Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/269929666

H2SandCO2generationmechanismsonasteaminjectionprojectonPetrocedeñoField,OrinocoBelt

CONFERENCEPAPER·JANUARY2011

CITATIONS

3

READS

105

Availablefrom:ViolaineLamoureux-Var

Retrievedon:25September2015

1

WHOC11-112

H2S and CO2 generation mechanisms on a steam injection project

on Petrocedeño Field, Orinoco Belt .

D, UZCATEGUI PDVSA, Petrocedeño

V, LAMOUREUX-VAR IFP Energies nouvelles

E, BERGER

PDVSA Petrocedeño, seconded by TOTAL

This paper has been selected for presentation and/or publication in the proceedings for the 2011 World Heavy Oil Congress [WHOC11]. The authors of this material have been cleared by all interested companies/employers/clients to authorize dmg events (Canada) Ltd., the congress producer, to make this material available to the attendees of WHOC11 and other relevant industry personnel.

Abstract Petrocedeño field is located in the Junin area of Orinoco Belt, East of Venezuela. The Oil in Place (STOIIP) volumes on this field exceeds 30 Gbbl. However, the nature of Petrocedeño crude (Extra Heavy Oil, 8°API) results in a limited recovery factor. An Enhanced Oil Recovery (EOR) Pilot Project is currently being implemented in order to quantify the additional recovery that can be expected with continuous steam injection techniques. Results will be used to design future applications in the Orinoco Belt. This article will briefly present what the Petrocedeño EOR project consists in. One of the main expected issues associated to steam injection is H2S and CO2 in situ production. Due to the detrimental effects of these gases and the constraints they impose on facilities design, it is of paramount importance for the Petrocedeño EOR Project to understand their origin, in order to be able to predict their concentrations in production fluids. This paper presents the results of the laboratory experiments that were carried out to fulfill such purpose. A geochemical characterization study was first carried out on several oil sand cores. Oil composition variability and sulfur content distribution in the oil sand were assessed, along with total and labile sulfur content of the crude oil. The matrix mineralogical composition was also determined in order to

investigate possible chemical interactions between oil and minerals. Finally, the results of several aquathermolysis experiments performed at 250°C on different oil sands are also presented, along with the H2S and CO2 produced quantities for increasing exposure time to steam. Keywords: H2S and CO2, thermal oil recovery, aquathermolysis, Orinoco Belt.

Introduction PDVSA Petrocedeño is a joint venture between the 3 oil companies PDVSA from Venezuela (with a 60% shareholding), Total from France (30.3%) and Statoil from Norway (9.7%). The main purpose of this association is to exploit and develop an 8°API extra-heavy oil reservoir located in JUNIN Block of the Orinoco’s Oil Belt. The field operations began in 1999, using a production scheme based on natural depletion also called “cold production”. Taking into consideration the type of reservoir, oil characteristics and the initial production scheme, it has been evaluated that the maximum ultimate recovery factor which can

