Differential Scanning Calorimetry: A New

Technique To Characterize Hydrate

Formation in Drilling Muds
Didier Dalmazzone, ENSTA, and Christine Dalmazzone and Benjamin Herzhaft, Institut Français du Pétrole

Summary cations related to this topic, and a certain number of operational

With the increasing number of deep offshore drilling operations, solutions exist. These solutions are classically based on the use of
operators and service companies are confronted by technical chal- thermodynamic inhibitor additives (mainly salt and glycol addi-
lenges ever growing in complexity. Extreme conditions encoun- tives), which displace the equilibrium point of hydrate formation.
tered at these depths require an adaptation of the drilling muds. In In deep offshore conditions (deeper than 2,000 ft), classical inhibi-
particular, the range of temperatures and pressures encountered (as tors may be ineffective, and other solutions may have to be pro-
low as –1°C and up to 400 bar) is favorable for the formation of posed. The first step is to define precisely the conditions of hydrate
gas hydrates. Gas hydrates are solid structures formed from water formation in drillings muds in order to identify the drilling phases
and gas; water contained in drilling muds under certain tempera- in which hydrate problems may occur. In this work, we present
ture and pressure conditions will form solid cages that entrap the experiments using DSC to determine the thermodynamic equilib-
gas molecules. Formation of these solid gas hydrates is liable to rium properties of methane hydrate in various media, from simple
plug kill and chokelines as well as the annular and may cause solutions to multiple-phase dispersed systems. After a review
interruption of the drilling operation and even destruction of rig of the operational procedures and the existing data related to hy-
equipment. Deep offshore drilling operators are aware of this prob- drate formation in drilling muds, the DSC technique will be de-
lem, and some operational solutions and drilling mud formulations scribed. In the third part of this paper, we will present the experi-
have been proposed and utilized, but when extreme water depths mental procedure and the first experimental results showing the
are attained, classical inhibitor solutions alone may be ineffective. possibility of characterizing gas hydrate formation in drilling muds
The usual way to determine the thermodynamic conditions of the with this technique.
formation of gas hydrates in drilling mud formulations is to use a
pressure/volume/temperature (PVT) cell. This technique requires
heavy instrumentation and often does not permit work with a State of the Art
whole formulation (especially in the presence of solids). More- Conditions of Gas Hydrate Formation and the Consequences
over, PVT cells do not give a quantitative evaluation of the kinetic of This Formation. Favorable conditions for gas hydrates forma-
properties of gas hydrate formation. tion are often encountered when drilling in deep offshore areas;
In this work, we present experiments using differential scan- very low temperatures are rapidly reached. For example, in the
ning calorimetry (DSC) to determine the thermodynamic equilib- Gulf of Mexico, it is common to have 8°C at 500 m depth and
rium properties of methane hydrate in calcium chloride solutions. 4.4°C after 900 m depth.2 As for the pressure values, 400 bar are
Additionally, thermograms of an emulsion and a complete drilling not uncommon when approaching extreme depths. If a certain
mud are presented that show the possibility of characterizing hy- quantity of gas penetrates into the well (during a gas kick situation,
drate thermodynamics in complex systems as synthetic mud for- for example), the drilling mud aqueous phase (continuous phase
mulations. This technique is easy to use, and, when correctly in- for water-based muds or dispersed phase for oil-based muds)
terpreted, it enables the dangerous zones for hydrate formation to comes in contact with the gas, leading to gas hydrate formation.
be determined. This risk is even higher when circulation is stopped. In well-
controlled situations, even if the gas leaves the reservoir at high
Introduction temperature, it may cool down to sea temperature when circulation
In recent years, deep offshore drilling operations have continu- is stopped for an extended period.3
ously set new depth records. This should continue in the near Consequences can be catastrophic; gas hydrates may plug sur-
future, particularly in the Gulf of Mexico or Gulf of Guinea areas, face equipment at the blowout preventer (BOP) level and/or in the
where numerous exploitation licenses have been taken to extreme kill and chokelines, where there is little fluid circulation. But it is
depths. A water depth approaching 3000 m is now a reality. These also possible to form hydrates in the annular, which will block the
depths are characterized by extreme conditions (poor consolidation drillstring, forcing the drilling operation to stop. When hydrates
of the surface formation, narrowness of the mud-weight window, are formed, part of the water present in the mud is consumed,
and shallow gas zones); temperatures down to –1°C and pressures which can lead to barite sedimentation or viscosity increase prob-
of up to 400 bar are not uncommon at the mudline. With these lems. Hydrate formation is, therefore, dangerous for several rea-
harsh conditions, more and more complex technical challenges sons and can halt operation. Safety problems must not be forgotten: 1
have to be overcome with adapted solutions. In particular, these m3 of hydrates can liberate 170 m3 of gas very rapidly. Propulsion of
ranges of temperatures and pressures are favorable conditions for gas hydrate plugs at very high velocity is also a hazard.
gas hydrates formation into the drilling muds. Gas hydrates are
solid structures formed from water and gas; these solid crystals Gas Hydrates in Drilling Muds: Field Cases. Deep offshore op-
may block the lines or the annular and cause serious damage to the erators are aware of this gas hydrates formation problem, as shown
rig equipment and even threaten the security of the crew.1 Opera- by the number of publications relative to this topic. However, there
tors are aware of this problem, as shown by the number of publi- are few publications reporting a field case of hydrate formation in
drilling muds during drilling operation, with most of the publications
referring to the same one case. This may be for several reasons1:
• Limited experience in deep offshore drilling.
Copyright © 2002 Society of Petroleum Engineers
• Low probability to meet all the conditions (pressure, tem-
This paper (SPE 78597) was revised for publication from paper SPE 62962, first presented perature, and quantities of gas and water).
at the 2000 SPE Annual Technical Conference and Exhibition, Dallas, 1–4 October. Original
manuscript received for review 8 November 2000. Revised manuscript received 10 Sep-
• Limited knowledge of hydrate formation phenomena may
tember 2001. Manuscript peer approved 29 January 2002. have caused some drilling problems to be attributed to other

