPE/IADC 67834.

3003

pubs.acs.org/EF

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

solutions are either saturated already or become extremely high density, which causes operational problems. Ethylene and propylene glycols and glycerols are more expensive and need to be added in large amounts to have significant effects.14,15 These low salinity, low density fluids with organic thermodynamic inhibitors are especially suited for use when low fracture gradients are encountered while drilling intervals where gas hydrate formation in the drilling fluid is possible. In addition, salts become less soluble as the percentage of glycol, or other organic solvent, in the aqueous phase increases, so there is a trade-off between these two classes of thermodynamic inhibitors. Consequently, there is a need to find improved methods to prevent hydrate formation in deep water drilling. The addition of KHIs to water-based drilling fluids or completion fluids could be a cost-effective alternative for hydrate control in ultra-deep-water drilling and completion. They could give the extra protection needed where the deep water drilling or completion fluid, containing thermodynamic inhibitors, is under-inhibited due to density or cost considerations. Some promising work has already been carried out by our own and other research groups.7,16-18 In our most recent work, we showed that some KHIs, particularly neutral or anionic polymers, could be used in WBMs if the activity of the clay/cuttings is low, probably due to negligible adsorption of the KHI polymer onto the clay/cuttings.7 This work was conducted at pressures up to 300 bar. As far as we are aware, no work on the performance of KHIs has been published at pressures higher than this. It was the objective of this work to investigate the possibility of using KHIs in completion fluids and WBMs at pressures up to 750 bar, typical of pressures found in ultradeep water.

Table 1. Composition of PB1 polysaccharide carboxymethylcellulose AVAS carboxymethylcellulose ADS tipoII KCl NaCl cationic polymer triazine NaOH surfactant

1 lb/bbl 1 lb/bbl 1.75 lb/bbl 16 lb/bbl 70 lb/bbl 6 lb/bbl 0.3 lb/bbl 0.75 lb/bbl 0.2 lb/bbl

There are several classes of KHIs currently commercially available. These are listed below: • N-Vinyl lactam polymers, the most common class of KHIs used today19-23 O Two common N-vinyl lactam-based KHIs are poly-N-vinyl caprolactam and N-vinyl caprolactam:N-vinyl pyrrolidone copolymer. • Hyperbranched poly(ester amide)s24-26 • Polyisopropylmethacrylamides1,27,28 • Polypyroglutamates29,30 It is known that polyisopropylmethacrylamide and poly-Nvinylcaprolactam have low cloud points in freshwater (ca. 30-35 °C) and are therefore not compatible with high salinity WBMs, so we excluded them from our project plans.1 In addition, we did not have a commercial sample of a poly(ester amide) with good solubility in high salinity brines at elevated temperatures. Polypyroglutamates are a recent commercial class of KHIs designed with high biodegradability but currently with fairly low performance. This left us with N-vinyl caprolactam:N-vinyl pyrrolidone copolymers which are commercially manufactured by two companies, BASF and ISP. We chose to study N-vinyl caprolactam:N-vinyl pyrrolidone copolymers from BASF, part of their Luvicap series, as they were available in varying molecular weights, had high cloud points, and we had success using them in earlier work on WBMs (Figure 1).7 Three samples were obtained, a low molecular weight 1:1 N-vinyl caprolactam:N-vinyl pyrrolidone copolymer called Luvicap 55W and two 1:2 N-vinyl caprolactam:N-vinyl pyrrolidone copolymers called Luvicap 21W 20k and Luvicap 21W 12k (the “k” values refer to the molecular weight). The 1:2 copolymers are expected to have a lower KHI performance than the 1:1 copolymer due to a lower percentage of N-vinyl caprolactam momomer. Very recently, ISP has reported proprietary N-vinyl caprolactam copolymers with high cloud points and good performance, but these were not available for our study.31

WBMs and KHIs Used in This Study Besides various brines containing NaCl or CaCl2, three WBMs were used in this work, coded muds PB1, PB2, and PB3. They were supplied by Petrobras. The composition of PB1 is as shown in Table 1. Muds PB1 and PB2 are similar in composition, except PB2 has an additional 5% monoethylene glycol (MEG). In terms of hydrate subcooling, the salt content alone in these muds is equivalent to a pure 19.0 wt % NaCl solution. Mud PB3 is a low salinity spudding mud and was only used in the hydrate equilibrium studies. All three WBMs are typical of what one can obtain from a range of drilling fluid companies. (14) Halliday, W.; Clapper, D. K.; Smalling, M. In Proceedings of the IADC/SPE Drilling Conference, March 3-6, 1998, Dallas, TX; SPE 39316. (15) Halliday, W.; Clapper, D. K.; Smalling M. U.S. Patent 6080704, 2000. (16) Power, D.; Slater, K.; Aldea, C.; Lattanzi, S. AADE 2003 National Technology Conference, April 1-3, 2003, Houston, TX; AADE-03-NTCE-48. (17) Dzialowski, A.; Patel, A.; Nordbo, K. In Proceedings of the Offshore Mediterranean Conference and Exhibition, March 28-30, 2001, Ravenna, Italy. (18) Trenery, J.; Fleyfel, F.; Shukla, K.; Hakimuddin M. Gas Hydrate Control for Deep Water Drilling Operations. Report from JIP at Intertek Westport Technology Center, Houston, TX, 1999-2000. (19) Sloan, E. D.; Christiansen, R. L.; Lederhos, J.; Panchalingam, V.; Du, Y.; Sum, A. K. W.; Ping, J. U.S. Patent 5639925, 1997. (20) Talley, L. D.; Mitchell, G. F. Application of Kinetic Hydrate Inhibitor in Black-Oil Flowlines, SPE 56770, SPE Annual Technical Conference and Exhibition, September 1999. (21) Fu, S. B.; Cenegy, L. M.; Neff, C. A Summary of Successful Field Applications of A Kinetic Hydrate Inhibitor, SPE 65022, SPE International Symposium on Oilfield Chemistry, February 13-16, 2001, Houston, TX.

