Experimental and Numerical Simulation of Oil Recovery from Oil Shales by Electrical Heating

Experimental and Numerical Simulation of Oil Recovery from OilShales by Electrical Heating

Berna Hascakir,†,‡,§ Tayfun Babadagli,*,‡ and Serhat Akin†

Department of Petroleum and Natural Gas Engineering, Middle East Technical UniVersity,Ankara 06531, Turkey, and School of Mining and Petroleum Engineering, Department of CiVil and

EnVironmental Engineering, UniVersity of Alberta, Edmonton, Alberta T6G 2W2, Canada

ReceiVed May 23, 2008. ReVised Manuscript ReceiVed July 24, 2008

The recovery characteristics of four different oil shale samples were tested experimentally using the retorttechnique. To accomplish efficient temperature distribution, the thermal conductivity of the oil shale sampleswas increased by the addition of three different iron powders. The doses of iron powders were optimized foreach oil shale sample based on the highest oil production value experimentally. The experiments were thenmodeled using the electrical heating option of a commercial reservoir simulator. The viscosity-temperaturerelationship was obtained by matching the experimentally obtained temperature distribution in the cores andproduction data to the numerical ones. After the other parameters needed for the numerical model were collectedand compiled, field-scale simulations were performed and a parametric analysis was performed for differentoil shale cases. The experimental and numerical results show that field-scale oil recovery from oil shales byelectrical heating could be technically and economically viable.

1. Introduction

To ensure current energy needs and meet future expectations,new techniques for efficient use of unconventional resources,exploration of new reserves, and evaluating the potentialalternatives have to be considered together.1,2 Oil shales are oneof the alternative fossil fuel resources. Shales are known as fine-grained sedimentary rocks that yield significant amounts of oilthrough pyrolysis, and these resources have been used sinceancient times.3-5 The modern industrial use of oil shales foroil extraction dates to the mid-19th century.6

Although information about many oil shale deposits isrudimentary and much exploratory drilling and analytical workneeds to be performed, the potential resources of oil shales inthe world are enormous.6 The total proven oil shale reserves ofthe world are about 80 000 million tons.7 Therefore, it isimportant to investigate the recovery characteristics of oil shaleresources.8

Oil shale is a general expression usually used for a fine-grained sedimentary rock, containing significant amounts ofkerogen, from which liquid hydrocarbons can be obtained.9,10

Kerogen is a mixture of organic chemical compounds that makeup a portion of the organic matter in sedimentary rocks. Becausekerogen has a huge molecular weight of its componentcompounds, it is insoluble in normal organic solvents. Bitumenand/or prebitumen, which are known as the soluble portion ofkerogen may also exist but in relatively lower amounts. If it isheated to the right temperatures in the Earth’s crust, some typesof kerogen release oil or gas, collectively known as hydrocar-bons (fossil fuels). If such kerogens are present in highconcentration in rocks, such as shale, and have not been heatedto a sufficient temperature to release their hydrocarbons, theymay form oil shale deposits.11

The key to produce oil from these resources is to reduce oilviscosity, and that is best accomplished by heating theseresources up to 500 °C, which is known as the pyrolysis process,which is known also as the decomposition of the kerogen.4,12

Oil retorting can be considered as the most effective methodfor extracting oil from oil shales.6 The easiest way to increasethe efficiency of this method is to increase the thermalconductivity of the system or increase the reduction of the oilviscosity by using some additives. Metallic additives cause

* To whom correspondence should be addressed. Fax: +1-780-492-0249.E-mail: [email protected].

† Middle East Technical University.‡ University of Alberta.§ Visiting Ph.D. student at the University of Alberta.(1) Greene, D. L.; Hopson, J. L.; Li, J. Have we run out of oil yet? Oil

peaking analysis from an optimist’s perspective. Energy Policy 2006, 34,515–531.

(2) Ivanhoe, L. F. Future world oil supplies: There is a limit. WorldOil, Nov, 1995; pp 91-94.

(3) Tissot, B. P.; Welte, D. H. Petroleum Formation and Occurrence;Springer: Verlag, Germany, 1984.

(4) Farouq Ali, S. M. Heavy oilsEvermore mobile. J. Pet. Sci. Eng.2003, 37, 5–9.