196 June 2002 SPE Journal

causes, such as mechanical problems or barite sag, when hydrates electrolytes, however, may cause problems; they increase mud
were responsible for these problems. density, and can promote corrosion.
• Operators having faced such problems may not be inclined to An electrolyte-only solution may not be adequate to suffi-
communicate about it. ciently inhibit hydrate formation in drilling muds. It is common to
The same common reference for almost every previous publi- add methanol or glycol to the formulation. Indeed, when dissolved
cation is an article by Barker and Gomez1 describing two cases of into an aqueous phase, alcohols and glycols create hydrogen bonds
drilling stop attributed to hydrates formation. The first one was with water molecules, hindering the hydrate formation mechanism.
reported in 1977 in California at a water depth of 350 m with a The concentration necessary may be very important; according to
mudline temperature of 7°C. A gas kick caused the closure of the Quigley and Hubbard,8 50% by weight of methanol is necessary to
BOP. After control operations, chokelines and kill lines were reach the inhibition degree for a 2887 m offshore drilling. Accord-
plugged. After a cementing job and BOP recovering, gas hydrates ing to Ebeltoft et al.,3 ethylene glycol is the most efficient glycol.
were found in kill and choke lines. The second case in the Gulf of Specific problems arise from glycol use, including toxicity, safety,
Mexico was similar at 945 m depth, with a mudline temperature of and compatibility with other mud additives.
4°C. These two examples show how hydrates can cause major Kinetic Inhibitors. Classical additives are thermodynamic in-
complications in offshore drilling. hibitors. They displace equilibrium temperature and pressure con-
It is difficult to find in official literature descriptions of real ditions of hydrate stability. Kinetic inhibitors will act on the ki-
field cases; however, hazards caused by gas hydrates exist and netics of the hydrate formation reaction. Mechanisms are various,
must be evaluated. According to Barker and Gomez,1 every drill- including slowing of the hydrate formation rate and increase of the
ing operation should evaluate and analyze the gas hydrate forma- induction time for nucleation or modification of the agglomeration
tion risk, considering every shut-in period. process. As far as we know, there is at the present time no attempt
to use such inhibitors in drilling muds, though they are being
Prevention of Hydrate Formation, Operational Procedures. The studied for use in production.9 According to Ebeltoft et al.,3
possibility of forming gas hydrates in drilling muds while drilling complementary studies are necessary before efficient kinetic in-
offshore has been a known hazard for about 10 years, and some hibitors should be proposed for drilling muds.
operational solutions (in varying degrees of effectiveness) exist
and are used by operators. Existing Experimental Data of Gas Hydrate Formation Prop-
The usual prevention solution consists of thermodynamically erties. When gas hydrates are formed at constant volume, the pres-
inhibiting hydrate formation with additives in drilling mud formu- sure of the environment begins to drop (as gas is consumed by the
lation. Inhibitors have to be perfectly compatible with other drill- hydrate). Because drilling muds are too opaque to visually follow
ing mud components. Some empirical rules exist in order to define hydrate formation, pressure measurement is the basic principle of
areas in which inhibition is necessary4: the major part of the existing experimental works. The mud for-
• Water depth of ⱕ1,000 ft (305 m): Probably no hydrates. mulation is put under gas pressure in a constant volume cell. The
• Water depth of ⱕ1,500 ft (457 m): Without inhibition, risk of cell temperature is then cooled down until the hydrate begins to
hydrates formation. form, leading to a strong pressure drop and a slight temperature
• Water depth of ⱕ2,000 ft (610 m): Without inhibition, hy- rise. After formation, the temperature is raised at a very slow rate
drates formation. until hydrates completely dissociate. The point of intersection be-