(22) Jurek, M.; Alexandre, M.; Bell, S. New Approaches to Low Dose Gas Hydrate Treatment. Chemicals in the Oil Industry; Royal Society of Chemistry: Manchester, U. K., November 5-7, 2007. (23) Rasch, A.; Mikalsen, A.; Austvik, T.; Gjertsen, L. H.; Li, X. Evaluation of a Kinetic Inhibitor with Focus on the Pressure and Fluid Dependency. Proceedings of the 4th International Conference on Gas Hydrates, May 19-23, 2002, Yokohama, Japan. (24) Klomp, U. C. WO Patent Application 01/77270, 2001. (25) Rivers, G. T.; Crosby, D. L. International Patent Application WO/2004/022909. (26) Kelland, M. A.; Del Villano, L. Chem. Eng. Sci. 2009, 64, 3197. (27) Talley, L. D.; Oelfke, R. H. International Patent Application WO97/07320, 1997. (28) Thieu, V.; Bakeev, K.; Shih, J. S. U.S. Patent 6451891, 2002. (29) Leinweber, D.; Feustel, M. International Patent Application WO/2007/054226. (30) Arjmandi, M.; Leinweber, D.; Allan, K.; Development of A New Class of Green Kinetic Hydrate Inhibitors. 19th International Oilfield Chemical Symposium, Geilo, Norway, March 2008. (31) Musa, O. M.; Lei, C.; Zheng, J.; Alexandre, M. M. Advances in Kinetic Gas Hydrate Inhibitors, RSC Chemistry in the Oil Industry XI; Manchester, U. K., November 2-4, 2009.

3004

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen Table 2. Cloud Points (Tcl) for the Three Different Luvicap Samples in Various Salinity Brines and Muds (WBMs)

Figure 1. Structure of N-vinyl pyrrolidone:N-vinyl caprolactam (VP:VCap) copolymers such as Luvicap 55W or 21W.

[NaCl] wt %

55W

21W 20k

21W 12k

0.0 3.6 10.0 15.0 20.0 Mud PB1 Mud PB2

85 72 52 39 27 28-40 28-40

100 80 71 49 38 26 26

100 95 74 58 46

Figure 2. Structure of Jeffamine D series. D230 (n = 2.5), D400 (n = 6.1).

In this study, we also investigated the use of several synergists which are known to enhance the performance of N-vinyl lactam-based KHIs. They are as follows: • 2-Butoxyethanol (monobutyl glycol ether or BGE) obtained from Sigma-Aldrich32 • Butyl diglycol ether (BDGE) and butyl triglycol ether (BTGE) from BP Chemicals32 • Tetrabutylammonium bromide (TBAB) obtained from Sigma-Aldrich33 • Polyetheramines (or diaminopolyoxypropylenes) called Jeffamine D230 and Jeffamine D400 obtained from Huntsman Corporation (Figure 2)34

Figure 3. Deposition points (Tdp) for Luvicap 55W and Luvicap 21W 12k in various salinity NaCl solutions.

circulation, mud temperatures can sometimes be considerably higher (70-90 °C), even in ultra-deep-water drilling. Hence, there is a need for KHI polymers that are compatible with high salinity muds at these temperatures. This is not easy to do. As you change the structure of the KHI polymer to be more soluble at high temperatures, the percentage of the active ingredient in the polymer, N-vinyl caprolactam, decreases, and so the KHI performance will decrease also. We approached BASF, the manufacturers of Luvicap 55W, about this issue. They prepared for us a new polymer called Luvicap 21W 20k which is a 2:1 N-vinyl pyrrolidone:N-vinyl caprolactam (VP:VCap) copolymer at 50 wt % in water with a molecular weight of ca. 20 000. The compatibility data for Luvicap 55W and Luvicap 21W 20k in various NaCl brines and muds (WBMs) are given in Table 2. For example, the Tcl of Luvicap 21W 20k is 49 °C compared with 38 °C for Luvicap 55W in 15 wt % NaCl. BASF also prepared a lower molecular weight version of Luvicap 21W 20k, called Luvicap 21W 12k. The lower molecular weight raises the Tcl and deposition point (Tdp) as shown in Table 2. Thus, Luvicap 21W 12k displays higher Tcl’s than Luvicap 21W 20k. Therefore, we checked the Tdp’s for Luvicap 21W 12k in various brines and compared them to those of Luvicap 55W. Tdp’s are sometimes similar but more often higher than the Tcl’s. The data are illustrated in the graph in Figure 3. For example, Luvicap 21W 12K has a Tdp of 58 °C in 20 wt % NaCl and 70 °C in 15 wt % NaCl brine. As mentioned earlier, in terms of hydrate subcooling, the salt content in muds PB1 and PB2 used in this project for KHI performance testing is equivalent to a pure 19.0 wt % NaCl solution. Thus, we can determine that the maximum operational temperature that one can safely use the Luvicap polymers in PB1 or PB2 without polymer deposition problems is