2

be expected from the natural depletion phase would be around 8-10%. In order to improve the recovery factor, Petrocedeño and its shareholders launched in 2006 a project to enhance oil recovery. After an extensive review of various technologies, EOR mechanisms based on thermal processes were recommended and approved by the shareholders. Petrocedeño EOR pilot project encompasses a 30 Km2 area (Figure 1) with the objective to evaluate recovery performances of 3 steam assisted recovery techniques: Steam Drive (SD), Steam Assisted Gravity Drainage (SAGD) and Horizontal Alternating Steam Drive (HASD). First steam injection is expected to occur in 2013. Steam and hot water assisted recovery from heavy oil reservoirs can induce physicochemical interactions between water, oil and rock. As a result, significant amounts of hydrogen sulfide, together with carbon dioxide, may be generated, increasing the risk of corrosion and health security problems during production (Burger et al., 1985; Fassihi et al., 1990; Thimm, 2001). For this reason, it is very important for the Petrocedeño EOR project to understand their origin, in order to be able to predict their concentration in produced fluids. Depending on the nature of sulfur present in reservoirs, three main types of reactions can potentially explain the formation of H2S under hydrothermal conditions. Thermochemical reduction of sulfates (TSR) is one possible process. TSR corresponds to the thermally-activated reduction of sulfates into free sulfur that rapidly reacts with hydrocarbons to form, among other products, large amounts of hydrogen sulfide (Hoffmann and Steinfatt, 1993; Worden and Smalley, 1996). Pyrite that commonly occurs in reservoir rocks is also a potential source of H2S. In contact with hot water or steam, this mineral can indeed be oxidized, generating sulfuric acid and sulfate ions that can promote TSR (Hutcheon, 1996). Finally, aquathermolysis of sulfur-rich heavy oils is a third process for the generation of H2S in reservoirs. Aquathermolysis is defined as chemical reactions between heavy oil and steam (Hyne et al., 1984). This paper presents the results of the laboratory experiments that were carried out to fulfill such purpose. In the first time, a geochemical and mineralogical characterization study was carried out on several oil sand cores from the EOR pilot area. Then, aquathermolysis experiments were carried out in IFPEn laboratories to reproduce in situ contact between reservoir rocksand steam.

Petrocedeño thermal EOR pilot The PETROCEDEÑO field, located in the core of the Venezuelan Orinoco Belt, hosts an Enhanced Oil Recovery (EOR) pilot project whose main objective is to demonstrate the possibility of increasing the recovery factor and extension of production plateau by testing different types of EOR techniques. After an extensive review of various technologies, the 3 steam injection based techniques - SAGD (Steam Assisted Gravity Drainage), HASD (Horizontal Alternating Steam Drive) and SD (Steam Drive) - were selected. A brief description of these techniques is given below. SAGD (Steam Assisted Gravity Drainage)

SAGD is a continuous steam injection technology that consists of two parallel horizontal wells, one injector and one producer, drilled with a vertical offset of 5 to 10 meters. The steam injected into the reservoir moves up from the injector

well (the upper well) and forms what is known as a “steam chamber” (a zone above the injector well with a high saturation of steam). The oil will be heated by the steam, will flow easier, and will be drained by gravity along the steam/oil interface back to the producer well. For such a chamber to develop, relatively thick (50 ft+) and permeable sands are needed. (Figure 2).

HASD (Horizontal Alternating Steam Drive) The HASD (Horizontal Alternating Steam Drive) process combines the advantages of steam injection processes like Steam Drive, SAGD and cyclic steam stimulation. The HASD configuration selected consists of 3 horizontal parallel wells drilled at the lower part of the reservoir sands. The central well is a permanent producer. The steam injection starts in one flank well and the production in the others. After a certain period the injection well is switched to a producer and the other flank producer is converted to an injector (Figure 3). The cycle time is a parameter that requires optimization for each reservoir (preliminary simulations indicate periods of 3 to 6 months for Petrocedeño). The HASD process involves three important aspects, (1) a high amount of steam being injected, (2) the alternating characteristic of cyclic steam soak and (3) the well known gravity drainage typical of SAGD. HASD is expected to be efficient in thinner sandy layers (30ft+) than SAGD, although it may target thick sands as well.

SD (Steam Drive)

Steam Drive, known as well as steam flooding, is a thermal recovery process that comprises continuous steam injection into the reservoir through vertical injection wells. The mechanism involves 3 main concepts; (1) the steam will push the oil to the producer well; (2) the steam injection will generate higher pressure and temperature, the latter lowering the viscosity of the heavy oil; and (3) the gravity effect will cause the steam to go up and the oil to drain down to the horizontal producers (Figure 4). Steam Drive may be envisaged in layers of less than 30ft, although applicable in sandy layers of 30ft+, but with a lower recovery factor than HASD.