(5) Moody, R. Oil and gas shales, definitions and distribution in timeand space, the history of on-shore hydrocarbon use in the U.K. Abstracts,April 20-22, 2007.

(6) Altun, N. E.; Hicyilmaz, C.; Hwang, J.-Y.; Bagci, S. Evaluation ofa Turkish low quality oil shale by flotation as a clean energy source: Materialcharacterization and determination of flotation behavior. 2006, 87, 783-791.

(7) World Energy Council (WEC). Survey of energy resources: Oil shale.WEC, London, U.K., 2001, also available at http://www.worldenergy.org/wec-geis/publications/reports/ser/shale/shale.asp.

(8) Campbell, C. J.; Laherrere, J. H. The end of cheap oil. Sci. Am.1998, March, 78–83.

(9) Tissot, B. P.; Welte, D. H. Petroleum Formation and Occurrence;Springer-Verlag: Berlin, Germany, 1984; p 699.

(10) Hutton, A. C. Petrographic classification of oil shales. Int. J. CoalGeol. 1987, 8, 203–231.

(11) Weber, G.; Green, J. Guide to oil shale. National Conference ofState Legislatures, Washington, D.C., 1981; p 21.

(12) Johannes, I.; Kruusement, K.; Veski, R. Evaluation of oil potentialand pyrolysis kinetics of renewable fuel and shale samples by Rock-Evalanalyzer. J. Anal. Appl. Pyrolysis 2007, 79, 183–190.

Energy & Fuels 2008, 22, 3976–39853976

10.1021/ef800389v CCC: $40.75 2008 American Chemical SocietyPublished on Web 09/05/2008

changes in the nature and the amount of fuel formed during insitu combustion. These changes appear to depend upon the typeof oil used. Various crude oils are affected differently bydifferent additives.13

In this study, the recovery characteristics of four differentoil shale samples obtained from different oil shale deposits inTurkey were investigated experimentally at laboratory conditionsby using the retort technique. To enhance the oil production byreducing oil viscosity and achieving the effective temperaturedistribution, three different iron powders and their three differentdoses were used. The experimental results were simulated usinga commercial reservoir simulator, where the data required forfield-scale simulation were obtained through history matchingof production data and temperature distribution inside the core.Note that the deposits were considered to be deeper than theminable range and, therefore, in situ recovery techniques areneeded for oil recovery. Finally, field-scale application ofelectrical heating was simulated, and a technical and economicevaluation of the applicability of the method was presented.

A few parameters such as viscosity-temperature variation,molecular weight, and relative permeabilities that are neededin the simulation of oil recovery from oil shale studies are highlydifficult to obtain (or not practically measurable) at laboratoryconditions unlike conventional reservoirs. The accuracy of thesimulation results, however, strongly depends upon theseparameters. Experimental verifications and supports are criticalin this regard, and we introduced a new approach in the field-scale simulation of the process based on the laboratory-scalemodeling of the process experimentally and numerically. Thiswould provide realistic data needed for simulation and, therefore,increase the accuracy of modeling studies.

2. Experimental Methodology

Samples and Characterization. Four different oil shale samples(OS1, OS2, OS3, and OS4) from four different oil shale depositswere used. The properties of those samples are given Table 1.

The OS1 basin is of the Neocene age. Its seam consists of morethan 50% liptinite, 20-50% huminite, and 0-20% inertinitemaceral groups and is characterized by its high organic content.15

The origin of the organic matter is mainly algae and plants.16 TheOS2 deposit is of the Paleocene-Eocene age. It contains around80% liptinite, 5-10% bituminite, and 5-10% huminite. The highliptinitic content shows that the organic matter originated mainlyfrom hydrogen-rich organic remains of algae and pollen. Calcite,dolomite, quartz, and smectite are the major inorganic constituents.17

OS3 is rich in quartz, dolomite, calcite, and clay minerals, such as

muscovite, illite, and smectite.18-20 Geochemical analysis revealedthat the organic content of this oil shale was mainly derived fromalgae, pollen, and planktonic algae, as well as bacteria. Also, tracesof liptodetrinite and humic organic material were observed.16,20 Itcontains a relatively high amount of ash (≈70%); however, thetotal sulfur content seldom exceeds 1.5%, with an average of lessthan 1%.16,19 The OS4 deposit underlies conglomeratic rocks, andthe average thickness of the oil shale bed is 13 m. The averagecalorific value and oil content of the deposit are 851 kcal/kg and13.7%, respectively.14,21