Compatibility Studies Cloud points and deposition points were measured by putting 10 mL of a 1 wt % solution of the KHI or synergist in a large test tube, which was then placed in a stirred oil bath on a heat regulator. The oil bath was heated at ca. 1 °C/min, but near the cloud point it was heated at ca. 0.2 °C/min. The cloud point (Tcl) was determined as the temperature at which the first sign of appreciable turbidity was detected, which was observed as a large and sharp change.35 The solution was heated further, and the deposition point (Tdp) was determined as the temperature for the first sign of solids on the walls of the test tube. If the Tdp value had not been reached, the solution was cooled below the cloud point until clear and the solution reheated and the cloud point determined again. The repeated cloud point values were all within (0.5 °C of the first values. In earlier work, we showed that Luvicap 55W had high Tcl values in high salinity brines and could be used in WBMs.7 For example, in 15 wt % NaCl brine, the Tcl is 39 °C. Switching some of the NaCl for KCl on a weight basis gave practically identical Tcl values. However, it is possible to use this polymer at up to 50 °C since it was shown that the polymer does not deposit from solution until above ca. 50 °C. However, during (32) Cohen, J. M.; Wolf, P. W.; Young, W. D. U.S. Patent Application 5723524, 1998. (33) Duncum, S.; Edwards, A. R.; Osborne, C. G. International Patent Application WO96/04462, 1996. (34) Hurd, D.; Pakulski, M. Uncovering a Dual Nature of Polyether Amines Hydrate Inhibitors. Proceedings of the 5th International Conference on Gas Hydrates, June 12-16, 2006, Trondheim, Norway. (35) Kjøniksen, A.-L.; Laukkanen, A.; Galant, C.; Knudsen, K. D.; Tenhu, H.; Nystr€ om, B. Macromolecules 2005, 38, 948.

• ca. 38 °C for Luvicap 55W • ca. 60 °C for Luvicap 21W 12k We also investigated the Tcl of the Luvicap polymers in calcium chloride (CaCl2) brines (Figure 4). In these brines, the Tcl is significantly higher. For example, at 19 wt % CaCl2, the 3005

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

Figure 4. Cloud points in CaCl2 brines for the three Luvicap polymers 55W, 21W 20k and 21W 12k.

Figure 6. Schematic outline of the titanium cell arrangement for KHI performance testing. Table 3. The Standard Natural Gas (SNG) Composition

Figure 5. Cloud points (Tcl) for Jeffamine D400 in NaCl brines.

component

mol. wt.

gas: mol %

N2 CO2 CH4 C2H6 C3H8 i-C4H10 n-C4H10 sum

28.013 44.010 16.043 30.070 44.097 58.124 58.124

0.105 1.820 80.426 10.350 4.980 1.590 0.729 100

Figure 6. For example, in 15 wt % NaCl, the solution became cloudy above approximately 42 °C, and at 30 °C in 20 wt % NaCl. Thus, D230 is much more suitable to be used with high salinity brines than Jeffamine D400. To summarize the compatibility studies, Luvicap 55W is compatible with high NaCl, NaCl/KCl, and CaCl2 brines and could be blended with TBAB, three butyl glycol ethers, or Jeffamine D400 to improve the performance, as these synergists are also compatible with these brines at elevated temperatures.

Tcl for Luvicap 55W is ca. 65 °C. This compares with 24 °C for Luvicap 55W in 19 wt % NaCl. For Luvicap 21W 20k, the cloud point is 84 °C in 19 wt % CaCl2 brine. Somewhat surprisingly, the cloud point for Luvicap 21W 12k is over 100 °C in all CaCl2 brines, even up to 20 wt %. More importantly, the deposition points in CaCl2 brine will be even higher for this polymer; i.e., they probably lie ca. 10 °C above the cloud points. This means one can use Luvicap 21W 12k in all CaCl2 brines irrespective of the brine concentration and temperature. In addition, one can use the higher performing KHI Luvicap 55W at up to ca. 80 °C in 10-20% CaCl2 brines, and up to 90 °C in lower CaCl2 salinities. Thus, this class of KHI polymer is easier to use operationally in CaCl2 brines than in NaCl or KCl brines due to better compatibility at high temperatures. We also investigated the compatibility of the synergists with high salinity brines. TBAB was soluble at all temperatures up to 100 °C in up to 23 wt % CaCl2 or NaCl brines. For the three butyl glycol ethers, BGE, BDGE, and BTGE, all three were completely miscible at 1 wt % in distilled water or NaCl brines of up to 20 wt % (Figure 5).36,37 Therefore, these butyl glycol ether synergists are compatible with low or high salinity WBMs. For the polyetheramines, Jeffamine D230 has a lower molecular weight than Jeffamine D400. Jeffamine D230 was completely miscible at 1 wt % in all brines up to 20 wt % NaCl. The cloud point data for Jeffamine D400 is given in

High Pressure Equilibrium Tests Test Procedure. Equilibrium and KHI performance tests were carried out using a titanium cell autoclave setup illustrated in Figure 6 and described previously.7 Tests were conducted at pressures up to 800 bar. The standard natural gas (SNG) composition, which forms Structure II hydrates, used in all experiments is given in Table 3. To determine the equilibrium temperature by the common method dissociation of hydrates, the brine or mud sample was loaded into the cell, which was closed and purged twice with SNG to remove residual air from the system. The cell was then loaded with SNG to a little higher pressure than the test pressure at ca. 22 °C. The cell was cooled down to approximately 2-5 °C to provoke hydrate production. In a closed cell system such as the sapphire cell arrangement, a sudden pressure drop (cf. point 2 in Figure 7) detects hydrate formation. When hydrates formed at the given pressure, the cell was heated to a temperature at least 5 °C below the hydrate formation temperature with a heating rate of 4 °C/h. The cooling bath was then programmed to continue the heating

(36) Ochi, K.; Tada, M.; Kojima, K. Fluid Phase Equilib. 1990, 56, 341. (37) Glycol Ethers. http://www.dow.com/PublishedLiterature/ dh_0032/0901b80380032bc8.pdf?filepath=oxysolvents/pdfs/noreg/ 110-00965.pdf&fromPage=GetDoc (accessed Apr 2010).

