In Canadian reservoirs at 5ºC and 100-200 meters depth, similar oil has a .... established within the National Iranian Oil Company (NIOC) to manage national HO ...
PAPER 2008-403
Case Histories for Heavy Oil Recovery in Naturally Fractured Carbonate Reservoirs in the Middle East A.S. BAGÇI Institute of Petroleum Engineering, Heriot-Watt University, Edinburgh, United Kingdom
A. SHAFIEI, M.B. DUSSEAULT Earth & Environmental Sciences, University of Waterloo, Waterloo, Ontario, Canada This paper has been selected for presentation and publication in the Proceedings for the World Heavy Oil Congress 2008. All papers selected will become the property of WHOC. The right to publish is retained by the WHOC’s Publications Committee. The authors agree to assign the right to publish the above-titled paper to WHOC, who have conveyed non-exclusive right to the Petroleum Society to publish, if it is selected.
fields in the Middle East. A review of production technologies application in HO NFCR’s is presented, although there are few field data available with a high level of detail. A review of numerical simulation and experimental studies in NFCR heavy oil reservoirs is presented, as well. Heavy oil will play a vital role in providing energy in the 20-100 yr time frame, and it seems the industry is on the verge of a steep learning experience in heavy oil production from NFCR’s.
Abstract Despite huge global reserves of heavy oil, worldwide production is only about 6% of the total oil production of about 84.5×109 b/d. Heavy oil resources in naturally fractured carbonate reservoir (NFCR’s) are estimated to be on the order of 1.50-2.2×1012 b, of which one third is in the Middle East. The technologies required to economically recover heavy oil in situ, particularly highly viscous bitumen and extra-heavy oil, have major differences compared to conventional oil. Flow instabilities (advective instabilities) such as viscous fingering and water or gas coning that plague production methods in conventional oil are far worse in heavy oil reservoirs because of the high viscosity (unfavorable mobility ratio). In fact, these instabilities are more severe in fractured carbonates, as compared to sandstones, because of the presence of pervasive interconnected fractures that have high fluid transmissivity characteristics. We highlight geographical distribution, occurrence, and geological setting and some petrophysical and chemical analysis of heavy oil from major NFCR heavy oil
Introduction Despite enormous world heavy oil endowment (HO - more viscous than 100 cP in situ) – estimated by the USGS to be on the order of 8.5-9.0 Tb – worldwide production is only about 6% of the total oil production of ~84.5×106 b/d. World HO resources in naturally fractured carbonate reservoirs (NFCR’s) are estimated to be 1.50-2.2×1012 b. The technologies required to economically recover HO in situ, particularly highly viscous bitumen and extra-heavy oil, have major differences compared to conventional oil. Flow instabilities (advective instabilities) such as viscous fingering and water or gas coning that plague
1
production methods in conventional oil are far worse in HO reservoirs because of the high viscosity (unfavorable mobility ratio). Such instabilities are more severe in NFCR’s, compared to sandstones, because of pervasive interconnected fractures with high fluid transmissivity characteristics. Heavy oil and extra heavy oil (HO and XHO) are defined by some as oil with density ρ = 0.934-1.0 g/cm3 and > 1.0 g/cm3 respectively,[1] although in production technology evaluation a definition based on in situ viscosity is preferred. The definition of HO is not consistent, varying between different authors. The word ‘heavy’ relates to the high density of the oil, and the API gravity scale is widely used. HO has been defined as oil with API gravity of 1.0 and μ < 104, such as the deeper Faja del Orinoco oil in Venezuela (typically 8.5-9ºAPI, μ 1000-4000 cP at 40-45ºC in situ). In Canadian reservoirs at 5ºC and 100-200 meters depth, similar oil has a viscosity over one million cP.[4] Oil viscosity and temperature control flow rate in thermal production; thus, viscosity under in situ conditions is far more important in economic assessment than API gravity. Many suggest that heavy oils be defined as having viscosities >100 and 10,000 cP. We will use these abbreviations: HO for heavy oil and XHO for extraheavy oil with a viscosity 100 cP), unless otherwise specified. Saidi[8] defined a naturally fractured reservoir (NFR) as one that contains fractures (planar discontinuities) created by natural processes (e.g. diagenesis, tectonics forces) and distributed as a consistent connected network throughout the reservoir. NFR’s are assumed to have interconnected fracture systems that provide the main flow paths (high permeability, low storativity), and a matrix that acts as the main hydrocarbon repository (low permeability, high storativity). The matrix system contains most of the oil, but flow of oil to the wells is through the high permeability fracture system, implying that it is the matrixfracture interaction that mainly controls oil recovery behaviour. As much as 50-60% of the world’s present proven conventional petroleum reserves are in naturally fractured carbonate reservoirs (NFCR’s).[9] Moreover, one-third of the HO and XHO endowment of the world is found in NFCR’s [5, 10] . Hence, NFCR evaluation has been a high priority for researchers and oil and gas sections for decades, but the challenges presented by these highly heterogeneous rocks seem to be never-ending. Despite the hydrocarbon wealth NFCR’s hold, carbonate rocks have a bad reputation for having highly complicated interrelationships between porosity, permeability and other reservoir properties, as well as a great deal of heterogeneity. Also, they are considered as stress-sensitive reservoirs [11], where flow properties change with changes in effective stresses as depletion or injection take place. In this article we highlight geographical distribution, occurrence and geological setting, and some petrophysical and chemical analysis of HO from major NFC HO fields in the Middle East. A review of production technologies application in HO NFCR’s is presented, although there are few reliable field data available. A review of numerical simulation and experimental studies in HO NFCR’s reservoirs is presented, as well. HO will play a vital role in providing energy in the 20-
100 yr time frame, and it seems the industry is on the verge of a steep learning experience in HO production from NFCR’s.
HO NFCR’s in the Middle East NFCR’s usually demonstrate an unusual production history; they initially appear highly productive, but then decline rapidly. They are well-known for early gas or water breakthrough as the low-viscosity phase migrates rapidly through the fractures. Nevertheless, carbonates include some of the largest, most productive reservoirs on Earth (e.g. Ghawar oil field in Saudi Arabia which produces 4.50×106 b/d). NFCR’s have extremes in flow behavior with open channels and large vugs with permeabilities of many Darcies, and matrix permeability in the milliDarcy range. Only slow seepage is possible in the latter, yet a network of open fractures and cavities can have flow rates like a small river. Heterogeneity and scale effects complicate reservoir evaluation, technology screening and hydrocarbon recovery itself. The complicated nature of this class of reservoirs is the motivation behind the oil industry’s current efforts to learn more about them and simulate their behavior. Oil production from HO NFCR’s is far more problematic than conventional oil NFCR’s because of the high viscosity (unfavorable mobility ratio). Advective instabilities are far more severe in HO in NFCR’s because of this. HO resources in NFCR’s in the Middle East are estimated to be 530-970 ×109 b, which is equal to 30-35% of the world HO in NFCR resource base [2, 5, 10]. Contribution to daily production is estimated to be 200-250×103 b/d, an almost negligible quantity compared to the daily light oil production in the region. Middle East HO production has lagged behind the giant light oil producers except in countries like Turkey, Oman and Egypt which have limited light oil reserves. Following oil price increases, countries with huge light oil reserves like Saudi Arabia, Iran and Kuwait became interested in developing their vast HO NFCR assets to boost their declining production rates. Productive HO-bearing NFCR’s in the Middle East are characterized by low matrix permeability and high fracture permeability. Large-scale oil flux is through the high permeability fracture system, whereas the matrix-fracture interaction mainly controls recovery efficiency and maintaining production levels; this is the type of production currently taking place from reservoirs in Oman, Iran, Iraq, Syria, Turkey and Egypt. In the following sections, HO resources in NFCR’s and HO recovery activities from HO NFCR’s and future plans in different countries in the Middle East are reviewed and discussed. The geographical distribution of HO in the Middle East, mainly occurring in NFCR’s, is presented in Figure 1.
Iran Iran (>72×106 people) is the 4th largest oil producer in the world ~4.2×106 b/d. According to the US-DOE EIA in January 2008, Iran has the third largest proven oil reserves in the world, ~138×109 b, after Saudi Arabia and Canada. It should be noted that this includes no HO reserves. Many believe that Iran has passed its peak oil production, others believe production in Iran could increase and peak in the next 15-20 years. Currently, production is declining 4-6%/yr while domestic consumption is increasing 6-8%/yr leading to decreasing net oil exports. This declining conventional oil availability, combined with a desire by Iran to sustain current production level as the major source of national income, is triggering interest in the HO resource.
