Carbon Dioxide Sequestration and Related Technologies

SECTION 1

DATA AND CORRELATION

Chapter 1

Prediction of Acid Gas Dew Points in the Presence of Water and Volatile Organic Compounds

Ray. A. Tomcej

Tomcej Engineering Inc. Edmonton, AB, Canada

Abstract

Aromatic hydrocarbons which are present in sour natural gas streams can be absorbed into the amine treating solution at the bottom of the contactor and exit in the rich amine stream. Depending on the process configuration, these dissolved hydrocarbons can end up in the acid gas leaving the amine regenerator. In acid gas injection facilities, trace amounts of heavy hydrocarbons in the acid gas may lead to the formation of a sour hydrocarbon liquid phase in the compressor interstage scrubbers.

In this exploratory work, a cubic equation-of-state (EOS) model was used to make predictions of non-aqueous (L1) dew points in acid gas systems. The objective was to develop a better understanding of the conditions under which this phenomenon can occur, and to reinforce the need for accurate experimental vapor-liquid equilibrium data to support cost effective design and model development.

1.1 Introduction

Benzene, toluene, ethyl benzene and xylene isomers are commonly referred to collectively as BTEX compounds. These compounds are known to be toxic to humans and their containment and disposal are of special interest to the hydrocarbon industry. BTEX environmental contamination is often linked to leakage from underground gasoline storage tanks or accidental spills. Awareness of this toxicity led to regulated clean air emission standards that directly impact the natural gas processing industry as trace amounts of BTEX compounds are associated with produced fluids such as natural gas.

Sour gas production generally involves a subsequent processing step in which the hydrogen sulphide (H2S) and carbon dioxide (CO2) are removed to produce an acid gas stream that may be a candidate for acid gas injection. Liquid solvents that are used to remove the H2S and CO2 from the gas stream are often aqueous solutions of organic chemicals that have a high affinity for the BTEX compounds.

Distribution of the BTEX compounds within the various streams of a natural gas processing plant is a complex phenomenon involving many interrelated process variables such as operating pressures and temperatures, amine composition, amine circulation rates, and others. Of particular interest in acid gas injection, is the amount of BTEX compounds that end up in the acid gas product leaving the amine regenerator.

The presence of trace quantities of BTEX compounds in the acid gas, if unaccounted for at the design stage, may lead to the unexpected formation of a sour non-aqueous liquid phase in the compressor train, and considerable operational difficulties. The objective of this work was to develop a better understanding of the conditions under which this phenomenon can occur, and to reinforce the need for accurate experimental vapor-liquid equilibrium data to support cost effective design and model development.

1.2 Previous Studies

In order to estimate the levels of BTEX compounds that will be present in the acid gas, there is a need for accurate vapor-liquid equilibria (VLE) and/or vapor-liquid-liquid equilibria (VLLE) data for BTEX and similar hydrocarbons in amine treating solutions under rich amine conditions. Operating plant data are also useful to verify the predictions of any thermodynamic model.

Ng et al. (1999) provided an overview of specific phase equilibria data and physical properties that are required for reliable design of acid gas injection facilities. Hegarty and Hawthorne (1999) presented valuable operating data for a Canadian gas plant using MDEA in which measured BTEX compositions were reported. McIntyre et al. (2001) and Bullin and Brown (2004) tabulated the experimental data available for hydrocarbon and BTEX solubility in amine treating solutions and demonstrated general trends in amine plant BTEX absorption using computer simulation. Valtz et al. (2002) presented a comprehensive set of fundamental solubility data for aromatic hydrocarbons in aqueous amine solutions. Miller and Hawthorne (2000) and Jou and Mather (2003) measured the solubility of BTEX compounds in water.

Clark et al. (2002) measured bubble and dew points for a nominal 10 mol% H2S/90 mol% CO2 mixture and regressed an equation of state to match the phase envelope. Satyro and van der Lee (2009) demonstrated that with suitable modification to interaction parameters, a cubic equation of state can provide reliable predictions of phase behavior in sour gas mixtures.