3006

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

Figure 7. A typical P vs T plot showing the hydrate formation;dissolution cycle of a typical run.

at lower and lower gradients down to 0.1 °C/h. We find that this very slow heating gives good results, although we note that one research group recommends stepwise heating only.38 Heating continued at this gradient up to a temperature of ca. 1-1.5 °C outside the hydrate region. The low heating gradient used in the vicinity of the hydrate equilibrium point was assumed sufficiently low to ensure equilibrium in the system during final hydrate dissolution. A typical pressure (P) vs temperature (T) plot for a run is shown in Figure 7. The temperature and pressure at final hydrate dissociation define the hydrate equilibrium point. In a closed system (i.e., isochoric system), the hydrate equilibrium point can be found as the final point of deflection on a P vs T plot (cf. point 6 in Figure 7). Point 6 in Figure 8 is reproducible, while the onset of hydrate formation may occur anywhere on the PT line between points 6 and 2. The hydrate dissolution point represents a true equilibrium property and is defined as the hydrate equilibrium point at the given pressure. The heating cycle was stopped ca.1.5 °C outside the hydrate region to ensure complete dissolution of the hydrates. Having reached the maximum temperature, the temperature was decreased to produce hydrates a second time, and the heating cycle as described above was repeated to dissolve the new hydrates. For each pressure examined, at least two parallel experiments were run. The three muds PB1, PB2, and PB3 were tested for their equilibrium properties at up to 800 bar. The results are given in Figure 8 on a logarithmic scale. Clearly, mud PB3 has much lower salinity and density than PB1 and PB2, as it gives much higher equilibrium temperatures compared to PB1 and PB2. PB3 is therefore much more susceptible to hydrate formation even with added KHI. For example, at 500 bar and 4 °C, the subcooling with mud PB3 is approximately 19.0 °C, whereas with mud PB1 the subcooling is 11.0 °C.

Figure 8. Pressure versus temperature phase diagram for muds PB1, PB2, and PB3. Equilibrium points determined for the three Petrobras (PB) Muds. PB1 is a high salinity mud, PB2 is the same high salinity mud but with added 5% MEG, and PB3 is a spudding mud with much lower salinity.

High Pressure KHI Performance Experiments The KHI performance test procedure is an isothermal method carried out as follows: 1. The KHI to be tested was dissolved in the brine or mud to the desired concentration, usually 6000 ppm. 2. The magnet housing of the cell was filled with the mud containing the additive to be tested. The magnet housing was then mounted in the bottom end piece of the cell, which thereafter was attached to the sapphire tube and the cell holder. 3. The desired amount of mud was put into the cell (above the cell bottom) using a pipet. The top end piece was mounted, and the cell was placed into the cooling bath (plastic cylinder). 4. The temperature of the cooling bath was adjusted to 2-3 °C outside the hydrate region at the pressure conditions to be used in the experiment.

(38) Østergaard, K. K.; Masoudi, R.; Tohidi, B.; Danesh, A.; Todd, A. C. J. Petr. Sci. Eng. 2005, 48, 70.

3007

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

Table 4. KHI Test Results with Luvicap 55W with and without 3000 ppm Synergists at 4 °C and Varying Pressures on Mud PB3 Luvicap 55W concentration (ppm)

pressure (bar)

subcooling (°C)

induction time, ti (h)

comments

6000 6000 10000 10000 6000 þ D230 6000 þ TBAB 6000 þ D230 6000 þ TBAB 10000 þ TBAB

290 626 290 626 290 290 626 626 290

17.3 19.9 17.3 19.9 17.3 17.3 19.9 19.9 17.3

0.1, 0.1, 0.1, 0.1 0.0, 0.0 0.3, 0.4, 0.3, 0.4 0.0, 0.0 0.1, 0.5, 0.5, 2.0 0.0, 0.3, 0.3, 0.2 0.0, 0.0 0.0, 0.2, 0.2 0.1, 0.1

rapid hydrate formation at onset rapid hydrate formation at onset rapid hydrate formation at onset rapid hydrate formation at onset rapid hydrate formation at onset rapid hydrate formation at 0.5-1.5 h rapid hydrate formation at onset rapid hydrate formation at onset rapid hydrate formation at 2.3 and 4 h

5. The cell was purged twice with the SNG used, and then the cell was loaded with SNG to the desired pressure stirring at 700 rpm. 6. The stirring was stopped and the cell cooled to the experimental temperature. When the temperature and pressure in the cell had stabilized, the stirring was started.

normally 6-7 h. This is perhaps surprising since, for example, at 600 bar and 4 °C, the subcooling is about 10.6 °C. Thus, the extra 5% MEG seems to make a lot of difference compared to mud PB1 at these pressures and subcoolings. However, in three experiments with PB2 at approximately 716 bar and 4 °C (12.3 °C subcooling), rapid pressure drops of 25-30 bar were observed after 1.7 h in one of the experiments and after 11 and 19 h in two further experiments. No slow pressure drop was observed before rapid gas uptake and hydrate formation ensued.

The induction time, ti, for hydrate formation was measured from the time of the start of stirring at the experimental temperature. The time from the start of hydrate formation to the time when rapid growth of hydrate ensues, usually with the formation of a hydrate plug in the cell, is called the crystal growth delay time, St-1. The results of all experiments were recorded by plotting the gas consumption (bars) as a function of time (minutes or hours).