2
The first discovery of HO in Iran goes back to early 1931 when wells drilled for gas reservoir evaluation in the southwest of Iran encountered a large HO occurrence. Systematic HO exploration in Iran started in 1982; in 1994 the Petroleum Engineering and Development Company (PEDEC) was established within the National Iranian Oil Company (NIOC) to manage national HO and XHO assets. Significant HO discoveries in the south and southwest of Iran, combined with high oil prices and difficulties in sustaining conventional oil production rates, have led to a greater interest in their development since 2000 [10]. According to the latest studies, Iran has over 60-90×109 b of HO, mostly occurring in NFCR’s (limestone and dolomite) and comprising >40-65% of Iran's proven oil reserves. Several NFCR HO fields have successfully cold produced oil qualities on the order of 6-18°API, but the HO NFCR’s reserves base remains undeveloped mainly because of lack of appropriate technology. Progressing cavity pumps were successfully field tested and implemented in an Iranian HO NFCR in 2007 and daily production of 3,500 b/d was recorded. It should be noted that current HO production in Iran is about 105 b/d, dominantly from a HO sandstone reservoir with in situ μ of ~600 cP. Iranian HO reservoirs are found in Upper Jurassic and Cretaceous to early Tertiary carbonate formations in the Persian Gulf and surrounding areas. Most reservoirs are composed of pelletal, oolitic, or bioclastic greenstones and reefal limestone that have high primary porosity and permeability. Reservoirs are sealed either by tight limestone, massive anhydrite, or by impermeable argillaceous rocks, and these seals are effective throughout most of the Persian Gulf and surrounding areas [10]. Most HO reservoir rocks in Iran are structural traps, i.e.: anticlinal structures, found in limestone and dolomite ranging in age from Cretaceous to Eocene. The most important HO fields in the southwest of Iran are Kuh-E Mond, Ferdowsi, Zaqeh and Paydar. Because of geological variability, production issues, and uncertainty in application of new production technologies in NFCR’s, these fields have not been systematically developed, with only a few wells here and there on slow production. The contribution of HO reserves in Iran to total oil production is negligible, and it remains to execute a technically-based assessment of these and other fields to choose optimum production approaches, or modify existing approaches to cope with particular geological circumstances [10]. It seems that the potential importance of new oil production technologies in Iran for HO production form HO NFCR’s is substantial.
Turkey has five major HO fields, mainly in the southeast (Figure 1): Bati Raman (OOIP ~1,850 ×106 b), Raman (OOIP ~400 ×106 b), Çamurlu (OOIP ~377 ×106 b), Garzan (OOIP ~163×106 b) and Bati Kozluca (OOIP ~138×106 b) with viscosities of 2,260-64,000 cP @ 20 °C and °API ranging from 12-18°. The NFC HO Bati Raman field discovered in 1961 is the largest oilfield in Turkey; it is a low-pressure field containing low-gravity (12°API [0.986-g/cm3]), high-viscosity (592 cP in situ) oil at an average depth of 1,311 m with a gross thickness of 64 meters and 1.85×109 b [294×106 stock-tank m3] OOIP. The producing formation is the Cretaceous Garzan limestone in an elongated east/west asymmetric anticline measuring 17 km by 4 km. The Garzan limestone has a reefal origin and a fractured, vuggy character exhibiting areal and vertical heterogeneities. The structure becomes chalky and thus tighter to the east. Average porosity is 18% and is mainly vugular and fissured in type. Average matrix permeability determined from cores is 16 mD. Well tests indicate 200-500 mD effective permeabilities, confirming the contribution of secondary porosity. In the western and central portions of the field, a secondary vugular porosity interconnected by fissures appears to be superimposed over low primary matrix porosity [12, 13, 14]. Turkey is the pioneer of CO2 HO production in the Middle East. Despite its complex and heterogeneous geology, the Bati Raman HO field is the only successfully commercialized immiscible CO2 HO production project in the world [15 after 16]. During almost four decades of HO production from Bati Raman, several HO production methods have been tried, such as immiscible carbon dioxide injection using vertical wells, horizontal wells and various pressure maintenance techniques. The main primary production mechanism is rock and fluid expansion. Water drive appears to be insignificant except for a very weak aquifer influence at the central north flank wells. The solution Gas Oil Ratio (GOR) is 18 scf/stb, resulting in a low bubble point pressure, ~1.10 MPa. Before CO2 injection, the reservoir pressure did not decrease below the bubble point pressure; therefore, in practice, there was no solution gas drive mechanism. The original reservoir pressure was 12.40 MPa, which dropped to an average of 2.76 MPa after cumulative production of 30×106 b, prior to CO2 application. Before the project began, 65 active producers were pumping, with a total production rate of 1,600 b/d compared with the year 1969 peak rate of 9,000 b/d. Initially, well production rates were up to 400 b/d; this decreased then to an average of 25 b/d before treatment [13, 14]. Primary recovery prospects were low mainly because of unfavorable oil properties (low °API gravity, low solution gas and high viscosity), low reservoir energy, and the type of driving mechanism. It was estimated that RF of 0.015 could be produced via cold production. The reservoir history and unfavorable properties caused rapid declines in reservoir pressure and production, suggesting the need for a suitable production technology to increase ultimate recovery from this vast reservoir. Since 1968, several pilot tests have been conducted, including cyclic steam injection, steam drive, air injection, and water flooding. The results of extensive laboratory, modeling, and engineering studies; the presence of a CO2 reservoir 55 miles from Bati Raman, and economic considerations led to an immiscible CO2 huff-'n'-puff-type method. The reservoir gas from Dodan field is almost pure CO2 (Dodan production capacity is 60 MMscf/d [13, 14]). Monitoring of oil production rate, injection pressures, gas/oil ratio (GOR), and gas utilization factor in the reservoir
Turkey Turkey (>71×106 people) has a rapidly growing economy and therefore growing energy demand; it is not a rich state in terms of oil reserves. Turkey’s proven recoverable light oil reserves are ~300×106 b (US-DOE EIA, January 2008), a figure that does not include any HO reserves. Total production in Turkey is only ~44×103 b/d, almost negligible compared to national consumption of 640×103 b/d. Major HO producing fields in Turkey are Bati Raman, Bati Kozluca, Garzan, Camurlu and Raman. A limited local light oil resource in Turkey was the main driving force for Turkish oil industry to consider HO production during the last half century. HO resource (OOIP) in Turkey is estimated to be ~6,650×106 b. Heavy oils in Turkey are mostly mobile under in situ reservoir conditions and can be produced with primary cold production but the RF (recovery factor) remains very low. Nevertheless, Turkey has been one of the leading countries in heavy oil production from NFCR’s in the region since 1963.
3
peaked in 1978 at 38×103 b/d, despite the high oil viscosity. Waterflooding was implemented in 1986 following a number of pilot project trials. The original development plan was waterflooding using an inverted nine-spot vertical well pattern, but early water breakthrough led to a change of drilling and development pattern in 1994-95 to a vertical line drive oriented parallel to the dominant NW/SE trend of faults and fractures in the field. During the initial cold HO production period, the oil contained in the highly permeable network of fractures was produced with a low RF of ~0.02. The major oil production mechanisms in the period 1975-1996 are emptying of the fracture system, fluid expansion due to pressure reduction, and isothermal GOGD. By 1979, a small secondary gas cap in the fracture network was formed because of the pressure drop near or slightly below the bubble point in the crest of the field and gas injection into the production wells for gas lift purposes. Simulation studies based on a history match of the first 20 years production indicated that half of the oil was likely supplied by emptying the fracture system, the other half coming from pressure reduction of the matrix hydrocarbon [21, 22]. Shahin et al. [23] investigated TA-GOGD application in Qarn Alam, hoping to increase the RF to 0.27%. TA-GOGD has been piloted successfully and is now being implemented in the field at full scale. Despite the fact that Qarn Alam is relatively shallow, ~360 m, with a higher matrix porosity and permeability in comparison to other NFC HO fields, the industry is learning much from this pioneer project.
after commencement of CO2 injection revealed three different periods in the history of Bati Raman HO field. 1986-1993 can be defined as the “fill-up” period, in which the injected gas filled the fractures and vugs in the reservoir. Because of the high compressibility of CO2, injection pressures are stable all over the field. 1993-1996 is the period of increased oil production rate; in which daily rate of 12,000 b/d was reached and remained almost constant. The last period is obviously the (current) decline period. The main cause of this is the high heterogeneity of the filed with combination of adverse factors and very low resistance pathways in which the recovery fluid can channel. Production rate declined substantially since 2000, and by 2003, the RF in Bati Raman reached 0.05 [17], still a very low value. The major reason for low RF is the heterogeneous nature of the reservoir reducing efficiency of most of the HO production processes. In 2004 a polymer gel treatment method (a fracture plugging polymer gel system) was implemented to impede or reverse the decline by increasing CO2 sweep efficiency; sweep efficiency was increased by 12% using this technique. The Bati Raman NFC HO field is still producing HO using CO2 immiscible injection and gel treatment techniques [18].