1.3 Thermodynamic Model

A rigorous treatment of the complex phase behavior in the H2S-CO2-water-BTEX system was beyond the scope of this work, which was intended to be exploratory in nature. The Peng-Robinson equation-of-state with classical van der Waals mixing rules was used in this study. The interaction parameter for the H2S-CO2 binary was set to 0.1 and all others were set to zero. Table 1 contains the critical properties used for the system components.

Table 1. Compo ne nt cri tica l proper ties.

The performance of the Peng-Robinson equation of state has been well documented in the literature. The model reproduced the dew point locus of Clark et al. (2002) to within 2.5%.

1.4 Calculation Results

The conditions of the calculations were chosen to encompass those normally found in acid gas injection compression: pressures from 150 kPa to 10 MPa, and temperatures above the hydrate formation curve from 0° to 100°C. Three different nominal acid gas compositions were considered: 20/80, 50/50, and 80 mole% H2S/20 mole% CO2. Hydrocarbon components studied included: benzene, toluene, ethyl benzene and dimethyl benzenes (xylenes).

The model was used to generate the phase envelope for each of the three nominal acid gas compositions. The influence of associated water on the location of the bubble and dew-point loci was not considered in this work. A typical injection profile was generated for each nominal composition using a starting pressure of 150 kPa and constant compression ratio. Temperatures in the compression process were restricted to remain under 150°C. Cooling temperature was set to 50°C. The final pressure was selected to be under 10 MPa but above the mixture critical point.

Initial calculations indicated that the phase behavior of the acid gas mixtures in the presence of each of the three xylene isomers was similar. For simplicity only o-xylene was considered in this study.

To establish a reasonable range of BTEX compositions, a sensitivity study was undertaken using pure H2S. The model was used to determine the L1 dew point temperature at 4000 kPa using various compositions of benzene and o-xylene ranging from 0 to 5000 ppmv. The results are shown in Figure 1.¹ Below concentrations of 100 ppmv, the aromatic compounds increase the dew point temperature by less than 1°C. Hegarty and Hawthorne (1999) reported BTEX content of up to 2500 ppmv in the acid gas of an operating MDEA plant. Using this as a guideline, non-aqueous liquid (L1) dew points were calculated for each of the three nominal acid gas compositions with 500-, 2000- and 5000 ppmv of each of the four aromatic compounds.

Figure 1. Effect of BTEX compounds on L1 dew point in pure H2S.

Clearly this range of calculated points generated a significant amount of data. The results for the 2000 ppmv cases are presented in Figures 2 through 4 and provide an adequate representation of the general trends that were observed. Note that curves labeled as organic compounds represent the dew point loci for the acid gas mixture with 2000 ppmv of only that organic compound.

Figure 2. Effect of BTEX compounds in 80% H2S - 20% CO2.

Figure 3. Effect of BTEX compounds in 50% H2S - 50% CO2.

Figure 4. Effect of BTEX compounds in 20% H2S - 80% CO2.

Using data from McIntyre et al. (2001) for BTEX component distribution in the acid gas from an MDEA plant as a guideline, flash calculations were performed at 50°C for the mixture given in Table 2. Identical calculations were performed for a mixture containing 80 mol% H2S and 20 mol% CO2. The results are shown in Table 3.

Table 2. Composition of mixture used for condensation study.

Table 3. Condensation study results at 50°C.

1.5 Discussion

In the absence of experimental data for dew point conditions in acid gases with contaminants, there can be no absolute conclusions drawn on the accuracy of the predictions. This exploratory study clearly emphasizes the importance of experimental research to provide fundamental information for process design and advanced model development. The results in Figures 2 through 4 illustrate that with conservative cooling temperatures and with BTEX contaminant levels in the range of those already measured in an operating MDEA plant, it is possible to enter the three-phase region in the higher pressure interstage coolers and separators in acid gas injection facilities. More aggressive cooling escalates the potential for three-phase conditions.