• Water depth of ⱖ2,000 ft (610 m): Poor experience; electro- tween the heating curve and the cooling curve is the equilibrium
lytes inhibitors alone are not sufficient. point of dissociation. A curve pressure/temperature, which is char-
Characteristics of a good inhibitor include the following5: acteristic of the thermodynamic equilibrium of gas hydrate in the
• Lowest mud density with highest inhibition capacity. mud formulation being studied, may be plotted by repeating this
• Compatibility with classical mud components. procedure at various gas pressures.
• Compatibility with most electrolytes. This technique has been used rarely with drilling muds, espe-
• No harm for the environment. cially when solids are present, because of the necessity of adapting
• Compatibility with good shale stability. the equipment. However, several studies of the influence of vari-
Inhibitors for Oil-Based Muds. Gas hydrates are composed of ous mud components have been carried out through the years.
water and gas, so an evident solution to prevent hydrate formation Ouar et al.4 systematically studied drilling mud components and
is to lower the water proportion in drilling muds by using oil-based concluded that salt and methanol had the major thermodynamic
muds. However, such drilling muds do contain at least 20% water effect on hydrate formation. Lai and Dzialowski2 and Kotkoskie
in volume, and it is therefore still possible to form hydrates in such et al.10 have carried out the same type of studies on water-based
systems under deep offshore conditions.6 Grigg and Lynes7 have mud components. Grigg and Lynes7 have studied oil-based muds
shown experimentally that hydrates could form in oil-based muds (without solids) and found that CaCl2 and oil have a strong inhibi-
with 20% water. Furthermore, the solubility of natural gas in an oil tor effect. Ebeltoft et al.3 have confirmed these results.
phase may lead to continuous hydrate formation. According to These techniques allow the thermodynamic characteristics of
Grigg and Lynes7 and Ebeltoft et al.,3 the organic phase of a several formulations to be measured, but not the kinetic properties.
synthetic mud acts as a light thermodynamic inhibitor but in- Furthermore, it often is not possible to work with complete for-
creases the formation rate and quantity of gas hydrate formed. mulations, knowing that solids cannot be inserted in classical cells.
Another consequence of hydrate formation in oil-based muds is A new technique is reported in this paper that allows measurement
the phase separation that may result from it. Indeed, hydrate crys- of thermodynamic as well as kinetic properties of hydrate forma-
tals will form at the interface between dispersed water drops and the tion in complete drilling-muds formulations.
continuous organic phase, changing the mechanical properties of the
interfacial film, and thus may lead to breakage of the emulsion. DSC
However, the classical solution for inhibiting hydrate formation Description of the Technique. DSC is one of the most commonly
in oil-based muds is to use an active aqueous phase with 20 to 30% used techniques of thermal characterization of physico-chemical
(in weight) of electrolytes. transformations. It consists of monitoring the heat exchanges be-
Inhibitors for Water-Based Muds. The most common practice tween the sample and a reference, either vs. time at constant tem-
to inhibit gas hydrates formation in water-based muds is to use perature or vs. temperature during a heating or cooling program.
high-salinity drilling muds comprised of sodium chloride or other Several types of DSC have been designed, and many apparatus are
electrolytes. A typical formulation is, for example, a 20 to 23% commercially available. All of these share the same principle,
NaCl/polymer system. Deep offshore wells (up to 7,500 ft) were briefly described hereafter.
successfully drilled with these formulations without problems.3 As shown in Fig. 1, the sample and the reference are contained
With equivalent mass concentration, NaCl is the most efficient in identical crucibles, which can be opened to atmosphere or
inhibitor, followed by KCl, CaCl2, and NaBr. Huge quantities of sealed, depending on the experimental conditions required. Both