Oman Oman is one of the pioneer countries in HO production from NFCR’s in the Middle East. Because of limited conventional oil reserves in the country compared to other oil-rich states in the region and following huge discoveries of HO accumulations mainly in clastic reservoirs and to a lesser extent in NFCR’s back in late 1960-70 [19, 20], Omani authorities became more interested in HO production. Petroleum Development Oman (PDO) supported numerous research projects and pilot tests on HO production from NFCR’s. Daily HO production in Oman from NFCR’s is about 50×103 b/d which is about 6% of daily oil production of 734×103 b/d from conventional oil reserves in the country. Oman is the third largest HO producing country after Iraq and Egypt in the Middle East. Proven conventional oil reserves in Oman are about 5.50×109 b (US-DOE EIA). Following major HO discoveries in Oman, estimated at ~10-15×109 b, different HO production strategies were investigated; some have been field tested and implemented such as steam drive, hot water injection, Thermally Assisted Gas-Oil Gravity Drainage (TA-GOGD) utilizing horizontal, multilateral and intelligent well technologies in a number of HO reservoirs. The Qarn Alam in Oman, a NFC HO field with OOIP of ~1.34×109 b [213×106 m3] and a 16°API and 200 cP in situ oil viscosity oil, has been the focus of HO production in Oman during the last decade. This Northern Oman field is an intensely fractured carbonate situated in an anticline structure of 6×3 kilometres with NNE - SSW orientation. The major oil bearing reservoirs, the Shuaiba and Kharaib formations, are separated by a very low permeability oil-bearing zone called the Hawar. Fracturing is reported through all zones, and it is believed that the fracture networks are in hydraulic communication with a very active aquifer. Initial oil saturation is ~0.95 and initial water saturation is the irreducible connate water content. The matrix porosity is about 30%, exceptionally high for carbonate reservoirs in the Middle East, while the matrix permeability ranges between 5 and 20 mD. Fracture
Egypt HO oil resources in Egypt (81×106 people) are ~16×109 b [2.55×109 m3], of which 90% is in clastic reservoirs, mainly sandstones and some conglomerate formations. Egypt is the 2nd largest HO producer in the Middle East from HO NFCR’s with producing 55×103 b/d. Proven conventional oil reserves in the country are ~3.70×109 b [0.59×109 m3] and daily conventional oil production is about 667 Mb/d (US-DOE EIA). Issaran and Bakr Amer HO fields are amongst the NFC HO fields in Egypt, and Issaran field of Miocene age with OOIP of 0.50×109 b [79×106 m3], was discovered in 1981. The producing strata are heterogeneous carbonate formations with average depth of 330-760 meters, the Upper and Lower Dolomite and the Nukhul Formations. The Nukhul Formation reservoir is a tight limestone matrix with a well-connected leached-fossil mold system with macro-pores, accounting for 15% of the Issaran field reserves. The average reservoir temperature is 85-100°F, the bottom-hole pressure ~650-700 psi, and at reservoir temperatures, oil densities are 0.964 to 0988 g/cm3, and viscosities are ~3,000 to 5,000 cP. Porosities range from 23% to 33%, and permeabilities from 1.3 to 104 mD. The matrix permeability for the Nukhul Formation is estimated to be very low, but fracture permeabilities exceed several Darcies. Before 1999, nine wells had been drilled in the field giving 450 b/d production. The field has been producing 1,800 b/d from five new wells drilled in the area during late 1999 and 2000. A ten-well CSS pilot test, 2004 to 2006, has led to a substantial increase in HO production, and the three best CSS wells produced an average of 230 b/d/well, also limited by available pumping equipment. Average production during 2006 in the field was 4,474 b/d from all wells, and 30 additional CSS wells and facilities were planned for 2007, and 50 more for 2008. Over 600 CSS wells and related thermal facilities are planned over a five year period [24].
permeabilities of 0.10-1000 D have been reported, based on pressure build-up tests in the field [21, 22]. The Qarn Alam field was discovered in 1968 and first put on primary cold production in 1976; the net oil production 4
Due to high reservoir heterogeneity and complex lithology in the Issaran NFC HO field, along with a low reservoir pressure, achieving a high RF is challenging. Better reservoir characterization with implementation of new production and well completion technologies will help in producing more oil from this complex and difficult NFCR.
measured at different temperatures. A decrease in the atomic H/C ratio with an increase in T was observed for all runs. The activation energies of Çamurlu (12°API), Bati Raman (12°API), and Raman (18°API) heavy oils for both fuel deposition and high-temperature oxidation reactions were similar. For medium-gravity oils (Adiyaman, 26°API and Garzan, 28°API), the activation energies for high-temperature oxidation is higher than for fuel deposition. For the LTO in carbonates, the activation energies are almost twice those of FD and HTO reactions for each crude oil, and are almost independent of oil densiy; i.e. the Arrhenius constant is not affected by the API gravity [27]. During fuel deposition, crude oil is coked and C is deposited on the solid matrix to be consumed as fuel by the advancing combustion zone. At low temperatures, oxygen in the produced carbon oxides (CO, CO2) is less than the oxygen consumed, indicating that some O2 is used in other reactions. At high temperatures, however, essentially all O2 is stoichiometrically consumed to produce CO2 and CO, indicating complete combustion. This suggests that LTO and cracking reactions are occurring in the lower-temperature region, and increased O2 consumption and greater production of CO2 and CO at lower temperatures for the heavier oil indicate a higher reactivity of these oils with O2 t low T in carbonate rocks. Another feature of the combustion of the HO NFCR’s is that CO2 concentration is much greater than that of CO in LTO and HTO regions. Thus, although in the light case the total fuel deposited or burnt is low, compared to that for the HO case, almost all of the oxygen consumed is accounted for, particularly in the HTO region. This suggests that less cracking has occurred during fuel deposition and combustion than with the HO. Sweep efficiency during in situ combustion is vital, yet poorly understood. Most laboratory tests are conducted in combustion tubes, which, because of their one-dimensional geometry, can not provide information on either areal or vertical sweep. Combustion sweep efficiency must be understood for assessing past performance, predicting future performance, and comparing processes for optimization. 3-D scaled physical models can provide such insight as well as combustion stability assessment over a range of operating conditions, helping in field performance predictions and numerical simulator verification. An experimental study was conducted with various configurations (vertical injector-horizontal side producer, vertical injector-diagonal horizontal producer, and vertical injector-vertical producer) in a 3-D model using two different heavy oils [28, 29] – Raman (18°API) and Bati Kozluca (12°API) oils. Because the igniter was located near the centre plane of the model, it was observed to be much warmer than the top and bottom planes. The occurrence of HTO at the vicinity of injection end in both runs was verified by examining the reaction kinetics. The heat front propagation in Raman oil was faster at the beginning and slowed down later in the tests. Peak temperatures observed in Raman oil were lower than those in Bati Kozluca oil, and a stabilized front was observed in Raman oil, absent in Bati Kozluca oil due to severe bypassing (advective instability). In Raman oil with a lower air rate, channelling was severe. It can be concluded that superficial burning is usual in vertical injector-vertical producer configurations, giving poor vertical sweep. With a vertical injector-horizontal producer configuration, a uniform temperature distribution throughout the centre plane in both oils was observed at the start of air injection. The front stabilized in both runs and proceeded in the direction of the producer. The creation of regular isotherms implied parallel to the side horizontal producer in both runs a
Kuwait Kuwait (2.5×106 people) is the 10th largest oil producing country in the world with ~2.67×106 b/d. Kuwait holds the 5th largest accredited oil reserves in the world, ~104×109 b, not including any potential HO reserves. Kuwait has >40-50×109 b of HO in both clastic and NFC reservoirs. Daily HO production in Kuwait is about 28×103 b/d, negligible compared to conventional oil production in the country. Among the clastic HO reservoirs in Kuwait, the Lower Fars (LF) sandstone reservoir in northern Kuwait is probably the single largest accumulation of heavy oil in Kuwait, containing ~12-15 ×109 b distributed over an area of ~1000 km2 northwest of Kuwait City against the Iraqi border [25]. Although a small resource in comparison to Canadian and Venezuelan HO resources of over 1012 ×109 b each, the LF represents a significant fraction of Kuwaiti HO resources. There are several HO fractured carbonate reservoirs in Kuwait with total OOIP of over 30×109 b. These are not yet developed and some are situated in the divided zone between Saudi Arabia and Kuwait. Wafra is the largest HO carbonate filed in the divided zone with two reservoirs with 13-20°API oil, depths of ~400-100 m, and good matrix porosity and permeability [26]. Cold production seems to be the first rational production option because of its lower CAPEX and OPEX compared to thermal methods. Subsequent thermal techniques based on horizontal well technology such as SAGD and CSS have to be investigated carefully for optimal strategies.
Production Technology in HO NFCR’s In this section some of the thermal and non-thermal commercialized and emerging HO production technologies, numerical simulation, modelling and experimental studies in HO NFCR’s are reviewed and discussed.