Experimental Setup and Procedure. Retort experiments wereconducted with a setup that consists of a stainless-steel cylindricalbody that houses the samples with an inner diameter of 10 cm andheight of 20 cm, respectively. The cylindrical body was wrappedwith a band heater (1000 W), and this heater was connected to atemperature controller to increase the temperature of the systemby using commercial software. A thermocouple was placed at thecenter of the sample holder and connected to the temperaturecontroller to record temperature values continuously by using thesame software. Then, the sample holder was placed in anothercylindrical cell, which has a 20 cm inner diameter and 25 cm height.To minimize the heat losses, the space between two cylinders wasfilled by crushed perlite [thermal conductivity of 75 °F (24 °C),0.27-0.41 Btu in. h-1 ft-2 °F-1 (0.04-0.06 W m-1 K-1)]. Fromthe help of software and the band heater, the temperature of thesystem could be increased to desired temperatures (Figure 1).

3. Experimental Results

To determine the optimum heating periods, eight differentexperiments were carried out with OS4. Optimum heating andsoaking periods were selected according to the highest oilproduction. These eight experiments were summarized in Table2. Because thermal cracking, which is also known as pyrolysis,takes place over 500 °C, for all of these experiments, 500 °Cwas selected as last temperature value, to produce shale oil.13

(13) Castanier, L. M.; Brigham, W. E. Upgrading of crude oil via insitu combustion. J. Pet. Sci. Eng. 2003, 39, 125–136.

(14) Senguler, I. Bituminous shales: Their origin, usage and importance.Bull. MTA Nat. Resour. Econ. 2007, 3 (in Turkish).

(15) Sener, M.; Gundogdu, M. N. Geochemical and petrographicinvestigation of Himmetogjlu oil shale field, Goynuk, Turkey. Fuel 1996,75 (11), 1313–1322.

(16) Putun, E.; Akar, A.; Ekinci, E.; Bartle, K. D. Chemistry andgeochemistry of Turkish oil shale kerogens. Fuel 1988, 67, 1106–1110.

(17) Sener, M.; Senguler, I. Geological, mineralogical, geochemicalcharacteristics of oil shale bearing deposits in the Hatıldagj oil shale field,Goynuk, Turkey. Fuel 1998, 77 (8), 871–880.

(18) Hufnagel, H. Investigation of oil shale deposits in western Turkey.Technical Report Part 2, Project 84.2127.3, BGR, Hannover, Germany,1991.

(19) Sener, M.; Senguler, I.; Kok, M. V. Geological considerations forthe economic evaluation of oil shale deposits in Turkey. Fuel 1995, 74,999–1003.

(20) Kok, M. V.; Senguler, I.; Hufnagel, H.; Sonel, N. Thermal andgeochemical investigation of Seyitomer oil shale. Thermochim. Acta 2001,371, 111–119.

(21) Altun, N. E. Beneficiation of Himmetogjlu and Beypazari oil shalesby flotation and their thermal characterization as an energy source. Ph.D.Thesis, Graduate School of Natural and Applied Sciences, Middle EastTechnical University, 2006; pp 5-29.

(22) Hascakir, B.; Demiral, B.; Akin, S. Experimental and numericalanalysis of oil shale production using retort technique. 15th InternationalPetroleum and Natural Gas Congress and Exhibition of Turkey, Ankara,Turkey, May 2005.

Table 1. Properties of Oil Shale Samples14

deposita calorific value (kcal/kg) oil shale resource (106 tons)

OS1 1390 66OS2 774 360OS3 860 122OS4 851 130b

a OS ) oil shale. b Probable reserve.

Figure 1. Retort setup.22

Oil RecoVery from Oil Shales by Electrical Heating Energy & Fuels, Vol. 22, No. 6, 2008 3977

Because the highest production was obtained after the secondexperiment, heat and soak periods belong to this experimentwere selected as optimum periods (Figure 2). The otherexperiments were accomplished by using this optimum condition.