June 2002 SPE Journal 197

 The power registered by the calorimeter, dQ/dt, is composed of
three terms: the first one is used to compensate the differences
R S between the heat capacities of the sample (Cs) and the reference
(Cr); the second one represents the transitory phenomena caused
by the release (or absorption) of heat by the sample, R being the
thermal resistance between the sample and the heater, and RCs
being the time constant of the cell containing the sample. The third
one represents the production of heat by the sample (h).
e = ε. δT Applications
R : reference cell S : sample cell This technique can be used in three types of applications12:
• Thermodynamics: thermodynamic applications include the
Fig. 1—Section of a plate DSC. measurement of heat capacities, transitions (first order transitions,
melting, crystallization, polymer transitions, electric and magnetic
transformations, order-disorder transitions, glass transitions, etc.),
crucibles are placed in a thermoregulated furnace that imposes the purity determinations, and the study of dispersed phases (ther-
temperature program. The variety of devices mainly differs by the moporosimetry, emulsions, etc.).
way thermal exchanges are measured. In the simplest DSCs, a • Kinetics: different types of measurements can be performed
thermocouple is used to measure the temperature difference be- using the relation between temperature, time and degree of ad-
tween the two crucibles at one point of their walls (generally the vancement of a reaction (isothermal studies, and measurement of
bottom). The heat exchange rate is derived from this temperature kinetics with a constant scanning rate or with different scanning
difference by calibration. rates). It is then possible to determine the order of the reaction and
The more complex DSC devices use the Calvet principle to the activation energy.
very precisely measure the heat exchanges. The two cylindrical • Analytical: the calorimetric signal can be linked in various
crucibles are placed into two independent heat sensors, formed cases to the transformation of a particular component. The mea-
from a large number of thermocouples surrounding the crucible. A surement of the corresponding energy allows the determination of
section of one of these sensors is shown in Fig. 2. Each thermo- the component mass. DSC is used, for example, to characterize
couple measures the temperature difference between the crucible silica in cements, polymorphic forms in pharmaceuticals, and dif-
and the furnace in the radial direction. This temperature difference ferent forms of polymers, and it can also be easily used for the
can be related to the local heat flow dqi/dt exchanged between the characterization of complex fluids such as diesel oils, bitumen, or
crucible and the furnace by crude oils.
␧ dqi
ei = , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ( 1) Thermogram Analysis. The baseline is the thermogram obtained
␭ dt in the absence of any thermal phenomenon. The form of the base-
where ei⳱the electrical force delivered by the thermocouple i, line depends on the evolution of the sample heat capacity. The
␣⳱its thermoelectrical constant, and ␭⳱the thermal conductivity surface area of the corresponding peak measures the energy of a
of the sensor material. All the thermocouples are connected so that thermal phenomenon. The study of the melting of pure com-
the sensor delivers a total electrical force E related to the overall pounds, for which the melting enthalpies are known, allows a
heat exchange rate dQ/dt by suitable calibration of the calorimeter.
Contrary to differential thermal analysis, temperatures of the
␧ dqi ␧ ␧ dQ
兺 e = 兺 ␭ dt = ␭ 兺 dt =
dqi samples are not directly recorded with DSC. The temperature cali-
E= . . . . . . . . . . . . . . . . . ( 2)
i
i
i i ␭ dt bration can be performed by the study of pure compounds melting.
The difference between the sample temperature Ts and the program
The differential measurement is achieved by connecting the refer- temperature Tp is related to the heating rate, the rate of heat pro-
ence and sample sensors in opposition. The exact relation between the duction dQ/dt, and the thermal properties of both the crucible and
heat flow and the electric power recorded is obtained by calibration. DSC sensor by the following equation:
Fundamental Equation. The peak area of the recorded signal dTp dQ
(generally called “thermogram”) leads to the heat exchange during Tp − Ts = RCs − R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ( 4)
dt dt
the thermal phenomenon. The fundamental equation of DSC is
the following11: Application to Hydrate Characterization. Virtually any
2 physico-chemical phenomenon featuring an enthalpy change
dQ dTp d Q dh
= 共Cs − Cr兲 − RCs 2 − . . . . . . . . . . . . . . . . . . . . . . ( 3) (chemical reaction, state transition, melting, or multiple phase
dt dt dt dt equilibrium) may be evidenced and measured by DSC. In most
cases, however, the extent of potential applications is limited by
the design of the apparatus. Because few DSC can be operated
under a controlled atmosphere, and even fewer under pressure,
the application of this technique to the characterization of hydrates
has been mostly limited to model hydrates, which can be syn-
thesized under atmospheric pressure. The study by DSC of hy-
drates of gases, requiring pressures of several tens of bar, have not
been reported.
State of the Art. Very few papers on the study of hydrates by
calorimetry can be found in the literature. Handa13–15 has devel-
oped a calorimetric technique in order to determine compositions,
dissociation enthalpies, and heat capacities of different hydrates
(xenon, krypton, methane, propane, ethane, isobutane, and natural
gas hydrates). A Calvet calorimeter from SETARAM (SETARAM
Scientific & Industrial Equipment, Caluire, France) was used for
this purpose.
Fig. 2—Section of a Calvet-type DSC showing one of the Koh et al.16 from the King’s College of London have recently
two sensors. used DSC in order to test hydrate inhibitors. The study was per-

198 June 2002 SPE Journal

 Junction for nitrogen sweeping
Junction for pressure inlet

DSC furnace

Thermal insulator
Crucible

Heat flow sensor

Crimped plug Fig. 4—View of the complete apparatus.