The formation of a second liquid phase in the compression interstage cooling system, in itself is not a problem, provided that the phase behavior phenomenon is understood at design time. The L1 phase is less dense than water, contains up to 20 mol% BTEX and, if formed, will accumulate in the interstage separators. As pointed out by Hegarty and Hawthorne (1999), it is extremely important to obtain an accurate inlet gas composition, including an extended analysis of the C6+ fraction to determine the aromatic content. Once the BTEX content, if any, is identified it can be accounted for in any process design, modeling, or operational troubleshooting of downstream processes such as acid gas injection.

In spite of the purely predictive nature of the calculated results, the following general observations can be made by analyzing Figures 2 through 4. The same behavior is observed in the 500 ppmv and 5000 ppmv calculated results.

At a given pressure, the presence of BTEX compounds in acid gas widens the phase envelope, with this effect being more pronounced in acid gases with higher CO2 content.

At a given pressure, the presence of BTEX compounds in acid gas increases the L1 dew point temperature, with this effect being more pronounced in acid gases with higher H2S content. This is, in part, a result of the shift of the acid gas phase envelope to higher temperatures in high H2S mixtures.

At equal concentration in the acid gas and at equal pressure, BTEX compounds increase the L1 dew point temperature in the order: benzene, toluene, ethyl benzene and o-xylene with o-xylene having the most pronounced effect.

In all cases, the possibility of non-aqueous L1 formation is highest in the separator before the final stage of compression.

If compressed acid gas is cooled to lower temperatures (e.g. 30°C) in the compressor facility, this increases the possibility of L1 formation.

If BTEX compounds are present in the acid gas at levels less than 100 ppmv, the acid gas dew point locus is relatively unaffected.

The dew point loci shown in Figures 2 through 4 indicate where the first droplet of L1 forms. Table 3 contains an example of the condensation behavior inside the phase envelope at constant temperature. Note that the condensation behavior of the BTEX mixture is similar to the BTEX-free system except for the deep depression of the dew point pressure. Lines of constant liquid volume % are widely spaced in this region of the phase envelope. This behavior is similar to the condensation behavior of rich gas systems. The location of the bubble point is relatively unaffected by the organic compounds.

References

Bullin, Jerry A. and William G. Brown, Hydrocarbons and BTEX Pickup and Control from Amine Systems, Proceedings of the 83rd Gas Processors Association Annual Convention, New Orleans, March 14–17, 2004.

Clark, M.A., W.Y. Svrcek, W.D. Monnery, A.K.M. Jamaluddin and E. Wichert, Acid Gas Water Content and Physical properties: Previously Unavailable Experimental Data for the Design of Cost Effective Acid gas Disposal Facilities, and Emission Free Alternative to Sulfur Recovery Plants, Hycal Energy Research Laboratories, 2002.

Hegarty, Mike and Dean Hawthorne, Application of BTEX/Amine VLE Data at Hanlan Robb Gas Plant, Proceedings of the 78th Gas Processors Association Annual Convention, Nashville, March 1–3, 1999.

Jou, Fang-Yuan and Alan E. Mather, Liquid-Liquid Equilibria for Binary Mixtures of Water+Benzene, Water+Toluene and Water+p-Xylene from 273K to 458K, J. Chem. Eng. Data, 48, 750–752 (2003)

McIntyre, G.D., V.N. Hernandez-Valencia and K.M. Lunsford, Recent GPA Data Improves BTEX Predictions for Amine Sweetening Facilities, Proceedings of the 80th Gas Processors Association Annual Convention, San Antonio, March 12–14, 2001.

Miller, David J. and Steven B. Hawthorne, Solubility of Liquid Organics of Environmental Interest in Subcritical (Hot/Liquid) Water from 298K to 473K, J. Chem. Eng. Data, 45, 78–81 (2000).

Ng, Heng-Joo, John J. Carroll and James Maddocks, Impact of Thermophysical Properties Research on Acid Gas Injection Process Design, Proceedings of the 78th Gas Processors Annual Convention, Nashville, March 1–3, 1999.