There are two different soaking periods determined as theoptimal experimental operation period (Figure 2). The first one

starts at 100 °C. This soaking period helps to vaporize water.The second soaking period begins at 500 °C, which is thepyrolysis temperature required for the decomposition of thekerogen. To vaporize the water or decompose the kerogenneeded, time is given with the help of soaking periods.

One can infer from experiment 7 that, below the pyrolysistemperature, it is impossible to produce oil from oil shalesamples. Although it is possible to produce oil from oil shalewhen the pyrolysis temperature is reached, the oil producedwould not be much. To enhance the oil production efficiencies,effective temperature distribution and viscosity reduction of theshale oil should be accomplished together. Therefore, threedifferent iron powders and their three different doses were testedon the shale oil production. The iron powders used are Fe,

Table 2. Determination of Optimum Heating Period for Retort Experiments22

first period second period

heating soaking heating soaking

number of experiments Da (s) Tb (°C) D (s) T (°C) D (s) T (°C) D (s) T (°C) petroleum production (cc)

1 300 100 3600 100 9000 500 32 1800 100 3600 100 5400 500 4500 500 43 1200 100 2700 100 6000 500 6000 500 0.24 1200 100 1800 100 900 580 15 3600 100 6000 520 16 600 100 5400 100 5400 500 0.27 1800 100 3600 100 12600 350 08 6000 500 3

a D ) duration. b T ) reached temperature values at the end of the heating or soaking period.

Figure 2. Temperature profile determined to be the optimum operation periods.22

Figure 3. Retort experiment results for oil shale samples.

Table 3. Density and the API Gravities of the Produced ShaleOil

sample oil density (g/cm3) API gravity

OS1 0.97 14.94OS1 plus 0.1% Fe 1.06 2.57OS2 0.89 27.49OS2 plus 0.1% Fe2O3 0.99 12.01OS3 1.03 5.88OS3 plus 0.5% Fe 1.02 7.40OS4 0.89 27.40OS4 plus 0.1% Fe2O3 0.92 21.97

Figure 4. Grid sizes used in simulations.

3978 Energy & Fuels, Vol. 22, No. 6, 2008 Hascakir et al.

Fe2O3, and FeCl3, and their doses are 0.1, 0.5, and 1% by weight.The average particle size of Fe and Fe2O3 is 10 µm. FeCl3 wascrushed by a porcelain mortar, and its average size is slightlygreater than this value. A total of four experiments wereconducted for raw oil shale samples, and 36 experiments wereconducted to find the optimum types and doses of iron powdersfor each oil shale sample. Oil production, after retorting of raw

oil shale samples and after retorting of samples containingoptimum types and doses of iron powders, is summarized inFigure 3.

Because iron powders help increase the thermal conductivityof the system, heat transfer was accomplished more efficiently,yielding increased oil production at laboratory conditions. Also,iron additives have a catalytic effect that increases the reaction

Figure 5. Water-oil relative permeability curves used in simulation for oil shale samples.25

Figure 6. Gas-liquid relative permeability curves used in simulation for oil shale samples.25

Figure 7. Temperature match for all oil shales (I, experimental study results; II, numerical study results).


speed. The chemical reactions between iron powders and shaleoil help to break the chemical bonds by increasing thetemperature and magnetic effect of iron powders on thereduction of oil viscosity, which caused an increase in the oilproduction after the addition of iron powders.13,23,24 API gravityand the oil densities with and without iron powder addition wereobtained experimentally and are given in Table 3.

4. Numerical Simulation

All simulation studies were performed using the electricalheating option of the CMG-STARS (steam, thermal, andadvanced processes reservoir simulator) commercial simulator(CMG). The domain was discretized into 20 × 1 × 10, 3Dradial blocks of varying size in the r direction and constant sizein the z direction. The dimensions of the reservoir for the fieldcase and grid size for both laboratory and field cases are givenin Figure 4.