KHI Experiments with WBMs with Added KHIs All KHI polymers were tested at either 6000 ppm or 10000 ppm active polymer in the WBM aqueous phase. In the first phase, we used only Luvicap 55W with and without synergists to gauge the performance at very high pressures and deepwater seabed temperatures. The results are summarized in Table 4. Phase 1: Tests with Mud PB3. We began by testing Luvicap 55W in mud PB3, as this mud was thermodynamically much less inhibited than either PB1 or PB2, so that the KHIs would probably fail at the high pressures and subcoolings used in this project. The first tests were at 290 bar and 4 °C. The subcooling using PB3 is approximately 17.3 °C, which is presumed to be well outside the field performance range for 6000 ppm of the main KHIs in our study, Luvicap 55W and 21W. Thus, in tests at 290 bar and 4 °C using 6000 ppm Luvicap 55W, we observed rapid hydrate formation after 5-6 min after the start of stirring in four experiments. We carried out further tests using 10 000 ppm Luvicap 55W at 290 bar and 4 °C. The results were only slightly better. Rapid hydrate formation occurred after 18-25 min in four experiments. We also investigated using an additional 3000 ppm of the synergists D230 or TBAB with 6000 ppm Luvicap 55W in PB1 at 290 bar and 4 °C. The results with added D230 were very similar to using 6000 ppm Luvicap 55W alone, with rapid hydrate formation detected after 6 min in one test and after 30, 35, and 120 min in three other tests. When using added TBAB, we observed a slow pressure drop after 2-20 min in 4 tests and a rapid pressure drop after 37, 50, 80, and 90 min. Using 10 000 ppm Luvicap 55W and 3000 ppm TBAB, we observed an almost immediate pressure drop and rapid hydrate formation after 140 and 240 min. At 626 bar, hydrate formation was immediate after the start of stirring using either 6000 or 10 000 ppm Luvicap 55W alone, or a mixture of 6000 ppm Luvicap 55W and 3000 ppm D230. Using 6000 ppm Luvicap 55W and 3000 ppm TBAB at the same pressure, rapid hydrates formed immediately in one experiment after the start of stirring and after 10 and 12 min in two further experiments. Clearly, these concentrations of Luvicap 55W with or without synergists are not recommended

KHI Experiments with WBMs without Added KHIs All three WBMs, PB1, PB2, and PB3, were tested for their ability to prevent hydrate prevention at various pressures and subcooling but always using isothermal conditions, beginning the stirring at 4 °C and holding at this temperature throughout the rest of the experiment. The results with PB3 contrast with those of PB1 and PB2. At all pressures from 150 to 658 bar in six experiments, we obtained rapid hydrate formation as soon as stirring was started. At 150 bar and 4 °C, the subcooling is ca. 14 °C, so it is not surprising that hydrate forms immediately. In fact, at 658 bar, hydrate formed at higher than the minimum temperature of 4 °C in both experiments during the nonstirring stage while cooling the cell. For PB1 at 150 bar and 4 °C, no hydrate was formed. The subcooling is roughly 3 °C in these tests, and in the laboratory we often find that we need at least 3-4 °C to get initiation of hydrate formation within a short time. This is because we are using a small “clean” cell, and the hydrate nucleation process is known to be stochastic in laboratory experiments.1-3 This is an important factor to remember when conducting KHI test in the laboratory and comparing to the field. In the field, we would expect plugging with hydrates to occur at 150 bar and 4 °C since there is still a hydrate subcooling of 3.5 °C. At 285 bar and 4 °C (8.4 °C subcooling), we obtained a rapid pressure drop and hydrate formation within 0-40 min in several experiments with PB1, and at 600 bar (12.2 °C subcooling) hydrates occurred immediately in all experiments. Clearly, this PB1 mud is capable of plugging rapidly at ultradeep-water conditions if gas breakthrough is experienced. PB2 contains additionally 5% MEG compared to no MEG in PB1. Thus, PB2 is more thermodynamically inhibited against hydrate formation. At 4 °C and 285 bar (three experiments), 500 bar (two experiments), and 600 bar (two experiments), we did not observe any pressure drop due to hydrate formation during the duration of the experiments, which was 3008

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

Table 5. KHI Test Results with Luvicap 55W with and without 3000 ppm Synergists at 4 °C and Varying Pressures on Mud PB1 Luvicap 55W pres- subcoolconcentration sure ing (ppm) (bar) (°C) 6000 6000 þ D230 6000 þ TBAB 6000 6000 þ D230 6000 þ TBAB 6000 6000 þ TBAB

290 290 290 600 600 600 738 750

8.4 8.4 8.4 12.2 12.2 12.2 13.4 13.5

induction time, ti (h) >8, >17 >8 >12, >12, >10, >17 5, 15 5, 35 >10, >11, >60 4, 4 >6, >6

Table 6. KHI Test Results with Luvicap 55W with and without 3000 ppm Synergists at 4 °C and Varying Pressures on Mud PB2

time to rapid hydrate growth (h)

15, 40 10, 50

Luvicap 55W concentration (ppm)

pressure (bar)

subcooling (°C)

induction time, ti (h)