Thermal Methods in Heavy Oil NFCR’s In Situ Combustion in Heavy Oil NFCR’s In situ combustion is a technique suitable for recovery of oil from medium viscosity HO reservoirs, but has yet to achieve success in high viscosity cases. Oil is ignited at the wellbore and a front is propagated by continued air or oxygen injection. The combustion front is sustained as long as enough coke is produced by cracking of crude oil to be consumed as fuel.Because it is potentially so valuable as a production technology (no water requirements, no steam costs, reduced CO2 emissions, less waste, less refining, no solvent use…), we discuss some experimental work on Turkish oils here. Sixteen experiments were conducted to study the combustion reaction kinetics of Turkish heavy oils (12-32°API) in carbonates. A mixture of the carbonate rock and HO was subjected to a controlled heating program under constant air flux. The produced gas was analysed for its oxygen and carbon oxides contents. Although the molar CO2/CO ratios vary during low-temperature oxidation (LTO), for fuel deposition (FD) and high-temperature oxidation (HTO), these values can be 5
distributed flow field at the centre plane. Early production of hot fluids that transport heat as a result of convection was noted, showing heat removal from the burned-out sand pack by vaporization of water behind the combustion front, which was deposited ahead of the burning zone by condensation in cooler regions of the reservoir. The horizontal production well conveyed heat downstream into the colder regions and no indication of flow channel or bypass was noted. Since horizontal producers have a larger areal contact, they can drain fluid without necessarily generating flow bypassing. With the same burned volume it was observed that horizontal producers recovered more oil than vertical producers. Although, dual horizontal producers recovered the highest amount of oil, the net recovery per producer was lower than that of a single producer placed alone at the boundary. The main reason for this is presumably the flow interference between the twin producers due to the restricted drainage volume in the model. A one dimensional simulation model was built using CMGSTARS and was calibrated with experimental data [30]. Parameters such as T profiles, cumulative oil and water produced, pressure profiles, amount of heavy and light oil produced were investigated and compared with results. In this model six components and pseudo components, water, heavy oil, light oil, oxygen, inert gas and coke, were considered. Simulated oil and water production lag behind actual oil and water production, respectively, during nitrogen injection. However, once air injection starts and combustion occurs, simulated oil production increases rapidly. Toward the end of the experiment, simulated oil production is about the same as the actual oil production. Combustion tube test results were also simulated using a model representing oil as two components: “heavy” and “light” oil fractions. Also, coke was described as a solid phase and four chemical reactions were used to describe cracking of heavy oil to light oil and coke, HO burning, light oil burning, and coke burning. In summary, combustion tube test results can be reasonably matched with a four-reaction model. Matching combustion tube test results yields some confidence in fluid model as well as chemical reaction descriptions, which might then be used in a field-scale model. Having obtained a reasonable history match of laboratory combustion tube data, the same fluid description and chemical reactions were used to simulate production performance for a field situation using a 1/8th element of symmetry for a 5.0 acre, inverted five-spot pattern. Homogeneous reservoir properties chosen were similar to the Bati Kozluca HO reservoir, and homogeneity was used at this early stage because of numerous modelling and process description uncertainties inherent in combustion simulation. A Cartesian, three-dimensional model with 12×12×5 grid blocks of ΔX = ΔY = ΔZ = 19.4 ft, and with six components was used (water, heavy oil, light oil, nitrogen, oxygen and coke). HO was represented as a two-component model oil of 12°API and molecular weight of 675. HO burning is represented by two chemical reactions corresponding to the air-fuel ratio requirements calculated from the laboratory observations. The chemical reaction uses activation energy and The frequency of 85,600 and 2×1010, respectively. homogeneous reservoir has [k]ij = 700 mD, φ = 0.25, thickness = 100 ft, and initial conditions of T = 77°F, oil saturation = 0.75, water saturation = 0.25, and pressure = 1000 psia at 4,000 ft depth. Constant air injection rate of 10×106 CF/d is maintained for a 5-yr simulation period. The producer operates under a constant flowing bottom hole pressure of 14.7 psia. The numerical model incorporated an external heater option to raise the temperature of the injector at the beginning. To
simulate adiabatic conditions, no external heat losses or gains were allowed. Several configurations and locations were tried while keeping well lengths constant: vertical injector - vertical producer, vertical injector - horizontal producer (right side) and vertical injector - horizontal producer (left side). Results are comforting. The reservoir centre plane is observed to be superior to the top and bottom regions for the combustion process. A uniform temperature distribution throughout the centre plane was observed shortly after achieving combustion front stabilization and it moves stably toward the producer in the VI-HP well configuration. The generated isotherms are parallel to the producer, which suggests an evenly distributed flow field at the centre plane. Early production of hot fluids, which transport heat as a result of convection, was noticed, but no indication of flow channelling or bypassing. Vertical advective instabilities are not severe, and the temperature levels at the top and bottom planes are comparable, implying uniform combustion in the reservoir. With the same burned volume, more oil was recovered by horizontal producers than that of vertical production wells. This was due to the fact that the horizontal wells remove oil without the creation of any extensive mobile flow path in the colder region, where as, because of their flow geometry, more volume should be burned to create a flow path between a vertical injector and a vertical producer with in which the fluid can flow readily. In general, for the progression of the simulation, the results tended to follow closely the fire tube results. These laboratory and numerical experiments of HO combustion in a carbonate suggest that simulation is feasible and therefore there is a chance that predictions in NFCR’s can be achieved if a suitable method of incorporating and analyzing heterogeneity at appropriate scales can be implemented in the simulation. Of course, this remains a huge challenge, but a necessary one, if the promise of the value of in situ combustion in NFCR’s is to be realized. Then, we will be in a far better position to realistically address issues such as well placement and rate control in gravity-segregated combustion process designed to achieve higher RF values. Steam Injection in Heavy Oil NFCR’s The most widely used thermal technique to extract residual oil from HO fields is steam injection. There is also an interest in using CO2 gas as an immiscible phase for recovering HO under appropriate p-T conditions. Research into the application of a simultaneous steam-CO2 drive process and the examination of vertical and horizontal injection-production well configurations was conducted in a physical model of a symmetry slice, 1/12th of an inverted regular seven-spot pattern, to determine recovery performance at Bati Kozluca (12°API) HO [31]. Vertical injection and production wells, vertical injection and horizontal production wells, and horizontal injection and production wells were explored. With steam alone, the vertical injector and horizontal producer scheme gave higher RF than the others; the lowest RF was obtained from the horizontal injector-horizontal producer well configuration. The co-injection of CO2 with steam increased RF and the production rate over steam alone, and again the RF of the horizontal injector-horizontal producer was the lowest, the vertical injector-horizontal producer the highest. When steam alone and steam-CO2 tests were compared, the oil recovery increased with increasing CO2/steam ratio till an optimum value, then a diminishing effect was observed. The distance between the wells also affected the efficiency of the process. 6
Addition of non-condensable gas to steam is known to have beneficial effects on HO production when conventional vertical wells are used. 1-D and 3-D physical models were used to examine the effects of simultaneous injection of CO2 and CH4 together with steam on the recovery of Bati Kozluca (12°API) HO mixed with unconsolidated carbonate rock [32]. Gas/steam ratio was found to be a significant variable in rate and RF values. CO2/steam ratio values above the optimum value led to a diminution of oil recoveries, and this was attributed to the increase in gas saturation along the model causing steam channelling through the production well and the increase in gas volume simply reducing the amount of steam injected. The steam injectivity was also less at higher CO2 concentrations. Lower residual oil saturations were obtained in gas-steam injection tests, compared to the values obtained with steam alone. The injected non-condensable gas created a permanent over-riding gas phase, reducing vertical heat losses, so high temperatures arrived at the producing well earlier than for steam alone. Depression of steam temperature was also observed due to the presence of non-condensable gases, CO2 and CH4. The higher temperature in the steam tests indicated that the steam overriding at the top of the model, and the bottom of the model was warmed conductively because the steam stayed at the top of the model. For steam-gas tests, temperatures increased and then decreased following the injection pressure behaviour changes (constant injection rate led to different injection pressures over time). However, for steam alone, the injection pressure stayed at the same level eased level during tests. Steam-oil-ratio (SOR) is often used to gauge steam methods’ success. SOR increases as the thickness of the bottom water zone increases for a horizontal producer. Because bottom water is more mobile than oil and is thus heated rapidly, it is displaced more rapidly by horizontal producers with larger areal contact, and as a result, more steam invades the bottom water zone. This demonstrates the improved injectivity of steam in the presence of bottom water, but also the degradation of thermal efficiency. In the absence of bottom water, the highest SOR in physical models is vertical injection-vertical production configurations, although the SOR decreased toward the end of the experiment and became almost similar to that observed in vertical injectionhorizontal production configurations. During the experimental program, well configurations and bottom water thickness were changed to determine effects on RF. Maximum RF was obtained by placing the horizontal producers along the hypotenuse of the triangular model; this configuration provided the best oil recovery even in the presence of bottom water. RF decreased with an increase in the thickness of the bottom water, and there are two important cases in using steam in the presence of bottom water. In the first case, the oil above the bottom water zone invades it and mostly remains as residual oil, showing the reason for lower RF; the second case leads to an increased tendency for coning of condensed steam and water into the vertical well.