Nitrogen evacuation
hydrate formation (or dissociation). The pressure was measured
Fig. 3—Section of a controlled pressure cell placed in the using a precision manometer graduated from 0 to 100 bar, with a
DSC111 furnace. resolution of 1 bar. The reference crucible was empty; 20 to 50
microgram samples were admitted into the sample crucible by
formed at atmospheric pressure on hydrates of tetrahydrofuran. means of a syringe.
The researchers measured supercooling degrees for different types Calibration. DSC requires not only careful calibration of the
of inhibitors, and they also performed isothermal studies in order heat flow sensor, but also precise temperature calibration, because
to observe crystallization of THF hydrates as a function of time. the sample temperature is never measured directly. It can only be
Some TTT curves (Time-Temperature-Transformation) were inferred by correcting the temperature measured in the DSC fur-
therefore plotted that allow the comparison of inhibitor properties nace, using a calibration curve obtained by scanning the melting of
on the kinetics of hydrate formation. substances the melting points of which are perfectly known. The
Hydrate formation was observed by DSC for the first time by best precision is obtained by calibrating the DSC with the cells
Fouconnier et al.,17 who studied the hydrate of trichlorofluo- being used and within the temperature range of interest for the
romethane in water-in-oil emulsions stabilized by a nonionic sur- measurements. For optimal accuracy, the influence of the heating
factant (Berol 26). Hydrates of trichlorofluoromethane, whose ki- rate also must be accounted for by the calibration.
netics of formation under the same conditions were investigated by We used a heating rate of 2°C/min in this study, and the phe-
Jakobsen et al.18 using dielectric spectroscopy, have the advantage nomena of interest occurred in the range of −50 to +10°C. Cali-
of being stable at atmospheric pressure. It was shown that the bration was thus performed at heating rates of 0.25, 0.5, 1, 2, and
amount of hydrate formed depends on the sample volume, the 4°C/min, using mercury (melting point −38.8°C, 99.99999% in
temperature reached during the cooling step, and the time. These purity) and gallium (melting point +29.8°C, 99.9999% in purity),
results confirmed the extreme importance of the hydrate formation leading to a very precise temperature correction under the condi-
process. It was also shown that ice crystals formed during cooling tions used for measurements.
could act as nucleating agents for hydrate. Furthermore, two poly- As a test, the pure water-melting peak was plotted, showing a
morphic forms of hydrates seem to coexist in the emulsion. temperature accuracy of 0.1°C and a heat accuracy of 1%, at any
Description of the Apparatus. For the study reported here, a heating rate from 0.25 to 4°C/min. Also, the melting of the water-
DSC111 (SETARAM) fitted with controlled pressure cells was used. calcium chloride eutectic was invariably recorded at 52°C, in good
The DSC111 works on the Calvet principle described above, and is agreement with known data.
well known as one of the most precise DSCs available. Its furnace can Hydrate Dissociation Equilibria in Water and Calcium Chlo-
be cooled down to –120°C by circulation of nitrogen vapor. ride Solutions. The thermograms reported in Fig. 5 were obtained
Fig. 3 shows a section of one cell. It is composed of a steel by warming up 50-␮m samples of distilled water at 2°C/min, after
cylindrical crucible having a capacity of 0.27 ml, connected at one rapid cooling down to −30°C to allow water crystallization, under
end to a thin steel tube ended by a junction and tightly sealed at the variable methane pressure. At lower pressures, only one peak can
other end by a steel plug crimped with an aluminium joint. When be seen, characterizing the ice melting. At pressures greater than
positioned into the DSC sensor, the crucible takes place exactly 60 bar, a second peak appears distinctly at a temperature that
within the heat flow sensitive region, while the junction of the tube increases rapidly with pressure, which we attributed to the hydrate
stands out of the furnace. A second junction is provided to sweep dissociation. The slight shift of the ice melting peak upon pressure
the sensor with nitrogen during the analysis, in order to avoid any increase of approximately −0.008°C per bar is in agreement with
water condensation at low temperature. A view of the complete the known pressure dependence of the ice melting point. At suf-
apparatus is given in Fig. 4. Methane pressure was applied to both ficiently high pressures, above 80 bar, the two peaks are totally
the sample and the reference cells via a pressure control panel disjointed, and it is possible to determine precisely at what tem-
mainly composed of a 0.4 L gas holder to compensate for any perature the hydrate dissociation begins. The equilibrium points
pressure variations owing to gas absorption (or desorption) during measured at several pressures are reported in Table 1. They are