The required input data for simulation are porosity, perme-ability, thermal conductivity, rock heat capacity, rock compress-ibility, viscosity, and relative permeabilities. The output data

of the simulation is time-dependent temperature distribution andoil production data, which were also determined experimentally.All required input data, except the viscosity-temperaturerelationships, were taken from the literature. All of the inputdata (porosity, permeability, rock properties, etc.), except thetemperature-viscosity relationship, were taken from the litera-ture for the numerical simulation runs. These data were compiledfrom the literature about the oil shales studied in this study andthe common oil shale fields around the world. The viscosity-temperature relationship was obtained through matching nu-merically obtained temperature distributions and production datato the experimental output. Another critical parameter in additionto the viscosity-temperature data is the relative permeabilities.It was observed that the effect of relative permeabilities is trivialcompared to the viscosity-temperature data. Relative perme-ability data used in the study were taken from another oil shalefield (Figures 6 and 7). Hence, relative permeabilities typicallysuggested for these types of simulations in the literature wereadapted and used.

Because shale oil is too viscous and the amount of producedoil at the laboratory conditions is very little, it was very difficultto determine the viscosity of the shale oil experimentally.Therefore, the only way to obtain this critical data was theexperimental matching exercise. Another critical point that

(23) Kershaw, J. R.; Barrass, G.; Gray, D. Chemical nature of coalhydrogenation oils part I. The effect of catalysts. Fuel Process. Technol.1980, 3-2, 115–129.

(24) Odenbach, S. Ferrofluids/magnetically controlled suspensions.Colloids Surf., A 2003, 217, 171–178.

Figure 8. Production match for all oil shales (I, experimental study results; II, numerical study results).

Figure 9. Viscosity variations of all shale oils.


entails this type of approach is to determine the level of heatingand corresponding temperature to have oil become flowable.

Using the shale oil viscosity values and literature data, theprocess was then simulated at field scale. In these numericalsimulation studies, the power of the system, operation time, andthe number of heaters were optimized by considering both oilproduction and economics of the project.

5. Numerical Simulation Results

During the experiments conducted with oil shale samples,temperature and production data were recorded continuously.As mentioned above, produced oil from oil shale is highlyviscous and very little in amount at the laboratory conditions.Thus, shale oil viscosities are difficult to measure. Moreover,there is little knowledge about oil shale properties in theliterature. In the simulation studies, literature data given inFigures 5 and 6 were used.

Using the viscosity as the adjustable parameter, the simulationoutputs were matched to the experimentally obtained temper-ature and production data. These matches are given in Figures

7 and 8. The viscosity-temperature relationship obtainedthrough this exercise is shown in Figure 9.

After 400 °C, a linear relationship was observed betweentemperature and shale oil viscosity on a semi-log plot, which meansthat oil starts to flow as a result of reaching the pyrolysistemperature. This type of trend observed after 400 °C is a typicalrelationship between viscosity and temperature during pyrolysis.4,12

Simulation studies were also performed for the verificationof the experimental studies carried out after the addition ofoptimum iron powder types and doses. These simulation matchesfor all oil shales containing optimum doses and types of ironpowders are given in Figures 10 and 11 for temperature andproduction, respectively. The viscosity-temperature relationshipobtained through the matching exercise is shown in Figure 12.As can be inferred from Figure 12, the iron powder additiondecreases the shale oil viscosity considerably.

The viscosity-temperature data obtained through matchingthe experimental results to the numerical output and the porosity,permeability, rock compressibility, rock heat capacity, rockthermal conductivity and the molecular weight of oil values (asgiven in Table 4) previously used for the simulation oflaboratory experiments were also used to simulate the field cases.When the data were kept the same as the laboratory simulationdata, the operation time, the power needed in the field

(25) Sarkar, A. K.; Sarathi, P. S. Feasibility of steam injection processin a thin, low-permeability heavy oil reservoir of ArkansassA numericalsimulation study. U.S. Department of Energy Assistant Secretary for FossilEnergy, 1993; p 42.

Figure 10. Temperature matches for all oil shale after the addition of optimum doses of iron powders (I, experimental study results; II, numericalstudy results).