6000 6000 6000 þ D230 6000 þ TBAB 6000 6000 þ TBAB

290 600 600 600 750 760

6.9 10.7 10.7 10.7 12.0 12.1

>6, >9 >9, >22 >20, >17 >6, >10 >10, >18 >4.5, >8

is perhaps surprising how good the results were at these pressures.4-7 This is also supported by the results with mud PB2 discussed below. Phase 1: Tests with Mud PB2. Mud PB2 contains an added 5% MEG compared to mud PB1. According to our equilibrium experiments, PB2 depresses SII hydrate formation by roughly an additional 1.5 °C compared to PB1. Hence, we should expect the results with PB2 and added KHIs to be a little better than with PB1 with added KHIs. Test results with mud PB2 are summarized in Table 6. In two experiments at each of the pressures 290 or 600 bar, no pressure drop or hydrate formation was observed. At 600 bar, the induction time was at least 9 h in one experiment and 22 h in another. In two tests at the same pressure using an additional 3000 ppm D230, we observed no pressure drop in 17 and 20 h. Finally, in two tests at the same pressure using an additional 3000 ppm TBAB, we observed no pressure drop in 6 and 10 h. These results are not significantly better than tests conducted without added KHIs at 600 bar, although more and longer experiments are needed to verify this. At 750 bar, we conducted two tests with 6000 ppm Luvicap 55W. No sign of a pressure drop was observed in 10 and 18 h in these two tests. At 760 bar, we conducted two tests with 6000 ppm Luvicap 55W and an additional 3000 ppm TBAB. No pressure drop or hydrate formation was observed in 4.5 and 8 h before termination of the tests, somewhat earlier than planned due to pressure sensor failure. Without KHIs, we observed slow hydrate formation after 1.7 h in one of the tests at 716 bar. Clearly, more and longer experiments are needed to verify that the use of KHI is giving additional inhibition at over 700 bar pressures, but the results using PB1 and PB2 certainly suggest that KHIs still function at these extreme pressures and deepwater seabed temperatures. However, since both PB1 and especially PB2 are strongly thermodynamically inhibited and the induction time results with and without added KHIs are in some cases indistinguishable, we decided to test Luvicap 55W at varying concentrations to see if this would better help determine the inhibitory effect of adding KHIs at these extreme pressures. We chose to use mud PB1 since this gave very short induction times for hydrate formation at 600-750 bar and so would be best for comparing with systems with added KHIs. Phase 2: KHI Tests with Luvicap 55W at Varying Concentrations Using Mud PB1. In phase 2, we first repeated the tests at 600 and 750 bar using Luvicap 55W at 6000 ppm, but now we varied the KHI concentration. The results are summarized in Table 7 and in Figure 9. Beginning with the same 6000 ppm Luvicap 55W concentration as in phase 1, we obtained comparable results to those in phase 1. No pressure drop and hydrate growth was observed in 14 h in one experiment and the start of slow hydrate growth after ca. 29 h in a second experiment. At lower concentrations of 2000 ppm and 1000 ppm, the results were fairly similar. The induction time was reduced to about

10, 18

for inhibiting PB3 at these substantial pressures, either 290 or 626 bar, and deepwater seabed temperatures. Phase 1: Tests with Mud PB1. Compared to mud PB3, muds PB1 and PB2 are much more thermodynamically inhibited and therefore much more relevant for the use of KHIs at very high pressures and deepwater seabed temperatures. Test results with mud PB1 are shown in Table 5. At 290 bar and 4 °C, the subcooling for mud PB1 is approximately 8.4 °C. Using 6000 ppm Luvicap 55W, we observed no pressure drop after 8 h in one test and 17 h in another before the experiments were terminated. The same observations were made when 3000 ppm D230 was added to 6000 ppm Luvicap 55W. No pressure drop occurred after 8 h. In two tests at 4 °C and 600 bar (12.2 °C subcooling), a slow pressure drop indicating slow hydrate growth was observed after about 5 and 15 h; however, neither test led to fast hydrate formation after 15 and 40 h, respectively. These long periods of slow growth were observed to be very common in tests using KHIs in high salinity muds such as PB1 and PB2. These results at 600 bar are much better than those of the test without added KHIs. When 3000 ppm D230 was added to the Luvicap 55W at 600 bar, we observed a slow pressure drop after 5 h in one test and 35 h in another, but again neither test gave rapid hydrate growth after 10 and 50 h, respectively. Two tests were conducted at 738 bar (13.4 °C subcooling) with only 6000 ppm Luvicap 55W. In both tests, we observed a slow pressure drop after 4 h but no rapid pressure drop after 10 and 18 h, at which time the experiments were terminated. The total pressure drop was no more than 3 bar in all these tests at 600 or 738 bar. Thus, even at 738 bar, Luvicap 55W is showing a clear effect as a KHI. The results using a combination of 6000 ppm Luvicap 55W and 3000 ppm TBAB were even more impressive. No hydrate formation was observed in four tests at 290 bar and 4 °C in at least 10 h (one test was terminated after 17 h without any sign of pressure drop). Three tests were conduced at 600 bar and 4 °C (12.2 °C subcooling). No pressure drop was observed in the first two tests after 10 and 11 h. Since these results were so good, the third test was continued for a longer period of 60 h. Again, we observed no pressure drop throughout the whole experiment. Two tests were conducted at 750 bar and 4 °C (13.5 °C subcooling), and no hydrate formation was observed in 6 h, at which point the tests were terminated. These results are better than those of the tests at 738 bar with Luvicap 55W alone. The mechanism of TBAB synergy with vinyl caprolactam polymers has been discussed.3 The subcooling at 600-750 bar is not so extreme (12.2-13.5 °C) for the use of a KHI, especially at a dosage of 6000 ppm with additive synergists, but as it is known that the absolute pressure also affects the performance of KHIs, it 3009