introduced into the oil-filled NFCR fracture system, leading to a gas-cap in the fractures. GOGD matrix drainage rate is low because of high viscosity and low permeability, and to overcome this heat must be added. Therefore cold gravitydominated drainage in Qarn Alam is enhanced thermally, leading to the acronym TA-GOGD. Heat transfer into the matrix from fractures is conduction-dominated initially, and as deeper and deeper oil is heated, it drains partly to the fractures, but partly down within the matrix block losing its excess heat to colder oil, slowing the drainage process (cold viscous oil represents a “drainage barrier”). Apparently, in a manner similar to horizontal shale flow baffles in gravity drainage, the cold oil barrier diverts the hot oil flowing vertically under gravity forces within the block to the block flanks, where it encounters the gas-filled fractures. Then, the oil can flow downward along the carbonate-gas interface. Of course, the oil will also cool as it drains vertically within the fractures to the oil interface (below the gas cap), and continued steam injection aids in sustaining the heat and the downward flux of liquids (oil and condensed water). The Qarn Alam pilot also had other goals, such as to demonstrate matrix and cap rock integrity during exposure to heat and condensed steam, sustainability of steam injection, and proof of economic considerations based on oil rates and RF values. The pilot project has demonstrated that the method works in heavily fractured regions (small block size), but in NFCR’s where the distance between fractures is large, such a method remains unproven. Nevertheless, the positive impact on the fractures in providing a shortened flow path for the hot oil in the matrix blocks, combined with the clear gravitational segregation aspects of this production mechanism, is a highly positive development in extraction of HO from NFCR’s [33, 34]. Hernandez and Trevisan[35] numerically studied drive mechanisms (solution gas drive, CO2 generation, steam distillation, capillary imbibition, gravitational drainage) for oil and gas recovery during steamflood of a live-oil HO NFCR model. Two numerical sub-models were used; one for the horizontal cross-sectional heating process, one for the vertical gravity-dominated flow cross-section, but both subjected to steady steam flooding. Steam distillation proved to be the most effective drive mechanism, totally recovering the CH4 and the light oil components (“pseudo-components”). Solution gas exsolution increases the pressure difference between the matrix and the fracture flow systems, and apparently the contribution of capillary imbibition to production is negligible because of the high viscosity of the residual oil components after steam distillation. Thermal expansion was as important as solution gas drive, and helped generate pressure gradients toward the fractures which drained the oil. Apparently, the gravitational drainage mechanism has little influence on oil recovery from the matrix, except in aiding the recovery of the heavy pseudocomponents. Of course, these conclusions remain to be verified in field conditions. SAGD in Heavy Oil NFCR’s Steam Assisted Gravity Drainage (SAGD) is a promising recovery process for producing HO and bitumen. The method ensures stable steam displacement by using gravity as the driving force. USusally, in current configurations, a pair of vertically offset horizontal wells are used for injection/production, but this is really only necessary if the oil in the fractures is almost fully immobile. In the case of low-μ HO (mobile) in NFCR’s, the location of the horizontal wells can be far more flexible, depending on the details of the process and the reservoir characteristics. It is even feaible to use
Thermally-Assisted-Gas-Oil-Gravity-Drainage (TA-GOGD) in the Qarn Alam fractured low permeability carbonate reservoir is being enhanced by steam injection; it is the world’s first full-field HO carbonate thermal development. Gravity drainage is typically used to drain heavily fractured low permeability reservoirs, particularly carbonates; the process relies on the density difference between matrix oil and fracture water (WOGD), or fracture gas and matrix oil (GOGD). Cold Gas-Oil-Gravity-Drainage (GOGD) occurs when gas is 7
vertical basal injectors, although to achieve economical production rates it is necessary to have long producing wells at the base of the oil zone (oil does not flow up hill in gravity processes). Steam is injected into one well and hot liquids produced from the other, generating a steam chamber which grows laterally by steam condensation and heat transfer at the lateral boundary, heating the oil and reducing its viscosity. Heated oil and water are drained by gravity along the chamber walls of the production well [36, 37, 38, 39], and the void space created by the drainage of the liquids is filled by non-condensed steam, exsolved gases (CH4 in solution), and injected inert gases. Coinjected hydrocarbon gases such as propane, butane or higher MW molecules also give a viscosity reduction component because they dissolve partially into the HO. SAGD may be an attractive recovery process for HO NFCR’s [40]. In these cases, thermal conduction allows heat to sweep the matrix volumes in the reservoir which may have little direct contact with steam because of capillarity blockage. Thermal expansion and the expansion of exsolved solution gas will be important recovery mechanisms. Gravity return flow of hot liquids on the inclined steam chamber flanks continues to transfer heat to the matrix blocks as liquids find pathways through the fractures toward the producer wells. Steam and non-condensed gases stay high in the zone, and cap rock integrity is essential to contain the heat and sustain thermal efficiency. Compared to unfractured homogeneous systems, vaporization ad steam distillation in fractured systems takes a longer time simply because steam flows at a high speed through the fractures, heating the reservoir matrix only slowly through conduction [41]. Apparently, including results reported herein, studies (experimental and numerical) have been limited to date to relatively homogeneous case and to unfractured sandstones. Thus, the effect of heterogeneity, anisotropy, and fracture systems on SAGD remains ill understood. The TA-GOGD results suggest that steam injection in fractured reservoirs may have economic potential because even though injected steam moves rapidly through fractures, the heat front moves uniformly if the process is carried out in the gravity-dominated regime, avoiding advective instabilities. Results to date indeed show that heat can be efficiently transferred from injected steam and hot liquids to the reservoir matrix. Bagci [42] presented an experimental investigation of the effect of fractures and well configurations on the steam–assisted gravity drainage (SAGD) process in a 3-D model, using Bati Kozulca HO (12°API) in carbonate rocks. Three different well configurations were used to assess the influence of fracture distribution on the steam-oil ratio (SOR) and oil recovery - a horizontal injection and production well pair, a vertical injection-vertical production well pair, and a vertical injectionhorizontal production well pair, with and without fractures that provided a vertical path through the horizontal producer. Results indicated that vertical fractures improved SAGD performance, other factors remaining equal. Maximum RF was observed in the double horizontal case in fractured rock, because of the favourable steam chamber geometry. Obviously, fracture location and connectivity affect process performance; very high permeability path (fractures) above the injection well actually aids the natural convection of steam resulting in efficient oil drainage. During the early stages of the experiments, the fractured model gave significantly higher SOR’s than the uniform permeability models because or more rapid fracture drainage. Apparently, whether heat communication between the wells is developed before SAGD is implemented is an important factor in fractured HO cases. Heat
communication influences the growth rate and geometry of the steam chamber. If full heat communication is established, better lateral growth behaviour is observed; because this also restricts the relative vertical growth rate, it reduces heat requirements. If heat communication does not from, the steam chamber grows rapidly in the vertical direction, increasing the temperature gradient in the model top, leading to increased heat losses and higher SOR values. Single-well SAGD techniques have attracted some interest. A number of oil companies are developing or evaluating SWSAGD because there are many HO reservoirs that are thin or underlain by water. If the process can be effectively applied, a large amount of resource currently considered uneconomical may be accessed. Although SW-SAGD and its variations are being field tested by several operators, a systematic evaluation of the process and its performance is not available. Some suggest that the SW-SAGD process is essentially a nearwellbore heating process effective for a finite (150 m) length of the horizontal section (up to 1000 m long?), yet others are more optimistic (as we are) because steam injection distributed mechanically over the entire well length uniformly will soon be developed. Akin and Bagci [43] investigated the optimization of a SW-SAGD start-up procedure in Bati Kozluca HO field. They experimentally studied two early-time processes to improve reservoir heating, cyclic steam injection and steam circulation, and compared results with other gravity drainage well configurations. The steam chamber volume for cyclic steam injection is slightly greater than that of steam circulation case, and should be favoured if SW-SAGD is implemented experimentally in a field case for HO in NFCR’s.
Non-Thermal Methods in HO NFCR’s CO2 Immiscible Injection in HO NFCR’s As a part of the EOR analysis of the Bati Raman HO field in Turkey, the implementation of steam displacement and in situ combustion have been examined as well. Although the theoretical recovery data were encouraging, in situ combustion is associated with high risks, uncertain technology and operational problems, so it was eliminated. Both CO2 and steam applications were deemed favourable. However, immiscible CO2 application appeared more so because of the nearby Dodan CO2 gas reserve and the high initial investment associated with steam injection. Based on laboratory successes and the vast amount of CO2 locally available, field-scale CO2 huff-and-puff was started in the early 1980’s. Due to rapid breakthrough of CO2 to offset wells, the project was converted to field-scale random pattern continuous injection. Over 20 years of injection, the HO production peaked at ~13,000 b/d and began to decline, reaching today’s ~7,000 b/d value. In the case of Bati Raman, at this mature state of the process, the injected CO2 is increasingly bypassing the remaining oil, and production is curtailed by high GOR’s due to the naturally fractured characteristics of the reservoir rock. From the field observations and the simulation studies carried out by using the performance data, the following results were interpreted as the early remarks: (1) a considerable amount of oil is produced due to the high flooding effect of gas, (2) reservoir behaviour is clearly dual porosity, and (3) diffusion of carbon dioxide into the oil is an effective drive mechanism [44] (swelling, some viscosity reduction, some displacement).
