ABSTRACT
Currently, the most common method for preventing hydrates in the oil and gas industry is the
injection of thermodynamic inhibitors (e.g., methanol and ethylene glycol). The total inhibitor
requirement can be calculated based on the amount required for treating the aqueous phase, plus
the amount of inhibitor lost to the hydrocarbon phase(s). Although a safety margin is usually
considered in this calculation, hydrates may still form in case of unexpected events. It is not
uncommon to use a large safety margin to reduce the risk of hydrate blockage. Excessive use of
an inhibitor has environmental and economical consequences and can result in significant
increase in capital and operational expenditures, in particular at high water-cut conditions.
While the effectiveness of thermodynamic inhibitors have been well proven and modelling of
hydrate thermodynamic properties in the presence of these inhibitors have reached a good degree
of reliability, only few researchers have addressed inhibitor optimisation and monitoring. This
paper presents a simple method for monitoring and optimising thermodynamic hydrate inhibitors
dosage in various production systems. The method is based on predicting the hydrate suppression
temperature from the measured water activity values of the downstream aqueous phase. Water
activity of the aqueous phase is measured using a simple commercial water activity meter.
Initially, a correlation was developed relating water activity values to the hydrate suppression
temperature using our in-house thermodynamic model. Next a large dataset was generated by
measuring water activities of a wide variety of aqueous solutions containing various salts and/or
organic inhibitors over a wide range of concentrations. The reliability of the above correlation
was examined by comparing its predictions with the actual measured water activity data. Finally,
the hydrate stability zones predicted by this approach were compared with independent
experimental data. The results are in good agreement, demonstrating the reliability of this
method.
Keywords: gas hydrates, kinetic inhibitors, thermodynamic inhibitor, water activity, phase
boundary, correlation
NOMENCLATURE INTRODUCTION
aw water activity Formation of hydrate is a problem to the oil and
C pressure effect constant gas industry because it can eventually block
P pressure [bar] pipelines and stop production and also leads to
T hydrate equilibrium temperature [oC] safety hazards to production/transportation
T0 hydrate equilibrium temperature of distilled systems and substantial financial losses.
water [oC] Furthermore, restarting the facility, especially in
T hydrate suppression temperature [oC] offshore environments, can be a long and costly
process since the remediation and dissociation of economical consequences. Excessive use of an
hydrate plugs is difficult to achieve. inhibitor has environmental and economical
Conventional hydrate flow assurance methods consequences and can result in significant increase
focus on preventing hydrate formation either by in capital and operational expenditure, in particular
modifying one of the elements required for hydrate at high water-cut conditions. While the
formation (i.e., dehydration of natural gas or effectiveness of thermodynamic inhibitors have
lowering the operating pressure) or by injecting been well proven and hydrate thermodynamic
thermodynamic hydrate inhibitors which shift properties in the presence of these inhibitors have
hydrate phase boundary away from the operating reached a good degree of reliability, only a limited
conditions. Thermal insulation and active heating effort has been devoted for optimizing and
methods are also used to increase the pipeline monitoring the concentration of inhibitors in
temperature outside the hydrate risk zone. pipelines [10, 11].
However, thermal methods are inherently very This paper presents a simple method for
expensive and are economical only for short monitoring and optimising thermodynamic hydrate
distances. New flow assurance methods are inhibitors dosage in the system. The method is
moving towards delaying hydrate formation and/or based on predicting the hydrate suppression
preventing their agglomeration by means of Low temperature of fluid from measured water activity
Dosage Hydrate Inhibitors (LDHIs) which are value of the downstream aqueous phase. Water
divided into Kinetic Hydrate Inhibitors (KHIs) and activity of the aqueous phase at constant
Anti-Agglomerants (AAs) respectively, the detail temperature depends on the concentration of
can be found in open literature [1], [2], [3]. thermodynamic inhibitors and salt present in the
The most common flow assurance method used in system, therefore monitoring of this parameter can
the oil and gas industry is injection of provide an indication of how far the system is
thermodynamic hydrate inhibitors, i.e. methanol from hydrate phase boundary. These inhibitors and
(MeOH) and mono ethylene glycol (MEG) at salts are able to bind to water molecules through
wellhead. Successful inhibition with methanol or intermolecular bonds, thus preventing bound water
glycol typically requires inhibitor concentrations molecules from forming hydrogen bonds with
from 15 to 50 mass% (or even higher) in the each other and with the remaining free water.