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

18-20 h at 2000 ppm. Surprisingly, even at a dosage of 1000 ppm Luvicap 55W, we still obtained long induction times. In one test, no hydrate formed in 12 h, while in the second test there is possibly slow hydrate growth after ca. 10 h as judged by a very small pressure drop after this time. However, after 19 h, we obtained fairly fast hydrate growth. The next set of tests with mud PB1 and Luvicap 55W were conducted at 4 °C and an even higher pressure of 750 bar. The subcooling is approximately 13.5 °C at these conditions, 1.4 °C more than at 600 bar. We tested Luvicap 55W at various concentrations. The results are summarized in Table 8 and Figure 10. As discussed earlier, no hydrates were detected using 6000 ppm Luvicap 55W. At 3000 ppm Luvicap 55W, gas uptake is observed clearly after 3 and 6 h in two tests, although the rate of hydrate growth was very slow for at least 18 h afterward in both tests. At 1000 ppm Luvicap 55W, there is still an induction time before any pressure drop occurs, but it is only about 1 h. From these results, we conclude that Luvicap 55W performs worse at 750 bar than 600 bar due to the higher pressure and subcooling. In the next group of experiments, we tested Luvicap 55W with added TBAB. Tests were done at 600 and 750 bar. The results are summarized in Table 9 and Figure 11. With 6000 ppm Luvicap 55W and 3000 ppm TBAB at 600 bar, we obtained equally good results as without TBAB, i.e., no hydrate formation in 20 h in two tests. With 1000 ppm Luvicap 55W and 500 ppm TBAB at 600 bar, we obtained an induction time of 23 h in one test. This is slightly better than Luvicap 55W at 1000 ppm without TBAB, but more tests would be needed to verify any improvement in performance. However, the synergistic performance enhancement of TBAB is clearer in tests at 750 bar. At this pressure we obtained induction times of 7 and 14 h in two experiments with 1000 ppm Luvicap and 500 ppm TBAB compared to 1 h in one experiment with Luvicap 55W alone at 1000 ppm.

Thus, addition of the cationic synergist TBAB does seem to improve the KHI performance. Since adding the synergist D230 to 6000 ppm Luvicap 55W did not show any marked improvement in performance in earlier tests, we did not carry out any further tests at other concentrations with this synergist. Phase 3: KHI Tests with Luvicap 21W 29k at Varying Concentrations Using Mud PB1. Although we have shown that Luvicap 55W is a useful KHI for giving extra subcooling protection in high salinity drilling or completion brines at

Table 7. Induction Times (ti) for KHI Tests with Luvicap 55W in Mud PB1 at 600 bar and 4 °C

Table 9. Tests with Luvicap 55W Plus TBAB and Varying Concentration and Pressures

expt. no. 1 2 3 4 5 6

KHI Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W

conc. [ppm] 6000 6000 2000 2000 1000 1000

Table 8. Tests with Luvicap 55W at 750 bar and Varying Concentration expt. no.

KHI

conc. [ppm]

ti [h]

7 8 9 10 11

Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W

6000 6000 3000 3000 1000

>18 >10 6 3 1

Figure 10. Tests with Luvicap 55W at 750 bar and varying concentration (1k = 1000 ppm).

ti [h] >14 29 20 18 >12 10-19

expt. no.

KHI

[polymer] [ppm]

[TBAB] [ppm]

P [bar]

ti [h]

12 13 14 15 16

Luvicap TBAB Luvicap TBAB Luvicap TBAB Luvicap TBAB Luvicap TBAB

6000 6000 1000 1000 1000

3000 3000 500 500 500

600 600 600 750 750

>20 >20 23 7 14

Figure 9. Experiments with Luvicap 55W in mud PB1 at 4 °C and 600 bar at various concentrations (1k = 1000 ppm).

3010

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

extreme pressures and deepwater seabed temperatures, its application range is limited by its solubility (cloud point) at elevated temperatures as discussed earlier. Therefore, we also wanted to investigate the KHI performance of Luvicap 21W, which is a more hydrophilic polymer and consequently has much higher cloud points in brines. Although we had two samples of Luvicap 21W available with different molecular weights, we only investigated Luvicap 21W 20k. Luvicap 21W 20k was expected to be a worse KHI than Luvicap 55W since the percentage of the most active component, N-vinyl caprolactam, in the copolymer was only 33 wt % compared to 50 wt % in Luvicap 55W. This was borne out in the experiments summarized in Table 10 and Figures 12 and 13. We obtained induction times of approximately 7 h in two experiments using 3000 ppm Luvicap 21W 20k and approximately 4 h in two experiments using 2000 ppm Luvicap 21W 20k. In comparison, 2000 ppm Luvicap 55W gave 18-20 h induction time under the same conditions. Using Luvicap 21W 20k, gas hydrates grew very slowly for several hours after initiation in all experiments; the longer, slow growth times were obtained using the higher KHI concentration as expected. For example, using 3000 ppm polymer at 600 bar, after the 7 h induction time, the growth of hydrate is very slow for at least 50 h. However, at 750 bar, Luvicap 21W 20k was not able to delay hydrate nucleation at all at 4000 ppm in two experiments, although neither test gave rapid hydrate growth but only a few bars of a pressure drop over 21 and 50 h. Phase 4: KHI tests in Mud PB1 with Added Clay. In an earlier report, we showed that some solids, particularly active clays, impaired the performance of KHIs such as Luvicap 55W probably due to adsorption of the KHI polymer onto the clay surface, thereby making it unavailable to perform its KHI function, or by promoting hydrate formation.7

Therefore, we carried out a series of tests using Luvicap 55W or Luvicap 21W 20k and added clay at 600 bar and 4 °C in mud PB1 (subcooling =12.2 °C). The results, with and without added clay for comparison, are summarized in Table 11. To start with, tests were carried out with Luvicap 55W but with added 5 wt % OCMA clay, supplied by Halliburton, to simulate drilling cuttings. This batch of OCMA clay was

Figure 12. KHI tests with Luvicap 21W 20k at varying pressure and concentration (1k = 1000 ppm).

Figure 13. KHI tests with Luvicap 55W and Luvicap 21W 20k at 2000 ppm (2k) and 600 bar. Table 11. KHI Experiments with Luvicap 55W with and without Added OCMA Clay at 600 bar and 4 °C in Mud PB1

Figure 11. Tests with Luvicap 55W (1000 ppm = 1k) plus TBAB (500 ppm) and varying pressure.

expt. no.