Satyro, Marco A. and James van der Lee, The Performance of State of the Art Industrial Thermodynamic Models for the Correlation and Prediction of Acid Gas Solubility in Water, Proceedings of the First International Acid Gas Injection Symposium, Calgary, Alberta, Canada, October 5–6, 2009.

Valtz, A., P. Guilbot and D. Richon, Amine BTEX Solubility, Gas Processors Association Research Report RR-180, 2002.

¹ Figures 1 through 4 appear at the end of this paper.

Chapter 2

Phase Behavior of China Reservoir Oil at Different CO2 Injected Concentrations

Fengguang Li Xin Yang Changyu Sun Guangjin Chen

State Key Laboratory of Heavy Oil Processing, China University of Petroleum Beijing, People’s Republic of China

Abstract

The phase behavior of China reservoir oil at different CO2 injected concentrations has been studied at the temperature of 339.2 K using a high-pressure PVT unit. Seven groups of reservoir fluids with CO2 molar contents of 0, 10.0, 34.1, 44.7, 48.9, 57.8, and 65.0 mol% have been prepared. Saturation pressure of reservoir fluids at seven CO2 injected contents were measured. The reservoir oil density and viscosity at different pressures under reservoir temperature were also obtained. The influence of CO2 molar contents on the interfacial tension of CO2 injected reservoir oil under stratum conditions was determined using a pendant drop method. The experimental data indicated that when CO2 content is lower than 45 mol%, the increase of bubble point pressure is slow. After that, the bubble point pressure value increases more sharply with the increase of CO2 molar concentrations. The reservoir viscosities decrease sharply with the increase of CO2 concentration when the system pressure is above the bubble point for different injection contents. The experimental results of interfacial tension for CO2 injected crude oil/stratum water show that it decreases with the increase of CO2 injected concentrations. The pressure has a slight effect on the interfacial tension value. These phase behavior data will be helpful for evaluating the effect of CO2 injected method to enhance oil recovery.

2.1 Introduction

The fluid phase behavior study is used as an important basis for miscible-slug process and predominant displacement mechanism, which is of critical importance during the miscible displacement process (1). The conventional fluid phase behavior test is usually conducted using PVT (Pressure-Volume-Temperature) unit. It is of great concern in many high-pressure technologies, such as fluid extraction process, exploration of near-critical gas condensate/volatile oil reservoir, and gas-injected enhanced oil recovery processes. CO2 displacement technology is recognized as a significant and well-established means for oil and gas enhanced recovery both at home and abroad. Miscible gas injection could minimize the trapping effect of capillary forces and is recognized as an economic enhanced oil recovery process.

Although some PVT fluid phase behavior data are available in the published papers, they are still insufficient because of the complexity of multi-component reservoir fluid. In this work, the phase behavior of China reservoir fluids collected from Jilin oil field were analyzed at different CO2 injected concentrations and pressures using a high-pressure PVT device. The density, bubble point pressure, viscosity, and interfacial tension properties of reservoir fluid at different CO2 injected mole percents and pressures under the stratum temperature were systematically measured.

2.2 Preparation of Reservoir Fluid

The reservoir fluid sample was collected from China Jilin oil field at reservoir conditions. The stratum temperature was 339.2K. The reservoir fluid arriving from the well was separated and flashed to standard condition. The molar composition of reservoir fluids was then obtained from analysis of the gas and oil samples. The gas phase was analyzed by HP6890 gas chromatograph. The liquid phase was analyzed by simulating distillation process using HP5890A. Afterwards, the reservoir fluid composition was obtained by combining the gas and liquid phase compositions using the gas–oil ratio (GOR). The measured composition for reservoir fluid was shown in Table 1. Molecular weights of the oil phase were determined by vapor pressure osmometer (VPO) and the determined molecular weight was 420 g/mol.

Table 1. The composition of reservoir fluid.