A southwestern area (around 30 km2) of the Petrocedeño field was selected to implement the pilot, taking into account the actual geological knowledge regarding sand thickness and main sedimentological axes. This area was also known as having a limited cold development plan, hence limited interferences between the Pilot Project and the existing wells. This zone was further delineated with 42 wells, drilled in 2 steps, in order to get a much more precise description of sand channels organization in space (vertically and laterally), and to optimize the final location of the EOR pilot. The first delineation campaign consisted of 30 vertical wells, drilled following a grid of more or less 1km, and with the most complete acquisition program (conventional and unconventional logging, coring (around 2000ft), production tests, vsp, minifrac). The second step consisted of 12 deviated wells drilled following a grid of 500m, and where full combo LWD (GR, resistivities, neutron/density) was acquired. Additionally, an intensive program of special laboratory studies was launched (Thermal PVT, geomechanics, aquathermolysis, and relative permeability). All of these studies were performed under high temperature conditions. The aquathermolysis study was required in order to assess the risk of H2S and CO2 generation upon steam injection and evaluate its effect on oil and gas composition. Outside of the safety and environmental

3

aspects, which are the most important, the results will also provide key information regarding the rate of corrosion to be expected for surface equipment. Five preserved core samples (full diameter) from the EOR area were sent to IFP Energies nouvelles (France) in order to perform this analysis. A possibly significant amount of H2S in the production stream had to be taken into account for the EOR pilot surface installations and wells design. As for EOR wells design, corrosion studies evidenced a significant risk of corrosion, mostly due to H2S and CO2 presence along with condensed water. After an exhaustive evaluation of different steel grades for tubings, liners, casings and wellheads, and balancing risks and costs, the use of carbon steel L80 was recommended for tubings and casings, due to the performance of this steel grade when it is exposed to a constrained thermal expansion and its reliability in sour environment. Blanket gas will be injected in the casing-tubing annulus as another way of corrosion protection for the production casing (single barrier). Same carbon steel concept recommendation was given for wellheads and X-Mas trees using EE class material for the equipment and parts in direct contact with the production flow and using a surface double mechanical barrier. A complete corrosion study was as well performed for surface facilities. Corrosion allowance and / or suitable materials were selected for process equipments and piping where corrosion problems were detected. Moreover, connection points for corrosion inhibitor injection and corrosion coupons were added in the design. A comprehensive monitoring of fluids H2S content will be as well performed with sampling points in various parts of the surface facilities, and with a H2S liquid analyzer located at the export point towards the main station. At last, a stripper will bring the emulsion H2S concentration down to the 12 ppm required for the export to the main station. The gas produced by the EOR pilot will be used as fuel gas for the boilers, along with natural gas coming from the main station. The expected high concentration of H2S in burnt gas will result in a significant amount of SO2 in exhaust gas, which will be mitigated through adequate chimney height.

Oil sands geochemistry Oil sand cores 6 cores issued from Petrocedeño Field, were provided by Petrocedeño Company for this study (Table 1). They were all cored during the delineation campaign in the target reservoirs of the EOR project.

Table 1: Origin and sedimentary environment of the cores (data from Petrocedeño)

Core well Reservoir type

1 1 fluvial

2 1 fluvial

3 1 fluvial

4 2 fluvial

5 3 Deltaic

6 4 seal

Prior to aquathermolysis experiment; it was very important to measure some properties of the cores, to get an idea of the variability of some relevant properties at the core scale (40cm to 90cm length). Aliquots were taken in different places on each core. For each aliquot, the heavy oil content (C14+), the NSO content (NSO = heavy organic molecules of oil containing Nitrogen, Sulfur and Oxygen) and the sulfur content were quantified. These properties were chosen because they are the most relevant with regard to the issue of H2S production.

Experimental protocols In order to quantify the heavy oil content (C14+) in the oil sands, the oil was extracted from rock with dichloromethane CH2Cl2. Then the obtained C14+ oil was dissolved in a mix of nC5 and CH2Cl2 and separated by liquid chromatography on Florisil into 2 fractions: the NSO and the maltenes C14+, which proportion was quantified. The maltenes C14+ are constituted by the Saturates C14+, the Aromatics C14+ and the light Resins. The mass of NSO was deduced. The sulfur content in oil was quantified by analyzing oil sands with the Rock-Eval designed for sulfur analysis. The organic compounds were analyzed as well, with an FID and an IR spectrometer.