8
practically achieved [48]. An optimization study was conducted to design and operate an efficient VAPEX process in a HO NFCR in Iran [49, 50]. A mechanistic model was developed for the study of the VAPEX process by considering the mass transfer and fluid flow mechanisms characteristics of the VAPEX process. The model is capable of predicting the drainage rates of HO in all stages of the process, including the pseudo-steady state. The model was further applied to a fractured system and effects of solvent injection rate, fracture and matrix permeability, matrix to fracture permeability ratio, and initial viscosity of the reservoir heavy oil on the performance of the VAPEX process were studied. Higher oil production rates were obtained due to higher injection rate of solvent in HO reservoir. The reservoir pressure was increased with increasing high rate and it resulted to early breakthrough of the solvent, especially in fractured reservoirs. Oil production rate increased with the increase in the matrix permeability for a given matrix-to-fracture permeability ratio. These results clearly show the importance of remaining in a gravitydominated domain in the case of VAPEX-type processes in HO NFCR’s.
Water-Alternating-Gas in HO NFCR’s Water-Alternating-Gas (WAG) injection was originally proposed as a method to improve the sweep efficiency of gas injection, mainly by using the water (and capillarity effects) to control the mobility of the displacement and to stabilize the front. The feasibility study was done for the Bati Kozluca HO field by carrying out necessary geological and engineering studies to determine the best method to increase the oil recovery and to determine the efficiency of CO2 injection as an EOR method for future development. Bati Kozluca is a HO NFCR having high oil viscosity and low aquifer support, the main constraints for production. Optimum injection pattern, number of injection wells; injection periods and injection rates were studied for WAG processes in Bati Kozluca HO field. After sensitivity analysis, 60 day of CO2 injection with 1×106 scf/d/well and 30 days of water injection with 800 stb/d/well was chosen as the best case. The effect of additional perforations and new wells were also studied. Since additional perforations both accelerate oil production and increase the cumulative oil production, additional perforations can be opened prior to the start of a project. However, infill drilling of new production wells only accelerating the oil production and have a limited effect on RF because of drainage area interference. With this study, it was predicted that applying WAG method cumulative oil production will be around 14.2×106 stb (10.3% of OOIP) compared to 7.5×106stb (5.5% of OOIP) without WAG application [45]. A potential severe constraint is rapid segregation of the water and gas alternating phases in the fracture system, leading to loss of effect
Results and Discussion Approximately ⅓ of known HO is in carbonate reservoirs and the Middle East has over 20% of the world’s known HO resource, found mainly in NFCR’s. Middle East HO and XHO resources are estimated to be ~500-900×109 b [90-142×109 m3], mostly in NFCR’s, and this number is the same order a the conventional oil reserves. Reservoir characteristics, geological data and fluid properties of some HO NFCR’s in the Middle East are in Table 1. Most reservoirs are undeveloped because of lack of suitable production technology; daily HO production is only 200-250×103 b/d, from the easiest of the HO NFCR’s. An average HO RF of 0.12 is required for the HO resources in the Middle East to equate heavy oil reserves with 8-10% of remaining conventional reserves to supply enough oil for total world oil consumption for 762 days. This is a low figure, as in some of the new oil production technologies, achieving higher RF is feasible, especially in gravity dominated approaches. Ultimate RF of >12% for HO is likely, given technological developments in Canada since 1985, and emerging and speculative technologies. Cumulative world consumption is approaching ~1,100×109 b [175×109 m3], thus HO, XHO and bitumen from NFCR’s are important long-term oil sources. Huge accumulations of HO in NFCR’s are reported in other Middle East countries: Saudi Arabia, Iraq, and Syria, and to a lesser extent in Yemen, Qatar and Bahrain. However, reliable documentation on HO resources in the Middle East remains pathetically incomplete. HO production is in its early days because of conventional oil absolute supremacy for over a century (and continuing). Implementing new oil production technologies along with improving reservoir characterization schemes in complex, challenging and heterogeneous HO NFCR’s are key factors in successful asset development. HO production in the Middle East will soon play a more important role in total production through implementing current and emerging new heavy oil production technologies in order to boost the declining oil production from light oil reservoirs. Vast HO NFCR’s in Iran, Kuwait, Iraq and Saudi Arabia will make up the bulk of HO production in the next 15-20 years. As shown in the Table 1, most of the HO NFCR’s remain undeveloped. Low matrix permeability, low porosity, medium to densely fractured media and low to medium fracture
VAPEX Process in Heavy Oil NFCR’s The VAPour EXtraction (VAPEX) process was originally proposed by Butler and Mokrys [46, 47] as a non-thermal alternative to the SAGD process. The process involves the injection of hydrocarbon gaseous phases to diffuse and dissolve into HO to reduce its viscosity. Because there are no issues arising from heat losses, the process can be applied to thin reservoirs, low-permeability carbonate reservoirs where the heat capacity per unit volume of contained oil is high, and reservoirs underlain by aquifers and/or gas cap, where application of the SAGD process leads to excessive heat losses to the underburden and overburden aquifer and/or gas cap. The fractures in carbonate reservoirs may cause different effects on the final oil recovery compared to a more homogeneous sandstone reservoir, but the miscibility of the gaseous solvent with the HO in the matrix blocks means that capillary barrier effects are reduced. The vertical fractures will have an upper gas-filled region and a lower liquid-filled (diluted oil) region, and production rate is controlled by the drawdown that guarantees that the gaseous phase is not entering the horizontal producing well. Because of good lateral permeability, the level of the gas-liquid interface in the fractured carbonate should be quite constant, thus helping to overcome inhomogeniy effects on the matrix blocks. Vertical fractures establish communication between layers and create high permeability flow paths for flow of solvent and/or diluted oil, so the effect of horizontal “barriers” or “baffles” on vertical low is limited to within the matrix blocks. hence improving the final oil recovery by VAPEX process. On the other hand, high permeability fractures combined with low permeability fractures may present some problems. In other words, the presence of the high permeability fractures may limit the production rate possible to remain in gravity drainage conditions, and economic constraints comined with a poor solvent-oil ration might arise, limiting the RF that can be 9
permeability are the most common characteristics of HO NFCR’s in this region. Iranian HO NFCR’s are the deepest reserves, but with low porosity and permeability. Low matrix permeability, low porosity and great depth are major constraints for implementation of currently commercialized HO production technologies in Iran. HO fields in Egypt are the heaviest HO reserves in this region in terms of viscosity. HO NFCR’s in Kuwait, Saudi Arabia and Iraq along with off-shore HO assets in the Persian Gulf are the richest HO reserves in the area. Commercialized technologies like SAGD, TA-GOGD, ISC, CSS, WAG, CO2 immiscible injection, CO2 miscible injection are applicable for HO production from HO NFCR’s depending on reservoir parameters, fluid properties and geological conditions. More field are for application of these technologies to have a better understanding of the processes involved, despite promising results from laboratory experiments, numerical simulations and modelling. Emerging technologies like VAPEX have to be field tested. The issue for the Middle Eastern HO base is to decide which production technologies are most suitable for which assets given the large range of properties, depths and conditions, and, to decide which assets are the most amenable for exploitation in terms of RF, speed, heat costs, national benefits, and so on. Before finalizing development technology selection, it is best to do a series of quick economic analyses of potential methods selected on the basis of general screening criteria. In choosing an optimum production technology or a combination of them, use of technical screening criteria is a necessary first-stage process; then, economic analyses can further refine the technology choice or plan for a careful sequence of extraction technologies. A discounted cash flow model is standard, but we point out that this prejudices the results toward profit maximization, and the possibility of high RF over a longer return period tends to be devalued. Whereas the private sector is necessarily profit-driven, a national oil company may be more interested in factors such as high RF, employment issues and infrastructure. Joint development means compromise.
o
o
o
o o
o
o
Documentation on the geographical distribution of HO NFCR’s in the Middle East and their reservoir parameters fluid properties and reservoir geology data is poor. Low matrix permeability, low porosity, medium to densely fractured media and low to medium fracture permeability are the most common fabric characteristics of HO NFCR’s. Iranian HO NFCR’s are the deepest HO reserves in the Middle East, their very low matrix permeability, low porosity and great depth are major constraints for implementation of currently commercialized HO production technologies. HO fields in Egypt are the most viscous HO reserves in this region. HO NFCR’s in Kuwait, Saudi Arabia and Iraq along with various off-shore HO assets in the Persian Gulf are the richest regional HO reserves. Cold production is a rational first production option to be applied in HO NFCR development because of lower CAPEX and OPEX compared with thermal methods. Horizontal well CP can be followed by thermal techniques as SAGD, SWSAGD, TA-GOGD and CSS, but this has to be investigated carefully on a case-by-case basis. HO in the Middle East will soon play a more important role in total hydrocarbon production from this part of the world as some commercialized and emerging HO production technologies are field proven.