KHI

conc. [ppm]

OCMA wt %

P [bar]

ti [h]

1 2 3 4 26 27 28

Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W Luvicap 55W

6000 6000 2000 2000 6000 6000 2000

0 0 0 0 5 5 5

600 600 600 600 600 600 600

>14 29 20 18 12 29 1

Table 10. KHI Tests with Luvicap 21W 20k at Varying Pressure and Concentration at 4 °C expt. no.

KHI

conc. [ppm]

P [bar]

ti [h]

comments

17 18 19 20 21 22 23 24 25

Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W

3000 3000 2000 2000 1000 3000 3000 4000 4000

600 600 600 600 600 750 750 750 750

7 7 4 4 0.3 0 0 0 0

1 bar slow growth over 50 h 2 bar slow growth over 60 h fast growth after 16 h fast growth after 16 h fast growth after 3 h fast growth after 37 h fast growth after 33 h about 2 bar growth over 21 h about 4 bar growth over 50 h

3011

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

Figure 15. Experiments with Luvicap 21W 20k with and without added clay at 600 bar.

Figure 14. Luvicap 55W (2000 ppm = 2k) with and without OCMA clay at 600 bar.

Table 12. Experiments with Luvicap 21W 20k with and without Added Clay at 600 bar expt. no.

KHI

OCMA wt %

conc. [ppm]

P [bar]

ti [h]

comments

17 18 29 30 31

Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W Luvicap 21W

0 0 5 5 5

3000 3000 6000 10000 10000

600 600 600 600 600

7 7 3 3 6

1 bar slow growth over 50 h 2 bar slow growth over 60 h 8 bar slow growth over 50 h 5 bar slow growth over 40 h 2 bar slow growth over 40 h

not the same as we used in a previous study and was known to be more active. The clay was added as a powder as 5 wt % compared to the aqueous phase. OCMA clay is calcium montmorillonite and has low activity compared to more active clays such as Wyoming bentonite. However, the surface area of OCMA clay is probably higher than real drilling cuttings and therefore can be taken as a “worst case” for most field conditions. The KHI polymer Luvicap 55W was tested at 6000 and 2000 ppm. In one test at 6000 ppm, we observed a significant reduction in the induction time compared to tests carried out without added clay. For example, at 6000 ppm Luvicap 55W, the induction dropped from at least 14 h (and probably as high as 29 h) to about 12 h. However, in a second test with added clay, we obtained a similar result of 29 h of induction as in a test without added clay. As these results were inconclusive, to check whether the clay really was impairing the performance of the KHI polymer, we ran an experiment with only 2000 ppm Luvicap 55W. This time, the performance dropped significantly compared to earlier tests without added clay (Figure 14). Thus, we obtained an induction time of only 1 h with clay compared to 18-20 h without added clay. The effect of the clay is more clear at low KHI concentrations probably because a higher percentage of the total polymer concentration will be adsorbed. An alternative theory for the poorer result is that the clay surface is promoting hydrate formation. Tests in the next section with Luvicap 21W confirmed the trends seen with Luvicap 55W. Table 12 summarizes experiments carried out with Luvicap 21W 20k with and without 5 wt % OCMA clay at 600 bar. With 6000 ppm Luvicap 21W 20k, we obtained an induction time of 3 h before slow growth of the hydrate began. This is a worse result than using 3000 ppm Luvicap 21W 20k at the same conditions without added clay. Therefore, this confirms the results obtained with Luvicap 55W that the clay is impairing the performance of the KHI polymer. When we increased the Luvicap 21W 20k concentration to 10 000 ppm with added clay, we obtained fairly similar results

to those at 3000 ppm without clay (Figure 15). Therefore, it is possible to increase the KHI concentration to obtain better results with added clay. However, this may be practically possible because the clay (cuttings) will be removed during each drilling circulation and will take with it some of the adsorbed KHI polymer. Thus, the concentration of KHI polymer will be reduced after each circulation until it is no longer at the concentration needed for safe hydrate inhibition. In a previous study, we showed that a cheaper sacrificial polymer, such as xanthan, could possibly be added to the drilling fluid which would preferentially bind to the clay, thus keeping the KHI in solution and at its required performance level.7 This could also be done with the WBMs in this study. Additionally, a lower molecular weight polymer than Luvicap 21W 20k could be used (such as Luvicap 21W 12k used in the compatibility studies) as the performance of many KHI classes generally increases as the molecular weight is decreased to values around 1500-2000.1-3 Conclusions In this paper, we have investigated whether KHIs can be used at very high pressures, particularly for use in ultra-deepwater completion and water-based drilling fluids (WBMs) with high salinity. Compatibility tests were carried out on various brines with several commercial KHIs. We found that N-vinyl caprolactam copolymers with 50 wt % and preferentially 67 wt % N-vinyl pyrrolidone had high cloud points in high salinity brines, making them suitable for use in deepwater WBMs. Laboratory hydrate equilibrium experiments were conducted on three WBMs up to 800 bar to determine at which subcoolings the KHIs would be investigated. KHI performance tests were carried out in a 500 mL stirred titanium autoclave at pressures up to 760 bar. To the best of our knowledge, this is well above the highest reported pressure at which the performance of KHIs has been tested. With the previous knowledge that not only the subcooling but the absolute pressure (70-150 bar) affects the performance of KHIs, our results indicate that, even at the highest pressures 3012

Energy Fuels 2010, 24, 3003–3013

: DOI:10.1021/ef9016152

Kelland and Iversen

investigated, KHIs can still give substantial delays in hydrate formation as long as the subcooling is not too high and the brine salinity is high. In addition, the very high salinity appears to cause very long slow hydrate growth periods even at the highest subcooling and pressure of 760 bar. This is a feature we have never observed with this class of polyvinyllactam KHIs in low salinity brines in similar subcoolings at high (up to 700 bar) or low (