June 2002 SPE Journal 199

 10
TABLE 1ÑTEMPERATURE OF DISSOCIATION OF METHANE

0 HYDRATE VS. PRESSURE IN PURE WATER AND VARIOUS

CALCIUM CHLORIDE SOLUTIONS

Ð10
CaCl2 CaCl2 CaCl2
Heat Flow, mW
Ð20
Pure Water 10 wt% 20 wt% 23 wt%
Ð30
T P T P T P T P
Ð40 (¼C) (bar) (¼C) (bar) (¼C) (bar) (¼C) (bar)
Ð50 7.1 50.0 4.8 60.0 Ð3.1 63.3 Ð9.2 62.0
Ð60 8.3 60.0 6.7 80.1 Ð1.9 74.5 Ð6.7 79.0
Ð70
9.9 70.0 8.2 100.0 Ð0.9 85.6 Ð5.8 90.0
Ð2 0 2 4 6 8 10 12 14 16 18 11.0 80.1 5.2 68.8 0.4 102.9 Ð5.4 99.5
Temperature, ¼C 11.4 90.0 6.8 77.8
1 bar 60 bar 93 bar 9.1 102.9
Fig. 5—Thermograms of pure water at variable methane pressure.
together with various solid compounds. As the mixture composi-
tion becomes more complex, the thermogram complexity also in-
represented in Fig. 6, where they can be compared with a large creases. Although some of the phenomena evidenced by these
number of experimental points from the literature.19 thermograms are not yet fully interpreted, the following observa-
Fig. 7 reports thermograms obtained under the same conditions tions can be made.
with a 20 wt% calcium chloride solution cooled down to −60°C. The hydrate peak surface, which is directly related to the
The aqueous phase melting appears to be divided into two peaks: amount of hydrate formed, is much larger in emulsion than in the
the eutectic melting, at −52°C, and the progressive melting of ice bulk. This is easily explained by the fact that methane is very
in equilibrium with the saline solution. At sufficient methane pres- poorly water soluble, and thus the hydrate appears only at the
sure, a third peak appears then at a temperature which is pressure- surface of the aqueous phase. In the bulk, this surface is limited to
dependent, and we attributed this to the hydrate dissociation. Be- the liquid/gas interface, while in water-in-oil emulsions, the meth-
cause the total water phase melting ends at very low temperatures ane is highly concentrated in the oil phase and available to form
in such concentrated solutions, it is possible to distinguish the hydrate at the surface of each water droplet.
hydrate peak at methane pressure as low as 25 bar. These mea- The hydrate dissociation in emulsions is not a monothermal
surements were repeated with 10%, 20%, and 23 wt% calcium transformation as in the bulk phase, but takes place over quite a
chloride solutions. The results are presented in Table 1 and plotted large range of temperatures. This indicates that the equilibrium
in Fig. 6. conditions vary during the dissociation. As the pressure is kept
Because there is no means of agitating the sample inside the constant during the experiments, the only possible variable is the
crucible, no quantitative measurement of the amount of hydrate salt concentration. During the cooling of the emulsion, large
formed could be done. This is obviously the main limitation of the amounts of hydrate are formed at the surface of each droplet,
technique. However, this is not a problem for equilibrium mea- consuming large amounts of water and causing the salt to concen-
surements, provided that the sample volume is small enough, and trate in the remaining water. Upon heating, the hydrate, in the
the heating rate slow enough, to ensure the sample homogeneity. presence of such a concentrated salt solution, begins to dissociate
The accuracy of the hydrate equilibrium point measured by DSC is at a very low temperature. The salt solution is then progressively
validated, at least in pure water, by the excellent agreement with diluted by the water released, and it finally returns to the initial
literature data. Unfortunately, we could not find any data on con- concentration when the hydrate is totally dissociated. In that case,
centrated calcium chloride solutions to compare with our results. the dissociation temperature should not be measured at the peak
Hydrate Dissociation Equilibria in a Water-in-Oil Emulsion onset, which is almost impossible to determine, but at the peak
and in a Drilling Mud. Fig. 8 reports a thermogram obtained maximum, which indicates the dissociation of the last particle
under the same conditions as above with a water-in-oil emulsion of hydrate.
sample at 90 bar. The aqueous phase is a 23 wt% calcium chloride
solution. Fig. 9 shows a thermogram obtained at 95 bar with a
drilling mud composed of the same emulsion as the one above, 10

120 0

110 Ð10
Heat Flow, mW
Methane Pressure, bar

100 Pure water: literature Ð20

90 Pure water: this work

CaCl2 10 wt% Ð30
80 CaCl2 20 wt%
CaCl2 23 wt% Ð40
70
60 Ð50

50 Ð60
Ð60 Ð50 Ð40 Ð30 Ð20 Ð10 0 10 20
40
Ð10 Ð5 0 5 10 15 Temperature, ¼C

Dissociation Temperature, ¼C 1 bar 25 bar 102 bar

Fig. 6—Thermograms of a 20 wt% CaCl2 solution at variable Fig. 7—Hydrate dissociation temperature vs. methane pressure
methane pressure. in various solutions.

200 June 2002 SPE Journal

 3 • Isothermal studies: hydrate formation at a fixed temperature;
Ice melting determination of the induction time and plotting of the TTT curves.
2 Hydrate • During heating at constant scanning rate: evolution of the
dissociation amount of hydrates formed after different thermal treatments (iso-