NOMENCLATURE Acronyms: Production Technologies CP CSS GOGD ISC SAGD SW-SAGD TA-GOGD VAPEX WAG WGOGD
Conclusion There is a substantial potential for HO production from NFCR’s in the Middle East employing commercialized and emerging HO production technologies, though the huge conventional oil resource has delayed such development. Geographical distribution, occurrence, geological setting and some petrophysical properties of HO from major NFC HO fields in the Middle East are presented in this article. Production technologies application in HO NFCR’s is reviewed, along with some experimental and modelling results, but few field data are available at a high level of detail. Our work indicates: o The Middle East has over 20% of the World HO endowment, mostly in NFCR’s but HO production from these reserves is negligible compared to regional daily light oil production. o Commercialized technologies like SAGD, TAGOGD, ISC, CSS, WAG, CO2 immiscible gas injection, and CO2 miscible injection are applicable for HO production from NFCR’s depending on reservoir parameters, fluid properties and geological conditions. o Most of the HO NFCR’s in the Middle East remain undeveloped mainly because of lack of demonstration of suitable production technology.
= = = = = = = = = =
Cold Production of heavy oil Cyclic Steam Stimulation (vertical wells) Gas Oil Gravity Drainage In Situ Combustion Steam-Assisted Gravity Drainage Single Well SAGD Thermally Assisted GOGD VAP-EXtraction Water Alternating Gas Water Gas Oil Gravity Drainage
Acronyms: Others °F API b CAPEX CF/d D DV EIA FD GOR H Hn HO HTO Kx KY Kz LTO m 10
= = = = = = = = = = = = = = = = = = =
Degrees Fahrenheit American Petroleum Institute Barrel of oil Capital expenditures (pre-production) Cubic foot/day Darcy Developed Energy Information Administration Fuel Deposition Gas Oil Ratio Reservoir thickness (meters) Reservoir net pay thickness (meters) Heavy Oil High Temperature Oxidation Horizontal permeability Horizontal permeability Vertical permeability Low Temperature Oxidation Meters
mD MPa ND NFCR’s NFR NIOC OOIP OPEX PDO PEDEC Psi/psia RF scf/d SOR stb USGS VI-HP XHO
= = = = = = = = = = = = = = = = =
Millidarcy Megapascal Not Developed Naturally Fractured Carbonate Reservoirs Naturally Fractured Reservoir National Iranian Oil Company Original Oil In Place Operating Expenses Petroleum Development Oman Petroleum Eng. and Development Company Pound per square inch Recovery Factor (fractional OOIP recovery) Standard cubic foot/day Steam-Oil Ratio Stock Tank Barrel of oil the United States Geological Survey Vertical Injector-Horizontal Producers Extra Heavy Oil
8. 9. 10.
11.
Symbols: Latin, then Greek So T z Δ μ φ
= = = = = =
12.
Oil Saturation [%] Temperature Depth in meters Difference or change in, as in Δp, Δρ, Δx Viscosity in centipoises [cP] Porosity in %
13.
Metric Conversion Factors 1 psi 1 psi/ft °F 1 barrel
= = = =
14.
6.8947 kPa 22.62 kPa/m or 22.62 MPa/km (°C*1.8) + 32 0.159 m3
15.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
16.
Head, I.M., Jones, D.M. and Larter, S.R., “Biological activity in the deep subsurface and the origin of heavy oil”, NATURE, Vol. 426, No. 20, pp. 344-352, November 2003. Meyer, R.F. and Dietzman, W.D., “World geography of heavy crude oils”, paper presented at the UNITAR Future of Heavy Crude and Tar Sands Conference, pp. 16-28, Edmonton, AB, Canada, 4-12 July 1979. Gibson, B.J., “Methods of classifying heavy crude oils using the UNITAR viscosity based definition”, paper presented at the Second International UNITAR Future of Heavy Crude and Tar Sands Conference, pp. 17-21, Caracas, Venezuela, 7-17 February 1982. Dusseault, M.B., “Comparing Venezuelan and Canadian heavy oil and tar sands”, paper 2001-061 presented at the Petroleum Society’s Canadian International Petroleum Conference 2001, Calgary, Alberta, Canada, 12-14 June 2001. Briggs, P.J., Baron, R.P., Fulleylove, R.J. and Wright, M.S., “Development of heavy oil reservoirs”, Journal of Petroleum Technology, pp. 206-214, February 1988. Dusseault, M.B., “Reservoir enhancement and production technology sequencing”, paper 2006-701 presented in the 1st World Heavy Oil Conference, Beijing, China, 12-15 November 2006. Dusseault, M.B., “Sequencing technologies to maximize recovery”, paper 2006-135 presented at 7th Canadian Int. Petr. Conf. (57th Annual Technical Meeting), Calgary, AB, Canada, 13-15 June 2006.
17.
18.
19.
20. 21.
22.
11
Saidi, A.M., “Reservoir engineering of fractured reservoirs - Fundamental and practical aspects”, TOTAL Edition Press, Paris, 1987. Roehl, P.O. and Choquette, P.W., (Editors), “Carbonate petroleum reservoirs”, Springer-Verlag, New York, 622 p, 1985. Shafiei, A., Dusseault, M.B., Memarian, H. and Samimi Sadeh, B., “Production technology selection for Iranian naturally fractured heavy oil reservoirs”, Paper No. 2007-145, Proc 8th Canadian Int. Petr. Conf. (58th Annual Technical Meeting), Calgary, AB, Canada, 12–14 June 2007. Lorenz, J.C. and Hill, R., “Measurement and Analysis of Fractures in Core”, In Schmoker JW, Coalson EB and Brown CA (eds): Geophysical Studies Relevant to Horizontal Drilling: Examples from North America. Denver: Rocky Mountain Association of Geologists, pp. 47–57, 1994. Spivak, A., Karaoguz, D. and Issever, K., “Simulation of immiscible CO2 injection in a fractured reservoir”, Bati Raman field, Turkey. SPE paper 18765 presented at the SPE California Regional Meeting, Bakersfield, California, 5-7 April, 1989. Karaoguz, D., Issever, K., Pamir, N. and Tirek, A., “Performance of a heavy-oil recovery process by an immiscible CO2 application, Bati Raman field”, SPE #18002, SPE Middle East Oil Technical Conf. and Exhibition, Manama, Bahrain, 11-14 March, 1989. Issever, K.; Pamir, N. and Tirek, A., “Performance of a heavy-oil field under CO2 injection, Bati Raman, Turkey”, SPE Reservoir Engineering, Vol. 8, No. 4, pp. 256-260, November 1993. Bagci, A.S., “Seven-spot steam injection experiments in heavy oil reservoirs having a bottom water zone”, Energy & Fuels, Vol. 19, No. 3, pp.1037-1046, 2005. Kanta, K. and Topkaya, I., “Development of oil and natural gas resources and prospects of enhanced oil recovery in Turkey”, Report, Turkish Petroleum Corp., 1983, pp. 21. Karaoguz, O.K., Nazan, N.T., Lane, R.H., Kalfa, U. and Celebioglu, D., “Improved sweep in Bati Raman heavy-oil CO2 flood: Bullhead flowing gel treatments plug natural fractures”, SPE #89400 presented at the 14th SPE/DOE Symp. on Improved Oil Recovery. Tulsa, Oklahoma, U.S.A., 17–21 April 2004. Karaoguz, O.K., Nazan, N.T., Lane, R.H., Kalfa, U. and Celebioglu, D., “Improving sweep in Bati Raman heavy-oil CO2 flood: Bullhead flowing gel treatments plug natural fractures”, SPE Reservoir Evaluation & Engineering, Vol. 10, No. 2, pp. 164-175, April 2007. Nadyal, M., Al-Hinai, K.M., Jaffer, N.A. and Thurber, S.S., “Development of heavy oil reserves in South Oman”, SPE #11474, 1983 Middle East Oil Exhibition, Manama, Bahrain, 14 March 1983. Al-Adawy, M.S. and Nadyal, M., “Status and scope for EOR development in Oman”, SPE #21407, 1991 SPE Middle East Oil Show, Bahrain, 2-5 March 1991. Macaulay, R.C., Krafft, J.M., Hartemink, M. and Escovedo, B., “Design of a steam pilot in a fractured carbonate reservoir - Qarn Alam field, Oman”, SPE #30300, International Heavy Oil Symposium, Calgary, AB, Canada, 19-21 June 1995. Al-Shizawi, A., Denby, P.G. and Marsden, G., “Heatfront monitoring in the Qarn Alam thermal GOGD
23.
24.
25.
26. 27. 28.
29.
30.
31. 32.
33.
34.