Results and Discussion The characterization of cores showed that the cores 2, 3 and 4 have homogeneous properties. On the contrary, there is geochemical heterogeneity in the reservoir at a scale lower than 30cm as seen in the cores 1 and 5. The core 1 showed a gradient of heavy oil content (from 15wt% to 19wt%, over 30cm), accompanied by a gradient of porosity and a gradient of sulfur content (from 4.0 to 4.8wt% in heavy oil, over 30 cm): at its top the core 1 is more porous, contains more oil and the oil is richer in sulfur. Consequently, it was decided to evaluate the effect of this gradient on gas production. Hence the upper part and the lower part of this core were selected to undergo aquathermolysis experiments separately. The core 6 appeared to be a green clay containing oil<1wt%. It was left apart from the study. For this reason 5 cores were selected for the study: 2, 3, 4 and 1 (top and base). Oil sand samples dedicated to aquathermolysis experiments It was relevant to study geochemical properties of the samples dedicated to the aquathermolysis experiments. Two points were considered: the oil composition and the mineralogical composition of the matrix.

Experimental protocols To ensure reproducible aquathermolysis experiments, one of the required conditions was that the oil sand samples were homogeneous. So the oil sand samples were mixed with a spoon or a spatula to homogenize them. For each oil sand sample, the C14+ oil was extracted with nC5 and then with CH2Cl2, at reflux. The nC5 extracts were fractionated into Saturates, Aromatics and Resins by liquid chromatography MPLC. The asphaltenes were separated by n-heptane precipitation. Hence the SARA (Saturates, Aromatics, Resins, Asphaltenes) composition of C14+ oil was deduced. For each SARA fraction, elemental oxygen and sulfur contents

4

were measured by direct coulometry, whereas elemental hydrogen and carbon contents were measured by direct thermal conductivity. The total and pyrolyzable sulfur contents in oil were quantified by the Rock Eval designed for sulfur analysis. This analysis enables discriminating labile sulfur from refractory sulfur: (Pyrolyzable sulfur = sulfur released by oil sand during Rock-Eval pyrolysis; Refractory sulfur = sulfur released by oil sand after Rock-Eval pyrolysis, during Rock-Eval oxidation).

The Matrix mineralogical composition was also evaluated by X-Ray Difractometry (XRD), Fourier Transform Infrared Spectrometry (FTIR), and direct elemental analysis by X-Ray Fluorescence Spectrometry (XRF).

Results and Discussion The geochemical study showed that the 5 oil sand samples have close geochemical properties. They contain 14 to 16 wt% of C14+ oil C14+, they have almost identical SARA composition (Table 3), and the elementary composition of the SARA fractions is very similar (Table 2).

Table 2: Elementary composition of the C14+ oils, from direct

elemental analysis

Sample %C %H %S %O %Total

1 Upper 83.6 10.6 4.3 1.3 99.7

1 lower 83.2 10.4 4.4 1.5 99.5

2 83.3 10.3 4.5 1.4 99.4

3 83.7 10.4 4.6 1.5 100.3

4 83.4 10.3 4.5 1.3 99.5

The mass balance of the elementary composition reaches more than 99%, indicating that the heavy oils were almost exclusively composed by C, H, S and O, N and heavy metals (Ni, V) representing less than 1wt%. Additionally, it can be noticed that the sulfur content is very high.

Table 3: SARA composition of the 5 heavy oil samples.

Sample Saturates

C14+ Aromatics

C14+ Resins Asphaltenes

nC7

1Upper 14% 25% 52% 8%

1 lower 14% 25% 52% 10%

2 13% 25% 52% 9%

3 13% 24% 55% 8%

4 13% 24% 52% 11%

It can be seen from Table 3 that the resins are the main fraction of C14+ oil, which is a typical property of the heavy oils. Moreover, the elementary composition of the SARA fractions showed that the resins were the main carrier of sulfur in the oil: they carry 65 to 69wt% of sulfur in the oil, whereas the aromatics carry 20 to 24wt% and the asphaltenes carry only 10 to 13wt% of sulfur in the oil.