Heat Flow, mW
1 therms, temperature gradients, quenching, etc.).
These studies could be performed with water- or oil-based
0 muds. Nucleation studies (determination of induction time and
degree of supercooling) are always difficult. Oil-based muds,
Ð1 which are basically water-in-oil emulsions, allow a statistical study
Eutectic of nucleation in the aqueous phase with a unique mud sample: as
Ð2 melting a matter of fact, each water droplet in the emulsion represents a
Ð3 sample.20 In the case of water-based muds, several experiments
Ð60 Ð50 Ð40 Ð30 Ð20 Ð10 0 10 20 can be necessary on the bulk sample in order to define the range of
temperatures at which nucleation occurs. It is noteworthy that
Temperature, ¼C
these constraints are specific to all types of nucleation studies, and
not especially to DSC. Another interesting study could involve the
Fig. 8—Thermogram of a water-in-oil emulsion at 90 bar. characterization of the oil-based muds’ stability after hydrates dis-
sociation in order to predict the emulsion breakage. The principle
These observations point out the need for further investigations of this kind of study is to measure the crystallization temperature
in order to fully elucidate the gas hydrate formation process in of water droplets during cooling at constant scanning rate before
complex mixtures. and after hydrate dissociation: the lower the temperature, the more
stable the emulsion.21
Conclusion and Perspectives Using this rapid and versatile technique, these experimental
DSC is a rapid and versatile technique that can be efficiently used studies as a whole should permit the rapid determination of dan-
for analytical, thermodynamic, and kinetic studies. Specific gas- gerous zones for hydrate appearance as a function of mud type in
tight pressure-controlled cells, which can accept any type of deep offshore drilling.
sample such as dense and viscous fluids that contain solids, are
commercially available. The present work has shown that DSC Nomenclature
could be a very useful tool for hydrate formation studies, espe- Cr ⳱ heat capacity of the reference
cially in complex fluids such as drilling muds. Hydrate formation Cs ⳱ heat capacity of the sample
in oil-based drilling muds has been evidenced. dh/dt ⳱ rate of enthalpy change of the sample
Some difficulties remain, however, among which the lack of dQ/dt ⳱ rate of heat production
agitation of the sample appears to be the most limiting. Conse- ei ⳱ electrical force delivered by the thermocouple i
quently, complementary calorimetric experiments can be rationally R ⳱ thermal resistance between the sample and the heater
considered, using two kinds of calorimeter:
RCs ⳱ time constant of the cell containing the sample
1. A rapid calorimeter such as DSC 111 (SETARAM) that allows Tp ⳱ program temperature
analyses under pressure with samples from 10 to 100 mg.
Ts ⳱ sample temperature
2. A “Calvet” calorimeter such as BT 215 (SETARAM) that ac-
cepts samples of a few grams and can be equipped with specific ␣ ⳱ thermoelectrical constant of thermocouple i
cells with gas circulation under pressure. This circulation of gas ␭ ⳱ thermal conductivity of the sensor material
should favor gas diffusion and agitation of the whole sample.
In addition, several methane hydrate dissociation points in so- Acknowledgments
dium chloride solutions will be determined, in order to allow fur- The authors thank V. Lachet and A. Audibert (Institut Français du
ther comparison with well established literature data. Pétrole) for fruitful discussions. The contribution of D. Clausse is
The following experimental studies can be therefore envisaged also gratefully acknowledged. This study has been realized
for different mud formulations (water- or oil-based muds) and through collaboration with TotalFinaElf. The authors wish to thank
different additives (thermodynamic or kinetic inhibitors): the ARTEP association and TotalFinaElf for permitting the pub-
1. Determination of dissociation points of hydrates as a function of lication of this work.
pressure (measurements under thermodynamic equilibrium).
2. Kinetic measurements. References
• During cooling at constant scanning rate: determination of 1. Barker, J.W. and Gomez, R.K.: “Formation of Hydrates During Deep-
the maximum degree of supercooling and, eventually, kinetics of water Drilling Operations,” JPT (March 1989) 41; Trans., AIME, 287.
hydrate formation after nucleation. 2. Lai, D.T. and Dzialowski, A.K.: “Investigation of Natural Gas Hy-
drates in Various Drilling Fluids,” paper SPE 18637 presented at the
1989 SPE/IADC Drilling Conference, New Orleans, 28 February–3
6
March.
4 3. Ebeltoft, H., Yousif, M., and Soergaard, E.: “Hydrate Control During
Deep Water Drilling: Overview and New Drilling Fluids Formations,”
Heat Flow, mW

2 paper SPE 38567 presented at the 1997 SPE Annual Technical Con-
ference and Exhibition, San Antonio, Texas, 5–8 October.
0 4. Ouar, H. et al.: “The Formation of Natural Gas Hydrates in Water-
Based Drilling Fluids,” Trans I. Chem. E. (1992) 70A, 48.
Ð2 5. Halliday, W., Clapper, D.K., and Smalling, M.: “New Gas Hydrate
Inhibitors for Deepwater Drilling Fluids,” paper SPE 39316 presented
Ð4 at the 1998 IADC/SPE Drilling Conference, Dallas, 3–6 March.
6. Szczepanski, R. et al.: “Research Provides Clues To Hydrate Forma-
Ð6 tion and Drilling-Hazard Solutions,” Oil and Gas J. (March 1998) 52.
Ð60 Ð50 Ð40 Ð30 Ð20 Ð10 0 10 20 7. Grigg, R.B. and Lynes, G.L.: “Oil-Based Drilling Mud as a Gas Hy-
Temperature, ¼C drate Inhibitor,” SPEDE (March 1992) 32.
8. Quigley, M.S. and Hubbard, J.C.: “Gas Hydrates Formation in Drilling
Fig. 9—Thermogram of a drilling mud at 95 bar. Fluids,” paper presented at the 1990 Drilling Managers Technical