Figure 11. Production matches for all oil shale after the addition of optimum doses of iron powders (I, experimental study results; II, numericalstudy results).


application, and the number of heaters were optimized. For thispurpose, an operation time of 60 and 90 days, the power valuesof 46 296, 34 722, and 23 148 W, and 5 or 10 heaters wereused. Furthermore, sensitivity analyses were conducted by usingtwo different reservoir depths, i.e., 500 and 3000 m, at threedifferent reservoir pressure values, i.e., 5000, 9000, and 10 000

kPa, and three different bottom hole pressures, i.e., 1000, 1800,and 2000 kPa. Because the simulation results yielded verysimilar oil production results (not shown here) for differentreservoir depths, pressures, and bottom hole pressures (even atextreme values of those reservoir properties), it can be stated

Figure 12. Viscosity variations for all oil shale after the addition of optimum doses of iron powders.

Table 4. Data Used for the Simulation of Laboratory and Field Cases4,6,26,27

oil shale

OS1 OS2 OS3 OS4

parameter unit raw 0.1% Fe raw 0.1% Fe2O3 raw 0.5% Fe raw 0.1% Fe2O3

porosity (%) 45 45 45 45 45 45 45 45permeability (md) 5000 5000 5000 5000 5000 5000 5000 5000rock compressibility (kPa-1) 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001rock heat capacity (J m-3 °C-1) 6.4 × 105 6.53 × 105 1.37 × 106 1.29 × 106 6.5 × 105 6.9 × 105 8.9 × 105 9.2 × 105

rock thermalconductivity

(J m-1 day-1 °C-1) 1.44 × 105 2.06 × 105 1.44 × 105 2.06 × 105 1.7 × 105 4.5 × 105 1.44 × 105 2.06 × 105

molecular weight (g mol-1) 650 650 600 600 600 597 600 600laboratory pressure (kPa) 101 101 101 101 101 101 101 101laboratory temperature (°C) 21 21 21 21 21 21 21 21reservoir field pressure (kPa, ×103) 5-10 5-10 5-10 5-10 5-10 5-10 5-10 5-10reservoir field

temperature(°C) 50 50 50 50 50 50 50 50

reservoir depth (m) 500-5000 500-5000 500-5000 500-5000 500-5000 500-5000 500-5000 500-5000initial water saturation

in the reservoir(%) 25 25 25 25 25 25 25 25

formation thickness (m) 30 30 30 30 30 30 30 30

Figure 13. Simulation results for OS1 (field case).


that electrical heating of oil shale reservoirs are not criticallyaffected by reservoir depth, pressure, and bottom hole pressurevalues.

In the simulation runs, oil was produced continuously whileheating the reservoir in the same well. When the soaking periodwas applied in the simulations, it was observed that oilproduction did not change significantly but the operation timeincreased, which resulted in an increase in the cost of theprocess.

For OS1 and OS2, three different field simulations wereconducted and the results were summarized in Figures 13 and14, respectively. While in the first runs, the operation time was60 days, in the second and third runs, operation times wereincreased to 90 days. The number of heaters for the third runwas reduced from 10 to 5, so that the power exerted for theseoperations was decreased. For both cases, the third run yieldedthe optimum oil production (highest recovery/power ratio). Itis observed that, while the iron powder addition caused anincrease in oil production for OS2, the opposite was observedfor OS1 when the power was reduced. The results for OS3 areshown in Figure 15. It was observed that the second run gavebetter oil production for the raw oil shale. However, after the

addition of 0.5% Fe, the oil production values increased sharplybecause of the effect of viscosity variation after the addition of0.5% Fe.23,24 All of the runs were performed for 60 days andwith 10 heaters.

For OS4, four different runs were carried out and the thirdrun yielded the highest oil production (Figure 16). Also thedifference between the raw and 0.1% Fe2O3 cases were verysimilar for the first three runs; the fourth run showed nosignificant change in production when Fe2O3 is added. For thefirst two and fourth runs, the operation times were 60 days; itwas 90 days for the third run. A total of 10 heaters were usedfor the first three runs, and for the forth run, the number ofheaters was decreased to 5. While keeping the number of heaters(i.e., 10) and the operation times (i.e., 60 days) constant, differentpowers were applied (i.e., 34 722 and 46 296 W) in the firstand second cases. The oil production increased from 49 to 109bbl for raw oil shale and from 148 to 250 bbl for additional oilshale.