35.
pilot, SPE #37781, Middle East Oil Show, Bahrain, 15-18 March 1997. Shahin Jr., G.T., Moosa, R., Kharusi, B. and Chilek, G., “The physics of steam injection in fractured carbonate reservoirs: Engineering development options that minimize risk”, SPE #102186, SPE Annual Technical Conference and Exhibition, San Antonio, Texas, U.S.A., 24-27 September 2006. Waheed, A., El-Assal, H., Negm, E., Sanad, M. and Sanad, O., “Practical methods to optimizing production in a heavy oil carbonate reservoir: Case study from Issaran field, Eastern desert, Egypt”, SPE #69730,SPE International Thermal Operations and Heavy-Oil Symposium, Porlamar, Margarita Island, Venezuela, March 12-14, 2001. Al-Qabandi, S., Al-Shatti, Y. and Gopalakrishnan, P., “Commercial heavy oil recovery by cyclic steam stimulation in Kuwait”, SPE paper 30288, presented at the SPE International Heavy Oil Symposium, Calgary, Alberta, Canada, 19-21 June 1995. Penner, S.S., “Assessment of research needs for oil recovery from heavy oil sources and tar sands”, DOE report, March 1982. Bagci, S. and Kok, M.V., “Combustion reaction kinetics studies of Turkish crude oils”, Energy & Fuels, Vol.18, 1472-1481, 2004. Akin, S., Kok, M.V., Bagci, S. and Karacan, O., “Oxidation of heavy oil and their SARA fractions: Its role in modelling in-situ combustion”, SPE #63230, SPE Annual Technical Conference and Exhibition, Dallas, Texas, 1-4 October 2000. Akin, S., Bagci, S. and Kok, M.V., “Dry forward combustion with diverse well configurations”, SPE #62551, SPE/AAPG Western Regional Meeting, Long Beach, California, 19-23 June 2000. Bagci. S., “Laboratory and simulation results of insitu combustion for heavy oil recovery from Bati Kozluca field, Turkey”, Paper 103, presented at Offshore Mediterranean Conference and Exhibition, Ravenna, Italy, March 28-30, 2007. Gumrah, F. and Bagci, S., “Steam-CO2 drive experiments using horizontal and vertical wells”, J. of Petroleum Sci. and Eng., Vol. 18, 113-129, 1997. Bagci, S. and Gumrah, F., “Effects of CO2 and CH4 addition to steam on recovery of West Kozluca heavy oil”, SPE #86953, SPE Int. Thermal Operations and Heavy Oil Symposium and Western Regional Meeting, Bakersfield, California, 16-18 March 2004. Shahin Jr., G.T., Moosa, R., Kharusi, B. and Chilek, G., “The physics of steam injection in fractured carbonate reservoirs: Engineering development options that minimize risk”, SPE #102816, SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 24-27 September 2006. Penney, S., Baqi Al-Lawati, S., Hinai, R., Van Ravesteijn, O., Rawsley, K., Putra, P., Geneau, M., Ikwumonu, A., Habsi, M. and Harrasy, H., “First full field steam injection in a fractured carbonate at Qarn Alam, Oman”, SPE #105406, 15th SPE Middle East Oil & Gas Show and Conf,, Bahrain, 11-14 March 2007. Mateo Hernandez, J.A. and Trevisan, O.V., “Heavy oil recovery mechanisms during steam injection in naturally fractured reservoirs”, SPE #107372, 2007 SPE Latin American and Caribbean Petroleum
36. 37.
38. 39. 40.
41.
42.
43.
44.
45. 46.
47.
48.
49.
12
Engineering Conference, Buenos Aires, Argentina, 15-18 April 2007. Butler, R.M. and Stephens, D.J., “The gravity drainage of steam heated to parallel horizontal wells”, J. Canadian Petrol. Tech., 90-96, April-June, 1981. Joshi, S.D. and Threlkeld, C.B., “Laboratory studies of thermally aided gravity drainage using horizontal wells”, AOSTRA Journal of Research, Vol.2, No.1, 11-19, 1985. Joshi, S.D., “A review of thermal oil recovery using horizontal wells”, IN SITU, Vol.2, No.1, 211-259, 1987. Butler, R.M., “Rise of interfering steam chambers”, Journal of Canadian Petroleum Technology, Vol. 26, No. 3, 70-75, 1987. Joshi, S.D., “A laboratory study of thermal oil recovery using horizontal wells”, SPE Paper 14916, presented at SPE/DOE 5th Symposium on Enhanced Oil recovery, Tulsa, OK, April 2-23, 1986. Hoffman, B.T. and Kovscek, A.R., “Light oil steam drive in fractured low-permeability reservoirs”, SPE #83491, SPE Western Regional/AAPG Pacific Section Joint Meeting, Long Beach, CA, May 19-24, 2003. Bagci, S., “Experimental and simulation studies of SAGD process in fractured reservoirs”, SPE #99920, presented at the SPE/DOE Symp. on Improved Oil Recovery, Tulsa, OK, 22-26 April 2006. Akin, S. and Bagci, S., “A laboratory study of singlewell steam-assisted gravity drainage process”, Journal of Petroleum Science and Engineering, Vol. 32, 2333, 2001.Sahin, S., Kalfa, U. and Celebioglu, D., “Bati Raman field immiscible CO2 application: Status quo and future plans”, SPE #106575, SPE Latin American and Caribbean Petroleum Engineering Conference, Buenos Aires, Argentina, 15-18 April 2007. Cobanoglu, M., “A numerical study to evaluate the use of WAG as an EOR method for oil production improvement at B. Kozluca field, Turkey”, SPE #72127, SPE Asia Pacific Improved Oil Recovery Conf., Kuala Lumpur, Malaysia, 8-9 October 2001. Butler, R.M. and Mokrys, I.J., “Solvent analogue model of steam-assisted gravity drainage”, AOSTRA Journal of Research, Vol.5, No.1, 17-32, 1989. Butler, R.M. and Mokrys, I.J., “A new process (VAPEX) for recovering heavy oils using hot water and hydrocarbon vapour”, J. Canadian Petroleum Tech., Vol.30, No.1, 1991. Singhal, A.K., Das, S.K., Leggitt, S.M., Kasraie, M. and Ito, J., “Screening of reservoirs for exploitation by application of steam-assisted gravity drainage/ VAPEX processes”, SPE #37144, SPE International Conference on Horizontal Well Technology, Calgary, Canada, November 18-20, 1996. Azin, R., Kharrat, R., Ghotbi, C. and Vossoughi, S., “Applicability of the VAPEX process to Iranian heavy oil reservoirs”, SPE 92720, 14th SPE Middle East Oil & Gas Show and Conference, Bahrain, 12-15 March 2005. Rostami, B., Azin, R. and Kharrat, R., “Investigation of the VAPEX process in high pressure fractured heavy oil reservoirs”, SPE #97766 SPE International Thermal Operations and Heavy Oil Symposium, Calgary, Alberta, Canada, 1-3 November 2005.
Figure 1: Geographical distribution of HO NFCR’s in the Middle East
13
Table 1: Geological, reservoir characteristics and fluid properties of selected HO NFCR’s in the Middle East. Country
NFC Heavy Oil Field Name OOIP 9 (×10 b)
Iran
K1
Iran
K2
Iran
P1
Iran
P2
Iran
Z
Production rate 3 (×10 b/d)
Lithology
Depth (z) (m)
Porosity (φ) (%)
Reservoir Characteristics Fracture Matrix permeability permeability (Kf)(mD) (Km) (mD)
Net pay thickness (Hn) (m)
Reservoir temperature (T) (◦F)
Oil saturation (So) (%)
Fluid Properties Viscosity API (μ) (cP) (°)
12
ND
Dolomite and dolomitic limestone
300 600
20
1
200-400
100
66
66
1,5002,000
814
1
ND
Limestone
1,150
16
1
200-300
176
110
46
600-700
1416
1
ND
Limestone and dolomite
2,900
27
1-2
100-200
40
175
80
250
18
0.10
ND
2.50
ND
Limestone Dolomitic and highly argillaceous limestone Limestone Limestone Limestone Limestone Limestone Limestone Limestone
4,000
12
1-2
100-150
25
115
75
120
1718
4,300
8
1
50-100
65
200-250
80
25-210
1516
34 ? ? ? 18 15-23
? ? ? ? 16 50-400
? ? ? ? ? ?
? ? ? ? 150 120
? ? ? ? ? ?
23-33
1.30-104
?
85-100
?
17 30 20-45
6 5-20 50-3,000
? ? ? ? 200-500 ? Several Darcies ? 1-10,000 ?
? ? ?
? ? ?
? ? ?
600 ? ? ? 592 935 3,0005,000 ? 200 80
17
300
?
?
?
?
50-70
Iran Iran Iran Iran Turkey Turkey Egypt
F1 F2 F3 F4 Bati Raman Ikiztepe Issaran
10 6 7 6 1.85 0.13
ND ND ND ND DV DV
0.50
DV
Egypt Oman Kuwait Kuwait
Bakr Amer Qarn Alam W1 W2
1.30 1.34 12
DV DV ND
Limestone Limestone Dolomite
? ? ? ? 1,311 1,375 330760 570 360 400
2
ND
Dolomite
670
14
16 ? 16 16 12 1014 ? 16 18 1320