June 2002 SPE Journal 201

 Meeting at Mobil Research and Development Corp., Dallas Research 20. Clausse, D.: “Research techniques utilizing emulsions,” Encyclopedia
Laboratory, 6–7 March. of emulsion technology, P. Becher (ed.), Dekker Publishers, New York
9. Sinquin, A.: “A Global Research Program on Hydrates Control: Role of City (1985) Vol. 2, 77.
Kinetic Inhibitors,” Proc., Third International Conference on Gas Hy- 21. Dalmazzone, C.: “Use of the DSC technique to characterize water-in-
drates, Park City, Utah, 18–22 July (1999). crude oil emulsions stability,” Proc., Second World Conference on
10. Kotkoskie, T.S. et al.: “Inhibition of Gas Hydrates in Water-Based Emulsion, Bordeaux, France, 23–26 September (1997) Vol. 2, 1–69.
Drilling Muds,” SPEDE (June 1992) 130.
11. Gray, A.P.: “A Simple Generalised Theory for the Analysis of Dy-
namic Thermal Measurements,” Analytical Calorimetry, R.J. Porter SI Metric Conversion Factors
and J.F. Johnson (eds.), Plenum Press, New York City (1968) Vol. 1, Bar × 1.0* E+05 ⳱ Pa
209. ft × 3.048* E–01 ⳱ m
12. Claudy, P.: “Differential Calorimetric Analysis (DSC)-Application to °F (°F – 32)/1.8 ⳱ °C
Chemistry,” L’Actualité Chimique (March 1999) 13.
*Conversion factor is exact.
13. Handa, Y.P.: “Compositions, Enthalpies of Dissociation and Heat Ca-
pacities in the Range 85 To 270 K for Clathrate Hydrates of Methane,
Ethane, Propane, and Enthalpy of Dissociation of Isobutane Hydrate, Didier Dalmazzone is currently an associate professor at ENSTA
As Determined by a Heat-Flow Calorimeter,” J. Chem. Thermodynam- (Ecole Nationale Supérieure de Techniques Avancées, Paris,
ics (1986a) 18, 915. France) and a researcher at the Chemistry and Process Labo-
ratory. e-mail: [email protected]. His research interests in-
14. Handa, Y.P.: “Calorimetric studies of laboratory synthesized and natu- clude thermochemistry, combustion, calorimetry, and process
rally occurring gas hydrates,” Proc., 1986 AIChE Annual Meeting, safety. Dalmazzone holds a degree in chemical engineering
Miami Beach, Florida, 2–7 November. from ESCOM, Paris, and a PhD in process engineering from the
15. Handa, Y.P.: “A calorimetric study of naturally occurring gas hy- U. of Technology of Compiègne, France. Christine Dalmaz-
drates,” Ind. Eng. Chem. Res. (1988) 27, 872. zone has been a research engineer at Inst. Français du Pétrole
16. Koh, C.A. et al.: “Low dosage natural gas hydrate inhibitor evalua- (IFP) since 1991. e-mail: [email protected]. Her re-
search interests include the physicochemical characterization
tion,” Proc., 1998 International Gas Research Conference, San Diego,
of complex fluids in drilling and production: oil-based muds,
8–11 November (1998) Vol. I, 194–200. hydrates formation in drilling fluids, crude oil emulsions, and
17. Fouconnier, B. et al.: “Formation of trichlorofluoromethane hydrate in crude oil/water separation. Dalmazzone holds a PhD in chemi-
w/o emulsions studied by DSC,” Progr. Colloid Polym. Sci. (1999) cal engineering from the U. of Technology of Compiègne
112, 105. (France). Benjamin Herzhaft has been a research engineer at
18. Jakobsen, T., Sjöblom, J., and Ruoff, P.: “Kinetics of gas hydrate IFP since 1997. e-mail: [email protected]. His research
formation in w/o emulsions: The model system trichlorofluoromethane/ interests include research projects on drilling-muds monitoring
and fluids for mature fields. He is coauthor of several SPE
water/non-ionic surfactant by means of dielectric spectroscopy,” Col-
papers, mainly on characterization of drilling fluids for dif-
loids and Surfaces, A: Physicochemical and Engineering Aspects ferent applications as foams for underbalanced drilling, the
(1996) 112, 73. rheology of oil-based muds, and hydrates in drilling muds.
19. Sloan, E. Jr.: Clathrate Hydrates of Natural Gases, Marcel Dekker Herzhaft holds a PhD degree in physics engineering from the U.
Publishers, New York City (1990). of Paris IV.

202 June 2002 SPE Journal