Finally, the sulfur contentin oil could be split into pyrolyzable sulfur and residual refractory sulfur, as presented in the Table 4.

Table 4: Rock-Eval "pyrolyzable" sulfur and refrac tory sulfur

Sample Sulfur

(Wt % in oil)

Pyrolyzable

Sulfur

(Wt % in oil)

Refractory Sulfur

(Wt % in oil)

1 Upper 4.27 3.65 0.62

1 lower 4.46 3.95 0.51

2 4.41 3.81 0.60

3 4.52 4.00 0.52

4 4.42 3.90 0.52

The residual sulfur content in the heavy oil varies between 0.51wt% and 0.62wt%, whereas the pyrolyzable sulfur content is much higher and varies on a wider range: between 3.65wt% and 4.00wt%. It is interesting to note that the pyrolyzable sulfur content increases with the total sulfur content (Figure 5). This result suggests that only 11% of the total sulfur in the oil is refractory to thermal treatment. This induces that the other 89% of the total sulfur in the oil might be a source of H2S upon steam treatment. The characterization of the matrix by X-Ray Diffraction investigation relieved that the 5 sand matrices contain quartz at approximately 90wt% and potassium feldspar at less than 10wt%. They also contain few percents of kaolinite and carbonates: calcite and possibly siderite and dolomite that can act as a CO2 source. Infra-Red spectrometry investigation allowed quantifying the relative carbonates content, showing different values according to the sample (Table 5).

Table 5: Carbonates relative content in the matrices, measured by

Infra Red Spectrometry

Sample Normaliced carbonates

content 1 Upper 1

1 lower 0.24

2 0.26

3 0.44

4 0.28

The metallic elements in presence are aluminum (0.85 to 1.52 wt%), potassium (0.02 to 1.14 wt%), calcium (0.08 to 0.17wt%) and iron (0.06 to 0.18wt%). They might catalyze the aquathermolysis process. The matrices contain a very small amount of sulfured coke, with a sulfur content of 1wt% of the matrix. As a contrary, mineral sulfur was bellow the quantification limit (<0.1%). So it was deduced that the matrices didn't contain any sulfate, pyrite or sulfide. This was an important point, which indicates that during steam treatment H2S originates from sulfur in bitumen only and there cannot be any thermochemical sulfate reduction (TSR), which is a major process for H2S production. It also indicates that no pyrite played any role in H2S production.

5

H2S and CO2 production from aquathermolysis laboratory experiments The aquathermolysis study was conducted aiming at assessing the capacity of the oil sands to generate gas, and especially H2S and CO2, upon steam treatment. For that, aquathermolysis experiments were designed to simulate in laboratory the chemical reactions occurring in the reservoir during steam injection. Experimental Protocols The aquathermolysis experiments consisted in heating few grams of oil sand with added distilled and free-oxygen water (1/1,(v/v)) in a closed gold tube, under inert gas and under external isotropic pressure. The temperature was isothermal at 250°C. The external pressure was maintained at 80 bar, which allowed water to be in the liquid phase. The experiments were carried out respectively for 14, 28 and 56 days. At the end of each experiment, the total gas was recovered and the produced CO2, H2S, H2, CH4, C2H6, C3H8, iC4H10 and nC4H10 were quantified in absolute amounts. The gas produced within each gold tube was recovered in a vacuum line equipped with a Tœpler pump. The amount of gas was quantified by pressure measurement in a calibrated volume. Then the molecular composition was determined by gas chromatography analysis (GC), using a calibrated thermal conductivity detector and a flame ionization detector. Knowing the total amount of the bulk gas, the percentage of each individual gas species was converted into absolute yield. Nitrogen and Oxygen where removed at the end of the calculation as the former was in excess in the experiments and the latter came from atmospheric pollution during the analysis, though it was in negligible quantity.