Seven groups of CO2 injected concentration (including 0% CO2) were chosen to study the reservoir fluid behavior under gas injection process. The CO2 injected crude oil was prepared using RUSKA PVT device.

2.3 PVT Phase Behavior for the CO2 Injected Crude Oil

Phase behavior of China reservoir oil was systematically investigated using a RUSKA high-pressure PVT system which was described in our previous papers (2,3). The PVT data at different CO2 injected molar components was measured to build the relationship between the volume and pressure of reservoir oil. The bubble point pressure and density of reservoir fluid at different pressures could then be determined according to the measured PVT data, which is useful to calculate the phase behavior properties such as the relatively volume, solubility of injected CO2 in oil, and so on.

The density of the CO2 injected reservoir fluid at different pressure under the strata temperature was plotted in Figure 1. From Figure 1, it can be found that there exists an inflexion for the curve of reservoir fluid density and pressure, showing the process of phase transition. When the CO2 concentration achieves 65.0 mol%, there is no significant difference between the gas phase and liquid phase, showing that there may exhibit first contact miscibility condition when the CO2 injected contents is 65.0 mol%.

Figure 1. Variation of reservoir oil density for CO2 injected crude oil at different CO2 mole percents and pressures.

The bubble point pressure at seven CO2 molar compositions determined from PVT measurement was shown in Figure 2. According to Figure 2, it shows that bubble point pressure increases with the increase of CO2 injected concentrations. When CO2 content is lower than 45 mol%, the increase of bubble point pressure is slow. However, when CO2 content is higher than 45 mol%, the bubble point pressure value increases more sharply with the increase of CO2 molar concentrations. The bubble point pressure data is also used to choose the suitable CO2 injected concentration.

Figure 2. Bubble point pressure at different CO2 injected concentrations for China reservoir crude oil.

2.4 Viscosity of the CO2 Injected Crude Oil

Viscosity is an important transport property in petroleum production and processing. RUSKA falling ball viscometer connected with RUSKA high-pressure PVT device was used in this work to investigate the viscosity of China Jilin oil samples after different CO2 content was injected under stratum conditions.

The basic principle of falling ball viscometer is based on Stokes law. The fluid viscosity could be exactly calculated by Stokes law according to the time of the ball travels through internal pipe from the top to the bottom. If the falling ball behaves to be laminar flow, the following equation was used:

(1) equation

where ρB and ρF are the density of the ball and fluid, respectively. t is the travel time. k is a constant value related to the diameter of the falling ball and the angel of the apparatus. Before the experiment, a falling ball was selected to measure the constant value k in Eqn. (1) using standard silicon oil for the viscometer. Thereafter, the reservoir crude oil viscosities were systematically measured with the same calibrated ball at different CO2 injected molar concentrations and pressures. The reservoir fluid viscosity was tested from higher pressure under single phase conditions until close to the saturation pressure. After the pressure was lower than the bubble point pressure, a gas exhaust valve was open to slowly reduce to the experimental pressure and the stable time was prolonged to 4–5 h. The measured viscosity for CO2 injected crude oil at different CO2 mole percents and pressures were plotted in Figure 3.

Figure 3. Variation of viscosity for CO2 injected crude oil at different CO2 mole percents and pressures.

As shown in Figure 3, the viscosity for CO2 injected crude oil decreased apparently with increasing of CO2 content. When the CO2 injected amount changed from 0 to 65.0 mol%, the reservoir oil viscosity value decreased greatly. At about 30 MPa, the viscosity value can decrease from 10.6 cP to 1.1 cP when 65 mol% CO2 was injected. It can be found that when the experimental pressure is higher than the saturated value, the reservoir oil viscosity increases with the increase of pressure; When it is lower than the saturated pressure, the reservoir oil viscosity increases with the decrease of pressure. With the decrease of pressure, more CO2 was released from the reservoir oil and induced the increase of viscosity of the residual oil.