aqueous phase. Methods for calculating the More specifically, these compounds reduce the
required inhibitor dosage are based on determining free-water concentration, which is considered as
the hydrate phase boundary for the fluid under water activity in the system. Water activity of the
consideration either experimentally or using aqueous phase in this work is measured using a
thermodynamic models. The appropriate injection simple commercial device, which is much easier
rate for each inhibitor to treat the free water phase than measuring hydrate dissociation point and can
depends on how far the phase boundary needs to significantly reduce the time and cost of the
be shifted for a specific system at a given pressure. experiments. A general correlation has been
The required inhibitor dosage (injection rate) to developed to relate the measured water activities
prevent hydrate formation is the amount of values to the hydrate suppression temperate by
inhibitor required to treat the free water, plus the generating pseudo-experimental data using a well-
amount of inhibitor lost to the vapour phase and proven thermodynamic model [5] [4]. To examine
the hydrocarbon liquid phase. Although a safety the reliability of the model in predicting water
margin is usually considered in this calculation, activity, experimental water activity data were
hydrates still form, which can be due to the measured using LabMaster-aw water activity
problems with injection pumps, a sudden increase meter. The water activity meter measures water
in the water-cut, a change in fluid flow rate, activity from change of humidity of the air around
emergency shut-downs, problems associated with the sample using a resistive electrolytic humidity
start-ups, lack of information on the composition sensor. A large database of different salt and
of produced water, changes in the system alcohols (methanol and MEG) solutions over a
temperature and pressure, etc. To overcome these wide range of concentrations has been created to
problems, in most cases too much inhibitor is used develop the correlation.
because if hydrate forms due to insufficient
inhibitor dosage, the problems caused by the THERMODYNAMIC MODELLING
formation of hydrates will have much greater
A general phase equilibrium model based on the
uniformity of component fugacities in all phases is 50
used to predict the water activity and the hydrate
suppression temperature of aqueous solutions of 40
T / C
30
o
thermodynamic model can be found elsewhere [4].
In summary, the statistical thermodynamics model
20 50 bar
100 bar
uses the Valderrama modification of the Patel and 10 200 bar
is modelled using the solid solution theory of van Figure 1 Hydrate suppression temperature vs
der Waals and Platteeuw [6], as developed by logarithm of the water activity
Parrish and Prausnitz [7]. The equation
recommended by Holder [8] is used to calculate
the heat capacity difference between the empty As can be seen from Figure 1, all the points
hydrate lattice and pure liquid water. The Kihara closely fit to a line suggesting a linear relation
model for spherical molecules is applied to between the hydrate suppression temperature and
calculate the potential function for compounds the Neperian logarithm of water activity of the
forming hydrate phases [9]. aqueous phase. To take into account the effect of
pressure on the correlation, an equation has also
CORRELATION DEVELOPMENT been developed as shown in Figure 2
A general correlation has been developed to relate
the measured water activities values to the hydrate P/ bar
suppression temperate by generating pseudo- 0 50 100 150 200 250 300
-70
experimental data using the above described
thermodynamic model [5]. Using pseudo- -72
0.9
a) NaCl
water activity /aw
0.8
1
0.7
0.95
0.7
0.6
b) KCl
0.5
water activity /aw
1.0
0.4
0.7 HWHYD aw @ 25 C
0.2
0.6
0.1
0.5
0.0
0.4 0 20 40 60 80 100
Measured aw @ 25 C
0.3 ethylene glycol /mass%
0.2 HWHYD aw @ 25 C
0.8
water activity /aw
0.7
0.6
0.7
0.5
0.4 0.6 Measured aw @ 25 C
Measured aw @ 25 C
0.3 HWHYD aw @ 25 C
HWHYD aw @ 25 C 0.5
0.2
0.1
0.4
0.0 0 10 20 30 40 50 60 70 80
0 20 40 60 80 100
ethylene glycol /mass%
ethylene glycol /mass%
Figure 6. Water activity of the real produced
water B in the presence of MEG. The solid line
presents the predicted water activity values.
c) MEG + Produced water F
0.8 500
450 10 mass% (NG1)
P / bar
0.6
0 20 40 60 80 100 250
methanol /mass% 200
150
100
50
b) MeOH + NaCl 0
-20 -10 0 10 20 30
1.0 T / oC
methanol /mass%
In this work a simple and robust
Figure 7. Water activity of methanol aqueous correlation has been developed based on
solutions in presence of salt (NaCl). Solid lines pseudo experimental data generated by an
present the model prediction. in-house thermodynamic model.