Results and Discussion Whatever the oil sand sample, the experimental results showed that the produced gas was composed mainly by CO2, amounting 70 to 90mol%. The produced H2S was in significant concentration, from 3 to 11mol%. H2 production reached 2 to 7mol%, and CH4 production reached 4 to 7 mol%. Hydrocarbon gases, C2 to C4, reached around 1mol% each, indicating a slight hydrocracking of the oil.

CO2 PRODUCTION

The experimental results showed that the gas production continuously increased as aquathermolysis occurred. Although having similar compositions, the 5 oil sand samples produced significantly different quantities of gas upon aquathermolysis: gas production varies from 1100 to 1800 [µg gas]/[g oil sand] after 56 days of aquathermolysis, according to the sample. This variability comes mainly from the variability of CO2 production, varying from 900 to 1600 [µg CO2]/[g oil sand] according to the sample (Figure 6).

CO2 production could be correlated to the mineral carbon content (Figure 7), to the relative carbonates content (Figure 8) and to the calcium content (Figure 9) in the sand matrix. This result is surprising, because the carbonates contents and calcium contents are very small -around 0.1wt% of calcium in the matrices- and, until now, carbonates were thought as a whole to have a negligible role in CO2 production during oil sand

steaming. However, such correlations suggest that carbonates and especially calcite in the sand matrix, even in very small quantities, induce significant CO2 generation under steam treatment at 250°C. Besides this mineral origin, an organic origin of CO2 exists. As a matter of fact, CO2 generation was observed from aquathermolysis experiments on the oil extract only. Moreover, similar aquathermolysis experiments performed at 300°C on thiolane and thiophene showed CO2 generation as well (Clark et al., 1983).

H2S PRODUCTION The experimental results showed that H2S production varies, according to the sample, from 0.08 to 0.12 [g H2S]/[kg oil sand] after 56 days of aquathermolysis (Figure 10).

It was seen from the geochemical characterization that, whatever the oil sand sample studied, they do not contain any mineral sulfur. As a consequence, it can be asserted that H2S generation is derived from sulfur occurring in oil. H2S production could be correlated to the sulfur content of C14+ oils (Figure 11), and even more clearly to the "Rock-Eval pyrolyzable sulfur" content (Figure 12). It is important to note that a difference of sulfur content in heavy oil by 0.1wt% only induces an augmentation of H2S production by 0.04 [g H2S]/[kg oil sand] after 56 days of aquathermolysis, which represents approximately 30% to 200% of the total H2S production. This shows that if we intend to foresee properly H2S production and its concentration, then a precise quantification of sulfur in oil is required (S absolute uncertainty <0.1wt%).

The elements Al, Fe and K, which were found in the matrices, are known to promote catalysis oxidation, decarboxylation and hydrogenation of organic compounds. In this respect, it is possible that they played a catalytic role on gas production such as H2S and CO2 during aquathermolysis.

SPATIAL VARIABILTIY OF GAS PRODUCTION

Another important point to note is that the maximum and the minimum of gas production generated by the 5 different oil sand samples are associated with two samples collected from the same core and separated by only 20cm. The top of the core produced approximately 50% more gas than the bottom. The other samples from other wells produced intermediate quantities of gas. This leads to underline that laboratory results should be interpreted keeping in mind that the heterogeneity in the reservoir can exist at a very small scale (20cm here) and may have significant effects. Hence this confirms that a small scale screening study is appropriate to estimate the spatial variability of the steam treatment impact. For example, a simple screening study could be directed only on sulfur and carbonates quantification, since these characteristics seem to be key points for gas production in these EOR and geological contexts. Accordingly the Rock-Eval appears as a good tool for this purpose, with a rapid time of analysis and with a level of precision for sulfur and mineral carbon quantification which is probably sufficient to screen the spatial relative risk of CO2 and H2S production.