The second and third runs compare the cases with differentoperation times. While the power (i.e., 46 296 W) and thenumber of heaters (i.e., 10) were kept constant for the secondand third runs, the oil production increased because of the




increasing of operation time, from 60 to 90 days. The increasein the oil production for the raw production case was 240%,and this value turned out to be 168% for the 0.5% Fe2O3 cases.

6. Economics

Economic evaluation was carried out by considering the costof electricity determined by the Turkish Electric Distribution

Incorporated Company (TEDAS).28 Table 5 summarizes theeconomic evaluation of this study for only the best oil productionresults.

Economic evaluation of this study shows that the recoveryof oil from these reserves by the retort technique can beapplicable for three oil shale reservoirs (OS1, OS2, and OS4)but not OS3. Note that only the heating cost was considered inthis analysis and the CAPEX was not included.

Finally, Figure 17 shows the time dependency of thecumulative oil production for the raw oil shales and oil shalesafter the addition of iron powders. The best result was obtainedfor the OS4 case with iron powder. Over the time period ofinvestigation (90 days), the increasing trend was still obviousand the plateau region had not been reached unlike the otherseven cases. It is interesting that its “raw version” also showeda similar trend but much less oil production. Note that OS4 hassignificantly higher rock heat capacity compared to the otherthree samples.

7. Conclusions

(1) Viscosity-temperature relationship as data to the simula-tor was observed as the most critical parameter, but it is noteasily measurable. Therefore, an experimental study was

(26) Bechtel SAIC Company, LLC. Heat capacity analysis report. U.S.Department of Energy Office of Civilian Radioactive Waste ManagementOffice of Repository Development, 2004; pp 6-23-6-24.


Figure 17. Cumulative oil production results for the optimum operation conditions of oil shales (I, raw oil shale; II, oil shale containing optimumtypes and doses of iron powders).

Table 5. Economic Evaluation of the Studya

samplename addition

number ofheaters

totaloperation

time(days)

oilproduction

(bbl)

cost ofthe study($/bbl)

OS1 5 90 296 15.55OS1 0.1% Fe 5 90 119 38.68OS2 5 90 145 30.46OS2 0.1% Fe2O3 5 90 180 24.62OS3 10 60 9 502.55OS3 0.5% Fe 10 60 55 79.85OS4 10 90 262 33.78OS4 0.1% Fe2O3 10 90 420 21.07

a The cost of electricity ) 0.088 US$/1 kWh.28


performed to obtain this input data by numerically simulatingthe experiments and matching the history for temperaturedistribution and production. (2) After 400 °C, the relationshipbetween viscosity and temperature was observed as linear on asemi-log plot. This is the pyrolysis temperature. (3) Introducingiron powder into the reservoir for practical applications is acritical issue, but we are not aware of any application orsuggestion in the literature in this regard. The addition of ironpowder could be achieved by injecting iron powders into thereservoir, after mixing them with petroleum-based fluids, suchas light oils or solvents. If the field is shallow enough for surfacemining, a better solution would be adding the iron powdersduring the extraction process. (4) The technical and economicfeasibility analyses showed that electrical heating is still a

marginal application, but the results proved that it is in anapplicable range. (5) All oil shales showed different recoverytrends. The production rate and the ultimate recovery from theoil shale case of OS4 were remarkably higher compared to theother three cases. This could be attributed to significantly higherrock heat capacity of this particular sample compared to theother three samples.

Acknowledgment. This work was funded by The Scientific andTechnological Research Council of Turkey (TUBITAK) and theFaculty Development Program-Middle East Technical University(OYP-METU). The numerical modeling part of the study wasperformed during the stay of the first author (B.H.) at the Universityof Alberta as a visiting Ph.D. student. We gratefully acknowledgethese supports. We also thank the Computer Modeling Group(CMG) for providing the simulation software package and itselectrical heating option for this research.

EF800389V

(27) Lide, D. R. CRC Handbook of Chemistry and Physics, 88th ed.;CRC Press: Boca Raton, FL, 2007-2008.

(28) http://www.tedas.gov.tr (accessed on Jan 2, 2008).


Documents

Experimental and Numerical Simulation of Oil Recovery from Oil Shales by Electrical Heating