This correlation could be used for systems
containing different salt and alcohol
As can be seen in Figure 7, in the very low MeOH (methanol and MEG).
concentration the measured water activities are not The model and the data have been
far away from the predictions. However, for the validated by a wide range of experimental
concentration greater than 5 mass% the data generated in this work and the results
measurements are not correct. In the systems with show a very good agreement.
salt present, large deviations are observed between
The developed correlation has been
the measurements and the predictions over the
validated by comparing the predictions
entire methanol range. Based on these results it is
with literature data and the results show a
concluded that the existing sensor of the
very good agreement between correlation
LabMaster device cannot measure the water
and experimental data.
activity of methanol aqueous solutions.
An executable program has been
developed which directly read the water
Validation of the developed correlation
activity data and calculate the hydrate
phase boundary using the correlation.
The developed correlation was validated by
This developed system could be used for
comparing the natural gas hydrate phase boundary
monitoring hydrate phase boundary.
correlated by Equations 1 and 2 with the
experimental data [12] and the model prediction
REFERENCES
data. The compositions of the natural gases used
[1]Behar, P., Kessel, D. and Sugier, A. Advances Table A.1 Composition of Produced waters
in Hydrate Control. 70th gas processors
Composition (mass%) of saline
association conference, San Antonio, 11-12 March water
1991. Salt (mass% )
[2]Frostman, L. M.,. Anti-Agglomerant Hydrate M P F S
Inhibitors for Prevention of Hydrate Plugs in NaCl 9.000 3.316 6.993 2.231
Deepwater Systems. SPE Annual Technical KCl 0.000 4.468 0.066 0.037
Conference and Exhibition. Dallas, Texas, 14 CaCl2 0.800 0.174 0.735 0.133
October 2000 MgCl2 0.200 0.014 0.186 0.041
Na2SO4 0.000 0.609 0.000 0.000
[3]Kelland, M. A. History of the Development of SrCl2 0.000 0.000 0.099 0.016
Low Dosage Hydrate Inhibitors. Energy and Fuels BaCl2 0.000 0.000 0.036 0.011
2006, 20, pp.825847 Total Salt 10.000 8.581 8.115 2.469
[4]Avlonitis, D. Prediction of VL and VLL
Equilibria of Mixtures Containing Petroleum
Reservoir Fluids and Methanol With a Cubic EoS,
Fluid Phase Equilibria 1994, 94, 181-216. Table A.2 Composition of natural gases
[5]Valderrama, J. O. A Generalized Patel-Teja
Equation of State for Polar and Nonpolar Fluids NG1 NG2
Component
and Their Mixtures, Journal of Chemical /Mole% /Mole%
Engineering of Japan, 1990 23(1), 87-91. C1 88.21 88.3
[6]Van der Waals. Clathrate Solutions, Advances C2 5.78 5.40
in Chemical Physics, 1959 2, 1-57. C3 1.78 1.50
[7]Parrish, W.R. Dissociation pressure of gas iC4 0.19 0.20
hydrates formed by gas mixtures, Industrial and nC4 0.30 0.30
iC5 0.06 0.10
Engineering Chemistry Proc. 1972 Des. Dev., 11,
nC5 0.07 0.09
26-35. C6 0.06 0.00
[8]Holder, G.D. Multiple-phase equilibria in C7 0.01 0.00
hydrates from methane, ethane, propane, and N2 1.40 2.39
water mixtures. AIChE J. 1982, 28, 440-447. CO2 2.15 1.72
[9]Kihara, T., Virial Coefficient and Models of Total 100.00 100.00
Molecules in Gases, Reviews of Modern Physics,
1953 25(4), 831-843.
[10]Tohidi, B., Chapoy, A. and Yang, J.,.
Developing a Hydrate-Monitoring System. SPE
Project, Facilities & Construction Journal, 4 (No.
1), pp:1-6 2009 .
[11]Tohidi, B., Yang, J. and Chapoy, A.,. Hydrate
Monitoring System. International Pat. WO
2010/004289 A2 2010.
[12] Haghighi H, Experimental and
thermodynamic modelling of systems containing
water and ethylene glycol: Application to flow
assurance and gas processing Fluid Phase
Equilibria 2009 276 2430
[12]Sloan, E.D. and Koh, C. A., 2008. Clathrate
Hydrates of Natural Gas. 3rd ed. New York: CRC
Press.
APPENDIX