Discussion of results and way forward

These aquathermolysis experiments evidenced the heavy oil as the main source for H2S generation, and possibly the matrix as the main source for CO2 generation although some doubts remain about the respective contribution of matrix carbonates

6

and heavy oil (all the experiments having been performed on oily sands, and not yet on oil and sands separately). The remaining issues are the effect of temperature on the H2S and CO2 generation, the search for H2S and CO2 plateau, and to confirm the origin of H2S and CO2 by performing experiments on heavy oil and cleaned matrix alone separately. Moreover, the scope of future experiments had to be adapted to account with the experimental dispersion observed during this first phase. At last, the need to integrate later on the experimental results in thermal dynamic simulation models had to be accounted for, in particular the need for preliminary kinetics for H2S and CO2 generation equations in these models.

Hence, it was decided to perform new aquathermolysis experiments on at least three different samples of oily sands, at different temperatures (250, 275 and if possible 300°C), and with three time steps (1, 2 and 3 months, this last one being the maximum duration possible with current equipments). Two experiments of one month duration, at 250°C were added, in order to investigate the behavior of matrix and heavy oil separately.

Lately, it was decided to drop the 300°C experiment, both for budgetary constraints and because it was feared that new reactions might take place when reaching such temperatures.

Conclusion .

Steam injection for EOR can lead to H2S and CO2 generation, in oil sand reservoirs at temperatures from 200°C, resulting from physicochemical transformations of sulfur compounds. Aiming at deciphering H2S and CO2 generation mechanisms, aquathermolysis laboratory experiments were carried out on various oil sand samples issued from Petrocedeño field of the Orinoco Belt in Venezuela.

The results showed a significant production of H2S and CO2, along with the production of H2, CH4, C2, C3, C4.

CO2 might be generated by organic and/or mineral source. The experimental results suggest that the matrix is a probable source. However, CO2 generation from the crude oil is possible as well. In order to verify this issue complementary work is in progress.

Like CO2, H2S might be generated by organic and/or mineral source. As the oil sand samples studied don't contain any mineral sulfur, it was deduced that H2S is generated from the sulfur in the crude oil. Consequently, the possibility of any thermochemical sulfate reduction (TSR), which is a major process for H2S production and induces high contents in H2S, was excluded. It also indicates that pyrite cannot be involved in the reaction and in the H2S fate. However, the catalytic effect of the minerals found in the matrix might also contribute to the generation of H2S.

Finally, H2S generation could be correlated to "pyrolyzable sulfur content" in oil, suggesting that only a labile part of sulfur in oil participates to H2S generation. Assuming that our experiments are representative of the aquathermolysis occurring in oil sands reservoirs during steam injection, the determination of the labile sulfur distribution within a reservoir would provide useful information for the evaluation of sour gas production risk during steam-assisted recovery projects.

Acknowledgement The authors want to express their gratitude to PDVSA

Petrocedeño and its shareholders TOTAL, and STATOIL, as well as to IFP Energies nouvelles for allowing the publication of this work. The authors also want to express their gratitude to the laboratory staff that performed the experimental work.

REFERENCES 1. BURGER J., SOURIEAU P., COMBARNOUS M.,

“Thermal methods of oil recovery“, Ed. Technip, Paris, France pp 430, 1985.

2. FASSIHI M.R., MEYERS K.O.,WELSBROD K.R.,

”Thermal Alteration of Viscous Crude Oils“, SPE 14225, 1990.

3. THIMM H.F., “Hydrogen sulfide measurements in

SAGD operations”, Journal of Canadian Petroleum Technology, Vol.40 N°1, pp.51-53, 2001.

4. HOFFMANN G.G., STEINFATT I.,

”Thermochemical sulfate reduction at steam flooding processes - a chemical approach.“ ,ACSPetroleum ChemistryDivision, Vol. 38,N°1, pp181-184,1993.

5. WORDEN R. H. AND SMALLEY P.C.,”H2S

producing reactions in deep carbonate gas reservoirs: Khüff Formation AbuDhabi“, Chem.Geol., Vol. 133,pp. 157-171, 1996.

6. HUTCHEON I.,” The potential role of pyrite

oxidation in corrosion and reservoir souring.“, 47th

Annual CIM PETROLSOCTECH MTG (Calgary,Canada,6/10-12/96) Proc. Vol 2, paper N°CIM 96-95, 1996.