From Figure 3, it can be concluded that CO2 injecting is significant in favor of the decrease of viscosity of Jilin reservoir crude oil. However, after CO2 content was higher than 44.7 mol%, the reservoir oil viscosity at single phase condition does not decrease significantly with the further increase of CO2 injecting concentration. Meanwhile, During the CO2 injecting concentration increases from 0 to 44.7 mol%, the bubble point pressure only increases from 11.28 MPa to 14.14 MPa. However, when the CO2 injected concentration increases from 44.7 mol% to 65.0 mol%, the bubble point pressure increases from 14.14 MPa to 25.0 MPa. Therefore, from the view of decrease of viscosity and bubble point pressure, there exists a suitable CO2 injecting concentration and high CO2 concentration is not needed.

2.5 Interfacial Tension for CO2 Injected Crude Oil/Strata Water

A great amount of reservoir water exists in the stratum after water displacement process of oil field. There is a special need for accurate interfacial tension estimation because the movement of reservoir fluids is influenced to a great extent by capillary forces. The CO2 injected concentration also plays an important role on the interfacial phenomena. In this work, the influence of CO2 molar contents on the interfacial tension of injected crude oil/water was systematically investigated using the JEFRI pendant drop high-pressure interfacial tension apparatus manufactured by D.B. Robinson (Canada), which the maximum working pressure is 34.5 MPa (5,000 psi) and the operating temperature range is 253–473 K. The experimental apparatus and procedures were detailed described in our previous papers (4,5). The interfacial tension measurement is based on the following principle:

If the drop is in equilibrium with its surroundings gas, the interfacial tension (γ) values can be calculated directly from an analysis of the stresses in the static, pendant drop, using the following equations developed by Andreas et al. (6):

(2) equation

(3) equation

where Δρ is the density difference between the two phases, De is the unmagnified equatorial diameter of the drop, g is the gravitational constant, ds is the diameter of the drop at a selected horizontal plane at height equal to the maximum diameter de. Andreas et al. have prepared a detailed table of 1/H as a function (ds/de).

The difference in density between reservoir oil and water could be calculated from the measured density data. The interfacial tension of CO2 injected crude oil/reservoir water were all measured under single-phase conditions at the stratum temperature. The measured interfacial tension data for CO2 injected reservoir oil/water at different CO2 injected molar concentrations and pressures are plotted in Figure 4.

As shown in Figure 4, the interfacial tension for CO2 injected oil/reservoir water decreased apparently with the increase of CO2 injected molar concentration when CO2 content varies from 0 to 65.0 mol%. The dissolvability of CO2 in oil has a significant influence on the interfacial tension value. The interfacial tension decreased by about one-third as the CO2 injected amount changed from 0 to 65.0 mol%. It also shows that the interfacial tension of the CO2 injected crude oil/water increased with increasing pressure. During the experiment process, the experimental pressure was always higher than the bubble point pressure at the corresponding CO2 injected condition. Compared with the effect of CO2 injected amounts, the pressure has only a slightly effect. When the CO2 composition was 65.0 mol%, the CO2 injected oil system approached complete miscibility and the interfacial tension data of CO2 injected crude oil/reservoir water changed a little with an increase in pressure.

Figure 4. Variation of interfacial tension for CO2 injected oil/reservoir water at different CO2 mole percents and pressures.

2.6 Conclusions

The phase behavior of reservoir oil collected from China Jilin oil field was systematically investigated by using a high-pressure RUSKA PVT device at different CO2 injected concentrations and pressures under strata temperature. Seven groups of CO2 injected concentrations varying from 0 to 65.0 mol% were prepared. The bubble point pressure increases from 11.28 MPa to 25.0 MPa when CO2 content increases from 0 to 65.0 mol%. When the CO2 concentration achieves 65.0 mol%, there is no significant difference between the gas phase and liquid phase, showing that there may exhibit first contact miscibility condition under the corresponding CO2 injected content. The viscosity for CO2 injected crude oil decreased apparently with increasing of CO2 content. CO2 injecting is significant in favor of the decrease of viscosity of Jilin reservoir crude oil. However, after CO2 content was higher than 44.7 mol%, the reservoir oil viscosity under single phase condition does not decrease significantly with the further increase of CO2 injecting concentration. The interfacial tension for CO2 injected oil/reservoir water decreased apparently with the increase of CO2 injected molar concentration when CO2 content varies from 0 to 65.0 mol%. When the CO2 composition was 65.0 mol%, the CO2 injected oil system approached complete miscibility and the interfacial tension data of CO2 injected crude oil/reservoir water changed a little with an increase in pressure.