7. HYNE J.B., GREIDANUS J.W., TYRER J.D.,

VERONA D., RIZEK C., CLARK P.D., CLARKE R.A., KOO J.,"Aquathermolysis Of Heavy Oils.", 2nd Int. Conf. The Future Of Heavy Crude And Tar Sands, McGraw Hill, New York, Chapter 45, pp. 404-411, 1984.

8. LAMOUREUX-VAR, V. and LORANT, F;

“Experimental Evaluation of H2S Yields in Reservoir Rocks submitted to steam Injection”. 13th European Symposium on Improved Oil Recovery,

EAGE, Budapest, Hungary, D08, 25-27 April 2005.

9. LORANT F., ANTONAS R., ESPITALIÉ J, ”Characterization of sulfur in reservoir rocks by

Rock-Eval analysis“, 13th

European Symposium on Improved Oil Recovery, EAGE, Budapest, Hungary, 25 -27, April 2005.

10. CLARK P.D., HYNE J. B., TYRER J. D.,

"Chemistry of organosulfur compounds types occurring in heavy oil sands - 1. High temperature hydrolysis and thermolysis of tetrahydrothiophene in relation to steam stimulation process", Fuel, Vol. 62, pp. 959-962, 1983.

7

Figures

Figure 1: EOR area location in the Petrocedeño block

Figure 2: SAGD (Steam Assisted Gravity Drainage): a concept requiring about 50 ft thick sands. Heated oil drains towards the lower producer by gravity along the steam/oil boundary (green area).

Figure 3: HASD 3 (Horizontal Alternating Steam Drive), with a permanent central producer, and two alternating producers/injectors.

Step n+1:

inject in right produce in left

Step n:

inject in left produce in right

Step n+2:

inject in left produce in right

Step n:

inject in left produce in right

8

Figure 4: SD (Steam Drive): Steam will be injected through vertical wells. In Petrocedeño EOR Pilot Project horizontal wells will be used as producers.

Figure 5: Pyrolyzable sulfur content and residual refractory sulfur content vs total sulfur content, for the five oil sand samples

STEAM ZONE

OIL +WATER

STEAM+HOT WATER

STEAM

PRODUCERINJECT OR

HEAT LOSS ES

GRAVITY

CONDENSATED STEAM HOT OILCOL

D OIL

PRODUCER ZONE

STEAM ZONE

OIL +WATER

STEAM+HOT WATER

STEAM

PRODUCERINJECT OR

HEAT LOSS ES

GRAVITYGRAVITY

CONDENSATED STEAM HOT OILCOL

D OIL

PRODUCER ZONE

STEAM ZONE

OIL +WATER

STEAM+HOT WATER

STEAM

PRODUCERINJECT OR

HEAT LOSS ES

GRAVITY

CONDENSATED STEAM HOT OILCOL

D OIL

PRODUCER ZONE

STEAM ZONE

OIL +WATER

STEAM+HOT WATER

STEAM

PRODUCERINJECT OR

HEAT LOSS ES

GRAVITYGRAVITY

CONDENSATED STEAM HOT OILCOL

D OIL

PRODUCER ZONE

9

Figure 6: CO2 production upon aquathermolysis experiments at 250C° and 80 bar of external pressure, vs duration.

Figure 7: CO2 production upon aquathermolysis at 250°C vs mineral carbon content.

10

Figure 8: CO2 production upon aquathermolysis at 250°C vs relative carbonates content

Figure 9: CO2 production upon aquathermolysis at 250°C vs Calcium content

11

Figure 10: H2S production upon aquathermolysis experiments at 250C° and 80 bar of external pressure, vs duration.

Figure 11: H2S production upon aquathermolysis experiments at 250°C, vs total sulfur content in oil measured by Rock-Eval in original oil sands.

12

Figure 12: H2S production upon aquathermolysis experiments at 250°C, versus "pyrolyzable sulfur" content in oil measured

by Rock-Eval in original oil sands.