Literature Cited

1. W. Yan, L.K. Wang, L.Y. Yang, T.M. Guo, Fluid Phase Equilibria, Vol. 190, p. 159–178, 2001.

2. M.X. Gu, Q. Li, X.Y. Zhou, W.D. Chen, T.M. Guo, Fluid Phase Equilibria, Vol. 82, p. 173–182, 1993.

3. H.Q. Pan, T. Yang, T.M. Guo, Fluid Phase Equilibria, Vol. 105, P. 259–271, 1995.

4. C.Y. Sun, G.J. Chen, L.Y. Yang, J. Chem. Eng. Data, Vol. 49, p. 1023–1025, 2004.

5. C.Y. Sun, G.J. Chen, J. Chem. Eng. Data, Vol. 50, p. 936–938, 2005.

6. J.H. Andreas, E.A. Hauser, W.B. Tucker, J. phys. Chem., Vol. 42, p. 1001–1019, 1938.

Chapter 3

Viscosity and Density Measurements for Sour Gas Fluids at High Temperatures and Pressures

B.R. Giri, P. Blais and R.A. Marriott‡

Alberta Sulphur Research Ltd. Department of Chemistry University of Calgary Calgary, AB, Canada

Abstract

Designing an acid gas injection scheme requires an accurate knowledge of the density and viscosity of the injected fluid as these properties are used to optimize compression, monitor transportation and model gas mobility in the reservoir. Fit-for-purpose models are developed based on the available literature data, which in some instances are either inaccurate or studied at industrially irrelevant temperatures and pressures. Moreover, the errors for predicted data at high pressures and temperatures can be as large as 20-50%.

An extensive literature search by Schmidt et al. [1] revealed that there are limited data for H2S and its mixtures available in the literature; most of which are limited to gaseous H2S and saturated liquids. The only existing data that extend to higher pressures (p = 10 to 50 MPa) and temperatures (T = 115 to 140°C) are from Monteil et al. [2] which were reported in the late 60’s, after which no measurements appeared to have been carried out. Expansion of the literature data to fill the void temperature and pressure regions, especially at relevant conditions for acid gas injection schemes (T = 0 to 150°C and p = 0.1 to 75 MPa) are desired so that the discrepancies of existing data sets can be resolved and reference viscosity models can be further tested and parameterised. It is worthwhile to note that during the recent development of the H2S viscosity model of Schmidt et al., [1] the data set from Monteil et al. [2] was excluded due to inconsistency. This further demonstrates the importance of additional experimental studies for the determination of H2S viscosity and density at elevated pressures and temperatures.

We have recently begun an experimental program aimed at measuring the high-pressure densities and viscosities of H2S and other acid gas mixtures using an Anton Parr vibrating tube densimeter and a Cambridge oscillating piston viscometer at p = 1 to 100 MPa and T = 0 to 150°C. This paper discusses how these instruments were commissioned, calibrated and operated. Interim CO2, CH4 and H2S results show the accuracy and reproducibility of the high-pressure measurements.

3.1 Introduction

Design of an acid gas injection (AGI), sour gas injection or CO2 injection scheme requires that the density and viscosity properties of the fluid be well known [1,3,4]. From pre-compression to the reservoir, the viscosity is required to assess frictional pressure drops and the density is required to calculate pressure gains due to static head. Expansion of the literature data to fill the applicable temperature and pressure